Merge commit '8156c2dfd2b8215d998cb4dac5fa05b62692ee13' into patch_branch Change-Id: I04dda9cd75a876a6332474e0536832f3146e351a
diff --git a/Documentation/gpu/drm-kms.rst b/Documentation/gpu/drm-kms.rst index 2dcf5b42..58534dd 100644 --- a/Documentation/gpu/drm-kms.rst +++ b/Documentation/gpu/drm-kms.rst
@@ -286,6 +286,9 @@ .. kernel-doc:: drivers/gpu/drm/drm_atomic.c :export: +.. kernel-doc:: drivers/gpu/drm/drm_atomic.c + :internal: + CRTC Abstraction ================
diff --git a/Documentation/gpu/kms-properties.csv b/Documentation/gpu/kms-properties.csv index 927b65e..6b28b01 100644 --- a/Documentation/gpu/kms-properties.csv +++ b/Documentation/gpu/kms-properties.csv
@@ -1,5 +1,4 @@ Owner Module/Drivers,Group,Property Name,Type,Property Values,Object attached,Description/Restrictions -,,“scaling mode”,ENUM,"{ ""None"", ""Full"", ""Center"", ""Full aspect"" }",Connector,"Supported by: amdgpu, gma500, i915, nouveau and radeon." ,DVI-I,“subconnector”,ENUM,"{ “Unknown”, “DVI-D”, “DVI-A” }",Connector,TBD ,,“select subconnector”,ENUM,"{ “Automatic”, “DVI-D”, “DVI-A” }",Connector,TBD ,TV,“subconnector”,ENUM,"{ ""Unknown"", ""Composite"", ""SVIDEO"", ""Component"", ""SCART"" }",Connector,TBD
diff --git a/MAINTAINERS b/MAINTAINERS index c861cd9..f15fa98 100644 --- a/MAINTAINERS +++ b/MAINTAINERS
@@ -755,6 +755,8 @@ F: drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h F: drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c F: drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c +F: drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c +F: drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c F: drivers/gpu/drm/amd/amdkfd/ F: drivers/gpu/drm/amd/include/cik_structs.h F: drivers/gpu/drm/amd/include/kgd_kfd_interface.h
diff --git a/drivers/dma-buf/reservation.c b/drivers/dma-buf/reservation.c index 750d0404..1bc15f61 100644 --- a/drivers/dma-buf/reservation.c +++ b/drivers/dma-buf/reservation.c
@@ -374,8 +374,9 @@ EXPORT_SYMBOL(reservation_object_copy_fences); * @pshared: the array of shared fence ptrs returned (array is krealloc'd to * the required size, and must be freed by caller) * - * RETURNS - * Zero or -errno + * Retrieve all fences from the reservation object. If the pointer for the + * exclusive fence is not specified the fence is put into the array of the + * shared fences as well. Returns either zero or -ENOMEM. */ int reservation_object_get_fences_rcu(struct reservation_object *obj, struct dma_fence **pfence_excl, @@ -389,8 +390,8 @@ int reservation_object_get_fences_rcu(struct reservation_object *obj, do { struct reservation_object_list *fobj; - unsigned seq; - unsigned int i; + unsigned int i, seq; + size_t sz = 0; shared_count = i = 0; @@ -402,9 +403,14 @@ int reservation_object_get_fences_rcu(struct reservation_object *obj, goto unlock; fobj = rcu_dereference(obj->fence); - if (fobj) { + if (fobj) + sz += sizeof(*shared) * fobj->shared_max; + + if (!pfence_excl && fence_excl) + sz += sizeof(*shared); + + if (sz) { struct dma_fence **nshared; - size_t sz = sizeof(*shared) * fobj->shared_max; nshared = krealloc(shared, sz, GFP_NOWAIT | __GFP_NOWARN); @@ -420,13 +426,19 @@ int reservation_object_get_fences_rcu(struct reservation_object *obj, break; } shared = nshared; - shared_count = fobj->shared_count; - + shared_count = fobj ? fobj->shared_count : 0; for (i = 0; i < shared_count; ++i) { shared[i] = rcu_dereference(fobj->shared[i]); if (!dma_fence_get_rcu(shared[i])) break; } + + if (!pfence_excl && fence_excl) { + shared[i] = fence_excl; + fence_excl = NULL; + ++i; + ++shared_count; + } } if (i != shared_count || read_seqcount_retry(&obj->seq, seq)) { @@ -448,7 +460,8 @@ int reservation_object_get_fences_rcu(struct reservation_object *obj, *pshared_count = shared_count; *pshared = shared; - *pfence_excl = fence_excl; + if (pfence_excl) + *pfence_excl = fence_excl; return ret; }
diff --git a/drivers/gpu/drm/amd/acp/include/acp_gfx_if.h b/drivers/gpu/drm/amd/acp/include/acp_gfx_if.h index a72ddb2f..feab8eb 100644 --- a/drivers/gpu/drm/amd/acp/include/acp_gfx_if.h +++ b/drivers/gpu/drm/amd/acp/include/acp_gfx_if.h
@@ -25,7 +25,6 @@ #define _ACP_GFX_IF_H #include <linux/types.h> -#include "cgs_linux.h" #include "cgs_common.h" int amd_acp_hw_init(struct cgs_device *cgs_device,
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index d6e5b72..2ca2b51 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -30,7 +30,6 @@ ccflags-y := -I$(FULL_AMD_PATH)/include/asic_reg \ -I$(FULL_AMD_PATH)/include \ -I$(FULL_AMD_PATH)/amdgpu \ - -I$(FULL_AMD_PATH)/scheduler \ -I$(FULL_AMD_PATH)/powerplay/inc \ -I$(FULL_AMD_PATH)/acp/include \ -I$(FULL_AMD_DISPLAY_PATH) \ @@ -63,7 +62,7 @@ amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o si_dpm.o si_smc.o amdgpu-y += \ - vi.o mxgpu_vi.o nbio_v6_1.o soc15.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o + vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o # add GMC block amdgpu-y += \ @@ -88,8 +87,7 @@ # add SMC block amdgpu-y += \ - amdgpu_dpm.o \ - amdgpu_powerplay.o + amdgpu_dpm.o # add DCE block amdgpu-y += \ @@ -130,6 +128,8 @@ # add amdkfd interfaces amdgpu-y += \ amdgpu_amdkfd.o \ + amdgpu_amdkfd_fence.o \ + amdgpu_amdkfd_gpuvm.o \ amdgpu_amdkfd_gfx_v8.o # add cgs
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 37f5f4a..6022f91 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -68,6 +68,7 @@ #include "amdgpu_vce.h" #include "amdgpu_vcn.h" #include "amdgpu_mn.h" +#include "amdgpu_gmc.h" #include "amdgpu_dm.h" #include "amdgpu_virt.h" #include "amdgpu_gart.h" @@ -127,6 +128,7 @@ extern int amdgpu_job_hang_limit; extern int amdgpu_lbpw; extern int amdgpu_compute_multipipe; extern int amdgpu_gpu_recovery; +extern int amdgpu_emu_mode; #ifdef CONFIG_DRM_AMDGPU_SI extern int amdgpu_si_support; @@ -180,10 +182,6 @@ extern int amdgpu_cik_support; #define CIK_CURSOR_WIDTH 128 #define CIK_CURSOR_HEIGHT 128 -/* GPU RESET flags */ -#define AMDGPU_RESET_INFO_VRAM_LOST (1 << 0) -#define AMDGPU_RESET_INFO_FULLRESET (1 << 1) - struct amdgpu_device; struct amdgpu_ib; struct amdgpu_cs_parser; @@ -319,13 +317,6 @@ struct amdgpu_vm_pte_funcs { void (*write_pte)(struct amdgpu_ib *ib, uint64_t pe, uint64_t value, unsigned count, uint32_t incr); - - /* maximum nums of PTEs/PDEs in a single operation */ - uint32_t set_max_nums_pte_pde; - - /* number of dw to reserve per operation */ - unsigned set_pte_pde_num_dw; - /* for linear pte/pde updates without addr mapping */ void (*set_pte_pde)(struct amdgpu_ib *ib, uint64_t pe, @@ -333,28 +324,6 @@ struct amdgpu_vm_pte_funcs { uint32_t incr, uint64_t flags); }; -/* provided by the gmc block */ -struct amdgpu_gart_funcs { - /* flush the vm tlb via mmio */ - void (*flush_gpu_tlb)(struct amdgpu_device *adev, - uint32_t vmid); - /* write pte/pde updates using the cpu */ - int (*set_pte_pde)(struct amdgpu_device *adev, - void *cpu_pt_addr, /* cpu addr of page table */ - uint32_t gpu_page_idx, /* pte/pde to update */ - uint64_t addr, /* addr to write into pte/pde */ - uint64_t flags); /* access flags */ - /* enable/disable PRT support */ - void (*set_prt)(struct amdgpu_device *adev, bool enable); - /* set pte flags based per asic */ - uint64_t (*get_vm_pte_flags)(struct amdgpu_device *adev, - uint32_t flags); - /* get the pde for a given mc addr */ - void (*get_vm_pde)(struct amdgpu_device *adev, int level, - u64 *dst, u64 *flags); - uint32_t (*get_invalidate_req)(unsigned int vmid); -}; - /* provided by the ih block */ struct amdgpu_ih_funcs { /* ring read/write ptr handling, called from interrupt context */ @@ -372,14 +341,6 @@ bool amdgpu_get_bios(struct amdgpu_device *adev); bool amdgpu_read_bios(struct amdgpu_device *adev); /* - * Dummy page - */ -struct amdgpu_dummy_page { - struct page *page; - dma_addr_t addr; -}; - -/* * Clocks */ @@ -421,8 +382,6 @@ struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev, int flags); struct drm_gem_object *amdgpu_gem_prime_import(struct drm_device *dev, struct dma_buf *dma_buf); -int amdgpu_gem_prime_pin(struct drm_gem_object *obj); -void amdgpu_gem_prime_unpin(struct drm_gem_object *obj); struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *); void *amdgpu_gem_prime_vmap(struct drm_gem_object *obj); void amdgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr); @@ -483,7 +442,7 @@ struct amdgpu_sa_bo { void amdgpu_gem_force_release(struct amdgpu_device *adev); int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size, int alignment, u32 initial_domain, - u64 flags, bool kernel, + u64 flags, enum ttm_bo_type type, struct reservation_object *resv, struct drm_gem_object **obj); @@ -497,56 +456,6 @@ int amdgpu_fence_slab_init(void); void amdgpu_fence_slab_fini(void); /* - * VMHUB structures, functions & helpers - */ -struct amdgpu_vmhub { - uint32_t ctx0_ptb_addr_lo32; - uint32_t ctx0_ptb_addr_hi32; - uint32_t vm_inv_eng0_req; - uint32_t vm_inv_eng0_ack; - uint32_t vm_context0_cntl; - uint32_t vm_l2_pro_fault_status; - uint32_t vm_l2_pro_fault_cntl; -}; - -/* - * GPU MC structures, functions & helpers - */ -struct amdgpu_mc { - resource_size_t aper_size; - resource_size_t aper_base; - resource_size_t agp_base; - /* for some chips with <= 32MB we need to lie - * about vram size near mc fb location */ - u64 mc_vram_size; - u64 visible_vram_size; - u64 gart_size; - u64 gart_start; - u64 gart_end; - u64 vram_start; - u64 vram_end; - unsigned vram_width; - u64 real_vram_size; - int vram_mtrr; - u64 mc_mask; - const struct firmware *fw; /* MC firmware */ - uint32_t fw_version; - struct amdgpu_irq_src vm_fault; - uint32_t vram_type; - uint32_t srbm_soft_reset; - bool prt_warning; - uint64_t stolen_size; - /* apertures */ - u64 shared_aperture_start; - u64 shared_aperture_end; - u64 private_aperture_start; - u64 private_aperture_end; - /* protects concurrent invalidation */ - spinlock_t invalidate_lock; - bool translate_further; -}; - -/* * GPU doorbell structures, functions & helpers */ typedef enum _AMDGPU_DOORBELL_ASSIGNMENT @@ -982,6 +891,7 @@ struct amdgpu_gfx_funcs { void (*read_wave_data)(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t *dst, int *no_fields); void (*read_wave_vgprs)(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t thread, uint32_t start, uint32_t size, uint32_t *dst); void (*read_wave_sgprs)(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t start, uint32_t size, uint32_t *dst); + void (*select_me_pipe_q)(struct amdgpu_device *adev, u32 me, u32 pipe, u32 queue); }; struct amdgpu_ngg_buf { @@ -1128,8 +1038,9 @@ struct amdgpu_job { void *owner; uint64_t fence_ctx; /* the fence_context this job uses */ bool vm_needs_flush; - unsigned vmid; uint64_t vm_pd_addr; + unsigned vmid; + unsigned pasid; uint32_t gds_base, gds_size; uint32_t gws_base, gws_size; uint32_t oa_base, oa_size; @@ -1172,8 +1083,6 @@ struct amdgpu_wb { int amdgpu_device_wb_get(struct amdgpu_device *adev, u32 *wb); void amdgpu_device_wb_free(struct amdgpu_device *adev, u32 wb); -void amdgpu_device_get_pcie_info(struct amdgpu_device *adev); - /* * SDMA */ @@ -1291,6 +1200,11 @@ struct amdgpu_asic_funcs { void (*set_pcie_lanes)(struct amdgpu_device *adev, int lanes); /* get config memsize register */ u32 (*get_config_memsize)(struct amdgpu_device *adev); + /* flush hdp write queue */ + void (*flush_hdp)(struct amdgpu_device *adev, struct amdgpu_ring *ring); + /* invalidate hdp read cache */ + void (*invalidate_hdp)(struct amdgpu_device *adev, + struct amdgpu_ring *ring); }; /* @@ -1434,7 +1348,7 @@ struct amdgpu_nbio_funcs { u32 (*get_pcie_data_offset)(struct amdgpu_device *adev); u32 (*get_rev_id)(struct amdgpu_device *adev); void (*mc_access_enable)(struct amdgpu_device *adev, bool enable); - void (*hdp_flush)(struct amdgpu_device *adev); + void (*hdp_flush)(struct amdgpu_device *adev, struct amdgpu_ring *ring); u32 (*get_memsize)(struct amdgpu_device *adev); void (*sdma_doorbell_range)(struct amdgpu_device *adev, int instance, bool use_doorbell, int doorbell_index); @@ -1466,6 +1380,7 @@ enum amd_hw_ip_block_type { ATHUB_HWIP, NBIO_HWIP, MP0_HWIP, + MP1_HWIP, UVD_HWIP, VCN_HWIP = UVD_HWIP, VCE_HWIP, @@ -1475,15 +1390,14 @@ enum amd_hw_ip_block_type { SMUIO_HWIP, PWR_HWIP, NBIF_HWIP, + THM_HWIP, MAX_HWIP }; #define HWIP_MAX_INSTANCE 6 struct amd_powerplay { - struct cgs_device *cgs_device; void *pp_handle; - const struct amd_ip_funcs *ip_funcs; const struct amd_pm_funcs *pp_funcs; }; @@ -1507,6 +1421,7 @@ struct amdgpu_device { const struct amdgpu_asic_funcs *asic_funcs; bool shutdown; bool need_dma32; + bool need_swiotlb; bool accel_working; struct work_struct reset_work; struct notifier_block acpi_nb; @@ -1576,9 +1491,9 @@ struct amdgpu_device { struct amdgpu_clock clock; /* MC */ - struct amdgpu_mc mc; + struct amdgpu_gmc gmc; struct amdgpu_gart gart; - struct amdgpu_dummy_page dummy_page; + dma_addr_t dummy_page_addr; struct amdgpu_vm_manager vm_manager; struct amdgpu_vmhub vmhub[AMDGPU_MAX_VMHUBS]; @@ -1717,6 +1632,9 @@ uint32_t amdgpu_mm_rreg(struct amdgpu_device *adev, uint32_t reg, uint32_t acc_flags); void amdgpu_mm_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, uint32_t acc_flags); +void amdgpu_mm_wreg8(struct amdgpu_device *adev, uint32_t offset, uint8_t value); +uint8_t amdgpu_mm_rreg8(struct amdgpu_device *adev, uint32_t offset); + u32 amdgpu_io_rreg(struct amdgpu_device *adev, u32 reg); void amdgpu_io_wreg(struct amdgpu_device *adev, u32 reg, u32 v); @@ -1728,6 +1646,8 @@ void amdgpu_mm_wdoorbell64(struct amdgpu_device *adev, u32 index, u64 v); bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type); bool amdgpu_device_has_dc_support(struct amdgpu_device *adev); +int emu_soc_asic_init(struct amdgpu_device *adev); + /* * Registers read & write functions. */ @@ -1738,6 +1658,9 @@ bool amdgpu_device_has_dc_support(struct amdgpu_device *adev); #define RREG32_NO_KIQ(reg) amdgpu_mm_rreg(adev, (reg), AMDGPU_REGS_NO_KIQ) #define WREG32_NO_KIQ(reg, v) amdgpu_mm_wreg(adev, (reg), (v), AMDGPU_REGS_NO_KIQ) +#define RREG8(reg) amdgpu_mm_rreg8(adev, (reg)) +#define WREG8(reg, v) amdgpu_mm_wreg8(adev, (reg), (v)) + #define RREG32(reg) amdgpu_mm_rreg(adev, (reg), 0) #define RREG32_IDX(reg) amdgpu_mm_rreg(adev, (reg), AMDGPU_REGS_IDX) #define DREG32(reg) printk(KERN_INFO "REGISTER: " #reg " : 0x%08X\n", amdgpu_mm_rreg(adev, (reg), 0)) @@ -1840,13 +1763,17 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_asic_read_bios_from_rom(adev, b, l) (adev)->asic_funcs->read_bios_from_rom((adev), (b), (l)) #define amdgpu_asic_read_register(adev, se, sh, offset, v)((adev)->asic_funcs->read_register((adev), (se), (sh), (offset), (v))) #define amdgpu_asic_get_config_memsize(adev) (adev)->asic_funcs->get_config_memsize((adev)) -#define amdgpu_gart_flush_gpu_tlb(adev, vmid) (adev)->gart.gart_funcs->flush_gpu_tlb((adev), (vmid)) -#define amdgpu_gart_set_pte_pde(adev, pt, idx, addr, flags) (adev)->gart.gart_funcs->set_pte_pde((adev), (pt), (idx), (addr), (flags)) -#define amdgpu_gart_get_vm_pde(adev, level, dst, flags) (adev)->gart.gart_funcs->get_vm_pde((adev), (level), (dst), (flags)) +#define amdgpu_asic_flush_hdp(adev, r) (adev)->asic_funcs->flush_hdp((adev), (r)) +#define amdgpu_asic_invalidate_hdp(adev, r) (adev)->asic_funcs->invalidate_hdp((adev), (r)) +#define amdgpu_gmc_flush_gpu_tlb(adev, vmid) (adev)->gmc.gmc_funcs->flush_gpu_tlb((adev), (vmid)) +#define amdgpu_gmc_emit_flush_gpu_tlb(r, vmid, addr) (r)->adev->gmc.gmc_funcs->emit_flush_gpu_tlb((r), (vmid), (addr)) +#define amdgpu_gmc_emit_pasid_mapping(r, vmid, pasid) (r)->adev->gmc.gmc_funcs->emit_pasid_mapping((r), (vmid), (pasid)) +#define amdgpu_gmc_set_pte_pde(adev, pt, idx, addr, flags) (adev)->gmc.gmc_funcs->set_pte_pde((adev), (pt), (idx), (addr), (flags)) +#define amdgpu_gmc_get_vm_pde(adev, level, dst, flags) (adev)->gmc.gmc_funcs->get_vm_pde((adev), (level), (dst), (flags)) +#define amdgpu_gmc_get_pte_flags(adev, flags) (adev)->gmc.gmc_funcs->get_vm_pte_flags((adev),(flags)) #define amdgpu_vm_copy_pte(adev, ib, pe, src, count) ((adev)->vm_manager.vm_pte_funcs->copy_pte((ib), (pe), (src), (count))) #define amdgpu_vm_write_pte(adev, ib, pe, value, count, incr) ((adev)->vm_manager.vm_pte_funcs->write_pte((ib), (pe), (value), (count), (incr))) #define amdgpu_vm_set_pte_pde(adev, ib, pe, addr, count, incr, flags) ((adev)->vm_manager.vm_pte_funcs->set_pte_pde((ib), (pe), (addr), (count), (incr), (flags))) -#define amdgpu_vm_get_pte_flags(adev, flags) (adev)->gart.gart_funcs->get_vm_pte_flags((adev),(flags)) #define amdgpu_ring_parse_cs(r, p, ib) ((r)->funcs->parse_cs((p), (ib))) #define amdgpu_ring_test_ring(r) (r)->funcs->test_ring((r)) #define amdgpu_ring_test_ib(r, t) (r)->funcs->test_ib((r), (t)) @@ -1859,11 +1786,11 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_ring_emit_fence(r, addr, seq, flags) (r)->funcs->emit_fence((r), (addr), (seq), (flags)) #define amdgpu_ring_emit_gds_switch(r, v, db, ds, wb, ws, ab, as) (r)->funcs->emit_gds_switch((r), (v), (db), (ds), (wb), (ws), (ab), (as)) #define amdgpu_ring_emit_hdp_flush(r) (r)->funcs->emit_hdp_flush((r)) -#define amdgpu_ring_emit_hdp_invalidate(r) (r)->funcs->emit_hdp_invalidate((r)) #define amdgpu_ring_emit_switch_buffer(r) (r)->funcs->emit_switch_buffer((r)) #define amdgpu_ring_emit_cntxcntl(r, d) (r)->funcs->emit_cntxcntl((r), (d)) #define amdgpu_ring_emit_rreg(r, d) (r)->funcs->emit_rreg((r), (d)) #define amdgpu_ring_emit_wreg(r, d, v) (r)->funcs->emit_wreg((r), (d), (v)) +#define amdgpu_ring_emit_reg_wait(r, d, v, m) (r)->funcs->emit_reg_wait((r), (d), (v), (m)) #define amdgpu_ring_emit_tmz(r, b) (r)->funcs->emit_tmz((r), (b)) #define amdgpu_ring_pad_ib(r, ib) ((r)->funcs->pad_ib((r), (ib))) #define amdgpu_ring_init_cond_exec(r) (r)->funcs->init_cond_exec((r)) @@ -1873,7 +1800,6 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_ih_decode_iv(adev, iv) (adev)->irq.ih_funcs->decode_iv((adev), (iv)) #define amdgpu_ih_set_rptr(adev) (adev)->irq.ih_funcs->set_rptr((adev)) #define amdgpu_display_vblank_get_counter(adev, crtc) (adev)->mode_info.funcs->vblank_get_counter((adev), (crtc)) -#define amdgpu_display_vblank_wait(adev, crtc) (adev)->mode_info.funcs->vblank_wait((adev), (crtc)) #define amdgpu_display_backlight_set_level(adev, e, l) (adev)->mode_info.funcs->backlight_set_level((e), (l)) #define amdgpu_display_backlight_get_level(adev, e) (adev)->mode_info.funcs->backlight_get_level((e)) #define amdgpu_display_hpd_sense(adev, h) (adev)->mode_info.funcs->hpd_sense((adev), (h)) @@ -1890,26 +1816,24 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_gfx_select_se_sh(adev, se, sh, instance) (adev)->gfx.funcs->select_se_sh((adev), (se), (sh), (instance)) #define amdgpu_gds_switch(adev, r, v, d, w, a) (adev)->gds.funcs->patch_gds_switch((r), (v), (d), (w), (a)) #define amdgpu_psp_check_fw_loading_status(adev, i) (adev)->firmware.funcs->check_fw_loading_status((adev), (i)) +#define amdgpu_gfx_select_me_pipe_q(adev, me, pipe, q) (adev)->gfx.funcs->select_me_pipe_q((adev), (me), (pipe), (q)) /* Common functions */ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, struct amdgpu_job* job, bool force); void amdgpu_device_pci_config_reset(struct amdgpu_device *adev); bool amdgpu_device_need_post(struct amdgpu_device *adev); -void amdgpu_update_display_priority(struct amdgpu_device *adev); +void amdgpu_display_update_priority(struct amdgpu_device *adev); void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev, u64 num_bytes, u64 num_vis_bytes); void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain); bool amdgpu_ttm_bo_is_amdgpu_bo(struct ttm_buffer_object *bo); void amdgpu_device_vram_location(struct amdgpu_device *adev, - struct amdgpu_mc *mc, u64 base); + struct amdgpu_gmc *mc, u64 base); void amdgpu_device_gart_location(struct amdgpu_device *adev, - struct amdgpu_mc *mc); + struct amdgpu_gmc *mc); int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev); -void amdgpu_ttm_set_active_vram_size(struct amdgpu_device *adev, u64 size); -int amdgpu_ttm_init(struct amdgpu_device *adev); -void amdgpu_ttm_fini(struct amdgpu_device *adev); void amdgpu_device_program_register_sequence(struct amdgpu_device *adev, const u32 *registers, const u32 array_size);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 1d605e1..4d36203 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -30,6 +30,8 @@ const struct kgd2kfd_calls *kgd2kfd; bool (*kgd2kfd_init_p)(unsigned int, const struct kgd2kfd_calls**); +static const unsigned int compute_vmid_bitmap = 0xFF00; + int amdgpu_amdkfd_init(void) { int ret; @@ -56,6 +58,7 @@ int amdgpu_amdkfd_init(void) #else ret = -ENOENT; #endif + amdgpu_amdkfd_gpuvm_init_mem_limits(); return ret; } @@ -78,10 +81,15 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev) switch (adev->asic_type) { #ifdef CONFIG_DRM_AMDGPU_CIK case CHIP_KAVERI: + case CHIP_HAWAII: kfd2kgd = amdgpu_amdkfd_gfx_7_get_functions(); break; #endif case CHIP_CARRIZO: + case CHIP_TONGA: + case CHIP_FIJI: + case CHIP_POLARIS10: + case CHIP_POLARIS11: kfd2kgd = amdgpu_amdkfd_gfx_8_0_get_functions(); break; default: @@ -132,9 +140,13 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev) int last_valid_bit; if (adev->kfd) { struct kgd2kfd_shared_resources gpu_resources = { - .compute_vmid_bitmap = 0xFF00, + .compute_vmid_bitmap = compute_vmid_bitmap, .num_pipe_per_mec = adev->gfx.mec.num_pipe_per_mec, - .num_queue_per_pipe = adev->gfx.mec.num_queue_per_pipe + .num_queue_per_pipe = adev->gfx.mec.num_queue_per_pipe, + .gpuvm_size = min(adev->vm_manager.max_pfn + << AMDGPU_GPU_PAGE_SHIFT, + AMDGPU_VA_HOLE_START), + .drm_render_minor = adev->ddev->render->index }; /* this is going to have a few of the MSBs set that we need to @@ -204,20 +216,14 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, void **cpu_ptr) { struct amdgpu_device *adev = (struct amdgpu_device *)kgd; - struct kgd_mem **mem = (struct kgd_mem **) mem_obj; + struct amdgpu_bo *bo = NULL; int r; + uint64_t gpu_addr_tmp = 0; + void *cpu_ptr_tmp = NULL; - BUG_ON(kgd == NULL); - BUG_ON(gpu_addr == NULL); - BUG_ON(cpu_ptr == NULL); - - *mem = kmalloc(sizeof(struct kgd_mem), GFP_KERNEL); - if ((*mem) == NULL) - return -ENOMEM; - - r = amdgpu_bo_create(adev, size, PAGE_SIZE, true, AMDGPU_GEM_DOMAIN_GTT, - AMDGPU_GEM_CREATE_CPU_GTT_USWC, NULL, NULL, 0, - &(*mem)->bo); + r = amdgpu_bo_create(adev, size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT, + AMDGPU_GEM_CREATE_CPU_GTT_USWC, ttm_bo_type_kernel, + NULL, &bo); if (r) { dev_err(adev->dev, "failed to allocate BO for amdkfd (%d)\n", r); @@ -225,54 +231,53 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, } /* map the buffer */ - r = amdgpu_bo_reserve((*mem)->bo, true); + r = amdgpu_bo_reserve(bo, true); if (r) { dev_err(adev->dev, "(%d) failed to reserve bo for amdkfd\n", r); goto allocate_mem_reserve_bo_failed; } - r = amdgpu_bo_pin((*mem)->bo, AMDGPU_GEM_DOMAIN_GTT, - &(*mem)->gpu_addr); + r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, + &gpu_addr_tmp); if (r) { dev_err(adev->dev, "(%d) failed to pin bo for amdkfd\n", r); goto allocate_mem_pin_bo_failed; } - *gpu_addr = (*mem)->gpu_addr; - r = amdgpu_bo_kmap((*mem)->bo, &(*mem)->cpu_ptr); + r = amdgpu_bo_kmap(bo, &cpu_ptr_tmp); if (r) { dev_err(adev->dev, "(%d) failed to map bo to kernel for amdkfd\n", r); goto allocate_mem_kmap_bo_failed; } - *cpu_ptr = (*mem)->cpu_ptr; - amdgpu_bo_unreserve((*mem)->bo); + *mem_obj = bo; + *gpu_addr = gpu_addr_tmp; + *cpu_ptr = cpu_ptr_tmp; + + amdgpu_bo_unreserve(bo); return 0; allocate_mem_kmap_bo_failed: - amdgpu_bo_unpin((*mem)->bo); + amdgpu_bo_unpin(bo); allocate_mem_pin_bo_failed: - amdgpu_bo_unreserve((*mem)->bo); + amdgpu_bo_unreserve(bo); allocate_mem_reserve_bo_failed: - amdgpu_bo_unref(&(*mem)->bo); + amdgpu_bo_unref(&bo); return r; } void free_gtt_mem(struct kgd_dev *kgd, void *mem_obj) { - struct kgd_mem *mem = (struct kgd_mem *) mem_obj; + struct amdgpu_bo *bo = (struct amdgpu_bo *) mem_obj; - BUG_ON(mem == NULL); - - amdgpu_bo_reserve(mem->bo, true); - amdgpu_bo_kunmap(mem->bo); - amdgpu_bo_unpin(mem->bo); - amdgpu_bo_unreserve(mem->bo); - amdgpu_bo_unref(&(mem->bo)); - kfree(mem); + amdgpu_bo_reserve(bo, true); + amdgpu_bo_kunmap(bo); + amdgpu_bo_unpin(bo); + amdgpu_bo_unreserve(bo); + amdgpu_bo_unref(&(bo)); } void get_local_mem_info(struct kgd_dev *kgd, @@ -281,24 +286,29 @@ void get_local_mem_info(struct kgd_dev *kgd, struct amdgpu_device *adev = (struct amdgpu_device *)kgd; uint64_t address_mask = adev->dev->dma_mask ? ~*adev->dev->dma_mask : ~((1ULL << 32) - 1); - resource_size_t aper_limit = adev->mc.aper_base + adev->mc.aper_size; + resource_size_t aper_limit = adev->gmc.aper_base + adev->gmc.aper_size; memset(mem_info, 0, sizeof(*mem_info)); - if (!(adev->mc.aper_base & address_mask || aper_limit & address_mask)) { - mem_info->local_mem_size_public = adev->mc.visible_vram_size; - mem_info->local_mem_size_private = adev->mc.real_vram_size - - adev->mc.visible_vram_size; + if (!(adev->gmc.aper_base & address_mask || aper_limit & address_mask)) { + mem_info->local_mem_size_public = adev->gmc.visible_vram_size; + mem_info->local_mem_size_private = adev->gmc.real_vram_size - + adev->gmc.visible_vram_size; } else { mem_info->local_mem_size_public = 0; - mem_info->local_mem_size_private = adev->mc.real_vram_size; + mem_info->local_mem_size_private = adev->gmc.real_vram_size; } - mem_info->vram_width = adev->mc.vram_width; + mem_info->vram_width = adev->gmc.vram_width; pr_debug("Address base: %pap limit %pap public 0x%llx private 0x%llx\n", - &adev->mc.aper_base, &aper_limit, + &adev->gmc.aper_base, &aper_limit, mem_info->local_mem_size_public, mem_info->local_mem_size_private); + if (amdgpu_emu_mode == 1) { + mem_info->mem_clk_max = 100; + return; + } + if (amdgpu_sriov_vf(adev)) mem_info->mem_clk_max = adev->clock.default_mclk / 100; else @@ -319,6 +329,9 @@ uint32_t get_max_engine_clock_in_mhz(struct kgd_dev *kgd) struct amdgpu_device *adev = (struct amdgpu_device *)kgd; /* the sclk is in quantas of 10kHz */ + if (amdgpu_emu_mode == 1) + return 100; + if (amdgpu_sriov_vf(adev)) return adev->clock.default_sclk / 100; @@ -354,3 +367,68 @@ uint64_t amdgpu_amdkfd_get_vram_usage(struct kgd_dev *kgd) return amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); } + +int amdgpu_amdkfd_submit_ib(struct kgd_dev *kgd, enum kgd_engine_type engine, + uint32_t vmid, uint64_t gpu_addr, + uint32_t *ib_cmd, uint32_t ib_len) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)kgd; + struct amdgpu_job *job; + struct amdgpu_ib *ib; + struct amdgpu_ring *ring; + struct dma_fence *f = NULL; + int ret; + + switch (engine) { + case KGD_ENGINE_MEC1: + ring = &adev->gfx.compute_ring[0]; + break; + case KGD_ENGINE_SDMA1: + ring = &adev->sdma.instance[0].ring; + break; + case KGD_ENGINE_SDMA2: + ring = &adev->sdma.instance[1].ring; + break; + default: + pr_err("Invalid engine in IB submission: %d\n", engine); + ret = -EINVAL; + goto err; + } + + ret = amdgpu_job_alloc(adev, 1, &job, NULL); + if (ret) + goto err; + + ib = &job->ibs[0]; + memset(ib, 0, sizeof(struct amdgpu_ib)); + + ib->gpu_addr = gpu_addr; + ib->ptr = ib_cmd; + ib->length_dw = ib_len; + /* This works for NO_HWS. TODO: need to handle without knowing VMID */ + job->vmid = vmid; + + ret = amdgpu_ib_schedule(ring, 1, ib, job, &f); + if (ret) { + DRM_ERROR("amdgpu: failed to schedule IB.\n"); + goto err_ib_sched; + } + + ret = dma_fence_wait(f, false); + +err_ib_sched: + dma_fence_put(f); + amdgpu_job_free(job); +err: + return ret; +} + +bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid) +{ + if (adev->kfd) { + if ((1 << vmid) & compute_vmid_bitmap) + return true; + } + + return false; +}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index e515ca0..c2c2bea 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -29,13 +29,68 @@ #include <linux/mm.h> #include <linux/mmu_context.h> #include <kgd_kfd_interface.h> +#include <drm/ttm/ttm_execbuf_util.h> +#include "amdgpu_sync.h" +#include "amdgpu_vm.h" + +extern const struct kgd2kfd_calls *kgd2kfd; struct amdgpu_device; +struct kfd_bo_va_list { + struct list_head bo_list; + struct amdgpu_bo_va *bo_va; + void *kgd_dev; + bool is_mapped; + uint64_t va; + uint64_t pte_flags; +}; + struct kgd_mem { + struct mutex lock; struct amdgpu_bo *bo; - uint64_t gpu_addr; - void *cpu_ptr; + struct list_head bo_va_list; + /* protected by amdkfd_process_info.lock */ + struct ttm_validate_buffer validate_list; + struct ttm_validate_buffer resv_list; + uint32_t domain; + unsigned int mapped_to_gpu_memory; + uint64_t va; + + uint32_t mapping_flags; + + struct amdkfd_process_info *process_info; + + struct amdgpu_sync sync; + + bool aql_queue; +}; + +/* KFD Memory Eviction */ +struct amdgpu_amdkfd_fence { + struct dma_fence base; + struct mm_struct *mm; + spinlock_t lock; + char timeline_name[TASK_COMM_LEN]; +}; + +struct amdgpu_amdkfd_fence *amdgpu_amdkfd_fence_create(u64 context, + struct mm_struct *mm); +bool amdkfd_fence_check_mm(struct dma_fence *f, struct mm_struct *mm); +struct amdgpu_amdkfd_fence *to_amdgpu_amdkfd_fence(struct dma_fence *f); + +struct amdkfd_process_info { + /* List head of all VMs that belong to a KFD process */ + struct list_head vm_list_head; + /* List head for all KFD BOs that belong to a KFD process. */ + struct list_head kfd_bo_list; + /* Lock to protect kfd_bo_list */ + struct mutex lock; + + /* Number of VMs */ + unsigned int n_vms; + /* Eviction Fence */ + struct amdgpu_amdkfd_fence *eviction_fence; }; int amdgpu_amdkfd_init(void); @@ -49,9 +104,15 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev); void amdgpu_amdkfd_device_init(struct amdgpu_device *adev); void amdgpu_amdkfd_device_fini(struct amdgpu_device *adev); +int amdgpu_amdkfd_submit_ib(struct kgd_dev *kgd, enum kgd_engine_type engine, + uint32_t vmid, uint64_t gpu_addr, + uint32_t *ib_cmd, uint32_t ib_len); + struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void); struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void); +bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid); + /* Shared API */ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, void **mem_obj, uint64_t *gpu_addr, @@ -80,4 +141,36 @@ uint64_t amdgpu_amdkfd_get_vram_usage(struct kgd_dev *kgd); valid; \ }) +/* GPUVM API */ +int amdgpu_amdkfd_gpuvm_create_process_vm(struct kgd_dev *kgd, void **vm, + void **process_info, + struct dma_fence **ef); +int amdgpu_amdkfd_gpuvm_acquire_process_vm(struct kgd_dev *kgd, + struct file *filp, + void **vm, void **process_info, + struct dma_fence **ef); +void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev, + struct amdgpu_vm *vm); +void amdgpu_amdkfd_gpuvm_destroy_process_vm(struct kgd_dev *kgd, void *vm); +uint32_t amdgpu_amdkfd_gpuvm_get_process_page_dir(void *vm); +int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( + struct kgd_dev *kgd, uint64_t va, uint64_t size, + void *vm, struct kgd_mem **mem, + uint64_t *offset, uint32_t flags); +int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( + struct kgd_dev *kgd, struct kgd_mem *mem); +int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( + struct kgd_dev *kgd, struct kgd_mem *mem, void *vm); +int amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu( + struct kgd_dev *kgd, struct kgd_mem *mem, void *vm); +int amdgpu_amdkfd_gpuvm_sync_memory( + struct kgd_dev *kgd, struct kgd_mem *mem, bool intr); +int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_dev *kgd, + struct kgd_mem *mem, void **kptr, uint64_t *size); +int amdgpu_amdkfd_gpuvm_restore_process_bos(void *process_info, + struct dma_fence **ef); + +void amdgpu_amdkfd_gpuvm_init_mem_limits(void); +void amdgpu_amdkfd_unreserve_system_memory_limit(struct amdgpu_bo *bo); + #endif /* AMDGPU_AMDKFD_H_INCLUDED */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c new file mode 100644 index 0000000..2c14025 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c
@@ -0,0 +1,179 @@ +/* + * Copyright 2016-2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include <linux/dma-fence.h> +#include <linux/spinlock.h> +#include <linux/atomic.h> +#include <linux/stacktrace.h> +#include <linux/sched.h> +#include <linux/slab.h> +#include <linux/sched/mm.h> +#include "amdgpu_amdkfd.h" + +static const struct dma_fence_ops amdkfd_fence_ops; +static atomic_t fence_seq = ATOMIC_INIT(0); + +/* Eviction Fence + * Fence helper functions to deal with KFD memory eviction. + * Big Idea - Since KFD submissions are done by user queues, a BO cannot be + * evicted unless all the user queues for that process are evicted. + * + * All the BOs in a process share an eviction fence. When process X wants + * to map VRAM memory but TTM can't find enough space, TTM will attempt to + * evict BOs from its LRU list. TTM checks if the BO is valuable to evict + * by calling ttm_bo_driver->eviction_valuable(). + * + * ttm_bo_driver->eviction_valuable() - will return false if the BO belongs + * to process X. Otherwise, it will return true to indicate BO can be + * evicted by TTM. + * + * If ttm_bo_driver->eviction_valuable returns true, then TTM will continue + * the evcition process for that BO by calling ttm_bo_evict --> amdgpu_bo_move + * --> amdgpu_copy_buffer(). This sets up job in GPU scheduler. + * + * GPU Scheduler (amd_sched_main) - sets up a cb (fence_add_callback) to + * nofity when the BO is free to move. fence_add_callback --> enable_signaling + * --> amdgpu_amdkfd_fence.enable_signaling + * + * amdgpu_amdkfd_fence.enable_signaling - Start a work item that will quiesce + * user queues and signal fence. The work item will also start another delayed + * work item to restore BOs + */ + +struct amdgpu_amdkfd_fence *amdgpu_amdkfd_fence_create(u64 context, + struct mm_struct *mm) +{ + struct amdgpu_amdkfd_fence *fence; + + fence = kzalloc(sizeof(*fence), GFP_KERNEL); + if (fence == NULL) + return NULL; + + /* This reference gets released in amdkfd_fence_release */ + mmgrab(mm); + fence->mm = mm; + get_task_comm(fence->timeline_name, current); + spin_lock_init(&fence->lock); + + dma_fence_init(&fence->base, &amdkfd_fence_ops, &fence->lock, + context, atomic_inc_return(&fence_seq)); + + return fence; +} + +struct amdgpu_amdkfd_fence *to_amdgpu_amdkfd_fence(struct dma_fence *f) +{ + struct amdgpu_amdkfd_fence *fence; + + if (!f) + return NULL; + + fence = container_of(f, struct amdgpu_amdkfd_fence, base); + if (fence && f->ops == &amdkfd_fence_ops) + return fence; + + return NULL; +} + +static const char *amdkfd_fence_get_driver_name(struct dma_fence *f) +{ + return "amdgpu_amdkfd_fence"; +} + +static const char *amdkfd_fence_get_timeline_name(struct dma_fence *f) +{ + struct amdgpu_amdkfd_fence *fence = to_amdgpu_amdkfd_fence(f); + + return fence->timeline_name; +} + +/** + * amdkfd_fence_enable_signaling - This gets called when TTM wants to evict + * a KFD BO and schedules a job to move the BO. + * If fence is already signaled return true. + * If fence is not signaled schedule a evict KFD process work item. + */ +static bool amdkfd_fence_enable_signaling(struct dma_fence *f) +{ + struct amdgpu_amdkfd_fence *fence = to_amdgpu_amdkfd_fence(f); + + if (!fence) + return false; + + if (dma_fence_is_signaled(f)) + return true; + + if (!kgd2kfd->schedule_evict_and_restore_process(fence->mm, f)) + return true; + + return false; +} + +/** + * amdkfd_fence_release - callback that fence can be freed + * + * @fence: fence + * + * This function is called when the reference count becomes zero. + * Drops the mm_struct reference and RCU schedules freeing up the fence. + */ +static void amdkfd_fence_release(struct dma_fence *f) +{ + struct amdgpu_amdkfd_fence *fence = to_amdgpu_amdkfd_fence(f); + + /* Unconditionally signal the fence. The process is getting + * terminated. + */ + if (WARN_ON(!fence)) + return; /* Not an amdgpu_amdkfd_fence */ + + mmdrop(fence->mm); + kfree_rcu(f, rcu); +} + +/** + * amdkfd_fence_check_mm - Check if @mm is same as that of the fence @f + * if same return TRUE else return FALSE. + * + * @f: [IN] fence + * @mm: [IN] mm that needs to be verified + */ +bool amdkfd_fence_check_mm(struct dma_fence *f, struct mm_struct *mm) +{ + struct amdgpu_amdkfd_fence *fence = to_amdgpu_amdkfd_fence(f); + + if (!fence) + return false; + else if (fence->mm == mm) + return true; + + return false; +} + +static const struct dma_fence_ops amdkfd_fence_ops = { + .get_driver_name = amdkfd_fence_get_driver_name, + .get_timeline_name = amdkfd_fence_get_timeline_name, + .enable_signaling = amdkfd_fence_enable_signaling, + .signaled = NULL, + .wait = dma_fence_default_wait, + .release = amdkfd_fence_release, +};
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c index 0a484b3..e1f21c5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
@@ -139,11 +139,14 @@ static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd, static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd, uint8_t vmid); static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd, uint8_t vmid); -static void write_vmid_invalidate_request(struct kgd_dev *kgd, uint8_t vmid); static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type); static void set_scratch_backing_va(struct kgd_dev *kgd, uint64_t va, uint32_t vmid); +static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid, + uint32_t page_table_base); +static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid); +static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid); /* Because of REG_GET_FIELD() being used, we put this function in the * asic specific file. @@ -196,12 +199,26 @@ static const struct kfd2kgd_calls kfd2kgd = { .address_watch_get_offset = kgd_address_watch_get_offset, .get_atc_vmid_pasid_mapping_pasid = get_atc_vmid_pasid_mapping_pasid, .get_atc_vmid_pasid_mapping_valid = get_atc_vmid_pasid_mapping_valid, - .write_vmid_invalidate_request = write_vmid_invalidate_request, .get_fw_version = get_fw_version, .set_scratch_backing_va = set_scratch_backing_va, .get_tile_config = get_tile_config, .get_cu_info = get_cu_info, - .get_vram_usage = amdgpu_amdkfd_get_vram_usage + .get_vram_usage = amdgpu_amdkfd_get_vram_usage, + .create_process_vm = amdgpu_amdkfd_gpuvm_create_process_vm, + .acquire_process_vm = amdgpu_amdkfd_gpuvm_acquire_process_vm, + .destroy_process_vm = amdgpu_amdkfd_gpuvm_destroy_process_vm, + .get_process_page_dir = amdgpu_amdkfd_gpuvm_get_process_page_dir, + .set_vm_context_page_table_base = set_vm_context_page_table_base, + .alloc_memory_of_gpu = amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu, + .free_memory_of_gpu = amdgpu_amdkfd_gpuvm_free_memory_of_gpu, + .map_memory_to_gpu = amdgpu_amdkfd_gpuvm_map_memory_to_gpu, + .unmap_memory_to_gpu = amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu, + .sync_memory = amdgpu_amdkfd_gpuvm_sync_memory, + .map_gtt_bo_to_kernel = amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel, + .restore_process_bos = amdgpu_amdkfd_gpuvm_restore_process_bos, + .invalidate_tlbs = invalidate_tlbs, + .invalidate_tlbs_vmid = invalidate_tlbs_vmid, + .submit_ib = amdgpu_amdkfd_submit_ib, }; struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void) @@ -787,14 +804,7 @@ static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd, struct amdgpu_device *adev = (struct amdgpu_device *) kgd; reg = RREG32(mmATC_VMID0_PASID_MAPPING + vmid); - return reg & ATC_VMID0_PASID_MAPPING__VALID_MASK; -} - -static void write_vmid_invalidate_request(struct kgd_dev *kgd, uint8_t vmid) -{ - struct amdgpu_device *adev = (struct amdgpu_device *) kgd; - - WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); + return reg & ATC_VMID0_PASID_MAPPING__PASID_MASK; } static void set_scratch_backing_va(struct kgd_dev *kgd, @@ -812,8 +822,6 @@ static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type) struct amdgpu_device *adev = (struct amdgpu_device *) kgd; const union amdgpu_firmware_header *hdr; - BUG_ON(kgd == NULL); - switch (type) { case KGD_ENGINE_PFP: hdr = (const union amdgpu_firmware_header *) @@ -866,3 +874,50 @@ static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type) return hdr->common.ucode_version; } +static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid, + uint32_t page_table_base) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + + if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) { + pr_err("trying to set page table base for wrong VMID\n"); + return; + } + WREG32(mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8, page_table_base); +} + +static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid) +{ + struct amdgpu_device *adev = (struct amdgpu_device *) kgd; + int vmid; + unsigned int tmp; + + for (vmid = 0; vmid < 16; vmid++) { + if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) + continue; + + tmp = RREG32(mmATC_VMID0_PASID_MAPPING + vmid); + if ((tmp & ATC_VMID0_PASID_MAPPING__VALID_MASK) && + (tmp & ATC_VMID0_PASID_MAPPING__PASID_MASK) == pasid) { + WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); + RREG32(mmVM_INVALIDATE_RESPONSE); + break; + } + } + + return 0; +} + +static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid) +{ + struct amdgpu_device *adev = (struct amdgpu_device *) kgd; + + if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) { + pr_err("non kfd vmid\n"); + return 0; + } + + WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); + RREG32(mmVM_INVALIDATE_RESPONSE); + return 0; +}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c index b127259..89264c9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
@@ -81,7 +81,6 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd, uint32_t queue_id); static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, unsigned int utimeout); -static void write_vmid_invalidate_request(struct kgd_dev *kgd, uint8_t vmid); static int kgd_address_watch_disable(struct kgd_dev *kgd); static int kgd_address_watch_execute(struct kgd_dev *kgd, unsigned int watch_point_id, @@ -99,10 +98,13 @@ static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd, uint8_t vmid); static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd, uint8_t vmid); -static void write_vmid_invalidate_request(struct kgd_dev *kgd, uint8_t vmid); static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type); static void set_scratch_backing_va(struct kgd_dev *kgd, uint64_t va, uint32_t vmid); +static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid, + uint32_t page_table_base); +static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid); +static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid); /* Because of REG_GET_FIELD() being used, we put this function in the * asic specific file. @@ -157,12 +159,26 @@ static const struct kfd2kgd_calls kfd2kgd = { get_atc_vmid_pasid_mapping_pasid, .get_atc_vmid_pasid_mapping_valid = get_atc_vmid_pasid_mapping_valid, - .write_vmid_invalidate_request = write_vmid_invalidate_request, .get_fw_version = get_fw_version, .set_scratch_backing_va = set_scratch_backing_va, .get_tile_config = get_tile_config, .get_cu_info = get_cu_info, - .get_vram_usage = amdgpu_amdkfd_get_vram_usage + .get_vram_usage = amdgpu_amdkfd_get_vram_usage, + .create_process_vm = amdgpu_amdkfd_gpuvm_create_process_vm, + .acquire_process_vm = amdgpu_amdkfd_gpuvm_acquire_process_vm, + .destroy_process_vm = amdgpu_amdkfd_gpuvm_destroy_process_vm, + .get_process_page_dir = amdgpu_amdkfd_gpuvm_get_process_page_dir, + .set_vm_context_page_table_base = set_vm_context_page_table_base, + .alloc_memory_of_gpu = amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu, + .free_memory_of_gpu = amdgpu_amdkfd_gpuvm_free_memory_of_gpu, + .map_memory_to_gpu = amdgpu_amdkfd_gpuvm_map_memory_to_gpu, + .unmap_memory_to_gpu = amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu, + .sync_memory = amdgpu_amdkfd_gpuvm_sync_memory, + .map_gtt_bo_to_kernel = amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel, + .restore_process_bos = amdgpu_amdkfd_gpuvm_restore_process_bos, + .invalidate_tlbs = invalidate_tlbs, + .invalidate_tlbs_vmid = invalidate_tlbs_vmid, + .submit_ib = amdgpu_amdkfd_submit_ib, }; struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void) @@ -704,14 +720,7 @@ static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd, struct amdgpu_device *adev = (struct amdgpu_device *) kgd; reg = RREG32(mmATC_VMID0_PASID_MAPPING + vmid); - return reg & ATC_VMID0_PASID_MAPPING__VALID_MASK; -} - -static void write_vmid_invalidate_request(struct kgd_dev *kgd, uint8_t vmid) -{ - struct amdgpu_device *adev = (struct amdgpu_device *) kgd; - - WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); + return reg & ATC_VMID0_PASID_MAPPING__PASID_MASK; } static int kgd_address_watch_disable(struct kgd_dev *kgd) @@ -775,8 +784,6 @@ static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type) struct amdgpu_device *adev = (struct amdgpu_device *) kgd; const union amdgpu_firmware_header *hdr; - BUG_ON(kgd == NULL); - switch (type) { case KGD_ENGINE_PFP: hdr = (const union amdgpu_firmware_header *) @@ -828,3 +835,51 @@ static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type) /* Only 12 bit in use*/ return hdr->common.ucode_version; } + +static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid, + uint32_t page_table_base) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + + if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) { + pr_err("trying to set page table base for wrong VMID\n"); + return; + } + WREG32(mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8, page_table_base); +} + +static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid) +{ + struct amdgpu_device *adev = (struct amdgpu_device *) kgd; + int vmid; + unsigned int tmp; + + for (vmid = 0; vmid < 16; vmid++) { + if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) + continue; + + tmp = RREG32(mmATC_VMID0_PASID_MAPPING + vmid); + if ((tmp & ATC_VMID0_PASID_MAPPING__VALID_MASK) && + (tmp & ATC_VMID0_PASID_MAPPING__PASID_MASK) == pasid) { + WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); + RREG32(mmVM_INVALIDATE_RESPONSE); + break; + } + } + + return 0; +} + +static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid) +{ + struct amdgpu_device *adev = (struct amdgpu_device *) kgd; + + if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) { + pr_err("non kfd vmid %d\n", vmid); + return -EINVAL; + } + + WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); + RREG32(mmVM_INVALIDATE_RESPONSE); + return 0; +}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c new file mode 100644 index 0000000..1d6e147 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -0,0 +1,1577 @@ +/* + * Copyright 2014-2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#define pr_fmt(fmt) "kfd2kgd: " fmt + +#include <linux/list.h> +#include <drm/drmP.h> +#include "amdgpu_object.h" +#include "amdgpu_vm.h" +#include "amdgpu_amdkfd.h" + +/* Special VM and GART address alignment needed for VI pre-Fiji due to + * a HW bug. + */ +#define VI_BO_SIZE_ALIGN (0x8000) + +/* Impose limit on how much memory KFD can use */ +static struct { + uint64_t max_system_mem_limit; + int64_t system_mem_used; + spinlock_t mem_limit_lock; +} kfd_mem_limit; + +/* Struct used for amdgpu_amdkfd_bo_validate */ +struct amdgpu_vm_parser { + uint32_t domain; + bool wait; +}; + +static const char * const domain_bit_to_string[] = { + "CPU", + "GTT", + "VRAM", + "GDS", + "GWS", + "OA" +}; + +#define domain_string(domain) domain_bit_to_string[ffs(domain)-1] + + + +static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd) +{ + return (struct amdgpu_device *)kgd; +} + +static bool check_if_add_bo_to_vm(struct amdgpu_vm *avm, + struct kgd_mem *mem) +{ + struct kfd_bo_va_list *entry; + + list_for_each_entry(entry, &mem->bo_va_list, bo_list) + if (entry->bo_va->base.vm == avm) + return false; + + return true; +} + +/* Set memory usage limits. Current, limits are + * System (kernel) memory - 3/8th System RAM + */ +void amdgpu_amdkfd_gpuvm_init_mem_limits(void) +{ + struct sysinfo si; + uint64_t mem; + + si_meminfo(&si); + mem = si.totalram - si.totalhigh; + mem *= si.mem_unit; + + spin_lock_init(&kfd_mem_limit.mem_limit_lock); + kfd_mem_limit.max_system_mem_limit = (mem >> 1) - (mem >> 3); + pr_debug("Kernel memory limit %lluM\n", + (kfd_mem_limit.max_system_mem_limit >> 20)); +} + +static int amdgpu_amdkfd_reserve_system_mem_limit(struct amdgpu_device *adev, + uint64_t size, u32 domain) +{ + size_t acc_size; + int ret = 0; + + acc_size = ttm_bo_dma_acc_size(&adev->mman.bdev, size, + sizeof(struct amdgpu_bo)); + + spin_lock(&kfd_mem_limit.mem_limit_lock); + if (domain == AMDGPU_GEM_DOMAIN_GTT) { + if (kfd_mem_limit.system_mem_used + (acc_size + size) > + kfd_mem_limit.max_system_mem_limit) { + ret = -ENOMEM; + goto err_no_mem; + } + kfd_mem_limit.system_mem_used += (acc_size + size); + } +err_no_mem: + spin_unlock(&kfd_mem_limit.mem_limit_lock); + return ret; +} + +static void unreserve_system_mem_limit(struct amdgpu_device *adev, + uint64_t size, u32 domain) +{ + size_t acc_size; + + acc_size = ttm_bo_dma_acc_size(&adev->mman.bdev, size, + sizeof(struct amdgpu_bo)); + + spin_lock(&kfd_mem_limit.mem_limit_lock); + if (domain == AMDGPU_GEM_DOMAIN_GTT) + kfd_mem_limit.system_mem_used -= (acc_size + size); + WARN_ONCE(kfd_mem_limit.system_mem_used < 0, + "kfd system memory accounting unbalanced"); + + spin_unlock(&kfd_mem_limit.mem_limit_lock); +} + +void amdgpu_amdkfd_unreserve_system_memory_limit(struct amdgpu_bo *bo) +{ + spin_lock(&kfd_mem_limit.mem_limit_lock); + + if (bo->preferred_domains == AMDGPU_GEM_DOMAIN_GTT) { + kfd_mem_limit.system_mem_used -= + (bo->tbo.acc_size + amdgpu_bo_size(bo)); + } + WARN_ONCE(kfd_mem_limit.system_mem_used < 0, + "kfd system memory accounting unbalanced"); + + spin_unlock(&kfd_mem_limit.mem_limit_lock); +} + + +/* amdgpu_amdkfd_remove_eviction_fence - Removes eviction fence(s) from BO's + * reservation object. + * + * @bo: [IN] Remove eviction fence(s) from this BO + * @ef: [IN] If ef is specified, then this eviction fence is removed if it + * is present in the shared list. + * @ef_list: [OUT] Returns list of eviction fences. These fences are removed + * from BO's reservation object shared list. + * @ef_count: [OUT] Number of fences in ef_list. + * + * NOTE: If called with ef_list, then amdgpu_amdkfd_add_eviction_fence must be + * called to restore the eviction fences and to avoid memory leak. This is + * useful for shared BOs. + * NOTE: Must be called with BO reserved i.e. bo->tbo.resv->lock held. + */ +static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo, + struct amdgpu_amdkfd_fence *ef, + struct amdgpu_amdkfd_fence ***ef_list, + unsigned int *ef_count) +{ + struct reservation_object_list *fobj; + struct reservation_object *resv; + unsigned int i = 0, j = 0, k = 0, shared_count; + unsigned int count = 0; + struct amdgpu_amdkfd_fence **fence_list; + + if (!ef && !ef_list) + return -EINVAL; + + if (ef_list) { + *ef_list = NULL; + *ef_count = 0; + } + + resv = bo->tbo.resv; + fobj = reservation_object_get_list(resv); + + if (!fobj) + return 0; + + preempt_disable(); + write_seqcount_begin(&resv->seq); + + /* Go through all the shared fences in the resevation object. If + * ef is specified and it exists in the list, remove it and reduce the + * count. If ef is not specified, then get the count of eviction fences + * present. + */ + shared_count = fobj->shared_count; + for (i = 0; i < shared_count; ++i) { + struct dma_fence *f; + + f = rcu_dereference_protected(fobj->shared[i], + reservation_object_held(resv)); + + if (ef) { + if (f->context == ef->base.context) { + dma_fence_put(f); + fobj->shared_count--; + } else { + RCU_INIT_POINTER(fobj->shared[j++], f); + } + } else if (to_amdgpu_amdkfd_fence(f)) + count++; + } + write_seqcount_end(&resv->seq); + preempt_enable(); + + if (ef || !count) + return 0; + + /* Alloc memory for count number of eviction fence pointers. Fill the + * ef_list array and ef_count + */ + fence_list = kcalloc(count, sizeof(struct amdgpu_amdkfd_fence *), + GFP_KERNEL); + if (!fence_list) + return -ENOMEM; + + preempt_disable(); + write_seqcount_begin(&resv->seq); + + j = 0; + for (i = 0; i < shared_count; ++i) { + struct dma_fence *f; + struct amdgpu_amdkfd_fence *efence; + + f = rcu_dereference_protected(fobj->shared[i], + reservation_object_held(resv)); + + efence = to_amdgpu_amdkfd_fence(f); + if (efence) { + fence_list[k++] = efence; + fobj->shared_count--; + } else { + RCU_INIT_POINTER(fobj->shared[j++], f); + } + } + + write_seqcount_end(&resv->seq); + preempt_enable(); + + *ef_list = fence_list; + *ef_count = k; + + return 0; +} + +/* amdgpu_amdkfd_add_eviction_fence - Adds eviction fence(s) back into BO's + * reservation object. + * + * @bo: [IN] Add eviction fences to this BO + * @ef_list: [IN] List of eviction fences to be added + * @ef_count: [IN] Number of fences in ef_list. + * + * NOTE: Must call amdgpu_amdkfd_remove_eviction_fence before calling this + * function. + */ +static void amdgpu_amdkfd_add_eviction_fence(struct amdgpu_bo *bo, + struct amdgpu_amdkfd_fence **ef_list, + unsigned int ef_count) +{ + int i; + + if (!ef_list || !ef_count) + return; + + for (i = 0; i < ef_count; i++) { + amdgpu_bo_fence(bo, &ef_list[i]->base, true); + /* Re-adding the fence takes an additional reference. Drop that + * reference. + */ + dma_fence_put(&ef_list[i]->base); + } + + kfree(ef_list); +} + +static int amdgpu_amdkfd_bo_validate(struct amdgpu_bo *bo, uint32_t domain, + bool wait) +{ + struct ttm_operation_ctx ctx = { false, false }; + int ret; + + if (WARN(amdgpu_ttm_tt_get_usermm(bo->tbo.ttm), + "Called with userptr BO")) + return -EINVAL; + + amdgpu_ttm_placement_from_domain(bo, domain); + + ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); + if (ret) + goto validate_fail; + if (wait) { + struct amdgpu_amdkfd_fence **ef_list; + unsigned int ef_count; + + ret = amdgpu_amdkfd_remove_eviction_fence(bo, NULL, &ef_list, + &ef_count); + if (ret) + goto validate_fail; + + ttm_bo_wait(&bo->tbo, false, false); + amdgpu_amdkfd_add_eviction_fence(bo, ef_list, ef_count); + } + +validate_fail: + return ret; +} + +static int amdgpu_amdkfd_validate(void *param, struct amdgpu_bo *bo) +{ + struct amdgpu_vm_parser *p = param; + + return amdgpu_amdkfd_bo_validate(bo, p->domain, p->wait); +} + +/* vm_validate_pt_pd_bos - Validate page table and directory BOs + * + * Page directories are not updated here because huge page handling + * during page table updates can invalidate page directory entries + * again. Page directories are only updated after updating page + * tables. + */ +static int vm_validate_pt_pd_bos(struct amdgpu_vm *vm) +{ + struct amdgpu_bo *pd = vm->root.base.bo; + struct amdgpu_device *adev = amdgpu_ttm_adev(pd->tbo.bdev); + struct amdgpu_vm_parser param; + uint64_t addr, flags = AMDGPU_PTE_VALID; + int ret; + + param.domain = AMDGPU_GEM_DOMAIN_VRAM; + param.wait = false; + + ret = amdgpu_vm_validate_pt_bos(adev, vm, amdgpu_amdkfd_validate, + ¶m); + if (ret) { + pr_err("amdgpu: failed to validate PT BOs\n"); + return ret; + } + + ret = amdgpu_amdkfd_validate(¶m, pd); + if (ret) { + pr_err("amdgpu: failed to validate PD\n"); + return ret; + } + + addr = amdgpu_bo_gpu_offset(vm->root.base.bo); + amdgpu_gmc_get_vm_pde(adev, -1, &addr, &flags); + vm->pd_phys_addr = addr; + + if (vm->use_cpu_for_update) { + ret = amdgpu_bo_kmap(pd, NULL); + if (ret) { + pr_err("amdgpu: failed to kmap PD, ret=%d\n", ret); + return ret; + } + } + + return 0; +} + +static int sync_vm_fence(struct amdgpu_device *adev, struct amdgpu_sync *sync, + struct dma_fence *f) +{ + int ret = amdgpu_sync_fence(adev, sync, f, false); + + /* Sync objects can't handle multiple GPUs (contexts) updating + * sync->last_vm_update. Fortunately we don't need it for + * KFD's purposes, so we can just drop that fence. + */ + if (sync->last_vm_update) { + dma_fence_put(sync->last_vm_update); + sync->last_vm_update = NULL; + } + + return ret; +} + +static int vm_update_pds(struct amdgpu_vm *vm, struct amdgpu_sync *sync) +{ + struct amdgpu_bo *pd = vm->root.base.bo; + struct amdgpu_device *adev = amdgpu_ttm_adev(pd->tbo.bdev); + int ret; + + ret = amdgpu_vm_update_directories(adev, vm); + if (ret) + return ret; + + return sync_vm_fence(adev, sync, vm->last_update); +} + +/* add_bo_to_vm - Add a BO to a VM + * + * Everything that needs to bo done only once when a BO is first added + * to a VM. It can later be mapped and unmapped many times without + * repeating these steps. + * + * 1. Allocate and initialize BO VA entry data structure + * 2. Add BO to the VM + * 3. Determine ASIC-specific PTE flags + * 4. Alloc page tables and directories if needed + * 4a. Validate new page tables and directories + */ +static int add_bo_to_vm(struct amdgpu_device *adev, struct kgd_mem *mem, + struct amdgpu_vm *vm, bool is_aql, + struct kfd_bo_va_list **p_bo_va_entry) +{ + int ret; + struct kfd_bo_va_list *bo_va_entry; + struct amdgpu_bo *pd = vm->root.base.bo; + struct amdgpu_bo *bo = mem->bo; + uint64_t va = mem->va; + struct list_head *list_bo_va = &mem->bo_va_list; + unsigned long bo_size = bo->tbo.mem.size; + + if (!va) { + pr_err("Invalid VA when adding BO to VM\n"); + return -EINVAL; + } + + if (is_aql) + va += bo_size; + + bo_va_entry = kzalloc(sizeof(*bo_va_entry), GFP_KERNEL); + if (!bo_va_entry) + return -ENOMEM; + + pr_debug("\t add VA 0x%llx - 0x%llx to vm %p\n", va, + va + bo_size, vm); + + /* Add BO to VM internal data structures*/ + bo_va_entry->bo_va = amdgpu_vm_bo_add(adev, vm, bo); + if (!bo_va_entry->bo_va) { + ret = -EINVAL; + pr_err("Failed to add BO object to VM. ret == %d\n", + ret); + goto err_vmadd; + } + + bo_va_entry->va = va; + bo_va_entry->pte_flags = amdgpu_gmc_get_pte_flags(adev, + mem->mapping_flags); + bo_va_entry->kgd_dev = (void *)adev; + list_add(&bo_va_entry->bo_list, list_bo_va); + + if (p_bo_va_entry) + *p_bo_va_entry = bo_va_entry; + + /* Allocate new page tables if needed and validate + * them. Clearing of new page tables and validate need to wait + * on move fences. We don't want that to trigger the eviction + * fence, so remove it temporarily. + */ + amdgpu_amdkfd_remove_eviction_fence(pd, + vm->process_info->eviction_fence, + NULL, NULL); + + ret = amdgpu_vm_alloc_pts(adev, vm, va, amdgpu_bo_size(bo)); + if (ret) { + pr_err("Failed to allocate pts, err=%d\n", ret); + goto err_alloc_pts; + } + + ret = vm_validate_pt_pd_bos(vm); + if (ret) { + pr_err("validate_pt_pd_bos() failed\n"); + goto err_alloc_pts; + } + + /* Add the eviction fence back */ + amdgpu_bo_fence(pd, &vm->process_info->eviction_fence->base, true); + + return 0; + +err_alloc_pts: + amdgpu_bo_fence(pd, &vm->process_info->eviction_fence->base, true); + amdgpu_vm_bo_rmv(adev, bo_va_entry->bo_va); + list_del(&bo_va_entry->bo_list); +err_vmadd: + kfree(bo_va_entry); + return ret; +} + +static void remove_bo_from_vm(struct amdgpu_device *adev, + struct kfd_bo_va_list *entry, unsigned long size) +{ + pr_debug("\t remove VA 0x%llx - 0x%llx in entry %p\n", + entry->va, + entry->va + size, entry); + amdgpu_vm_bo_rmv(adev, entry->bo_va); + list_del(&entry->bo_list); + kfree(entry); +} + +static void add_kgd_mem_to_kfd_bo_list(struct kgd_mem *mem, + struct amdkfd_process_info *process_info) +{ + struct ttm_validate_buffer *entry = &mem->validate_list; + struct amdgpu_bo *bo = mem->bo; + + INIT_LIST_HEAD(&entry->head); + entry->shared = true; + entry->bo = &bo->tbo; + mutex_lock(&process_info->lock); + list_add_tail(&entry->head, &process_info->kfd_bo_list); + mutex_unlock(&process_info->lock); +} + +/* Reserving a BO and its page table BOs must happen atomically to + * avoid deadlocks. Some operations update multiple VMs at once. Track + * all the reservation info in a context structure. Optionally a sync + * object can track VM updates. + */ +struct bo_vm_reservation_context { + struct amdgpu_bo_list_entry kfd_bo; /* BO list entry for the KFD BO */ + unsigned int n_vms; /* Number of VMs reserved */ + struct amdgpu_bo_list_entry *vm_pd; /* Array of VM BO list entries */ + struct ww_acquire_ctx ticket; /* Reservation ticket */ + struct list_head list, duplicates; /* BO lists */ + struct amdgpu_sync *sync; /* Pointer to sync object */ + bool reserved; /* Whether BOs are reserved */ +}; + +enum bo_vm_match { + BO_VM_NOT_MAPPED = 0, /* Match VMs where a BO is not mapped */ + BO_VM_MAPPED, /* Match VMs where a BO is mapped */ + BO_VM_ALL, /* Match all VMs a BO was added to */ +}; + +/** + * reserve_bo_and_vm - reserve a BO and a VM unconditionally. + * @mem: KFD BO structure. + * @vm: the VM to reserve. + * @ctx: the struct that will be used in unreserve_bo_and_vms(). + */ +static int reserve_bo_and_vm(struct kgd_mem *mem, + struct amdgpu_vm *vm, + struct bo_vm_reservation_context *ctx) +{ + struct amdgpu_bo *bo = mem->bo; + int ret; + + WARN_ON(!vm); + + ctx->reserved = false; + ctx->n_vms = 1; + ctx->sync = &mem->sync; + + INIT_LIST_HEAD(&ctx->list); + INIT_LIST_HEAD(&ctx->duplicates); + + ctx->vm_pd = kcalloc(ctx->n_vms, sizeof(*ctx->vm_pd), GFP_KERNEL); + if (!ctx->vm_pd) + return -ENOMEM; + + ctx->kfd_bo.robj = bo; + ctx->kfd_bo.priority = 0; + ctx->kfd_bo.tv.bo = &bo->tbo; + ctx->kfd_bo.tv.shared = true; + ctx->kfd_bo.user_pages = NULL; + list_add(&ctx->kfd_bo.tv.head, &ctx->list); + + amdgpu_vm_get_pd_bo(vm, &ctx->list, &ctx->vm_pd[0]); + + ret = ttm_eu_reserve_buffers(&ctx->ticket, &ctx->list, + false, &ctx->duplicates); + if (!ret) + ctx->reserved = true; + else { + pr_err("Failed to reserve buffers in ttm\n"); + kfree(ctx->vm_pd); + ctx->vm_pd = NULL; + } + + return ret; +} + +/** + * reserve_bo_and_cond_vms - reserve a BO and some VMs conditionally + * @mem: KFD BO structure. + * @vm: the VM to reserve. If NULL, then all VMs associated with the BO + * is used. Otherwise, a single VM associated with the BO. + * @map_type: the mapping status that will be used to filter the VMs. + * @ctx: the struct that will be used in unreserve_bo_and_vms(). + * + * Returns 0 for success, negative for failure. + */ +static int reserve_bo_and_cond_vms(struct kgd_mem *mem, + struct amdgpu_vm *vm, enum bo_vm_match map_type, + struct bo_vm_reservation_context *ctx) +{ + struct amdgpu_bo *bo = mem->bo; + struct kfd_bo_va_list *entry; + unsigned int i; + int ret; + + ctx->reserved = false; + ctx->n_vms = 0; + ctx->vm_pd = NULL; + ctx->sync = &mem->sync; + + INIT_LIST_HEAD(&ctx->list); + INIT_LIST_HEAD(&ctx->duplicates); + + list_for_each_entry(entry, &mem->bo_va_list, bo_list) { + if ((vm && vm != entry->bo_va->base.vm) || + (entry->is_mapped != map_type + && map_type != BO_VM_ALL)) + continue; + + ctx->n_vms++; + } + + if (ctx->n_vms != 0) { + ctx->vm_pd = kcalloc(ctx->n_vms, sizeof(*ctx->vm_pd), + GFP_KERNEL); + if (!ctx->vm_pd) + return -ENOMEM; + } + + ctx->kfd_bo.robj = bo; + ctx->kfd_bo.priority = 0; + ctx->kfd_bo.tv.bo = &bo->tbo; + ctx->kfd_bo.tv.shared = true; + ctx->kfd_bo.user_pages = NULL; + list_add(&ctx->kfd_bo.tv.head, &ctx->list); + + i = 0; + list_for_each_entry(entry, &mem->bo_va_list, bo_list) { + if ((vm && vm != entry->bo_va->base.vm) || + (entry->is_mapped != map_type + && map_type != BO_VM_ALL)) + continue; + + amdgpu_vm_get_pd_bo(entry->bo_va->base.vm, &ctx->list, + &ctx->vm_pd[i]); + i++; + } + + ret = ttm_eu_reserve_buffers(&ctx->ticket, &ctx->list, + false, &ctx->duplicates); + if (!ret) + ctx->reserved = true; + else + pr_err("Failed to reserve buffers in ttm.\n"); + + if (ret) { + kfree(ctx->vm_pd); + ctx->vm_pd = NULL; + } + + return ret; +} + +/** + * unreserve_bo_and_vms - Unreserve BO and VMs from a reservation context + * @ctx: Reservation context to unreserve + * @wait: Optionally wait for a sync object representing pending VM updates + * @intr: Whether the wait is interruptible + * + * Also frees any resources allocated in + * reserve_bo_and_(cond_)vm(s). Returns the status from + * amdgpu_sync_wait. + */ +static int unreserve_bo_and_vms(struct bo_vm_reservation_context *ctx, + bool wait, bool intr) +{ + int ret = 0; + + if (wait) + ret = amdgpu_sync_wait(ctx->sync, intr); + + if (ctx->reserved) + ttm_eu_backoff_reservation(&ctx->ticket, &ctx->list); + kfree(ctx->vm_pd); + + ctx->sync = NULL; + + ctx->reserved = false; + ctx->vm_pd = NULL; + + return ret; +} + +static int unmap_bo_from_gpuvm(struct amdgpu_device *adev, + struct kfd_bo_va_list *entry, + struct amdgpu_sync *sync) +{ + struct amdgpu_bo_va *bo_va = entry->bo_va; + struct amdgpu_vm *vm = bo_va->base.vm; + struct amdgpu_bo *pd = vm->root.base.bo; + + /* Remove eviction fence from PD (and thereby from PTs too as + * they share the resv. object). Otherwise during PT update + * job (see amdgpu_vm_bo_update_mapping), eviction fence would + * get added to job->sync object and job execution would + * trigger the eviction fence. + */ + amdgpu_amdkfd_remove_eviction_fence(pd, + vm->process_info->eviction_fence, + NULL, NULL); + amdgpu_vm_bo_unmap(adev, bo_va, entry->va); + + amdgpu_vm_clear_freed(adev, vm, &bo_va->last_pt_update); + + /* Add the eviction fence back */ + amdgpu_bo_fence(pd, &vm->process_info->eviction_fence->base, true); + + sync_vm_fence(adev, sync, bo_va->last_pt_update); + + return 0; +} + +static int update_gpuvm_pte(struct amdgpu_device *adev, + struct kfd_bo_va_list *entry, + struct amdgpu_sync *sync) +{ + int ret; + struct amdgpu_vm *vm; + struct amdgpu_bo_va *bo_va; + struct amdgpu_bo *bo; + + bo_va = entry->bo_va; + vm = bo_va->base.vm; + bo = bo_va->base.bo; + + /* Update the page tables */ + ret = amdgpu_vm_bo_update(adev, bo_va, false); + if (ret) { + pr_err("amdgpu_vm_bo_update failed\n"); + return ret; + } + + return sync_vm_fence(adev, sync, bo_va->last_pt_update); +} + +static int map_bo_to_gpuvm(struct amdgpu_device *adev, + struct kfd_bo_va_list *entry, struct amdgpu_sync *sync) +{ + int ret; + + /* Set virtual address for the allocation */ + ret = amdgpu_vm_bo_map(adev, entry->bo_va, entry->va, 0, + amdgpu_bo_size(entry->bo_va->base.bo), + entry->pte_flags); + if (ret) { + pr_err("Failed to map VA 0x%llx in vm. ret %d\n", + entry->va, ret); + return ret; + } + + ret = update_gpuvm_pte(adev, entry, sync); + if (ret) { + pr_err("update_gpuvm_pte() failed\n"); + goto update_gpuvm_pte_failed; + } + + return 0; + +update_gpuvm_pte_failed: + unmap_bo_from_gpuvm(adev, entry, sync); + return ret; +} + +static int process_validate_vms(struct amdkfd_process_info *process_info) +{ + struct amdgpu_vm *peer_vm; + int ret; + + list_for_each_entry(peer_vm, &process_info->vm_list_head, + vm_list_node) { + ret = vm_validate_pt_pd_bos(peer_vm); + if (ret) + return ret; + } + + return 0; +} + +static int process_update_pds(struct amdkfd_process_info *process_info, + struct amdgpu_sync *sync) +{ + struct amdgpu_vm *peer_vm; + int ret; + + list_for_each_entry(peer_vm, &process_info->vm_list_head, + vm_list_node) { + ret = vm_update_pds(peer_vm, sync); + if (ret) + return ret; + } + + return 0; +} + +static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info, + struct dma_fence **ef) +{ + struct amdkfd_process_info *info = NULL; + int ret; + + if (!*process_info) { + info = kzalloc(sizeof(*info), GFP_KERNEL); + if (!info) + return -ENOMEM; + + mutex_init(&info->lock); + INIT_LIST_HEAD(&info->vm_list_head); + INIT_LIST_HEAD(&info->kfd_bo_list); + + info->eviction_fence = + amdgpu_amdkfd_fence_create(dma_fence_context_alloc(1), + current->mm); + if (!info->eviction_fence) { + pr_err("Failed to create eviction fence\n"); + ret = -ENOMEM; + goto create_evict_fence_fail; + } + + *process_info = info; + *ef = dma_fence_get(&info->eviction_fence->base); + } + + vm->process_info = *process_info; + + /* Validate page directory and attach eviction fence */ + ret = amdgpu_bo_reserve(vm->root.base.bo, true); + if (ret) + goto reserve_pd_fail; + ret = vm_validate_pt_pd_bos(vm); + if (ret) { + pr_err("validate_pt_pd_bos() failed\n"); + goto validate_pd_fail; + } + ret = ttm_bo_wait(&vm->root.base.bo->tbo, false, false); + if (ret) + goto wait_pd_fail; + amdgpu_bo_fence(vm->root.base.bo, + &vm->process_info->eviction_fence->base, true); + amdgpu_bo_unreserve(vm->root.base.bo); + + /* Update process info */ + mutex_lock(&vm->process_info->lock); + list_add_tail(&vm->vm_list_node, + &(vm->process_info->vm_list_head)); + vm->process_info->n_vms++; + mutex_unlock(&vm->process_info->lock); + + return 0; + +wait_pd_fail: +validate_pd_fail: + amdgpu_bo_unreserve(vm->root.base.bo); +reserve_pd_fail: + vm->process_info = NULL; + if (info) { + /* Two fence references: one in info and one in *ef */ + dma_fence_put(&info->eviction_fence->base); + dma_fence_put(*ef); + *ef = NULL; + *process_info = NULL; +create_evict_fence_fail: + mutex_destroy(&info->lock); + kfree(info); + } + return ret; +} + +int amdgpu_amdkfd_gpuvm_create_process_vm(struct kgd_dev *kgd, void **vm, + void **process_info, + struct dma_fence **ef) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + struct amdgpu_vm *new_vm; + int ret; + + new_vm = kzalloc(sizeof(*new_vm), GFP_KERNEL); + if (!new_vm) + return -ENOMEM; + + /* Initialize AMDGPU part of the VM */ + ret = amdgpu_vm_init(adev, new_vm, AMDGPU_VM_CONTEXT_COMPUTE, 0); + if (ret) { + pr_err("Failed init vm ret %d\n", ret); + goto amdgpu_vm_init_fail; + } + + /* Initialize KFD part of the VM and process info */ + ret = init_kfd_vm(new_vm, process_info, ef); + if (ret) + goto init_kfd_vm_fail; + + *vm = (void *) new_vm; + + return 0; + +init_kfd_vm_fail: + amdgpu_vm_fini(adev, new_vm); +amdgpu_vm_init_fail: + kfree(new_vm); + return ret; +} + +int amdgpu_amdkfd_gpuvm_acquire_process_vm(struct kgd_dev *kgd, + struct file *filp, + void **vm, void **process_info, + struct dma_fence **ef) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + struct drm_file *drm_priv = filp->private_data; + struct amdgpu_fpriv *drv_priv = drm_priv->driver_priv; + struct amdgpu_vm *avm = &drv_priv->vm; + int ret; + + /* Already a compute VM? */ + if (avm->process_info) + return -EINVAL; + + /* Convert VM into a compute VM */ + ret = amdgpu_vm_make_compute(adev, avm); + if (ret) + return ret; + + /* Initialize KFD part of the VM and process info */ + ret = init_kfd_vm(avm, process_info, ef); + if (ret) + return ret; + + *vm = (void *)avm; + + return 0; +} + +void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev, + struct amdgpu_vm *vm) +{ + struct amdkfd_process_info *process_info = vm->process_info; + struct amdgpu_bo *pd = vm->root.base.bo; + + if (!process_info) + return; + + /* Release eviction fence from PD */ + amdgpu_bo_reserve(pd, false); + amdgpu_bo_fence(pd, NULL, false); + amdgpu_bo_unreserve(pd); + + /* Update process info */ + mutex_lock(&process_info->lock); + process_info->n_vms--; + list_del(&vm->vm_list_node); + mutex_unlock(&process_info->lock); + + /* Release per-process resources when last compute VM is destroyed */ + if (!process_info->n_vms) { + WARN_ON(!list_empty(&process_info->kfd_bo_list)); + + dma_fence_put(&process_info->eviction_fence->base); + mutex_destroy(&process_info->lock); + kfree(process_info); + } +} + +void amdgpu_amdkfd_gpuvm_destroy_process_vm(struct kgd_dev *kgd, void *vm) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + struct amdgpu_vm *avm = (struct amdgpu_vm *)vm; + + if (WARN_ON(!kgd || !vm)) + return; + + pr_debug("Destroying process vm %p\n", vm); + + /* Release the VM context */ + amdgpu_vm_fini(adev, avm); + kfree(vm); +} + +uint32_t amdgpu_amdkfd_gpuvm_get_process_page_dir(void *vm) +{ + struct amdgpu_vm *avm = (struct amdgpu_vm *)vm; + + return avm->pd_phys_addr >> AMDGPU_GPU_PAGE_SHIFT; +} + +int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( + struct kgd_dev *kgd, uint64_t va, uint64_t size, + void *vm, struct kgd_mem **mem, + uint64_t *offset, uint32_t flags) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + struct amdgpu_vm *avm = (struct amdgpu_vm *)vm; + struct amdgpu_bo *bo; + int byte_align; + u32 alloc_domain; + u64 alloc_flags; + uint32_t mapping_flags; + int ret; + + /* + * Check on which domain to allocate BO + */ + if (flags & ALLOC_MEM_FLAGS_VRAM) { + alloc_domain = AMDGPU_GEM_DOMAIN_VRAM; + alloc_flags = AMDGPU_GEM_CREATE_VRAM_CLEARED; + alloc_flags |= (flags & ALLOC_MEM_FLAGS_PUBLIC) ? + AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED : + AMDGPU_GEM_CREATE_NO_CPU_ACCESS; + } else if (flags & ALLOC_MEM_FLAGS_GTT) { + alloc_domain = AMDGPU_GEM_DOMAIN_GTT; + alloc_flags = 0; + } else { + return -EINVAL; + } + + *mem = kzalloc(sizeof(struct kgd_mem), GFP_KERNEL); + if (!*mem) + return -ENOMEM; + INIT_LIST_HEAD(&(*mem)->bo_va_list); + mutex_init(&(*mem)->lock); + (*mem)->aql_queue = !!(flags & ALLOC_MEM_FLAGS_AQL_QUEUE_MEM); + + /* Workaround for AQL queue wraparound bug. Map the same + * memory twice. That means we only actually allocate half + * the memory. + */ + if ((*mem)->aql_queue) + size = size >> 1; + + /* Workaround for TLB bug on older VI chips */ + byte_align = (adev->family == AMDGPU_FAMILY_VI && + adev->asic_type != CHIP_FIJI && + adev->asic_type != CHIP_POLARIS10 && + adev->asic_type != CHIP_POLARIS11) ? + VI_BO_SIZE_ALIGN : 1; + + mapping_flags = AMDGPU_VM_PAGE_READABLE; + if (flags & ALLOC_MEM_FLAGS_WRITABLE) + mapping_flags |= AMDGPU_VM_PAGE_WRITEABLE; + if (flags & ALLOC_MEM_FLAGS_EXECUTABLE) + mapping_flags |= AMDGPU_VM_PAGE_EXECUTABLE; + if (flags & ALLOC_MEM_FLAGS_COHERENT) + mapping_flags |= AMDGPU_VM_MTYPE_UC; + else + mapping_flags |= AMDGPU_VM_MTYPE_NC; + (*mem)->mapping_flags = mapping_flags; + + amdgpu_sync_create(&(*mem)->sync); + + ret = amdgpu_amdkfd_reserve_system_mem_limit(adev, size, alloc_domain); + if (ret) { + pr_debug("Insufficient system memory\n"); + goto err_reserve_system_mem; + } + + pr_debug("\tcreate BO VA 0x%llx size 0x%llx domain %s\n", + va, size, domain_string(alloc_domain)); + + ret = amdgpu_bo_create(adev, size, byte_align, + alloc_domain, alloc_flags, ttm_bo_type_device, NULL, &bo); + if (ret) { + pr_debug("Failed to create BO on domain %s. ret %d\n", + domain_string(alloc_domain), ret); + goto err_bo_create; + } + bo->kfd_bo = *mem; + (*mem)->bo = bo; + + (*mem)->va = va; + (*mem)->domain = alloc_domain; + (*mem)->mapped_to_gpu_memory = 0; + (*mem)->process_info = avm->process_info; + add_kgd_mem_to_kfd_bo_list(*mem, avm->process_info); + + if (offset) + *offset = amdgpu_bo_mmap_offset(bo); + + return 0; + +err_bo_create: + unreserve_system_mem_limit(adev, size, alloc_domain); +err_reserve_system_mem: + mutex_destroy(&(*mem)->lock); + kfree(*mem); + return ret; +} + +int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( + struct kgd_dev *kgd, struct kgd_mem *mem) +{ + struct amdkfd_process_info *process_info = mem->process_info; + unsigned long bo_size = mem->bo->tbo.mem.size; + struct kfd_bo_va_list *entry, *tmp; + struct bo_vm_reservation_context ctx; + struct ttm_validate_buffer *bo_list_entry; + int ret; + + mutex_lock(&mem->lock); + + if (mem->mapped_to_gpu_memory > 0) { + pr_debug("BO VA 0x%llx size 0x%lx is still mapped.\n", + mem->va, bo_size); + mutex_unlock(&mem->lock); + return -EBUSY; + } + + mutex_unlock(&mem->lock); + /* lock is not needed after this, since mem is unused and will + * be freed anyway + */ + + /* Make sure restore workers don't access the BO any more */ + bo_list_entry = &mem->validate_list; + mutex_lock(&process_info->lock); + list_del(&bo_list_entry->head); + mutex_unlock(&process_info->lock); + + ret = reserve_bo_and_cond_vms(mem, NULL, BO_VM_ALL, &ctx); + if (unlikely(ret)) + return ret; + + /* The eviction fence should be removed by the last unmap. + * TODO: Log an error condition if the bo still has the eviction fence + * attached + */ + amdgpu_amdkfd_remove_eviction_fence(mem->bo, + process_info->eviction_fence, + NULL, NULL); + pr_debug("Release VA 0x%llx - 0x%llx\n", mem->va, + mem->va + bo_size * (1 + mem->aql_queue)); + + /* Remove from VM internal data structures */ + list_for_each_entry_safe(entry, tmp, &mem->bo_va_list, bo_list) + remove_bo_from_vm((struct amdgpu_device *)entry->kgd_dev, + entry, bo_size); + + ret = unreserve_bo_and_vms(&ctx, false, false); + + /* Free the sync object */ + amdgpu_sync_free(&mem->sync); + + /* Free the BO*/ + amdgpu_bo_unref(&mem->bo); + mutex_destroy(&mem->lock); + kfree(mem); + + return ret; +} + +int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( + struct kgd_dev *kgd, struct kgd_mem *mem, void *vm) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + struct amdgpu_vm *avm = (struct amdgpu_vm *)vm; + int ret; + struct amdgpu_bo *bo; + uint32_t domain; + struct kfd_bo_va_list *entry; + struct bo_vm_reservation_context ctx; + struct kfd_bo_va_list *bo_va_entry = NULL; + struct kfd_bo_va_list *bo_va_entry_aql = NULL; + unsigned long bo_size; + + /* Make sure restore is not running concurrently. + */ + mutex_lock(&mem->process_info->lock); + + mutex_lock(&mem->lock); + + bo = mem->bo; + + if (!bo) { + pr_err("Invalid BO when mapping memory to GPU\n"); + ret = -EINVAL; + goto out; + } + + domain = mem->domain; + bo_size = bo->tbo.mem.size; + + pr_debug("Map VA 0x%llx - 0x%llx to vm %p domain %s\n", + mem->va, + mem->va + bo_size * (1 + mem->aql_queue), + vm, domain_string(domain)); + + ret = reserve_bo_and_vm(mem, vm, &ctx); + if (unlikely(ret)) + goto out; + + if (check_if_add_bo_to_vm(avm, mem)) { + ret = add_bo_to_vm(adev, mem, avm, false, + &bo_va_entry); + if (ret) + goto add_bo_to_vm_failed; + if (mem->aql_queue) { + ret = add_bo_to_vm(adev, mem, avm, + true, &bo_va_entry_aql); + if (ret) + goto add_bo_to_vm_failed_aql; + } + } else { + ret = vm_validate_pt_pd_bos(avm); + if (unlikely(ret)) + goto add_bo_to_vm_failed; + } + + if (mem->mapped_to_gpu_memory == 0) { + /* Validate BO only once. The eviction fence gets added to BO + * the first time it is mapped. Validate will wait for all + * background evictions to complete. + */ + ret = amdgpu_amdkfd_bo_validate(bo, domain, true); + if (ret) { + pr_debug("Validate failed\n"); + goto map_bo_to_gpuvm_failed; + } + } + + list_for_each_entry(entry, &mem->bo_va_list, bo_list) { + if (entry->bo_va->base.vm == vm && !entry->is_mapped) { + pr_debug("\t map VA 0x%llx - 0x%llx in entry %p\n", + entry->va, entry->va + bo_size, + entry); + + ret = map_bo_to_gpuvm(adev, entry, ctx.sync); + if (ret) { + pr_err("Failed to map radeon bo to gpuvm\n"); + goto map_bo_to_gpuvm_failed; + } + + ret = vm_update_pds(vm, ctx.sync); + if (ret) { + pr_err("Failed to update page directories\n"); + goto map_bo_to_gpuvm_failed; + } + + entry->is_mapped = true; + mem->mapped_to_gpu_memory++; + pr_debug("\t INC mapping count %d\n", + mem->mapped_to_gpu_memory); + } + } + + if (!amdgpu_ttm_tt_get_usermm(bo->tbo.ttm) && !bo->pin_count) + amdgpu_bo_fence(bo, + &avm->process_info->eviction_fence->base, + true); + ret = unreserve_bo_and_vms(&ctx, false, false); + + goto out; + +map_bo_to_gpuvm_failed: + if (bo_va_entry_aql) + remove_bo_from_vm(adev, bo_va_entry_aql, bo_size); +add_bo_to_vm_failed_aql: + if (bo_va_entry) + remove_bo_from_vm(adev, bo_va_entry, bo_size); +add_bo_to_vm_failed: + unreserve_bo_and_vms(&ctx, false, false); +out: + mutex_unlock(&mem->process_info->lock); + mutex_unlock(&mem->lock); + return ret; +} + +int amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu( + struct kgd_dev *kgd, struct kgd_mem *mem, void *vm) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + struct amdkfd_process_info *process_info = + ((struct amdgpu_vm *)vm)->process_info; + unsigned long bo_size = mem->bo->tbo.mem.size; + struct kfd_bo_va_list *entry; + struct bo_vm_reservation_context ctx; + int ret; + + mutex_lock(&mem->lock); + + ret = reserve_bo_and_cond_vms(mem, vm, BO_VM_MAPPED, &ctx); + if (unlikely(ret)) + goto out; + /* If no VMs were reserved, it means the BO wasn't actually mapped */ + if (ctx.n_vms == 0) { + ret = -EINVAL; + goto unreserve_out; + } + + ret = vm_validate_pt_pd_bos((struct amdgpu_vm *)vm); + if (unlikely(ret)) + goto unreserve_out; + + pr_debug("Unmap VA 0x%llx - 0x%llx from vm %p\n", + mem->va, + mem->va + bo_size * (1 + mem->aql_queue), + vm); + + list_for_each_entry(entry, &mem->bo_va_list, bo_list) { + if (entry->bo_va->base.vm == vm && entry->is_mapped) { + pr_debug("\t unmap VA 0x%llx - 0x%llx from entry %p\n", + entry->va, + entry->va + bo_size, + entry); + + ret = unmap_bo_from_gpuvm(adev, entry, ctx.sync); + if (ret == 0) { + entry->is_mapped = false; + } else { + pr_err("failed to unmap VA 0x%llx\n", + mem->va); + goto unreserve_out; + } + + mem->mapped_to_gpu_memory--; + pr_debug("\t DEC mapping count %d\n", + mem->mapped_to_gpu_memory); + } + } + + /* If BO is unmapped from all VMs, unfence it. It can be evicted if + * required. + */ + if (mem->mapped_to_gpu_memory == 0 && + !amdgpu_ttm_tt_get_usermm(mem->bo->tbo.ttm) && !mem->bo->pin_count) + amdgpu_amdkfd_remove_eviction_fence(mem->bo, + process_info->eviction_fence, + NULL, NULL); + +unreserve_out: + unreserve_bo_and_vms(&ctx, false, false); +out: + mutex_unlock(&mem->lock); + return ret; +} + +int amdgpu_amdkfd_gpuvm_sync_memory( + struct kgd_dev *kgd, struct kgd_mem *mem, bool intr) +{ + struct amdgpu_sync sync; + int ret; + + amdgpu_sync_create(&sync); + + mutex_lock(&mem->lock); + amdgpu_sync_clone(&mem->sync, &sync); + mutex_unlock(&mem->lock); + + ret = amdgpu_sync_wait(&sync, intr); + amdgpu_sync_free(&sync); + return ret; +} + +int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_dev *kgd, + struct kgd_mem *mem, void **kptr, uint64_t *size) +{ + int ret; + struct amdgpu_bo *bo = mem->bo; + + if (amdgpu_ttm_tt_get_usermm(bo->tbo.ttm)) { + pr_err("userptr can't be mapped to kernel\n"); + return -EINVAL; + } + + /* delete kgd_mem from kfd_bo_list to avoid re-validating + * this BO in BO's restoring after eviction. + */ + mutex_lock(&mem->process_info->lock); + + ret = amdgpu_bo_reserve(bo, true); + if (ret) { + pr_err("Failed to reserve bo. ret %d\n", ret); + goto bo_reserve_failed; + } + + ret = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL); + if (ret) { + pr_err("Failed to pin bo. ret %d\n", ret); + goto pin_failed; + } + + ret = amdgpu_bo_kmap(bo, kptr); + if (ret) { + pr_err("Failed to map bo to kernel. ret %d\n", ret); + goto kmap_failed; + } + + amdgpu_amdkfd_remove_eviction_fence( + bo, mem->process_info->eviction_fence, NULL, NULL); + list_del_init(&mem->validate_list.head); + + if (size) + *size = amdgpu_bo_size(bo); + + amdgpu_bo_unreserve(bo); + + mutex_unlock(&mem->process_info->lock); + return 0; + +kmap_failed: + amdgpu_bo_unpin(bo); +pin_failed: + amdgpu_bo_unreserve(bo); +bo_reserve_failed: + mutex_unlock(&mem->process_info->lock); + + return ret; +} + +/** amdgpu_amdkfd_gpuvm_restore_process_bos - Restore all BOs for the given + * KFD process identified by process_info + * + * @process_info: amdkfd_process_info of the KFD process + * + * After memory eviction, restore thread calls this function. The function + * should be called when the Process is still valid. BO restore involves - + * + * 1. Release old eviction fence and create new one + * 2. Get two copies of PD BO list from all the VMs. Keep one copy as pd_list. + * 3 Use the second PD list and kfd_bo_list to create a list (ctx.list) of + * BOs that need to be reserved. + * 4. Reserve all the BOs + * 5. Validate of PD and PT BOs. + * 6. Validate all KFD BOs using kfd_bo_list and Map them and add new fence + * 7. Add fence to all PD and PT BOs. + * 8. Unreserve all BOs + */ +int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence **ef) +{ + struct amdgpu_bo_list_entry *pd_bo_list; + struct amdkfd_process_info *process_info = info; + struct amdgpu_vm *peer_vm; + struct kgd_mem *mem; + struct bo_vm_reservation_context ctx; + struct amdgpu_amdkfd_fence *new_fence; + int ret = 0, i; + struct list_head duplicate_save; + struct amdgpu_sync sync_obj; + + INIT_LIST_HEAD(&duplicate_save); + INIT_LIST_HEAD(&ctx.list); + INIT_LIST_HEAD(&ctx.duplicates); + + pd_bo_list = kcalloc(process_info->n_vms, + sizeof(struct amdgpu_bo_list_entry), + GFP_KERNEL); + if (!pd_bo_list) + return -ENOMEM; + + i = 0; + mutex_lock(&process_info->lock); + list_for_each_entry(peer_vm, &process_info->vm_list_head, + vm_list_node) + amdgpu_vm_get_pd_bo(peer_vm, &ctx.list, &pd_bo_list[i++]); + + /* Reserve all BOs and page tables/directory. Add all BOs from + * kfd_bo_list to ctx.list + */ + list_for_each_entry(mem, &process_info->kfd_bo_list, + validate_list.head) { + + list_add_tail(&mem->resv_list.head, &ctx.list); + mem->resv_list.bo = mem->validate_list.bo; + mem->resv_list.shared = mem->validate_list.shared; + } + + ret = ttm_eu_reserve_buffers(&ctx.ticket, &ctx.list, + false, &duplicate_save); + if (ret) { + pr_debug("Memory eviction: TTM Reserve Failed. Try again\n"); + goto ttm_reserve_fail; + } + + amdgpu_sync_create(&sync_obj); + + /* Validate PDs and PTs */ + ret = process_validate_vms(process_info); + if (ret) + goto validate_map_fail; + + /* Wait for PD/PTs validate to finish */ + /* FIXME: I think this isn't needed */ + list_for_each_entry(peer_vm, &process_info->vm_list_head, + vm_list_node) { + struct amdgpu_bo *bo = peer_vm->root.base.bo; + + ttm_bo_wait(&bo->tbo, false, false); + } + + /* Validate BOs and map them to GPUVM (update VM page tables). */ + list_for_each_entry(mem, &process_info->kfd_bo_list, + validate_list.head) { + + struct amdgpu_bo *bo = mem->bo; + uint32_t domain = mem->domain; + struct kfd_bo_va_list *bo_va_entry; + + ret = amdgpu_amdkfd_bo_validate(bo, domain, false); + if (ret) { + pr_debug("Memory eviction: Validate BOs failed. Try again\n"); + goto validate_map_fail; + } + + list_for_each_entry(bo_va_entry, &mem->bo_va_list, + bo_list) { + ret = update_gpuvm_pte((struct amdgpu_device *) + bo_va_entry->kgd_dev, + bo_va_entry, + &sync_obj); + if (ret) { + pr_debug("Memory eviction: update PTE failed. Try again\n"); + goto validate_map_fail; + } + } + } + + /* Update page directories */ + ret = process_update_pds(process_info, &sync_obj); + if (ret) { + pr_debug("Memory eviction: update PDs failed. Try again\n"); + goto validate_map_fail; + } + + amdgpu_sync_wait(&sync_obj, false); + + /* Release old eviction fence and create new one, because fence only + * goes from unsignaled to signaled, fence cannot be reused. + * Use context and mm from the old fence. + */ + new_fence = amdgpu_amdkfd_fence_create( + process_info->eviction_fence->base.context, + process_info->eviction_fence->mm); + if (!new_fence) { + pr_err("Failed to create eviction fence\n"); + ret = -ENOMEM; + goto validate_map_fail; + } + dma_fence_put(&process_info->eviction_fence->base); + process_info->eviction_fence = new_fence; + *ef = dma_fence_get(&new_fence->base); + + /* Wait for validate to finish and attach new eviction fence */ + list_for_each_entry(mem, &process_info->kfd_bo_list, + validate_list.head) + ttm_bo_wait(&mem->bo->tbo, false, false); + list_for_each_entry(mem, &process_info->kfd_bo_list, + validate_list.head) + amdgpu_bo_fence(mem->bo, + &process_info->eviction_fence->base, true); + + /* Attach eviction fence to PD / PT BOs */ + list_for_each_entry(peer_vm, &process_info->vm_list_head, + vm_list_node) { + struct amdgpu_bo *bo = peer_vm->root.base.bo; + + amdgpu_bo_fence(bo, &process_info->eviction_fence->base, true); + } + +validate_map_fail: + ttm_eu_backoff_reservation(&ctx.ticket, &ctx.list); + amdgpu_sync_free(&sync_obj); +ttm_reserve_fail: + mutex_unlock(&process_info->lock); + kfree(pd_bo_list); + return ret; +}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c index ff8efd0..a0f48cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
@@ -114,6 +114,9 @@ union igp_info { struct atom_integrated_system_info_v1_11 v11; }; +union umc_info { + struct atom_umc_info_v3_1 v31; +}; /* * Return vram width from integrated system info table, if available, * or 0 if not. @@ -143,6 +146,94 @@ int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev) return 0; } +static int convert_atom_mem_type_to_vram_type (struct amdgpu_device *adev, + int atom_mem_type) +{ + int vram_type; + + if (adev->flags & AMD_IS_APU) { + switch (atom_mem_type) { + case Ddr2MemType: + case LpDdr2MemType: + vram_type = AMDGPU_VRAM_TYPE_DDR2; + break; + case Ddr3MemType: + case LpDdr3MemType: + vram_type = AMDGPU_VRAM_TYPE_DDR3; + break; + case Ddr4MemType: + case LpDdr4MemType: + vram_type = AMDGPU_VRAM_TYPE_DDR4; + break; + default: + vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; + break; + } + } else { + switch (atom_mem_type) { + case ATOM_DGPU_VRAM_TYPE_GDDR5: + vram_type = AMDGPU_VRAM_TYPE_GDDR5; + break; + case ATOM_DGPU_VRAM_TYPE_HBM: + vram_type = AMDGPU_VRAM_TYPE_HBM; + break; + default: + vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; + break; + } + } + + return vram_type; +} +/* + * Return vram type from either integrated system info table + * or umc info table, if available, or 0 (TYPE_UNKNOWN) if not + */ +int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev) +{ + struct amdgpu_mode_info *mode_info = &adev->mode_info; + int index; + u16 data_offset, size; + union igp_info *igp_info; + union umc_info *umc_info; + u8 frev, crev; + u8 mem_type; + + if (adev->flags & AMD_IS_APU) + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + integratedsysteminfo); + else + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + umc_info); + if (amdgpu_atom_parse_data_header(mode_info->atom_context, + index, &size, + &frev, &crev, &data_offset)) { + if (adev->flags & AMD_IS_APU) { + igp_info = (union igp_info *) + (mode_info->atom_context->bios + data_offset); + switch (crev) { + case 11: + mem_type = igp_info->v11.memorytype; + return convert_atom_mem_type_to_vram_type(adev, mem_type); + default: + return 0; + } + } else { + umc_info = (union umc_info *) + (mode_info->atom_context->bios + data_offset); + switch (crev) { + case 1: + mem_type = umc_info->v31.vram_type; + return convert_atom_mem_type_to_vram_type(adev, mem_type); + default: + return 0; + } + } + } + + return 0; +} + union firmware_info { struct atom_firmware_info_v3_1 v31; }; @@ -151,10 +242,6 @@ union smu_info { struct atom_smu_info_v3_1 v31; }; -union umc_info { - struct atom_umc_info_v3_1 v31; -}; - int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev) { struct amdgpu_mode_info *mode_info = &adev->mode_info;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h index 288b97e..7689c96 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
@@ -28,6 +28,7 @@ bool amdgpu_atomfirmware_gpu_supports_virtualization(struct amdgpu_device *adev) void amdgpu_atomfirmware_scratch_regs_init(struct amdgpu_device *adev); int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev); int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev); +int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev); int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev); #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c index 63ec1e1..02b849b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
@@ -80,8 +80,8 @@ static void amdgpu_benchmark_move(struct amdgpu_device *adev, unsigned size, int time; n = AMDGPU_BENCHMARK_ITERATIONS; - r = amdgpu_bo_create(adev, size, PAGE_SIZE, true, sdomain, 0, NULL, - NULL, 0, &sobj); + r = amdgpu_bo_create(adev, size, PAGE_SIZE,sdomain, 0, + ttm_bo_type_kernel, NULL, &sobj); if (r) { goto out_cleanup; } @@ -93,8 +93,8 @@ static void amdgpu_benchmark_move(struct amdgpu_device *adev, unsigned size, if (r) { goto out_cleanup; } - r = amdgpu_bo_create(adev, size, PAGE_SIZE, true, ddomain, 0, NULL, - NULL, 0, &dobj); + r = amdgpu_bo_create(adev, size, PAGE_SIZE, ddomain, 0, + ttm_bo_type_kernel, NULL, &dobj); if (r) { goto out_cleanup; }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c index 1034d86..8f9ad4fd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -24,12 +24,10 @@ #include <linux/list.h> #include <linux/slab.h> #include <linux/pci.h> -#include <linux/acpi.h> #include <drm/drmP.h> #include <linux/firmware.h> #include <drm/amdgpu_drm.h> #include "amdgpu.h" -#include "cgs_linux.h" #include "atom.h" #include "amdgpu_ucode.h" @@ -42,152 +40,6 @@ struct amdgpu_cgs_device { struct amdgpu_device *adev = \ ((struct amdgpu_cgs_device *)cgs_device)->adev -static void *amdgpu_cgs_register_pp_handle(struct cgs_device *cgs_device, - int (*call_back_func)(struct amd_pp_init *, void **)) -{ - CGS_FUNC_ADEV; - struct amd_pp_init pp_init; - struct amd_powerplay *amd_pp; - - if (call_back_func == NULL) - return NULL; - - amd_pp = &(adev->powerplay); - pp_init.chip_family = adev->family; - pp_init.chip_id = adev->asic_type; - pp_init.pm_en = (amdgpu_dpm != 0 && !amdgpu_sriov_vf(adev)) ? true : false; - pp_init.feature_mask = amdgpu_pp_feature_mask; - pp_init.device = cgs_device; - if (call_back_func(&pp_init, &(amd_pp->pp_handle))) - return NULL; - - return adev->powerplay.pp_handle; -} - -static int amdgpu_cgs_alloc_gpu_mem(struct cgs_device *cgs_device, - enum cgs_gpu_mem_type type, - uint64_t size, uint64_t align, - cgs_handle_t *handle) -{ - CGS_FUNC_ADEV; - uint16_t flags = 0; - int ret = 0; - uint32_t domain = 0; - struct amdgpu_bo *obj; - - /* fail if the alignment is not a power of 2 */ - if (((align != 1) && (align & (align - 1))) - || size == 0 || align == 0) - return -EINVAL; - - - switch(type) { - case CGS_GPU_MEM_TYPE__VISIBLE_CONTIG_FB: - case CGS_GPU_MEM_TYPE__VISIBLE_FB: - flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | - AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; - domain = AMDGPU_GEM_DOMAIN_VRAM; - break; - case CGS_GPU_MEM_TYPE__INVISIBLE_CONTIG_FB: - case CGS_GPU_MEM_TYPE__INVISIBLE_FB: - flags = AMDGPU_GEM_CREATE_NO_CPU_ACCESS | - AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; - domain = AMDGPU_GEM_DOMAIN_VRAM; - break; - case CGS_GPU_MEM_TYPE__GART_CACHEABLE: - domain = AMDGPU_GEM_DOMAIN_GTT; - break; - case CGS_GPU_MEM_TYPE__GART_WRITECOMBINE: - flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC; - domain = AMDGPU_GEM_DOMAIN_GTT; - break; - default: - return -EINVAL; - } - - - *handle = 0; - - ret = amdgpu_bo_create(adev, size, align, true, domain, flags, - NULL, NULL, 0, &obj); - if (ret) { - DRM_ERROR("(%d) bo create failed\n", ret); - return ret; - } - *handle = (cgs_handle_t)obj; - - return ret; -} - -static int amdgpu_cgs_free_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t handle) -{ - struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; - - if (obj) { - int r = amdgpu_bo_reserve(obj, true); - if (likely(r == 0)) { - amdgpu_bo_kunmap(obj); - amdgpu_bo_unpin(obj); - amdgpu_bo_unreserve(obj); - } - amdgpu_bo_unref(&obj); - - } - return 0; -} - -static int amdgpu_cgs_gmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t handle, - uint64_t *mcaddr) -{ - int r; - struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; - - WARN_ON_ONCE(obj->placement.num_placement > 1); - - r = amdgpu_bo_reserve(obj, true); - if (unlikely(r != 0)) - return r; - r = amdgpu_bo_pin(obj, obj->preferred_domains, mcaddr); - amdgpu_bo_unreserve(obj); - return r; -} - -static int amdgpu_cgs_gunmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t handle) -{ - int r; - struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; - r = amdgpu_bo_reserve(obj, true); - if (unlikely(r != 0)) - return r; - r = amdgpu_bo_unpin(obj); - amdgpu_bo_unreserve(obj); - return r; -} - -static int amdgpu_cgs_kmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t handle, - void **map) -{ - int r; - struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; - r = amdgpu_bo_reserve(obj, true); - if (unlikely(r != 0)) - return r; - r = amdgpu_bo_kmap(obj, map); - amdgpu_bo_unreserve(obj); - return r; -} - -static int amdgpu_cgs_kunmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t handle) -{ - int r; - struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; - r = amdgpu_bo_reserve(obj, true); - if (unlikely(r != 0)) - return r; - amdgpu_bo_kunmap(obj); - amdgpu_bo_unreserve(obj); - return r; -} static uint32_t amdgpu_cgs_read_register(struct cgs_device *cgs_device, unsigned offset) { @@ -329,109 +181,6 @@ static int amdgpu_cgs_atom_exec_cmd_table(struct cgs_device *cgs_device, unsigne adev->mode_info.atom_context, table, args); } -struct cgs_irq_params { - unsigned src_id; - cgs_irq_source_set_func_t set; - cgs_irq_handler_func_t handler; - void *private_data; -}; - -static int cgs_set_irq_state(struct amdgpu_device *adev, - struct amdgpu_irq_src *src, - unsigned type, - enum amdgpu_interrupt_state state) -{ - struct cgs_irq_params *irq_params = - (struct cgs_irq_params *)src->data; - if (!irq_params) - return -EINVAL; - if (!irq_params->set) - return -EINVAL; - return irq_params->set(irq_params->private_data, - irq_params->src_id, - type, - (int)state); -} - -static int cgs_process_irq(struct amdgpu_device *adev, - struct amdgpu_irq_src *source, - struct amdgpu_iv_entry *entry) -{ - struct cgs_irq_params *irq_params = - (struct cgs_irq_params *)source->data; - if (!irq_params) - return -EINVAL; - if (!irq_params->handler) - return -EINVAL; - return irq_params->handler(irq_params->private_data, - irq_params->src_id, - entry->iv_entry); -} - -static const struct amdgpu_irq_src_funcs cgs_irq_funcs = { - .set = cgs_set_irq_state, - .process = cgs_process_irq, -}; - -static int amdgpu_cgs_add_irq_source(void *cgs_device, - unsigned client_id, - unsigned src_id, - unsigned num_types, - cgs_irq_source_set_func_t set, - cgs_irq_handler_func_t handler, - void *private_data) -{ - CGS_FUNC_ADEV; - int ret = 0; - struct cgs_irq_params *irq_params; - struct amdgpu_irq_src *source = - kzalloc(sizeof(struct amdgpu_irq_src), GFP_KERNEL); - if (!source) - return -ENOMEM; - irq_params = - kzalloc(sizeof(struct cgs_irq_params), GFP_KERNEL); - if (!irq_params) { - kfree(source); - return -ENOMEM; - } - source->num_types = num_types; - source->funcs = &cgs_irq_funcs; - irq_params->src_id = src_id; - irq_params->set = set; - irq_params->handler = handler; - irq_params->private_data = private_data; - source->data = (void *)irq_params; - ret = amdgpu_irq_add_id(adev, client_id, src_id, source); - if (ret) { - kfree(irq_params); - kfree(source); - } - - return ret; -} - -static int amdgpu_cgs_irq_get(void *cgs_device, unsigned client_id, - unsigned src_id, unsigned type) -{ - CGS_FUNC_ADEV; - - if (!adev->irq.client[client_id].sources) - return -EINVAL; - - return amdgpu_irq_get(adev, adev->irq.client[client_id].sources[src_id], type); -} - -static int amdgpu_cgs_irq_put(void *cgs_device, unsigned client_id, - unsigned src_id, unsigned type) -{ - CGS_FUNC_ADEV; - - if (!adev->irq.client[client_id].sources) - return -EINVAL; - - return amdgpu_irq_put(adev, adev->irq.client[client_id].sources[src_id], type); -} - static int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device, enum amd_ip_block_type block_type, enum amd_clockgating_state state) @@ -803,11 +552,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, else strcpy(fw_name, "amdgpu/vega10_smc.bin"); break; - case CHIP_CARRIZO: - case CHIP_STONEY: - case CHIP_RAVEN: - adev->pm.fw_version = info->version; - return 0; + case CHIP_VEGA12: + strcpy(fw_name, "amdgpu/vega12_smc.bin"); + break; default: DRM_ERROR("SMC firmware not supported\n"); return -EINVAL; @@ -859,61 +606,6 @@ static int amdgpu_cgs_is_virtualization_enabled(void *cgs_device) return amdgpu_sriov_vf(adev); } -static int amdgpu_cgs_query_system_info(struct cgs_device *cgs_device, - struct cgs_system_info *sys_info) -{ - CGS_FUNC_ADEV; - - if (NULL == sys_info) - return -ENODEV; - - if (sizeof(struct cgs_system_info) != sys_info->size) - return -ENODEV; - - switch (sys_info->info_id) { - case CGS_SYSTEM_INFO_ADAPTER_BDF_ID: - sys_info->value = adev->pdev->devfn | (adev->pdev->bus->number << 8); - break; - case CGS_SYSTEM_INFO_PCIE_GEN_INFO: - sys_info->value = adev->pm.pcie_gen_mask; - break; - case CGS_SYSTEM_INFO_PCIE_MLW: - sys_info->value = adev->pm.pcie_mlw_mask; - break; - case CGS_SYSTEM_INFO_PCIE_DEV: - sys_info->value = adev->pdev->device; - break; - case CGS_SYSTEM_INFO_PCIE_REV: - sys_info->value = adev->pdev->revision; - break; - case CGS_SYSTEM_INFO_CG_FLAGS: - sys_info->value = adev->cg_flags; - break; - case CGS_SYSTEM_INFO_PG_FLAGS: - sys_info->value = adev->pg_flags; - break; - case CGS_SYSTEM_INFO_GFX_CU_INFO: - sys_info->value = adev->gfx.cu_info.number; - break; - case CGS_SYSTEM_INFO_GFX_SE_INFO: - sys_info->value = adev->gfx.config.max_shader_engines; - break; - case CGS_SYSTEM_INFO_PCIE_SUB_SYS_ID: - sys_info->value = adev->pdev->subsystem_device; - break; - case CGS_SYSTEM_INFO_PCIE_SUB_SYS_VENDOR_ID: - sys_info->value = adev->pdev->subsystem_vendor; - break; - case CGS_SYSTEM_INFO_PCIE_BUS_DEVFN: - sys_info->value = adev->pdev->devfn; - break; - default: - return -ENODEV; - } - - return 0; -} - static int amdgpu_cgs_get_active_displays_info(struct cgs_device *cgs_device, struct cgs_display_info *info) { @@ -924,12 +616,9 @@ static int amdgpu_cgs_get_active_displays_info(struct cgs_device *cgs_device, return -EINVAL; mode_info = info->mode_info; - if (mode_info) { + if (mode_info) /* if the displays are off, vblank time is max */ mode_info->vblank_time_us = 0xffffffff; - /* always set the reference clock */ - mode_info->ref_clock = adev->clock.spll.reference_freq; - } if (!amdgpu_device_has_dc_support(adev)) { struct amdgpu_crtc *amdgpu_crtc; @@ -955,6 +644,11 @@ static int amdgpu_cgs_get_active_displays_info(struct cgs_device *cgs_device, (amdgpu_crtc->v_border * 2); mode_info->vblank_time_us = vblank_lines * line_time_us; mode_info->refresh_rate = drm_mode_vrefresh(&amdgpu_crtc->hw_mode); + /* we have issues with mclk switching with refresh rates + * over 120 hz on the non-DC code. + */ + if (mode_info->refresh_rate > 120) + mode_info->vblank_time_us = 0; mode_info = NULL; } } @@ -979,223 +673,7 @@ static int amdgpu_cgs_notify_dpm_enabled(struct cgs_device *cgs_device, bool ena return 0; } -/** \brief evaluate acpi namespace object, handle or pathname must be valid - * \param cgs_device - * \param info input/output arguments for the control method - * \return status - */ - -#if defined(CONFIG_ACPI) -static int amdgpu_cgs_acpi_eval_object(struct cgs_device *cgs_device, - struct cgs_acpi_method_info *info) -{ - CGS_FUNC_ADEV; - acpi_handle handle; - struct acpi_object_list input; - struct acpi_buffer output = { ACPI_ALLOCATE_BUFFER, NULL }; - union acpi_object *params, *obj; - uint8_t name[5] = {'\0'}; - struct cgs_acpi_method_argument *argument; - uint32_t i, count; - acpi_status status; - int result; - - handle = ACPI_HANDLE(&adev->pdev->dev); - if (!handle) - return -ENODEV; - - memset(&input, 0, sizeof(struct acpi_object_list)); - - /* validate input info */ - if (info->size != sizeof(struct cgs_acpi_method_info)) - return -EINVAL; - - input.count = info->input_count; - if (info->input_count > 0) { - if (info->pinput_argument == NULL) - return -EINVAL; - argument = info->pinput_argument; - for (i = 0; i < info->input_count; i++) { - if (((argument->type == ACPI_TYPE_STRING) || - (argument->type == ACPI_TYPE_BUFFER)) && - (argument->pointer == NULL)) - return -EINVAL; - argument++; - } - } - - if (info->output_count > 0) { - if (info->poutput_argument == NULL) - return -EINVAL; - argument = info->poutput_argument; - for (i = 0; i < info->output_count; i++) { - if (((argument->type == ACPI_TYPE_STRING) || - (argument->type == ACPI_TYPE_BUFFER)) - && (argument->pointer == NULL)) - return -EINVAL; - argument++; - } - } - - /* The path name passed to acpi_evaluate_object should be null terminated */ - if ((info->field & CGS_ACPI_FIELD_METHOD_NAME) != 0) { - strncpy(name, (char *)&(info->name), sizeof(uint32_t)); - name[4] = '\0'; - } - - /* parse input parameters */ - if (input.count > 0) { - input.pointer = params = - kzalloc(sizeof(union acpi_object) * input.count, GFP_KERNEL); - if (params == NULL) - return -EINVAL; - - argument = info->pinput_argument; - - for (i = 0; i < input.count; i++) { - params->type = argument->type; - switch (params->type) { - case ACPI_TYPE_INTEGER: - params->integer.value = argument->value; - break; - case ACPI_TYPE_STRING: - params->string.length = argument->data_length; - params->string.pointer = argument->pointer; - break; - case ACPI_TYPE_BUFFER: - params->buffer.length = argument->data_length; - params->buffer.pointer = argument->pointer; - break; - default: - break; - } - params++; - argument++; - } - } - - /* parse output info */ - count = info->output_count; - argument = info->poutput_argument; - - /* evaluate the acpi method */ - status = acpi_evaluate_object(handle, name, &input, &output); - - if (ACPI_FAILURE(status)) { - result = -EIO; - goto free_input; - } - - /* return the output info */ - obj = output.pointer; - - if (count > 1) { - if ((obj->type != ACPI_TYPE_PACKAGE) || - (obj->package.count != count)) { - result = -EIO; - goto free_obj; - } - params = obj->package.elements; - } else - params = obj; - - if (params == NULL) { - result = -EIO; - goto free_obj; - } - - for (i = 0; i < count; i++) { - if (argument->type != params->type) { - result = -EIO; - goto free_obj; - } - switch (params->type) { - case ACPI_TYPE_INTEGER: - argument->value = params->integer.value; - break; - case ACPI_TYPE_STRING: - if ((params->string.length != argument->data_length) || - (params->string.pointer == NULL)) { - result = -EIO; - goto free_obj; - } - strncpy(argument->pointer, - params->string.pointer, - params->string.length); - break; - case ACPI_TYPE_BUFFER: - if (params->buffer.pointer == NULL) { - result = -EIO; - goto free_obj; - } - memcpy(argument->pointer, - params->buffer.pointer, - argument->data_length); - break; - default: - break; - } - argument++; - params++; - } - - result = 0; -free_obj: - kfree(obj); -free_input: - kfree((void *)input.pointer); - return result; -} -#else -static int amdgpu_cgs_acpi_eval_object(struct cgs_device *cgs_device, - struct cgs_acpi_method_info *info) -{ - return -EIO; -} -#endif - -static int amdgpu_cgs_call_acpi_method(struct cgs_device *cgs_device, - uint32_t acpi_method, - uint32_t acpi_function, - void *pinput, void *poutput, - uint32_t output_count, - uint32_t input_size, - uint32_t output_size) -{ - struct cgs_acpi_method_argument acpi_input[2] = { {0}, {0} }; - struct cgs_acpi_method_argument acpi_output = {0}; - struct cgs_acpi_method_info info = {0}; - - acpi_input[0].type = CGS_ACPI_TYPE_INTEGER; - acpi_input[0].data_length = sizeof(uint32_t); - acpi_input[0].value = acpi_function; - - acpi_input[1].type = CGS_ACPI_TYPE_BUFFER; - acpi_input[1].data_length = input_size; - acpi_input[1].pointer = pinput; - - acpi_output.type = CGS_ACPI_TYPE_BUFFER; - acpi_output.data_length = output_size; - acpi_output.pointer = poutput; - - info.size = sizeof(struct cgs_acpi_method_info); - info.field = CGS_ACPI_FIELD_METHOD_NAME | CGS_ACPI_FIELD_INPUT_ARGUMENT_COUNT; - info.input_count = 2; - info.name = acpi_method; - info.pinput_argument = acpi_input; - info.output_count = output_count; - info.poutput_argument = &acpi_output; - - return amdgpu_cgs_acpi_eval_object(cgs_device, &info); -} - static const struct cgs_ops amdgpu_cgs_ops = { - .alloc_gpu_mem = amdgpu_cgs_alloc_gpu_mem, - .free_gpu_mem = amdgpu_cgs_free_gpu_mem, - .gmap_gpu_mem = amdgpu_cgs_gmap_gpu_mem, - .gunmap_gpu_mem = amdgpu_cgs_gunmap_gpu_mem, - .kmap_gpu_mem = amdgpu_cgs_kmap_gpu_mem, - .kunmap_gpu_mem = amdgpu_cgs_kunmap_gpu_mem, .read_register = amdgpu_cgs_read_register, .write_register = amdgpu_cgs_write_register, .read_ind_register = amdgpu_cgs_read_ind_register, @@ -1210,18 +688,9 @@ static const struct cgs_ops amdgpu_cgs_ops = { .set_clockgating_state = amdgpu_cgs_set_clockgating_state, .get_active_displays_info = amdgpu_cgs_get_active_displays_info, .notify_dpm_enabled = amdgpu_cgs_notify_dpm_enabled, - .call_acpi_method = amdgpu_cgs_call_acpi_method, - .query_system_info = amdgpu_cgs_query_system_info, .is_virtualization_enabled = amdgpu_cgs_is_virtualization_enabled, .enter_safe_mode = amdgpu_cgs_enter_safe_mode, .lock_grbm_idx = amdgpu_cgs_lock_grbm_idx, - .register_pp_handle = amdgpu_cgs_register_pp_handle, -}; - -static const struct cgs_os_ops amdgpu_cgs_os_ops = { - .add_irq_source = amdgpu_cgs_add_irq_source, - .irq_get = amdgpu_cgs_irq_get, - .irq_put = amdgpu_cgs_irq_put }; struct cgs_device *amdgpu_cgs_create_device(struct amdgpu_device *adev) @@ -1235,7 +704,6 @@ struct cgs_device *amdgpu_cgs_create_device(struct amdgpu_device *adev) } cgs_device->base.ops = &amdgpu_cgs_ops; - cgs_device->base.os_ops = &amdgpu_cgs_os_ops; cgs_device->adev = adev; return (struct cgs_device *)cgs_device;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c index 7a073ac..96501ff 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
@@ -878,7 +878,7 @@ amdgpu_connector_vga_detect(struct drm_connector *connector, bool force) ret = connector_status_disconnected; if (amdgpu_connector->ddc_bus) - dret = amdgpu_ddc_probe(amdgpu_connector, false); + dret = amdgpu_display_ddc_probe(amdgpu_connector, false); if (dret) { amdgpu_connector->detected_by_load = false; amdgpu_connector_free_edid(connector); @@ -1003,7 +1003,7 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force) } if (amdgpu_connector->ddc_bus) - dret = amdgpu_ddc_probe(amdgpu_connector, false); + dret = amdgpu_display_ddc_probe(amdgpu_connector, false); if (dret) { amdgpu_connector->detected_by_load = false; amdgpu_connector_free_edid(connector); @@ -1410,7 +1410,8 @@ amdgpu_connector_dp_detect(struct drm_connector *connector, bool force) /* setup ddc on the bridge */ amdgpu_atombios_encoder_setup_ext_encoder_ddc(encoder); /* bridge chips are always aux */ - if (amdgpu_ddc_probe(amdgpu_connector, true)) /* try DDC */ + /* try DDC */ + if (amdgpu_display_ddc_probe(amdgpu_connector, true)) ret = connector_status_connected; else if (amdgpu_connector->dac_load_detect) { /* try load detection */ const struct drm_encoder_helper_funcs *encoder_funcs = encoder->helper_private; @@ -1430,7 +1431,8 @@ amdgpu_connector_dp_detect(struct drm_connector *connector, bool force) ret = connector_status_connected; } else { /* try non-aux ddc (DP to DVI/HDMI/etc. adapter) */ - if (amdgpu_ddc_probe(amdgpu_connector, false)) + if (amdgpu_display_ddc_probe(amdgpu_connector, + false)) ret = connector_status_connected; } }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 0200eb7..b290f27d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -264,7 +264,7 @@ static void amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev, return; } - total_vram = adev->mc.real_vram_size - adev->vram_pin_size; + total_vram = adev->gmc.real_vram_size - adev->vram_pin_size; used_vram = amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); free_vram = used_vram >= total_vram ? 0 : total_vram - used_vram; @@ -309,8 +309,8 @@ static void amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev, *max_bytes = us_to_bytes(adev, adev->mm_stats.accum_us); /* Do the same for visible VRAM if half of it is free */ - if (adev->mc.visible_vram_size < adev->mc.real_vram_size) { - u64 total_vis_vram = adev->mc.visible_vram_size; + if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size) { + u64 total_vis_vram = adev->gmc.visible_vram_size; u64 used_vis_vram = amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); @@ -353,8 +353,8 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, struct ttm_operation_ctx ctx = { .interruptible = true, .no_wait_gpu = false, - .allow_reserved_eviction = false, - .resv = bo->tbo.resv + .resv = bo->tbo.resv, + .flags = 0 }; uint32_t domain; int r; @@ -366,7 +366,7 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, * to move it. Don't move anything if the threshold is zero. */ if (p->bytes_moved < p->bytes_moved_threshold) { - if (adev->mc.visible_vram_size < adev->mc.real_vram_size && + if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size && (bo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)) { /* And don't move a CPU_ACCESS_REQUIRED BO to limited * visible VRAM if we've depleted our allowance to do @@ -388,9 +388,9 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); p->bytes_moved += ctx.bytes_moved; - if (adev->mc.visible_vram_size < adev->mc.real_vram_size && + if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size && bo->tbo.mem.mem_type == TTM_PL_VRAM && - bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT) + bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT) p->bytes_moved_vis += ctx.bytes_moved; if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) { @@ -444,9 +444,9 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p, /* Good we can try to move this BO somewhere else */ amdgpu_ttm_placement_from_domain(bo, other); update_bytes_moved_vis = - adev->mc.visible_vram_size < adev->mc.real_vram_size && + adev->gmc.visible_vram_size < adev->gmc.real_vram_size && bo->tbo.mem.mem_type == TTM_PL_VRAM && - bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT; + bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT; initial_bytes_moved = atomic64_read(&adev->num_bytes_moved); r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); bytes_moved = atomic64_read(&adev->num_bytes_moved) -
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c index 09d35051..3fabf9f9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -419,9 +419,11 @@ int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, unsigned ring_id) if (other) { signed long r; - r = dma_fence_wait_timeout(other, false, MAX_SCHEDULE_TIMEOUT); + r = dma_fence_wait(other, true); if (r < 0) { - DRM_ERROR("Error (%ld) waiting for fence!\n", r); + if (r != -ERESTARTSYS) + DRM_ERROR("Error (%ld) waiting for fence!\n", r); + return r; } }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index ee76b46..448d69f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
@@ -64,16 +64,21 @@ int amdgpu_debugfs_add_files(struct amdgpu_device *adev, #if defined(CONFIG_DEBUG_FS) -static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, - size_t size, loff_t *pos) + +static int amdgpu_debugfs_process_reg_op(bool read, struct file *f, + char __user *buf, size_t size, loff_t *pos) { struct amdgpu_device *adev = file_inode(f)->i_private; ssize_t result = 0; int r; - bool pm_pg_lock, use_bank; - unsigned instance_bank, sh_bank, se_bank; + bool pm_pg_lock, use_bank, use_ring; + unsigned instance_bank, sh_bank, se_bank, me, pipe, queue; - if (size & 0x3 || *pos & 0x3) + pm_pg_lock = use_bank = use_ring = false; + instance_bank = sh_bank = se_bank = me = pipe = queue = 0; + + if (size & 0x3 || *pos & 0x3 || + ((*pos & (1ULL << 62)) && (*pos & (1ULL << 61)))) return -EINVAL; /* are we reading registers for which a PG lock is necessary? */ @@ -91,8 +96,15 @@ static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, if (instance_bank == 0x3FF) instance_bank = 0xFFFFFFFF; use_bank = 1; + } else if (*pos & (1ULL << 61)) { + + me = (*pos & GENMASK_ULL(33, 24)) >> 24; + pipe = (*pos & GENMASK_ULL(43, 34)) >> 34; + queue = (*pos & GENMASK_ULL(53, 44)) >> 44; + + use_ring = 1; } else { - use_bank = 0; + use_bank = use_ring = 0; } *pos &= (1UL << 22) - 1; @@ -104,6 +116,9 @@ static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, mutex_lock(&adev->grbm_idx_mutex); amdgpu_gfx_select_se_sh(adev, se_bank, sh_bank, instance_bank); + } else if (use_ring) { + mutex_lock(&adev->srbm_mutex); + amdgpu_gfx_select_me_pipe_q(adev, me, pipe, queue); } if (pm_pg_lock) @@ -115,8 +130,14 @@ static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, if (*pos > adev->rmmio_size) goto end; - value = RREG32(*pos >> 2); - r = put_user(value, (uint32_t *)buf); + if (read) { + value = RREG32(*pos >> 2); + r = put_user(value, (uint32_t *)buf); + } else { + r = get_user(value, (uint32_t *)buf); + if (!r) + WREG32(*pos >> 2, value); + } if (r) { result = r; goto end; @@ -132,6 +153,9 @@ static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, if (use_bank) { amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); + } else if (use_ring) { + amdgpu_gfx_select_me_pipe_q(adev, 0, 0, 0); + mutex_unlock(&adev->srbm_mutex); } if (pm_pg_lock) @@ -140,78 +164,17 @@ static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, return result; } + +static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + return amdgpu_debugfs_process_reg_op(true, f, buf, size, pos); +} + static ssize_t amdgpu_debugfs_regs_write(struct file *f, const char __user *buf, size_t size, loff_t *pos) { - struct amdgpu_device *adev = file_inode(f)->i_private; - ssize_t result = 0; - int r; - bool pm_pg_lock, use_bank; - unsigned instance_bank, sh_bank, se_bank; - - if (size & 0x3 || *pos & 0x3) - return -EINVAL; - - /* are we reading registers for which a PG lock is necessary? */ - pm_pg_lock = (*pos >> 23) & 1; - - if (*pos & (1ULL << 62)) { - se_bank = (*pos & GENMASK_ULL(33, 24)) >> 24; - sh_bank = (*pos & GENMASK_ULL(43, 34)) >> 34; - instance_bank = (*pos & GENMASK_ULL(53, 44)) >> 44; - - if (se_bank == 0x3FF) - se_bank = 0xFFFFFFFF; - if (sh_bank == 0x3FF) - sh_bank = 0xFFFFFFFF; - if (instance_bank == 0x3FF) - instance_bank = 0xFFFFFFFF; - use_bank = 1; - } else { - use_bank = 0; - } - - *pos &= (1UL << 22) - 1; - - if (use_bank) { - if ((sh_bank != 0xFFFFFFFF && sh_bank >= adev->gfx.config.max_sh_per_se) || - (se_bank != 0xFFFFFFFF && se_bank >= adev->gfx.config.max_shader_engines)) - return -EINVAL; - mutex_lock(&adev->grbm_idx_mutex); - amdgpu_gfx_select_se_sh(adev, se_bank, - sh_bank, instance_bank); - } - - if (pm_pg_lock) - mutex_lock(&adev->pm.mutex); - - while (size) { - uint32_t value; - - if (*pos > adev->rmmio_size) - return result; - - r = get_user(value, (uint32_t *)buf); - if (r) - return r; - - WREG32(*pos >> 2, value); - - result += 4; - buf += 4; - *pos += 4; - size -= 4; - } - - if (use_bank) { - amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); - mutex_unlock(&adev->grbm_idx_mutex); - } - - if (pm_pg_lock) - mutex_unlock(&adev->pm.mutex); - - return result; + return amdgpu_debugfs_process_reg_op(false, f, (char __user *)buf, size, pos); } static ssize_t amdgpu_debugfs_regs_pcie_read(struct file *f, char __user *buf, @@ -767,10 +730,21 @@ static int amdgpu_debugfs_evict_vram(struct seq_file *m, void *data) return 0; } +static int amdgpu_debugfs_evict_gtt(struct seq_file *m, void *data) +{ + struct drm_info_node *node = (struct drm_info_node *)m->private; + struct drm_device *dev = node->minor->dev; + struct amdgpu_device *adev = dev->dev_private; + + seq_printf(m, "(%d)\n", ttm_bo_evict_mm(&adev->mman.bdev, TTM_PL_TT)); + return 0; +} + static const struct drm_info_list amdgpu_debugfs_list[] = { {"amdgpu_vbios", amdgpu_debugfs_get_vbios_dump}, {"amdgpu_test_ib", &amdgpu_debugfs_test_ib}, - {"amdgpu_evict_vram", &amdgpu_debugfs_evict_vram} + {"amdgpu_evict_vram", &amdgpu_debugfs_evict_vram}, + {"amdgpu_evict_gtt", &amdgpu_debugfs_evict_gtt}, }; int amdgpu_debugfs_init(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 80efcd6..87e92db 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -59,6 +59,7 @@ #include "amdgpu_pm.h" MODULE_FIRMWARE("amdgpu/vega10_gpu_info.bin"); +MODULE_FIRMWARE("amdgpu/vega12_gpu_info.bin"); MODULE_FIRMWARE("amdgpu/raven_gpu_info.bin"); #define AMDGPU_RESUME_MS 2000 @@ -83,10 +84,21 @@ static const char *amdgpu_asic_name[] = { "POLARIS11", "POLARIS12", "VEGA10", + "VEGA12", "RAVEN", "LAST", }; +static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev); + +/** + * amdgpu_device_is_px - Is the device is a dGPU with HG/PX power control + * + * @dev: drm_device pointer + * + * Returns true if the device is a dGPU with HG/PX power control, + * otherwise return false. + */ bool amdgpu_device_is_px(struct drm_device *dev) { struct amdgpu_device *adev = dev->dev_private; @@ -99,6 +111,15 @@ bool amdgpu_device_is_px(struct drm_device *dev) /* * MMIO register access helper functions. */ +/** + * amdgpu_mm_rreg - read a memory mapped IO register + * + * @adev: amdgpu_device pointer + * @reg: dword aligned register offset + * @acc_flags: access flags which require special behavior + * + * Returns the 32 bit value from the offset specified. + */ uint32_t amdgpu_mm_rreg(struct amdgpu_device *adev, uint32_t reg, uint32_t acc_flags) { @@ -121,6 +142,58 @@ uint32_t amdgpu_mm_rreg(struct amdgpu_device *adev, uint32_t reg, return ret; } +/* + * MMIO register read with bytes helper functions + * @offset:bytes offset from MMIO start + * +*/ + +/** + * amdgpu_mm_rreg8 - read a memory mapped IO register + * + * @adev: amdgpu_device pointer + * @offset: byte aligned register offset + * + * Returns the 8 bit value from the offset specified. + */ +uint8_t amdgpu_mm_rreg8(struct amdgpu_device *adev, uint32_t offset) { + if (offset < adev->rmmio_size) + return (readb(adev->rmmio + offset)); + BUG(); +} + +/* + * MMIO register write with bytes helper functions + * @offset:bytes offset from MMIO start + * @value: the value want to be written to the register + * +*/ +/** + * amdgpu_mm_wreg8 - read a memory mapped IO register + * + * @adev: amdgpu_device pointer + * @offset: byte aligned register offset + * @value: 8 bit value to write + * + * Writes the value specified to the offset specified. + */ +void amdgpu_mm_wreg8(struct amdgpu_device *adev, uint32_t offset, uint8_t value) { + if (offset < adev->rmmio_size) + writeb(value, adev->rmmio + offset); + else + BUG(); +} + +/** + * amdgpu_mm_wreg - write to a memory mapped IO register + * + * @adev: amdgpu_device pointer + * @reg: dword aligned register offset + * @v: 32 bit value to write to the register + * @acc_flags: access flags which require special behavior + * + * Writes the value specified to the offset specified. + */ void amdgpu_mm_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, uint32_t acc_flags) { @@ -149,6 +222,14 @@ void amdgpu_mm_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, } } +/** + * amdgpu_io_rreg - read an IO register + * + * @adev: amdgpu_device pointer + * @reg: dword aligned register offset + * + * Returns the 32 bit value from the offset specified. + */ u32 amdgpu_io_rreg(struct amdgpu_device *adev, u32 reg) { if ((reg * 4) < adev->rio_mem_size) @@ -159,6 +240,15 @@ u32 amdgpu_io_rreg(struct amdgpu_device *adev, u32 reg) } } +/** + * amdgpu_io_wreg - write to an IO register + * + * @adev: amdgpu_device pointer + * @reg: dword aligned register offset + * @v: 32 bit value to write to the register + * + * Writes the value specified to the offset specified. + */ void amdgpu_io_wreg(struct amdgpu_device *adev, u32 reg, u32 v) { if (adev->asic_type >= CHIP_VEGA10 && reg == 0) { @@ -327,6 +417,14 @@ static void amdgpu_block_invalid_wreg(struct amdgpu_device *adev, BUG(); } +/** + * amdgpu_device_vram_scratch_init - allocate the VRAM scratch page + * + * @adev: amdgpu device pointer + * + * Allocates a scratch page of VRAM for use by various things in the + * driver. + */ static int amdgpu_device_vram_scratch_init(struct amdgpu_device *adev) { return amdgpu_bo_create_kernel(adev, AMDGPU_GPU_PAGE_SIZE, @@ -336,6 +434,13 @@ static int amdgpu_device_vram_scratch_init(struct amdgpu_device *adev) (void **)&adev->vram_scratch.ptr); } +/** + * amdgpu_device_vram_scratch_fini - Free the VRAM scratch page + * + * @adev: amdgpu device pointer + * + * Frees the VRAM scratch page. + */ static void amdgpu_device_vram_scratch_fini(struct amdgpu_device *adev) { amdgpu_bo_free_kernel(&adev->vram_scratch.robj, NULL, NULL); @@ -377,6 +482,14 @@ void amdgpu_device_program_register_sequence(struct amdgpu_device *adev, } } +/** + * amdgpu_device_pci_config_reset - reset the GPU + * + * @adev: amdgpu_device pointer + * + * Resets the GPU using the pci config reset sequence. + * Only applicable to asics prior to vega10. + */ void amdgpu_device_pci_config_reset(struct amdgpu_device *adev) { pci_write_config_dword(adev->pdev, 0x7c, AMDGPU_ASIC_RESET_DATA); @@ -537,6 +650,7 @@ void amdgpu_device_wb_free(struct amdgpu_device *adev, u32 wb) /** * amdgpu_device_vram_location - try to find VRAM location + * * @adev: amdgpu device structure holding all necessary informations * @mc: memory controller structure holding memory informations * @base: base address at which to put VRAM @@ -545,7 +659,7 @@ void amdgpu_device_wb_free(struct amdgpu_device *adev, u32 wb) * as parameter. */ void amdgpu_device_vram_location(struct amdgpu_device *adev, - struct amdgpu_mc *mc, u64 base) + struct amdgpu_gmc *mc, u64 base) { uint64_t limit = (uint64_t)amdgpu_vram_limit << 20; @@ -560,6 +674,7 @@ void amdgpu_device_vram_location(struct amdgpu_device *adev, /** * amdgpu_device_gart_location - try to find GTT location + * * @adev: amdgpu device structure holding all necessary informations * @mc: memory controller structure holding memory informations * @@ -571,11 +686,11 @@ void amdgpu_device_vram_location(struct amdgpu_device *adev, * FIXME: when reducing GTT size align new size on power of 2. */ void amdgpu_device_gart_location(struct amdgpu_device *adev, - struct amdgpu_mc *mc) + struct amdgpu_gmc *mc) { u64 size_af, size_bf; - size_af = adev->mc.mc_mask - mc->vram_end; + size_af = adev->gmc.mc_mask - mc->vram_end; size_bf = mc->vram_start; if (size_bf > size_af) { if (mc->gart_size > size_bf) { @@ -609,7 +724,7 @@ void amdgpu_device_gart_location(struct amdgpu_device *adev, */ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev) { - u64 space_needed = roundup_pow_of_two(adev->mc.real_vram_size); + u64 space_needed = roundup_pow_of_two(adev->gmc.real_vram_size); u32 rbar_size = order_base_2(((space_needed >> 20) | 1)) - 1; struct pci_bus *root; struct resource *res; @@ -746,6 +861,16 @@ static unsigned int amdgpu_device_vga_set_decode(void *cookie, bool state) return VGA_RSRC_NORMAL_IO | VGA_RSRC_NORMAL_MEM; } +/** + * amdgpu_device_check_block_size - validate the vm block size + * + * @adev: amdgpu_device pointer + * + * Validates the vm block size specified via module parameter. + * The vm block size defines number of bits in page table versus page directory, + * a page is 4KB so we have 12 bits offset, minimum 9 bits in the + * page table and the remaining bits are in the page directory. + */ static void amdgpu_device_check_block_size(struct amdgpu_device *adev) { /* defines number of bits in page table versus page directory, @@ -761,6 +886,14 @@ static void amdgpu_device_check_block_size(struct amdgpu_device *adev) } } +/** + * amdgpu_device_check_vm_size - validate the vm size + * + * @adev: amdgpu_device pointer + * + * Validates the vm size in GB specified via module parameter. + * The VM size is the size of the GPU virtual memory space in GB. + */ static void amdgpu_device_check_vm_size(struct amdgpu_device *adev) { /* no need to check the default value */ @@ -830,6 +963,8 @@ static void amdgpu_device_check_arguments(struct amdgpu_device *adev) dev_warn(adev->dev, "lockup_timeout msut be > 0, adjusting to 10000\n"); amdgpu_lockup_timeout = 10000; } + + adev->firmware.load_type = amdgpu_ucode_get_load_type(adev, amdgpu_fw_load_type); } /** @@ -893,6 +1028,17 @@ static const struct vga_switcheroo_client_ops amdgpu_switcheroo_ops = { .can_switch = amdgpu_switcheroo_can_switch, }; +/** + * amdgpu_device_ip_set_clockgating_state - set the CG state + * + * @adev: amdgpu_device pointer + * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) + * @state: clockgating state (gate or ungate) + * + * Sets the requested clockgating state for all instances of + * the hardware IP specified. + * Returns the error code from the last instance. + */ int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev, enum amd_ip_block_type block_type, enum amd_clockgating_state state) @@ -915,6 +1061,17 @@ int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev, return r; } +/** + * amdgpu_device_ip_set_powergating_state - set the PG state + * + * @adev: amdgpu_device pointer + * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) + * @state: powergating state (gate or ungate) + * + * Sets the requested powergating state for all instances of + * the hardware IP specified. + * Returns the error code from the last instance. + */ int amdgpu_device_ip_set_powergating_state(struct amdgpu_device *adev, enum amd_ip_block_type block_type, enum amd_powergating_state state) @@ -937,6 +1094,17 @@ int amdgpu_device_ip_set_powergating_state(struct amdgpu_device *adev, return r; } +/** + * amdgpu_device_ip_get_clockgating_state - get the CG state + * + * @adev: amdgpu_device pointer + * @flags: clockgating feature flags + * + * Walks the list of IPs on the device and updates the clockgating + * flags for each IP. + * Updates @flags with the feature flags for each hardware IP where + * clockgating is enabled. + */ void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev, u32 *flags) { @@ -950,6 +1118,15 @@ void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev, } } +/** + * amdgpu_device_ip_wait_for_idle - wait for idle + * + * @adev: amdgpu_device pointer + * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) + * + * Waits for the request hardware IP to be idle. + * Returns 0 for success or a negative error code on failure. + */ int amdgpu_device_ip_wait_for_idle(struct amdgpu_device *adev, enum amd_ip_block_type block_type) { @@ -969,6 +1146,15 @@ int amdgpu_device_ip_wait_for_idle(struct amdgpu_device *adev, } +/** + * amdgpu_device_ip_is_idle - is the hardware IP idle + * + * @adev: amdgpu_device pointer + * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) + * + * Check if the hardware IP is idle or not. + * Returns true if it the IP is idle, false if not. + */ bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev, enum amd_ip_block_type block_type) { @@ -984,6 +1170,15 @@ bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev, } +/** + * amdgpu_device_ip_get_ip_block - get a hw IP pointer + * + * @adev: amdgpu_device pointer + * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) + * + * Returns a pointer to the hardware IP block structure + * if it exists for the asic, otherwise NULL. + */ struct amdgpu_ip_block * amdgpu_device_ip_get_ip_block(struct amdgpu_device *adev, enum amd_ip_block_type type) @@ -1037,7 +1232,7 @@ int amdgpu_device_ip_block_add(struct amdgpu_device *adev, if (!ip_block_version) return -EINVAL; - DRM_DEBUG("add ip block number %d <%s>\n", adev->num_ip_blocks, + DRM_INFO("add ip block number %d <%s>\n", adev->num_ip_blocks, ip_block_version->funcs->name); adev->ip_blocks[adev->num_ip_blocks++].version = ip_block_version; @@ -1045,6 +1240,18 @@ int amdgpu_device_ip_block_add(struct amdgpu_device *adev, return 0; } +/** + * amdgpu_device_enable_virtual_display - enable virtual display feature + * + * @adev: amdgpu_device pointer + * + * Enabled the virtual display feature if the user has enabled it via + * the module parameter virtual_display. This feature provides a virtual + * display hardware on headless boards or in virtualized environments. + * This function parses and validates the configuration string specified by + * the user and configues the virtual display configuration (number of + * virtual connectors, crtcs, etc.) specified. + */ static void amdgpu_device_enable_virtual_display(struct amdgpu_device *adev) { adev->enable_virtual_display = false; @@ -1090,6 +1297,16 @@ static void amdgpu_device_enable_virtual_display(struct amdgpu_device *adev) } } +/** + * amdgpu_device_parse_gpu_info_fw - parse gpu info firmware + * + * @adev: amdgpu_device pointer + * + * Parses the asic configuration parameters specified in the gpu info + * firmware and makes them availale to the driver for use in configuring + * the asic. + * Returns 0 on success, -EINVAL on failure. + */ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev) { const char *chip_name; @@ -1127,6 +1344,9 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev) case CHIP_VEGA10: chip_name = "vega10"; break; + case CHIP_VEGA12: + chip_name = "vega12"; + break; case CHIP_RAVEN: chip_name = "raven"; break; @@ -1188,6 +1408,16 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev) return err; } +/** + * amdgpu_device_ip_early_init - run early init for hardware IPs + * + * @adev: amdgpu_device pointer + * + * Early initialization pass for hardware IPs. The hardware IPs that make + * up each asic are discovered each IP's early_init callback is run. This + * is the first stage in initializing the asic. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) { int i, r; @@ -1240,8 +1470,9 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) return r; break; #endif - case CHIP_VEGA10: - case CHIP_RAVEN: + case CHIP_VEGA10: + case CHIP_VEGA12: + case CHIP_RAVEN: if (adev->asic_type == CHIP_RAVEN) adev->family = AMDGPU_FAMILY_RV; else @@ -1297,6 +1528,17 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_ip_init - run init for hardware IPs + * + * @adev: amdgpu_device pointer + * + * Main initialization pass for hardware IPs. The list of all the hardware + * IPs that make up the asic is walked and the sw_init and hw_init callbacks + * are run. sw_init initializes the software state associated with each IP + * and hw_init initializes the hardware associated with each IP. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_init(struct amdgpu_device *adev) { int i, r; @@ -1311,6 +1553,7 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) return r; } adev->ip_blocks[i].status.sw = true; + /* need to do gmc hw init early so we can allocate gpu mem */ if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC) { r = amdgpu_device_vram_scratch_init(adev); @@ -1344,8 +1587,7 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) for (i = 0; i < adev->num_ip_blocks; i++) { if (!adev->ip_blocks[i].status.sw) continue; - /* gmc hw init is done early */ - if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC) + if (adev->ip_blocks[i].status.hw) continue; r = adev->ip_blocks[i].version->funcs->hw_init((void *)adev); if (r) { @@ -1364,11 +1606,30 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_fill_reset_magic - writes reset magic to gart pointer + * + * @adev: amdgpu_device pointer + * + * Writes a reset magic value to the gart pointer in VRAM. The driver calls + * this function before a GPU reset. If the value is retained after a + * GPU reset, VRAM has not been lost. Some GPU resets may destry VRAM contents. + */ static void amdgpu_device_fill_reset_magic(struct amdgpu_device *adev) { memcpy(adev->reset_magic, adev->gart.ptr, AMDGPU_RESET_MAGIC_NUM); } +/** + * amdgpu_device_check_vram_lost - check if vram is valid + * + * @adev: amdgpu_device pointer + * + * Checks the reset magic value written to the gart pointer in VRAM. + * The driver calls this after a GPU reset to see if the contents of + * VRAM is lost or now. + * returns true if vram is lost, false if not. + */ static bool amdgpu_device_check_vram_lost(struct amdgpu_device *adev) { return !!memcmp(adev->gart.ptr, adev->reset_magic, @@ -1392,13 +1653,17 @@ static int amdgpu_device_set_cg_state(struct amdgpu_device *adev, { int i, j, r; + if (amdgpu_emu_mode == 1) + return 0; + for (j = 0; j < adev->num_ip_blocks; j++) { i = state == AMD_CG_STATE_GATE ? j : adev->num_ip_blocks - j - 1; if (!adev->ip_blocks[i].status.valid) continue; /* skip CG for VCE/UVD, it's handled specially */ if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD && - adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE) { + adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE && + adev->ip_blocks[i].version->funcs->set_clockgating_state) { /* enable clockgating to save power */ r = adev->ip_blocks[i].version->funcs->set_clockgating_state((void *)adev, state); @@ -1458,6 +1723,18 @@ static int amdgpu_device_set_pg_state(struct amdgpu_device *adev, enum amd_power return 0; } +/** + * amdgpu_device_ip_late_init - run late init for hardware IPs + * + * @adev: amdgpu_device pointer + * + * Late initialization pass for hardware IPs. The list of all the hardware + * IPs that make up the asic is walked and the late_init callbacks are run. + * late_init covers any special initialization that an IP requires + * after all of the have been initialized or something that needs to happen + * late in the init process. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev) { int i = 0, r; @@ -1487,6 +1764,17 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_ip_fini - run fini for hardware IPs + * + * @adev: amdgpu_device pointer + * + * Main teardown pass for hardware IPs. The list of all the hardware + * IPs that make up the asic is walked and the hw_fini and sw_fini callbacks + * are run. hw_fini tears down the hardware associated with each IP + * and sw_fini tears down any software state associated with each IP. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) { int i, r; @@ -1513,9 +1801,6 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) } for (i = adev->num_ip_blocks - 1; i >= 0; i--) { - if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC && - adev->firmware.load_type == AMDGPU_FW_LOAD_SMU) - amdgpu_ucode_fini_bo(adev); if (!adev->ip_blocks[i].status.hw) continue; @@ -1529,6 +1814,7 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) adev->ip_blocks[i].status.hw = false; } + for (i = adev->num_ip_blocks - 1; i >= 0; i--) { if (!adev->ip_blocks[i].status.sw) continue; @@ -1707,6 +1993,8 @@ static int amdgpu_device_ip_reinit_early_sriov(struct amdgpu_device *adev) r = block->version->funcs->hw_init(adev); DRM_INFO("RE-INIT: %s %s\n", block->version->funcs->name, r?"failed":"successed"); + if (r) + return r; } } @@ -1740,12 +2028,26 @@ static int amdgpu_device_ip_reinit_late_sriov(struct amdgpu_device *adev) r = block->version->funcs->hw_init(adev); DRM_INFO("RE-INIT: %s %s\n", block->version->funcs->name, r?"failed":"successed"); + if (r) + return r; } } return 0; } +/** + * amdgpu_device_ip_resume_phase1 - run resume for hardware IPs + * + * @adev: amdgpu_device pointer + * + * First resume function for hardware IPs. The list of all the hardware + * IPs that make up the asic is walked and the resume callbacks are run for + * COMMON, GMC, and IH. resume puts the hardware into a functional state + * after a suspend and updates the software state as necessary. This + * function is also used for restoring the GPU after a GPU reset. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev) { int i, r; @@ -1754,9 +2056,8 @@ static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev) if (!adev->ip_blocks[i].status.valid) continue; if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC || - adev->ip_blocks[i].version->type == - AMD_IP_BLOCK_TYPE_IH) { + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC || + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH) { r = adev->ip_blocks[i].version->funcs->resume(adev); if (r) { DRM_ERROR("resume of IP block <%s> failed %d\n", @@ -1769,6 +2070,19 @@ static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_ip_resume_phase2 - run resume for hardware IPs + * + * @adev: amdgpu_device pointer + * + * First resume function for hardware IPs. The list of all the hardware + * IPs that make up the asic is walked and the resume callbacks are run for + * all blocks except COMMON, GMC, and IH. resume puts the hardware into a + * functional state after a suspend and updates the software state as + * necessary. This function is also used for restoring the GPU after a GPU + * reset. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_resume_phase2(struct amdgpu_device *adev) { int i, r; @@ -1777,8 +2091,8 @@ static int amdgpu_device_ip_resume_phase2(struct amdgpu_device *adev) if (!adev->ip_blocks[i].status.valid) continue; if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH ) + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC || + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH) continue; r = adev->ip_blocks[i].version->funcs->resume(adev); if (r) { @@ -1791,6 +2105,18 @@ static int amdgpu_device_ip_resume_phase2(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_ip_resume - run resume for hardware IPs + * + * @adev: amdgpu_device pointer + * + * Main resume function for hardware IPs. The hardware IPs + * are split into two resume functions because they are + * are also used in in recovering from a GPU reset and some additional + * steps need to be take between them. In this case (S3/S4) they are + * run sequentially. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_resume(struct amdgpu_device *adev) { int r; @@ -1803,6 +2129,13 @@ static int amdgpu_device_ip_resume(struct amdgpu_device *adev) return r; } +/** + * amdgpu_device_detect_sriov_bios - determine if the board supports SR-IOV + * + * @adev: amdgpu_device pointer + * + * Query the VBIOS data tables to determine if the board supports SR-IOV. + */ static void amdgpu_device_detect_sriov_bios(struct amdgpu_device *adev) { if (amdgpu_sriov_vf(adev)) { @@ -1819,6 +2152,14 @@ static void amdgpu_device_detect_sriov_bios(struct amdgpu_device *adev) } } +/** + * amdgpu_device_asic_has_dc_support - determine if DC supports the asic + * + * @asic_type: AMD asic type + * + * Check if there is DC (new modesetting infrastructre) support for an asic. + * returns true if DC has support, false if not. + */ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type) { switch (asic_type) { @@ -1836,6 +2177,7 @@ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type) case CHIP_TONGA: case CHIP_FIJI: case CHIP_VEGA10: + case CHIP_VEGA12: #if defined(CONFIG_DRM_AMD_DC_DCN1_0) case CHIP_RAVEN: #endif @@ -1889,14 +2231,16 @@ int amdgpu_device_init(struct amdgpu_device *adev, adev->flags = flags; adev->asic_type = flags & AMD_ASIC_MASK; adev->usec_timeout = AMDGPU_MAX_USEC_TIMEOUT; - adev->mc.gart_size = 512 * 1024 * 1024; + if (amdgpu_emu_mode == 1) + adev->usec_timeout *= 2; + adev->gmc.gart_size = 512 * 1024 * 1024; adev->accel_working = false; adev->num_rings = 0; adev->mman.buffer_funcs = NULL; adev->mman.buffer_funcs_ring = NULL; adev->vm_manager.vm_pte_funcs = NULL; adev->vm_manager.vm_pte_num_rings = 0; - adev->gart.gart_funcs = NULL; + adev->gmc.gmc_funcs = NULL; adev->fence_context = dma_fence_context_alloc(AMDGPU_MAX_RINGS); bitmap_zero(adev->gfx.pipe_reserve_bitmap, AMDGPU_MAX_COMPUTE_QUEUES); @@ -1985,6 +2329,8 @@ int amdgpu_device_init(struct amdgpu_device *adev, if (adev->rio_mem == NULL) DRM_INFO("PCI I/O BAR is not found.\n"); + amdgpu_device_get_pcie_info(adev); + /* early init functions */ r = amdgpu_device_ip_early_init(adev); if (r) @@ -2003,6 +2349,12 @@ int amdgpu_device_init(struct amdgpu_device *adev, if (runtime) vga_switcheroo_init_domain_pm_ops(adev->dev, &adev->vga_pm_domain); + if (amdgpu_emu_mode == 1) { + /* post the asic on emulation mode */ + emu_soc_asic_init(adev); + goto fence_driver_init; + } + /* Read BIOS */ if (!amdgpu_get_bios(adev)) { r = -EINVAL; @@ -2055,6 +2407,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, amdgpu_atombios_i2c_init(adev); } +fence_driver_init: /* Fence driver */ r = amdgpu_fence_driver_init(adev); if (r) { @@ -2082,7 +2435,6 @@ int amdgpu_device_init(struct amdgpu_device *adev, } dev_err(adev->dev, "amdgpu_device_ip_init failed\n"); amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, 0, 0); - amdgpu_device_ip_fini(adev); goto failed; } @@ -2177,6 +2529,8 @@ void amdgpu_device_fini(struct amdgpu_device *adev) DRM_INFO("amdgpu: finishing device.\n"); adev->shutdown = true; + /* disable all interrupts */ + amdgpu_irq_disable_all(adev); if (adev->mode_info.mode_config_initialized){ if (!amdgpu_device_has_dc_support(adev)) drm_crtc_force_disable_all(adev->ddev); @@ -2185,6 +2539,7 @@ void amdgpu_device_fini(struct amdgpu_device *adev) } amdgpu_ib_pool_fini(adev); amdgpu_fence_driver_fini(adev); + amdgpu_pm_sysfs_fini(adev); amdgpu_fbdev_fini(adev); r = amdgpu_device_ip_fini(adev); if (adev->firmware.gpu_info_fw) { @@ -2196,7 +2551,10 @@ void amdgpu_device_fini(struct amdgpu_device *adev) /* free i2c buses */ if (!amdgpu_device_has_dc_support(adev)) amdgpu_i2c_fini(adev); - amdgpu_atombios_fini(adev); + + if (amdgpu_emu_mode != 1) + amdgpu_atombios_fini(adev); + kfree(adev->bios); adev->bios = NULL; if (!pci_is_thunderbolt_attached(adev->pdev)) @@ -2210,7 +2568,6 @@ void amdgpu_device_fini(struct amdgpu_device *adev) iounmap(adev->rmmio); adev->rmmio = NULL; amdgpu_device_doorbell_fini(adev); - amdgpu_pm_sysfs_fini(adev); amdgpu_debugfs_regs_cleanup(adev); } @@ -2429,6 +2786,16 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) return 0; } +/** + * amdgpu_device_ip_check_soft_reset - did soft reset succeed + * + * @adev: amdgpu_device pointer + * + * The list of all the hardware IPs that make up the asic is walked and + * the check_soft_reset callbacks are run. check_soft_reset determines + * if the asic is still hung or not. + * Returns true if any of the IPs are still in a hung state, false if not. + */ static bool amdgpu_device_ip_check_soft_reset(struct amdgpu_device *adev) { int i; @@ -2451,6 +2818,17 @@ static bool amdgpu_device_ip_check_soft_reset(struct amdgpu_device *adev) return asic_hang; } +/** + * amdgpu_device_ip_pre_soft_reset - prepare for soft reset + * + * @adev: amdgpu_device pointer + * + * The list of all the hardware IPs that make up the asic is walked and the + * pre_soft_reset callbacks are run if the block is hung. pre_soft_reset + * handles any IP specific hardware or software state changes that are + * necessary for a soft reset to succeed. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_pre_soft_reset(struct amdgpu_device *adev) { int i, r = 0; @@ -2469,6 +2847,15 @@ static int amdgpu_device_ip_pre_soft_reset(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_ip_need_full_reset - check if a full asic reset is needed + * + * @adev: amdgpu_device pointer + * + * Some hardware IPs cannot be soft reset. If they are hung, a full gpu + * reset is necessary to recover. + * Returns true if a full asic reset is required, false if not. + */ static bool amdgpu_device_ip_need_full_reset(struct amdgpu_device *adev) { int i; @@ -2490,6 +2877,17 @@ static bool amdgpu_device_ip_need_full_reset(struct amdgpu_device *adev) return false; } +/** + * amdgpu_device_ip_soft_reset - do a soft reset + * + * @adev: amdgpu_device pointer + * + * The list of all the hardware IPs that make up the asic is walked and the + * soft_reset callbacks are run if the block is hung. soft_reset handles any + * IP specific hardware or software state changes that are necessary to soft + * reset the IP. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_soft_reset(struct amdgpu_device *adev) { int i, r = 0; @@ -2508,6 +2906,17 @@ static int amdgpu_device_ip_soft_reset(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_ip_post_soft_reset - clean up from soft reset + * + * @adev: amdgpu_device pointer + * + * The list of all the hardware IPs that make up the asic is walked and the + * post_soft_reset callbacks are run if the asic was hung. post_soft_reset + * handles any IP specific hardware or software state changes that are + * necessary after the IP has been soft reset. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_ip_post_soft_reset(struct amdgpu_device *adev) { int i, r = 0; @@ -2525,6 +2934,19 @@ static int amdgpu_device_ip_post_soft_reset(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_recover_vram_from_shadow - restore shadowed VRAM buffers + * + * @adev: amdgpu_device pointer + * @ring: amdgpu_ring for the engine handling the buffer operations + * @bo: amdgpu_bo buffer whose shadow is being restored + * @fence: dma_fence associated with the operation + * + * Restores the VRAM buffer contents from the shadow in GTT. Used to + * restore things like GPUVM page tables after a GPU reset where + * the contents of VRAM might be lost. + * Returns 0 on success, negative error code on failure. + */ static int amdgpu_device_recover_vram_from_shadow(struct amdgpu_device *adev, struct amdgpu_ring *ring, struct amdgpu_bo *bo, @@ -2560,17 +2982,81 @@ static int amdgpu_device_recover_vram_from_shadow(struct amdgpu_device *adev, return r; } -/* +/** + * amdgpu_device_handle_vram_lost - Handle the loss of VRAM contents + * + * @adev: amdgpu_device pointer + * + * Restores the contents of VRAM buffers from the shadows in GTT. Used to + * restore things like GPUVM page tables after a GPU reset where + * the contents of VRAM might be lost. + * Returns 0 on success, 1 on failure. + */ +static int amdgpu_device_handle_vram_lost(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring; + struct amdgpu_bo *bo, *tmp; + struct dma_fence *fence = NULL, *next = NULL; + long r = 1; + int i = 0; + long tmo; + + if (amdgpu_sriov_runtime(adev)) + tmo = msecs_to_jiffies(amdgpu_lockup_timeout); + else + tmo = msecs_to_jiffies(100); + + DRM_INFO("recover vram bo from shadow start\n"); + mutex_lock(&adev->shadow_list_lock); + list_for_each_entry_safe(bo, tmp, &adev->shadow_list, shadow_list) { + next = NULL; + amdgpu_device_recover_vram_from_shadow(adev, ring, bo, &next); + if (fence) { + r = dma_fence_wait_timeout(fence, false, tmo); + if (r == 0) + pr_err("wait fence %p[%d] timeout\n", fence, i); + else if (r < 0) + pr_err("wait fence %p[%d] interrupted\n", fence, i); + if (r < 1) { + dma_fence_put(fence); + fence = next; + break; + } + i++; + } + + dma_fence_put(fence); + fence = next; + } + mutex_unlock(&adev->shadow_list_lock); + + if (fence) { + r = dma_fence_wait_timeout(fence, false, tmo); + if (r == 0) + pr_err("wait fence %p[%d] timeout\n", fence, i); + else if (r < 0) + pr_err("wait fence %p[%d] interrupted\n", fence, i); + + } + dma_fence_put(fence); + + if (r > 0) + DRM_INFO("recover vram bo from shadow done\n"); + else + DRM_ERROR("recover vram bo from shadow failed\n"); + + return (r > 0) ? 0 : 1; +} + +/** * amdgpu_device_reset - reset ASIC/GPU for bare-metal or passthrough * * @adev: amdgpu device pointer - * @reset_flags: output param tells caller the reset result * * attempt to do soft-reset or full-reset and reinitialize Asic * return 0 means successed otherwise failed -*/ -static int amdgpu_device_reset(struct amdgpu_device *adev, - uint64_t* reset_flags) + */ +static int amdgpu_device_reset(struct amdgpu_device *adev) { bool need_full_reset, vram_lost = 0; int r; @@ -2585,7 +3071,6 @@ static int amdgpu_device_reset(struct amdgpu_device *adev, DRM_INFO("soft reset failed, will fallback to full reset!\n"); need_full_reset = true; } - } if (need_full_reset) { @@ -2634,28 +3119,21 @@ static int amdgpu_device_reset(struct amdgpu_device *adev, } } - if (reset_flags) { - if (vram_lost) - (*reset_flags) |= AMDGPU_RESET_INFO_VRAM_LOST; - - if (need_full_reset) - (*reset_flags) |= AMDGPU_RESET_INFO_FULLRESET; - } + if (!r && ((need_full_reset && !(adev->flags & AMD_IS_APU)) || vram_lost)) + r = amdgpu_device_handle_vram_lost(adev); return r; } -/* +/** * amdgpu_device_reset_sriov - reset ASIC for SR-IOV vf * * @adev: amdgpu device pointer - * @reset_flags: output param tells caller the reset result * * do VF FLR and reinitialize Asic * return 0 means successed otherwise failed -*/ + */ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, - uint64_t *reset_flags, bool from_hypervisor) { int r; @@ -2677,27 +3155,19 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, /* now we are okay to resume SMC/CP/SDMA */ r = amdgpu_device_ip_reinit_late_sriov(adev); + amdgpu_virt_release_full_gpu(adev, true); if (r) goto error; amdgpu_irq_gpu_reset_resume_helper(adev); r = amdgpu_ib_ring_tests(adev); - if (r) - dev_err(adev->dev, "[GPU_RESET] ib ring test failed (%d).\n", r); + + if (!r && adev->virt.gim_feature & AMDGIM_FEATURE_GIM_FLR_VRAMLOST) { + atomic_inc(&adev->vram_lost_counter); + r = amdgpu_device_handle_vram_lost(adev); + } error: - /* release full control of GPU after ib test */ - amdgpu_virt_release_full_gpu(adev, true); - - if (reset_flags) { - if (adev->virt.gim_feature & AMDGIM_FEATURE_GIM_FLR_VRAMLOST) { - (*reset_flags) |= AMDGPU_RESET_INFO_VRAM_LOST; - atomic_inc(&adev->vram_lost_counter); - } - - /* VF FLR or hotlink reset is always full-reset */ - (*reset_flags) |= AMDGPU_RESET_INFO_FULLRESET; - } return r; } @@ -2716,7 +3186,6 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, struct amdgpu_job *job, bool force) { struct drm_atomic_state *state = NULL; - uint64_t reset_flags = 0; int i, r, resched; if (!force && !amdgpu_device_ip_check_soft_reset(adev)) { @@ -2738,22 +3207,23 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, /* block TTM */ resched = ttm_bo_lock_delayed_workqueue(&adev->mman.bdev); + /* store modesetting */ if (amdgpu_device_has_dc_support(adev)) state = drm_atomic_helper_suspend(adev->ddev); - /* block scheduler */ + /* block all schedulers and reset given job's ring */ for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { struct amdgpu_ring *ring = adev->rings[i]; if (!ring || !ring->sched.thread) continue; - /* only focus on the ring hit timeout if &job not NULL */ + kthread_park(ring->sched.thread); + if (job && job->ring->idx != i) continue; - kthread_park(ring->sched.thread); drm_sched_hw_job_reset(&ring->sched, &job->base); /* after all hw jobs are reset, hw fence is meaningless, so force_completion */ @@ -2761,68 +3231,24 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, } if (amdgpu_sriov_vf(adev)) - r = amdgpu_device_reset_sriov(adev, &reset_flags, job ? false : true); + r = amdgpu_device_reset_sriov(adev, job ? false : true); else - r = amdgpu_device_reset(adev, &reset_flags); + r = amdgpu_device_reset(adev); - if (!r) { - if (((reset_flags & AMDGPU_RESET_INFO_FULLRESET) && !(adev->flags & AMD_IS_APU)) || - (reset_flags & AMDGPU_RESET_INFO_VRAM_LOST)) { - struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring; - struct amdgpu_bo *bo, *tmp; - struct dma_fence *fence = NULL, *next = NULL; + for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { + struct amdgpu_ring *ring = adev->rings[i]; - DRM_INFO("recover vram bo from shadow\n"); - mutex_lock(&adev->shadow_list_lock); - list_for_each_entry_safe(bo, tmp, &adev->shadow_list, shadow_list) { - next = NULL; - amdgpu_device_recover_vram_from_shadow(adev, ring, bo, &next); - if (fence) { - r = dma_fence_wait(fence, false); - if (r) { - WARN(r, "recovery from shadow isn't completed\n"); - break; - } - } + if (!ring || !ring->sched.thread) + continue; - dma_fence_put(fence); - fence = next; - } - mutex_unlock(&adev->shadow_list_lock); - if (fence) { - r = dma_fence_wait(fence, false); - if (r) - WARN(r, "recovery from shadow isn't completed\n"); - } - dma_fence_put(fence); - } - - for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { - struct amdgpu_ring *ring = adev->rings[i]; - - if (!ring || !ring->sched.thread) - continue; - - /* only focus on the ring hit timeout if &job not NULL */ - if (job && job->ring->idx != i) - continue; - + /* only need recovery sched of the given job's ring + * or all rings (in the case @job is NULL) + * after above amdgpu_reset accomplished + */ + if ((!job || job->ring->idx == i) && !r) drm_sched_job_recovery(&ring->sched); - kthread_unpark(ring->sched.thread); - } - } else { - for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { - struct amdgpu_ring *ring = adev->rings[i]; - if (!ring || !ring->sched.thread) - continue; - - /* only focus on the ring hit timeout if &job not NULL */ - if (job && job->ring->idx != i) - continue; - - kthread_unpark(adev->rings[i]->sched.thread); - } + kthread_unpark(ring->sched.thread); } if (amdgpu_device_has_dc_support(adev)) { @@ -2848,7 +3274,16 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, return r; } -void amdgpu_device_get_pcie_info(struct amdgpu_device *adev) +/** + * amdgpu_device_get_pcie_info - fence pcie info about the PCIE slot + * + * @adev: amdgpu_device pointer + * + * Fetchs and stores in the driver the PCIE capabilities (gen speed + * and lanes) of the slot the device is in. Handles APUs and + * virtualized environments where PCIE config space may not be available. + */ +static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev) { u32 mask; int ret;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index 9f2cca1..03614a8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
@@ -37,7 +37,8 @@ #include <drm/drm_edid.h> #include <drm/drm_fb_helper.h> -static void amdgpu_flip_callback(struct dma_fence *f, struct dma_fence_cb *cb) +static void amdgpu_display_flip_callback(struct dma_fence *f, + struct dma_fence_cb *cb) { struct amdgpu_flip_work *work = container_of(cb, struct amdgpu_flip_work, cb); @@ -46,8 +47,8 @@ static void amdgpu_flip_callback(struct dma_fence *f, struct dma_fence_cb *cb) schedule_work(&work->flip_work.work); } -static bool amdgpu_flip_handle_fence(struct amdgpu_flip_work *work, - struct dma_fence **f) +static bool amdgpu_display_flip_handle_fence(struct amdgpu_flip_work *work, + struct dma_fence **f) { struct dma_fence *fence= *f; @@ -56,14 +57,15 @@ static bool amdgpu_flip_handle_fence(struct amdgpu_flip_work *work, *f = NULL; - if (!dma_fence_add_callback(fence, &work->cb, amdgpu_flip_callback)) + if (!dma_fence_add_callback(fence, &work->cb, + amdgpu_display_flip_callback)) return true; dma_fence_put(fence); return false; } -static void amdgpu_flip_work_func(struct work_struct *__work) +static void amdgpu_display_flip_work_func(struct work_struct *__work) { struct delayed_work *delayed_work = container_of(__work, struct delayed_work, work); @@ -77,20 +79,20 @@ static void amdgpu_flip_work_func(struct work_struct *__work) unsigned i; int vpos, hpos; - if (amdgpu_flip_handle_fence(work, &work->excl)) + if (amdgpu_display_flip_handle_fence(work, &work->excl)) return; for (i = 0; i < work->shared_count; ++i) - if (amdgpu_flip_handle_fence(work, &work->shared[i])) + if (amdgpu_display_flip_handle_fence(work, &work->shared[i])) return; /* Wait until we're out of the vertical blank period before the one * targeted by the flip */ if (amdgpu_crtc->enabled && - (amdgpu_get_crtc_scanoutpos(adev->ddev, work->crtc_id, 0, - &vpos, &hpos, NULL, NULL, - &crtc->hwmode) + (amdgpu_display_get_crtc_scanoutpos(adev->ddev, work->crtc_id, 0, + &vpos, &hpos, NULL, NULL, + &crtc->hwmode) & (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK)) == (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK) && (int)(work->target_vblank - @@ -118,7 +120,7 @@ static void amdgpu_flip_work_func(struct work_struct *__work) /* * Handle unpin events outside the interrupt handler proper. */ -static void amdgpu_unpin_work_func(struct work_struct *__work) +static void amdgpu_display_unpin_work_func(struct work_struct *__work) { struct amdgpu_flip_work *work = container_of(__work, struct amdgpu_flip_work, unpin_work); @@ -140,11 +142,11 @@ static void amdgpu_unpin_work_func(struct work_struct *__work) kfree(work); } -int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - struct drm_pending_vblank_event *event, - uint32_t page_flip_flags, uint32_t target, - struct drm_modeset_acquire_ctx *ctx) +int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc, + struct drm_framebuffer *fb, + struct drm_pending_vblank_event *event, + uint32_t page_flip_flags, uint32_t target, + struct drm_modeset_acquire_ctx *ctx) { struct drm_device *dev = crtc->dev; struct amdgpu_device *adev = dev->dev_private; @@ -163,8 +165,8 @@ int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, if (work == NULL) return -ENOMEM; - INIT_DELAYED_WORK(&work->flip_work, amdgpu_flip_work_func); - INIT_WORK(&work->unpin_work, amdgpu_unpin_work_func); + INIT_DELAYED_WORK(&work->flip_work, amdgpu_display_flip_work_func); + INIT_WORK(&work->unpin_work, amdgpu_display_unpin_work_func); work->event = event; work->adev = adev; @@ -208,7 +210,7 @@ int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, amdgpu_bo_unreserve(new_abo); work->base = base; - work->target_vblank = target - drm_crtc_vblank_count(crtc) + + work->target_vblank = target - (uint32_t)drm_crtc_vblank_count(crtc) + amdgpu_get_vblank_counter_kms(dev, work->crtc_id); /* we borrow the event spin lock for protecting flip_wrok */ @@ -229,7 +231,7 @@ int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, /* update crtc fb */ crtc->primary->fb = fb; spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - amdgpu_flip_work_func(&work->flip_work.work); + amdgpu_display_flip_work_func(&work->flip_work.work); return 0; pflip_cleanup: @@ -255,8 +257,8 @@ int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, return r; } -int amdgpu_crtc_set_config(struct drm_mode_set *set, - struct drm_modeset_acquire_ctx *ctx) +int amdgpu_display_crtc_set_config(struct drm_mode_set *set, + struct drm_modeset_acquire_ctx *ctx) { struct drm_device *dev; struct amdgpu_device *adev; @@ -353,7 +355,7 @@ static const char *hpd_names[6] = { "HPD6", }; -void amdgpu_print_display_setup(struct drm_device *dev) +void amdgpu_display_print_display_setup(struct drm_device *dev) { struct drm_connector *connector; struct amdgpu_connector *amdgpu_connector; @@ -430,11 +432,11 @@ void amdgpu_print_display_setup(struct drm_device *dev) } /** - * amdgpu_ddc_probe + * amdgpu_display_ddc_probe * */ -bool amdgpu_ddc_probe(struct amdgpu_connector *amdgpu_connector, - bool use_aux) +bool amdgpu_display_ddc_probe(struct amdgpu_connector *amdgpu_connector, + bool use_aux) { u8 out = 0x0; u8 buf[8]; @@ -480,7 +482,7 @@ bool amdgpu_ddc_probe(struct amdgpu_connector *amdgpu_connector, return true; } -static void amdgpu_user_framebuffer_destroy(struct drm_framebuffer *fb) +static void amdgpu_display_user_framebuffer_destroy(struct drm_framebuffer *fb) { struct amdgpu_framebuffer *amdgpu_fb = to_amdgpu_framebuffer(fb); @@ -489,9 +491,10 @@ static void amdgpu_user_framebuffer_destroy(struct drm_framebuffer *fb) kfree(amdgpu_fb); } -static int amdgpu_user_framebuffer_create_handle(struct drm_framebuffer *fb, - struct drm_file *file_priv, - unsigned int *handle) +static int amdgpu_display_user_framebuffer_create_handle( + struct drm_framebuffer *fb, + struct drm_file *file_priv, + unsigned int *handle) { struct amdgpu_framebuffer *amdgpu_fb = to_amdgpu_framebuffer(fb); @@ -499,8 +502,8 @@ static int amdgpu_user_framebuffer_create_handle(struct drm_framebuffer *fb, } static const struct drm_framebuffer_funcs amdgpu_fb_funcs = { - .destroy = amdgpu_user_framebuffer_destroy, - .create_handle = amdgpu_user_framebuffer_create_handle, + .destroy = amdgpu_display_user_framebuffer_destroy, + .create_handle = amdgpu_display_user_framebuffer_create_handle, }; uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev) @@ -517,11 +520,10 @@ uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev) return domain; } -int -amdgpu_framebuffer_init(struct drm_device *dev, - struct amdgpu_framebuffer *rfb, - const struct drm_mode_fb_cmd2 *mode_cmd, - struct drm_gem_object *obj) +int amdgpu_display_framebuffer_init(struct drm_device *dev, + struct amdgpu_framebuffer *rfb, + const struct drm_mode_fb_cmd2 *mode_cmd, + struct drm_gem_object *obj) { int ret; rfb->obj = obj; @@ -535,9 +537,9 @@ amdgpu_framebuffer_init(struct drm_device *dev, } struct drm_framebuffer * -amdgpu_user_framebuffer_create(struct drm_device *dev, - struct drm_file *file_priv, - const struct drm_mode_fb_cmd2 *mode_cmd) +amdgpu_display_user_framebuffer_create(struct drm_device *dev, + struct drm_file *file_priv, + const struct drm_mode_fb_cmd2 *mode_cmd) { struct drm_gem_object *obj; struct amdgpu_framebuffer *amdgpu_fb; @@ -562,7 +564,7 @@ amdgpu_user_framebuffer_create(struct drm_device *dev, return ERR_PTR(-ENOMEM); } - ret = amdgpu_framebuffer_init(dev, amdgpu_fb, mode_cmd, obj); + ret = amdgpu_display_framebuffer_init(dev, amdgpu_fb, mode_cmd, obj); if (ret) { kfree(amdgpu_fb); drm_gem_object_put_unlocked(obj); @@ -573,7 +575,7 @@ amdgpu_user_framebuffer_create(struct drm_device *dev, } const struct drm_mode_config_funcs amdgpu_mode_funcs = { - .fb_create = amdgpu_user_framebuffer_create, + .fb_create = amdgpu_display_user_framebuffer_create, .output_poll_changed = drm_fb_helper_output_poll_changed, }; @@ -595,7 +597,7 @@ static const struct drm_prop_enum_list amdgpu_dither_enum_list[] = { AMDGPU_FMT_DITHER_ENABLE, "on" }, }; -int amdgpu_modeset_create_props(struct amdgpu_device *adev) +int amdgpu_display_modeset_create_props(struct amdgpu_device *adev) { int sz; @@ -644,7 +646,7 @@ int amdgpu_modeset_create_props(struct amdgpu_device *adev) return 0; } -void amdgpu_update_display_priority(struct amdgpu_device *adev) +void amdgpu_display_update_priority(struct amdgpu_device *adev) { /* adjustment options for the display watermarks */ if ((amdgpu_disp_priority == 0) || (amdgpu_disp_priority > 2)) @@ -654,7 +656,7 @@ void amdgpu_update_display_priority(struct amdgpu_device *adev) } -static bool is_hdtv_mode(const struct drm_display_mode *mode) +static bool amdgpu_display_is_hdtv_mode(const struct drm_display_mode *mode) { /* try and guess if this is a tv or a monitor */ if ((mode->vdisplay == 480 && mode->hdisplay == 720) || /* 480p */ @@ -666,9 +668,9 @@ static bool is_hdtv_mode(const struct drm_display_mode *mode) return false; } -bool amdgpu_crtc_scaling_mode_fixup(struct drm_crtc *crtc, - const struct drm_display_mode *mode, - struct drm_display_mode *adjusted_mode) +bool amdgpu_display_crtc_scaling_mode_fixup(struct drm_crtc *crtc, + const struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode) { struct drm_device *dev = crtc->dev; struct drm_encoder *encoder; @@ -711,7 +713,7 @@ bool amdgpu_crtc_scaling_mode_fixup(struct drm_crtc *crtc, ((amdgpu_encoder->underscan_type == UNDERSCAN_ON) || ((amdgpu_encoder->underscan_type == UNDERSCAN_AUTO) && drm_detect_hdmi_monitor(amdgpu_connector_edid(connector)) && - is_hdtv_mode(mode)))) { + amdgpu_display_is_hdtv_mode(mode)))) { if (amdgpu_encoder->underscan_hborder != 0) amdgpu_crtc->h_border = amdgpu_encoder->underscan_hborder; else @@ -779,10 +781,10 @@ bool amdgpu_crtc_scaling_mode_fixup(struct drm_crtc *crtc, * unknown small number of scanlines wrt. real scanout position. * */ -int amdgpu_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, - unsigned int flags, int *vpos, int *hpos, - ktime_t *stime, ktime_t *etime, - const struct drm_display_mode *mode) +int amdgpu_display_get_crtc_scanoutpos(struct drm_device *dev, + unsigned int pipe, unsigned int flags, int *vpos, + int *hpos, ktime_t *stime, ktime_t *etime, + const struct drm_display_mode *mode) { u32 vbl = 0, position = 0; int vbl_start, vbl_end, vtotal, ret = 0; @@ -874,7 +876,7 @@ int amdgpu_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, return ret; } -int amdgpu_crtc_idx_to_irq_type(struct amdgpu_device *adev, int crtc) +int amdgpu_display_crtc_idx_to_irq_type(struct amdgpu_device *adev, int crtc) { if (crtc < 0 || crtc >= adev->mode_info.num_crtc) return AMDGPU_CRTC_IRQ_NONE;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h index 6f93668d..f66e3e3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h
@@ -25,8 +25,8 @@ uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev); struct drm_framebuffer * -amdgpu_user_framebuffer_create(struct drm_device *dev, - struct drm_file *file_priv, - const struct drm_mode_fb_cmd2 *mode_cmd); +amdgpu_display_user_framebuffer_create(struct drm_device *dev, + struct drm_file *file_priv, + const struct drm_mode_fb_cmd2 *mode_cmd); #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h index 1e08f75..74ef2a9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -265,9 +265,6 @@ enum amdgpu_pcie_gen { #define amdgpu_dpm_read_sensor(adev, idx, value, size) \ ((adev)->powerplay.pp_funcs->read_sensor((adev)->powerplay.pp_handle, (idx), (value), (size))) -#define amdgpu_dpm_get_temperature(adev) \ - ((adev)->powerplay.pp_funcs->get_temperature((adev)->powerplay.pp_handle)) - #define amdgpu_dpm_set_fan_control_mode(adev, m) \ ((adev)->powerplay.pp_funcs->set_fan_control_mode((adev)->powerplay.pp_handle, (m))) @@ -322,8 +319,8 @@ enum amdgpu_pcie_gen { #define amdgpu_dpm_set_mclk_od(adev, value) \ ((adev)->powerplay.pp_funcs->set_mclk_od((adev)->powerplay.pp_handle, value)) -#define amdgpu_dpm_dispatch_task(adev, task_id, input, output) \ - ((adev)->powerplay.pp_funcs->dispatch_tasks)((adev)->powerplay.pp_handle, (task_id), (input), (output)) +#define amdgpu_dpm_dispatch_task(adev, task_id, user_state) \ + ((adev)->powerplay.pp_funcs->dispatch_tasks)((adev)->powerplay.pp_handle, (task_id), (user_state)) #define amdgpu_dpm_check_state_equal(adev, cps, rps, equal) \ ((adev)->powerplay.pp_funcs->check_state_equal((adev)->powerplay.pp_handle, (cps), (rps), (equal))) @@ -338,17 +335,9 @@ enum amdgpu_pcie_gen { ((adev)->powerplay.pp_funcs->reset_power_profile_state(\ (adev)->powerplay.pp_handle, request)) -#define amdgpu_dpm_get_power_profile_state(adev, query) \ - ((adev)->powerplay.pp_funcs->get_power_profile_state(\ - (adev)->powerplay.pp_handle, query)) - -#define amdgpu_dpm_set_power_profile_state(adev, request) \ - ((adev)->powerplay.pp_funcs->set_power_profile_state(\ - (adev)->powerplay.pp_handle, request)) - -#define amdgpu_dpm_switch_power_profile(adev, type) \ +#define amdgpu_dpm_switch_power_profile(adev, type, en) \ ((adev)->powerplay.pp_funcs->switch_power_profile(\ - (adev)->powerplay.pp_handle, type)) + (adev)->powerplay.pp_handle, type, en)) #define amdgpu_dpm_set_clockgating_by_smu(adev, msg_id) \ ((adev)->powerplay.pp_funcs->set_clockgating_by_smu(\ @@ -360,6 +349,18 @@ enum amdgpu_pcie_gen { (adev)->powerplay.pp_handle, virtual_addr_low, \ virtual_addr_hi, mc_addr_low, mc_addr_hi, size) +#define amdgpu_dpm_get_power_profile_mode(adev, buf) \ + ((adev)->powerplay.pp_funcs->get_power_profile_mode(\ + (adev)->powerplay.pp_handle, buf)) + +#define amdgpu_dpm_set_power_profile_mode(adev, parameter, size) \ + ((adev)->powerplay.pp_funcs->set_power_profile_mode(\ + (adev)->powerplay.pp_handle, parameter, size)) + +#define amdgpu_dpm_odn_edit_dpm_table(adev, type, parameter, size) \ + ((adev)->powerplay.pp_funcs->odn_edit_dpm_table(\ + (adev)->powerplay.pp_handle, type, parameter, size)) + #define amdgpu_dpm_set_powergating_by_smu(adev, block_type, gate) \ ((adev)->powerplay.pp_funcs->set_powergating_by_smu(\ (adev)->powerplay.pp_handle, block_type, gate))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 5cd887f..79d9261 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -73,9 +73,11 @@ * - 3.21.0 - Add DRM_AMDGPU_FENCE_TO_HANDLE ioctl * - 3.22.0 - Add DRM_AMDGPU_SCHED ioctl * - 3.23.0 - Add query for VRAM lost counter + * - 3.24.0 - Add high priority compute support for gfx9 + * - 3.25.0 - Add support for sensor query info (stable pstate sclk/mclk). */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 23 +#define KMS_DRIVER_MINOR 25 #define KMS_DRIVER_PATCHLEVEL 0 int amdgpu_vram_limit = 0; @@ -119,7 +121,7 @@ uint amdgpu_pg_mask = 0xffffffff; uint amdgpu_sdma_phase_quantum = 32; char *amdgpu_disable_cu = NULL; char *amdgpu_virtual_display = NULL; -uint amdgpu_pp_feature_mask = 0xffffffff; +uint amdgpu_pp_feature_mask = 0xffffbfff; int amdgpu_ngg = 0; int amdgpu_prim_buf_per_se = 0; int amdgpu_pos_buf_per_se = 0; @@ -129,6 +131,7 @@ int amdgpu_job_hang_limit = 0; int amdgpu_lbpw = -1; int amdgpu_compute_multipipe = -1; int amdgpu_gpu_recovery = -1; /* auto */ +int amdgpu_emu_mode = 0; MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes"); module_param_named(vramlimit, amdgpu_vram_limit, int, 0600); @@ -281,9 +284,12 @@ module_param_named(lbpw, amdgpu_lbpw, int, 0444); MODULE_PARM_DESC(compute_multipipe, "Force compute queues to be spread across pipes (1 = enable, 0 = disable, -1 = auto)"); module_param_named(compute_multipipe, amdgpu_compute_multipipe, int, 0444); -MODULE_PARM_DESC(gpu_recovery, "Enable GPU recovery mechanism, (1 = enable, 0 = disable, -1 = auto"); +MODULE_PARM_DESC(gpu_recovery, "Enable GPU recovery mechanism, (1 = enable, 0 = disable, -1 = auto)"); module_param_named(gpu_recovery, amdgpu_gpu_recovery, int, 0444); +MODULE_PARM_DESC(emu_mode, "Emulation mode, (1 = enable, 0 = disable)"); +module_param_named(emu_mode, amdgpu_emu_mode, int, 0444); + #ifdef CONFIG_DRM_AMDGPU_SI #if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE) @@ -539,6 +545,12 @@ static const struct pci_device_id pciidlist[] = { {0x1002, 0x6868, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10}, {0x1002, 0x686c, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10}, {0x1002, 0x687f, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10}, + /* Vega 12 */ + {0x1002, 0x69A0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12}, + {0x1002, 0x69A1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12}, + {0x1002, 0x69A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12}, + {0x1002, 0x69A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12}, + {0x1002, 0x69AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12}, /* Raven */ {0x1002, 0x15dd, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RAVEN|AMD_IS_APU}, @@ -732,7 +744,6 @@ static int amdgpu_pmops_runtime_suspend(struct device *dev) drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING; drm_kms_helper_poll_disable(drm_dev); - vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_OFF); ret = amdgpu_device_suspend(drm_dev, false, false); pci_save_state(pdev); @@ -769,7 +780,6 @@ static int amdgpu_pmops_runtime_resume(struct device *dev) ret = amdgpu_device_resume(drm_dev, false, false); drm_kms_helper_poll_enable(drm_dev); - vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_ON); drm_dev->switch_power_state = DRM_SWITCH_POWER_ON; return 0; } @@ -847,8 +857,8 @@ amdgpu_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe, ktime_t *stime, ktime_t *etime, const struct drm_display_mode *mode) { - return amdgpu_get_crtc_scanoutpos(dev, pipe, 0, vpos, hpos, - stime, etime, mode); + return amdgpu_display_get_crtc_scanoutpos(dev, pipe, 0, vpos, hpos, + stime, etime, mode); } static struct drm_driver kms_driver = { @@ -866,9 +876,6 @@ static struct drm_driver kms_driver = { .disable_vblank = amdgpu_disable_vblank_kms, .get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos, .get_scanout_position = amdgpu_get_crtc_scanout_position, - .irq_preinstall = amdgpu_irq_preinstall, - .irq_postinstall = amdgpu_irq_postinstall, - .irq_uninstall = amdgpu_irq_uninstall, .irq_handler = amdgpu_irq_handler, .ioctls = amdgpu_ioctls_kms, .gem_free_object_unlocked = amdgpu_gem_object_free, @@ -882,8 +889,6 @@ static struct drm_driver kms_driver = { .prime_fd_to_handle = drm_gem_prime_fd_to_handle, .gem_prime_export = amdgpu_gem_prime_export, .gem_prime_import = amdgpu_gem_prime_import, - .gem_prime_pin = amdgpu_gem_prime_pin, - .gem_prime_unpin = amdgpu_gem_prime_unpin, .gem_prime_res_obj = amdgpu_gem_prime_res_obj, .gem_prime_get_sg_table = amdgpu_gem_prime_get_sg_table, .gem_prime_import_sg_table = amdgpu_gem_prime_import_sg_table, @@ -917,6 +922,11 @@ static int __init amdgpu_init(void) { int r; + if (vgacon_text_force()) { + DRM_ERROR("VGACON disables amdgpu kernel modesetting.\n"); + return -EINVAL; + } + r = amdgpu_sync_init(); if (r) goto error_sync; @@ -925,10 +935,6 @@ static int __init amdgpu_init(void) if (r) goto error_fence; - if (vgacon_text_force()) { - DRM_ERROR("VGACON disables amdgpu kernel modesetting.\n"); - return -EINVAL; - } DRM_INFO("amdgpu kernel modesetting enabled.\n"); driver = &kms_driver; pdriver = &amdgpu_kms_pci_driver;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c index 2baa2ce..f9d1ca2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
@@ -227,7 +227,8 @@ static int amdgpufb_create(struct drm_fb_helper *helper, info->par = rfbdev; info->skip_vt_switch = true; - ret = amdgpu_framebuffer_init(adev->ddev, &rfbdev->rfb, &mode_cmd, gobj); + ret = amdgpu_display_framebuffer_init(adev->ddev, &rfbdev->rfb, + &mode_cmd, gobj); if (ret) { DRM_ERROR("failed to initialize framebuffer %d\n", ret); goto out; @@ -244,8 +245,8 @@ static int amdgpufb_create(struct drm_fb_helper *helper, info->fbops = &amdgpufb_ops; - tmp = amdgpu_bo_gpu_offset(abo) - adev->mc.vram_start; - info->fix.smem_start = adev->mc.aper_base + tmp; + tmp = amdgpu_bo_gpu_offset(abo) - adev->gmc.vram_start; + info->fix.smem_start = adev->gmc.aper_base + tmp; info->fix.smem_len = amdgpu_bo_size(abo); info->screen_base = amdgpu_bo_kptr(abo); info->screen_size = amdgpu_bo_size(abo); @@ -254,7 +255,7 @@ static int amdgpufb_create(struct drm_fb_helper *helper, /* setup aperture base/size for vesafb takeover */ info->apertures->ranges[0].base = adev->ddev->mode_config.fb_base; - info->apertures->ranges[0].size = adev->mc.aper_size; + info->apertures->ranges[0].size = adev->gmc.aper_size; /* Use default scratch pixmap (info->pixmap.flags = FB_PIXMAP_SYSTEM) */ @@ -264,7 +265,7 @@ static int amdgpufb_create(struct drm_fb_helper *helper, } DRM_INFO("fb mappable at 0x%lX\n", info->fix.smem_start); - DRM_INFO("vram apper at 0x%lX\n", (unsigned long)adev->mc.aper_base); + DRM_INFO("vram apper at 0x%lX\n", (unsigned long)adev->gmc.aper_base); DRM_INFO("size %lu\n", (unsigned long)amdgpu_bo_size(abo)); DRM_INFO("fb depth is %d\n", fb->format->depth); DRM_INFO(" pitch is %d\n", fb->pitches[0]); @@ -321,7 +322,7 @@ int amdgpu_fbdev_init(struct amdgpu_device *adev) return 0; /* select 8 bpp console on low vram cards */ - if (adev->mc.real_vram_size <= (32*1024*1024)) + if (adev->gmc.real_vram_size <= (32*1024*1024)) bpp_sel = 8; rfbdev = kzalloc(sizeof(struct amdgpu_fbdev), GFP_KERNEL);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index 9e2c225..52a2368d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -409,6 +409,7 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring, int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring, unsigned num_hw_submission) { + long timeout; int r; /* Check that num_hw_submission is a power of two */ @@ -433,9 +434,16 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring, /* No need to setup the GPU scheduler for KIQ ring */ if (ring->funcs->type != AMDGPU_RING_TYPE_KIQ) { + /* for non-sriov case, no timeout enforce on compute ring */ + if ((ring->funcs->type == AMDGPU_RING_TYPE_COMPUTE) + && !amdgpu_sriov_vf(ring->adev)) + timeout = MAX_SCHEDULE_TIMEOUT; + else + timeout = msecs_to_jiffies(amdgpu_lockup_timeout); + r = drm_sched_init(&ring->sched, &amdgpu_sched_ops, num_hw_submission, amdgpu_job_hang_limit, - msecs_to_jiffies(amdgpu_lockup_timeout), ring->name); + timeout, ring->name); if (r) { DRM_ERROR("Failed to create scheduler on ring %s.\n", ring->name);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c index 0a4f34a..cf0f186 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
@@ -68,17 +68,15 @@ */ static int amdgpu_gart_dummy_page_init(struct amdgpu_device *adev) { - if (adev->dummy_page.page) + struct page *dummy_page = adev->mman.bdev.glob->dummy_read_page; + + if (adev->dummy_page_addr) return 0; - adev->dummy_page.page = alloc_page(GFP_DMA32 | GFP_KERNEL | __GFP_ZERO); - if (adev->dummy_page.page == NULL) - return -ENOMEM; - adev->dummy_page.addr = pci_map_page(adev->pdev, adev->dummy_page.page, - 0, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); - if (pci_dma_mapping_error(adev->pdev, adev->dummy_page.addr)) { + adev->dummy_page_addr = pci_map_page(adev->pdev, dummy_page, 0, + PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); + if (pci_dma_mapping_error(adev->pdev, adev->dummy_page_addr)) { dev_err(&adev->pdev->dev, "Failed to DMA MAP the dummy page\n"); - __free_page(adev->dummy_page.page); - adev->dummy_page.page = NULL; + adev->dummy_page_addr = 0; return -ENOMEM; } return 0; @@ -93,12 +91,11 @@ static int amdgpu_gart_dummy_page_init(struct amdgpu_device *adev) */ static void amdgpu_gart_dummy_page_fini(struct amdgpu_device *adev) { - if (adev->dummy_page.page == NULL) + if (!adev->dummy_page_addr) return; - pci_unmap_page(adev->pdev, adev->dummy_page.addr, - PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); - __free_page(adev->dummy_page.page); - adev->dummy_page.page = NULL; + pci_unmap_page(adev->pdev, adev->dummy_page_addr, + PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); + adev->dummy_page_addr = 0; } /** @@ -116,11 +113,12 @@ int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev) int r; if (adev->gart.robj == NULL) { - r = amdgpu_bo_create(adev, adev->gart.table_size, - PAGE_SIZE, true, AMDGPU_GEM_DOMAIN_VRAM, + r = amdgpu_bo_create(adev, adev->gart.table_size, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, - NULL, NULL, 0, &adev->gart.robj); + ttm_bo_type_kernel, NULL, + &adev->gart.robj); if (r) { return r; } @@ -236,18 +234,19 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset, #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS adev->gart.pages[p] = NULL; #endif - page_base = adev->dummy_page.addr; + page_base = adev->dummy_page_addr; if (!adev->gart.ptr) continue; for (j = 0; j < (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); j++, t++) { - amdgpu_gart_set_pte_pde(adev, adev->gart.ptr, - t, page_base, flags); + amdgpu_gmc_set_pte_pde(adev, adev->gart.ptr, + t, page_base, flags); page_base += AMDGPU_GPU_PAGE_SIZE; } } mb(); - amdgpu_gart_flush_gpu_tlb(adev, 0); + amdgpu_asic_flush_hdp(adev, NULL); + amdgpu_gmc_flush_gpu_tlb(adev, 0); return 0; } @@ -279,7 +278,7 @@ int amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset, for (i = 0; i < pages; i++) { page_base = dma_addr[i]; for (j = 0; j < (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); j++, t++) { - amdgpu_gart_set_pte_pde(adev, dst, t, page_base, flags); + amdgpu_gmc_set_pte_pde(adev, dst, t, page_base, flags); page_base += AMDGPU_GPU_PAGE_SIZE; } } @@ -317,7 +316,7 @@ int amdgpu_gart_bind(struct amdgpu_device *adev, uint64_t offset, t = offset / AMDGPU_GPU_PAGE_SIZE; p = t / (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); for (i = 0; i < pages; i++, p++) - adev->gart.pages[p] = pagelist[i]; + adev->gart.pages[p] = pagelist ? pagelist[i] : NULL; #endif if (!adev->gart.ptr) @@ -329,7 +328,8 @@ int amdgpu_gart_bind(struct amdgpu_device *adev, uint64_t offset, return r; mb(); - amdgpu_gart_flush_gpu_tlb(adev, 0); + amdgpu_asic_flush_hdp(adev, NULL); + amdgpu_gmc_flush_gpu_tlb(adev, 0); return 0; } @@ -345,7 +345,7 @@ int amdgpu_gart_init(struct amdgpu_device *adev) { int r; - if (adev->dummy_page.page) + if (adev->dummy_page_addr) return 0; /* We need PAGE_SIZE >= AMDGPU_GPU_PAGE_SIZE */ @@ -357,8 +357,8 @@ int amdgpu_gart_init(struct amdgpu_device *adev) if (r) return r; /* Compute table size */ - adev->gart.num_cpu_pages = adev->mc.gart_size / PAGE_SIZE; - adev->gart.num_gpu_pages = adev->mc.gart_size / AMDGPU_GPU_PAGE_SIZE; + adev->gart.num_cpu_pages = adev->gmc.gart_size / PAGE_SIZE; + adev->gart.num_gpu_pages = adev->gmc.gart_size / AMDGPU_GPU_PAGE_SIZE; DRM_INFO("GART: num cpu pages %u, num gpu pages %u\n", adev->gart.num_cpu_pages, adev->gart.num_gpu_pages);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h index d4a4330..456295c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h
@@ -31,7 +31,6 @@ */ struct amdgpu_device; struct amdgpu_bo; -struct amdgpu_gart_funcs; #define AMDGPU_GPU_PAGE_SIZE 4096 #define AMDGPU_GPU_PAGE_MASK (AMDGPU_GPU_PAGE_SIZE - 1) @@ -52,8 +51,6 @@ struct amdgpu_gart { /* Asic default pte flags */ uint64_t gart_pte_flags; - - const struct amdgpu_gart_funcs *gart_funcs; }; int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 239a868..92266d8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -44,7 +44,7 @@ void amdgpu_gem_object_free(struct drm_gem_object *gobj) int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size, int alignment, u32 initial_domain, - u64 flags, bool kernel, + u64 flags, enum ttm_bo_type type, struct reservation_object *resv, struct drm_gem_object **obj) { @@ -58,8 +58,8 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size, } retry: - r = amdgpu_bo_create(adev, size, alignment, kernel, initial_domain, - flags, NULL, resv, 0, &bo); + r = amdgpu_bo_create(adev, size, alignment, initial_domain, + flags, type, resv, &bo); if (r) { if (r != -ERESTARTSYS) { if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) { @@ -522,12 +522,13 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, goto error; if (operation == AMDGPU_VA_OP_MAP || - operation == AMDGPU_VA_OP_REPLACE) + operation == AMDGPU_VA_OP_REPLACE) { r = amdgpu_vm_bo_update(adev, bo_va, false); + if (r) + goto error; + } r = amdgpu_vm_update_directories(adev, vm); - if (r) - goto error; error: if (r && r != -ERESTARTSYS) @@ -633,7 +634,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, if (r) goto error_backoff; - va_flags = amdgpu_vm_get_pte_flags(adev, args->flags); + va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags); r = amdgpu_vm_bo_map(adev, bo_va, args->va_address, args->offset_in_bo, args->map_size, va_flags); @@ -653,7 +654,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, if (r) goto error_backoff; - va_flags = amdgpu_vm_get_pte_flags(adev, args->flags); + va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags); r = amdgpu_vm_bo_replace_map(adev, bo_va, args->va_address, args->offset_in_bo, args->map_size, va_flags);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h new file mode 100644 index 0000000..893c249 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
@@ -0,0 +1,112 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * All Rights Reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the + * "Software"), to deal in the Software without restriction, including + * without limitation the rights to use, copy, modify, merge, publish, + * distribute, sub license, and/or sell copies of the Software, and to + * permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, + * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR + * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE + * USE OR OTHER DEALINGS IN THE SOFTWARE. + * + * The above copyright notice and this permission notice (including the + * next paragraph) shall be included in all copies or substantial portions + * of the Software. + * + */ +#ifndef __AMDGPU_GMC_H__ +#define __AMDGPU_GMC_H__ + +#include <linux/types.h> + +#include "amdgpu_irq.h" + +struct firmware; + +/* + * VMHUB structures, functions & helpers + */ +struct amdgpu_vmhub { + uint32_t ctx0_ptb_addr_lo32; + uint32_t ctx0_ptb_addr_hi32; + uint32_t vm_inv_eng0_req; + uint32_t vm_inv_eng0_ack; + uint32_t vm_context0_cntl; + uint32_t vm_l2_pro_fault_status; + uint32_t vm_l2_pro_fault_cntl; +}; + +/* + * GPU MC structures, functions & helpers + */ +struct amdgpu_gmc_funcs { + /* flush the vm tlb via mmio */ + void (*flush_gpu_tlb)(struct amdgpu_device *adev, + uint32_t vmid); + /* flush the vm tlb via ring */ + uint64_t (*emit_flush_gpu_tlb)(struct amdgpu_ring *ring, unsigned vmid, + uint64_t pd_addr); + /* Change the VMID -> PASID mapping */ + void (*emit_pasid_mapping)(struct amdgpu_ring *ring, unsigned vmid, + unsigned pasid); + /* write pte/pde updates using the cpu */ + int (*set_pte_pde)(struct amdgpu_device *adev, + void *cpu_pt_addr, /* cpu addr of page table */ + uint32_t gpu_page_idx, /* pte/pde to update */ + uint64_t addr, /* addr to write into pte/pde */ + uint64_t flags); /* access flags */ + /* enable/disable PRT support */ + void (*set_prt)(struct amdgpu_device *adev, bool enable); + /* set pte flags based per asic */ + uint64_t (*get_vm_pte_flags)(struct amdgpu_device *adev, + uint32_t flags); + /* get the pde for a given mc addr */ + void (*get_vm_pde)(struct amdgpu_device *adev, int level, + u64 *dst, u64 *flags); +}; + +struct amdgpu_gmc { + resource_size_t aper_size; + resource_size_t aper_base; + /* for some chips with <= 32MB we need to lie + * about vram size near mc fb location */ + u64 mc_vram_size; + u64 visible_vram_size; + u64 gart_size; + u64 gart_start; + u64 gart_end; + u64 vram_start; + u64 vram_end; + unsigned vram_width; + u64 real_vram_size; + int vram_mtrr; + u64 mc_mask; + const struct firmware *fw; /* MC firmware */ + uint32_t fw_version; + struct amdgpu_irq_src vm_fault; + uint32_t vram_type; + uint32_t srbm_soft_reset; + bool prt_warning; + uint64_t stolen_size; + /* apertures */ + u64 shared_aperture_start; + u64 shared_aperture_end; + u64 private_aperture_start; + u64 private_aperture_end; + /* protects concurrent invalidation */ + spinlock_t invalidate_lock; + bool translate_further; + + const struct amdgpu_gmc_funcs *gmc_funcs; +}; + +#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c index 7c2be32..da7b1b9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
@@ -56,7 +56,7 @@ static int amdgpu_gtt_mgr_init(struct ttm_mem_type_manager *man, return -ENOMEM; start = AMDGPU_GTT_MAX_TRANSFER_SIZE * AMDGPU_GTT_NUM_TRANSFER_WINDOWS; - size = (adev->mc.gart_size >> PAGE_SHIFT) - start; + size = (adev->gmc.gart_size >> PAGE_SHIFT) - start; drm_mm_init(&mgr->mm, start, size); spin_lock_init(&mgr->lock); atomic64_set(&mgr->available, p_size);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index c2ca70a..5cf08c4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -184,12 +184,15 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, if (job && ring->funcs->init_cond_exec) patch_offset = amdgpu_ring_init_cond_exec(ring); - if (ring->funcs->emit_hdp_flush #ifdef CONFIG_X86_64 - && !(adev->flags & AMD_IS_APU) + if (!(adev->flags & AMD_IS_APU)) #endif - ) - amdgpu_ring_emit_hdp_flush(ring); + { + if (ring->funcs->emit_hdp_flush) + amdgpu_ring_emit_hdp_flush(ring); + else + amdgpu_asic_flush_hdp(adev, ring); + } skip_preamble = ring->current_ctx == fence_ctx; need_ctx_switch = ring->current_ctx != fence_ctx; @@ -219,12 +222,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, if (ring->funcs->emit_tmz) amdgpu_ring_emit_tmz(ring, false); - if (ring->funcs->emit_hdp_invalidate #ifdef CONFIG_X86_64 - && !(adev->flags & AMD_IS_APU) + if (!(adev->flags & AMD_IS_APU)) #endif - ) - amdgpu_ring_emit_hdp_invalidate(ring); + amdgpu_asic_invalidate_hdp(adev, ring); r = amdgpu_fence_emit(ring, f); if (r) { @@ -278,11 +279,6 @@ int amdgpu_ib_pool_init(struct amdgpu_device *adev) return r; } - r = amdgpu_sa_bo_manager_start(adev, &adev->ring_tmp_bo); - if (r) { - return r; - } - adev->ib_pool_ready = true; if (amdgpu_debugfs_sa_init(adev)) { dev_err(adev->dev, "failed to register debugfs file for SA\n"); @@ -301,7 +297,6 @@ int amdgpu_ib_pool_init(struct amdgpu_device *adev) void amdgpu_ib_pool_fini(struct amdgpu_device *adev) { if (adev->ib_pool_ready) { - amdgpu_sa_bo_manager_suspend(adev, &adev->ring_tmp_bo); amdgpu_sa_bo_manager_fini(adev, &adev->ring_tmp_bo); adev->ib_pool_ready = false; }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c index 16884a0..a1c78f9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
@@ -40,6 +40,12 @@ */ static DEFINE_IDA(amdgpu_pasid_ida); +/* Helper to free pasid from a fence callback */ +struct amdgpu_pasid_cb { + struct dma_fence_cb cb; + unsigned int pasid; +}; + /** * amdgpu_pasid_alloc - Allocate a PASID * @bits: Maximum width of the PASID in bits, must be at least 1 @@ -63,6 +69,9 @@ int amdgpu_pasid_alloc(unsigned int bits) break; } + if (pasid >= 0) + trace_amdgpu_pasid_allocated(pasid); + return pasid; } @@ -72,9 +81,86 @@ int amdgpu_pasid_alloc(unsigned int bits) */ void amdgpu_pasid_free(unsigned int pasid) { + trace_amdgpu_pasid_freed(pasid); ida_simple_remove(&amdgpu_pasid_ida, pasid); } +static void amdgpu_pasid_free_cb(struct dma_fence *fence, + struct dma_fence_cb *_cb) +{ + struct amdgpu_pasid_cb *cb = + container_of(_cb, struct amdgpu_pasid_cb, cb); + + amdgpu_pasid_free(cb->pasid); + dma_fence_put(fence); + kfree(cb); +} + +/** + * amdgpu_pasid_free_delayed - free pasid when fences signal + * + * @resv: reservation object with the fences to wait for + * @pasid: pasid to free + * + * Free the pasid only after all the fences in resv are signaled. + */ +void amdgpu_pasid_free_delayed(struct reservation_object *resv, + unsigned int pasid) +{ + struct dma_fence *fence, **fences; + struct amdgpu_pasid_cb *cb; + unsigned count; + int r; + + r = reservation_object_get_fences_rcu(resv, NULL, &count, &fences); + if (r) + goto fallback; + + if (count == 0) { + amdgpu_pasid_free(pasid); + return; + } + + if (count == 1) { + fence = fences[0]; + kfree(fences); + } else { + uint64_t context = dma_fence_context_alloc(1); + struct dma_fence_array *array; + + array = dma_fence_array_create(count, fences, context, + 1, false); + if (!array) { + kfree(fences); + goto fallback; + } + fence = &array->base; + } + + cb = kmalloc(sizeof(*cb), GFP_KERNEL); + if (!cb) { + /* Last resort when we are OOM */ + dma_fence_wait(fence, false); + dma_fence_put(fence); + amdgpu_pasid_free(pasid); + } else { + cb->pasid = pasid; + if (dma_fence_add_callback(fence, &cb->cb, + amdgpu_pasid_free_cb)) + amdgpu_pasid_free_cb(fence, &cb->cb); + } + + return; + +fallback: + /* Not enough memory for the delayed delete, as last resort + * block for all the fences to complete. + */ + reservation_object_wait_timeout_rcu(resv, true, false, + MAX_SCHEDULE_TIMEOUT); + amdgpu_pasid_free(pasid); +} + /* * VMID manager * @@ -96,64 +182,210 @@ bool amdgpu_vmid_had_gpu_reset(struct amdgpu_device *adev, atomic_read(&adev->gpu_reset_counter); } -/* idr_mgr->lock must be held */ -static int amdgpu_vmid_grab_reserved_locked(struct amdgpu_vm *vm, - struct amdgpu_ring *ring, - struct amdgpu_sync *sync, - struct dma_fence *fence, - struct amdgpu_job *job) +/** + * amdgpu_vm_grab_idle - grab idle VMID + * + * @vm: vm to allocate id for + * @ring: ring we want to submit job to + * @sync: sync object where we add dependencies + * @idle: resulting idle VMID + * + * Try to find an idle VMID, if none is idle add a fence to wait to the sync + * object. Returns -ENOMEM when we are out of memory. + */ +static int amdgpu_vmid_grab_idle(struct amdgpu_vm *vm, + struct amdgpu_ring *ring, + struct amdgpu_sync *sync, + struct amdgpu_vmid **idle) +{ + struct amdgpu_device *adev = ring->adev; + unsigned vmhub = ring->funcs->vmhub; + struct amdgpu_vmid_mgr *id_mgr = &adev->vm_manager.id_mgr[vmhub]; + struct dma_fence **fences; + unsigned i; + int r; + + if (ring->vmid_wait && !dma_fence_is_signaled(ring->vmid_wait)) + return amdgpu_sync_fence(adev, sync, ring->vmid_wait, false); + + fences = kmalloc_array(sizeof(void *), id_mgr->num_ids, GFP_KERNEL); + if (!fences) + return -ENOMEM; + + /* Check if we have an idle VMID */ + i = 0; + list_for_each_entry((*idle), &id_mgr->ids_lru, list) { + fences[i] = amdgpu_sync_peek_fence(&(*idle)->active, ring); + if (!fences[i]) + break; + ++i; + } + + /* If we can't find a idle VMID to use, wait till one becomes available */ + if (&(*idle)->list == &id_mgr->ids_lru) { + u64 fence_context = adev->vm_manager.fence_context + ring->idx; + unsigned seqno = ++adev->vm_manager.seqno[ring->idx]; + struct dma_fence_array *array; + unsigned j; + + *idle = NULL; + for (j = 0; j < i; ++j) + dma_fence_get(fences[j]); + + array = dma_fence_array_create(i, fences, fence_context, + seqno, true); + if (!array) { + for (j = 0; j < i; ++j) + dma_fence_put(fences[j]); + kfree(fences); + return -ENOMEM; + } + + r = amdgpu_sync_fence(adev, sync, &array->base, false); + dma_fence_put(ring->vmid_wait); + ring->vmid_wait = &array->base; + return r; + } + kfree(fences); + + return 0; +} + +/** + * amdgpu_vm_grab_reserved - try to assign reserved VMID + * + * @vm: vm to allocate id for + * @ring: ring we want to submit job to + * @sync: sync object where we add dependencies + * @fence: fence protecting ID from reuse + * @job: job who wants to use the VMID + * + * Try to assign a reserved VMID. + */ +static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm, + struct amdgpu_ring *ring, + struct amdgpu_sync *sync, + struct dma_fence *fence, + struct amdgpu_job *job, + struct amdgpu_vmid **id) { struct amdgpu_device *adev = ring->adev; unsigned vmhub = ring->funcs->vmhub; uint64_t fence_context = adev->fence_context + ring->idx; - struct amdgpu_vmid *id = vm->reserved_vmid[vmhub]; - struct amdgpu_vmid_mgr *id_mgr = &adev->vm_manager.id_mgr[vmhub]; struct dma_fence *updates = sync->last_vm_update; - int r = 0; - struct dma_fence *flushed, *tmp; bool needs_flush = vm->use_cpu_for_update; + int r = 0; - flushed = id->flushed_updates; - if ((amdgpu_vmid_had_gpu_reset(adev, id)) || - (atomic64_read(&id->owner) != vm->entity.fence_context) || - (job->vm_pd_addr != id->pd_gpu_addr) || - (updates && (!flushed || updates->context != flushed->context || - dma_fence_is_later(updates, flushed))) || - (!id->last_flush || (id->last_flush->context != fence_context && - !dma_fence_is_signaled(id->last_flush)))) { - needs_flush = true; + *id = vm->reserved_vmid[vmhub]; + if (updates && (*id)->flushed_updates && + updates->context == (*id)->flushed_updates->context && + !dma_fence_is_later(updates, (*id)->flushed_updates)) + updates = NULL; + + if ((*id)->owner != vm->entity.fence_context || + job->vm_pd_addr != (*id)->pd_gpu_addr || + updates || !(*id)->last_flush || + ((*id)->last_flush->context != fence_context && + !dma_fence_is_signaled((*id)->last_flush))) { + struct dma_fence *tmp; + /* to prevent one context starved by another context */ - id->pd_gpu_addr = 0; - tmp = amdgpu_sync_peek_fence(&id->active, ring); + (*id)->pd_gpu_addr = 0; + tmp = amdgpu_sync_peek_fence(&(*id)->active, ring); if (tmp) { + *id = NULL; r = amdgpu_sync_fence(adev, sync, tmp, false); return r; } + needs_flush = true; } /* Good we can use this VMID. Remember this submission as * user of the VMID. */ - r = amdgpu_sync_fence(ring->adev, &id->active, fence, false); + r = amdgpu_sync_fence(ring->adev, &(*id)->active, fence, false); if (r) - goto out; + return r; - if (updates && (!flushed || updates->context != flushed->context || - dma_fence_is_later(updates, flushed))) { - dma_fence_put(id->flushed_updates); - id->flushed_updates = dma_fence_get(updates); + if (updates) { + dma_fence_put((*id)->flushed_updates); + (*id)->flushed_updates = dma_fence_get(updates); } - id->pd_gpu_addr = job->vm_pd_addr; - atomic64_set(&id->owner, vm->entity.fence_context); job->vm_needs_flush = needs_flush; - if (needs_flush) { - dma_fence_put(id->last_flush); - id->last_flush = NULL; + return 0; +} + +/** + * amdgpu_vm_grab_used - try to reuse a VMID + * + * @vm: vm to allocate id for + * @ring: ring we want to submit job to + * @sync: sync object where we add dependencies + * @fence: fence protecting ID from reuse + * @job: job who wants to use the VMID + * @id: resulting VMID + * + * Try to reuse a VMID for this submission. + */ +static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm, + struct amdgpu_ring *ring, + struct amdgpu_sync *sync, + struct dma_fence *fence, + struct amdgpu_job *job, + struct amdgpu_vmid **id) +{ + struct amdgpu_device *adev = ring->adev; + unsigned vmhub = ring->funcs->vmhub; + struct amdgpu_vmid_mgr *id_mgr = &adev->vm_manager.id_mgr[vmhub]; + uint64_t fence_context = adev->fence_context + ring->idx; + struct dma_fence *updates = sync->last_vm_update; + int r; + + job->vm_needs_flush = vm->use_cpu_for_update; + + /* Check if we can use a VMID already assigned to this VM */ + list_for_each_entry_reverse((*id), &id_mgr->ids_lru, list) { + bool needs_flush = vm->use_cpu_for_update; + struct dma_fence *flushed; + + /* Check all the prerequisites to using this VMID */ + if ((*id)->owner != vm->entity.fence_context) + continue; + + if ((*id)->pd_gpu_addr != job->vm_pd_addr) + continue; + + if (!(*id)->last_flush || + ((*id)->last_flush->context != fence_context && + !dma_fence_is_signaled((*id)->last_flush))) + needs_flush = true; + + flushed = (*id)->flushed_updates; + if (updates && (!flushed || dma_fence_is_later(updates, flushed))) + needs_flush = true; + + /* Concurrent flushes are only possible starting with Vega10 */ + if (adev->asic_type < CHIP_VEGA10 && needs_flush) + continue; + + /* Good, we can use this VMID. Remember this submission as + * user of the VMID. + */ + r = amdgpu_sync_fence(ring->adev, &(*id)->active, fence, false); + if (r) + return r; + + if (updates && (!flushed || dma_fence_is_later(updates, flushed))) { + dma_fence_put((*id)->flushed_updates); + (*id)->flushed_updates = dma_fence_get(updates); + } + + job->vm_needs_flush |= needs_flush; + return 0; } - job->vmid = id - id_mgr->ids; - trace_amdgpu_vm_grab_id(vm, ring, job); -out: - return r; + + *id = NULL; + return 0; } /** @@ -163,6 +395,7 @@ static int amdgpu_vmid_grab_reserved_locked(struct amdgpu_vm *vm, * @ring: ring we want to submit job to * @sync: sync object where we add dependencies * @fence: fence protecting ID from reuse + * @job: job who wants to use the VMID * * Allocate an id for the vm, adding fences to the sync obj as necessary. */ @@ -173,135 +406,53 @@ int amdgpu_vmid_grab(struct amdgpu_vm *vm, struct amdgpu_ring *ring, struct amdgpu_device *adev = ring->adev; unsigned vmhub = ring->funcs->vmhub; struct amdgpu_vmid_mgr *id_mgr = &adev->vm_manager.id_mgr[vmhub]; - uint64_t fence_context = adev->fence_context + ring->idx; - struct dma_fence *updates = sync->last_vm_update; - struct amdgpu_vmid *id, *idle; - struct dma_fence **fences; - unsigned i; + struct amdgpu_vmid *idle = NULL; + struct amdgpu_vmid *id = NULL; int r = 0; mutex_lock(&id_mgr->lock); - if (vm->reserved_vmid[vmhub]) { - r = amdgpu_vmid_grab_reserved_locked(vm, ring, sync, fence, job); - mutex_unlock(&id_mgr->lock); - return r; - } - fences = kmalloc_array(sizeof(void *), id_mgr->num_ids, GFP_KERNEL); - if (!fences) { - mutex_unlock(&id_mgr->lock); - return -ENOMEM; - } - /* Check if we have an idle VMID */ - i = 0; - list_for_each_entry(idle, &id_mgr->ids_lru, list) { - fences[i] = amdgpu_sync_peek_fence(&idle->active, ring); - if (!fences[i]) - break; - ++i; - } - - /* If we can't find a idle VMID to use, wait till one becomes available */ - if (&idle->list == &id_mgr->ids_lru) { - u64 fence_context = adev->vm_manager.fence_context + ring->idx; - unsigned seqno = ++adev->vm_manager.seqno[ring->idx]; - struct dma_fence_array *array; - unsigned j; - - for (j = 0; j < i; ++j) - dma_fence_get(fences[j]); - - array = dma_fence_array_create(i, fences, fence_context, - seqno, true); - if (!array) { - for (j = 0; j < i; ++j) - dma_fence_put(fences[j]); - kfree(fences); - r = -ENOMEM; - goto error; - } - - - r = amdgpu_sync_fence(ring->adev, sync, &array->base, false); - dma_fence_put(&array->base); - if (r) - goto error; - - mutex_unlock(&id_mgr->lock); - return 0; - - } - kfree(fences); - - job->vm_needs_flush = vm->use_cpu_for_update; - /* Check if we can use a VMID already assigned to this VM */ - list_for_each_entry_reverse(id, &id_mgr->ids_lru, list) { - struct dma_fence *flushed; - bool needs_flush = vm->use_cpu_for_update; - - /* Check all the prerequisites to using this VMID */ - if (amdgpu_vmid_had_gpu_reset(adev, id)) - continue; - - if (atomic64_read(&id->owner) != vm->entity.fence_context) - continue; - - if (job->vm_pd_addr != id->pd_gpu_addr) - continue; - - if (!id->last_flush || - (id->last_flush->context != fence_context && - !dma_fence_is_signaled(id->last_flush))) - needs_flush = true; - - flushed = id->flushed_updates; - if (updates && (!flushed || dma_fence_is_later(updates, flushed))) - needs_flush = true; - - /* Concurrent flushes are only possible starting with Vega10 */ - if (adev->asic_type < CHIP_VEGA10 && needs_flush) - continue; - - /* Good we can use this VMID. Remember this submission as - * user of the VMID. - */ - r = amdgpu_sync_fence(ring->adev, &id->active, fence, false); - if (r) - goto error; - - if (updates && (!flushed || dma_fence_is_later(updates, flushed))) { - dma_fence_put(id->flushed_updates); - id->flushed_updates = dma_fence_get(updates); - } - - if (needs_flush) - goto needs_flush; - else - goto no_flush_needed; - - }; - - /* Still no ID to use? Then use the idle one found earlier */ - id = idle; - - /* Remember this submission as user of the VMID */ - r = amdgpu_sync_fence(ring->adev, &id->active, fence, false); - if (r) + r = amdgpu_vmid_grab_idle(vm, ring, sync, &idle); + if (r || !idle) goto error; + if (vm->reserved_vmid[vmhub]) { + r = amdgpu_vmid_grab_reserved(vm, ring, sync, fence, job, &id); + if (r || !id) + goto error; + } else { + r = amdgpu_vmid_grab_used(vm, ring, sync, fence, job, &id); + if (r) + goto error; + + if (!id) { + struct dma_fence *updates = sync->last_vm_update; + + /* Still no ID to use? Then use the idle one found earlier */ + id = idle; + + /* Remember this submission as user of the VMID */ + r = amdgpu_sync_fence(ring->adev, &id->active, + fence, false); + if (r) + goto error; + + dma_fence_put(id->flushed_updates); + id->flushed_updates = dma_fence_get(updates); + job->vm_needs_flush = true; + } + + list_move_tail(&id->list, &id_mgr->ids_lru); + } + id->pd_gpu_addr = job->vm_pd_addr; - dma_fence_put(id->flushed_updates); - id->flushed_updates = dma_fence_get(updates); - atomic64_set(&id->owner, vm->entity.fence_context); + id->owner = vm->entity.fence_context; -needs_flush: - job->vm_needs_flush = true; - dma_fence_put(id->last_flush); - id->last_flush = NULL; - -no_flush_needed: - list_move_tail(&id->list, &id_mgr->ids_lru); - + if (job->vm_needs_flush) { + dma_fence_put(id->last_flush); + id->last_flush = NULL; + } job->vmid = id - id_mgr->ids; + job->pasid = vm->pasid; trace_amdgpu_vm_grab_id(vm, ring, job); error: @@ -370,13 +521,15 @@ void amdgpu_vmid_reset(struct amdgpu_device *adev, unsigned vmhub, struct amdgpu_vmid_mgr *id_mgr = &adev->vm_manager.id_mgr[vmhub]; struct amdgpu_vmid *id = &id_mgr->ids[vmid]; - atomic64_set(&id->owner, 0); + mutex_lock(&id_mgr->lock); + id->owner = 0; id->gds_base = 0; id->gds_size = 0; id->gws_base = 0; id->gws_size = 0; id->oa_base = 0; id->oa_size = 0; + mutex_unlock(&id_mgr->lock); } /** @@ -454,6 +607,7 @@ void amdgpu_vmid_mgr_fini(struct amdgpu_device *adev) amdgpu_sync_free(&id->active); dma_fence_put(id->flushed_updates); dma_fence_put(id->last_flush); + dma_fence_put(id->pasid_mapping); } } }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h index ad931fa..7625419 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h
@@ -43,7 +43,7 @@ struct amdgpu_vmid { struct list_head list; struct amdgpu_sync active; struct dma_fence *last_flush; - atomic64_t owner; + uint64_t owner; uint64_t pd_gpu_addr; /* last flushed PD/PT update */ @@ -57,6 +57,9 @@ struct amdgpu_vmid { uint32_t gws_size; uint32_t oa_base; uint32_t oa_size; + + unsigned pasid; + struct dma_fence *pasid_mapping; }; struct amdgpu_vmid_mgr { @@ -69,6 +72,8 @@ struct amdgpu_vmid_mgr { int amdgpu_pasid_alloc(unsigned int bits); void amdgpu_pasid_free(unsigned int pasid); +void amdgpu_pasid_free_delayed(struct reservation_object *resv, + unsigned int pasid); bool amdgpu_vmid_had_gpu_reset(struct amdgpu_device *adev, struct amdgpu_vmid *id);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h index 29cf109..0e01f11 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h
@@ -25,51 +25,12 @@ #define __AMDGPU_IH_H__ #include <linux/chash.h> +#include "soc15_ih_clientid.h" struct amdgpu_device; - /* - * vega10+ IH clients - */ -enum amdgpu_ih_clientid -{ - AMDGPU_IH_CLIENTID_IH = 0x00, - AMDGPU_IH_CLIENTID_ACP = 0x01, - AMDGPU_IH_CLIENTID_ATHUB = 0x02, - AMDGPU_IH_CLIENTID_BIF = 0x03, - AMDGPU_IH_CLIENTID_DCE = 0x04, - AMDGPU_IH_CLIENTID_ISP = 0x05, - AMDGPU_IH_CLIENTID_PCIE0 = 0x06, - AMDGPU_IH_CLIENTID_RLC = 0x07, - AMDGPU_IH_CLIENTID_SDMA0 = 0x08, - AMDGPU_IH_CLIENTID_SDMA1 = 0x09, - AMDGPU_IH_CLIENTID_SE0SH = 0x0a, - AMDGPU_IH_CLIENTID_SE1SH = 0x0b, - AMDGPU_IH_CLIENTID_SE2SH = 0x0c, - AMDGPU_IH_CLIENTID_SE3SH = 0x0d, - AMDGPU_IH_CLIENTID_SYSHUB = 0x0e, - AMDGPU_IH_CLIENTID_THM = 0x0f, - AMDGPU_IH_CLIENTID_UVD = 0x10, - AMDGPU_IH_CLIENTID_VCE0 = 0x11, - AMDGPU_IH_CLIENTID_VMC = 0x12, - AMDGPU_IH_CLIENTID_XDMA = 0x13, - AMDGPU_IH_CLIENTID_GRBM_CP = 0x14, - AMDGPU_IH_CLIENTID_ATS = 0x15, - AMDGPU_IH_CLIENTID_ROM_SMUIO = 0x16, - AMDGPU_IH_CLIENTID_DF = 0x17, - AMDGPU_IH_CLIENTID_VCE1 = 0x18, - AMDGPU_IH_CLIENTID_PWR = 0x19, - AMDGPU_IH_CLIENTID_UTCL2 = 0x1b, - AMDGPU_IH_CLIENTID_EA = 0x1c, - AMDGPU_IH_CLIENTID_UTCL2LOG = 0x1d, - AMDGPU_IH_CLIENTID_MP0 = 0x1e, - AMDGPU_IH_CLIENTID_MP1 = 0x1f, - - AMDGPU_IH_CLIENTID_MAX, - - AMDGPU_IH_CLIENTID_VCN = AMDGPU_IH_CLIENTID_UVD -}; #define AMDGPU_IH_CLIENTID_LEGACY 0 +#define AMDGPU_IH_CLIENTID_MAX SOC15_IH_CLIENTID_MAX #define AMDGPU_PAGEFAULT_HASH_BITS 8 struct amdgpu_retryfault_hashtable { @@ -109,7 +70,7 @@ struct amdgpu_iv_entry { unsigned vmid_src; uint64_t timestamp; unsigned timestamp_src; - unsigned pas_id; + unsigned pasid; unsigned pasid_src; unsigned src_data[AMDGPU_IH_SRC_DATA_MAX_SIZE_DW]; const uint32_t *iv_entry;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c index 36483e0..3a5ca462 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
@@ -92,7 +92,7 @@ static void amdgpu_irq_reset_work_func(struct work_struct *work) } /* Disable *all* interrupts */ -static void amdgpu_irq_disable_all(struct amdgpu_device *adev) +void amdgpu_irq_disable_all(struct amdgpu_device *adev) { unsigned long irqflags; unsigned i, j, k; @@ -123,55 +123,6 @@ static void amdgpu_irq_disable_all(struct amdgpu_device *adev) } /** - * amdgpu_irq_preinstall - drm irq preinstall callback - * - * @dev: drm dev pointer - * - * Gets the hw ready to enable irqs (all asics). - * This function disables all interrupt sources on the GPU. - */ -void amdgpu_irq_preinstall(struct drm_device *dev) -{ - struct amdgpu_device *adev = dev->dev_private; - - /* Disable *all* interrupts */ - amdgpu_irq_disable_all(adev); - /* Clear bits */ - amdgpu_ih_process(adev); -} - -/** - * amdgpu_irq_postinstall - drm irq preinstall callback - * - * @dev: drm dev pointer - * - * Handles stuff to be done after enabling irqs (all asics). - * Returns 0 on success. - */ -int amdgpu_irq_postinstall(struct drm_device *dev) -{ - dev->max_vblank_count = 0x00ffffff; - return 0; -} - -/** - * amdgpu_irq_uninstall - drm irq uninstall callback - * - * @dev: drm dev pointer - * - * This function disables all interrupt sources on the GPU (all asics). - */ -void amdgpu_irq_uninstall(struct drm_device *dev) -{ - struct amdgpu_device *adev = dev->dev_private; - - if (adev == NULL) { - return; - } - amdgpu_irq_disable_all(adev); -} - -/** * amdgpu_irq_handler - irq handler * * @int irq, void *arg: args @@ -262,6 +213,7 @@ int amdgpu_irq_init(struct amdgpu_device *adev) cancel_work_sync(&adev->reset_work); return r; } + adev->ddev->max_vblank_count = 0x00ffffff; DRM_DEBUG("amdgpu: irq initialized.\n"); return 0; @@ -307,6 +259,7 @@ void amdgpu_irq_fini(struct amdgpu_device *adev) } } kfree(adev->irq.client[i].sources); + adev->irq.client[i].sources = NULL; } }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h index 0610cc4..3375ad7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h
@@ -78,9 +78,7 @@ struct amdgpu_irq { uint32_t srbm_soft_reset; }; -void amdgpu_irq_preinstall(struct drm_device *dev); -int amdgpu_irq_postinstall(struct drm_device *dev); -void amdgpu_irq_uninstall(struct drm_device *dev); +void amdgpu_irq_disable_all(struct amdgpu_device *adev); irqreturn_t amdgpu_irq_handler(int irq, void *arg); int amdgpu_irq_init(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index bfdeacb..0e40f34 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -190,8 +190,12 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info, fw_info->ver = adev->uvd.fw_version; fw_info->feature = 0; break; + case AMDGPU_INFO_FW_VCN: + fw_info->ver = adev->vcn.fw_version; + fw_info->feature = 0; + break; case AMDGPU_INFO_FW_GMC: - fw_info->ver = adev->mc.fw_version; + fw_info->ver = adev->gmc.fw_version; fw_info->feature = 0; break; case AMDGPU_INFO_FW_GFX_ME: @@ -473,9 +477,9 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file case AMDGPU_INFO_VRAM_GTT: { struct drm_amdgpu_info_vram_gtt vram_gtt; - vram_gtt.vram_size = adev->mc.real_vram_size; + vram_gtt.vram_size = adev->gmc.real_vram_size; vram_gtt.vram_size -= adev->vram_pin_size; - vram_gtt.vram_cpu_accessible_size = adev->mc.visible_vram_size; + vram_gtt.vram_cpu_accessible_size = adev->gmc.visible_vram_size; vram_gtt.vram_cpu_accessible_size -= (adev->vram_pin_size - adev->invisible_pin_size); vram_gtt.gtt_size = adev->mman.bdev.man[TTM_PL_TT].size; vram_gtt.gtt_size *= PAGE_SIZE; @@ -487,17 +491,17 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file struct drm_amdgpu_memory_info mem; memset(&mem, 0, sizeof(mem)); - mem.vram.total_heap_size = adev->mc.real_vram_size; + mem.vram.total_heap_size = adev->gmc.real_vram_size; mem.vram.usable_heap_size = - adev->mc.real_vram_size - adev->vram_pin_size; + adev->gmc.real_vram_size - adev->vram_pin_size; mem.vram.heap_usage = amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); mem.vram.max_allocation = mem.vram.usable_heap_size * 3 / 4; mem.cpu_accessible_vram.total_heap_size = - adev->mc.visible_vram_size; + adev->gmc.visible_vram_size; mem.cpu_accessible_vram.usable_heap_size = - adev->mc.visible_vram_size - + adev->gmc.visible_vram_size - (adev->vram_pin_size - adev->invisible_pin_size); mem.cpu_accessible_vram.heap_usage = amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); @@ -583,11 +587,16 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file dev_info.ids_flags |= AMDGPU_IDS_FLAGS_PREEMPTION; vm_size = adev->vm_manager.max_pfn * AMDGPU_GPU_PAGE_SIZE; + vm_size -= AMDGPU_VA_RESERVED_SIZE; + + /* Older VCE FW versions are buggy and can handle only 40bits */ + if (adev->vce.fw_version < AMDGPU_VCE_FW_53_45) + vm_size = min(vm_size, 1ULL << 40); + dev_info.virtual_address_offset = AMDGPU_VA_RESERVED_SIZE; dev_info.virtual_address_max = min(vm_size, AMDGPU_VA_HOLE_START); - vm_size -= AMDGPU_VA_RESERVED_SIZE; if (vm_size > AMDGPU_VA_HOLE_START) { dev_info.high_va_offset = AMDGPU_VA_HOLE_END; dev_info.high_va_max = AMDGPU_VA_HOLE_END | vm_size; @@ -602,8 +611,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file sizeof(adev->gfx.cu_info.ao_cu_bitmap)); memcpy(&dev_info.cu_bitmap[0], &adev->gfx.cu_info.bitmap[0], sizeof(adev->gfx.cu_info.bitmap)); - dev_info.vram_type = adev->mc.vram_type; - dev_info.vram_bit_width = adev->mc.vram_width; + dev_info.vram_type = adev->gmc.vram_type; + dev_info.vram_bit_width = adev->gmc.vram_width; dev_info.vce_harvest_config = adev->vce.harvest_config; dev_info.gc_double_offchip_lds_buf = adev->gfx.config.double_offchip_lds_buf; @@ -761,6 +770,24 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file return -EINVAL; } break; + case AMDGPU_INFO_SENSOR_STABLE_PSTATE_GFX_SCLK: + /* get stable pstate sclk in Mhz */ + if (amdgpu_dpm_read_sensor(adev, + AMDGPU_PP_SENSOR_STABLE_PSTATE_SCLK, + (void *)&ui32, &ui32_size)) { + return -EINVAL; + } + ui32 /= 100; + break; + case AMDGPU_INFO_SENSOR_STABLE_PSTATE_GFX_MCLK: + /* get stable pstate mclk in Mhz */ + if (amdgpu_dpm_read_sensor(adev, + AMDGPU_PP_SENSOR_STABLE_PSTATE_MCLK, + (void *)&ui32, &ui32_size)) { + return -EINVAL; + } + ui32 /= 100; + break; default: DRM_DEBUG_KMS("Invalid request %d\n", info->sensor_info.type); @@ -808,7 +835,7 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) { struct amdgpu_device *adev = dev->dev_private; struct amdgpu_fpriv *fpriv; - int r; + int r, pasid; file_priv->driver_priv = NULL; @@ -822,28 +849,25 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) goto out_suspend; } - r = amdgpu_vm_init(adev, &fpriv->vm, - AMDGPU_VM_CONTEXT_GFX, 0); - if (r) { - kfree(fpriv); - goto out_suspend; + pasid = amdgpu_pasid_alloc(16); + if (pasid < 0) { + dev_warn(adev->dev, "No more PASIDs available!"); + pasid = 0; } + r = amdgpu_vm_init(adev, &fpriv->vm, AMDGPU_VM_CONTEXT_GFX, pasid); + if (r) + goto error_pasid; fpriv->prt_va = amdgpu_vm_bo_add(adev, &fpriv->vm, NULL); if (!fpriv->prt_va) { r = -ENOMEM; - amdgpu_vm_fini(adev, &fpriv->vm); - kfree(fpriv); - goto out_suspend; + goto error_vm; } if (amdgpu_sriov_vf(adev)) { r = amdgpu_map_static_csa(adev, &fpriv->vm, &fpriv->csa_va); - if (r) { - amdgpu_vm_fini(adev, &fpriv->vm); - kfree(fpriv); - goto out_suspend; - } + if (r) + goto error_vm; } mutex_init(&fpriv->bo_list_lock); @@ -852,6 +876,16 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) amdgpu_ctx_mgr_init(&fpriv->ctx_mgr); file_priv->driver_priv = fpriv; + goto out_suspend; + +error_vm: + amdgpu_vm_fini(adev, &fpriv->vm); + +error_pasid: + if (pasid) + amdgpu_pasid_free(pasid); + + kfree(fpriv); out_suspend: pm_runtime_mark_last_busy(dev->dev); @@ -874,6 +908,8 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, struct amdgpu_device *adev = dev->dev_private; struct amdgpu_fpriv *fpriv = file_priv->driver_priv; struct amdgpu_bo_list *list; + struct amdgpu_bo *pd; + unsigned int pasid; int handle; if (!fpriv) @@ -898,7 +934,13 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, amdgpu_bo_unreserve(adev->virt.csa_obj); } + pasid = fpriv->vm.pasid; + pd = amdgpu_bo_ref(fpriv->vm.root.base.bo); + amdgpu_vm_fini(adev, &fpriv->vm); + if (pasid) + amdgpu_pasid_free_delayed(pd->tbo.resv, pasid); + amdgpu_bo_unref(&pd); idr_for_each_entry(&fpriv->bo_list_handles, list, handle) amdgpu_bo_list_free(list); @@ -950,11 +992,11 @@ u32 amdgpu_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe) */ do { count = amdgpu_display_vblank_get_counter(adev, pipe); - /* Ask amdgpu_get_crtc_scanoutpos to return vpos as - * distance to start of vblank, instead of regular - * vertical scanout pos. + /* Ask amdgpu_display_get_crtc_scanoutpos to return + * vpos as distance to start of vblank, instead of + * regular vertical scanout pos. */ - stat = amdgpu_get_crtc_scanoutpos( + stat = amdgpu_display_get_crtc_scanoutpos( dev, pipe, GET_DISTANCE_TO_VBLANKSTART, &vpos, &hpos, NULL, NULL, &adev->mode_info.crtcs[pipe]->base.hwmode); @@ -995,7 +1037,7 @@ u32 amdgpu_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe) int amdgpu_enable_vblank_kms(struct drm_device *dev, unsigned int pipe) { struct amdgpu_device *adev = dev->dev_private; - int idx = amdgpu_crtc_idx_to_irq_type(adev, pipe); + int idx = amdgpu_display_crtc_idx_to_irq_type(adev, pipe); return amdgpu_irq_get(adev, &adev->crtc_irq, idx); } @@ -1011,7 +1053,7 @@ int amdgpu_enable_vblank_kms(struct drm_device *dev, unsigned int pipe) void amdgpu_disable_vblank_kms(struct drm_device *dev, unsigned int pipe) { struct amdgpu_device *adev = dev->dev_private; - int idx = amdgpu_crtc_idx_to_irq_type(adev, pipe); + int idx = amdgpu_display_crtc_idx_to_irq_type(adev, pipe); amdgpu_irq_put(adev, &adev->crtc_irq, idx); } @@ -1163,6 +1205,14 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data) i, fw_info.feature, fw_info.ver); } + /* VCN */ + query_fw.fw_type = AMDGPU_INFO_FW_VCN; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "VCN feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + return 0; }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 2264c5c..d6416ee 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
@@ -267,8 +267,6 @@ struct amdgpu_display_funcs { void (*bandwidth_update)(struct amdgpu_device *adev); /* get frame count */ u32 (*vblank_get_counter)(struct amdgpu_device *adev, int crtc); - /* wait for vblank */ - void (*vblank_wait)(struct amdgpu_device *adev, int crtc); /* set backlight level */ void (*backlight_set_level)(struct amdgpu_encoder *amdgpu_encoder, u8 level); @@ -553,14 +551,6 @@ struct amdgpu_connector { /* we need to mind the EDID between detect and get modes due to analog/digital/tvencoder */ struct edid *edid; - /* number of modes generated from EDID at 'dc_sink' */ - int num_modes; - /* The 'old' sink - before an HPD. - * The 'current' sink is in dc_link->sink. */ - struct dc_sink *dc_sink; - struct dc_link *dc_link; - struct dc_sink *dc_em_sink; - const struct dc_stream *stream; void *con_priv; bool dac_load_detect; bool detected_by_load; /* if the connection status was determined by load */ @@ -571,27 +561,6 @@ struct amdgpu_connector { enum amdgpu_connector_audio audio; enum amdgpu_connector_dither dither; unsigned pixelclock_for_modeset; - - struct drm_dp_mst_topology_mgr mst_mgr; - struct amdgpu_dm_dp_aux dm_dp_aux; - struct drm_dp_mst_port *port; - struct amdgpu_connector *mst_port; - struct amdgpu_encoder *mst_encoder; - struct semaphore mst_sem; - - /* TODO see if we can merge with ddc_bus or make a dm_connector */ - struct amdgpu_i2c_adapter *i2c; - - /* Monitor range limits */ - int min_vfreq ; - int max_vfreq ; - int pixel_clock_mhz; - - /*freesync caps*/ - struct mod_freesync_caps caps; - - struct mutex hpd_lock; - }; /* TODO: start to use this struct and remove same field from base one */ @@ -609,7 +578,7 @@ struct amdgpu_mst_connector { #define ENCODER_MODE_IS_DP(em) (((em) == ATOM_ENCODER_MODE_DP) || \ ((em) == ATOM_ENCODER_MODE_DP_MST)) -/* Driver internal use only flags of amdgpu_get_crtc_scanoutpos() */ +/* Driver internal use only flags of amdgpu_display_get_crtc_scanoutpos() */ #define DRM_SCANOUTPOS_VALID (1 << 0) #define DRM_SCANOUTPOS_IN_VBLANK (1 << 1) #define DRM_SCANOUTPOS_ACCURATE (1 << 2) @@ -628,30 +597,31 @@ bool amdgpu_dig_monitor_is_duallink(struct drm_encoder *encoder, u16 amdgpu_encoder_get_dp_bridge_encoder_id(struct drm_encoder *encoder); struct drm_encoder *amdgpu_get_external_encoder(struct drm_encoder *encoder); -bool amdgpu_ddc_probe(struct amdgpu_connector *amdgpu_connector, bool use_aux); +bool amdgpu_display_ddc_probe(struct amdgpu_connector *amdgpu_connector, + bool use_aux); void amdgpu_encoder_set_active_device(struct drm_encoder *encoder); -int amdgpu_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, - unsigned int flags, int *vpos, int *hpos, - ktime_t *stime, ktime_t *etime, - const struct drm_display_mode *mode); +int amdgpu_display_get_crtc_scanoutpos(struct drm_device *dev, + unsigned int pipe, unsigned int flags, int *vpos, + int *hpos, ktime_t *stime, ktime_t *etime, + const struct drm_display_mode *mode); -int amdgpu_framebuffer_init(struct drm_device *dev, - struct amdgpu_framebuffer *rfb, - const struct drm_mode_fb_cmd2 *mode_cmd, - struct drm_gem_object *obj); +int amdgpu_display_framebuffer_init(struct drm_device *dev, + struct amdgpu_framebuffer *rfb, + const struct drm_mode_fb_cmd2 *mode_cmd, + struct drm_gem_object *obj); int amdgpufb_remove(struct drm_device *dev, struct drm_framebuffer *fb); void amdgpu_enc_destroy(struct drm_encoder *encoder); void amdgpu_copy_fb(struct drm_device *dev, struct drm_gem_object *dst_obj); -bool amdgpu_crtc_scaling_mode_fixup(struct drm_crtc *crtc, - const struct drm_display_mode *mode, - struct drm_display_mode *adjusted_mode); +bool amdgpu_display_crtc_scaling_mode_fixup(struct drm_crtc *crtc, + const struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode); void amdgpu_panel_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *adjusted_mode); -int amdgpu_crtc_idx_to_irq_type(struct amdgpu_device *adev, int crtc); +int amdgpu_display_crtc_idx_to_irq_type(struct amdgpu_device *adev, int crtc); /* fbdev layer */ int amdgpu_fbdev_init(struct amdgpu_device *adev); @@ -663,15 +633,15 @@ bool amdgpu_fbdev_robj_is_fb(struct amdgpu_device *adev, struct amdgpu_bo *robj) int amdgpu_align_pitch(struct amdgpu_device *adev, int width, int bpp, bool tiled); /* amdgpu_display.c */ -void amdgpu_print_display_setup(struct drm_device *dev); -int amdgpu_modeset_create_props(struct amdgpu_device *adev); -int amdgpu_crtc_set_config(struct drm_mode_set *set, - struct drm_modeset_acquire_ctx *ctx); -int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - struct drm_pending_vblank_event *event, - uint32_t page_flip_flags, uint32_t target, - struct drm_modeset_acquire_ctx *ctx); +void amdgpu_display_print_display_setup(struct drm_device *dev); +int amdgpu_display_modeset_create_props(struct amdgpu_device *adev); +int amdgpu_display_crtc_set_config(struct drm_mode_set *set, + struct drm_modeset_acquire_ctx *ctx); +int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc, + struct drm_framebuffer *fb, + struct drm_pending_vblank_event *event, + uint32_t page_flip_flags, uint32_t target, + struct drm_modeset_acquire_ctx *ctx); extern const struct drm_mode_config_funcs amdgpu_mode_funcs; #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 47c0f73..d0687d2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -36,6 +36,7 @@ #include <drm/drm_cache.h> #include "amdgpu.h" #include "amdgpu_trace.h" +#include "amdgpu_amdkfd.h" static bool amdgpu_need_backup(struct amdgpu_device *adev) { @@ -54,6 +55,9 @@ static void amdgpu_ttm_bo_destroy(struct ttm_buffer_object *tbo) struct amdgpu_device *adev = amdgpu_ttm_adev(tbo->bdev); struct amdgpu_bo *bo = ttm_to_amdgpu_bo(tbo); + if (bo->kfd_bo) + amdgpu_amdkfd_unreserve_system_memory_limit(bo); + amdgpu_bo_kunmap(bo); if (bo->gem_base.import_attach) @@ -85,7 +89,7 @@ void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain) u32 c = 0; if (domain & AMDGPU_GEM_DOMAIN_VRAM) { - unsigned visible_pfn = adev->mc.visible_vram_size >> PAGE_SHIFT; + unsigned visible_pfn = adev->gmc.visible_vram_size >> PAGE_SHIFT; places[c].fpfn = 0; places[c].lpfn = 0; @@ -105,7 +109,7 @@ void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain) if (domain & AMDGPU_GEM_DOMAIN_GTT) { places[c].fpfn = 0; if (flags & AMDGPU_GEM_CREATE_SHADOW) - places[c].lpfn = adev->mc.gart_size >> PAGE_SHIFT; + places[c].lpfn = adev->gmc.gart_size >> PAGE_SHIFT; else places[c].lpfn = 0; places[c].flags = TTM_PL_FLAG_TT; @@ -171,13 +175,15 @@ void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain) * @size: size for the new BO * @align: alignment for the new BO * @domain: where to place it - * @bo_ptr: resulting BO + * @bo_ptr: used to initialize BOs in structures * @gpu_addr: GPU addr of the pinned BO * @cpu_addr: optional CPU address mapping * * Allocates and pins a BO for kernel internal use, and returns it still * reserved. * + * Note: For bo_ptr new BO is only created if bo_ptr points to NULL. + * * Returns 0 on success, negative error code otherwise. */ int amdgpu_bo_create_reserved(struct amdgpu_device *adev, @@ -189,10 +195,10 @@ int amdgpu_bo_create_reserved(struct amdgpu_device *adev, int r; if (!*bo_ptr) { - r = amdgpu_bo_create(adev, size, align, true, domain, + r = amdgpu_bo_create(adev, size, align, domain, AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, - NULL, NULL, 0, bo_ptr); + ttm_bo_type_kernel, NULL, bo_ptr); if (r) { dev_err(adev->dev, "(%d) failed to allocate kernel bo\n", r); @@ -240,12 +246,14 @@ int amdgpu_bo_create_reserved(struct amdgpu_device *adev, * @size: size for the new BO * @align: alignment for the new BO * @domain: where to place it - * @bo_ptr: resulting BO + * @bo_ptr: used to initialize BOs in structures * @gpu_addr: GPU addr of the pinned BO * @cpu_addr: optional CPU address mapping * * Allocates and pins a BO for kernel internal use. * + * Note: For bo_ptr new BO is only created if bo_ptr points to NULL. + * * Returns 0 on success, negative error code otherwise. */ int amdgpu_bo_create_kernel(struct amdgpu_device *adev, @@ -333,22 +341,19 @@ static bool amdgpu_bo_validate_size(struct amdgpu_device *adev, return false; } -static int amdgpu_bo_do_create(struct amdgpu_device *adev, - unsigned long size, int byte_align, - bool kernel, u32 domain, u64 flags, - struct sg_table *sg, +static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size, + int byte_align, u32 domain, + u64 flags, enum ttm_bo_type type, struct reservation_object *resv, - uint64_t init_value, struct amdgpu_bo **bo_ptr) { struct ttm_operation_ctx ctx = { - .interruptible = !kernel, + .interruptible = (type != ttm_bo_type_kernel), .no_wait_gpu = false, - .allow_reserved_eviction = true, - .resv = resv + .resv = resv, + .flags = TTM_OPT_FLAG_ALLOW_RES_EVICT }; struct amdgpu_bo *bo; - enum ttm_bo_type type; unsigned long page_align; size_t acc_size; int r; @@ -359,13 +364,6 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, if (!amdgpu_bo_validate_size(adev, size, domain)) return -ENOMEM; - if (kernel) { - type = ttm_bo_type_kernel; - } else if (sg) { - type = ttm_bo_type_sg; - } else { - type = ttm_bo_type_device; - } *bo_ptr = NULL; acc_size = ttm_bo_dma_acc_size(&adev->mman.bdev, size, @@ -374,11 +372,7 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, bo = kzalloc(sizeof(struct amdgpu_bo), GFP_KERNEL); if (bo == NULL) return -ENOMEM; - r = drm_gem_object_init(adev->ddev, &bo->gem_base, size); - if (unlikely(r)) { - kfree(bo); - return r; - } + drm_gem_private_object_init(adev->ddev, &bo->gem_base, size); INIT_LIST_HEAD(&bo->shadow_list); INIT_LIST_HEAD(&bo->va); bo->preferred_domains = domain & (AMDGPU_GEM_DOMAIN_VRAM | @@ -388,7 +382,8 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, AMDGPU_GEM_DOMAIN_GWS | AMDGPU_GEM_DOMAIN_OA); bo->allowed_domains = bo->preferred_domains; - if (!kernel && bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM) + if (type != ttm_bo_type_kernel && + bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM) bo->allowed_domains |= AMDGPU_GEM_DOMAIN_GTT; bo->flags = flags; @@ -425,27 +420,27 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, amdgpu_ttm_placement_from_domain(bo, domain); r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type, - &bo->placement, page_align, &ctx, NULL, - acc_size, sg, resv, &amdgpu_ttm_bo_destroy); + &bo->placement, page_align, &ctx, acc_size, + NULL, resv, &amdgpu_ttm_bo_destroy); if (unlikely(r != 0)) return r; - if (adev->mc.visible_vram_size < adev->mc.real_vram_size && + if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size && bo->tbo.mem.mem_type == TTM_PL_VRAM && - bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT) + bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT) amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, ctx.bytes_moved); else amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, 0); - if (kernel) + if (type == ttm_bo_type_kernel) bo->tbo.priority = 1; if (flags & AMDGPU_GEM_CREATE_VRAM_CLEARED && bo->tbo.mem.placement & TTM_PL_FLAG_VRAM) { struct dma_fence *fence; - r = amdgpu_fill_buffer(bo, init_value, bo->tbo.resv, &fence); + r = amdgpu_fill_buffer(bo, 0, bo->tbo.resv, &fence); if (unlikely(r)) goto fail_unreserve; @@ -482,12 +477,11 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev, if (bo->shadow) return 0; - r = amdgpu_bo_do_create(adev, size, byte_align, true, - AMDGPU_GEM_DOMAIN_GTT, + r = amdgpu_bo_do_create(adev, size, byte_align, AMDGPU_GEM_DOMAIN_GTT, AMDGPU_GEM_CREATE_CPU_GTT_USWC | AMDGPU_GEM_CREATE_SHADOW, - NULL, bo->tbo.resv, 0, - &bo->shadow); + ttm_bo_type_kernel, + bo->tbo.resv, &bo->shadow); if (!r) { bo->shadow->parent = amdgpu_bo_ref(bo); mutex_lock(&adev->shadow_list_lock); @@ -498,22 +492,17 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev, return r; } -/* init_value will only take effect when flags contains - * AMDGPU_GEM_CREATE_VRAM_CLEARED. - */ -int amdgpu_bo_create(struct amdgpu_device *adev, - unsigned long size, int byte_align, - bool kernel, u32 domain, u64 flags, - struct sg_table *sg, +int amdgpu_bo_create(struct amdgpu_device *adev, unsigned long size, + int byte_align, u32 domain, + u64 flags, enum ttm_bo_type type, struct reservation_object *resv, - uint64_t init_value, struct amdgpu_bo **bo_ptr) { uint64_t parent_flags = flags & ~AMDGPU_GEM_CREATE_SHADOW; int r; - r = amdgpu_bo_do_create(adev, size, byte_align, kernel, domain, - parent_flags, sg, resv, init_value, bo_ptr); + r = amdgpu_bo_do_create(adev, size, byte_align, domain, + parent_flags, type, resv, bo_ptr); if (r) return r; @@ -835,23 +824,24 @@ static const char *amdgpu_vram_names[] = { "GDDR4", "GDDR5", "HBM", - "DDR3" + "DDR3", + "DDR4", }; int amdgpu_bo_init(struct amdgpu_device *adev) { /* reserve PAT memory space to WC for VRAM */ - arch_io_reserve_memtype_wc(adev->mc.aper_base, - adev->mc.aper_size); + arch_io_reserve_memtype_wc(adev->gmc.aper_base, + adev->gmc.aper_size); /* Add an MTRR for the VRAM */ - adev->mc.vram_mtrr = arch_phys_wc_add(adev->mc.aper_base, - adev->mc.aper_size); + adev->gmc.vram_mtrr = arch_phys_wc_add(adev->gmc.aper_base, + adev->gmc.aper_size); DRM_INFO("Detected VRAM RAM=%lluM, BAR=%lluM\n", - adev->mc.mc_vram_size >> 20, - (unsigned long long)adev->mc.aper_size >> 20); + adev->gmc.mc_vram_size >> 20, + (unsigned long long)adev->gmc.aper_size >> 20); DRM_INFO("RAM width %dbits %s\n", - adev->mc.vram_width, amdgpu_vram_names[adev->mc.vram_type]); + adev->gmc.vram_width, amdgpu_vram_names[adev->gmc.vram_type]); return amdgpu_ttm_init(adev); } @@ -865,8 +855,8 @@ int amdgpu_bo_late_init(struct amdgpu_device *adev) void amdgpu_bo_fini(struct amdgpu_device *adev) { amdgpu_ttm_fini(adev); - arch_phys_wc_del(adev->mc.vram_mtrr); - arch_io_free_memtype_wc(adev->mc.aper_base, adev->mc.aper_size); + arch_phys_wc_del(adev->gmc.vram_mtrr); + arch_io_free_memtype_wc(adev->gmc.aper_base, adev->gmc.aper_size); } int amdgpu_bo_fbdev_mmap(struct amdgpu_bo *bo, @@ -996,7 +986,7 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) size = bo->mem.num_pages << PAGE_SHIFT; offset = bo->mem.start << PAGE_SHIFT; - if ((offset + size) <= adev->mc.visible_vram_size) + if ((offset + size) <= adev->gmc.visible_vram_size) return 0; /* Can't move a pinned BO to visible VRAM */ @@ -1019,7 +1009,7 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) offset = bo->mem.start << PAGE_SHIFT; /* this should never happen */ if (bo->mem.mem_type == TTM_PL_VRAM && - (offset + size) > adev->mc.visible_vram_size) + (offset + size) > adev->gmc.visible_vram_size) return -EINVAL; return 0; @@ -1072,7 +1062,7 @@ uint32_t amdgpu_bo_get_preferred_pin_domain(struct amdgpu_device *adev, { if (domain == (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT)) { domain = AMDGPU_GEM_DOMAIN_VRAM; - if (adev->mc.real_vram_size <= AMDGPU_SG_THRESHOLD) + if (adev->gmc.real_vram_size <= AMDGPU_SG_THRESHOLD) domain = AMDGPU_GEM_DOMAIN_GTT; } return domain;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index 8e47dfd..b301ef9b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -92,6 +92,8 @@ struct amdgpu_bo { struct list_head mn_list; struct list_head shadow_list; }; + + struct kgd_mem *kfd_bo; }; static inline struct amdgpu_bo *ttm_to_amdgpu_bo(struct ttm_buffer_object *tbo) @@ -201,13 +203,11 @@ static inline bool amdgpu_bo_explicit_sync(struct amdgpu_bo *bo) return bo->flags & AMDGPU_GEM_CREATE_EXPLICIT_SYNC; } -int amdgpu_bo_create(struct amdgpu_device *adev, - unsigned long size, int byte_align, - bool kernel, u32 domain, u64 flags, - struct sg_table *sg, - struct reservation_object *resv, - uint64_t init_value, - struct amdgpu_bo **bo_ptr); +int amdgpu_bo_create(struct amdgpu_device *adev, unsigned long size, + int byte_align, u32 domain, + u64 flags, enum ttm_bo_type type, + struct reservation_object *resv, + struct amdgpu_bo **bo_ptr); int amdgpu_bo_create_reserved(struct amdgpu_device *adev, unsigned long size, int align, u32 domain, struct amdgpu_bo **bo_ptr, @@ -284,8 +284,6 @@ void amdgpu_sa_bo_manager_fini(struct amdgpu_device *adev, struct amdgpu_sa_manager *sa_manager); int amdgpu_sa_bo_manager_start(struct amdgpu_device *adev, struct amdgpu_sa_manager *sa_manager); -int amdgpu_sa_bo_manager_suspend(struct amdgpu_device *adev, - struct amdgpu_sa_manager *sa_manager); int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager, struct amdgpu_sa_bo **sa_bo, unsigned size, unsigned align);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c index 30a8843..865c5b3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -118,7 +118,7 @@ static ssize_t amdgpu_set_dpm_state(struct device *dev, } if (adev->powerplay.pp_funcs->dispatch_tasks) { - amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_ENABLE_USER_STATE, &state, NULL); + amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_ENABLE_USER_STATE, &state); } else { mutex_lock(&adev->pm.mutex); adev->pm.dpm.user_state = state; @@ -318,7 +318,7 @@ static ssize_t amdgpu_set_pp_force_state(struct device *dev, if (state != POWER_STATE_TYPE_INTERNAL_BOOT && state != POWER_STATE_TYPE_DEFAULT) { amdgpu_dpm_dispatch_task(adev, - AMD_PP_TASK_ENABLE_USER_STATE, &state, NULL); + AMD_PP_TASK_ENABLE_USER_STATE, &state); adev->pp_force_state_enabled = true; } } @@ -362,6 +362,90 @@ static ssize_t amdgpu_set_pp_table(struct device *dev, return count; } +static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t count) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + int ret; + uint32_t parameter_size = 0; + long parameter[64]; + char buf_cpy[128]; + char *tmp_str; + char *sub_str; + const char delimiter[3] = {' ', '\n', '\0'}; + uint32_t type; + + if (count > 127) + return -EINVAL; + + if (*buf == 's') + type = PP_OD_EDIT_SCLK_VDDC_TABLE; + else if (*buf == 'm') + type = PP_OD_EDIT_MCLK_VDDC_TABLE; + else if(*buf == 'r') + type = PP_OD_RESTORE_DEFAULT_TABLE; + else if (*buf == 'c') + type = PP_OD_COMMIT_DPM_TABLE; + else + return -EINVAL; + + memcpy(buf_cpy, buf, count+1); + + tmp_str = buf_cpy; + + while (isspace(*++tmp_str)); + + while (tmp_str[0]) { + sub_str = strsep(&tmp_str, delimiter); + ret = kstrtol(sub_str, 0, ¶meter[parameter_size]); + if (ret) + return -EINVAL; + parameter_size++; + + while (isspace(*tmp_str)) + tmp_str++; + } + + if (adev->powerplay.pp_funcs->odn_edit_dpm_table) + ret = amdgpu_dpm_odn_edit_dpm_table(adev, type, + parameter, parameter_size); + + if (ret) + return -EINVAL; + + if (type == PP_OD_COMMIT_DPM_TABLE) { + if (adev->powerplay.pp_funcs->dispatch_tasks) { + amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); + return count; + } else { + return -EINVAL; + } + } + + return count; +} + +static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + uint32_t size = 0; + + if (adev->powerplay.pp_funcs->print_clock_levels) { + size = amdgpu_dpm_print_clock_levels(adev, OD_SCLK, buf); + size += amdgpu_dpm_print_clock_levels(adev, OD_MCLK, buf+size); + return size; + } else { + return snprintf(buf, PAGE_SIZE, "\n"); + } + +} + static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev, struct device_attribute *attr, char *buf) @@ -532,7 +616,7 @@ static ssize_t amdgpu_set_pp_sclk_od(struct device *dev, amdgpu_dpm_set_sclk_od(adev, (uint32_t)value); if (adev->powerplay.pp_funcs->dispatch_tasks) { - amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL, NULL); + amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); } else { adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps; amdgpu_pm_compute_clocks(adev); @@ -576,7 +660,7 @@ static ssize_t amdgpu_set_pp_mclk_od(struct device *dev, amdgpu_dpm_set_mclk_od(adev, (uint32_t)value); if (adev->powerplay.pp_funcs->dispatch_tasks) { - amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL, NULL); + amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); } else { adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps; amdgpu_pm_compute_clocks(adev); @@ -586,159 +670,70 @@ static ssize_t amdgpu_set_pp_mclk_od(struct device *dev, return count; } -static ssize_t amdgpu_get_pp_power_profile(struct device *dev, - char *buf, struct amd_pp_profile *query) +static ssize_t amdgpu_get_pp_power_profile_mode(struct device *dev, + struct device_attribute *attr, + char *buf) { struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - int ret = 0xff; - if (adev->powerplay.pp_funcs->get_power_profile_state) - ret = amdgpu_dpm_get_power_profile_state( - adev, query); + if (adev->powerplay.pp_funcs->get_power_profile_mode) + return amdgpu_dpm_get_power_profile_mode(adev, buf); - if (ret) - return ret; - - return snprintf(buf, PAGE_SIZE, - "%d %d %d %d %d\n", - query->min_sclk / 100, - query->min_mclk / 100, - query->activity_threshold, - query->up_hyst, - query->down_hyst); + return snprintf(buf, PAGE_SIZE, "\n"); } -static ssize_t amdgpu_get_pp_gfx_power_profile(struct device *dev, + +static ssize_t amdgpu_set_pp_power_profile_mode(struct device *dev, struct device_attribute *attr, - char *buf) -{ - struct amd_pp_profile query = {0}; - - query.type = AMD_PP_GFX_PROFILE; - - return amdgpu_get_pp_power_profile(dev, buf, &query); -} - -static ssize_t amdgpu_get_pp_compute_power_profile(struct device *dev, - struct device_attribute *attr, - char *buf) -{ - struct amd_pp_profile query = {0}; - - query.type = AMD_PP_COMPUTE_PROFILE; - - return amdgpu_get_pp_power_profile(dev, buf, &query); -} - -static ssize_t amdgpu_set_pp_power_profile(struct device *dev, const char *buf, - size_t count, - struct amd_pp_profile *request) + size_t count) { + int ret = 0xff; struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - uint32_t loop = 0; - char *sub_str, buf_cpy[128], *tmp_str; + uint32_t parameter_size = 0; + long parameter[64]; + char *sub_str, buf_cpy[128]; + char *tmp_str; + uint32_t i = 0; + char tmp[2]; + long int profile_mode = 0; const char delimiter[3] = {' ', '\n', '\0'}; - long int value; - int ret = 0xff; - if (strncmp("reset", buf, strlen("reset")) == 0) { - if (adev->powerplay.pp_funcs->reset_power_profile_state) - ret = amdgpu_dpm_reset_power_profile_state( - adev, request); - if (ret) { - count = -EINVAL; - goto fail; - } - return count; - } - - if (strncmp("set", buf, strlen("set")) == 0) { - if (adev->powerplay.pp_funcs->set_power_profile_state) - ret = amdgpu_dpm_set_power_profile_state( - adev, request); - - if (ret) { - count = -EINVAL; - goto fail; - } - return count; - } - - if (count + 1 >= 128) { - count = -EINVAL; - goto fail; - } - - memcpy(buf_cpy, buf, count + 1); - tmp_str = buf_cpy; - - while (tmp_str[0]) { - sub_str = strsep(&tmp_str, delimiter); - ret = kstrtol(sub_str, 0, &value); - if (ret) { - count = -EINVAL; - goto fail; - } - - switch (loop) { - case 0: - /* input unit MHz convert to dpm table unit 10KHz*/ - request->min_sclk = (uint32_t)value * 100; - break; - case 1: - /* input unit MHz convert to dpm table unit 10KHz*/ - request->min_mclk = (uint32_t)value * 100; - break; - case 2: - request->activity_threshold = (uint16_t)value; - break; - case 3: - request->up_hyst = (uint8_t)value; - break; - case 4: - request->down_hyst = (uint8_t)value; - break; - default: - break; - } - - loop++; - } - if (adev->powerplay.pp_funcs->set_power_profile_state) - ret = amdgpu_dpm_set_power_profile_state(adev, request); - + tmp[0] = *(buf); + tmp[1] = '\0'; + ret = kstrtol(tmp, 0, &profile_mode); if (ret) - count = -EINVAL; + goto fail; + if (profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { + if (count < 2 || count > 127) + return -EINVAL; + while (isspace(*++buf)) + i++; + memcpy(buf_cpy, buf, count-i); + tmp_str = buf_cpy; + while (tmp_str[0]) { + sub_str = strsep(&tmp_str, delimiter); + ret = kstrtol(sub_str, 0, ¶meter[parameter_size]); + if (ret) { + count = -EINVAL; + goto fail; + } + parameter_size++; + while (isspace(*tmp_str)) + tmp_str++; + } + } + parameter[parameter_size] = profile_mode; + if (adev->powerplay.pp_funcs->set_power_profile_mode) + ret = amdgpu_dpm_set_power_profile_mode(adev, parameter, parameter_size); + + if (!ret) + return count; fail: - return count; -} - -static ssize_t amdgpu_set_pp_gfx_power_profile(struct device *dev, - struct device_attribute *attr, - const char *buf, - size_t count) -{ - struct amd_pp_profile request = {0}; - - request.type = AMD_PP_GFX_PROFILE; - - return amdgpu_set_pp_power_profile(dev, buf, count, &request); -} - -static ssize_t amdgpu_set_pp_compute_power_profile(struct device *dev, - struct device_attribute *attr, - const char *buf, - size_t count) -{ - struct amd_pp_profile request = {0}; - - request.type = AMD_PP_COMPUTE_PROFILE; - - return amdgpu_set_pp_power_profile(dev, buf, count, &request); + return -EINVAL; } static DEVICE_ATTR(power_dpm_state, S_IRUGO | S_IWUSR, amdgpu_get_dpm_state, amdgpu_set_dpm_state); @@ -768,12 +763,12 @@ static DEVICE_ATTR(pp_sclk_od, S_IRUGO | S_IWUSR, static DEVICE_ATTR(pp_mclk_od, S_IRUGO | S_IWUSR, amdgpu_get_pp_mclk_od, amdgpu_set_pp_mclk_od); -static DEVICE_ATTR(pp_gfx_power_profile, S_IRUGO | S_IWUSR, - amdgpu_get_pp_gfx_power_profile, - amdgpu_set_pp_gfx_power_profile); -static DEVICE_ATTR(pp_compute_power_profile, S_IRUGO | S_IWUSR, - amdgpu_get_pp_compute_power_profile, - amdgpu_set_pp_compute_power_profile); +static DEVICE_ATTR(pp_power_profile_mode, S_IRUGO | S_IWUSR, + amdgpu_get_pp_power_profile_mode, + amdgpu_set_pp_power_profile_mode); +static DEVICE_ATTR(pp_od_clk_voltage, S_IRUGO | S_IWUSR, + amdgpu_get_pp_od_clk_voltage, + amdgpu_set_pp_od_clk_voltage); static ssize_t amdgpu_hwmon_show_temp(struct device *dev, struct device_attribute *attr, @@ -781,17 +776,23 @@ static ssize_t amdgpu_hwmon_show_temp(struct device *dev, { struct amdgpu_device *adev = dev_get_drvdata(dev); struct drm_device *ddev = adev->ddev; - int temp; + int r, temp, size = sizeof(temp); /* Can't get temperature when the card is off */ if ((adev->flags & AMD_IS_PX) && (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - if (!adev->powerplay.pp_funcs->get_temperature) - temp = 0; - else - temp = amdgpu_dpm_get_temperature(adev); + /* sanity check PP is enabled */ + if (!(adev->powerplay.pp_funcs && + adev->powerplay.pp_funcs->read_sensor)) + return -EINVAL; + + /* get the temperature */ + r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP, + (void *)&temp, &size); + if (r) + return r; return snprintf(buf, PAGE_SIZE, "%d\n", temp); } @@ -836,6 +837,11 @@ static ssize_t amdgpu_hwmon_set_pwm1_enable(struct device *dev, int err; int value; + /* Can't adjust fan when the card is off */ + if ((adev->flags & AMD_IS_PX) && + (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + if (!adev->powerplay.pp_funcs->set_fan_control_mode) return -EINVAL; @@ -870,6 +876,11 @@ static ssize_t amdgpu_hwmon_set_pwm1(struct device *dev, int err; u32 value; + /* Can't adjust fan when the card is off */ + if ((adev->flags & AMD_IS_PX) && + (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + err = kstrtou32(buf, 10, &value); if (err) return err; @@ -893,6 +904,11 @@ static ssize_t amdgpu_hwmon_get_pwm1(struct device *dev, int err; u32 speed = 0; + /* Can't adjust fan when the card is off */ + if ((adev->flags & AMD_IS_PX) && + (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + if (adev->powerplay.pp_funcs->get_fan_speed_percent) { err = amdgpu_dpm_get_fan_speed_percent(adev, &speed); if (err) @@ -912,6 +928,11 @@ static ssize_t amdgpu_hwmon_get_fan1_input(struct device *dev, int err; u32 speed = 0; + /* Can't adjust fan when the card is off */ + if ((adev->flags & AMD_IS_PX) && + (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + if (adev->powerplay.pp_funcs->get_fan_speed_rpm) { err = amdgpu_dpm_get_fan_speed_rpm(adev, &speed); if (err) @@ -921,6 +942,175 @@ static ssize_t amdgpu_hwmon_get_fan1_input(struct device *dev, return sprintf(buf, "%i\n", speed); } +static ssize_t amdgpu_hwmon_show_vddgfx(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct amdgpu_device *adev = dev_get_drvdata(dev); + struct drm_device *ddev = adev->ddev; + u32 vddgfx; + int r, size = sizeof(vddgfx); + + /* Can't get voltage when the card is off */ + if ((adev->flags & AMD_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + + /* sanity check PP is enabled */ + if (!(adev->powerplay.pp_funcs && + adev->powerplay.pp_funcs->read_sensor)) + return -EINVAL; + + /* get the voltage */ + r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDGFX, + (void *)&vddgfx, &size); + if (r) + return r; + + return snprintf(buf, PAGE_SIZE, "%d\n", vddgfx); +} + +static ssize_t amdgpu_hwmon_show_vddgfx_label(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return snprintf(buf, PAGE_SIZE, "vddgfx\n"); +} + +static ssize_t amdgpu_hwmon_show_vddnb(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct amdgpu_device *adev = dev_get_drvdata(dev); + struct drm_device *ddev = adev->ddev; + u32 vddnb; + int r, size = sizeof(vddnb); + + /* only APUs have vddnb */ + if (adev->flags & AMD_IS_APU) + return -EINVAL; + + /* Can't get voltage when the card is off */ + if ((adev->flags & AMD_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + + /* sanity check PP is enabled */ + if (!(adev->powerplay.pp_funcs && + adev->powerplay.pp_funcs->read_sensor)) + return -EINVAL; + + /* get the voltage */ + r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDNB, + (void *)&vddnb, &size); + if (r) + return r; + + return snprintf(buf, PAGE_SIZE, "%d\n", vddnb); +} + +static ssize_t amdgpu_hwmon_show_vddnb_label(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return snprintf(buf, PAGE_SIZE, "vddnb\n"); +} + +static ssize_t amdgpu_hwmon_show_power_avg(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct amdgpu_device *adev = dev_get_drvdata(dev); + struct drm_device *ddev = adev->ddev; + struct pp_gpu_power query = {0}; + int r, size = sizeof(query); + unsigned uw; + + /* Can't get power when the card is off */ + if ((adev->flags & AMD_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + + /* sanity check PP is enabled */ + if (!(adev->powerplay.pp_funcs && + adev->powerplay.pp_funcs->read_sensor)) + return -EINVAL; + + /* get the voltage */ + r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_POWER, + (void *)&query, &size); + if (r) + return r; + + /* convert to microwatts */ + uw = (query.average_gpu_power >> 8) * 1000000; + + return snprintf(buf, PAGE_SIZE, "%u\n", uw); +} + +static ssize_t amdgpu_hwmon_show_power_cap_min(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return sprintf(buf, "%i\n", 0); +} + +static ssize_t amdgpu_hwmon_show_power_cap_max(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct amdgpu_device *adev = dev_get_drvdata(dev); + uint32_t limit = 0; + + if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) { + adev->powerplay.pp_funcs->get_power_limit(adev->powerplay.pp_handle, &limit, true); + return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000); + } else { + return snprintf(buf, PAGE_SIZE, "\n"); + } +} + +static ssize_t amdgpu_hwmon_show_power_cap(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct amdgpu_device *adev = dev_get_drvdata(dev); + uint32_t limit = 0; + + if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) { + adev->powerplay.pp_funcs->get_power_limit(adev->powerplay.pp_handle, &limit, false); + return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000); + } else { + return snprintf(buf, PAGE_SIZE, "\n"); + } +} + + +static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t count) +{ + struct amdgpu_device *adev = dev_get_drvdata(dev); + int err; + u32 value; + + err = kstrtou32(buf, 10, &value); + if (err) + return err; + + value = value / 1000000; /* convert to Watt */ + if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->set_power_limit) { + err = adev->powerplay.pp_funcs->set_power_limit(adev->powerplay.pp_handle, value); + if (err) + return err; + } else { + return -EINVAL; + } + + return count; +} + static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, 0); static SENSOR_DEVICE_ATTR(temp1_crit, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 0); static SENSOR_DEVICE_ATTR(temp1_crit_hyst, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 1); @@ -929,6 +1119,14 @@ static SENSOR_DEVICE_ATTR(pwm1_enable, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1_ static SENSOR_DEVICE_ATTR(pwm1_min, S_IRUGO, amdgpu_hwmon_get_pwm1_min, NULL, 0); static SENSOR_DEVICE_ATTR(pwm1_max, S_IRUGO, amdgpu_hwmon_get_pwm1_max, NULL, 0); static SENSOR_DEVICE_ATTR(fan1_input, S_IRUGO, amdgpu_hwmon_get_fan1_input, NULL, 0); +static SENSOR_DEVICE_ATTR(in0_input, S_IRUGO, amdgpu_hwmon_show_vddgfx, NULL, 0); +static SENSOR_DEVICE_ATTR(in0_label, S_IRUGO, amdgpu_hwmon_show_vddgfx_label, NULL, 0); +static SENSOR_DEVICE_ATTR(in1_input, S_IRUGO, amdgpu_hwmon_show_vddnb, NULL, 0); +static SENSOR_DEVICE_ATTR(in1_label, S_IRUGO, amdgpu_hwmon_show_vddnb_label, NULL, 0); +static SENSOR_DEVICE_ATTR(power1_average, S_IRUGO, amdgpu_hwmon_show_power_avg, NULL, 0); +static SENSOR_DEVICE_ATTR(power1_cap_max, S_IRUGO, amdgpu_hwmon_show_power_cap_max, NULL, 0); +static SENSOR_DEVICE_ATTR(power1_cap_min, S_IRUGO, amdgpu_hwmon_show_power_cap_min, NULL, 0); +static SENSOR_DEVICE_ATTR(power1_cap, S_IRUGO | S_IWUSR, amdgpu_hwmon_show_power_cap, amdgpu_hwmon_set_power_cap, 0); static struct attribute *hwmon_attributes[] = { &sensor_dev_attr_temp1_input.dev_attr.attr, @@ -939,6 +1137,14 @@ static struct attribute *hwmon_attributes[] = { &sensor_dev_attr_pwm1_min.dev_attr.attr, &sensor_dev_attr_pwm1_max.dev_attr.attr, &sensor_dev_attr_fan1_input.dev_attr.attr, + &sensor_dev_attr_in0_input.dev_attr.attr, + &sensor_dev_attr_in0_label.dev_attr.attr, + &sensor_dev_attr_in1_input.dev_attr.attr, + &sensor_dev_attr_in1_label.dev_attr.attr, + &sensor_dev_attr_power1_average.dev_attr.attr, + &sensor_dev_attr_power1_cap_max.dev_attr.attr, + &sensor_dev_attr_power1_cap_min.dev_attr.attr, + &sensor_dev_attr_power1_cap.dev_attr.attr, NULL }; @@ -949,9 +1155,19 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, struct amdgpu_device *adev = dev_get_drvdata(dev); umode_t effective_mode = attr->mode; - /* no skipping for powerplay */ - if (adev->powerplay.cgs_device) - return effective_mode; + /* handle non-powerplay limitations */ + if (!adev->powerplay.pp_handle) { + /* Skip fan attributes if fan is not present */ + if (adev->pm.no_fan && + (attr == &sensor_dev_attr_pwm1.dev_attr.attr || + attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr || + attr == &sensor_dev_attr_pwm1_max.dev_attr.attr || + attr == &sensor_dev_attr_pwm1_min.dev_attr.attr)) + return 0; + /* requires powerplay */ + if (attr == &sensor_dev_attr_fan1_input.dev_attr.attr) + return 0; + } /* Skip limit attributes if DPM is not enabled */ if (!adev->pm.dpm_enabled && @@ -963,14 +1179,6 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, attr == &sensor_dev_attr_pwm1_min.dev_attr.attr)) return 0; - /* Skip fan attributes if fan is not present */ - if (adev->pm.no_fan && - (attr == &sensor_dev_attr_pwm1.dev_attr.attr || - attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr || - attr == &sensor_dev_attr_pwm1_max.dev_attr.attr || - attr == &sensor_dev_attr_pwm1_min.dev_attr.attr)) - return 0; - /* mask fan attributes if we have no bindings for this asic to expose */ if ((!adev->powerplay.pp_funcs->get_fan_speed_percent && attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't query fan */ @@ -984,6 +1192,12 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't manage state */ effective_mode &= ~S_IWUSR; + if ((adev->flags & AMD_IS_APU) && + (attr == &sensor_dev_attr_power1_cap_max.dev_attr.attr || + attr == &sensor_dev_attr_power1_cap_min.dev_attr.attr|| + attr == &sensor_dev_attr_power1_cap.dev_attr.attr)) + return 0; + /* hide max/min values if we can't both query and manage the fan */ if ((!adev->powerplay.pp_funcs->set_fan_speed_percent && !adev->powerplay.pp_funcs->get_fan_speed_percent) && @@ -991,8 +1205,10 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, attr == &sensor_dev_attr_pwm1_min.dev_attr.attr)) return 0; - /* requires powerplay */ - if (attr == &sensor_dev_attr_fan1_input.dev_attr.attr) + /* only APUs have vddnb */ + if (!(adev->flags & AMD_IS_APU) && + (attr == &sensor_dev_attr_in1_input.dev_attr.attr || + attr == &sensor_dev_attr_in1_label.dev_attr.attr)) return 0; return effective_mode; @@ -1015,13 +1231,15 @@ void amdgpu_dpm_thermal_work_handler(struct work_struct *work) pm.dpm.thermal.work); /* switch to the thermal state */ enum amd_pm_state_type dpm_state = POWER_STATE_TYPE_INTERNAL_THERMAL; + int temp, size = sizeof(temp); if (!adev->pm.dpm_enabled) return; - if (adev->powerplay.pp_funcs->get_temperature) { - int temp = amdgpu_dpm_get_temperature(adev); - + if (adev->powerplay.pp_funcs && + adev->powerplay.pp_funcs->read_sensor && + !amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP, + (void *)&temp, &size)) { if (temp < adev->pm.dpm.thermal.min_temp) /* switch back the user state */ dpm_state = adev->pm.dpm.user_state; @@ -1267,17 +1485,13 @@ void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable) /* enable/disable Low Memory PState for UVD (4k videos) */ if (adev->asic_type == CHIP_STONEY && adev->uvd.decode_image_width >= WIDTH_4K) { - struct pp_hwmgr *hwmgr; - struct pp_instance *pp_handle = - (struct pp_instance *)adev->powerplay.pp_handle; - if (pp_handle) { - hwmgr = pp_handle->hwmgr; - if (hwmgr && hwmgr->hwmgr_func && - hwmgr->hwmgr_func->update_nbdpm_pstate) - hwmgr->hwmgr_func->update_nbdpm_pstate(hwmgr, - !enable, - true); - } + struct pp_hwmgr *hwmgr = + (struct pp_hwmgr *)adev->powerplay.pp_handle; + if (hwmgr && hwmgr->hwmgr_func && + hwmgr->hwmgr_func->update_nbdpm_pstate) + hwmgr->hwmgr_func->update_nbdpm_pstate(hwmgr, + !enable, + true); } } @@ -1336,9 +1550,6 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev) if (adev->pm.dpm_enabled == 0) return 0; - if (adev->powerplay.pp_funcs->get_temperature == NULL) - return 0; - adev->pm.int_hwmon_dev = hwmon_device_register_with_groups(adev->dev, DRIVER_NAME, adev, hwmon_groups); @@ -1408,20 +1619,19 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev) return ret; } ret = device_create_file(adev->dev, - &dev_attr_pp_gfx_power_profile); + &dev_attr_pp_power_profile_mode); if (ret) { DRM_ERROR("failed to create device file " - "pp_gfx_power_profile\n"); + "pp_power_profile_mode\n"); return ret; } ret = device_create_file(adev->dev, - &dev_attr_pp_compute_power_profile); + &dev_attr_pp_od_clk_voltage); if (ret) { DRM_ERROR("failed to create device file " - "pp_compute_power_profile\n"); + "pp_od_clk_voltage\n"); return ret; } - ret = amdgpu_debugfs_pm_init(adev); if (ret) { DRM_ERROR("Failed to register debugfs file for dpm!\n"); @@ -1454,9 +1664,9 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev) device_remove_file(adev->dev, &dev_attr_pp_sclk_od); device_remove_file(adev->dev, &dev_attr_pp_mclk_od); device_remove_file(adev->dev, - &dev_attr_pp_gfx_power_profile); + &dev_attr_pp_power_profile_mode); device_remove_file(adev->dev, - &dev_attr_pp_compute_power_profile); + &dev_attr_pp_od_clk_voltage); } void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) @@ -1479,7 +1689,7 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) } if (adev->powerplay.pp_funcs->dispatch_tasks) { - amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL, NULL); + amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL); } else { mutex_lock(&adev->pm.mutex); adev->pm.dpm.new_active_crtcs = 0; @@ -1529,6 +1739,10 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a seq_printf(m, "\t%u MHz (MCLK)\n", value/100); if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GFX_SCLK, (void *)&value, &size)) seq_printf(m, "\t%u MHz (SCLK)\n", value/100); + if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_STABLE_PSTATE_SCLK, (void *)&value, &size)) + seq_printf(m, "\t%u MHz (PSTATE_SCLK)\n", value/100); + if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_STABLE_PSTATE_MCLK, (void *)&value, &size)) + seq_printf(m, "\t%u MHz (PSTATE_MCLK)\n", value/100); if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDGFX, (void *)&value, &size)) seq_printf(m, "\t%u mV (VDDGFX)\n", value); if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDNB, (void *)&value, &size))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c deleted file mode 100644 index 03f108c..0000000 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c +++ /dev/null
@@ -1,291 +0,0 @@ -/* - * Copyright 2015 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ -#include "atom.h" -#include "amdgpu.h" -#include "amd_shared.h" -#include <linux/module.h> -#include <linux/moduleparam.h> -#include "amdgpu_pm.h" -#include <drm/amdgpu_drm.h> -#include "amdgpu_powerplay.h" -#include "si_dpm.h" -#include "cik_dpm.h" -#include "vi_dpm.h" - -static int amdgpu_pp_early_init(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - struct amd_powerplay *amd_pp; - int ret = 0; - - amd_pp = &(adev->powerplay); - amd_pp->pp_handle = (void *)adev; - - switch (adev->asic_type) { - case CHIP_POLARIS11: - case CHIP_POLARIS10: - case CHIP_POLARIS12: - case CHIP_TONGA: - case CHIP_FIJI: - case CHIP_TOPAZ: - case CHIP_CARRIZO: - case CHIP_STONEY: - case CHIP_VEGA10: - case CHIP_RAVEN: - amd_pp->cgs_device = amdgpu_cgs_create_device(adev); - amd_pp->ip_funcs = &pp_ip_funcs; - amd_pp->pp_funcs = &pp_dpm_funcs; - break; - /* These chips don't have powerplay implemenations */ -#ifdef CONFIG_DRM_AMDGPU_SI - case CHIP_TAHITI: - case CHIP_PITCAIRN: - case CHIP_VERDE: - case CHIP_OLAND: - case CHIP_HAINAN: - amd_pp->ip_funcs = &si_dpm_ip_funcs; - amd_pp->pp_funcs = &si_dpm_funcs; - break; -#endif -#ifdef CONFIG_DRM_AMDGPU_CIK - case CHIP_BONAIRE: - case CHIP_HAWAII: - if (amdgpu_dpm == -1) { - amd_pp->ip_funcs = &ci_dpm_ip_funcs; - amd_pp->pp_funcs = &ci_dpm_funcs; - } else { - amd_pp->cgs_device = amdgpu_cgs_create_device(adev); - amd_pp->ip_funcs = &pp_ip_funcs; - amd_pp->pp_funcs = &pp_dpm_funcs; - } - break; - case CHIP_KABINI: - case CHIP_MULLINS: - case CHIP_KAVERI: - amd_pp->ip_funcs = &kv_dpm_ip_funcs; - amd_pp->pp_funcs = &kv_dpm_funcs; - break; -#endif - default: - ret = -EINVAL; - break; - } - - if (adev->powerplay.ip_funcs->early_init) - ret = adev->powerplay.ip_funcs->early_init( - amd_pp->cgs_device ? amd_pp->cgs_device : - amd_pp->pp_handle); - - return ret; -} - - -static int amdgpu_pp_late_init(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->late_init) - ret = adev->powerplay.ip_funcs->late_init( - adev->powerplay.pp_handle); - - return ret; -} - -static int amdgpu_pp_sw_init(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->sw_init) - ret = adev->powerplay.ip_funcs->sw_init( - adev->powerplay.pp_handle); - - return ret; -} - -static int amdgpu_pp_sw_fini(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->sw_fini) - ret = adev->powerplay.ip_funcs->sw_fini( - adev->powerplay.pp_handle); - if (ret) - return ret; - - return ret; -} - -static int amdgpu_pp_hw_init(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->firmware.load_type == AMDGPU_FW_LOAD_SMU) - amdgpu_ucode_init_bo(adev); - - /* Setting default min and max temperature */ - adev->pm.dpm.thermal.min_temp = MIN_TEMP; - adev->pm.dpm.thermal.max_temp = MAX_TEMP; - - if (adev->powerplay.ip_funcs->hw_init) - ret = adev->powerplay.ip_funcs->hw_init( - adev->powerplay.pp_handle); - - return ret; -} - -static int amdgpu_pp_hw_fini(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->hw_fini) - ret = adev->powerplay.ip_funcs->hw_fini( - adev->powerplay.pp_handle); - - return ret; -} - -static void amdgpu_pp_late_fini(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->late_fini) - adev->powerplay.ip_funcs->late_fini( - adev->powerplay.pp_handle); - - if (adev->powerplay.cgs_device) - amdgpu_cgs_destroy_device(adev->powerplay.cgs_device); -} - -static int amdgpu_pp_suspend(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->suspend) - ret = adev->powerplay.ip_funcs->suspend( - adev->powerplay.pp_handle); - return ret; -} - -static int amdgpu_pp_resume(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->resume) - ret = adev->powerplay.ip_funcs->resume( - adev->powerplay.pp_handle); - return ret; -} - -static int amdgpu_pp_set_clockgating_state(void *handle, - enum amd_clockgating_state state) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->set_clockgating_state) - ret = adev->powerplay.ip_funcs->set_clockgating_state( - adev->powerplay.pp_handle, state); - return ret; -} - -static int amdgpu_pp_set_powergating_state(void *handle, - enum amd_powergating_state state) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->set_powergating_state) - ret = adev->powerplay.ip_funcs->set_powergating_state( - adev->powerplay.pp_handle, state); - return ret; -} - - -static bool amdgpu_pp_is_idle(void *handle) -{ - bool ret = true; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->is_idle) - ret = adev->powerplay.ip_funcs->is_idle( - adev->powerplay.pp_handle); - return ret; -} - -static int amdgpu_pp_wait_for_idle(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->wait_for_idle) - ret = adev->powerplay.ip_funcs->wait_for_idle( - adev->powerplay.pp_handle); - return ret; -} - -static int amdgpu_pp_soft_reset(void *handle) -{ - int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (adev->powerplay.ip_funcs->soft_reset) - ret = adev->powerplay.ip_funcs->soft_reset( - adev->powerplay.pp_handle); - return ret; -} - -static const struct amd_ip_funcs amdgpu_pp_ip_funcs = { - .name = "amdgpu_powerplay", - .early_init = amdgpu_pp_early_init, - .late_init = amdgpu_pp_late_init, - .sw_init = amdgpu_pp_sw_init, - .sw_fini = amdgpu_pp_sw_fini, - .hw_init = amdgpu_pp_hw_init, - .hw_fini = amdgpu_pp_hw_fini, - .late_fini = amdgpu_pp_late_fini, - .suspend = amdgpu_pp_suspend, - .resume = amdgpu_pp_resume, - .is_idle = amdgpu_pp_is_idle, - .wait_for_idle = amdgpu_pp_wait_for_idle, - .soft_reset = amdgpu_pp_soft_reset, - .set_clockgating_state = amdgpu_pp_set_clockgating_state, - .set_powergating_state = amdgpu_pp_set_powergating_state, -}; - -const struct amdgpu_ip_block_version amdgpu_pp_ip_block = -{ - .type = AMD_IP_BLOCK_TYPE_SMC, - .major = 1, - .minor = 0, - .rev = 0, - .funcs = &amdgpu_pp_ip_funcs, -};
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.h deleted file mode 100644 index ebc584d..0000000 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.h +++ /dev/null
@@ -1,36 +0,0 @@ -/* - * Copyright 2015 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ - -#ifndef __AMDGPU_POWERPLAY_H__ -#define __AMDGPU_POWERPLAY_H__ - -#include "amd_shared.h" - -#define MIN_TEMP 0 -#define MAX_TEMP 95 - -extern const struct amdgpu_ip_block_version amdgpu_pp_ip_block; - -#endif /* __AMDGPU_POWERPLAY_H__ */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c index ac71aed..cf0749f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
@@ -30,6 +30,8 @@ #include <drm/amdgpu_drm.h> #include <linux/dma-buf.h> +static const struct dma_buf_ops amdgpu_dmabuf_ops; + struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj) { struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); @@ -103,59 +105,95 @@ amdgpu_gem_prime_import_sg_table(struct drm_device *dev, int ret; ww_mutex_lock(&resv->lock, NULL); - ret = amdgpu_bo_create(adev, attach->dmabuf->size, PAGE_SIZE, false, - AMDGPU_GEM_DOMAIN_GTT, 0, sg, resv, 0, &bo); - ww_mutex_unlock(&resv->lock); + ret = amdgpu_bo_create(adev, attach->dmabuf->size, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_CPU, 0, ttm_bo_type_sg, + resv, &bo); if (ret) - return ERR_PTR(ret); + goto error; - bo->prime_shared_count = 1; + bo->tbo.sg = sg; + bo->tbo.ttm->sg = sg; + bo->allowed_domains = AMDGPU_GEM_DOMAIN_GTT; + bo->preferred_domains = AMDGPU_GEM_DOMAIN_GTT; + if (attach->dmabuf->ops != &amdgpu_dmabuf_ops) + bo->prime_shared_count = 1; + + ww_mutex_unlock(&resv->lock); return &bo->gem_base; + +error: + ww_mutex_unlock(&resv->lock); + return ERR_PTR(ret); } -int amdgpu_gem_prime_pin(struct drm_gem_object *obj) +static int amdgpu_gem_map_attach(struct dma_buf *dma_buf, + struct device *target_dev, + struct dma_buf_attachment *attach) { + struct drm_gem_object *obj = dma_buf->priv; struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); - long ret = 0; + struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); + long r; - ret = amdgpu_bo_reserve(bo, false); - if (unlikely(ret != 0)) - return ret; + r = drm_gem_map_attach(dma_buf, target_dev, attach); + if (r) + return r; - /* - * Wait for all shared fences to complete before we switch to future - * use of exclusive fence on this prime shared bo. - */ - ret = reservation_object_wait_timeout_rcu(bo->tbo.resv, true, false, - MAX_SCHEDULE_TIMEOUT); - if (unlikely(ret < 0)) { - DRM_DEBUG_PRIME("Fence wait failed: %li\n", ret); - amdgpu_bo_unreserve(bo); - return ret; + r = amdgpu_bo_reserve(bo, false); + if (unlikely(r != 0)) + goto error_detach; + + + if (attach->dev->driver != adev->dev->driver) { + /* + * Wait for all shared fences to complete before we switch to future + * use of exclusive fence on this prime shared bo. + */ + r = reservation_object_wait_timeout_rcu(bo->tbo.resv, + true, false, + MAX_SCHEDULE_TIMEOUT); + if (unlikely(r < 0)) { + DRM_DEBUG_PRIME("Fence wait failed: %li\n", r); + goto error_unreserve; + } } /* pin buffer into GTT */ - ret = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL); - if (likely(ret == 0)) + r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL); + if (r) + goto error_unreserve; + + if (attach->dev->driver != adev->dev->driver) bo->prime_shared_count++; +error_unreserve: amdgpu_bo_unreserve(bo); - return ret; + +error_detach: + if (r) + drm_gem_map_detach(dma_buf, attach); + return r; } -void amdgpu_gem_prime_unpin(struct drm_gem_object *obj) +static void amdgpu_gem_map_detach(struct dma_buf *dma_buf, + struct dma_buf_attachment *attach) { + struct drm_gem_object *obj = dma_buf->priv; struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); + struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); int ret = 0; ret = amdgpu_bo_reserve(bo, true); if (unlikely(ret != 0)) - return; + goto error; amdgpu_bo_unpin(bo); - if (bo->prime_shared_count) + if (attach->dev->driver != adev->dev->driver && bo->prime_shared_count) bo->prime_shared_count--; amdgpu_bo_unreserve(bo); + +error: + drm_gem_map_detach(dma_buf, attach); } struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *obj) @@ -194,8 +232,8 @@ static int amdgpu_gem_begin_cpu_access(struct dma_buf *dma_buf, } static const struct dma_buf_ops amdgpu_dmabuf_ops = { - .attach = drm_gem_map_attach, - .detach = drm_gem_map_detach, + .attach = amdgpu_gem_map_attach, + .detach = amdgpu_gem_map_detach, .map_dma_buf = drm_gem_map_dma_buf, .unmap_dma_buf = drm_gem_unmap_dma_buf, .release = drm_gem_dmabuf_release,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 80deddc..28be0aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -51,29 +51,11 @@ static int psp_sw_init(void *handle) switch (adev->asic_type) { case CHIP_VEGA10: - psp->init_microcode = psp_v3_1_init_microcode; - psp->bootloader_load_sysdrv = psp_v3_1_bootloader_load_sysdrv; - psp->bootloader_load_sos = psp_v3_1_bootloader_load_sos; - psp->prep_cmd_buf = psp_v3_1_prep_cmd_buf; - psp->ring_init = psp_v3_1_ring_init; - psp->ring_create = psp_v3_1_ring_create; - psp->ring_stop = psp_v3_1_ring_stop; - psp->ring_destroy = psp_v3_1_ring_destroy; - psp->cmd_submit = psp_v3_1_cmd_submit; - psp->compare_sram_data = psp_v3_1_compare_sram_data; - psp->smu_reload_quirk = psp_v3_1_smu_reload_quirk; - psp->mode1_reset = psp_v3_1_mode1_reset; + case CHIP_VEGA12: + psp_v3_1_set_psp_funcs(psp); break; case CHIP_RAVEN: - psp->init_microcode = psp_v10_0_init_microcode; - psp->prep_cmd_buf = psp_v10_0_prep_cmd_buf; - psp->ring_init = psp_v10_0_ring_init; - psp->ring_create = psp_v10_0_ring_create; - psp->ring_stop = psp_v10_0_ring_stop; - psp->ring_destroy = psp_v10_0_ring_destroy; - psp->cmd_submit = psp_v10_0_cmd_submit; - psp->compare_sram_data = psp_v10_0_compare_sram_data; - psp->mode1_reset = psp_v10_0_mode1_reset; + psp_v10_0_set_psp_funcs(psp); break; default: return -EINVAL; @@ -81,6 +63,9 @@ static int psp_sw_init(void *handle) psp->adev = adev; + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) + return 0; + ret = psp_init_microcode(psp); if (ret) { DRM_ERROR("Failed to load psp firmware!\n"); @@ -94,6 +79,9 @@ static int psp_sw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) + return 0; + release_firmware(adev->psp.sos_fw); adev->psp.sos_fw = NULL; release_firmware(adev->psp.asd_fw); @@ -477,6 +465,9 @@ static int psp_suspend(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; struct psp_context *psp = &adev->psp; + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) + return 0; + ret = psp_ring_stop(psp, PSP_RING_TYPE__KM); if (ret) { DRM_ERROR("PSP ring stop failed\n"); @@ -517,19 +508,11 @@ static int psp_resume(void *handle) return ret; } -static bool psp_check_reset(void* handle) +int psp_gpu_reset(struct amdgpu_device *adev) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) + return 0; - if (adev->flags & AMD_IS_APU) - return true; - - return false; -} - -static int psp_reset(void* handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; return psp_mode1_reset(&adev->psp); } @@ -576,9 +559,9 @@ const struct amd_ip_funcs psp_ip_funcs = { .suspend = psp_suspend, .resume = psp_resume, .is_idle = NULL, - .check_soft_reset = psp_check_reset, + .check_soft_reset = NULL, .wait_for_idle = NULL, - .soft_reset = psp_reset, + .soft_reset = NULL, .set_clockgating_state = psp_set_clockgating_state, .set_powergating_state = psp_set_powergating_state, };
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index ce465455..1292096 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
@@ -33,6 +33,8 @@ #define PSP_ASD_SHARED_MEM_SIZE 0x4000 #define PSP_1_MEG 0x100000 +struct psp_context; + enum psp_ring_type { PSP_RING_TYPE__INVALID = 0, @@ -53,12 +55,8 @@ struct psp_ring uint32_t ring_size; }; -struct psp_context +struct psp_funcs { - struct amdgpu_device *adev; - struct psp_ring km_ring; - struct psp_gfx_cmd_resp *cmd; - int (*init_microcode)(struct psp_context *psp); int (*bootloader_load_sysdrv)(struct psp_context *psp); int (*bootloader_load_sos)(struct psp_context *psp); @@ -77,6 +75,15 @@ struct psp_context enum AMDGPU_UCODE_ID ucode_type); bool (*smu_reload_quirk)(struct psp_context *psp); int (*mode1_reset)(struct psp_context *psp); +}; + +struct psp_context +{ + struct amdgpu_device *adev; + struct psp_ring km_ring; + struct psp_gfx_cmd_resp *cmd; + + const struct psp_funcs *funcs; /* fence buffer */ struct amdgpu_bo *fw_pri_bo; @@ -123,25 +130,25 @@ struct amdgpu_psp_funcs { enum AMDGPU_UCODE_ID); }; -#define psp_prep_cmd_buf(ucode, type) (psp)->prep_cmd_buf((ucode), (type)) -#define psp_ring_init(psp, type) (psp)->ring_init((psp), (type)) -#define psp_ring_create(psp, type) (psp)->ring_create((psp), (type)) -#define psp_ring_stop(psp, type) (psp)->ring_stop((psp), (type)) -#define psp_ring_destroy(psp, type) ((psp)->ring_destroy((psp), (type))) +#define psp_prep_cmd_buf(ucode, type) (psp)->funcs->prep_cmd_buf((ucode), (type)) +#define psp_ring_init(psp, type) (psp)->funcs->ring_init((psp), (type)) +#define psp_ring_create(psp, type) (psp)->funcs->ring_create((psp), (type)) +#define psp_ring_stop(psp, type) (psp)->funcs->ring_stop((psp), (type)) +#define psp_ring_destroy(psp, type) ((psp)->funcs->ring_destroy((psp), (type))) #define psp_cmd_submit(psp, ucode, cmd_mc, fence_mc, index) \ - (psp)->cmd_submit((psp), (ucode), (cmd_mc), (fence_mc), (index)) + (psp)->funcs->cmd_submit((psp), (ucode), (cmd_mc), (fence_mc), (index)) #define psp_compare_sram_data(psp, ucode, type) \ - (psp)->compare_sram_data((psp), (ucode), (type)) + (psp)->funcs->compare_sram_data((psp), (ucode), (type)) #define psp_init_microcode(psp) \ - ((psp)->init_microcode ? (psp)->init_microcode((psp)) : 0) + ((psp)->funcs->init_microcode ? (psp)->funcs->init_microcode((psp)) : 0) #define psp_bootloader_load_sysdrv(psp) \ - ((psp)->bootloader_load_sysdrv ? (psp)->bootloader_load_sysdrv((psp)) : 0) + ((psp)->funcs->bootloader_load_sysdrv ? (psp)->funcs->bootloader_load_sysdrv((psp)) : 0) #define psp_bootloader_load_sos(psp) \ - ((psp)->bootloader_load_sos ? (psp)->bootloader_load_sos((psp)) : 0) + ((psp)->funcs->bootloader_load_sos ? (psp)->funcs->bootloader_load_sos((psp)) : 0) #define psp_smu_reload_quirk(psp) \ - ((psp)->smu_reload_quirk ? (psp)->smu_reload_quirk((psp)) : false) + ((psp)->funcs->smu_reload_quirk ? (psp)->funcs->smu_reload_quirk((psp)) : false) #define psp_mode1_reset(psp) \ - ((psp)->mode1_reset ? (psp)->mode1_reset((psp)) : false) + ((psp)->funcs->mode1_reset ? (psp)->funcs->mode1_reset((psp)) : false) extern const struct amd_ip_funcs psp_ip_funcs; @@ -151,4 +158,6 @@ extern int psp_wait_for(struct psp_context *psp, uint32_t reg_index, extern const struct amdgpu_ip_block_version psp_v10_0_ip_block; +int psp_gpu_reset(struct amdgpu_device *adev); + #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 4e0d8bd..bcd56ad 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -360,6 +360,9 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring) amdgpu_debugfs_ring_fini(ring); + dma_fence_put(ring->vmid_wait); + ring->vmid_wait = NULL; + ring->adev->rings[ring->idx] = NULL; }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index e7db1a8..f4093ab 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
@@ -26,6 +26,7 @@ #include <drm/amdgpu_drm.h> #include <drm/gpu_scheduler.h> +#include <drm/drm_print.h> /* max number of rings */ #define AMDGPU_MAX_RINGS 18 @@ -35,8 +36,9 @@ #define AMDGPU_MAX_UVD_ENC_RINGS 2 /* some special values for the owner field */ -#define AMDGPU_FENCE_OWNER_UNDEFINED ((void*)0ul) -#define AMDGPU_FENCE_OWNER_VM ((void*)1ul) +#define AMDGPU_FENCE_OWNER_UNDEFINED ((void *)0ul) +#define AMDGPU_FENCE_OWNER_VM ((void *)1ul) +#define AMDGPU_FENCE_OWNER_KFD ((void *)2ul) #define AMDGPU_FENCE_FLAG_64BIT (1 << 0) #define AMDGPU_FENCE_FLAG_INT (1 << 1) @@ -128,7 +130,6 @@ struct amdgpu_ring_funcs { void (*emit_vm_flush)(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr); void (*emit_hdp_flush)(struct amdgpu_ring *ring); - void (*emit_hdp_invalidate)(struct amdgpu_ring *ring); void (*emit_gds_switch)(struct amdgpu_ring *ring, uint32_t vmid, uint32_t gds_base, uint32_t gds_size, uint32_t gws_base, uint32_t gws_size, @@ -151,6 +152,8 @@ struct amdgpu_ring_funcs { void (*emit_cntxcntl) (struct amdgpu_ring *ring, uint32_t flags); void (*emit_rreg)(struct amdgpu_ring *ring, uint32_t reg); void (*emit_wreg)(struct amdgpu_ring *ring, uint32_t reg, uint32_t val); + void (*emit_reg_wait)(struct amdgpu_ring *ring, uint32_t reg, + uint32_t val, uint32_t mask); void (*emit_tmz)(struct amdgpu_ring *ring, bool start); /* priority functions */ void (*set_priority) (struct amdgpu_ring *ring, @@ -197,6 +200,7 @@ struct amdgpu_ring { u64 cond_exe_gpu_addr; volatile u32 *cond_exe_cpu_addr; unsigned vm_inv_eng; + struct dma_fence *vmid_wait; bool has_compute_vm_bug; atomic_t num_jobs[DRM_SCHED_PRIORITY_MAX];
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c index 3144400..fb1667b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c
@@ -63,21 +63,27 @@ int amdgpu_sa_bo_manager_init(struct amdgpu_device *adev, for (i = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i) INIT_LIST_HEAD(&sa_manager->flist[i]); - r = amdgpu_bo_create(adev, size, align, true, domain, - 0, NULL, NULL, 0, &sa_manager->bo); + r = amdgpu_bo_create_kernel(adev, size, align, domain, &sa_manager->bo, + &sa_manager->gpu_addr, &sa_manager->cpu_ptr); if (r) { dev_err(adev->dev, "(%d) failed to allocate bo for manager\n", r); return r; } + memset(sa_manager->cpu_ptr, 0, sa_manager->size); return r; } void amdgpu_sa_bo_manager_fini(struct amdgpu_device *adev, - struct amdgpu_sa_manager *sa_manager) + struct amdgpu_sa_manager *sa_manager) { struct amdgpu_sa_bo *sa_bo, *tmp; + if (sa_manager->bo == NULL) { + dev_err(adev->dev, "no bo for sa manager\n"); + return; + } + if (!list_empty(&sa_manager->olist)) { sa_manager->hole = &sa_manager->olist, amdgpu_sa_bo_try_free(sa_manager); @@ -88,57 +94,11 @@ void amdgpu_sa_bo_manager_fini(struct amdgpu_device *adev, list_for_each_entry_safe(sa_bo, tmp, &sa_manager->olist, olist) { amdgpu_sa_bo_remove_locked(sa_bo); } - amdgpu_bo_unref(&sa_manager->bo); + + amdgpu_bo_free_kernel(&sa_manager->bo, &sa_manager->gpu_addr, &sa_manager->cpu_ptr); sa_manager->size = 0; } -int amdgpu_sa_bo_manager_start(struct amdgpu_device *adev, - struct amdgpu_sa_manager *sa_manager) -{ - int r; - - if (sa_manager->bo == NULL) { - dev_err(adev->dev, "no bo for sa manager\n"); - return -EINVAL; - } - - /* map the buffer */ - r = amdgpu_bo_reserve(sa_manager->bo, false); - if (r) { - dev_err(adev->dev, "(%d) failed to reserve manager bo\n", r); - return r; - } - r = amdgpu_bo_pin(sa_manager->bo, sa_manager->domain, &sa_manager->gpu_addr); - if (r) { - amdgpu_bo_unreserve(sa_manager->bo); - dev_err(adev->dev, "(%d) failed to pin manager bo\n", r); - return r; - } - r = amdgpu_bo_kmap(sa_manager->bo, &sa_manager->cpu_ptr); - memset(sa_manager->cpu_ptr, 0, sa_manager->size); - amdgpu_bo_unreserve(sa_manager->bo); - return r; -} - -int amdgpu_sa_bo_manager_suspend(struct amdgpu_device *adev, - struct amdgpu_sa_manager *sa_manager) -{ - int r; - - if (sa_manager->bo == NULL) { - dev_err(adev->dev, "no bo for sa manager\n"); - return -EINVAL; - } - - r = amdgpu_bo_reserve(sa_manager->bo, true); - if (!r) { - amdgpu_bo_kunmap(sa_manager->bo); - amdgpu_bo_unpin(sa_manager->bo); - amdgpu_bo_unreserve(sa_manager->bo); - } - return r; -} - static void amdgpu_sa_bo_remove_locked(struct amdgpu_sa_bo *sa_bo) { struct amdgpu_sa_manager *sa_manager = sa_bo->manager;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c index df65c66..2d6f5ec 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c
@@ -31,6 +31,7 @@ #include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_trace.h" +#include "amdgpu_amdkfd.h" struct amdgpu_sync_entry { struct hlist_node node; @@ -85,11 +86,20 @@ static bool amdgpu_sync_same_dev(struct amdgpu_device *adev, */ static void *amdgpu_sync_get_owner(struct dma_fence *f) { - struct drm_sched_fence *s_fence = to_drm_sched_fence(f); + struct drm_sched_fence *s_fence; + struct amdgpu_amdkfd_fence *kfd_fence; + if (!f) + return AMDGPU_FENCE_OWNER_UNDEFINED; + + s_fence = to_drm_sched_fence(f); if (s_fence) return s_fence->owner; + kfd_fence = to_amdgpu_amdkfd_fence(f); + if (kfd_fence) + return AMDGPU_FENCE_OWNER_KFD; + return AMDGPU_FENCE_OWNER_UNDEFINED; } @@ -204,11 +214,18 @@ int amdgpu_sync_resv(struct amdgpu_device *adev, for (i = 0; i < flist->shared_count; ++i) { f = rcu_dereference_protected(flist->shared[i], reservation_object_held(resv)); + /* We only want to trigger KFD eviction fences on + * evict or move jobs. Skip KFD fences otherwise. + */ + fence_owner = amdgpu_sync_get_owner(f); + if (fence_owner == AMDGPU_FENCE_OWNER_KFD && + owner != AMDGPU_FENCE_OWNER_UNDEFINED) + continue; + if (amdgpu_sync_same_dev(adev, f)) { /* VM updates are only interesting * for other VM updates and moves. */ - fence_owner = amdgpu_sync_get_owner(f); if ((owner != AMDGPU_FENCE_OWNER_UNDEFINED) && (fence_owner != AMDGPU_FENCE_OWNER_UNDEFINED) && ((owner == AMDGPU_FENCE_OWNER_VM) != @@ -305,6 +322,41 @@ struct dma_fence *amdgpu_sync_get_fence(struct amdgpu_sync *sync, bool *explicit return NULL; } +/** + * amdgpu_sync_clone - clone a sync object + * + * @source: sync object to clone + * @clone: pointer to destination sync object + * + * Adds references to all unsignaled fences in @source to @clone. Also + * removes signaled fences from @source while at it. + */ +int amdgpu_sync_clone(struct amdgpu_sync *source, struct amdgpu_sync *clone) +{ + struct amdgpu_sync_entry *e; + struct hlist_node *tmp; + struct dma_fence *f; + int i, r; + + hash_for_each_safe(source->fences, i, tmp, e, node) { + f = e->fence; + if (!dma_fence_is_signaled(f)) { + r = amdgpu_sync_fence(NULL, clone, f, e->explicit); + if (r) + return r; + } else { + hash_del(&e->node); + dma_fence_put(f); + kmem_cache_free(amdgpu_sync_slab, e); + } + } + + dma_fence_put(clone->last_vm_update); + clone->last_vm_update = dma_fence_get(source->last_vm_update); + + return 0; +} + int amdgpu_sync_wait(struct amdgpu_sync *sync, bool intr) { struct amdgpu_sync_entry *e;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h index 7aba38d..10cf23a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h
@@ -50,6 +50,7 @@ int amdgpu_sync_resv(struct amdgpu_device *adev, struct dma_fence *amdgpu_sync_peek_fence(struct amdgpu_sync *sync, struct amdgpu_ring *ring); struct dma_fence *amdgpu_sync_get_fence(struct amdgpu_sync *sync, bool *explicit); +int amdgpu_sync_clone(struct amdgpu_sync *source, struct amdgpu_sync *clone); int amdgpu_sync_wait(struct amdgpu_sync *sync, bool intr); void amdgpu_sync_free(struct amdgpu_sync *sync); int amdgpu_sync_init(void);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c index ed8c373..2dbe875 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
@@ -42,7 +42,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) /* Number of tests = * (Total GTT - IB pool - writeback page - ring buffers) / test size */ - n = adev->mc.gart_size - AMDGPU_IB_POOL_SIZE*64*1024; + n = adev->gmc.gart_size - AMDGPU_IB_POOL_SIZE*64*1024; for (i = 0; i < AMDGPU_MAX_RINGS; ++i) if (adev->rings[i]) n -= adev->rings[i]->ring_size; @@ -59,9 +59,8 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) goto out_cleanup; } - r = amdgpu_bo_create(adev, size, PAGE_SIZE, true, - AMDGPU_GEM_DOMAIN_VRAM, 0, - NULL, NULL, 0, &vram_obj); + r = amdgpu_bo_create(adev, size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, 0, + ttm_bo_type_kernel, NULL, &vram_obj); if (r) { DRM_ERROR("Failed to create VRAM object\n"); goto out_cleanup; @@ -80,9 +79,9 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) void **vram_start, **vram_end; struct dma_fence *fence = NULL; - r = amdgpu_bo_create(adev, size, PAGE_SIZE, true, - AMDGPU_GEM_DOMAIN_GTT, 0, NULL, - NULL, 0, gtt_obj + i); + r = amdgpu_bo_create(adev, size, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_GTT, 0, + ttm_bo_type_kernel, NULL, gtt_obj + i); if (r) { DRM_ERROR("Failed to create GTT object %d\n", i); goto out_lclean; @@ -142,10 +141,10 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) "0x%16llx/0x%16llx)\n", i, *vram_start, gart_start, (unsigned long long) - (gart_addr - adev->mc.gart_start + + (gart_addr - adev->gmc.gart_start + (void*)gart_start - gtt_map), (unsigned long long) - (vram_addr - adev->mc.vram_start + + (vram_addr - adev->gmc.vram_start + (void*)gart_start - gtt_map)); amdgpu_bo_kunmap(vram_obj); goto out_lclean_unpin; @@ -187,10 +186,10 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) "0x%16llx/0x%16llx)\n", i, *gart_start, vram_start, (unsigned long long) - (vram_addr - adev->mc.vram_start + + (vram_addr - adev->gmc.vram_start + (void*)vram_start - vram_map), (unsigned long long) - (gart_addr - adev->mc.gart_start + + (gart_addr - adev->gmc.gart_start + (void*)vram_start - vram_map)); amdgpu_bo_kunmap(gtt_obj[i]); goto out_lclean_unpin; @@ -200,7 +199,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) amdgpu_bo_kunmap(gtt_obj[i]); DRM_INFO("Tested GTT->VRAM and VRAM->GTT copy for GTT offset 0x%llx\n", - gart_addr - adev->mc.gart_start); + gart_addr - adev->gmc.gart_start); continue; out_lclean_unpin:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index cace7a9..532263a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
@@ -86,7 +86,7 @@ TRACE_EVENT(amdgpu_iv, __field(unsigned, vmid_src) __field(uint64_t, timestamp) __field(unsigned, timestamp_src) - __field(unsigned, pas_id) + __field(unsigned, pasid) __array(unsigned, src_data, 4) ), TP_fast_assign( @@ -97,16 +97,16 @@ TRACE_EVENT(amdgpu_iv, __entry->vmid_src = iv->vmid_src; __entry->timestamp = iv->timestamp; __entry->timestamp_src = iv->timestamp_src; - __entry->pas_id = iv->pas_id; + __entry->pasid = iv->pasid; __entry->src_data[0] = iv->src_data[0]; __entry->src_data[1] = iv->src_data[1]; __entry->src_data[2] = iv->src_data[2]; __entry->src_data[3] = iv->src_data[3]; ), - TP_printk("client_id:%u src_id:%u ring:%u vmid:%u timestamp: %llu pas_id:%u src_data: %08x %08x %08x %08x\n", + TP_printk("client_id:%u src_id:%u ring:%u vmid:%u timestamp: %llu pasid:%u src_data: %08x %08x %08x %08x\n", __entry->client_id, __entry->src_id, __entry->ring_id, __entry->vmid, - __entry->timestamp, __entry->pas_id, + __entry->timestamp, __entry->pasid, __entry->src_data[0], __entry->src_data[1], __entry->src_data[2], __entry->src_data[3]) ); @@ -217,7 +217,7 @@ TRACE_EVENT(amdgpu_vm_grab_id, struct amdgpu_job *job), TP_ARGS(vm, ring, job), TP_STRUCT__entry( - __field(struct amdgpu_vm *, vm) + __field(u32, pasid) __field(u32, ring) __field(u32, vmid) __field(u32, vm_hub) @@ -226,15 +226,15 @@ TRACE_EVENT(amdgpu_vm_grab_id, ), TP_fast_assign( - __entry->vm = vm; + __entry->pasid = vm->pasid; __entry->ring = ring->idx; __entry->vmid = job->vmid; __entry->vm_hub = ring->funcs->vmhub, __entry->pd_addr = job->vm_pd_addr; __entry->needs_flush = job->vm_needs_flush; ), - TP_printk("vm=%p, ring=%u, id=%u, hub=%u, pd_addr=%010Lx needs_flush=%u", - __entry->vm, __entry->ring, __entry->vmid, + TP_printk("pasid=%d, ring=%u, id=%u, hub=%u, pd_addr=%010Lx needs_flush=%u", + __entry->pasid, __entry->ring, __entry->vmid, __entry->vm_hub, __entry->pd_addr, __entry->needs_flush) ); @@ -378,6 +378,28 @@ TRACE_EVENT(amdgpu_vm_flush, __entry->vm_hub,__entry->pd_addr) ); +DECLARE_EVENT_CLASS(amdgpu_pasid, + TP_PROTO(unsigned pasid), + TP_ARGS(pasid), + TP_STRUCT__entry( + __field(unsigned, pasid) + ), + TP_fast_assign( + __entry->pasid = pasid; + ), + TP_printk("pasid=%u", __entry->pasid) +); + +DEFINE_EVENT(amdgpu_pasid, amdgpu_pasid_allocated, + TP_PROTO(unsigned pasid), + TP_ARGS(pasid) +); + +DEFINE_EVENT(amdgpu_pasid, amdgpu_pasid_freed, + TP_PROTO(unsigned pasid), + TP_ARGS(pasid) +); + TRACE_EVENT(amdgpu_bo_list_set, TP_PROTO(struct amdgpu_bo_list *list, struct amdgpu_bo *bo), TP_ARGS(list, bo),
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 2e74357..89c23aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -46,6 +46,7 @@ #include "amdgpu.h" #include "amdgpu_object.h" #include "amdgpu_trace.h" +#include "amdgpu_amdkfd.h" #include "bif/bif_4_1_d.h" #define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT) @@ -161,7 +162,7 @@ static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, break; case TTM_PL_TT: man->func = &amdgpu_gtt_mgr_func; - man->gpu_offset = adev->mc.gart_start; + man->gpu_offset = adev->gmc.gart_start; man->available_caching = TTM_PL_MASK_CACHING; man->default_caching = TTM_PL_FLAG_CACHED; man->flags = TTM_MEMTYPE_FLAG_MAPPABLE | TTM_MEMTYPE_FLAG_CMA; @@ -169,7 +170,7 @@ static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, case TTM_PL_VRAM: /* "On-card" video ram */ man->func = &amdgpu_vram_mgr_func; - man->gpu_offset = adev->mc.vram_start; + man->gpu_offset = adev->gmc.vram_start; man->flags = TTM_MEMTYPE_FLAG_FIXED | TTM_MEMTYPE_FLAG_MAPPABLE; man->available_caching = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_WC; @@ -203,6 +204,12 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, .flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM }; + if (bo->type == ttm_bo_type_sg) { + placement->num_placement = 0; + placement->num_busy_placement = 0; + return; + } + if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) { placement->placement = &placements; placement->busy_placement = &placements; @@ -213,13 +220,11 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, abo = ttm_to_amdgpu_bo(bo); switch (bo->mem.mem_type) { case TTM_PL_VRAM: - if (adev->mman.buffer_funcs && - adev->mman.buffer_funcs_ring && - adev->mman.buffer_funcs_ring->ready == false) { + if (!adev->mman.buffer_funcs_enabled) { amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU); - } else if (adev->mc.visible_vram_size < adev->mc.real_vram_size && + } else if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size && !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)) { - unsigned fpfn = adev->mc.visible_vram_size >> PAGE_SHIFT; + unsigned fpfn = adev->gmc.visible_vram_size >> PAGE_SHIFT; struct drm_mm_node *node = bo->mem.mm_node; unsigned long pages_left; @@ -260,6 +265,13 @@ static int amdgpu_verify_access(struct ttm_buffer_object *bo, struct file *filp) { struct amdgpu_bo *abo = ttm_to_amdgpu_bo(bo); + /* + * Don't verify access for KFD BOs. They don't have a GEM + * object associated with them. + */ + if (abo->kfd_bo) + return 0; + if (amdgpu_ttm_tt_get_usermm(bo->ttm)) return -EPERM; return drm_vma_node_verify_access(&abo->gem_base.vma_node, @@ -331,7 +343,7 @@ int amdgpu_ttm_copy_mem_to_mem(struct amdgpu_device *adev, const uint64_t GTT_MAX_BYTES = (AMDGPU_GTT_MAX_TRANSFER_SIZE * AMDGPU_GPU_PAGE_SIZE); - if (!ring->ready) { + if (!adev->mman.buffer_funcs_enabled) { DRM_ERROR("Trying to move memory with ring turned off.\n"); return -EINVAL; } @@ -577,12 +589,9 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, amdgpu_move_null(bo, new_mem); return 0; } - if (adev->mman.buffer_funcs == NULL || - adev->mman.buffer_funcs_ring == NULL || - !adev->mman.buffer_funcs_ring->ready) { - /* use memcpy */ + + if (!adev->mman.buffer_funcs_enabled) goto memcpy; - } if (old_mem->mem_type == TTM_PL_VRAM && new_mem->mem_type == TTM_PL_SYSTEM) { @@ -621,6 +630,7 @@ static int amdgpu_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_ { struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type]; struct amdgpu_device *adev = amdgpu_ttm_adev(bdev); + struct drm_mm_node *mm_node = mem->mm_node; mem->bus.addr = NULL; mem->bus.offset = 0; @@ -638,9 +648,18 @@ static int amdgpu_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_ case TTM_PL_VRAM: mem->bus.offset = mem->start << PAGE_SHIFT; /* check if it's visible */ - if ((mem->bus.offset + mem->bus.size) > adev->mc.visible_vram_size) + if ((mem->bus.offset + mem->bus.size) > adev->gmc.visible_vram_size) return -EINVAL; - mem->bus.base = adev->mc.aper_base; + /* Only physically contiguous buffers apply. In a contiguous + * buffer, size of the first mm_node would match the number of + * pages in ttm_mem_reg. + */ + if (adev->mman.aper_base_kaddr && + (mm_node->size == mem->num_pages)) + mem->bus.addr = (u8 *)adev->mman.aper_base_kaddr + + mem->bus.offset; + + mem->bus.base = adev->gmc.aper_base; mem->bus.is_iomem = true; break; default: @@ -674,7 +693,6 @@ struct amdgpu_ttm_gup_task_list { struct amdgpu_ttm_tt { struct ttm_dma_tt ttm; - struct amdgpu_device *adev; u64 offset; uint64_t userptr; struct mm_struct *usermm; @@ -832,6 +850,7 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm) static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem) { + struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev); struct amdgpu_ttm_tt *gtt = (void*)ttm; uint64_t flags; int r = 0; @@ -858,9 +877,9 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, return 0; } - flags = amdgpu_ttm_tt_pte_flags(gtt->adev, ttm, bo_mem); + flags = amdgpu_ttm_tt_pte_flags(adev, ttm, bo_mem); gtt->offset = (u64)bo_mem->start << PAGE_SHIFT; - r = amdgpu_gart_bind(gtt->adev, gtt->offset, ttm->num_pages, + r = amdgpu_gart_bind(adev, gtt->offset, ttm->num_pages, ttm->pages, gtt->ttm.dma_address, flags); if (r) @@ -892,7 +911,7 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo) placement.num_busy_placement = 1; placement.busy_placement = &placements; placements.fpfn = 0; - placements.lpfn = adev->mc.gart_size >> PAGE_SHIFT; + placements.lpfn = adev->gmc.gart_size >> PAGE_SHIFT; placements.flags = (bo->mem.placement & ~TTM_PL_MASK_MEM) | TTM_PL_FLAG_TT; @@ -938,6 +957,7 @@ int amdgpu_ttm_recover_gart(struct ttm_buffer_object *tbo) static int amdgpu_ttm_backend_unbind(struct ttm_tt *ttm) { + struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev); struct amdgpu_ttm_tt *gtt = (void *)ttm; int r; @@ -948,7 +968,7 @@ static int amdgpu_ttm_backend_unbind(struct ttm_tt *ttm) return 0; /* unbind shouldn't be done for GDS/GWS/OA in ttm_bo_clean_mm */ - r = amdgpu_gart_unbind(gtt->adev, gtt->offset, ttm->num_pages); + r = amdgpu_gart_unbind(adev, gtt->offset, ttm->num_pages); if (r) DRM_ERROR("failed to unbind %lu pages at 0x%08llX\n", gtt->ttm.ttm.num_pages, gtt->offset); @@ -969,22 +989,20 @@ static struct ttm_backend_func amdgpu_backend_func = { .destroy = &amdgpu_ttm_backend_destroy, }; -static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_bo_device *bdev, - unsigned long size, uint32_t page_flags, - struct page *dummy_read_page) +static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_buffer_object *bo, + uint32_t page_flags) { struct amdgpu_device *adev; struct amdgpu_ttm_tt *gtt; - adev = amdgpu_ttm_adev(bdev); + adev = amdgpu_ttm_adev(bo->bdev); gtt = kzalloc(sizeof(struct amdgpu_ttm_tt), GFP_KERNEL); if (gtt == NULL) { return NULL; } gtt->ttm.ttm.func = &amdgpu_backend_func; - gtt->adev = adev; - if (ttm_dma_tt_init(>t->ttm, bdev, size, page_flags, dummy_read_page)) { + if (ttm_sg_tt_init(>t->ttm, bo, page_flags)) { kfree(gtt); return NULL; } @@ -998,9 +1016,6 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm, struct amdgpu_ttm_tt *gtt = (void *)ttm; bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG); - if (ttm->state != tt_unpopulated) - return 0; - if (gtt && gtt->userptr) { ttm->sg = kzalloc(sizeof(struct sg_table), GFP_KERNEL); if (!ttm->sg) @@ -1013,13 +1028,14 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm, if (slave && ttm->sg) { drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages, - gtt->ttm.dma_address, ttm->num_pages); + gtt->ttm.dma_address, + ttm->num_pages); ttm->state = tt_unbound; return 0; } #ifdef CONFIG_SWIOTLB - if (swiotlb_nr_tbl()) { + if (adev->need_swiotlb && swiotlb_nr_tbl()) { return ttm_dma_populate(>t->ttm, adev->dev, ctx); } #endif @@ -1046,7 +1062,7 @@ static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm) adev = amdgpu_ttm_adev(ttm->bdev); #ifdef CONFIG_SWIOTLB - if (swiotlb_nr_tbl()) { + if (adev->need_swiotlb && swiotlb_nr_tbl()) { ttm_dma_unpopulate(>t->ttm, adev->dev); return; } @@ -1171,6 +1187,23 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo, { unsigned long num_pages = bo->mem.num_pages; struct drm_mm_node *node = bo->mem.mm_node; + struct reservation_object_list *flist; + struct dma_fence *f; + int i; + + /* If bo is a KFD BO, check if the bo belongs to the current process. + * If true, then return false as any KFD process needs all its BOs to + * be resident to run successfully + */ + flist = reservation_object_get_list(bo->resv); + if (flist) { + for (i = 0; i < flist->shared_count; ++i) { + f = rcu_dereference_protected(flist->shared[i], + reservation_object_held(bo->resv)); + if (amdkfd_fence_check_mm(f, current->mm)) + return false; + } + } switch (bo->mem.mem_type) { case TTM_PL_TT: @@ -1213,7 +1246,7 @@ static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo, nodes = amdgpu_find_mm_node(&abo->tbo.mem, &offset); pos = (nodes->start << PAGE_SHIFT) + offset; - while (len && pos < adev->mc.mc_vram_size) { + while (len && pos < adev->gmc.mc_vram_size) { uint64_t aligned_pos = pos & ~(uint64_t)3; uint32_t bytes = 4 - (pos & 3); uint32_t shift = (pos & 3) * 8; @@ -1299,7 +1332,7 @@ static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev) struct ttm_operation_ctx ctx = { false, false }; int r = 0; int i; - u64 vram_size = adev->mc.visible_vram_size; + u64 vram_size = adev->gmc.visible_vram_size; u64 offset = adev->fw_vram_usage.start_offset; u64 size = adev->fw_vram_usage.size; struct amdgpu_bo *bo; @@ -1310,11 +1343,12 @@ static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev) if (adev->fw_vram_usage.size > 0 && adev->fw_vram_usage.size <= vram_size) { - r = amdgpu_bo_create(adev, adev->fw_vram_usage.size, - PAGE_SIZE, true, AMDGPU_GEM_DOMAIN_VRAM, - AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | - AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, NULL, NULL, 0, - &adev->fw_vram_usage.reserved_bo); + r = amdgpu_bo_create(adev, adev->fw_vram_usage.size, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | + AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, + ttm_bo_type_kernel, NULL, + &adev->fw_vram_usage.reserved_bo); if (r) goto error_create; @@ -1388,8 +1422,12 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) return r; } adev->mman.initialized = true; + + /* We opt to avoid OOM on system pages allocations */ + adev->mman.bdev.no_retry = true; + r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_VRAM, - adev->mc.real_vram_size >> PAGE_SHIFT); + adev->gmc.real_vram_size >> PAGE_SHIFT); if (r) { DRM_ERROR("Failed initializing VRAM heap.\n"); return r; @@ -1398,11 +1436,15 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) /* Reduce size of CPU-visible VRAM if requested */ vis_vram_limit = (u64)amdgpu_vis_vram_limit * 1024 * 1024; if (amdgpu_vis_vram_limit > 0 && - vis_vram_limit <= adev->mc.visible_vram_size) - adev->mc.visible_vram_size = vis_vram_limit; + vis_vram_limit <= adev->gmc.visible_vram_size) + adev->gmc.visible_vram_size = vis_vram_limit; /* Change the size here instead of the init above so only lpfn is affected */ - amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, false); +#ifdef CONFIG_64BIT + adev->mman.aper_base_kaddr = ioremap_wc(adev->gmc.aper_base, + adev->gmc.visible_vram_size); +#endif /* *The reserved vram for firmware must be pinned to the specified @@ -1413,8 +1455,8 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) return r; } - if (adev->mc.stolen_size) { - r = amdgpu_bo_create_kernel(adev, adev->mc.stolen_size, PAGE_SIZE, + if (adev->gmc.stolen_size) { + r = amdgpu_bo_create_kernel(adev, adev->gmc.stolen_size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, &adev->stolen_vga_memory, NULL, NULL); @@ -1423,14 +1465,14 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) } DRM_INFO("amdgpu: %uM of VRAM memory ready\n", - (unsigned) (adev->mc.real_vram_size / (1024 * 1024))); + (unsigned) (adev->gmc.real_vram_size / (1024 * 1024))); if (amdgpu_gtt_size == -1) { struct sysinfo si; si_meminfo(&si); gtt_size = min(max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), - adev->mc.mc_vram_size), + adev->gmc.mc_vram_size), ((uint64_t)si.totalram * si.mem_unit * 3/4)); } else @@ -1502,6 +1544,9 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev) amdgpu_ttm_debugfs_fini(adev); amdgpu_ttm_fw_reserve_vram_fini(adev); + if (adev->mman.aper_base_kaddr) + iounmap(adev->mman.aper_base_kaddr); + adev->mman.aper_base_kaddr = NULL; ttm_bo_clean_mm(&adev->mman.bdev, TTM_PL_VRAM); ttm_bo_clean_mm(&adev->mman.bdev, TTM_PL_TT); @@ -1517,18 +1562,30 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev) DRM_INFO("amdgpu: ttm finalized\n"); } -/* this should only be called at bootup or when userspace - * isn't running */ -void amdgpu_ttm_set_active_vram_size(struct amdgpu_device *adev, u64 size) +/** + * amdgpu_ttm_set_buffer_funcs_status - enable/disable use of buffer functions + * + * @adev: amdgpu_device pointer + * @enable: true when we can use buffer functions. + * + * Enable/disable use of buffer functions during suspend/resume. This should + * only be called at bootup or when userspace isn't running. + */ +void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev, bool enable) { - struct ttm_mem_type_manager *man; + struct ttm_mem_type_manager *man = &adev->mman.bdev.man[TTM_PL_VRAM]; + uint64_t size; - if (!adev->mman.initialized) + if (!adev->mman.initialized || adev->in_gpu_reset) return; - man = &adev->mman.bdev.man[TTM_PL_VRAM]; /* this just adjusts TTM size idea, which sets lpfn to the correct value */ + if (enable) + size = adev->gmc.real_vram_size; + else + size = adev->gmc.visible_vram_size; man->size = size >> PAGE_SHIFT; + adev->mman.buffer_funcs_enabled = enable; } int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma) @@ -1567,7 +1624,7 @@ static int amdgpu_map_buffer(struct ttm_buffer_object *bo, BUG_ON(adev->mman.buffer_funcs->copy_max_bytes < AMDGPU_GTT_MAX_TRANSFER_SIZE * 8); - *addr = adev->mc.gart_start; + *addr = adev->gmc.gart_start; *addr += (u64)window * AMDGPU_GTT_MAX_TRANSFER_SIZE * AMDGPU_GPU_PAGE_SIZE; @@ -1627,6 +1684,11 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, unsigned i; int r; + if (direct_submit && !ring->ready) { + DRM_ERROR("Trying to move memory with ring turned off.\n"); + return -EINVAL; + } + max_bytes = adev->mman.buffer_funcs->copy_max_bytes; num_loops = DIV_ROUND_UP(byte_count, max_bytes); num_dw = num_loops * adev->mman.buffer_funcs->copy_num_dw; @@ -1685,13 +1747,12 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, } int amdgpu_fill_buffer(struct amdgpu_bo *bo, - uint64_t src_data, + uint32_t src_data, struct reservation_object *resv, struct dma_fence **fence) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - uint32_t max_bytes = 8 * - adev->vm_manager.vm_pte_funcs->set_max_nums_pte_pde; + uint32_t max_bytes = adev->mman.buffer_funcs->fill_max_bytes; struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring; struct drm_mm_node *mm_node; @@ -1701,7 +1762,7 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, struct amdgpu_job *job; int r; - if (!ring->ready) { + if (!adev->mman.buffer_funcs_enabled) { DRM_ERROR("Trying to clear memory with ring turned off.\n"); return -EINVAL; } @@ -1722,9 +1783,7 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, num_pages -= mm_node->size; ++mm_node; } - - /* num of dwords for each SDMA_OP_PTEPDE cmd */ - num_dw = num_loops * adev->vm_manager.vm_pte_funcs->set_pte_pde_num_dw; + num_dw = num_loops * adev->mman.buffer_funcs->fill_num_dw; /* for IB padding */ num_dw += 64; @@ -1749,16 +1808,12 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, uint32_t byte_count = mm_node->size << PAGE_SHIFT; uint64_t dst_addr; - WARN_ONCE(byte_count & 0x7, "size should be a multiple of 8"); - dst_addr = amdgpu_mm_node_addr(&bo->tbo, mm_node, &bo->tbo.mem); while (byte_count) { uint32_t cur_size_in_bytes = min(byte_count, max_bytes); - amdgpu_vm_set_pte_pde(adev, &job->ibs[0], - dst_addr, 0, - cur_size_in_bytes >> 3, 0, - src_data); + amdgpu_emit_fill_buffer(adev, &job->ibs[0], src_data, + dst_addr, cur_size_in_bytes); dst_addr += cur_size_in_bytes; byte_count -= cur_size_in_bytes; @@ -1819,14 +1874,14 @@ static ssize_t amdgpu_ttm_vram_read(struct file *f, char __user *buf, if (size & 0x3 || *pos & 0x3) return -EINVAL; - if (*pos >= adev->mc.mc_vram_size) + if (*pos >= adev->gmc.mc_vram_size) return -ENXIO; while (size) { unsigned long flags; uint32_t value; - if (*pos >= adev->mc.mc_vram_size) + if (*pos >= adev->gmc.mc_vram_size) return result; spin_lock_irqsave(&adev->mmio_idx_lock, flags); @@ -1858,14 +1913,14 @@ static ssize_t amdgpu_ttm_vram_write(struct file *f, const char __user *buf, if (size & 0x3 || *pos & 0x3) return -EINVAL; - if (*pos >= adev->mc.mc_vram_size) + if (*pos >= adev->gmc.mc_vram_size) return -ENXIO; while (size) { unsigned long flags; uint32_t value; - if (*pos >= adev->mc.mc_vram_size) + if (*pos >= adev->gmc.mc_vram_size) return result; r = get_user(value, (uint32_t *)buf); @@ -1943,38 +1998,98 @@ static const struct file_operations amdgpu_ttm_gtt_fops = { #endif -static ssize_t amdgpu_iova_to_phys_read(struct file *f, char __user *buf, - size_t size, loff_t *pos) +static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) { struct amdgpu_device *adev = file_inode(f)->i_private; - int r; - uint64_t phys; struct iommu_domain *dom; - - // always return 8 bytes - if (size != 8) - return -EINVAL; - - // only accept page addresses - if (*pos & 0xFFF) - return -EINVAL; + ssize_t result = 0; + int r; dom = iommu_get_domain_for_dev(adev->dev); - if (dom) - phys = iommu_iova_to_phys(dom, *pos); - else - phys = *pos; - r = copy_to_user(buf, &phys, 8); - if (r) - return -EFAULT; + while (size) { + phys_addr_t addr = *pos & PAGE_MASK; + loff_t off = *pos & ~PAGE_MASK; + size_t bytes = PAGE_SIZE - off; + unsigned long pfn; + struct page *p; + void *ptr; - return 8; + bytes = bytes < size ? bytes : size; + + addr = dom ? iommu_iova_to_phys(dom, addr) : addr; + + pfn = addr >> PAGE_SHIFT; + if (!pfn_valid(pfn)) + return -EPERM; + + p = pfn_to_page(pfn); + if (p->mapping != adev->mman.bdev.dev_mapping) + return -EPERM; + + ptr = kmap(p); + r = copy_to_user(buf, ptr + off, bytes); + kunmap(p); + if (r) + return -EFAULT; + + size -= bytes; + *pos += bytes; + result += bytes; + } + + return result; } -static const struct file_operations amdgpu_ttm_iova_fops = { +static ssize_t amdgpu_iomem_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_device *adev = file_inode(f)->i_private; + struct iommu_domain *dom; + ssize_t result = 0; + int r; + + dom = iommu_get_domain_for_dev(adev->dev); + + while (size) { + phys_addr_t addr = *pos & PAGE_MASK; + loff_t off = *pos & ~PAGE_MASK; + size_t bytes = PAGE_SIZE - off; + unsigned long pfn; + struct page *p; + void *ptr; + + bytes = bytes < size ? bytes : size; + + addr = dom ? iommu_iova_to_phys(dom, addr) : addr; + + pfn = addr >> PAGE_SHIFT; + if (!pfn_valid(pfn)) + return -EPERM; + + p = pfn_to_page(pfn); + if (p->mapping != adev->mman.bdev.dev_mapping) + return -EPERM; + + ptr = kmap(p); + r = copy_from_user(ptr + off, buf, bytes); + kunmap(p); + if (r) + return -EFAULT; + + size -= bytes; + *pos += bytes; + result += bytes; + } + + return result; +} + +static const struct file_operations amdgpu_ttm_iomem_fops = { .owner = THIS_MODULE, - .read = amdgpu_iova_to_phys_read, + .read = amdgpu_iomem_read, + .write = amdgpu_iomem_write, .llseek = default_llseek }; @@ -1987,7 +2102,7 @@ static const struct { #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS { "amdgpu_gtt", &amdgpu_ttm_gtt_fops, TTM_PL_TT }, #endif - { "amdgpu_iova", &amdgpu_ttm_iova_fops, TTM_PL_SYSTEM }, + { "amdgpu_iomem", &amdgpu_ttm_iomem_fops, TTM_PL_SYSTEM }, }; #endif @@ -2009,16 +2124,16 @@ static int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev) if (IS_ERR(ent)) return PTR_ERR(ent); if (ttm_debugfs_entries[count].domain == TTM_PL_VRAM) - i_size_write(ent->d_inode, adev->mc.mc_vram_size); + i_size_write(ent->d_inode, adev->gmc.mc_vram_size); else if (ttm_debugfs_entries[count].domain == TTM_PL_TT) - i_size_write(ent->d_inode, adev->mc.gart_size); + i_size_write(ent->d_inode, adev->gmc.gart_size); adev->mman.debugfs_entries[count] = ent; } count = ARRAY_SIZE(amdgpu_ttm_debugfs_list); #ifdef CONFIG_SWIOTLB - if (!swiotlb_nr_tbl()) + if (!(adev->need_swiotlb && swiotlb_nr_tbl())) --count; #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index 1bfc46db..685a7b8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
@@ -44,6 +44,7 @@ struct amdgpu_mman { struct ttm_bo_device bdev; bool mem_global_referenced; bool initialized; + void __iomem *aper_base_kaddr; #if defined(CONFIG_DEBUG_FS) struct dentry *debugfs_entries[8]; @@ -52,6 +53,7 @@ struct amdgpu_mman { /* buffer handling */ const struct amdgpu_buffer_funcs *buffer_funcs; struct amdgpu_ring *buffer_funcs_ring; + bool buffer_funcs_enabled; struct mutex gtt_window_lock; /* Scheduler entity for buffer moves */ @@ -75,6 +77,10 @@ u64 amdgpu_vram_mgr_bo_invisible_size(struct amdgpu_bo *bo); uint64_t amdgpu_vram_mgr_usage(struct ttm_mem_type_manager *man); uint64_t amdgpu_vram_mgr_vis_usage(struct ttm_mem_type_manager *man); +int amdgpu_ttm_init(struct amdgpu_device *adev); +void amdgpu_ttm_fini(struct amdgpu_device *adev); +void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev, + bool enable); void amdgpu_ttm_late_init(struct amdgpu_device *adev); int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, @@ -88,7 +94,7 @@ int amdgpu_ttm_copy_mem_to_mem(struct amdgpu_device *adev, struct reservation_object *resv, struct dma_fence **f); int amdgpu_fill_buffer(struct amdgpu_bo *bo, - uint64_t src_data, + uint32_t src_data, struct reservation_object *resv, struct dma_fence **fence);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c index 180d58b6..2786847 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -272,6 +272,7 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type) return AMDGPU_FW_LOAD_SMU; case CHIP_VEGA10: case CHIP_RAVEN: + case CHIP_VEGA12: if (!load_type) return AMDGPU_FW_LOAD_DIRECT; else
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c index 85d1d7b..c643ca7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -68,6 +68,7 @@ #define FIRMWARE_POLARIS12 "amdgpu/polaris12_uvd.bin" #define FIRMWARE_VEGA10 "amdgpu/vega10_uvd.bin" +#define FIRMWARE_VEGA12 "amdgpu/vega12_uvd.bin" #define mmUVD_GPCOM_VCPU_DATA0_VEGA10 (0x03c4 + 0x7e00) #define mmUVD_GPCOM_VCPU_DATA1_VEGA10 (0x03c5 + 0x7e00) @@ -110,6 +111,7 @@ MODULE_FIRMWARE(FIRMWARE_POLARIS11); MODULE_FIRMWARE(FIRMWARE_POLARIS12); MODULE_FIRMWARE(FIRMWARE_VEGA10); +MODULE_FIRMWARE(FIRMWARE_VEGA12); static void amdgpu_uvd_idle_work_handler(struct work_struct *work); @@ -161,11 +163,14 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev) case CHIP_POLARIS11: fw_name = FIRMWARE_POLARIS11; break; + case CHIP_POLARIS12: + fw_name = FIRMWARE_POLARIS12; + break; case CHIP_VEGA10: fw_name = FIRMWARE_VEGA10; break; - case CHIP_POLARIS12: - fw_name = FIRMWARE_POLARIS12; + case CHIP_VEGA12: + fw_name = FIRMWARE_VEGA12; break; default: return -EINVAL; @@ -957,37 +962,28 @@ int amdgpu_uvd_ring_parse_cs(struct amdgpu_cs_parser *parser, uint32_t ib_idx) static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo, bool direct, struct dma_fence **fence) { - struct ttm_operation_ctx ctx = { true, false }; - struct ttm_validate_buffer tv; - struct ww_acquire_ctx ticket; - struct list_head head; + struct amdgpu_device *adev = ring->adev; + struct dma_fence *f = NULL; struct amdgpu_job *job; struct amdgpu_ib *ib; - struct dma_fence *f = NULL; - struct amdgpu_device *adev = ring->adev; - uint64_t addr; uint32_t data[4]; - int i, r; + uint64_t addr; + long r; + int i; - memset(&tv, 0, sizeof(tv)); - tv.bo = &bo->tbo; - - INIT_LIST_HEAD(&head); - list_add(&tv.head, &head); - - r = ttm_eu_reserve_buffers(&ticket, &head, true, NULL); - if (r) - return r; + amdgpu_bo_kunmap(bo); + amdgpu_bo_unpin(bo); if (!ring->adev->uvd.address_64_bit) { + struct ttm_operation_ctx ctx = { true, false }; + amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM); amdgpu_uvd_force_into_uvd_segment(bo); + r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); + if (r) + goto err; } - r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); - if (r) - goto err; - r = amdgpu_job_alloc_with_ib(adev, 64, &job); if (r) goto err; @@ -1019,6 +1015,14 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo, ib->length_dw = 16; if (direct) { + r = reservation_object_wait_timeout_rcu(bo->tbo.resv, + true, false, + msecs_to_jiffies(10)); + if (r == 0) + r = -ETIMEDOUT; + if (r < 0) + goto err_free; + r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); job->fence = dma_fence_get(f); if (r) @@ -1026,17 +1030,23 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo, amdgpu_job_free(job); } else { + r = amdgpu_sync_resv(adev, &job->sync, bo->tbo.resv, + AMDGPU_FENCE_OWNER_UNDEFINED, false); + if (r) + goto err_free; + r = amdgpu_job_submit(job, ring, &adev->uvd.entity, AMDGPU_FENCE_OWNER_UNDEFINED, &f); if (r) goto err_free; } - ttm_eu_fence_buffer_objects(&ticket, &head, f); + amdgpu_bo_fence(bo, f, false); + amdgpu_bo_unreserve(bo); + amdgpu_bo_unref(&bo); if (fence) *fence = dma_fence_get(f); - amdgpu_bo_unref(&bo); dma_fence_put(f); return 0; @@ -1045,7 +1055,8 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo, amdgpu_job_free(job); err: - ttm_eu_backoff_reservation(&ticket, &head); + amdgpu_bo_unreserve(bo); + amdgpu_bo_unref(&bo); return r; } @@ -1056,31 +1067,16 @@ int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle, struct dma_fence **fence) { struct amdgpu_device *adev = ring->adev; - struct amdgpu_bo *bo; + struct amdgpu_bo *bo = NULL; uint32_t *msg; int r, i; - r = amdgpu_bo_create(adev, 1024, PAGE_SIZE, true, - AMDGPU_GEM_DOMAIN_VRAM, - AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | - AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, - NULL, NULL, 0, &bo); + r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &bo, NULL, (void **)&msg); if (r) return r; - r = amdgpu_bo_reserve(bo, false); - if (r) { - amdgpu_bo_unref(&bo); - return r; - } - - r = amdgpu_bo_kmap(bo, (void **)&msg); - if (r) { - amdgpu_bo_unreserve(bo); - amdgpu_bo_unref(&bo); - return r; - } - /* stitch together an UVD create msg */ msg[0] = cpu_to_le32(0x00000de4); msg[1] = cpu_to_le32(0x00000000); @@ -1096,9 +1092,6 @@ int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle, for (i = 11; i < 1024; ++i) msg[i] = cpu_to_le32(0x0); - amdgpu_bo_kunmap(bo); - amdgpu_bo_unreserve(bo); - return amdgpu_uvd_send_msg(ring, bo, true, fence); } @@ -1106,31 +1099,16 @@ int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle, bool direct, struct dma_fence **fence) { struct amdgpu_device *adev = ring->adev; - struct amdgpu_bo *bo; + struct amdgpu_bo *bo = NULL; uint32_t *msg; int r, i; - r = amdgpu_bo_create(adev, 1024, PAGE_SIZE, true, - AMDGPU_GEM_DOMAIN_VRAM, - AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | - AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, - NULL, NULL, 0, &bo); + r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &bo, NULL, (void **)&msg); if (r) return r; - r = amdgpu_bo_reserve(bo, false); - if (r) { - amdgpu_bo_unref(&bo); - return r; - } - - r = amdgpu_bo_kmap(bo, (void **)&msg); - if (r) { - amdgpu_bo_unreserve(bo); - amdgpu_bo_unref(&bo); - return r; - } - /* stitch together an UVD destroy msg */ msg[0] = cpu_to_le32(0x00000de4); msg[1] = cpu_to_le32(0x00000002); @@ -1139,9 +1117,6 @@ int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle, for (i = 4; i < 1024; ++i) msg[i] = cpu_to_le32(0x0); - amdgpu_bo_kunmap(bo); - amdgpu_bo_unreserve(bo); - return amdgpu_uvd_send_msg(ring, bo, direct, fence); } @@ -1151,9 +1126,6 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work) container_of(work, struct amdgpu_device, uvd.idle_work.work); unsigned fences = amdgpu_fence_count_emitted(&adev->uvd.ring); - if (amdgpu_sriov_vf(adev)) - return; - if (fences == 0) { if (adev->pm.dpm_enabled) { amdgpu_dpm_enable_uvd(adev, false); @@ -1173,11 +1145,12 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work) void amdgpu_uvd_ring_begin_use(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - bool set_clocks = !cancel_delayed_work_sync(&adev->uvd.idle_work); + bool set_clocks; if (amdgpu_sriov_vf(adev)) return; + set_clocks = !cancel_delayed_work_sync(&adev->uvd.idle_work); if (set_clocks) { if (adev->pm.dpm_enabled) { amdgpu_dpm_enable_uvd(adev, true); @@ -1193,7 +1166,8 @@ void amdgpu_uvd_ring_begin_use(struct amdgpu_ring *ring) void amdgpu_uvd_ring_end_use(struct amdgpu_ring *ring) { - schedule_delayed_work(&ring->adev->uvd.idle_work, UVD_IDLE_TIMEOUT); + if (!amdgpu_sriov_vf(ring->adev)) + schedule_delayed_work(&ring->adev->uvd.idle_work, UVD_IDLE_TIMEOUT); } /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c index f18bbae..7ad622e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -55,6 +55,7 @@ #define FIRMWARE_POLARIS12 "amdgpu/polaris12_vce.bin" #define FIRMWARE_VEGA10 "amdgpu/vega10_vce.bin" +#define FIRMWARE_VEGA12 "amdgpu/vega12_vce.bin" #ifdef CONFIG_DRM_AMDGPU_CIK MODULE_FIRMWARE(FIRMWARE_BONAIRE); @@ -72,6 +73,7 @@ MODULE_FIRMWARE(FIRMWARE_POLARIS11); MODULE_FIRMWARE(FIRMWARE_POLARIS12); MODULE_FIRMWARE(FIRMWARE_VEGA10); +MODULE_FIRMWARE(FIRMWARE_VEGA12); static void amdgpu_vce_idle_work_handler(struct work_struct *work); @@ -127,11 +129,14 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size) case CHIP_POLARIS11: fw_name = FIRMWARE_POLARIS11; break; + case CHIP_POLARIS12: + fw_name = FIRMWARE_POLARIS12; + break; case CHIP_VEGA10: fw_name = FIRMWARE_VEGA10; break; - case CHIP_POLARIS12: - fw_name = FIRMWARE_POLARIS12; + case CHIP_VEGA12: + fw_name = FIRMWARE_VEGA12; break; default: @@ -301,9 +306,6 @@ static void amdgpu_vce_idle_work_handler(struct work_struct *work) container_of(work, struct amdgpu_device, vce.idle_work.work); unsigned i, count = 0; - if (amdgpu_sriov_vf(adev)) - return; - for (i = 0; i < adev->vce.num_rings; i++) count += amdgpu_fence_count_emitted(&adev->vce.ring[i]); @@ -363,7 +365,8 @@ void amdgpu_vce_ring_begin_use(struct amdgpu_ring *ring) */ void amdgpu_vce_ring_end_use(struct amdgpu_ring *ring) { - schedule_delayed_work(&ring->adev->vce.idle_work, VCE_IDLE_TIMEOUT); + if (!amdgpu_sriov_vf(ring->adev)) + schedule_delayed_work(&ring->adev->vce.idle_work, VCE_IDLE_TIMEOUT); } /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h index 0fd378a..7178126 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
@@ -30,6 +30,8 @@ #define AMDGPU_VCE_HARVEST_VCE0 (1 << 0) #define AMDGPU_VCE_HARVEST_VCE1 (1 << 1) +#define AMDGPU_VCE_FW_53_45 ((53 << 24) | (45 << 16)) + struct amdgpu_vce { struct amdgpu_bo *vcpu_bo; uint64_t gpu_addr;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index 052426a..27209cd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -274,34 +274,17 @@ int amdgpu_vcn_dec_ring_test_ring(struct amdgpu_ring *ring) return r; } -static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo, - bool direct, struct dma_fence **fence) +static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring, + struct amdgpu_bo *bo, bool direct, + struct dma_fence **fence) { - struct ttm_operation_ctx ctx = { true, false }; - struct ttm_validate_buffer tv; - struct ww_acquire_ctx ticket; - struct list_head head; + struct amdgpu_device *adev = ring->adev; + struct dma_fence *f = NULL; struct amdgpu_job *job; struct amdgpu_ib *ib; - struct dma_fence *f = NULL; - struct amdgpu_device *adev = ring->adev; uint64_t addr; int i, r; - memset(&tv, 0, sizeof(tv)); - tv.bo = &bo->tbo; - - INIT_LIST_HEAD(&head); - list_add(&tv.head, &head); - - r = ttm_eu_reserve_buffers(&ticket, &head, true, NULL); - if (r) - return r; - - r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); - if (r) - goto err; - r = amdgpu_job_alloc_with_ib(adev, 64, &job); if (r) goto err; @@ -334,11 +317,12 @@ static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *b goto err_free; } - ttm_eu_fence_buffer_objects(&ticket, &head, f); + amdgpu_bo_fence(bo, f, false); + amdgpu_bo_unreserve(bo); + amdgpu_bo_unref(&bo); if (fence) *fence = dma_fence_get(f); - amdgpu_bo_unref(&bo); dma_fence_put(f); return 0; @@ -347,7 +331,8 @@ static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *b amdgpu_job_free(job); err: - ttm_eu_backoff_reservation(&ticket, &head); + amdgpu_bo_unreserve(bo); + amdgpu_bo_unref(&bo); return r; } @@ -355,31 +340,16 @@ static int amdgpu_vcn_dec_get_create_msg(struct amdgpu_ring *ring, uint32_t hand struct dma_fence **fence) { struct amdgpu_device *adev = ring->adev; - struct amdgpu_bo *bo; + struct amdgpu_bo *bo = NULL; uint32_t *msg; int r, i; - r = amdgpu_bo_create(adev, 1024, PAGE_SIZE, true, - AMDGPU_GEM_DOMAIN_VRAM, - AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | - AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, - NULL, NULL, 0, &bo); + r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &bo, NULL, (void **)&msg); if (r) return r; - r = amdgpu_bo_reserve(bo, false); - if (r) { - amdgpu_bo_unref(&bo); - return r; - } - - r = amdgpu_bo_kmap(bo, (void **)&msg); - if (r) { - amdgpu_bo_unreserve(bo); - amdgpu_bo_unref(&bo); - return r; - } - msg[0] = cpu_to_le32(0x00000028); msg[1] = cpu_to_le32(0x00000038); msg[2] = cpu_to_le32(0x00000001); @@ -397,9 +367,6 @@ static int amdgpu_vcn_dec_get_create_msg(struct amdgpu_ring *ring, uint32_t hand for (i = 14; i < 1024; ++i) msg[i] = cpu_to_le32(0x0); - amdgpu_bo_kunmap(bo); - amdgpu_bo_unreserve(bo); - return amdgpu_vcn_dec_send_msg(ring, bo, true, fence); } @@ -407,31 +374,16 @@ static int amdgpu_vcn_dec_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han bool direct, struct dma_fence **fence) { struct amdgpu_device *adev = ring->adev; - struct amdgpu_bo *bo; + struct amdgpu_bo *bo = NULL; uint32_t *msg; int r, i; - r = amdgpu_bo_create(adev, 1024, PAGE_SIZE, true, - AMDGPU_GEM_DOMAIN_VRAM, - AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | - AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, - NULL, NULL, 0, &bo); + r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &bo, NULL, (void **)&msg); if (r) return r; - r = amdgpu_bo_reserve(bo, false); - if (r) { - amdgpu_bo_unref(&bo); - return r; - } - - r = amdgpu_bo_kmap(bo, (void **)&msg); - if (r) { - amdgpu_bo_unreserve(bo); - amdgpu_bo_unref(&bo); - return r; - } - msg[0] = cpu_to_le32(0x00000028); msg[1] = cpu_to_le32(0x00000018); msg[2] = cpu_to_le32(0x00000000); @@ -441,9 +393,6 @@ static int amdgpu_vcn_dec_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han for (i = 6; i < 1024; ++i) msg[i] = cpu_to_le32(0x0); - amdgpu_bo_kunmap(bo); - amdgpu_bo_unreserve(bo); - return amdgpu_vcn_dec_send_msg(ring, bo, direct, fence); }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index e7dfb7b..21adb1b6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
@@ -22,7 +22,21 @@ */ #include "amdgpu.h" -#define MAX_KIQ_REG_WAIT 100000000 /* in usecs */ +#define MAX_KIQ_REG_WAIT 5000 /* in usecs, 5ms */ +#define MAX_KIQ_REG_BAILOUT_INTERVAL 5 /* in msecs, 5ms */ +#define MAX_KIQ_REG_TRY 20 + +uint64_t amdgpu_csa_vaddr(struct amdgpu_device *adev) +{ + uint64_t addr = adev->vm_manager.max_pfn << AMDGPU_GPU_PAGE_SHIFT; + + addr -= AMDGPU_VA_RESERVED_SIZE; + + if (addr >= AMDGPU_VA_HOLE_START) + addr |= AMDGPU_VA_HOLE_END; + + return addr; +} bool amdgpu_virt_mmio_blocked(struct amdgpu_device *adev) { @@ -55,14 +69,14 @@ void amdgpu_free_static_csa(struct amdgpu_device *adev) { /* * amdgpu_map_static_csa should be called during amdgpu_vm_init - * it maps virtual address "AMDGPU_VA_RESERVED_SIZE - AMDGPU_CSA_SIZE" - * to this VM, and each command submission of GFX should use this virtual - * address within META_DATA init package to support SRIOV gfx preemption. + * it maps virtual address amdgpu_csa_vaddr() to this VM, and each command + * submission of GFX should use this virtual address within META_DATA init + * package to support SRIOV gfx preemption. */ - int amdgpu_map_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm, struct amdgpu_bo_va **bo_va) { + uint64_t csa_addr = amdgpu_csa_vaddr(adev) & AMDGPU_VA_HOLE_MASK; struct ww_acquire_ctx ticket; struct list_head list; struct amdgpu_bo_list_entry pd; @@ -90,7 +104,7 @@ int amdgpu_map_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm, return -ENOMEM; } - r = amdgpu_vm_alloc_pts(adev, (*bo_va)->base.vm, AMDGPU_CSA_VADDR, + r = amdgpu_vm_alloc_pts(adev, (*bo_va)->base.vm, csa_addr, AMDGPU_CSA_SIZE); if (r) { DRM_ERROR("failed to allocate pts for static CSA, err=%d\n", r); @@ -99,7 +113,7 @@ int amdgpu_map_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm, return r; } - r = amdgpu_vm_bo_map(adev, *bo_va, AMDGPU_CSA_VADDR, 0, AMDGPU_CSA_SIZE, + r = amdgpu_vm_bo_map(adev, *bo_va, csa_addr, 0, AMDGPU_CSA_SIZE, AMDGPU_PTE_READABLE | AMDGPU_PTE_WRITEABLE | AMDGPU_PTE_EXECUTABLE); @@ -125,9 +139,9 @@ void amdgpu_virt_init_setting(struct amdgpu_device *adev) uint32_t amdgpu_virt_kiq_rreg(struct amdgpu_device *adev, uint32_t reg) { - signed long r; + signed long r, cnt = 0; unsigned long flags; - uint32_t val, seq; + uint32_t seq; struct amdgpu_kiq *kiq = &adev->gfx.kiq; struct amdgpu_ring *ring = &kiq->ring; @@ -141,18 +155,39 @@ uint32_t amdgpu_virt_kiq_rreg(struct amdgpu_device *adev, uint32_t reg) spin_unlock_irqrestore(&kiq->ring_lock, flags); r = amdgpu_fence_wait_polling(ring, seq, MAX_KIQ_REG_WAIT); - if (r < 1) { - DRM_ERROR("wait for kiq fence error: %ld\n", r); - return ~0; - } - val = adev->wb.wb[adev->virt.reg_val_offs]; - return val; + /* don't wait anymore for gpu reset case because this way may + * block gpu_recover() routine forever, e.g. this virt_kiq_rreg + * is triggered in TTM and ttm_bo_lock_delayed_workqueue() will + * never return if we keep waiting in virt_kiq_rreg, which cause + * gpu_recover() hang there. + * + * also don't wait anymore for IRQ context + * */ + if (r < 1 && (adev->in_gpu_reset || in_interrupt())) + goto failed_kiq_read; + + if (in_interrupt()) + might_sleep(); + + while (r < 1 && cnt++ < MAX_KIQ_REG_TRY) { + msleep(MAX_KIQ_REG_BAILOUT_INTERVAL); + r = amdgpu_fence_wait_polling(ring, seq, MAX_KIQ_REG_WAIT); + } + + if (cnt > MAX_KIQ_REG_TRY) + goto failed_kiq_read; + + return adev->wb.wb[adev->virt.reg_val_offs]; + +failed_kiq_read: + pr_err("failed to read reg:%x\n", reg); + return ~0; } void amdgpu_virt_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v) { - signed long r; + signed long r, cnt = 0; unsigned long flags; uint32_t seq; struct amdgpu_kiq *kiq = &adev->gfx.kiq; @@ -168,8 +203,34 @@ void amdgpu_virt_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v) spin_unlock_irqrestore(&kiq->ring_lock, flags); r = amdgpu_fence_wait_polling(ring, seq, MAX_KIQ_REG_WAIT); - if (r < 1) - DRM_ERROR("wait for kiq fence error: %ld\n", r); + + /* don't wait anymore for gpu reset case because this way may + * block gpu_recover() routine forever, e.g. this virt_kiq_rreg + * is triggered in TTM and ttm_bo_lock_delayed_workqueue() will + * never return if we keep waiting in virt_kiq_rreg, which cause + * gpu_recover() hang there. + * + * also don't wait anymore for IRQ context + * */ + if (r < 1 && (adev->in_gpu_reset || in_interrupt())) + goto failed_kiq_write; + + if (in_interrupt()) + might_sleep(); + + while (r < 1 && cnt++ < MAX_KIQ_REG_TRY) { + + msleep(MAX_KIQ_REG_BAILOUT_INTERVAL); + r = amdgpu_fence_wait_polling(ring, seq, MAX_KIQ_REG_WAIT); + } + + if (cnt > MAX_KIQ_REG_TRY) + goto failed_kiq_write; + + return; + +failed_kiq_write: + pr_err("failed to write reg:%x\n", reg); } /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h index 6a83425..880ac113 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
@@ -251,8 +251,7 @@ struct amdgpu_virt { uint32_t gim_feature; }; -#define AMDGPU_CSA_SIZE (8 * 1024) -#define AMDGPU_CSA_VADDR (AMDGPU_VA_RESERVED_SIZE - AMDGPU_CSA_SIZE) +#define AMDGPU_CSA_SIZE (8 * 1024) #define amdgpu_sriov_enabled(adev) \ ((adev)->virt.caps & AMDGPU_SRIOV_CAPS_ENABLE_IOV) @@ -279,6 +278,8 @@ static inline bool is_virtual_machine(void) } struct amdgpu_vm; + +uint64_t amdgpu_csa_vaddr(struct amdgpu_device *adev); bool amdgpu_virt_mmio_blocked(struct amdgpu_device *adev); int amdgpu_allocate_static_csa(struct amdgpu_device *adev); int amdgpu_map_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 5afbc5e..da55a78 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -32,6 +32,7 @@ #include <drm/amdgpu_drm.h> #include "amdgpu.h" #include "amdgpu_trace.h" +#include "amdgpu_amdkfd.h" /* * GPUVM @@ -75,7 +76,8 @@ struct amdgpu_pte_update_params { /* indirect buffer to fill with commands */ struct amdgpu_ib *ib; /* Function which actually does the update */ - void (*func)(struct amdgpu_pte_update_params *params, uint64_t pe, + void (*func)(struct amdgpu_pte_update_params *params, + struct amdgpu_bo *bo, uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, uint64_t flags); /* The next two are used during VM update by CPU @@ -257,6 +259,104 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm) } /** + * amdgpu_vm_clear_bo - initially clear the PDs/PTs + * + * @adev: amdgpu_device pointer + * @bo: BO to clear + * @level: level this BO is at + * + * Root PD needs to be reserved when calling this. + */ +static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, + struct amdgpu_vm *vm, struct amdgpu_bo *bo, + unsigned level, bool pte_support_ats) +{ + struct ttm_operation_ctx ctx = { true, false }; + struct dma_fence *fence = NULL; + unsigned entries, ats_entries; + struct amdgpu_ring *ring; + struct amdgpu_job *job; + uint64_t addr; + int r; + + addr = amdgpu_bo_gpu_offset(bo); + entries = amdgpu_bo_size(bo) / 8; + + if (pte_support_ats) { + if (level == adev->vm_manager.root_level) { + ats_entries = amdgpu_vm_level_shift(adev, level); + ats_entries += AMDGPU_GPU_PAGE_SHIFT; + ats_entries = AMDGPU_VA_HOLE_START >> ats_entries; + ats_entries = min(ats_entries, entries); + entries -= ats_entries; + } else { + ats_entries = entries; + entries = 0; + } + } else { + ats_entries = 0; + } + + ring = container_of(vm->entity.sched, struct amdgpu_ring, sched); + + r = reservation_object_reserve_shared(bo->tbo.resv); + if (r) + return r; + + r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); + if (r) + goto error; + + r = amdgpu_job_alloc_with_ib(adev, 64, &job); + if (r) + goto error; + + if (ats_entries) { + uint64_t ats_value; + + ats_value = AMDGPU_PTE_DEFAULT_ATC; + if (level != AMDGPU_VM_PTB) + ats_value |= AMDGPU_PDE_PTE; + + amdgpu_vm_set_pte_pde(adev, &job->ibs[0], addr, 0, + ats_entries, 0, ats_value); + addr += ats_entries * 8; + } + + if (entries) + amdgpu_vm_set_pte_pde(adev, &job->ibs[0], addr, 0, + entries, 0, 0); + + amdgpu_ring_pad_ib(ring, &job->ibs[0]); + + WARN_ON(job->ibs[0].length_dw > 64); + r = amdgpu_sync_resv(adev, &job->sync, bo->tbo.resv, + AMDGPU_FENCE_OWNER_UNDEFINED, false); + if (r) + goto error_free; + + r = amdgpu_job_submit(job, ring, &vm->entity, + AMDGPU_FENCE_OWNER_UNDEFINED, &fence); + if (r) + goto error_free; + + amdgpu_bo_fence(bo, fence, true); + dma_fence_put(fence); + + if (bo->shadow) + return amdgpu_vm_clear_bo(adev, vm, bo->shadow, + level, pte_support_ats); + + return 0; + +error_free: + amdgpu_job_free(job); + +error: + return r; +} + +/** * amdgpu_vm_alloc_levels - allocate the PD/PT levels * * @adev: amdgpu_device pointer @@ -270,13 +370,12 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, struct amdgpu_vm *vm, struct amdgpu_vm_pt *parent, uint64_t saddr, uint64_t eaddr, - unsigned level) + unsigned level, bool ats) { unsigned shift = amdgpu_vm_level_shift(adev, level); unsigned pt_idx, from, to; - int r; u64 flags; - uint64_t init_value = 0; + int r; if (!parent->entries) { unsigned num_entries = amdgpu_vm_num_entries(adev, level); @@ -299,21 +398,13 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, saddr = saddr & ((1 << shift) - 1); eaddr = eaddr & ((1 << shift) - 1); - flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS | - AMDGPU_GEM_CREATE_VRAM_CLEARED; + flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; if (vm->use_cpu_for_update) flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; else flags |= (AMDGPU_GEM_CREATE_NO_CPU_ACCESS | AMDGPU_GEM_CREATE_SHADOW); - if (vm->pte_support_ats) { - init_value = AMDGPU_PTE_DEFAULT_ATC; - if (level != AMDGPU_VM_PTB) - init_value |= AMDGPU_PDE_PTE; - - } - /* walk over the address space and allocate the page tables */ for (pt_idx = from; pt_idx <= to; ++pt_idx) { struct reservation_object *resv = vm->root.base.bo->tbo.resv; @@ -323,16 +414,23 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, if (!entry->base.bo) { r = amdgpu_bo_create(adev, amdgpu_vm_bo_size(adev, level), - AMDGPU_GPU_PAGE_SIZE, true, - AMDGPU_GEM_DOMAIN_VRAM, - flags, - NULL, resv, init_value, &pt); + AMDGPU_GPU_PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, flags, + ttm_bo_type_kernel, resv, &pt); if (r) return r; + r = amdgpu_vm_clear_bo(adev, vm, pt, level, ats); + if (r) { + amdgpu_bo_unref(&pt->shadow); + amdgpu_bo_unref(&pt); + return r; + } + if (vm->use_cpu_for_update) { r = amdgpu_bo_kmap(pt, NULL); if (r) { + amdgpu_bo_unref(&pt->shadow); amdgpu_bo_unref(&pt); return r; } @@ -356,7 +454,7 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, uint64_t sub_eaddr = (pt_idx == to) ? eaddr : ((1 << shift) - 1); r = amdgpu_vm_alloc_levels(adev, vm, entry, sub_saddr, - sub_eaddr, level); + sub_eaddr, level, ats); if (r) return r; } @@ -379,26 +477,29 @@ int amdgpu_vm_alloc_pts(struct amdgpu_device *adev, struct amdgpu_vm *vm, uint64_t saddr, uint64_t size) { - uint64_t last_pfn; uint64_t eaddr; + bool ats = false; /* validate the parameters */ if (saddr & AMDGPU_GPU_PAGE_MASK || size & AMDGPU_GPU_PAGE_MASK) return -EINVAL; eaddr = saddr + size - 1; - last_pfn = eaddr / AMDGPU_GPU_PAGE_SIZE; - if (last_pfn >= adev->vm_manager.max_pfn) { - dev_err(adev->dev, "va above limit (0x%08llX >= 0x%08llX)\n", - last_pfn, adev->vm_manager.max_pfn); - return -EINVAL; - } + + if (vm->pte_support_ats) + ats = saddr < AMDGPU_VA_HOLE_START; saddr /= AMDGPU_GPU_PAGE_SIZE; eaddr /= AMDGPU_GPU_PAGE_SIZE; + if (eaddr >= adev->vm_manager.max_pfn) { + dev_err(adev->dev, "va above limit (0x%08llX >= 0x%08llX)\n", + eaddr, adev->vm_manager.max_pfn); + return -EINVAL; + } + return amdgpu_vm_alloc_levels(adev, vm, &vm->root, saddr, eaddr, - adev->vm_manager.root_level); + adev->vm_manager.root_level, ats); } /** @@ -465,7 +566,7 @@ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring, static bool amdgpu_vm_is_large_bar(struct amdgpu_device *adev) { - return (adev->mc.real_vram_size == adev->mc.visible_vram_size); + return (adev->gmc.real_vram_size == adev->gmc.visible_vram_size); } /** @@ -491,14 +592,24 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_ id->oa_base != job->oa_base || id->oa_size != job->oa_size); bool vm_flush_needed = job->vm_needs_flush; + bool pasid_mapping_needed = id->pasid != job->pasid || + !id->pasid_mapping || + !dma_fence_is_signaled(id->pasid_mapping); + struct dma_fence *fence = NULL; unsigned patch_offset = 0; int r; if (amdgpu_vmid_had_gpu_reset(adev, id)) { gds_switch_needed = true; vm_flush_needed = true; + pasid_mapping_needed = true; } + gds_switch_needed &= !!ring->funcs->emit_gds_switch; + vm_flush_needed &= !!ring->funcs->emit_vm_flush; + pasid_mapping_needed &= adev->gmc.gmc_funcs->emit_pasid_mapping && + ring->funcs->emit_wreg; + if (!vm_flush_needed && !gds_switch_needed && !need_pipe_sync) return 0; @@ -508,23 +619,36 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_ if (need_pipe_sync) amdgpu_ring_emit_pipeline_sync(ring); - if (ring->funcs->emit_vm_flush && vm_flush_needed) { - struct dma_fence *fence; - + if (vm_flush_needed) { trace_amdgpu_vm_flush(ring, job->vmid, job->vm_pd_addr); amdgpu_ring_emit_vm_flush(ring, job->vmid, job->vm_pd_addr); + } + if (pasid_mapping_needed) + amdgpu_gmc_emit_pasid_mapping(ring, job->vmid, job->pasid); + + if (vm_flush_needed || pasid_mapping_needed) { r = amdgpu_fence_emit(ring, &fence); if (r) return r; + } + if (vm_flush_needed) { mutex_lock(&id_mgr->lock); dma_fence_put(id->last_flush); - id->last_flush = fence; - id->current_gpu_reset_count = atomic_read(&adev->gpu_reset_counter); + id->last_flush = dma_fence_get(fence); + id->current_gpu_reset_count = + atomic_read(&adev->gpu_reset_counter); mutex_unlock(&id_mgr->lock); } + if (pasid_mapping_needed) { + id->pasid = job->pasid; + dma_fence_put(id->pasid_mapping); + id->pasid_mapping = dma_fence_get(fence); + } + dma_fence_put(fence); + if (ring->funcs->emit_gds_switch && gds_switch_needed) { id->gds_base = job->gds_base; id->gds_size = job->gds_size; @@ -578,6 +702,7 @@ struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm, * amdgpu_vm_do_set_ptes - helper to call the right asic function * * @params: see amdgpu_pte_update_params definition + * @bo: PD/PT to update * @pe: addr of the page entry * @addr: dst addr to write into pe * @count: number of page entries to update @@ -588,10 +713,12 @@ struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm, * to setup the page table using the DMA. */ static void amdgpu_vm_do_set_ptes(struct amdgpu_pte_update_params *params, + struct amdgpu_bo *bo, uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, uint64_t flags) { + pe += amdgpu_bo_gpu_offset(bo); trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags); if (count < 3) { @@ -608,6 +735,7 @@ static void amdgpu_vm_do_set_ptes(struct amdgpu_pte_update_params *params, * amdgpu_vm_do_copy_ptes - copy the PTEs from the GART * * @params: see amdgpu_pte_update_params definition + * @bo: PD/PT to update * @pe: addr of the page entry * @addr: dst addr to write into pe * @count: number of page entries to update @@ -617,13 +745,14 @@ static void amdgpu_vm_do_set_ptes(struct amdgpu_pte_update_params *params, * Traces the parameters and calls the DMA function to copy the PTEs. */ static void amdgpu_vm_do_copy_ptes(struct amdgpu_pte_update_params *params, + struct amdgpu_bo *bo, uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, uint64_t flags) { uint64_t src = (params->src + (addr >> 12) * 8); - + pe += amdgpu_bo_gpu_offset(bo); trace_amdgpu_vm_copy_ptes(pe, src, count); amdgpu_vm_copy_pte(params->adev, params->ib, pe, src, count); @@ -657,6 +786,7 @@ static uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr) * amdgpu_vm_cpu_set_ptes - helper to update page tables via CPU * * @params: see amdgpu_pte_update_params definition + * @bo: PD/PT to update * @pe: kmap addr of the page entry * @addr: dst addr to write into pe * @count: number of page entries to update @@ -666,6 +796,7 @@ static uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr) * Write count number of PT/PD entries directly. */ static void amdgpu_vm_cpu_set_ptes(struct amdgpu_pte_update_params *params, + struct amdgpu_bo *bo, uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, uint64_t flags) @@ -673,14 +804,16 @@ static void amdgpu_vm_cpu_set_ptes(struct amdgpu_pte_update_params *params, unsigned int i; uint64_t value; + pe += (unsigned long)amdgpu_bo_kptr(bo); + trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags); for (i = 0; i < count; i++) { value = params->pages_addr ? amdgpu_vm_map_gart(params->pages_addr, addr) : addr; - amdgpu_gart_set_pte_pde(params->adev, (void *)(uintptr_t)pe, - i, value, flags); + amdgpu_gmc_set_pte_pde(params->adev, (void *)(uintptr_t)pe, + i, value, flags); addr += incr; } } @@ -714,8 +847,7 @@ static void amdgpu_vm_update_pde(struct amdgpu_pte_update_params *params, struct amdgpu_vm_pt *parent, struct amdgpu_vm_pt *entry) { - struct amdgpu_bo *bo = entry->base.bo, *shadow = NULL, *pbo; - uint64_t pd_addr, shadow_addr = 0; + struct amdgpu_bo *bo = parent->base.bo, *pbo; uint64_t pde, pt, flags; unsigned level; @@ -723,29 +855,17 @@ static void amdgpu_vm_update_pde(struct amdgpu_pte_update_params *params, if (entry->huge) return; - if (vm->use_cpu_for_update) { - pd_addr = (unsigned long)amdgpu_bo_kptr(parent->base.bo); - } else { - pd_addr = amdgpu_bo_gpu_offset(parent->base.bo); - shadow = parent->base.bo->shadow; - if (shadow) - shadow_addr = amdgpu_bo_gpu_offset(shadow); - } - - for (level = 0, pbo = parent->base.bo->parent; pbo; ++level) + for (level = 0, pbo = bo->parent; pbo; ++level) pbo = pbo->parent; level += params->adev->vm_manager.root_level; - pt = amdgpu_bo_gpu_offset(bo); + pt = amdgpu_bo_gpu_offset(entry->base.bo); flags = AMDGPU_PTE_VALID; - amdgpu_gart_get_vm_pde(params->adev, level, &pt, &flags); - if (shadow) { - pde = shadow_addr + (entry - parent->entries) * 8; - params->func(params, pde, pt, 1, 0, flags); - } - - pde = pd_addr + (entry - parent->entries) * 8; - params->func(params, pde, pt, 1, 0, flags); + amdgpu_gmc_get_vm_pde(params->adev, level, &pt, &flags); + pde = (entry - parent->entries) * 8; + if (bo->shadow) + params->func(params, bo->shadow, pde, pt, 1, 0, flags); + params->func(params, bo, pde, pt, 1, 0, flags); } /* @@ -856,7 +976,7 @@ int amdgpu_vm_update_directories(struct amdgpu_device *adev, if (vm->use_cpu_for_update) { /* Flush HDP */ mb(); - amdgpu_gart_flush_gpu_tlb(adev, 0); + amdgpu_asic_flush_hdp(adev, NULL); } else if (params.ib->length_dw == 0) { amdgpu_job_free(job); } else { @@ -870,11 +990,6 @@ int amdgpu_vm_update_directories(struct amdgpu_device *adev, amdgpu_ring_pad_ib(ring, params.ib); amdgpu_sync_resv(adev, &job->sync, root->tbo.resv, AMDGPU_FENCE_OWNER_VM, false); - if (root->shadow) - amdgpu_sync_resv(adev, &job->sync, - root->shadow->tbo.resv, - AMDGPU_FENCE_OWNER_VM, false); - WARN_ON(params.ib->length_dw > ndw); r = amdgpu_job_submit(job, ring, &vm->entity, AMDGPU_FENCE_OWNER_VM, &fence); @@ -946,7 +1061,7 @@ static void amdgpu_vm_handle_huge_pages(struct amdgpu_pte_update_params *p, unsigned nptes, uint64_t dst, uint64_t flags) { - uint64_t pd_addr, pde; + uint64_t pde; /* In the case of a mixed PT the PDE must point to it*/ if (p->adev->asic_type >= CHIP_VEGA10 && !p->src && @@ -967,21 +1082,12 @@ static void amdgpu_vm_handle_huge_pages(struct amdgpu_pte_update_params *p, } entry->huge = true; - amdgpu_gart_get_vm_pde(p->adev, AMDGPU_VM_PDB0, - &dst, &flags); + amdgpu_gmc_get_vm_pde(p->adev, AMDGPU_VM_PDB0, &dst, &flags); - if (p->func == amdgpu_vm_cpu_set_ptes) { - pd_addr = (unsigned long)amdgpu_bo_kptr(parent->base.bo); - } else { - if (parent->base.bo->shadow) { - pd_addr = amdgpu_bo_gpu_offset(parent->base.bo->shadow); - pde = pd_addr + (entry - parent->entries) * 8; - p->func(p, pde, dst, 1, 0, flags); - } - pd_addr = amdgpu_bo_gpu_offset(parent->base.bo); - } - pde = pd_addr + (entry - parent->entries) * 8; - p->func(p, pde, dst, 1, 0, flags); + pde = (entry - parent->entries) * 8; + if (parent->base.bo->shadow) + p->func(p, parent->base.bo->shadow, pde, dst, 1, 0, flags); + p->func(p, parent->base.bo, pde, dst, 1, 0, flags); } /** @@ -1007,7 +1113,6 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, uint64_t addr, pe_start; struct amdgpu_bo *pt; unsigned nptes; - bool use_cpu_update = (params->func == amdgpu_vm_cpu_set_ptes); /* walk over the address space and update the page tables */ for (addr = start; addr < end; addr += nptes, @@ -1030,20 +1135,11 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, continue; pt = entry->base.bo; - if (use_cpu_update) { - pe_start = (unsigned long)amdgpu_bo_kptr(pt); - } else { - if (pt->shadow) { - pe_start = amdgpu_bo_gpu_offset(pt->shadow); - pe_start += (addr & mask) * 8; - params->func(params, pe_start, dst, nptes, - AMDGPU_GPU_PAGE_SIZE, flags); - } - pe_start = amdgpu_bo_gpu_offset(pt); - } - - pe_start += (addr & mask) * 8; - params->func(params, pe_start, dst, nptes, + pe_start = (addr & mask) * 8; + if (pt->shadow) + params->func(params, pt->shadow, pe_start, dst, nptes, + AMDGPU_GPU_PAGE_SIZE, flags); + params->func(params, pt, pe_start, dst, nptes, AMDGPU_GPU_PAGE_SIZE, flags); } @@ -1204,11 +1300,10 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, } else { /* set page commands needed */ - ndw += ncmds * adev->vm_manager.vm_pte_funcs->set_pte_pde_num_dw; + ndw += ncmds * 10; /* extra commands for begin/end fragments */ - ndw += 2 * adev->vm_manager.vm_pte_funcs->set_pte_pde_num_dw - * adev->vm_manager.fragment_size; + ndw += 2 * 10 * adev->vm_manager.fragment_size; params.func = amdgpu_vm_do_set_ptes; } @@ -1457,7 +1552,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, if (vm->use_cpu_for_update) { /* Flush HDP */ mb(); - amdgpu_gart_flush_gpu_tlb(adev, 0); + amdgpu_asic_flush_hdp(adev, NULL); } spin_lock(&vm->status_lock); @@ -1485,7 +1580,7 @@ static void amdgpu_vm_update_prt_state(struct amdgpu_device *adev) spin_lock_irqsave(&adev->vm_manager.prt_lock, flags); enable = !!atomic_read(&adev->vm_manager.num_prt_users); - adev->gart.gart_funcs->set_prt(adev, enable); + adev->gmc.gmc_funcs->set_prt(adev, enable); spin_unlock_irqrestore(&adev->vm_manager.prt_lock, flags); } @@ -1494,7 +1589,7 @@ static void amdgpu_vm_update_prt_state(struct amdgpu_device *adev) */ static void amdgpu_vm_prt_get(struct amdgpu_device *adev) { - if (!adev->gart.gart_funcs->set_prt) + if (!adev->gmc.gmc_funcs->set_prt) return; if (atomic_inc_return(&adev->vm_manager.num_prt_users) == 1) @@ -1529,7 +1624,7 @@ static void amdgpu_vm_add_prt_cb(struct amdgpu_device *adev, { struct amdgpu_prt_cb *cb; - if (!adev->gart.gart_funcs->set_prt) + if (!adev->gmc.gmc_funcs->set_prt) return; cb = kmalloc(sizeof(struct amdgpu_prt_cb), GFP_KERNEL); @@ -1623,16 +1718,16 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, struct dma_fence **fence) { struct amdgpu_bo_va_mapping *mapping; + uint64_t init_pte_value = 0; struct dma_fence *f = NULL; int r; - uint64_t init_pte_value = 0; while (!list_empty(&vm->freed)) { mapping = list_first_entry(&vm->freed, struct amdgpu_bo_va_mapping, list); list_del(&mapping->list); - if (vm->pte_support_ats) + if (vm->pte_support_ats && mapping->start < AMDGPU_VA_HOLE_START) init_pte_value = AMDGPU_PTE_DEFAULT_ATC; r = amdgpu_vm_bo_update_mapping(adev, NULL, NULL, vm, @@ -2262,11 +2357,11 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, { const unsigned align = min(AMDGPU_VM_PTB_ALIGN_SIZE, AMDGPU_VM_PTE_COUNT(adev) * 8); - uint64_t init_pde_value = 0, flags; unsigned ring_instance; struct amdgpu_ring *ring; struct drm_sched_rq *rq; unsigned long size; + uint64_t flags; int r, i; vm->va = RB_ROOT_CACHED; @@ -2295,33 +2390,27 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_COMPUTE); - if (adev->asic_type == CHIP_RAVEN) { + if (adev->asic_type == CHIP_RAVEN) vm->pte_support_ats = true; - init_pde_value = AMDGPU_PTE_DEFAULT_ATC - | AMDGPU_PDE_PTE; - - } - } else + } else { vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_GFX); + } DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update & !amdgpu_vm_is_large_bar(adev)), "CPU update of VM recommended only for large BAR system\n"); vm->last_update = NULL; - flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS | - AMDGPU_GEM_CREATE_VRAM_CLEARED; + flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; if (vm->use_cpu_for_update) flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; else - flags |= (AMDGPU_GEM_CREATE_NO_CPU_ACCESS | - AMDGPU_GEM_CREATE_SHADOW); + flags |= AMDGPU_GEM_CREATE_SHADOW; size = amdgpu_vm_bo_size(adev, adev->vm_manager.root_level); - r = amdgpu_bo_create(adev, size, align, true, AMDGPU_GEM_DOMAIN_VRAM, - flags, NULL, NULL, init_pde_value, - &vm->root.base.bo); + r = amdgpu_bo_create(adev, size, align, AMDGPU_GEM_DOMAIN_VRAM, flags, + ttm_bo_type_kernel, NULL, &vm->root.base.bo); if (r) goto error_free_sched_entity; @@ -2329,6 +2418,12 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, if (r) goto error_free_root; + r = amdgpu_vm_clear_bo(adev, vm, vm->root.base.bo, + adev->vm_manager.root_level, + vm->pte_support_ats); + if (r) + goto error_unreserve; + vm->root.base.vm = vm; list_add_tail(&vm->root.base.bo_list, &vm->root.base.bo->va); list_add_tail(&vm->root.base.vm_status, &vm->evicted); @@ -2352,6 +2447,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, return 0; +error_unreserve: + amdgpu_bo_unreserve(vm->root.base.bo); + error_free_root: amdgpu_bo_unref(&vm->root.base.bo->shadow); amdgpu_bo_unref(&vm->root.base.bo); @@ -2364,6 +2462,73 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, } /** + * amdgpu_vm_make_compute - Turn a GFX VM into a compute VM + * + * This only works on GFX VMs that don't have any BOs added and no + * page tables allocated yet. + * + * Changes the following VM parameters: + * - use_cpu_for_update + * - pte_supports_ats + * - pasid (old PASID is released, because compute manages its own PASIDs) + * + * Reinitializes the page directory to reflect the changed ATS + * setting. May leave behind an unused shadow BO for the page + * directory when switching from SDMA updates to CPU updates. + * + * Returns 0 for success, -errno for errors. + */ +int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm) +{ + bool pte_support_ats = (adev->asic_type == CHIP_RAVEN); + int r; + + r = amdgpu_bo_reserve(vm->root.base.bo, true); + if (r) + return r; + + /* Sanity checks */ + if (!RB_EMPTY_ROOT(&vm->va.rb_root) || vm->root.entries) { + r = -EINVAL; + goto error; + } + + /* Check if PD needs to be reinitialized and do it before + * changing any other state, in case it fails. + */ + if (pte_support_ats != vm->pte_support_ats) { + r = amdgpu_vm_clear_bo(adev, vm, vm->root.base.bo, + adev->vm_manager.root_level, + pte_support_ats); + if (r) + goto error; + } + + /* Update VM state */ + vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & + AMDGPU_VM_USE_CPU_FOR_COMPUTE); + vm->pte_support_ats = pte_support_ats; + DRM_DEBUG_DRIVER("VM update mode is %s\n", + vm->use_cpu_for_update ? "CPU" : "SDMA"); + WARN_ONCE((vm->use_cpu_for_update & !amdgpu_vm_is_large_bar(adev)), + "CPU update of VM recommended only for large BAR system\n"); + + if (vm->pasid) { + unsigned long flags; + + spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags); + idr_remove(&adev->vm_manager.pasid_idr, vm->pasid); + spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); + + vm->pasid = 0; + } + +error: + amdgpu_bo_unreserve(vm->root.base.bo); + return r; +} + +/** * amdgpu_vm_free_levels - free PD/PT levels * * @adev: amdgpu device structure @@ -2405,11 +2570,13 @@ static void amdgpu_vm_free_levels(struct amdgpu_device *adev, void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) { struct amdgpu_bo_va_mapping *mapping, *tmp; - bool prt_fini_needed = !!adev->gart.gart_funcs->set_prt; + bool prt_fini_needed = !!adev->gmc.gmc_funcs->set_prt; struct amdgpu_bo *root; u64 fault; int i, r; + amdgpu_amdkfd_gpuvm_destroy_cb(adev, vm); + /* Clear pending page faults from IH when the VM is destroyed */ while (kfifo_get(&vm->faults, &fault)) amdgpu_ih_clear_fault(adev, fault);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 21a80f1..30f0803 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -28,6 +28,7 @@ #include <linux/kfifo.h> #include <linux/rbtree.h> #include <drm/gpu_scheduler.h> +#include <drm/drm_file.h> #include "amdgpu_sync.h" #include "amdgpu_ring.h" @@ -99,7 +100,7 @@ struct amdgpu_bo_list_entry; #define AMDGPU_MMHUB 1 /* hardcode that limit for now */ -#define AMDGPU_VA_RESERVED_SIZE (8ULL << 20) +#define AMDGPU_VA_RESERVED_SIZE (1ULL << 20) /* VA hole for 48bit addresses on Vega10 */ #define AMDGPU_VA_HOLE_START 0x0000800000000000ULL @@ -206,6 +207,15 @@ struct amdgpu_vm { /* Limit non-retry fault storms */ unsigned int fault_credit; + + /* Points to the KFD process VM info */ + struct amdkfd_process_info *process_info; + + /* List node in amdkfd_process_info.vm_list_head */ + struct list_head vm_list_node; + + /* Valid while the PD is reserved or fenced */ + uint64_t pd_phys_addr; }; struct amdgpu_vm_manager { @@ -250,6 +260,7 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev); void amdgpu_vm_manager_fini(struct amdgpu_device *adev); int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, int vm_context, unsigned int pasid); +int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm); void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm); bool amdgpu_vm_pasid_fault_credit(struct amdgpu_device *adev, unsigned int pasid);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c index c9f3336..ae0049c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
@@ -89,11 +89,11 @@ static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev, uint64_t start = node->start << PAGE_SHIFT; uint64_t end = (node->size + node->start) << PAGE_SHIFT; - if (start >= adev->mc.visible_vram_size) + if (start >= adev->gmc.visible_vram_size) return 0; - return (end > adev->mc.visible_vram_size ? - adev->mc.visible_vram_size : end) - start; + return (end > adev->gmc.visible_vram_size ? + adev->gmc.visible_vram_size : end) - start; } /**
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c index 6a73763..484f333 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
@@ -65,6 +65,8 @@ MODULE_FIRMWARE("radeon/hawaii_k_smc.bin"); #define VOLTAGE_VID_OFFSET_SCALE1 625 #define VOLTAGE_VID_OFFSET_SCALE2 100 +static const struct amd_pm_funcs ci_dpm_funcs; + static const struct ci_pt_defaults defaults_hawaii_xt = { 1, 0xF, 0xFD, 0x19, 5, 0x14, 0, 0xB0000, @@ -905,7 +907,7 @@ static bool ci_dpm_vblank_too_short(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; u32 vblank_time = amdgpu_dpm_get_vblank_time(adev); - u32 switch_limit = adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5 ? 450 : 300; + u32 switch_limit = adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5 ? 450 : 300; /* disable mclk switching if the refresh is >120Hz, even if the * blanking period would allow it @@ -2954,7 +2956,7 @@ static int ci_calculate_mclk_params(struct amdgpu_device *adev, mpll_ad_func_cntl &= ~MPLL_AD_FUNC_CNTL__YCLK_POST_DIV_MASK; mpll_ad_func_cntl |= (mpll_param.post_div << MPLL_AD_FUNC_CNTL__YCLK_POST_DIV__SHIFT); - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { mpll_dq_func_cntl &= ~(MPLL_DQ_FUNC_CNTL__YCLK_SEL_MASK | MPLL_AD_FUNC_CNTL__YCLK_POST_DIV_MASK); mpll_dq_func_cntl |= (mpll_param.yclk_sel << MPLL_DQ_FUNC_CNTL__YCLK_SEL__SHIFT) | @@ -3077,7 +3079,7 @@ static int ci_populate_single_memory_level(struct amdgpu_device *adev, (memory_clock <= pi->mclk_strobe_mode_threshold)) memory_level->StrobeEnable = 1; - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { memory_level->StrobeRatio = ci_get_mclk_frequency_ratio(memory_clock, memory_level->StrobeEnable); if (pi->mclk_edc_enable_threshold && @@ -3695,40 +3697,6 @@ static int ci_find_boot_level(struct ci_single_dpm_table *table, return ret; } -static void ci_save_default_power_profile(struct amdgpu_device *adev) -{ - struct ci_power_info *pi = ci_get_pi(adev); - struct SMU7_Discrete_GraphicsLevel *levels = - pi->smc_state_table.GraphicsLevel; - uint32_t min_level = 0; - - pi->default_gfx_power_profile.activity_threshold = - be16_to_cpu(levels[0].ActivityLevel); - pi->default_gfx_power_profile.up_hyst = levels[0].UpH; - pi->default_gfx_power_profile.down_hyst = levels[0].DownH; - pi->default_gfx_power_profile.type = AMD_PP_GFX_PROFILE; - - pi->default_compute_power_profile = pi->default_gfx_power_profile; - pi->default_compute_power_profile.type = AMD_PP_COMPUTE_PROFILE; - - /* Optimize compute power profile: Use only highest - * 2 power levels (if more than 2 are available), Hysteresis: - * 0ms up, 5ms down - */ - if (pi->smc_state_table.GraphicsDpmLevelCount > 2) - min_level = pi->smc_state_table.GraphicsDpmLevelCount - 2; - else if (pi->smc_state_table.GraphicsDpmLevelCount == 2) - min_level = 1; - pi->default_compute_power_profile.min_sclk = - be32_to_cpu(levels[min_level].SclkFrequency); - - pi->default_compute_power_profile.up_hyst = 0; - pi->default_compute_power_profile.down_hyst = 5; - - pi->gfx_power_profile = pi->default_gfx_power_profile; - pi->compute_power_profile = pi->default_compute_power_profile; -} - static int ci_init_smc_table(struct amdgpu_device *adev) { struct ci_power_info *pi = ci_get_pi(adev); @@ -3752,7 +3720,7 @@ static int ci_init_smc_table(struct amdgpu_device *adev) if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_STEPVDDC) table->SystemFlags |= PPSMC_SYSTEMFLAG_STEPVDDC; - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) table->SystemFlags |= PPSMC_SYSTEMFLAG_GDDR5; if (ulv->supported) { @@ -3874,8 +3842,6 @@ static int ci_init_smc_table(struct amdgpu_device *adev) if (ret) return ret; - ci_save_default_power_profile(adev); - return 0; } @@ -4549,12 +4515,12 @@ static int ci_set_mc_special_registers(struct amdgpu_device *adev, for (k = 0; k < table->num_entries; k++) { table->mc_reg_table_entry[k].mc_data[j] = (temp_reg & 0xffff0000) | (table->mc_reg_table_entry[k].mc_data[i] & 0x0000ffff); - if (adev->mc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) + if (adev->gmc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) table->mc_reg_table_entry[k].mc_data[j] |= 0x100; } j++; - if (adev->mc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) { + if (adev->gmc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) { if (j >= SMU7_DISCRETE_MC_REGISTER_ARRAY_SIZE) return -EINVAL; table->mc_reg_address[j].s1 = mmMC_PMG_AUTO_CMD; @@ -6277,6 +6243,8 @@ static int ci_dpm_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + adev->powerplay.pp_funcs = &ci_dpm_funcs; + adev->powerplay.pp_handle = adev; ci_dpm_set_irq_funcs(adev); return 0; @@ -6639,9 +6607,10 @@ static int ci_dpm_force_clock_level(void *handle, struct amdgpu_device *adev = (struct amdgpu_device *)handle; struct ci_power_info *pi = ci_get_pi(adev); - if (adev->pm.dpm.forced_level & (AMD_DPM_FORCED_LEVEL_AUTO | - AMD_DPM_FORCED_LEVEL_LOW | - AMD_DPM_FORCED_LEVEL_HIGH)) + if (adev->pm.dpm.forced_level != AMD_DPM_FORCED_LEVEL_MANUAL) + return -EINVAL; + + if (mask == 0) return -EINVAL; switch (type) { @@ -6662,15 +6631,15 @@ static int ci_dpm_force_clock_level(void *handle, case PP_PCIE: { uint32_t tmp = mask & pi->dpm_level_enable_mask.pcie_dpm_enable_mask; - uint32_t level = 0; - while (tmp >>= 1) - level++; - - if (!pi->pcie_dpm_key_disabled) - amdgpu_ci_send_msg_to_smc_with_parameter(adev, + if (!pi->pcie_dpm_key_disabled) { + if (fls(tmp) != ffs(tmp)) + amdgpu_ci_send_msg_to_smc(adev, PPSMC_MSG_PCIeDPM_UnForceLevel); + else + amdgpu_ci_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_PCIeDPM_ForceLevel, - level); + fls(tmp) - 1); + } break; } default: @@ -6752,222 +6721,6 @@ static int ci_dpm_set_mclk_od(void *handle, uint32_t value) return 0; } -static int ci_dpm_get_power_profile_state(void *handle, - struct amd_pp_profile *query) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - struct ci_power_info *pi = ci_get_pi(adev); - - if (!pi || !query) - return -EINVAL; - - if (query->type == AMD_PP_GFX_PROFILE) - memcpy(query, &pi->gfx_power_profile, - sizeof(struct amd_pp_profile)); - else if (query->type == AMD_PP_COMPUTE_PROFILE) - memcpy(query, &pi->compute_power_profile, - sizeof(struct amd_pp_profile)); - else - return -EINVAL; - - return 0; -} - -static int ci_populate_requested_graphic_levels(struct amdgpu_device *adev, - struct amd_pp_profile *request) -{ - struct ci_power_info *pi = ci_get_pi(adev); - struct ci_dpm_table *dpm_table = &(pi->dpm_table); - struct SMU7_Discrete_GraphicsLevel *levels = - pi->smc_state_table.GraphicsLevel; - uint32_t array = pi->dpm_table_start + - offsetof(SMU7_Discrete_DpmTable, GraphicsLevel); - uint32_t array_size = sizeof(struct SMU7_Discrete_GraphicsLevel) * - SMU7_MAX_LEVELS_GRAPHICS; - uint32_t i; - - for (i = 0; i < dpm_table->sclk_table.count; i++) { - levels[i].ActivityLevel = - cpu_to_be16(request->activity_threshold); - levels[i].EnabledForActivity = 1; - levels[i].UpH = request->up_hyst; - levels[i].DownH = request->down_hyst; - } - - return amdgpu_ci_copy_bytes_to_smc(adev, array, (uint8_t *)levels, - array_size, pi->sram_end); -} - -static void ci_find_min_clock_masks(struct amdgpu_device *adev, - uint32_t *sclk_mask, uint32_t *mclk_mask, - uint32_t min_sclk, uint32_t min_mclk) -{ - struct ci_power_info *pi = ci_get_pi(adev); - struct ci_dpm_table *dpm_table = &(pi->dpm_table); - uint32_t i; - - for (i = 0; i < dpm_table->sclk_table.count; i++) { - if (dpm_table->sclk_table.dpm_levels[i].enabled && - dpm_table->sclk_table.dpm_levels[i].value >= min_sclk) - *sclk_mask |= 1 << i; - } - - for (i = 0; i < dpm_table->mclk_table.count; i++) { - if (dpm_table->mclk_table.dpm_levels[i].enabled && - dpm_table->mclk_table.dpm_levels[i].value >= min_mclk) - *mclk_mask |= 1 << i; - } -} - -static int ci_set_power_profile_state(struct amdgpu_device *adev, - struct amd_pp_profile *request) -{ - struct ci_power_info *pi = ci_get_pi(adev); - int tmp_result, result = 0; - uint32_t sclk_mask = 0, mclk_mask = 0; - - tmp_result = ci_freeze_sclk_mclk_dpm(adev); - if (tmp_result) { - DRM_ERROR("Failed to freeze SCLK MCLK DPM!"); - result = tmp_result; - } - - tmp_result = ci_populate_requested_graphic_levels(adev, - request); - if (tmp_result) { - DRM_ERROR("Failed to populate requested graphic levels!"); - result = tmp_result; - } - - tmp_result = ci_unfreeze_sclk_mclk_dpm(adev); - if (tmp_result) { - DRM_ERROR("Failed to unfreeze SCLK MCLK DPM!"); - result = tmp_result; - } - - ci_find_min_clock_masks(adev, &sclk_mask, &mclk_mask, - request->min_sclk, request->min_mclk); - - if (sclk_mask) { - if (!pi->sclk_dpm_key_disabled) - amdgpu_ci_send_msg_to_smc_with_parameter( - adev, - PPSMC_MSG_SCLKDPM_SetEnabledMask, - pi->dpm_level_enable_mask. - sclk_dpm_enable_mask & - sclk_mask); - } - - if (mclk_mask) { - if (!pi->mclk_dpm_key_disabled) - amdgpu_ci_send_msg_to_smc_with_parameter( - adev, - PPSMC_MSG_MCLKDPM_SetEnabledMask, - pi->dpm_level_enable_mask. - mclk_dpm_enable_mask & - mclk_mask); - } - - - return result; -} - -static int ci_dpm_set_power_profile_state(void *handle, - struct amd_pp_profile *request) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - struct ci_power_info *pi = ci_get_pi(adev); - int ret = -1; - - if (!pi || !request) - return -EINVAL; - - if (adev->pm.dpm.forced_level != - AMD_DPM_FORCED_LEVEL_AUTO) - return -EINVAL; - - if (request->min_sclk || - request->min_mclk || - request->activity_threshold || - request->up_hyst || - request->down_hyst) { - if (request->type == AMD_PP_GFX_PROFILE) - memcpy(&pi->gfx_power_profile, request, - sizeof(struct amd_pp_profile)); - else if (request->type == AMD_PP_COMPUTE_PROFILE) - memcpy(&pi->compute_power_profile, request, - sizeof(struct amd_pp_profile)); - else - return -EINVAL; - - if (request->type == pi->current_power_profile) - ret = ci_set_power_profile_state( - adev, - request); - } else { - /* set power profile if it exists */ - switch (request->type) { - case AMD_PP_GFX_PROFILE: - ret = ci_set_power_profile_state( - adev, - &pi->gfx_power_profile); - break; - case AMD_PP_COMPUTE_PROFILE: - ret = ci_set_power_profile_state( - adev, - &pi->compute_power_profile); - break; - default: - return -EINVAL; - } - } - - if (!ret) - pi->current_power_profile = request->type; - - return 0; -} - -static int ci_dpm_reset_power_profile_state(void *handle, - struct amd_pp_profile *request) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - struct ci_power_info *pi = ci_get_pi(adev); - - if (!pi || !request) - return -EINVAL; - - if (request->type == AMD_PP_GFX_PROFILE) { - pi->gfx_power_profile = pi->default_gfx_power_profile; - return ci_dpm_set_power_profile_state(adev, - &pi->gfx_power_profile); - } else if (request->type == AMD_PP_COMPUTE_PROFILE) { - pi->compute_power_profile = - pi->default_compute_power_profile; - return ci_dpm_set_power_profile_state(adev, - &pi->compute_power_profile); - } else - return -EINVAL; -} - -static int ci_dpm_switch_power_profile(void *handle, - enum amd_pp_profile_type type) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - struct ci_power_info *pi = ci_get_pi(adev); - struct amd_pp_profile request = {0}; - - if (!pi) - return -EINVAL; - - if (pi->current_power_profile != type) { - request.type = type; - return ci_dpm_set_power_profile_state(adev, &request); - } - - return 0; -} - static int ci_dpm_read_sensor(void *handle, int idx, void *value, int *size) { @@ -7024,7 +6777,7 @@ static int ci_set_powergating_by_smu(void *handle, return 0; } -const struct amd_ip_funcs ci_dpm_ip_funcs = { +static const struct amd_ip_funcs ci_dpm_ip_funcs = { .name = "ci_dpm", .early_init = ci_dpm_early_init, .late_init = ci_dpm_late_init, @@ -7041,8 +6794,16 @@ const struct amd_ip_funcs ci_dpm_ip_funcs = { .set_powergating_state = ci_dpm_set_powergating_state, }; -const struct amd_pm_funcs ci_dpm_funcs = { - .get_temperature = &ci_dpm_get_temp, +const struct amdgpu_ip_block_version ci_smu_ip_block = +{ + .type = AMD_IP_BLOCK_TYPE_SMC, + .major = 7, + .minor = 0, + .rev = 0, + .funcs = &ci_dpm_ip_funcs, +}; + +static const struct amd_pm_funcs ci_dpm_funcs = { .pre_set_power_state = &ci_dpm_pre_set_power_state, .set_power_state = &ci_dpm_set_power_state, .post_set_power_state = &ci_dpm_post_set_power_state, @@ -7066,10 +6827,6 @@ const struct amd_pm_funcs ci_dpm_funcs = { .set_mclk_od = ci_dpm_set_mclk_od, .check_state_equal = ci_check_state_equal, .get_vce_clock_state = amdgpu_get_vce_clock_state, - .get_power_profile_state = ci_dpm_get_power_profile_state, - .set_power_profile_state = ci_dpm_set_power_profile_state, - .reset_power_profile_state = ci_dpm_reset_power_profile_state, - .switch_power_profile = ci_dpm_switch_power_profile, .read_sensor = ci_dpm_read_sensor, };
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.h b/drivers/gpu/drm/amd/amdgpu/ci_dpm.h index 84cbc9c..91be2996 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.h +++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.h
@@ -295,13 +295,6 @@ struct ci_power_info { bool fan_is_controlled_by_smc; u32 t_min; u32 fan_ctrl_default_mode; - - /* power profile */ - struct amd_pp_profile gfx_power_profile; - struct amd_pp_profile compute_power_profile; - struct amd_pp_profile default_gfx_power_profile; - struct amd_pp_profile default_compute_power_profile; - enum amd_pp_profile_type current_power_profile; }; #define CISLANDS_VOLTAGE_CONTROL_NONE 0x0
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c index 8e59e65..0df2203 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.c +++ b/drivers/gpu/drm/amd/amdgpu/cik.c
@@ -67,7 +67,6 @@ #include "amdgpu_dm.h" #include "amdgpu_amdkfd.h" -#include "amdgpu_powerplay.h" #include "dce_virtual.h" /* @@ -1715,6 +1714,27 @@ static void cik_detect_hw_virtualization(struct amdgpu_device *adev) adev->virt.caps |= AMDGPU_PASSTHROUGH_MODE; } +static void cik_flush_hdp(struct amdgpu_device *adev, struct amdgpu_ring *ring) +{ + if (!ring || !ring->funcs->emit_wreg) { + WREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 1); + RREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL); + } else { + amdgpu_ring_emit_wreg(ring, mmHDP_MEM_COHERENCY_FLUSH_CNTL, 1); + } +} + +static void cik_invalidate_hdp(struct amdgpu_device *adev, + struct amdgpu_ring *ring) +{ + if (!ring || !ring->funcs->emit_wreg) { + WREG32(mmHDP_DEBUG0, 1); + RREG32(mmHDP_DEBUG0); + } else { + amdgpu_ring_emit_wreg(ring, mmHDP_DEBUG0, 1); + } +} + static const struct amdgpu_asic_funcs cik_asic_funcs = { .read_disabled_bios = &cik_read_disabled_bios, @@ -1726,6 +1746,8 @@ static const struct amdgpu_asic_funcs cik_asic_funcs = .set_uvd_clocks = &cik_set_uvd_clocks, .set_vce_clocks = &cik_set_vce_clocks, .get_config_memsize = &cik_get_config_memsize, + .flush_hdp = &cik_flush_hdp, + .invalidate_hdp = &cik_invalidate_hdp, }; static int cik_common_early_init(void *handle) @@ -1864,10 +1886,6 @@ static int cik_common_early_init(void *handle) return -EINVAL; } - adev->firmware.load_type = amdgpu_ucode_get_load_type(adev, amdgpu_fw_load_type); - - amdgpu_device_get_pcie_info(adev); - return 0; } @@ -1977,7 +1995,10 @@ int cik_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &cik_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block); amdgpu_device_ip_block_add(adev, &cik_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + if (amdgpu_dpm == -1) + amdgpu_device_ip_block_add(adev, &ci_smu_ip_block); + else + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -1995,7 +2016,10 @@ int cik_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &cik_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block); amdgpu_device_ip_block_add(adev, &cik_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + if (amdgpu_dpm == -1) + amdgpu_device_ip_block_add(adev, &ci_smu_ip_block); + else + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -2013,7 +2037,7 @@ int cik_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &cik_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block); amdgpu_device_ip_block_add(adev, &cik_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &kv_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -2032,7 +2056,7 @@ int cik_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &cik_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block); amdgpu_device_ip_block_add(adev, &cik_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &kv_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC)
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.h b/drivers/gpu/drm/amd/amdgpu/cik.h index c4989f5..e49c6f1 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.h +++ b/drivers/gpu/drm/amd/amdgpu/cik.h
@@ -24,6 +24,8 @@ #ifndef __CIK_H__ #define __CIK_H__ +#define CIK_FLUSH_GPU_TLB_NUM_WREG 3 + void cik_srbm_select(struct amdgpu_device *adev, u32 me, u32 pipe, u32 queue, u32 vmid); int cik_set_ip_blocks(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/cik_dpm.h b/drivers/gpu/drm/amd/amdgpu/cik_dpm.h index c7b4349..2a08661 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_dpm.h +++ b/drivers/gpu/drm/amd/amdgpu/cik_dpm.h
@@ -24,8 +24,7 @@ #ifndef __CIK_DPM_H__ #define __CIK_DPM_H__ -extern const struct amd_ip_funcs ci_dpm_ip_funcs; -extern const struct amd_ip_funcs kv_dpm_ip_funcs; -extern const struct amd_pm_funcs ci_dpm_funcs; -extern const struct amd_pm_funcs kv_dpm_funcs; +extern const struct amdgpu_ip_block_version ci_smu_ip_block; +extern const struct amdgpu_ip_block_version kv_smu_ip_block; + #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/cik_ih.c b/drivers/gpu/drm/amd/amdgpu/cik_ih.c index d5a05c1..44d10c21 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_ih.c
@@ -111,7 +111,7 @@ static int cik_ih_irq_init(struct amdgpu_device *adev) cik_ih_disable_interrupts(adev); /* setup interrupt control */ - WREG32(mmINTERRUPT_CNTL2, adev->dummy_page.addr >> 8); + WREG32(mmINTERRUPT_CNTL2, adev->dummy_page_addr >> 8); interrupt_cntl = RREG32(mmINTERRUPT_CNTL); /* INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=0 - dummy read disabled with msi, enabled without msi * INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=1 - dummy read controlled by IH_DUMMY_RD_EN @@ -281,7 +281,7 @@ static void cik_ih_decode_iv(struct amdgpu_device *adev, entry->src_data[0] = dw[1] & 0xfffffff; entry->ring_id = dw[2] & 0xff; entry->vmid = (dw[2] >> 8) & 0xff; - entry->pas_id = (dw[2] >> 16) & 0xffff; + entry->pasid = (dw[2] >> 16) & 0xffff; /* wptr/rptr are in bytes! */ adev->irq.ih.rptr += 16;
diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index 0066da3..a757625 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
@@ -261,13 +261,6 @@ static void cik_sdma_ring_emit_hdp_flush(struct amdgpu_ring *ring) amdgpu_ring_write(ring, (0xfff << 16) | 10); /* retry count, poll interval */ } -static void cik_sdma_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000)); - amdgpu_ring_write(ring, mmHDP_DEBUG0); - amdgpu_ring_write(ring, 1); -} - /** * cik_sdma_ring_emit_fence - emit a fence on the DMA ring * @@ -317,7 +310,7 @@ static void cik_sdma_gfx_stop(struct amdgpu_device *adev) if ((adev->mman.buffer_funcs_ring == sdma0) || (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, false); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32(mmSDMA0_GFX_RB_CNTL + sdma_offsets[i]); @@ -517,7 +510,7 @@ static int cik_sdma_gfx_resume(struct amdgpu_device *adev) } if (adev->mman.buffer_funcs_ring == ring) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.real_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, true); } return 0; @@ -885,18 +878,7 @@ static void cik_sdma_ring_emit_vm_flush(struct amdgpu_ring *ring, u32 extra_bits = (SDMA_POLL_REG_MEM_EXTRA_OP(0) | SDMA_POLL_REG_MEM_EXTRA_FUNC(0)); /* always */ - amdgpu_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000)); - if (vmid < 8) { - amdgpu_ring_write(ring, (mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid)); - } else { - amdgpu_ring_write(ring, (mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8)); - } - amdgpu_ring_write(ring, pd_addr >> 12); - - /* flush TLB */ - amdgpu_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000)); - amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); amdgpu_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_POLL_REG_MEM, 0, extra_bits)); amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST << 2); @@ -906,6 +888,14 @@ static void cik_sdma_ring_emit_vm_flush(struct amdgpu_ring *ring, amdgpu_ring_write(ring, (0xfff << 16) | 10); /* retry count, poll interval */ } +static void cik_sdma_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ + amdgpu_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000)); + amdgpu_ring_write(ring, reg); + amdgpu_ring_write(ring, val); +} + static void cik_enable_sdma_mgcg(struct amdgpu_device *adev, bool enable) { @@ -1279,9 +1269,9 @@ static const struct amdgpu_ring_funcs cik_sdma_ring_funcs = { .set_wptr = cik_sdma_ring_set_wptr, .emit_frame_size = 6 + /* cik_sdma_ring_emit_hdp_flush */ - 3 + /* cik_sdma_ring_emit_hdp_invalidate */ + 3 + /* hdp invalidate */ 6 + /* cik_sdma_ring_emit_pipeline_sync */ - 12 + /* cik_sdma_ring_emit_vm_flush */ + CIK_FLUSH_GPU_TLB_NUM_WREG * 3 + 6 + /* cik_sdma_ring_emit_vm_flush */ 9 + 9 + 9, /* cik_sdma_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 7 + 4, /* cik_sdma_ring_emit_ib */ .emit_ib = cik_sdma_ring_emit_ib, @@ -1289,11 +1279,11 @@ static const struct amdgpu_ring_funcs cik_sdma_ring_funcs = { .emit_pipeline_sync = cik_sdma_ring_emit_pipeline_sync, .emit_vm_flush = cik_sdma_ring_emit_vm_flush, .emit_hdp_flush = cik_sdma_ring_emit_hdp_flush, - .emit_hdp_invalidate = cik_sdma_ring_emit_hdp_invalidate, .test_ring = cik_sdma_ring_test_ring, .test_ib = cik_sdma_ring_test_ib, .insert_nop = cik_sdma_ring_insert_nop, .pad_ib = cik_sdma_ring_pad_ib, + .emit_wreg = cik_sdma_ring_emit_wreg, }; static void cik_sdma_set_ring_funcs(struct amdgpu_device *adev) @@ -1391,9 +1381,6 @@ static const struct amdgpu_vm_pte_funcs cik_sdma_vm_pte_funcs = { .copy_pte = cik_sdma_vm_copy_pte, .write_pte = cik_sdma_vm_write_pte, - - .set_max_nums_pte_pde = 0x1fffff >> 3, - .set_pte_pde_num_dw = 10, .set_pte_pde = cik_sdma_vm_set_pte_pde, };
diff --git a/drivers/gpu/drm/amd/amdgpu/cz_ih.c b/drivers/gpu/drm/amd/amdgpu/cz_ih.c index f576e9c..960c29e 100644 --- a/drivers/gpu/drm/amd/amdgpu/cz_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/cz_ih.c
@@ -111,7 +111,7 @@ static int cz_ih_irq_init(struct amdgpu_device *adev) cz_ih_disable_interrupts(adev); /* setup interrupt control */ - WREG32(mmINTERRUPT_CNTL2, adev->dummy_page.addr >> 8); + WREG32(mmINTERRUPT_CNTL2, adev->dummy_page_addr >> 8); interrupt_cntl = RREG32(mmINTERRUPT_CNTL); /* INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=0 - dummy read disabled with msi, enabled without msi * INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=1 - dummy read controlled by IH_DUMMY_RD_EN @@ -260,7 +260,7 @@ static void cz_ih_decode_iv(struct amdgpu_device *adev, entry->src_data[0] = dw[1] & 0xfffffff; entry->ring_id = dw[2] & 0xff; entry->vmid = (dw[2] >> 8) & 0xff; - entry->pas_id = (dw[2] >> 16) & 0xffff; + entry->pasid = (dw[2] >> 16) & 0xffff; /* wptr/rptr are in bytes! */ adev->irq.ih.rptr += 16;
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index 022f303..452f88e 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -190,66 +190,6 @@ static void dce_v10_0_audio_endpt_wreg(struct amdgpu_device *adev, spin_unlock_irqrestore(&adev->audio_endpt_idx_lock, flags); } -static bool dce_v10_0_is_in_vblank(struct amdgpu_device *adev, int crtc) -{ - if (RREG32(mmCRTC_STATUS + crtc_offsets[crtc]) & - CRTC_V_BLANK_START_END__CRTC_V_BLANK_START_MASK) - return true; - else - return false; -} - -static bool dce_v10_0_is_counter_moving(struct amdgpu_device *adev, int crtc) -{ - u32 pos1, pos2; - - pos1 = RREG32(mmCRTC_STATUS_POSITION + crtc_offsets[crtc]); - pos2 = RREG32(mmCRTC_STATUS_POSITION + crtc_offsets[crtc]); - - if (pos1 != pos2) - return true; - else - return false; -} - -/** - * dce_v10_0_vblank_wait - vblank wait asic callback. - * - * @adev: amdgpu_device pointer - * @crtc: crtc to wait for vblank on - * - * Wait for vblank on the requested crtc (evergreen+). - */ -static void dce_v10_0_vblank_wait(struct amdgpu_device *adev, int crtc) -{ - unsigned i = 100; - - if (crtc >= adev->mode_info.num_crtc) - return; - - if (!(RREG32(mmCRTC_CONTROL + crtc_offsets[crtc]) & CRTC_CONTROL__CRTC_MASTER_EN_MASK)) - return; - - /* depending on when we hit vblank, we may be close to active; if so, - * wait for another frame. - */ - while (dce_v10_0_is_in_vblank(adev, crtc)) { - if (i++ == 100) { - i = 0; - if (!dce_v10_0_is_counter_moving(adev, crtc)) - break; - } - } - - while (!dce_v10_0_is_in_vblank(adev, crtc)) { - if (i++ == 100) { - i = 0; - if (!dce_v10_0_is_counter_moving(adev, crtc)) - break; - } - } -} - static u32 dce_v10_0_vblank_get_counter(struct amdgpu_device *adev, int crtc) { if (crtc >= adev->mode_info.num_crtc) @@ -1205,7 +1145,7 @@ static void dce_v10_0_bandwidth_update(struct amdgpu_device *adev) u32 num_heads = 0, lb_size; int i; - amdgpu_update_display_priority(adev); + amdgpu_display_update_priority(adev); for (i = 0; i < adev->mode_info.num_crtc; i++) { if (adev->mode_info.crtcs[i]->base.enabled) @@ -2517,9 +2457,9 @@ static const struct drm_crtc_funcs dce_v10_0_crtc_funcs = { .cursor_set2 = dce_v10_0_crtc_cursor_set2, .cursor_move = dce_v10_0_crtc_cursor_move, .gamma_set = dce_v10_0_crtc_gamma_set, - .set_config = amdgpu_crtc_set_config, + .set_config = amdgpu_display_crtc_set_config, .destroy = dce_v10_0_crtc_destroy, - .page_flip_target = amdgpu_crtc_page_flip_target, + .page_flip_target = amdgpu_display_crtc_page_flip_target, }; static void dce_v10_0_crtc_dpms(struct drm_crtc *crtc, int mode) @@ -2537,7 +2477,8 @@ static void dce_v10_0_crtc_dpms(struct drm_crtc *crtc, int mode) amdgpu_atombios_crtc_blank(crtc, ATOM_DISABLE); dce_v10_0_vga_enable(crtc, false); /* Make sure VBLANK and PFLIP interrupts are still enabled */ - type = amdgpu_crtc_idx_to_irq_type(adev, amdgpu_crtc->crtc_id); + type = amdgpu_display_crtc_idx_to_irq_type(adev, + amdgpu_crtc->crtc_id); amdgpu_irq_update(adev, &adev->crtc_irq, type); amdgpu_irq_update(adev, &adev->pageflip_irq, type); drm_crtc_vblank_on(crtc); @@ -2676,7 +2617,7 @@ static bool dce_v10_0_crtc_mode_fixup(struct drm_crtc *crtc, amdgpu_crtc->connector = NULL; return false; } - if (!amdgpu_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) + if (!amdgpu_display_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) return false; if (amdgpu_atombios_crtc_prepare_pll(crtc, adjusted_mode)) return false; @@ -2824,9 +2765,9 @@ static int dce_v10_0_sw_init(void *handle) adev->ddev->mode_config.preferred_depth = 24; adev->ddev->mode_config.prefer_shadow = 1; - adev->ddev->mode_config.fb_base = adev->mc.aper_base; + adev->ddev->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_modeset_create_props(adev); + r = amdgpu_display_modeset_create_props(adev); if (r) return r; @@ -2841,7 +2782,7 @@ static int dce_v10_0_sw_init(void *handle) } if (amdgpu_atombios_get_connector_info_from_object_table(adev)) - amdgpu_print_display_setup(adev->ddev); + amdgpu_display_print_display_setup(adev->ddev); else return -EINVAL; @@ -3257,7 +3198,7 @@ static int dce_v10_0_crtc_irq(struct amdgpu_device *adev, { unsigned crtc = entry->src_id - 1; uint32_t disp_int = RREG32(interrupt_status_offsets[crtc].reg); - unsigned irq_type = amdgpu_crtc_idx_to_irq_type(adev, crtc); + unsigned int irq_type = amdgpu_display_crtc_idx_to_irq_type(adev, crtc); switch (entry->src_data[0]) { case 0: /* vblank */ @@ -3609,7 +3550,6 @@ static void dce_v10_0_encoder_add(struct amdgpu_device *adev, static const struct amdgpu_display_funcs dce_v10_0_display_funcs = { .bandwidth_update = &dce_v10_0_bandwidth_update, .vblank_get_counter = &dce_v10_0_vblank_get_counter, - .vblank_wait = &dce_v10_0_vblank_wait, .backlight_set_level = &amdgpu_atombios_encoder_set_backlight_level, .backlight_get_level = &amdgpu_atombios_encoder_get_backlight_level, .hpd_sense = &dce_v10_0_hpd_sense,
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index 800a9f3..a7c1c58 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -207,66 +207,6 @@ static void dce_v11_0_audio_endpt_wreg(struct amdgpu_device *adev, spin_unlock_irqrestore(&adev->audio_endpt_idx_lock, flags); } -static bool dce_v11_0_is_in_vblank(struct amdgpu_device *adev, int crtc) -{ - if (RREG32(mmCRTC_STATUS + crtc_offsets[crtc]) & - CRTC_V_BLANK_START_END__CRTC_V_BLANK_START_MASK) - return true; - else - return false; -} - -static bool dce_v11_0_is_counter_moving(struct amdgpu_device *adev, int crtc) -{ - u32 pos1, pos2; - - pos1 = RREG32(mmCRTC_STATUS_POSITION + crtc_offsets[crtc]); - pos2 = RREG32(mmCRTC_STATUS_POSITION + crtc_offsets[crtc]); - - if (pos1 != pos2) - return true; - else - return false; -} - -/** - * dce_v11_0_vblank_wait - vblank wait asic callback. - * - * @adev: amdgpu_device pointer - * @crtc: crtc to wait for vblank on - * - * Wait for vblank on the requested crtc (evergreen+). - */ -static void dce_v11_0_vblank_wait(struct amdgpu_device *adev, int crtc) -{ - unsigned i = 100; - - if (crtc < 0 || crtc >= adev->mode_info.num_crtc) - return; - - if (!(RREG32(mmCRTC_CONTROL + crtc_offsets[crtc]) & CRTC_CONTROL__CRTC_MASTER_EN_MASK)) - return; - - /* depending on when we hit vblank, we may be close to active; if so, - * wait for another frame. - */ - while (dce_v11_0_is_in_vblank(adev, crtc)) { - if (i++ == 100) { - i = 0; - if (!dce_v11_0_is_counter_moving(adev, crtc)) - break; - } - } - - while (!dce_v11_0_is_in_vblank(adev, crtc)) { - if (i++ == 100) { - i = 0; - if (!dce_v11_0_is_counter_moving(adev, crtc)) - break; - } - } -} - static u32 dce_v11_0_vblank_get_counter(struct amdgpu_device *adev, int crtc) { if (crtc < 0 || crtc >= adev->mode_info.num_crtc) @@ -1229,7 +1169,7 @@ static void dce_v11_0_bandwidth_update(struct amdgpu_device *adev) u32 num_heads = 0, lb_size; int i; - amdgpu_update_display_priority(adev); + amdgpu_display_update_priority(adev); for (i = 0; i < adev->mode_info.num_crtc; i++) { if (adev->mode_info.crtcs[i]->base.enabled) @@ -2592,9 +2532,9 @@ static const struct drm_crtc_funcs dce_v11_0_crtc_funcs = { .cursor_set2 = dce_v11_0_crtc_cursor_set2, .cursor_move = dce_v11_0_crtc_cursor_move, .gamma_set = dce_v11_0_crtc_gamma_set, - .set_config = amdgpu_crtc_set_config, + .set_config = amdgpu_display_crtc_set_config, .destroy = dce_v11_0_crtc_destroy, - .page_flip_target = amdgpu_crtc_page_flip_target, + .page_flip_target = amdgpu_display_crtc_page_flip_target, }; static void dce_v11_0_crtc_dpms(struct drm_crtc *crtc, int mode) @@ -2612,7 +2552,8 @@ static void dce_v11_0_crtc_dpms(struct drm_crtc *crtc, int mode) amdgpu_atombios_crtc_blank(crtc, ATOM_DISABLE); dce_v11_0_vga_enable(crtc, false); /* Make sure VBLANK and PFLIP interrupts are still enabled */ - type = amdgpu_crtc_idx_to_irq_type(adev, amdgpu_crtc->crtc_id); + type = amdgpu_display_crtc_idx_to_irq_type(adev, + amdgpu_crtc->crtc_id); amdgpu_irq_update(adev, &adev->crtc_irq, type); amdgpu_irq_update(adev, &adev->pageflip_irq, type); drm_crtc_vblank_on(crtc); @@ -2779,7 +2720,7 @@ static bool dce_v11_0_crtc_mode_fixup(struct drm_crtc *crtc, amdgpu_crtc->connector = NULL; return false; } - if (!amdgpu_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) + if (!amdgpu_display_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) return false; if (amdgpu_atombios_crtc_prepare_pll(crtc, adjusted_mode)) return false; @@ -2939,9 +2880,9 @@ static int dce_v11_0_sw_init(void *handle) adev->ddev->mode_config.preferred_depth = 24; adev->ddev->mode_config.prefer_shadow = 1; - adev->ddev->mode_config.fb_base = adev->mc.aper_base; + adev->ddev->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_modeset_create_props(adev); + r = amdgpu_display_modeset_create_props(adev); if (r) return r; @@ -2957,7 +2898,7 @@ static int dce_v11_0_sw_init(void *handle) } if (amdgpu_atombios_get_connector_info_from_object_table(adev)) - amdgpu_print_display_setup(adev->ddev); + amdgpu_display_print_display_setup(adev->ddev); else return -EINVAL; @@ -3376,7 +3317,8 @@ static int dce_v11_0_crtc_irq(struct amdgpu_device *adev, { unsigned crtc = entry->src_id - 1; uint32_t disp_int = RREG32(interrupt_status_offsets[crtc].reg); - unsigned irq_type = amdgpu_crtc_idx_to_irq_type(adev, crtc); + unsigned int irq_type = amdgpu_display_crtc_idx_to_irq_type(adev, + crtc); switch (entry->src_data[0]) { case 0: /* vblank */ @@ -3733,7 +3675,6 @@ static void dce_v11_0_encoder_add(struct amdgpu_device *adev, static const struct amdgpu_display_funcs dce_v11_0_display_funcs = { .bandwidth_update = &dce_v11_0_bandwidth_update, .vblank_get_counter = &dce_v11_0_vblank_get_counter, - .vblank_wait = &dce_v11_0_vblank_wait, .backlight_set_level = &amdgpu_atombios_encoder_set_backlight_level, .backlight_get_level = &amdgpu_atombios_encoder_get_backlight_level, .hpd_sense = &dce_v11_0_hpd_sense,
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index b8368f6..9f67b7f 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
@@ -142,64 +142,6 @@ static void dce_v6_0_audio_endpt_wreg(struct amdgpu_device *adev, spin_unlock_irqrestore(&adev->audio_endpt_idx_lock, flags); } -static bool dce_v6_0_is_in_vblank(struct amdgpu_device *adev, int crtc) -{ - if (RREG32(mmCRTC_STATUS + crtc_offsets[crtc]) & CRTC_STATUS__CRTC_V_BLANK_MASK) - return true; - else - return false; -} - -static bool dce_v6_0_is_counter_moving(struct amdgpu_device *adev, int crtc) -{ - u32 pos1, pos2; - - pos1 = RREG32(mmCRTC_STATUS_POSITION + crtc_offsets[crtc]); - pos2 = RREG32(mmCRTC_STATUS_POSITION + crtc_offsets[crtc]); - - if (pos1 != pos2) - return true; - else - return false; -} - -/** - * dce_v6_0_wait_for_vblank - vblank wait asic callback. - * - * @crtc: crtc to wait for vblank on - * - * Wait for vblank on the requested crtc (evergreen+). - */ -static void dce_v6_0_vblank_wait(struct amdgpu_device *adev, int crtc) -{ - unsigned i = 100; - - if (crtc >= adev->mode_info.num_crtc) - return; - - if (!(RREG32(mmCRTC_CONTROL + crtc_offsets[crtc]) & CRTC_CONTROL__CRTC_MASTER_EN_MASK)) - return; - - /* depending on when we hit vblank, we may be close to active; if so, - * wait for another frame. - */ - while (dce_v6_0_is_in_vblank(adev, crtc)) { - if (i++ == 100) { - i = 0; - if (!dce_v6_0_is_counter_moving(adev, crtc)) - break; - } - } - - while (!dce_v6_0_is_in_vblank(adev, crtc)) { - if (i++ == 100) { - i = 0; - if (!dce_v6_0_is_counter_moving(adev, crtc)) - break; - } - } -} - static u32 dce_v6_0_vblank_get_counter(struct amdgpu_device *adev, int crtc) { if (crtc >= adev->mode_info.num_crtc) @@ -1108,7 +1050,7 @@ static void dce_v6_0_bandwidth_update(struct amdgpu_device *adev) if (!adev->mode_info.mode_config_initialized) return; - amdgpu_update_display_priority(adev); + amdgpu_display_update_priority(adev); for (i = 0; i < adev->mode_info.num_crtc; i++) { if (adev->mode_info.crtcs[i]->base.enabled) @@ -2407,9 +2349,9 @@ static const struct drm_crtc_funcs dce_v6_0_crtc_funcs = { .cursor_set2 = dce_v6_0_crtc_cursor_set2, .cursor_move = dce_v6_0_crtc_cursor_move, .gamma_set = dce_v6_0_crtc_gamma_set, - .set_config = amdgpu_crtc_set_config, + .set_config = amdgpu_display_crtc_set_config, .destroy = dce_v6_0_crtc_destroy, - .page_flip_target = amdgpu_crtc_page_flip_target, + .page_flip_target = amdgpu_display_crtc_page_flip_target, }; static void dce_v6_0_crtc_dpms(struct drm_crtc *crtc, int mode) @@ -2425,7 +2367,8 @@ static void dce_v6_0_crtc_dpms(struct drm_crtc *crtc, int mode) amdgpu_atombios_crtc_enable(crtc, ATOM_ENABLE); amdgpu_atombios_crtc_blank(crtc, ATOM_DISABLE); /* Make sure VBLANK and PFLIP interrupts are still enabled */ - type = amdgpu_crtc_idx_to_irq_type(adev, amdgpu_crtc->crtc_id); + type = amdgpu_display_crtc_idx_to_irq_type(adev, + amdgpu_crtc->crtc_id); amdgpu_irq_update(adev, &adev->crtc_irq, type); amdgpu_irq_update(adev, &adev->pageflip_irq, type); drm_crtc_vblank_on(crtc); @@ -2562,7 +2505,7 @@ static bool dce_v6_0_crtc_mode_fixup(struct drm_crtc *crtc, amdgpu_crtc->connector = NULL; return false; } - if (!amdgpu_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) + if (!amdgpu_display_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) return false; if (amdgpu_atombios_crtc_prepare_pll(crtc, adjusted_mode)) return false; @@ -2693,9 +2636,9 @@ static int dce_v6_0_sw_init(void *handle) adev->ddev->mode_config.max_height = 16384; adev->ddev->mode_config.preferred_depth = 24; adev->ddev->mode_config.prefer_shadow = 1; - adev->ddev->mode_config.fb_base = adev->mc.aper_base; + adev->ddev->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_modeset_create_props(adev); + r = amdgpu_display_modeset_create_props(adev); if (r) return r; @@ -2711,7 +2654,7 @@ static int dce_v6_0_sw_init(void *handle) ret = amdgpu_atombios_get_connector_info_from_object_table(adev); if (ret) - amdgpu_print_display_setup(adev->ddev); + amdgpu_display_print_display_setup(adev->ddev); else return -EINVAL; @@ -2974,7 +2917,8 @@ static int dce_v6_0_crtc_irq(struct amdgpu_device *adev, { unsigned crtc = entry->src_id - 1; uint32_t disp_int = RREG32(interrupt_status_offsets[crtc].reg); - unsigned irq_type = amdgpu_crtc_idx_to_irq_type(adev, crtc); + unsigned int irq_type = amdgpu_display_crtc_idx_to_irq_type(adev, + crtc); switch (entry->src_data[0]) { case 0: /* vblank */ @@ -3415,7 +3359,6 @@ static void dce_v6_0_encoder_add(struct amdgpu_device *adev, static const struct amdgpu_display_funcs dce_v6_0_display_funcs = { .bandwidth_update = &dce_v6_0_bandwidth_update, .vblank_get_counter = &dce_v6_0_vblank_get_counter, - .vblank_wait = &dce_v6_0_vblank_wait, .backlight_set_level = &amdgpu_atombios_encoder_set_backlight_level, .backlight_get_level = &amdgpu_atombios_encoder_get_backlight_level, .hpd_sense = &dce_v6_0_hpd_sense,
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index 012e0a9..f55422c 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -140,66 +140,6 @@ static void dce_v8_0_audio_endpt_wreg(struct amdgpu_device *adev, spin_unlock_irqrestore(&adev->audio_endpt_idx_lock, flags); } -static bool dce_v8_0_is_in_vblank(struct amdgpu_device *adev, int crtc) -{ - if (RREG32(mmCRTC_STATUS + crtc_offsets[crtc]) & - CRTC_V_BLANK_START_END__CRTC_V_BLANK_START_MASK) - return true; - else - return false; -} - -static bool dce_v8_0_is_counter_moving(struct amdgpu_device *adev, int crtc) -{ - u32 pos1, pos2; - - pos1 = RREG32(mmCRTC_STATUS_POSITION + crtc_offsets[crtc]); - pos2 = RREG32(mmCRTC_STATUS_POSITION + crtc_offsets[crtc]); - - if (pos1 != pos2) - return true; - else - return false; -} - -/** - * dce_v8_0_vblank_wait - vblank wait asic callback. - * - * @adev: amdgpu_device pointer - * @crtc: crtc to wait for vblank on - * - * Wait for vblank on the requested crtc (evergreen+). - */ -static void dce_v8_0_vblank_wait(struct amdgpu_device *adev, int crtc) -{ - unsigned i = 100; - - if (crtc >= adev->mode_info.num_crtc) - return; - - if (!(RREG32(mmCRTC_CONTROL + crtc_offsets[crtc]) & CRTC_CONTROL__CRTC_MASTER_EN_MASK)) - return; - - /* depending on when we hit vblank, we may be close to active; if so, - * wait for another frame. - */ - while (dce_v8_0_is_in_vblank(adev, crtc)) { - if (i++ == 100) { - i = 0; - if (!dce_v8_0_is_counter_moving(adev, crtc)) - break; - } - } - - while (!dce_v8_0_is_in_vblank(adev, crtc)) { - if (i++ == 100) { - i = 0; - if (!dce_v8_0_is_counter_moving(adev, crtc)) - break; - } - } -} - static u32 dce_v8_0_vblank_get_counter(struct amdgpu_device *adev, int crtc) { if (crtc >= adev->mode_info.num_crtc) @@ -1144,7 +1084,7 @@ static void dce_v8_0_bandwidth_update(struct amdgpu_device *adev) u32 num_heads = 0, lb_size; int i; - amdgpu_update_display_priority(adev); + amdgpu_display_update_priority(adev); for (i = 0; i < adev->mode_info.num_crtc; i++) { if (adev->mode_info.crtcs[i]->base.enabled) @@ -2421,9 +2361,9 @@ static const struct drm_crtc_funcs dce_v8_0_crtc_funcs = { .cursor_set2 = dce_v8_0_crtc_cursor_set2, .cursor_move = dce_v8_0_crtc_cursor_move, .gamma_set = dce_v8_0_crtc_gamma_set, - .set_config = amdgpu_crtc_set_config, + .set_config = amdgpu_display_crtc_set_config, .destroy = dce_v8_0_crtc_destroy, - .page_flip_target = amdgpu_crtc_page_flip_target, + .page_flip_target = amdgpu_display_crtc_page_flip_target, }; static void dce_v8_0_crtc_dpms(struct drm_crtc *crtc, int mode) @@ -2441,7 +2381,8 @@ static void dce_v8_0_crtc_dpms(struct drm_crtc *crtc, int mode) amdgpu_atombios_crtc_blank(crtc, ATOM_DISABLE); dce_v8_0_vga_enable(crtc, false); /* Make sure VBLANK and PFLIP interrupts are still enabled */ - type = amdgpu_crtc_idx_to_irq_type(adev, amdgpu_crtc->crtc_id); + type = amdgpu_display_crtc_idx_to_irq_type(adev, + amdgpu_crtc->crtc_id); amdgpu_irq_update(adev, &adev->crtc_irq, type); amdgpu_irq_update(adev, &adev->pageflip_irq, type); drm_crtc_vblank_on(crtc); @@ -2587,7 +2528,7 @@ static bool dce_v8_0_crtc_mode_fixup(struct drm_crtc *crtc, amdgpu_crtc->connector = NULL; return false; } - if (!amdgpu_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) + if (!amdgpu_display_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) return false; if (amdgpu_atombios_crtc_prepare_pll(crtc, adjusted_mode)) return false; @@ -2724,9 +2665,9 @@ static int dce_v8_0_sw_init(void *handle) adev->ddev->mode_config.preferred_depth = 24; adev->ddev->mode_config.prefer_shadow = 1; - adev->ddev->mode_config.fb_base = adev->mc.aper_base; + adev->ddev->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_modeset_create_props(adev); + r = amdgpu_display_modeset_create_props(adev); if (r) return r; @@ -2741,7 +2682,7 @@ static int dce_v8_0_sw_init(void *handle) } if (amdgpu_atombios_get_connector_info_from_object_table(adev)) - amdgpu_print_display_setup(adev->ddev); + amdgpu_display_print_display_setup(adev->ddev); else return -EINVAL; @@ -3071,7 +3012,8 @@ static int dce_v8_0_crtc_irq(struct amdgpu_device *adev, { unsigned crtc = entry->src_id - 1; uint32_t disp_int = RREG32(interrupt_status_offsets[crtc].reg); - unsigned irq_type = amdgpu_crtc_idx_to_irq_type(adev, crtc); + unsigned int irq_type = amdgpu_display_crtc_idx_to_irq_type(adev, + crtc); switch (entry->src_data[0]) { case 0: /* vblank */ @@ -3499,7 +3441,6 @@ static void dce_v8_0_encoder_add(struct amdgpu_device *adev, static const struct amdgpu_display_funcs dce_v8_0_display_funcs = { .bandwidth_update = &dce_v8_0_bandwidth_update, .vblank_get_counter = &dce_v8_0_vblank_get_counter, - .vblank_wait = &dce_v8_0_vblank_wait, .backlight_set_level = &amdgpu_atombios_encoder_set_backlight_level, .backlight_get_level = &amdgpu_atombios_encoder_get_backlight_level, .hpd_sense = &dce_v8_0_hpd_sense,
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c index 120dd3b..b51f05d 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -48,19 +48,6 @@ static void dce_virtual_set_crtc_vblank_interrupt_state(struct amdgpu_device *ad int crtc, enum amdgpu_interrupt_state state); -/** - * dce_virtual_vblank_wait - vblank wait asic callback. - * - * @adev: amdgpu_device pointer - * @crtc: crtc to wait for vblank on - * - * Wait for vblank on the requested crtc (evergreen+). - */ -static void dce_virtual_vblank_wait(struct amdgpu_device *adev, int crtc) -{ - return; -} - static u32 dce_virtual_vblank_get_counter(struct amdgpu_device *adev, int crtc) { return 0; @@ -130,9 +117,9 @@ static const struct drm_crtc_funcs dce_virtual_crtc_funcs = { .cursor_set2 = NULL, .cursor_move = NULL, .gamma_set = dce_virtual_crtc_gamma_set, - .set_config = amdgpu_crtc_set_config, + .set_config = amdgpu_display_crtc_set_config, .destroy = dce_virtual_crtc_destroy, - .page_flip_target = amdgpu_crtc_page_flip_target, + .page_flip_target = amdgpu_display_crtc_page_flip_target, }; static void dce_virtual_crtc_dpms(struct drm_crtc *crtc, int mode) @@ -149,7 +136,8 @@ static void dce_virtual_crtc_dpms(struct drm_crtc *crtc, int mode) case DRM_MODE_DPMS_ON: amdgpu_crtc->enabled = true; /* Make sure VBLANK interrupts are still enabled */ - type = amdgpu_crtc_idx_to_irq_type(adev, amdgpu_crtc->crtc_id); + type = amdgpu_display_crtc_idx_to_irq_type(adev, + amdgpu_crtc->crtc_id); amdgpu_irq_update(adev, &adev->crtc_irq, type); drm_crtc_vblank_on(crtc); break; @@ -406,9 +394,9 @@ static int dce_virtual_sw_init(void *handle) adev->ddev->mode_config.preferred_depth = 24; adev->ddev->mode_config.prefer_shadow = 1; - adev->ddev->mode_config.fb_base = adev->mc.aper_base; + adev->ddev->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_modeset_create_props(adev); + r = amdgpu_display_modeset_create_props(adev); if (r) return r; @@ -485,6 +473,7 @@ static int dce_virtual_hw_init(void *handle) /* no DCE */ break; case CHIP_VEGA10: + case CHIP_VEGA12: break; default: DRM_ERROR("Virtual display unsupported ASIC type: 0x%X\n", adev->asic_type); @@ -653,7 +642,6 @@ static int dce_virtual_connector_encoder_init(struct amdgpu_device *adev, static const struct amdgpu_display_funcs dce_virtual_display_funcs = { .bandwidth_update = &dce_virtual_bandwidth_update, .vblank_get_counter = &dce_virtual_vblank_get_counter, - .vblank_wait = &dce_virtual_vblank_wait, .backlight_set_level = NULL, .backlight_get_level = NULL, .hpd_sense = &dce_virtual_hpd_sense,
diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_acpi.h b/drivers/gpu/drm/amd/amdgpu/emu_soc.c similarity index 81% rename from drivers/gpu/drm/amd/powerplay/inc/pp_acpi.h rename to drivers/gpu/drm/amd/amdgpu/emu_soc.c index 8fe8ba9..d72c25c 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/pp_acpi.h +++ b/drivers/gpu/drm/amd/amdgpu/emu_soc.c
@@ -1,5 +1,5 @@ /* - * Copyright 2015 Advanced Micro Devices, Inc. + * Copyright 2018 Advanced Micro Devices, Inc. * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), @@ -20,7 +20,14 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ +#include "amdgpu.h" +#include "soc15.h" -bool acpi_atcs_functions_supported(void *device, uint32_t index); -int acpi_pcie_perf_request(void *device, uint8_t perf_req, bool advertise); -bool acpi_atcs_notify_pcie_device_ready(void *device); +#include "soc15_common.h" +#include "soc15_hw_ip.h" + +int emu_soc_asic_init(struct amdgpu_device *adev) +{ + return 0; +} +
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index 9870d83..cd6bf29 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
@@ -38,6 +38,7 @@ #include "dce/dce_6_0_sh_mask.h" #include "gca/gfx_7_2_enum.h" #include "si_enums.h" +#include "si.h" static void gfx_v6_0_set_ring_funcs(struct amdgpu_device *adev); static void gfx_v6_0_set_irq_funcs(struct amdgpu_device *adev); @@ -1808,17 +1809,6 @@ static int gfx_v6_0_ring_test_ring(struct amdgpu_ring *ring) return r; } -static void gfx_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) -{ - /* flush hdp cache */ - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) | - WRITE_DATA_DST_SEL(0))); - amdgpu_ring_write(ring, mmHDP_MEM_COHERENCY_FLUSH_CNTL); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 0x1); -} - static void gfx_v6_0_ring_emit_vgt_flush(struct amdgpu_ring *ring) { amdgpu_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE, 0)); @@ -1826,24 +1816,6 @@ static void gfx_v6_0_ring_emit_vgt_flush(struct amdgpu_ring *ring) EVENT_INDEX(0)); } -/** - * gfx_v6_0_ring_emit_hdp_invalidate - emit an hdp invalidate on the cp - * - * @adev: amdgpu_device pointer - * @ridx: amdgpu ring index - * - * Emits an hdp invalidate on the cp. - */ -static void gfx_v6_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) | - WRITE_DATA_DST_SEL(0))); - amdgpu_ring_write(ring, mmHDP_DEBUG0); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 0x1); -} - static void gfx_v6_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq, unsigned flags) { @@ -2358,25 +2330,7 @@ static void gfx_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring, { int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); - /* write new base address */ - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) | - WRITE_DATA_DST_SEL(0))); - if (vmid < 8) { - amdgpu_ring_write(ring, (mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid )); - } else { - amdgpu_ring_write(ring, (mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (vmid - 8))); - } - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, pd_addr >> 12); - - /* bits 0-15 are the VM contexts0-15 */ - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) | - WRITE_DATA_DST_SEL(0))); - amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); /* wait for the invalidate to complete */ amdgpu_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5)); @@ -2401,6 +2355,18 @@ static void gfx_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring, } } +static void gfx_v6_0_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ + int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); + + amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); + amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(usepfp) | + WRITE_DATA_DST_SEL(0))); + amdgpu_ring_write(ring, reg); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, val); +} static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev) { @@ -3095,11 +3061,18 @@ static void gfx_v6_0_read_wave_sgprs(struct amdgpu_device *adev, uint32_t simd, start + SQIND_WAVE_SGPRS_OFFSET, size, dst); } +static void gfx_v6_0_select_me_pipe_q(struct amdgpu_device *adev, + u32 me, u32 pipe, u32 q) +{ + DRM_INFO("Not implemented\n"); +} + static const struct amdgpu_gfx_funcs gfx_v6_0_gfx_funcs = { .get_gpu_clock_counter = &gfx_v6_0_get_gpu_clock_counter, .select_se_sh = &gfx_v6_0_select_se_sh, .read_wave_data = &gfx_v6_0_read_wave_data, .read_wave_sgprs = &gfx_v6_0_read_wave_sgprs, + .select_me_pipe_q = &gfx_v6_0_select_me_pipe_q }; static int gfx_v6_0_early_init(void *handle) @@ -3511,23 +3484,21 @@ static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_gfx = { .get_wptr = gfx_v6_0_ring_get_wptr, .set_wptr = gfx_v6_0_ring_set_wptr_gfx, .emit_frame_size = - 5 + /* gfx_v6_0_ring_emit_hdp_flush */ - 5 + /* gfx_v6_0_ring_emit_hdp_invalidate */ + 5 + 5 + /* hdp flush / invalidate */ 14 + 14 + 14 + /* gfx_v6_0_ring_emit_fence x3 for user fence, vm fence */ 7 + 4 + /* gfx_v6_0_ring_emit_pipeline_sync */ - 17 + 6 + /* gfx_v6_0_ring_emit_vm_flush */ + SI_FLUSH_GPU_TLB_NUM_WREG * 5 + 7 + 6 + /* gfx_v6_0_ring_emit_vm_flush */ 3 + 2, /* gfx_v6_ring_emit_cntxcntl including vgt flush */ .emit_ib_size = 6, /* gfx_v6_0_ring_emit_ib */ .emit_ib = gfx_v6_0_ring_emit_ib, .emit_fence = gfx_v6_0_ring_emit_fence, .emit_pipeline_sync = gfx_v6_0_ring_emit_pipeline_sync, .emit_vm_flush = gfx_v6_0_ring_emit_vm_flush, - .emit_hdp_flush = gfx_v6_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v6_0_ring_emit_hdp_invalidate, .test_ring = gfx_v6_0_ring_test_ring, .test_ib = gfx_v6_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .emit_cntxcntl = gfx_v6_ring_emit_cntxcntl, + .emit_wreg = gfx_v6_0_ring_emit_wreg, }; static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_compute = { @@ -3538,21 +3509,19 @@ static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_compute = { .get_wptr = gfx_v6_0_ring_get_wptr, .set_wptr = gfx_v6_0_ring_set_wptr_compute, .emit_frame_size = - 5 + /* gfx_v6_0_ring_emit_hdp_flush */ - 5 + /* gfx_v6_0_ring_emit_hdp_invalidate */ + 5 + 5 + /* hdp flush / invalidate */ 7 + /* gfx_v6_0_ring_emit_pipeline_sync */ - 17 + /* gfx_v6_0_ring_emit_vm_flush */ + SI_FLUSH_GPU_TLB_NUM_WREG * 5 + 7 + /* gfx_v6_0_ring_emit_vm_flush */ 14 + 14 + 14, /* gfx_v6_0_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 6, /* gfx_v6_0_ring_emit_ib */ .emit_ib = gfx_v6_0_ring_emit_ib, .emit_fence = gfx_v6_0_ring_emit_fence, .emit_pipeline_sync = gfx_v6_0_ring_emit_pipeline_sync, .emit_vm_flush = gfx_v6_0_ring_emit_vm_flush, - .emit_hdp_flush = gfx_v6_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v6_0_ring_emit_hdp_invalidate, .test_ring = gfx_v6_0_ring_test_ring, .test_ib = gfx_v6_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, + .emit_wreg = gfx_v6_0_ring_emit_wreg, }; static void gfx_v6_0_set_ring_funcs(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index a430969..42b6144 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
@@ -1946,7 +1946,7 @@ static void gfx_v7_0_gpu_init(struct amdgpu_device *adev) if (i == 0) sh_mem_base = 0; else - sh_mem_base = adev->mc.shared_aperture_start >> 48; + sh_mem_base = adev->gmc.shared_aperture_start >> 48; cik_srbm_select(adev, 0, 0, 0, i); /* CP and shaders */ WREG32(mmSH_MEM_CONFIG, sh_mem_cfg); @@ -2147,26 +2147,6 @@ static void gfx_v7_0_ring_emit_vgt_flush(struct amdgpu_ring *ring) EVENT_INDEX(0)); } - -/** - * gfx_v7_0_ring_emit_hdp_invalidate - emit an hdp invalidate on the cp - * - * @adev: amdgpu_device pointer - * @ridx: amdgpu ring index - * - * Emits an hdp invalidate on the cp. - */ -static void gfx_v7_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) | - WRITE_DATA_DST_SEL(0) | - WR_CONFIRM)); - amdgpu_ring_write(ring, mmHDP_DEBUG0); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 1); -} - /** * gfx_v7_0_ring_emit_fence_gfx - emit a fence on the gfx ring * @@ -3243,26 +3223,7 @@ static void gfx_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring, { int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(usepfp) | - WRITE_DATA_DST_SEL(0))); - if (vmid < 8) { - amdgpu_ring_write(ring, - (mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid)); - } else { - amdgpu_ring_write(ring, - (mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8)); - } - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, pd_addr >> 12); - - /* bits 0-15 are the VM contexts0-15 */ - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) | - WRITE_DATA_DST_SEL(0))); - amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); /* wait for the invalidate to complete */ amdgpu_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5)); @@ -3289,6 +3250,19 @@ static void gfx_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring, } } +static void gfx_v7_0_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ + int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); + + amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); + amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(usepfp) | + WRITE_DATA_DST_SEL(0))); + amdgpu_ring_write(ring, reg); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, val); +} + /* * RLC * The RLC is a multi-purpose microengine that handles a @@ -4296,11 +4270,18 @@ static void gfx_v7_0_read_wave_sgprs(struct amdgpu_device *adev, uint32_t simd, start + SQIND_WAVE_SGPRS_OFFSET, size, dst); } +static void gfx_v7_0_select_me_pipe_q(struct amdgpu_device *adev, + u32 me, u32 pipe, u32 q) +{ + cik_srbm_select(adev, me, pipe, q, 0); +} + static const struct amdgpu_gfx_funcs gfx_v7_0_gfx_funcs = { .get_gpu_clock_counter = &gfx_v7_0_get_gpu_clock_counter, .select_se_sh = &gfx_v7_0_select_se_sh, .read_wave_data = &gfx_v7_0_read_wave_data, .read_wave_sgprs = &gfx_v7_0_read_wave_sgprs, + .select_me_pipe_q = &gfx_v7_0_select_me_pipe_q }; static const struct amdgpu_rlc_funcs gfx_v7_0_rlc_funcs = { @@ -5089,10 +5070,10 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = { .emit_frame_size = 20 + /* gfx_v7_0_ring_emit_gds_switch */ 7 + /* gfx_v7_0_ring_emit_hdp_flush */ - 5 + /* gfx_v7_0_ring_emit_hdp_invalidate */ + 5 + /* hdp invalidate */ 12 + 12 + 12 + /* gfx_v7_0_ring_emit_fence_gfx x3 for user fence, vm fence */ 7 + 4 + /* gfx_v7_0_ring_emit_pipeline_sync */ - 17 + 6 + /* gfx_v7_0_ring_emit_vm_flush */ + CIK_FLUSH_GPU_TLB_NUM_WREG * 5 + 7 + 6 + /* gfx_v7_0_ring_emit_vm_flush */ 3 + 4, /* gfx_v7_ring_emit_cntxcntl including vgt flush*/ .emit_ib_size = 4, /* gfx_v7_0_ring_emit_ib_gfx */ .emit_ib = gfx_v7_0_ring_emit_ib_gfx, @@ -5101,12 +5082,12 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = { .emit_vm_flush = gfx_v7_0_ring_emit_vm_flush, .emit_gds_switch = gfx_v7_0_ring_emit_gds_switch, .emit_hdp_flush = gfx_v7_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v7_0_ring_emit_hdp_invalidate, .test_ring = gfx_v7_0_ring_test_ring, .test_ib = gfx_v7_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, .emit_cntxcntl = gfx_v7_ring_emit_cntxcntl, + .emit_wreg = gfx_v7_0_ring_emit_wreg, }; static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = { @@ -5120,9 +5101,9 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = { .emit_frame_size = 20 + /* gfx_v7_0_ring_emit_gds_switch */ 7 + /* gfx_v7_0_ring_emit_hdp_flush */ - 5 + /* gfx_v7_0_ring_emit_hdp_invalidate */ + 5 + /* hdp invalidate */ 7 + /* gfx_v7_0_ring_emit_pipeline_sync */ - 17 + /* gfx_v7_0_ring_emit_vm_flush */ + CIK_FLUSH_GPU_TLB_NUM_WREG * 5 + 7 + /* gfx_v7_0_ring_emit_vm_flush */ 7 + 7 + 7, /* gfx_v7_0_ring_emit_fence_compute x3 for user fence, vm fence */ .emit_ib_size = 4, /* gfx_v7_0_ring_emit_ib_compute */ .emit_ib = gfx_v7_0_ring_emit_ib_compute, @@ -5131,11 +5112,11 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = { .emit_vm_flush = gfx_v7_0_ring_emit_vm_flush, .emit_gds_switch = gfx_v7_0_ring_emit_gds_switch, .emit_hdp_flush = gfx_v7_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v7_0_ring_emit_hdp_invalidate, .test_ring = gfx_v7_0_ring_test_ring, .test_ib = gfx_v7_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, + .emit_wreg = gfx_v7_0_ring_emit_wreg, }; static void gfx_v7_0_set_ring_funcs(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 9cf53fc..f459e00 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -3479,6 +3479,12 @@ static void gfx_v8_0_select_se_sh(struct amdgpu_device *adev, WREG32(mmGRBM_GFX_INDEX, data); } +static void gfx_v8_0_select_me_pipe_q(struct amdgpu_device *adev, + u32 me, u32 pipe, u32 q) +{ + vi_srbm_select(adev, me, pipe, q, 0); +} + static u32 gfx_v8_0_get_rb_active_bitmap(struct amdgpu_device *adev) { u32 data, mask; @@ -3800,7 +3806,7 @@ static void gfx_v8_0_gpu_init(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, SH_MEM_CONFIG, ALIGNMENT_MODE, SH_MEM_ALIGNMENT_MODE_UNALIGNED); WREG32(mmSH_MEM_CONFIG, tmp); - tmp = adev->mc.shared_aperture_start >> 48; + tmp = adev->gmc.shared_aperture_start >> 48; WREG32(mmSH_MEM_BASES, tmp); } @@ -4832,6 +4838,9 @@ static int gfx_v8_0_kcq_init_queue(struct amdgpu_ring *ring) /* reset MQD to a clean status */ if (adev->gfx.mec.mqd_backup[mqd_idx]) memcpy(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(struct vi_mqd_allocation)); + /* reset ring buffer */ + ring->wptr = 0; + amdgpu_ring_clear_ring(ring); } else { amdgpu_ring_clear_ring(ring); } @@ -4906,13 +4915,6 @@ static int gfx_v8_0_kiq_resume(struct amdgpu_device *adev) /* Test KCQs */ for (i = 0; i < adev->gfx.num_compute_rings; i++) { ring = &adev->gfx.compute_ring[i]; - if (adev->in_gpu_reset) { - /* move reset ring buffer to here to workaround - * compute ring test failed - */ - ring->wptr = 0; - amdgpu_ring_clear_ring(ring); - } ring->ready = true; r = amdgpu_ring_test_ring(ring); if (r) @@ -5459,6 +5461,7 @@ static const struct amdgpu_gfx_funcs gfx_v8_0_gfx_funcs = { .select_se_sh = &gfx_v8_0_select_se_sh, .read_wave_data = &gfx_v8_0_read_wave_data, .read_wave_sgprs = &gfx_v8_0_read_wave_sgprs, + .select_me_pipe_q = &gfx_v8_0_select_me_pipe_q }; static int gfx_v8_0_early_init(void *handle) @@ -6248,19 +6251,6 @@ static void gfx_v8_0_ring_emit_vgt_flush(struct amdgpu_ring *ring) EVENT_INDEX(0)); } - -static void gfx_v8_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) | - WRITE_DATA_DST_SEL(0) | - WR_CONFIRM)); - amdgpu_ring_write(ring, mmHDP_DEBUG0); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 1); - -} - static void gfx_v8_0_ring_emit_ib_gfx(struct amdgpu_ring *ring, struct amdgpu_ib *ib, unsigned vmid, bool ctx_switch) @@ -6350,28 +6340,7 @@ static void gfx_v8_0_ring_emit_vm_flush(struct amdgpu_ring *ring, { int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(usepfp) | - WRITE_DATA_DST_SEL(0)) | - WR_CONFIRM); - if (vmid < 8) { - amdgpu_ring_write(ring, - (mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid)); - } else { - amdgpu_ring_write(ring, - (mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8)); - } - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, pd_addr >> 12); - - /* bits 0-15 are the VM contexts0-15 */ - /* invalidate the cache */ - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) | - WRITE_DATA_DST_SEL(0))); - amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); /* wait for the invalidate to complete */ amdgpu_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5)); @@ -6635,8 +6604,22 @@ static void gfx_v8_0_ring_emit_rreg(struct amdgpu_ring *ring, uint32_t reg) static void gfx_v8_0_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val) { + uint32_t cmd; + + switch (ring->funcs->type) { + case AMDGPU_RING_TYPE_GFX: + cmd = WRITE_DATA_ENGINE_SEL(1) | WR_CONFIRM; + break; + case AMDGPU_RING_TYPE_KIQ: + cmd = 1 << 16; /* no inc addr */ + break; + default: + cmd = WR_CONFIRM; + break; + } + amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (1 << 16)); /* no inc addr */ + amdgpu_ring_write(ring, cmd); amdgpu_ring_write(ring, reg); amdgpu_ring_write(ring, 0); amdgpu_ring_write(ring, val); @@ -6855,7 +6838,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { .emit_frame_size = /* maximum 215dw if count 16 IBs in */ 5 + /* COND_EXEC */ 7 + /* PIPELINE_SYNC */ - 19 + /* VM_FLUSH */ + VI_FLUSH_GPU_TLB_NUM_WREG * 5 + 9 + /* VM_FLUSH */ 8 + /* FENCE for VM_FLUSH */ 20 + /* GDS switch */ 4 + /* double SWITCH_BUFFER, @@ -6877,7 +6860,6 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { .emit_vm_flush = gfx_v8_0_ring_emit_vm_flush, .emit_gds_switch = gfx_v8_0_ring_emit_gds_switch, .emit_hdp_flush = gfx_v8_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v8_0_ring_emit_hdp_invalidate, .test_ring = gfx_v8_0_ring_test_ring, .test_ib = gfx_v8_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, @@ -6886,6 +6868,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { .emit_cntxcntl = gfx_v8_ring_emit_cntxcntl, .init_cond_exec = gfx_v8_0_ring_emit_init_cond_exec, .patch_cond_exec = gfx_v8_0_ring_emit_patch_cond_exec, + .emit_wreg = gfx_v8_0_ring_emit_wreg, .soft_recovery = gfx_v8_0_ring_soft_recovery, }; @@ -6900,9 +6883,9 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_compute = { .emit_frame_size = 20 + /* gfx_v8_0_ring_emit_gds_switch */ 7 + /* gfx_v8_0_ring_emit_hdp_flush */ - 5 + /* gfx_v8_0_ring_emit_hdp_invalidate */ + 5 + /* hdp_invalidate */ 7 + /* gfx_v8_0_ring_emit_pipeline_sync */ - 17 + /* gfx_v8_0_ring_emit_vm_flush */ + VI_FLUSH_GPU_TLB_NUM_WREG * 5 + 7 + /* gfx_v8_0_ring_emit_vm_flush */ 7 + 7 + 7, /* gfx_v8_0_ring_emit_fence_compute x3 for user fence, vm fence */ .emit_ib_size = 4, /* gfx_v8_0_ring_emit_ib_compute */ .emit_ib = gfx_v8_0_ring_emit_ib_compute, @@ -6911,12 +6894,12 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_compute = { .emit_vm_flush = gfx_v8_0_ring_emit_vm_flush, .emit_gds_switch = gfx_v8_0_ring_emit_gds_switch, .emit_hdp_flush = gfx_v8_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v8_0_ring_emit_hdp_invalidate, .test_ring = gfx_v8_0_ring_test_ring, .test_ib = gfx_v8_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, .set_priority = gfx_v8_0_ring_set_priority_compute, + .emit_wreg = gfx_v8_0_ring_emit_wreg, }; static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_kiq = { @@ -6930,7 +6913,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_kiq = { .emit_frame_size = 20 + /* gfx_v8_0_ring_emit_gds_switch */ 7 + /* gfx_v8_0_ring_emit_hdp_flush */ - 5 + /* gfx_v8_0_ring_emit_hdp_invalidate */ + 5 + /* hdp_invalidate */ 7 + /* gfx_v8_0_ring_emit_pipeline_sync */ 17 + /* gfx_v8_0_ring_emit_vm_flush */ 7 + 7 + 7, /* gfx_v8_0_ring_emit_fence_kiq x3 for user fence, vm fence */ @@ -7128,12 +7111,12 @@ static void gfx_v8_0_ring_emit_ce_meta(struct amdgpu_ring *ring) } ce_payload = {}; if (ring->adev->virt.chained_ib_support) { - ce_payload_addr = AMDGPU_VA_RESERVED_SIZE - 2 * 4096 + - offsetof(struct vi_gfx_meta_data_chained_ib, ce_payload); + ce_payload_addr = amdgpu_csa_vaddr(ring->adev) + + offsetof(struct vi_gfx_meta_data_chained_ib, ce_payload); cnt_ce = (sizeof(ce_payload.chained) >> 2) + 4 - 2; } else { - ce_payload_addr = AMDGPU_VA_RESERVED_SIZE - 2 * 4096 + - offsetof(struct vi_gfx_meta_data, ce_payload); + ce_payload_addr = amdgpu_csa_vaddr(ring->adev) + + offsetof(struct vi_gfx_meta_data, ce_payload); cnt_ce = (sizeof(ce_payload.regular) >> 2) + 4 - 2; } @@ -7156,7 +7139,7 @@ static void gfx_v8_0_ring_emit_de_meta(struct amdgpu_ring *ring) struct vi_de_ib_state_chained_ib chained; } de_payload = {}; - csa_addr = AMDGPU_VA_RESERVED_SIZE - 2 * 4096; + csa_addr = amdgpu_csa_vaddr(ring->adev); gds_addr = csa_addr + 4096; if (ring->adev->virt.chained_ib_support) { de_payload.chained.gds_backup_addrlo = lower_32_bits(gds_addr);
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 6125e6c..154b9f6 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -57,6 +57,13 @@ MODULE_FIRMWARE("amdgpu/vega10_mec.bin"); MODULE_FIRMWARE("amdgpu/vega10_mec2.bin"); MODULE_FIRMWARE("amdgpu/vega10_rlc.bin"); +MODULE_FIRMWARE("amdgpu/vega12_ce.bin"); +MODULE_FIRMWARE("amdgpu/vega12_pfp.bin"); +MODULE_FIRMWARE("amdgpu/vega12_me.bin"); +MODULE_FIRMWARE("amdgpu/vega12_mec.bin"); +MODULE_FIRMWARE("amdgpu/vega12_mec2.bin"); +MODULE_FIRMWARE("amdgpu/vega12_rlc.bin"); + MODULE_FIRMWARE("amdgpu/raven_ce.bin"); MODULE_FIRMWARE("amdgpu/raven_pfp.bin"); MODULE_FIRMWARE("amdgpu/raven_me.bin"); @@ -144,7 +151,42 @@ static const struct soc15_reg_golden golden_settings_gc_9_x_common[] = SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x2544c382) }; +static const struct soc15_reg_golden golden_settings_gc_9_2_1[] = +{ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG2, 0xf00fffff, 0x00000420), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_GPU_ID, 0x0000000f, 0x00000000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_EVENT_CNTL_3, 0x00000003, 0x82400024), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE, 0x3fffffff, 0x00000001), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSH_MEM_CONFIG, 0x00001000, 0x00001000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_CU_0, 0x0007ffff, 0x00000800), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_CU_1, 0x0007ffff, 0x00000800), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_0, 0x01ffffff, 0x0000ff87), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_1, 0x01ffffff, 0x0000ff8f), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQC_CONFIG, 0x03000000, 0x020a2000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfffffeef, 0x010b0000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_HI, 0xffffffff, 0x4a2c0e68), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_LO, 0xffffffff, 0xb5d3f197), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_CACHE_INVALIDATION, 0x3fff3af3, 0x19200000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000003ff) +}; + +static const struct soc15_reg_golden golden_settings_gc_9_2_1_vg12[] = +{ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_DCC_CONFIG, 0x00000080, 0x04000080), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL, 0xfffdf3cf, 0x00014104), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_2, 0x0f000000, 0x0a000000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0xffff77ff, 0x24104041), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG_READ, 0xffff77ff, 0x24104041), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0xffffffff, 0x04040000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL_1, 0xffff03ff, 0x01000107), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_HI, 0xffffffff, 0x00000000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_LO, 0xffffffff, 0x76325410), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x01bd9f33, 0x01000000) +}; + #define VEGA10_GB_ADDR_CONFIG_GOLDEN 0x2a114042 +#define VEGA12_GB_ADDR_CONFIG_GOLDEN 0x24104041 #define RAVEN_GB_ADDR_CONFIG_GOLDEN 0x24000042 static void gfx_v9_0_set_ring_funcs(struct amdgpu_device *adev); @@ -168,6 +210,14 @@ static void gfx_v9_0_init_golden_registers(struct amdgpu_device *adev) golden_settings_gc_9_0_vg10, ARRAY_SIZE(golden_settings_gc_9_0_vg10)); break; + case CHIP_VEGA12: + soc15_program_register_sequence(adev, + golden_settings_gc_9_2_1, + ARRAY_SIZE(golden_settings_gc_9_2_1)); + soc15_program_register_sequence(adev, + golden_settings_gc_9_2_1_vg12, + ARRAY_SIZE(golden_settings_gc_9_2_1_vg12)); + break; case CHIP_RAVEN: soc15_program_register_sequence(adev, golden_settings_gc_9_1, @@ -271,58 +321,65 @@ static int gfx_v9_0_ring_test_ring(struct amdgpu_ring *ring) static int gfx_v9_0_ring_test_ib(struct amdgpu_ring *ring, long timeout) { - struct amdgpu_device *adev = ring->adev; - struct amdgpu_ib ib; - struct dma_fence *f = NULL; - uint32_t scratch; - uint32_t tmp = 0; - long r; + struct amdgpu_device *adev = ring->adev; + struct amdgpu_ib ib; + struct dma_fence *f = NULL; - r = amdgpu_gfx_scratch_get(adev, &scratch); - if (r) { - DRM_ERROR("amdgpu: failed to get scratch reg (%ld).\n", r); - return r; - } - WREG32(scratch, 0xCAFEDEAD); - memset(&ib, 0, sizeof(ib)); - r = amdgpu_ib_get(adev, NULL, 256, &ib); - if (r) { - DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r); - goto err1; - } - ib.ptr[0] = PACKET3(PACKET3_SET_UCONFIG_REG, 1); - ib.ptr[1] = ((scratch - PACKET3_SET_UCONFIG_REG_START)); - ib.ptr[2] = 0xDEADBEEF; - ib.length_dw = 3; + unsigned index; + uint64_t gpu_addr; + uint32_t tmp; + long r; - r = amdgpu_ib_schedule(ring, 1, &ib, NULL, &f); - if (r) - goto err2; + r = amdgpu_device_wb_get(adev, &index); + if (r) { + dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r); + return r; + } - r = dma_fence_wait_timeout(f, false, timeout); - if (r == 0) { - DRM_ERROR("amdgpu: IB test timed out.\n"); - r = -ETIMEDOUT; - goto err2; - } else if (r < 0) { - DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); - goto err2; - } - tmp = RREG32(scratch); - if (tmp == 0xDEADBEEF) { - DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx); - r = 0; - } else { - DRM_ERROR("amdgpu: ib test failed (scratch(0x%04X)=0x%08X)\n", - scratch, tmp); - r = -EINVAL; - } + gpu_addr = adev->wb.gpu_addr + (index * 4); + adev->wb.wb[index] = cpu_to_le32(0xCAFEDEAD); + memset(&ib, 0, sizeof(ib)); + r = amdgpu_ib_get(adev, NULL, 16, &ib); + if (r) { + DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r); + goto err1; + } + ib.ptr[0] = PACKET3(PACKET3_WRITE_DATA, 3); + ib.ptr[1] = WRITE_DATA_DST_SEL(5) | WR_CONFIRM; + ib.ptr[2] = lower_32_bits(gpu_addr); + ib.ptr[3] = upper_32_bits(gpu_addr); + ib.ptr[4] = 0xDEADBEEF; + ib.length_dw = 5; + + r = amdgpu_ib_schedule(ring, 1, &ib, NULL, &f); + if (r) + goto err2; + + r = dma_fence_wait_timeout(f, false, timeout); + if (r == 0) { + DRM_ERROR("amdgpu: IB test timed out.\n"); + r = -ETIMEDOUT; + goto err2; + } else if (r < 0) { + DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); + goto err2; + } + + tmp = adev->wb.wb[index]; + if (tmp == 0xDEADBEEF) { + DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx); + r = 0; + } else { + DRM_ERROR("ib test on ring %d failed\n", ring->idx); + r = -EINVAL; + } + err2: - amdgpu_ib_free(adev, &ib, NULL); - dma_fence_put(f); + amdgpu_ib_free(adev, &ib, NULL); + dma_fence_put(f); err1: - amdgpu_gfx_scratch_free(adev, scratch); - return r; + amdgpu_device_wb_free(adev, index); + return r; } @@ -362,6 +419,9 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) case CHIP_VEGA10: chip_name = "vega10"; break; + case CHIP_VEGA12: + chip_name = "vega12"; + break; case CHIP_RAVEN: chip_name = "raven"; break; @@ -938,12 +998,19 @@ static void gfx_v9_0_read_wave_vgprs(struct amdgpu_device *adev, uint32_t simd, start + SQIND_WAVE_VGPRS_OFFSET, size, dst); } +static void gfx_v9_0_select_me_pipe_q(struct amdgpu_device *adev, + u32 me, u32 pipe, u32 q) +{ + soc15_grbm_select(adev, me, pipe, q, 0); +} + static const struct amdgpu_gfx_funcs gfx_v9_0_gfx_funcs = { .get_gpu_clock_counter = &gfx_v9_0_get_gpu_clock_counter, .select_se_sh = &gfx_v9_0_select_se_sh, .read_wave_data = &gfx_v9_0_read_wave_data, .read_wave_sgprs = &gfx_v9_0_read_wave_sgprs, .read_wave_vgprs = &gfx_v9_0_read_wave_vgprs, + .select_me_pipe_q = &gfx_v9_0_select_me_pipe_q }; static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev) @@ -961,6 +1028,15 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev) adev->gfx.config.sc_earlyz_tile_fifo_size = 0x4C0; gb_addr_config = VEGA10_GB_ADDR_CONFIG_GOLDEN; break; + case CHIP_VEGA12: + adev->gfx.config.max_hw_contexts = 8; + adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; + adev->gfx.config.sc_prim_fifo_size_backend = 0x100; + adev->gfx.config.sc_hiz_tile_fifo_size = 0x30; + adev->gfx.config.sc_earlyz_tile_fifo_size = 0x4C0; + gb_addr_config = VEGA12_GB_ADDR_CONFIG_GOLDEN; + DRM_INFO("fix gfx.config for vega12\n"); + break; case CHIP_RAVEN: adev->gfx.config.max_hw_contexts = 8; adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; @@ -1242,6 +1318,7 @@ static int gfx_v9_0_sw_init(void *handle) switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: case CHIP_RAVEN: adev->gfx.mec.num_mec = 2; break; @@ -1254,18 +1331,18 @@ static int gfx_v9_0_sw_init(void *handle) adev->gfx.mec.num_queue_per_pipe = 8; /* EOP Event */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_GRBM_CP, 181, &adev->gfx.eop_irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 181, &adev->gfx.eop_irq); if (r) return r; /* Privileged reg */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_GRBM_CP, 184, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 184, &adev->gfx.priv_reg_irq); if (r) return r; /* Privileged inst */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_GRBM_CP, 185, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 185, &adev->gfx.priv_inst_irq); if (r) return r; @@ -1534,7 +1611,7 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev) tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE, SH_MEM_ALIGNMENT_MODE_UNALIGNED); WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp); - tmp = adev->mc.shared_aperture_start >> 48; + tmp = adev->gmc.shared_aperture_start >> 48; WREG32_SOC15(GC, 0, mmSH_MEM_BASES, tmp); } } @@ -2683,6 +2760,45 @@ static int gfx_v9_0_kiq_init_register(struct amdgpu_ring *ring) return 0; } +static int gfx_v9_0_kiq_fini_register(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + int j; + + /* disable the queue if it's active */ + if (RREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE) & 1) { + + WREG32_SOC15(GC, 0, mmCP_HQD_DEQUEUE_REQUEST, 1); + + for (j = 0; j < adev->usec_timeout; j++) { + if (!(RREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE) & 1)) + break; + udelay(1); + } + + if (j == AMDGPU_MAX_USEC_TIMEOUT) { + DRM_DEBUG("KIQ dequeue request failed.\n"); + + /* Manual disable if dequeue request times out */ + WREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE, 0); + } + + WREG32_SOC15(GC, 0, mmCP_HQD_DEQUEUE_REQUEST, + 0); + } + + WREG32_SOC15(GC, 0, mmCP_HQD_IQ_TIMER, 0); + WREG32_SOC15(GC, 0, mmCP_HQD_IB_CONTROL, 0); + WREG32_SOC15(GC, 0, mmCP_HQD_PERSISTENT_STATE, 0); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL, 0x40000000); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL, 0); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_RPTR, 0); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_HI, 0); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_LO, 0); + + return 0; +} + static int gfx_v9_0_kiq_init_queue(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; @@ -2936,7 +3052,6 @@ static int gfx_v9_0_kcq_disable(struct amdgpu_ring *kiq_ring,struct amdgpu_ring return r; } - static int gfx_v9_0_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -2959,6 +3074,20 @@ static int gfx_v9_0_hw_fini(void *handle) WREG32_FIELD15(GC, 0, CP_PQ_WPTR_POLL_CNTL, EN, 0); return 0; } + + /* Use deinitialize sequence from CAIL when unbinding device from driver, + * otherwise KIQ is hanging when binding back + */ + if (!adev->in_gpu_reset && !adev->gfx.in_suspend) { + mutex_lock(&adev->srbm_mutex); + soc15_grbm_select(adev, adev->gfx.kiq.ring.me, + adev->gfx.kiq.ring.pipe, + adev->gfx.kiq.ring.queue, 0); + gfx_v9_0_kiq_fini_register(&adev->gfx.kiq.ring); + soc15_grbm_select(adev, 0, 0, 0, 0); + mutex_unlock(&adev->srbm_mutex); + } + gfx_v9_0_cp_enable(adev, false); gfx_v9_0_rlc_stop(adev); @@ -3471,6 +3600,7 @@ static int gfx_v9_0_set_clockgating_state(void *handle, switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: case CHIP_RAVEN: gfx_v9_0_update_gfx_clock_gating(adev, state == AMD_CG_STATE_GATE ? true : false); @@ -3587,14 +3717,6 @@ static void gfx_v9_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) ref_and_mask, ref_and_mask, 0x20); } -static void gfx_v9_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - struct amdgpu_device *adev = ring->adev; - - gfx_v9_0_write_data_to_reg(ring, 0, true, - SOC15_REG_OFFSET(HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 1); -} - static void gfx_v9_0_ring_emit_ib_gfx(struct amdgpu_ring *ring, struct amdgpu_ib *ib, unsigned vmid, bool ctx_switch) @@ -3688,32 +3810,10 @@ static void gfx_v9_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void gfx_v9_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; - int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); - uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vmid); - uint64_t flags = AMDGPU_PTE_VALID; - unsigned eng = ring->vm_inv_eng; - - amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags); - pd_addr |= flags; - - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->ctx0_ptb_addr_lo32 + (2 * vmid), - lower_32_bits(pd_addr)); - - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->ctx0_ptb_addr_hi32 + (2 * vmid), - upper_32_bits(pd_addr)); - - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->vm_inv_eng0_req + eng, req); - - /* wait for the invalidate to complete */ - gfx_v9_0_wait_reg_mem(ring, 0, 0, 0, hub->vm_inv_eng0_ack + - eng, 0, 1 << vmid, 1 << vmid, 0x20); + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); /* compute doesn't have PFP */ - if (usepfp) { + if (ring->funcs->type == AMDGPU_RING_TYPE_GFX) { /* sync PFP to ME, otherwise we might get invalid PFP reads */ amdgpu_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0)); amdgpu_ring_write(ring, 0x0); @@ -3737,6 +3837,105 @@ static u64 gfx_v9_0_ring_get_wptr_compute(struct amdgpu_ring *ring) return wptr; } +static void gfx_v9_0_ring_set_pipe_percent(struct amdgpu_ring *ring, + bool acquire) +{ + struct amdgpu_device *adev = ring->adev; + int pipe_num, tmp, reg; + int pipe_percent = acquire ? SPI_WCL_PIPE_PERCENT_GFX__VALUE_MASK : 0x1; + + pipe_num = ring->me * adev->gfx.mec.num_pipe_per_mec + ring->pipe; + + /* first me only has 2 entries, GFX and HP3D */ + if (ring->me > 0) + pipe_num -= 2; + + reg = SOC15_REG_OFFSET(GC, 0, mmSPI_WCL_PIPE_PERCENT_GFX) + pipe_num; + tmp = RREG32(reg); + tmp = REG_SET_FIELD(tmp, SPI_WCL_PIPE_PERCENT_GFX, VALUE, pipe_percent); + WREG32(reg, tmp); +} + +static void gfx_v9_0_pipe_reserve_resources(struct amdgpu_device *adev, + struct amdgpu_ring *ring, + bool acquire) +{ + int i, pipe; + bool reserve; + struct amdgpu_ring *iring; + + mutex_lock(&adev->gfx.pipe_reserve_mutex); + pipe = amdgpu_gfx_queue_to_bit(adev, ring->me, ring->pipe, 0); + if (acquire) + set_bit(pipe, adev->gfx.pipe_reserve_bitmap); + else + clear_bit(pipe, adev->gfx.pipe_reserve_bitmap); + + if (!bitmap_weight(adev->gfx.pipe_reserve_bitmap, AMDGPU_MAX_COMPUTE_QUEUES)) { + /* Clear all reservations - everyone reacquires all resources */ + for (i = 0; i < adev->gfx.num_gfx_rings; ++i) + gfx_v9_0_ring_set_pipe_percent(&adev->gfx.gfx_ring[i], + true); + + for (i = 0; i < adev->gfx.num_compute_rings; ++i) + gfx_v9_0_ring_set_pipe_percent(&adev->gfx.compute_ring[i], + true); + } else { + /* Lower all pipes without a current reservation */ + for (i = 0; i < adev->gfx.num_gfx_rings; ++i) { + iring = &adev->gfx.gfx_ring[i]; + pipe = amdgpu_gfx_queue_to_bit(adev, + iring->me, + iring->pipe, + 0); + reserve = test_bit(pipe, adev->gfx.pipe_reserve_bitmap); + gfx_v9_0_ring_set_pipe_percent(iring, reserve); + } + + for (i = 0; i < adev->gfx.num_compute_rings; ++i) { + iring = &adev->gfx.compute_ring[i]; + pipe = amdgpu_gfx_queue_to_bit(adev, + iring->me, + iring->pipe, + 0); + reserve = test_bit(pipe, adev->gfx.pipe_reserve_bitmap); + gfx_v9_0_ring_set_pipe_percent(iring, reserve); + } + } + + mutex_unlock(&adev->gfx.pipe_reserve_mutex); +} + +static void gfx_v9_0_hqd_set_priority(struct amdgpu_device *adev, + struct amdgpu_ring *ring, + bool acquire) +{ + uint32_t pipe_priority = acquire ? 0x2 : 0x0; + uint32_t queue_priority = acquire ? 0xf : 0x0; + + mutex_lock(&adev->srbm_mutex); + soc15_grbm_select(adev, ring->me, ring->pipe, ring->queue, 0); + + WREG32_SOC15(GC, 0, mmCP_HQD_PIPE_PRIORITY, pipe_priority); + WREG32_SOC15(GC, 0, mmCP_HQD_QUEUE_PRIORITY, queue_priority); + + soc15_grbm_select(adev, 0, 0, 0, 0); + mutex_unlock(&adev->srbm_mutex); +} + +static void gfx_v9_0_ring_set_priority_compute(struct amdgpu_ring *ring, + enum drm_sched_priority priority) +{ + struct amdgpu_device *adev = ring->adev; + bool acquire = priority == DRM_SCHED_PRIORITY_HIGH_HW; + + if (ring->funcs->type != AMDGPU_RING_TYPE_COMPUTE) + return; + + gfx_v9_0_hqd_set_priority(adev, ring, acquire); + gfx_v9_0_pipe_reserve_resources(adev, ring, acquire); +} + static void gfx_v9_0_ring_set_wptr_compute(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; @@ -3790,7 +3989,7 @@ static void gfx_v9_0_ring_emit_ce_meta(struct amdgpu_ring *ring) int cnt; cnt = (sizeof(ce_payload) >> 2) + 4 - 2; - csa_addr = AMDGPU_VA_RESERVED_SIZE - 2 * 4096; + csa_addr = amdgpu_csa_vaddr(ring->adev); amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, cnt)); amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(2) | @@ -3808,7 +4007,7 @@ static void gfx_v9_0_ring_emit_de_meta(struct amdgpu_ring *ring) uint64_t csa_addr, gds_addr; int cnt; - csa_addr = AMDGPU_VA_RESERVED_SIZE - 2 * 4096; + csa_addr = amdgpu_csa_vaddr(ring->adev); gds_addr = csa_addr + 4096; de_payload.gds_backup_addrlo = lower_32_bits(gds_addr); de_payload.gds_backup_addrhi = upper_32_bits(gds_addr); @@ -3906,15 +4105,34 @@ static void gfx_v9_0_ring_emit_rreg(struct amdgpu_ring *ring, uint32_t reg) } static void gfx_v9_0_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, - uint32_t val) + uint32_t val) { + uint32_t cmd = 0; + + switch (ring->funcs->type) { + case AMDGPU_RING_TYPE_GFX: + cmd = WRITE_DATA_ENGINE_SEL(1) | WR_CONFIRM; + break; + case AMDGPU_RING_TYPE_KIQ: + cmd = (1 << 16); /* no inc addr */ + break; + default: + cmd = WR_CONFIRM; + break; + } amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, (1 << 16)); /* no inc addr */ + amdgpu_ring_write(ring, cmd); amdgpu_ring_write(ring, reg); amdgpu_ring_write(ring, 0); amdgpu_ring_write(ring, val); } +static void gfx_v9_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg, + uint32_t val, uint32_t mask) +{ + gfx_v9_0_wait_reg_mem(ring, 0, 0, 0, reg, 0, val, mask, 0x20); +} + static void gfx_v9_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev, enum amdgpu_interrupt_state state) { @@ -4139,7 +4357,9 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = { .emit_frame_size = /* totally 242 maximum if 16 IBs */ 5 + /* COND_EXEC */ 7 + /* PIPELINE_SYNC */ - 24 + /* VM_FLUSH */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + + 2 + /* VM_FLUSH */ 8 + /* FENCE for VM_FLUSH */ 20 + /* GDS switch */ 4 + /* double SWITCH_BUFFER, @@ -4161,7 +4381,6 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = { .emit_vm_flush = gfx_v9_0_ring_emit_vm_flush, .emit_gds_switch = gfx_v9_0_ring_emit_gds_switch, .emit_hdp_flush = gfx_v9_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v9_0_ring_emit_hdp_invalidate, .test_ring = gfx_v9_0_ring_test_ring, .test_ib = gfx_v9_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, @@ -4171,6 +4390,8 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = { .init_cond_exec = gfx_v9_0_ring_emit_init_cond_exec, .patch_cond_exec = gfx_v9_0_ring_emit_patch_cond_exec, .emit_tmz = gfx_v9_0_ring_emit_tmz, + .emit_wreg = gfx_v9_0_ring_emit_wreg, + .emit_reg_wait = gfx_v9_0_ring_emit_reg_wait, }; static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = { @@ -4185,9 +4406,11 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = { .emit_frame_size = 20 + /* gfx_v9_0_ring_emit_gds_switch */ 7 + /* gfx_v9_0_ring_emit_hdp_flush */ - 5 + /* gfx_v9_0_ring_emit_hdp_invalidate */ + 5 + /* hdp invalidate */ 7 + /* gfx_v9_0_ring_emit_pipeline_sync */ - 24 + /* gfx_v9_0_ring_emit_vm_flush */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + + 2 + /* gfx_v9_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v9_0_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */ .emit_ib = gfx_v9_0_ring_emit_ib_compute, @@ -4196,11 +4419,13 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = { .emit_vm_flush = gfx_v9_0_ring_emit_vm_flush, .emit_gds_switch = gfx_v9_0_ring_emit_gds_switch, .emit_hdp_flush = gfx_v9_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v9_0_ring_emit_hdp_invalidate, .test_ring = gfx_v9_0_ring_test_ring, .test_ib = gfx_v9_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, + .set_priority = gfx_v9_0_ring_set_priority_compute, + .emit_wreg = gfx_v9_0_ring_emit_wreg, + .emit_reg_wait = gfx_v9_0_ring_emit_reg_wait, }; static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = { @@ -4215,9 +4440,11 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = { .emit_frame_size = 20 + /* gfx_v9_0_ring_emit_gds_switch */ 7 + /* gfx_v9_0_ring_emit_hdp_flush */ - 5 + /* gfx_v9_0_ring_emit_hdp_invalidate */ + 5 + /* hdp invalidate */ 7 + /* gfx_v9_0_ring_emit_pipeline_sync */ - 24 + /* gfx_v9_0_ring_emit_vm_flush */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + + 2 + /* gfx_v9_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v9_0_ring_emit_fence_kiq x3 for user fence, vm fence */ .emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */ .emit_ib = gfx_v9_0_ring_emit_ib_compute, @@ -4228,6 +4455,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = { .pad_ib = amdgpu_ring_generic_pad_ib, .emit_rreg = gfx_v9_0_ring_emit_rreg, .emit_wreg = gfx_v9_0_ring_emit_wreg, + .emit_reg_wait = gfx_v9_0_ring_emit_reg_wait, }; static void gfx_v9_0_set_ring_funcs(struct amdgpu_device *adev) @@ -4274,6 +4502,7 @@ static void gfx_v9_0_set_rlc_funcs(struct amdgpu_device *adev) { switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: case CHIP_RAVEN: adev->gfx.rlc.funcs = &gfx_v9_0_rlc_funcs; break;
diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c index 56f5fe4..acfbd2d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c
@@ -40,7 +40,7 @@ static void gfxhub_v1_0_init_gart_pt_regs(struct amdgpu_device *adev) uint64_t value; BUG_ON(adev->gart.table_addr & (~0x0000FFFFFFFFF000ULL)); - value = adev->gart.table_addr - adev->mc.vram_start + value = adev->gart.table_addr - adev->gmc.vram_start + adev->vm_manager.vram_base_offset; value &= 0x0000FFFFFFFFF000ULL; value |= 0x1; /*valid bit*/ @@ -57,14 +57,14 @@ static void gfxhub_v1_0_init_gart_aperture_regs(struct amdgpu_device *adev) gfxhub_v1_0_init_gart_pt_regs(adev); WREG32_SOC15(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32, - (u32)(adev->mc.gart_start >> 12)); + (u32)(adev->gmc.gart_start >> 12)); WREG32_SOC15(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32, - (u32)(adev->mc.gart_start >> 44)); + (u32)(adev->gmc.gart_start >> 44)); WREG32_SOC15(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32, - (u32)(adev->mc.gart_end >> 12)); + (u32)(adev->gmc.gart_end >> 12)); WREG32_SOC15(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32, - (u32)(adev->mc.gart_end >> 44)); + (u32)(adev->gmc.gart_end >> 44)); } static void gfxhub_v1_0_init_system_aperture_regs(struct amdgpu_device *adev) @@ -78,12 +78,12 @@ static void gfxhub_v1_0_init_system_aperture_regs(struct amdgpu_device *adev) /* Program the system aperture low logical page number. */ WREG32_SOC15(GC, 0, mmMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->mc.vram_start >> 18); + adev->gmc.vram_start >> 18); WREG32_SOC15(GC, 0, mmMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->mc.vram_end >> 18); + adev->gmc.vram_end >> 18); /* Set default page address. */ - value = adev->vram_scratch.gpu_addr - adev->mc.vram_start + value = adev->vram_scratch.gpu_addr - adev->gmc.vram_start + adev->vm_manager.vram_base_offset; WREG32_SOC15(GC, 0, mmMC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB, (u32)(value >> 12)); @@ -92,9 +92,9 @@ static void gfxhub_v1_0_init_system_aperture_regs(struct amdgpu_device *adev) /* Program "protection fault". */ WREG32_SOC15(GC, 0, mmVM_L2_PROTECTION_FAULT_DEFAULT_ADDR_LO32, - (u32)(adev->dummy_page.addr >> 12)); + (u32)(adev->dummy_page_addr >> 12)); WREG32_SOC15(GC, 0, mmVM_L2_PROTECTION_FAULT_DEFAULT_ADDR_HI32, - (u32)((u64)adev->dummy_page.addr >> 44)); + (u32)((u64)adev->dummy_page_addr >> 44)); WREG32_FIELD15(GC, 0, VM_L2_PROTECTION_FAULT_CNTL2, ACTIVE_PAGE_MIGRATION_PTE_READ_RETRY, 1); @@ -143,7 +143,7 @@ static void gfxhub_v1_0_init_cache_regs(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, mmVM_L2_CNTL2, tmp); tmp = mmVM_L2_CNTL3_DEFAULT; - if (adev->mc.translate_further) { + if (adev->gmc.translate_further) { tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 12); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, L2_CACHE_BIGK_FRAGMENT_SIZE, 9); @@ -195,7 +195,7 @@ static void gfxhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) num_level = adev->vm_manager.num_level; block_size = adev->vm_manager.block_size; - if (adev->mc.translate_further) + if (adev->gmc.translate_further) num_level -= 1; else block_size -= 9; @@ -257,9 +257,9 @@ int gfxhub_v1_0_gart_enable(struct amdgpu_device *adev) * SRIOV driver need to program them */ WREG32_SOC15(GC, 0, mmMC_VM_FB_LOCATION_BASE, - adev->mc.vram_start >> 24); + adev->gmc.vram_start >> 24); WREG32_SOC15(GC, 0, mmMC_VM_FB_LOCATION_TOP, - adev->mc.vram_end >> 24); + adev->gmc.vram_end >> 24); } /* GART Enable. */
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index 749c902..79f9ac29 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -22,6 +22,7 @@ */ #include <linux/firmware.h> #include <drm/drmP.h> +#include <drm/drm_cache.h> #include "amdgpu.h" #include "gmc_v6_0.h" #include "amdgpu_ucode.h" @@ -36,7 +37,7 @@ #include "dce/dce_6_0_sh_mask.h" #include "si_enums.h" -static void gmc_v6_0_set_gart_funcs(struct amdgpu_device *adev); +static void gmc_v6_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev); static int gmc_v6_0_wait_for_idle(void *handle); @@ -136,19 +137,19 @@ static int gmc_v6_0_init_microcode(struct amdgpu_device *adev) snprintf(fw_name, sizeof(fw_name), "radeon/si58_mc.bin"); else snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name); - err = request_firmware(&adev->mc.fw, fw_name, adev->dev); + err = request_firmware(&adev->gmc.fw, fw_name, adev->dev); if (err) goto out; - err = amdgpu_ucode_validate(adev->mc.fw); + err = amdgpu_ucode_validate(adev->gmc.fw); out: if (err) { dev_err(adev->dev, "si_mc: Failed to load firmware \"%s\"\n", fw_name); - release_firmware(adev->mc.fw); - adev->mc.fw = NULL; + release_firmware(adev->gmc.fw); + adev->gmc.fw = NULL; } return err; } @@ -161,20 +162,20 @@ static int gmc_v6_0_mc_load_microcode(struct amdgpu_device *adev) int i, regs_size, ucode_size; const struct mc_firmware_header_v1_0 *hdr; - if (!adev->mc.fw) + if (!adev->gmc.fw) return -EINVAL; - hdr = (const struct mc_firmware_header_v1_0 *)adev->mc.fw->data; + hdr = (const struct mc_firmware_header_v1_0 *)adev->gmc.fw->data; amdgpu_ucode_print_mc_hdr(&hdr->header); - adev->mc.fw_version = le32_to_cpu(hdr->header.ucode_version); + adev->gmc.fw_version = le32_to_cpu(hdr->header.ucode_version); regs_size = le32_to_cpu(hdr->io_debug_size_bytes) / (4 * 2); new_io_mc_regs = (const __le32 *) - (adev->mc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); + (adev->gmc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4; new_fw_data = (const __le32 *) - (adev->mc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + (adev->gmc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); running = RREG32(mmMC_SEQ_SUP_CNTL) & MC_SEQ_SUP_CNTL__RUN_MASK; @@ -217,12 +218,12 @@ static int gmc_v6_0_mc_load_microcode(struct amdgpu_device *adev) } static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev, - struct amdgpu_mc *mc) + struct amdgpu_gmc *mc) { u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; base <<= 24; - amdgpu_device_vram_location(adev, &adev->mc, base); + amdgpu_device_vram_location(adev, &adev->gmc, base); amdgpu_device_gart_location(adev, mc); } @@ -259,9 +260,9 @@ static void gmc_v6_0_mc_program(struct amdgpu_device *adev) } /* Update configuration */ WREG32(mmMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->mc.vram_start >> 12); + adev->gmc.vram_start >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->mc.vram_end >> 12); + adev->gmc.vram_end >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_DEFAULT_ADDR, adev->vram_scratch.gpu_addr >> 12); WREG32(mmMC_VM_AGP_BASE, 0); @@ -319,56 +320,69 @@ static int gmc_v6_0_mc_init(struct amdgpu_device *adev) numchan = 16; break; } - adev->mc.vram_width = numchan * chansize; + adev->gmc.vram_width = numchan * chansize; /* size in MB on si */ - adev->mc.mc_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; - adev->mc.real_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; + adev->gmc.mc_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; + adev->gmc.real_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; if (!(adev->flags & AMD_IS_APU)) { r = amdgpu_device_resize_fb_bar(adev); if (r) return r; } - adev->mc.aper_base = pci_resource_start(adev->pdev, 0); - adev->mc.aper_size = pci_resource_len(adev->pdev, 0); - adev->mc.visible_vram_size = adev->mc.aper_size; + adev->gmc.aper_base = pci_resource_start(adev->pdev, 0); + adev->gmc.aper_size = pci_resource_len(adev->pdev, 0); + adev->gmc.visible_vram_size = adev->gmc.aper_size; /* set the gart size */ if (amdgpu_gart_size == -1) { switch (adev->asic_type) { case CHIP_HAINAN: /* no MM engines */ default: - adev->mc.gart_size = 256ULL << 20; + adev->gmc.gart_size = 256ULL << 20; break; case CHIP_VERDE: /* UVD, VCE do not support GPUVM */ case CHIP_TAHITI: /* UVD, VCE do not support GPUVM */ case CHIP_PITCAIRN: /* UVD, VCE do not support GPUVM */ case CHIP_OLAND: /* UVD, VCE do not support GPUVM */ - adev->mc.gart_size = 1024ULL << 20; + adev->gmc.gart_size = 1024ULL << 20; break; } } else { - adev->mc.gart_size = (u64)amdgpu_gart_size << 20; + adev->gmc.gart_size = (u64)amdgpu_gart_size << 20; } - gmc_v6_0_vram_gtt_location(adev, &adev->mc); + gmc_v6_0_vram_gtt_location(adev, &adev->gmc); return 0; } -static void gmc_v6_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, - uint32_t vmid) +static void gmc_v6_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid) { - WREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0); - WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); } -static int gmc_v6_0_gart_set_pte_pde(struct amdgpu_device *adev, - void *cpu_pt_addr, - uint32_t gpu_page_idx, - uint64_t addr, - uint64_t flags) +static uint64_t gmc_v6_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, + unsigned vmid, uint64_t pd_addr) +{ + uint32_t reg; + + /* write new base address */ + if (vmid < 8) + reg = mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid; + else + reg = mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (vmid - 8); + amdgpu_ring_emit_wreg(ring, reg, pd_addr >> 12); + + /* bits 0-15 are the VM contexts0-15 */ + amdgpu_ring_emit_wreg(ring, mmVM_INVALIDATE_REQUEST, 1 << vmid); + + return pd_addr; +} + +static int gmc_v6_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, + uint32_t gpu_page_idx, uint64_t addr, + uint64_t flags) { void __iomem *ptr = (void *)cpu_pt_addr; uint64_t value; @@ -432,9 +446,9 @@ static void gmc_v6_0_set_prt(struct amdgpu_device *adev, bool enable) { u32 tmp; - if (enable && !adev->mc.prt_warning) { + if (enable && !adev->gmc.prt_warning) { dev_warn(adev->dev, "Disabling VM faults because of PRT request!\n"); - adev->mc.prt_warning = true; + adev->gmc.prt_warning = true; } tmp = RREG32(mmVM_PRT_CNTL); @@ -454,7 +468,8 @@ static void gmc_v6_0_set_prt(struct amdgpu_device *adev, bool enable) if (enable) { uint32_t low = AMDGPU_VA_RESERVED_SIZE >> AMDGPU_GPU_PAGE_SHIFT; - uint32_t high = adev->vm_manager.max_pfn; + uint32_t high = adev->vm_manager.max_pfn - + (AMDGPU_VA_RESERVED_SIZE >> AMDGPU_GPU_PAGE_SHIFT); WREG32(mmVM_PRT_APERTURE0_LOW_ADDR, low); WREG32(mmVM_PRT_APERTURE1_LOW_ADDR, low); @@ -514,11 +529,11 @@ static int gmc_v6_0_gart_enable(struct amdgpu_device *adev) (field << VM_L2_CNTL3__BANK_SELECT__SHIFT) | (field << VM_L2_CNTL3__L2_CACHE_BIGK_FRAGMENT_SIZE__SHIFT)); /* setup context0 */ - WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gart_start >> 12); - WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gart_end >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->gmc.gart_start >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->gmc.gart_end >> 12); WREG32(mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR, adev->gart.table_addr >> 12); WREG32(mmVM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, - (u32)(adev->dummy_page.addr >> 12)); + (u32)(adev->dummy_page_addr >> 12)); WREG32(mmVM_CONTEXT0_CNTL2, 0); WREG32(mmVM_CONTEXT0_CNTL, VM_CONTEXT0_CNTL__ENABLE_CONTEXT_MASK | @@ -548,7 +563,7 @@ static int gmc_v6_0_gart_enable(struct amdgpu_device *adev) /* enable context1-15 */ WREG32(mmVM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR, - (u32)(adev->dummy_page.addr >> 12)); + (u32)(adev->dummy_page_addr >> 12)); WREG32(mmVM_CONTEXT1_CNTL2, 4); WREG32(mmVM_CONTEXT1_CNTL, VM_CONTEXT1_CNTL__ENABLE_CONTEXT_MASK | @@ -560,9 +575,9 @@ static int gmc_v6_0_gart_enable(struct amdgpu_device *adev) else gmc_v6_0_set_fault_enable_default(adev, true); - gmc_v6_0_gart_flush_gpu_tlb(adev, 0); + gmc_v6_0_flush_gpu_tlb(adev, 0); dev_info(adev->dev, "PCIE GART of %uM enabled (table at 0x%016llX).\n", - (unsigned)(adev->mc.gart_size >> 20), + (unsigned)(adev->gmc.gart_size >> 20), (unsigned long long)adev->gart.table_addr); adev->gart.ready = true; return 0; @@ -794,7 +809,7 @@ static int gmc_v6_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - gmc_v6_0_set_gart_funcs(adev); + gmc_v6_0_set_gmc_funcs(adev); gmc_v6_0_set_irq_funcs(adev); return 0; @@ -807,7 +822,7 @@ static int gmc_v6_0_late_init(void *handle) amdgpu_bo_late_init(adev); if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS) - return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0); + return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); else return 0; } @@ -826,7 +841,7 @@ static unsigned gmc_v6_0_get_vbios_fb_size(struct amdgpu_device *adev) 4); } /* return 0 if the pre-OS buffer uses up most of vram */ - if ((adev->mc.real_vram_size - size) < (8 * 1024 * 1024)) + if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024)) return 0; return size; } @@ -838,24 +853,24 @@ static int gmc_v6_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (adev->flags & AMD_IS_APU) { - adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; + adev->gmc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; } else { u32 tmp = RREG32(mmMC_SEQ_MISC0); tmp &= MC_SEQ_MISC0__MT__MASK; - adev->mc.vram_type = gmc_v6_0_convert_vram_type(tmp); + adev->gmc.vram_type = gmc_v6_0_convert_vram_type(tmp); } - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->mc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->gmc.vm_fault); if (r) return r; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->mc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->gmc.vm_fault); if (r) return r; amdgpu_vm_adjust_size(adev, 64, 9, 1, 40); - adev->mc.mc_mask = 0xffffffffffULL; + adev->gmc.mc_mask = 0xffffffffffULL; adev->need_dma32 = false; dma_bits = adev->need_dma32 ? 32 : 40; @@ -870,6 +885,7 @@ static int gmc_v6_0_sw_init(void *handle) pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32)); dev_warn(adev->dev, "amdgpu: No coherent DMA available.\n"); } + adev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits); r = gmc_v6_0_init_microcode(adev); if (r) { @@ -881,7 +897,7 @@ static int gmc_v6_0_sw_init(void *handle) if (r) return r; - adev->mc.stolen_size = gmc_v6_0_get_vbios_fb_size(adev); + adev->gmc.stolen_size = gmc_v6_0_get_vbios_fb_size(adev); r = amdgpu_bo_init(adev); if (r) @@ -921,8 +937,8 @@ static int gmc_v6_0_sw_fini(void *handle) amdgpu_vm_manager_fini(adev); gmc_v6_0_gart_fini(adev); amdgpu_bo_fini(adev); - release_firmware(adev->mc.fw); - adev->mc.fw = NULL; + release_firmware(adev->gmc.fw); + adev->gmc.fw = NULL; return 0; } @@ -953,7 +969,7 @@ static int gmc_v6_0_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - amdgpu_irq_put(adev, &adev->mc.vm_fault, 0); + amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); gmc_v6_0_gart_disable(adev); return 0; @@ -1148,9 +1164,10 @@ static const struct amd_ip_funcs gmc_v6_0_ip_funcs = { .set_powergating_state = gmc_v6_0_set_powergating_state, }; -static const struct amdgpu_gart_funcs gmc_v6_0_gart_funcs = { - .flush_gpu_tlb = gmc_v6_0_gart_flush_gpu_tlb, - .set_pte_pde = gmc_v6_0_gart_set_pte_pde, +static const struct amdgpu_gmc_funcs gmc_v6_0_gmc_funcs = { + .flush_gpu_tlb = gmc_v6_0_flush_gpu_tlb, + .emit_flush_gpu_tlb = gmc_v6_0_emit_flush_gpu_tlb, + .set_pte_pde = gmc_v6_0_set_pte_pde, .set_prt = gmc_v6_0_set_prt, .get_vm_pde = gmc_v6_0_get_vm_pde, .get_vm_pte_flags = gmc_v6_0_get_vm_pte_flags @@ -1161,16 +1178,16 @@ static const struct amdgpu_irq_src_funcs gmc_v6_0_irq_funcs = { .process = gmc_v6_0_process_interrupt, }; -static void gmc_v6_0_set_gart_funcs(struct amdgpu_device *adev) +static void gmc_v6_0_set_gmc_funcs(struct amdgpu_device *adev) { - if (adev->gart.gart_funcs == NULL) - adev->gart.gart_funcs = &gmc_v6_0_gart_funcs; + if (adev->gmc.gmc_funcs == NULL) + adev->gmc.gmc_funcs = &gmc_v6_0_gmc_funcs; } static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev) { - adev->mc.vm_fault.num_types = 1; - adev->mc.vm_fault.funcs = &gmc_v6_0_irq_funcs; + adev->gmc.vm_fault.num_types = 1; + adev->gmc.vm_fault.funcs = &gmc_v6_0_irq_funcs; } const struct amdgpu_ip_block_version gmc_v6_0_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index c8b5da72..7147bfe 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -22,6 +22,7 @@ */ #include <linux/firmware.h> #include <drm/drmP.h> +#include <drm/drm_cache.h> #include "amdgpu.h" #include "cikd.h" #include "cik.h" @@ -42,7 +43,7 @@ #include "amdgpu_atombios.h" -static void gmc_v7_0_set_gart_funcs(struct amdgpu_device *adev); +static void gmc_v7_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v7_0_set_irq_funcs(struct amdgpu_device *adev); static int gmc_v7_0_wait_for_idle(void *handle); @@ -151,16 +152,16 @@ static int gmc_v7_0_init_microcode(struct amdgpu_device *adev) else snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name); - err = request_firmware(&adev->mc.fw, fw_name, adev->dev); + err = request_firmware(&adev->gmc.fw, fw_name, adev->dev); if (err) goto out; - err = amdgpu_ucode_validate(adev->mc.fw); + err = amdgpu_ucode_validate(adev->gmc.fw); out: if (err) { pr_err("cik_mc: Failed to load firmware \"%s\"\n", fw_name); - release_firmware(adev->mc.fw); - adev->mc.fw = NULL; + release_firmware(adev->gmc.fw); + adev->gmc.fw = NULL; } return err; } @@ -181,19 +182,19 @@ static int gmc_v7_0_mc_load_microcode(struct amdgpu_device *adev) u32 running; int i, ucode_size, regs_size; - if (!adev->mc.fw) + if (!adev->gmc.fw) return -EINVAL; - hdr = (const struct mc_firmware_header_v1_0 *)adev->mc.fw->data; + hdr = (const struct mc_firmware_header_v1_0 *)adev->gmc.fw->data; amdgpu_ucode_print_mc_hdr(&hdr->header); - adev->mc.fw_version = le32_to_cpu(hdr->header.ucode_version); + adev->gmc.fw_version = le32_to_cpu(hdr->header.ucode_version); regs_size = le32_to_cpu(hdr->io_debug_size_bytes) / (4 * 2); io_mc_regs = (const __le32 *) - (adev->mc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); + (adev->gmc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4; fw_data = (const __le32 *) - (adev->mc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + (adev->gmc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); running = REG_GET_FIELD(RREG32(mmMC_SEQ_SUP_CNTL), MC_SEQ_SUP_CNTL, RUN); @@ -235,12 +236,12 @@ static int gmc_v7_0_mc_load_microcode(struct amdgpu_device *adev) } static void gmc_v7_0_vram_gtt_location(struct amdgpu_device *adev, - struct amdgpu_mc *mc) + struct amdgpu_gmc *mc) { u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; base <<= 24; - amdgpu_device_vram_location(adev, &adev->mc, base); + amdgpu_device_vram_location(adev, &adev->gmc, base); amdgpu_device_gart_location(adev, mc); } @@ -283,9 +284,9 @@ static void gmc_v7_0_mc_program(struct amdgpu_device *adev) } /* Update configuration */ WREG32(mmMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->mc.vram_start >> 12); + adev->gmc.vram_start >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->mc.vram_end >> 12); + adev->gmc.vram_end >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_DEFAULT_ADDR, adev->vram_scratch.gpu_addr >> 12); WREG32(mmMC_VM_AGP_BASE, 0); @@ -318,8 +319,8 @@ static int gmc_v7_0_mc_init(struct amdgpu_device *adev) { int r; - adev->mc.vram_width = amdgpu_atombios_get_vram_width(adev); - if (!adev->mc.vram_width) { + adev->gmc.vram_width = amdgpu_atombios_get_vram_width(adev); + if (!adev->gmc.vram_width) { u32 tmp; int chansize, numchan; @@ -361,38 +362,38 @@ static int gmc_v7_0_mc_init(struct amdgpu_device *adev) numchan = 16; break; } - adev->mc.vram_width = numchan * chansize; + adev->gmc.vram_width = numchan * chansize; } /* size in MB on si */ - adev->mc.mc_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; - adev->mc.real_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; + adev->gmc.mc_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; + adev->gmc.real_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; if (!(adev->flags & AMD_IS_APU)) { r = amdgpu_device_resize_fb_bar(adev); if (r) return r; } - adev->mc.aper_base = pci_resource_start(adev->pdev, 0); - adev->mc.aper_size = pci_resource_len(adev->pdev, 0); + adev->gmc.aper_base = pci_resource_start(adev->pdev, 0); + adev->gmc.aper_size = pci_resource_len(adev->pdev, 0); #ifdef CONFIG_X86_64 if (adev->flags & AMD_IS_APU) { - adev->mc.aper_base = ((u64)RREG32(mmMC_VM_FB_OFFSET)) << 22; - adev->mc.aper_size = adev->mc.real_vram_size; + adev->gmc.aper_base = ((u64)RREG32(mmMC_VM_FB_OFFSET)) << 22; + adev->gmc.aper_size = adev->gmc.real_vram_size; } #endif /* In case the PCI BAR is larger than the actual amount of vram */ - adev->mc.visible_vram_size = adev->mc.aper_size; - if (adev->mc.visible_vram_size > adev->mc.real_vram_size) - adev->mc.visible_vram_size = adev->mc.real_vram_size; + adev->gmc.visible_vram_size = adev->gmc.aper_size; + if (adev->gmc.visible_vram_size > adev->gmc.real_vram_size) + adev->gmc.visible_vram_size = adev->gmc.real_vram_size; /* set the gart size */ if (amdgpu_gart_size == -1) { switch (adev->asic_type) { case CHIP_TOPAZ: /* no MM engines */ default: - adev->mc.gart_size = 256ULL << 20; + adev->gmc.gart_size = 256ULL << 20; break; #ifdef CONFIG_DRM_AMDGPU_CIK case CHIP_BONAIRE: /* UVD, VCE do not support GPUVM */ @@ -400,15 +401,15 @@ static int gmc_v7_0_mc_init(struct amdgpu_device *adev) case CHIP_KAVERI: /* UVD, VCE do not support GPUVM */ case CHIP_KABINI: /* UVD, VCE do not support GPUVM */ case CHIP_MULLINS: /* UVD, VCE do not support GPUVM */ - adev->mc.gart_size = 1024ULL << 20; + adev->gmc.gart_size = 1024ULL << 20; break; #endif } } else { - adev->mc.gart_size = (u64)amdgpu_gart_size << 20; + adev->gmc.gart_size = (u64)amdgpu_gart_size << 20; } - gmc_v7_0_vram_gtt_location(adev, &adev->mc); + gmc_v7_0_vram_gtt_location(adev, &adev->gmc); return 0; } @@ -421,25 +422,44 @@ static int gmc_v7_0_mc_init(struct amdgpu_device *adev) */ /** - * gmc_v7_0_gart_flush_gpu_tlb - gart tlb flush callback + * gmc_v7_0_flush_gpu_tlb - gart tlb flush callback * * @adev: amdgpu_device pointer * @vmid: vm instance to flush * * Flush the TLB for the requested page table (CIK). */ -static void gmc_v7_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, - uint32_t vmid) +static void gmc_v7_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid) { - /* flush hdp cache */ - WREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0); - /* bits 0-15 are the VM contexts0-15 */ WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); } +static uint64_t gmc_v7_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, + unsigned vmid, uint64_t pd_addr) +{ + uint32_t reg; + + if (vmid < 8) + reg = mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid; + else + reg = mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8; + amdgpu_ring_emit_wreg(ring, reg, pd_addr >> 12); + + /* bits 0-15 are the VM contexts0-15 */ + amdgpu_ring_emit_wreg(ring, mmVM_INVALIDATE_REQUEST, 1 << vmid); + + return pd_addr; +} + +static void gmc_v7_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid, + unsigned pasid) +{ + amdgpu_ring_emit_wreg(ring, mmIH_VMID_0_LUT + vmid, pasid); +} + /** - * gmc_v7_0_gart_set_pte_pde - update the page tables using MMIO + * gmc_v7_0_set_pte_pde - update the page tables using MMIO * * @adev: amdgpu_device pointer * @cpu_pt_addr: cpu address of the page table @@ -449,11 +469,9 @@ static void gmc_v7_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, * * Update the page tables using the CPU. */ -static int gmc_v7_0_gart_set_pte_pde(struct amdgpu_device *adev, - void *cpu_pt_addr, - uint32_t gpu_page_idx, - uint64_t addr, - uint64_t flags) +static int gmc_v7_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, + uint32_t gpu_page_idx, uint64_t addr, + uint64_t flags) { void __iomem *ptr = (void *)cpu_pt_addr; uint64_t value; @@ -523,9 +541,9 @@ static void gmc_v7_0_set_prt(struct amdgpu_device *adev, bool enable) { uint32_t tmp; - if (enable && !adev->mc.prt_warning) { + if (enable && !adev->gmc.prt_warning) { dev_warn(adev->dev, "Disabling VM faults because of PRT request!\n"); - adev->mc.prt_warning = true; + adev->gmc.prt_warning = true; } tmp = RREG32(mmVM_PRT_CNTL); @@ -547,7 +565,8 @@ static void gmc_v7_0_set_prt(struct amdgpu_device *adev, bool enable) if (enable) { uint32_t low = AMDGPU_VA_RESERVED_SIZE >> AMDGPU_GPU_PAGE_SHIFT; - uint32_t high = adev->vm_manager.max_pfn; + uint32_t high = adev->vm_manager.max_pfn - + (AMDGPU_VA_RESERVED_SIZE >> AMDGPU_GPU_PAGE_SHIFT); WREG32(mmVM_PRT_APERTURE0_LOW_ADDR, low); WREG32(mmVM_PRT_APERTURE1_LOW_ADDR, low); @@ -621,11 +640,11 @@ static int gmc_v7_0_gart_enable(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, L2_CACHE_BIGK_FRAGMENT_SIZE, field); WREG32(mmVM_L2_CNTL3, tmp); /* setup context0 */ - WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gart_start >> 12); - WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gart_end >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->gmc.gart_start >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->gmc.gart_end >> 12); WREG32(mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR, adev->gart.table_addr >> 12); WREG32(mmVM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, - (u32)(adev->dummy_page.addr >> 12)); + (u32)(adev->dummy_page_addr >> 12)); WREG32(mmVM_CONTEXT0_CNTL2, 0); tmp = RREG32(mmVM_CONTEXT0_CNTL); tmp = REG_SET_FIELD(tmp, VM_CONTEXT0_CNTL, ENABLE_CONTEXT, 1); @@ -655,7 +674,7 @@ static int gmc_v7_0_gart_enable(struct amdgpu_device *adev) /* enable context1-15 */ WREG32(mmVM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR, - (u32)(adev->dummy_page.addr >> 12)); + (u32)(adev->dummy_page_addr >> 12)); WREG32(mmVM_CONTEXT1_CNTL2, 4); tmp = RREG32(mmVM_CONTEXT1_CNTL); tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, ENABLE_CONTEXT, 1); @@ -674,9 +693,9 @@ static int gmc_v7_0_gart_enable(struct amdgpu_device *adev) WREG32(mmCHUB_CONTROL, tmp); } - gmc_v7_0_gart_flush_gpu_tlb(adev, 0); + gmc_v7_0_flush_gpu_tlb(adev, 0); DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n", - (unsigned)(adev->mc.gart_size >> 20), + (unsigned)(adev->gmc.gart_size >> 20), (unsigned long long)adev->gart.table_addr); adev->gart.ready = true; return 0; @@ -749,21 +768,21 @@ static void gmc_v7_0_gart_fini(struct amdgpu_device *adev) * * Print human readable fault information (CIK). */ -static void gmc_v7_0_vm_decode_fault(struct amdgpu_device *adev, - u32 status, u32 addr, u32 mc_client) +static void gmc_v7_0_vm_decode_fault(struct amdgpu_device *adev, u32 status, + u32 addr, u32 mc_client, unsigned pasid) { - u32 mc_id; u32 vmid = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, VMID); u32 protections = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, PROTECTIONS); char block[5] = { mc_client >> 24, (mc_client >> 16) & 0xff, (mc_client >> 8) & 0xff, mc_client & 0xff, 0 }; + u32 mc_id; mc_id = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, MEMORY_CLIENT_ID); - dev_err(adev->dev, "VM fault (0x%02x, vmid %d) at page %u, %s from '%s' (0x%08x) (%d)\n", - protections, vmid, addr, + dev_err(adev->dev, "VM fault (0x%02x, vmid %d, pasid %d) at page %u, %s from '%s' (0x%08x) (%d)\n", + protections, vmid, pasid, addr, REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, MEMORY_CLIENT_RW) ? "write" : "read", block, mc_client, mc_id); @@ -921,16 +940,16 @@ static int gmc_v7_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - gmc_v7_0_set_gart_funcs(adev); + gmc_v7_0_set_gmc_funcs(adev); gmc_v7_0_set_irq_funcs(adev); - adev->mc.shared_aperture_start = 0x2000000000000000ULL; - adev->mc.shared_aperture_end = - adev->mc.shared_aperture_start + (4ULL << 30) - 1; - adev->mc.private_aperture_start = - adev->mc.shared_aperture_end + 1; - adev->mc.private_aperture_end = - adev->mc.private_aperture_start + (4ULL << 30) - 1; + adev->gmc.shared_aperture_start = 0x2000000000000000ULL; + adev->gmc.shared_aperture_end = + adev->gmc.shared_aperture_start + (4ULL << 30) - 1; + adev->gmc.private_aperture_start = + adev->gmc.shared_aperture_end + 1; + adev->gmc.private_aperture_end = + adev->gmc.private_aperture_start + (4ULL << 30) - 1; return 0; } @@ -942,7 +961,7 @@ static int gmc_v7_0_late_init(void *handle) amdgpu_bo_late_init(adev); if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS) - return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0); + return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); else return 0; } @@ -961,7 +980,7 @@ static unsigned gmc_v7_0_get_vbios_fb_size(struct amdgpu_device *adev) 4); } /* return 0 if the pre-OS buffer uses up most of vram */ - if ((adev->mc.real_vram_size - size) < (8 * 1024 * 1024)) + if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024)) return 0; return size; } @@ -973,18 +992,18 @@ static int gmc_v7_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (adev->flags & AMD_IS_APU) { - adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; + adev->gmc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; } else { u32 tmp = RREG32(mmMC_SEQ_MISC0); tmp &= MC_SEQ_MISC0__MT__MASK; - adev->mc.vram_type = gmc_v7_0_convert_vram_type(tmp); + adev->gmc.vram_type = gmc_v7_0_convert_vram_type(tmp); } - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->mc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->gmc.vm_fault); if (r) return r; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->mc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->gmc.vm_fault); if (r) return r; @@ -998,7 +1017,7 @@ static int gmc_v7_0_sw_init(void *handle) * This is the max address of the GPU's * internal address space. */ - adev->mc.mc_mask = 0xffffffffffULL; /* 40 bit MC */ + adev->gmc.mc_mask = 0xffffffffffULL; /* 40 bit MC */ /* set DMA mask + need_dma32 flags. * PCIE - can handle 40-bits. @@ -1018,6 +1037,7 @@ static int gmc_v7_0_sw_init(void *handle) pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32)); pr_warn("amdgpu: No coherent DMA available\n"); } + adev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits); r = gmc_v7_0_init_microcode(adev); if (r) { @@ -1029,7 +1049,7 @@ static int gmc_v7_0_sw_init(void *handle) if (r) return r; - adev->mc.stolen_size = gmc_v7_0_get_vbios_fb_size(adev); + adev->gmc.stolen_size = gmc_v7_0_get_vbios_fb_size(adev); /* Memory manager */ r = amdgpu_bo_init(adev); @@ -1070,8 +1090,8 @@ static int gmc_v7_0_sw_fini(void *handle) amdgpu_vm_manager_fini(adev); gmc_v7_0_gart_fini(adev); amdgpu_bo_fini(adev); - release_firmware(adev->mc.fw); - adev->mc.fw = NULL; + release_firmware(adev->gmc.fw); + adev->gmc.fw = NULL; return 0; } @@ -1104,7 +1124,7 @@ static int gmc_v7_0_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - amdgpu_irq_put(adev, &adev->mc.vm_fault, 0); + amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); gmc_v7_0_gart_disable(adev); return 0; @@ -1278,7 +1298,8 @@ static int gmc_v7_0_process_interrupt(struct amdgpu_device *adev, addr); dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n", status); - gmc_v7_0_vm_decode_fault(adev, status, addr, mc_client); + gmc_v7_0_vm_decode_fault(adev, status, addr, mc_client, + entry->pasid); } return 0; @@ -1327,9 +1348,11 @@ static const struct amd_ip_funcs gmc_v7_0_ip_funcs = { .set_powergating_state = gmc_v7_0_set_powergating_state, }; -static const struct amdgpu_gart_funcs gmc_v7_0_gart_funcs = { - .flush_gpu_tlb = gmc_v7_0_gart_flush_gpu_tlb, - .set_pte_pde = gmc_v7_0_gart_set_pte_pde, +static const struct amdgpu_gmc_funcs gmc_v7_0_gmc_funcs = { + .flush_gpu_tlb = gmc_v7_0_flush_gpu_tlb, + .emit_flush_gpu_tlb = gmc_v7_0_emit_flush_gpu_tlb, + .emit_pasid_mapping = gmc_v7_0_emit_pasid_mapping, + .set_pte_pde = gmc_v7_0_set_pte_pde, .set_prt = gmc_v7_0_set_prt, .get_vm_pte_flags = gmc_v7_0_get_vm_pte_flags, .get_vm_pde = gmc_v7_0_get_vm_pde @@ -1340,16 +1363,16 @@ static const struct amdgpu_irq_src_funcs gmc_v7_0_irq_funcs = { .process = gmc_v7_0_process_interrupt, }; -static void gmc_v7_0_set_gart_funcs(struct amdgpu_device *adev) +static void gmc_v7_0_set_gmc_funcs(struct amdgpu_device *adev) { - if (adev->gart.gart_funcs == NULL) - adev->gart.gart_funcs = &gmc_v7_0_gart_funcs; + if (adev->gmc.gmc_funcs == NULL) + adev->gmc.gmc_funcs = &gmc_v7_0_gmc_funcs; } static void gmc_v7_0_set_irq_funcs(struct amdgpu_device *adev) { - adev->mc.vm_fault.num_types = 1; - adev->mc.vm_fault.funcs = &gmc_v7_0_irq_funcs; + adev->gmc.vm_fault.num_types = 1; + adev->gmc.vm_fault.funcs = &gmc_v7_0_irq_funcs; } const struct amdgpu_ip_block_version gmc_v7_0_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index c8fb7e6..57ee13a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -22,6 +22,7 @@ */ #include <linux/firmware.h> #include <drm/drmP.h> +#include <drm/drm_cache.h> #include "amdgpu.h" #include "gmc_v8_0.h" #include "amdgpu_ucode.h" @@ -44,7 +45,7 @@ #include "amdgpu_atombios.h" -static void gmc_v8_0_set_gart_funcs(struct amdgpu_device *adev); +static void gmc_v8_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v8_0_set_irq_funcs(struct amdgpu_device *adev); static int gmc_v8_0_wait_for_idle(void *handle); @@ -264,16 +265,16 @@ static int gmc_v8_0_init_microcode(struct amdgpu_device *adev) } snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mc.bin", chip_name); - err = request_firmware(&adev->mc.fw, fw_name, adev->dev); + err = request_firmware(&adev->gmc.fw, fw_name, adev->dev); if (err) goto out; - err = amdgpu_ucode_validate(adev->mc.fw); + err = amdgpu_ucode_validate(adev->gmc.fw); out: if (err) { pr_err("mc: Failed to load firmware \"%s\"\n", fw_name); - release_firmware(adev->mc.fw); - adev->mc.fw = NULL; + release_firmware(adev->gmc.fw); + adev->gmc.fw = NULL; } return err; } @@ -302,19 +303,19 @@ static int gmc_v8_0_tonga_mc_load_microcode(struct amdgpu_device *adev) if (amdgpu_sriov_bios(adev)) return 0; - if (!adev->mc.fw) + if (!adev->gmc.fw) return -EINVAL; - hdr = (const struct mc_firmware_header_v1_0 *)adev->mc.fw->data; + hdr = (const struct mc_firmware_header_v1_0 *)adev->gmc.fw->data; amdgpu_ucode_print_mc_hdr(&hdr->header); - adev->mc.fw_version = le32_to_cpu(hdr->header.ucode_version); + adev->gmc.fw_version = le32_to_cpu(hdr->header.ucode_version); regs_size = le32_to_cpu(hdr->io_debug_size_bytes) / (4 * 2); io_mc_regs = (const __le32 *) - (adev->mc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); + (adev->gmc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4; fw_data = (const __le32 *) - (adev->mc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + (adev->gmc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); running = REG_GET_FIELD(RREG32(mmMC_SEQ_SUP_CNTL), MC_SEQ_SUP_CNTL, RUN); @@ -378,19 +379,19 @@ static int gmc_v8_0_polaris_mc_load_microcode(struct amdgpu_device *adev) if (vbios_version == 0) return 0; - if (!adev->mc.fw) + if (!adev->gmc.fw) return -EINVAL; - hdr = (const struct mc_firmware_header_v1_0 *)adev->mc.fw->data; + hdr = (const struct mc_firmware_header_v1_0 *)adev->gmc.fw->data; amdgpu_ucode_print_mc_hdr(&hdr->header); - adev->mc.fw_version = le32_to_cpu(hdr->header.ucode_version); + adev->gmc.fw_version = le32_to_cpu(hdr->header.ucode_version); regs_size = le32_to_cpu(hdr->io_debug_size_bytes) / (4 * 2); io_mc_regs = (const __le32 *) - (adev->mc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); + (adev->gmc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4; fw_data = (const __le32 *) - (adev->mc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + (adev->gmc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); data = RREG32(mmMC_SEQ_MISC0); data &= ~(0x40); @@ -426,7 +427,7 @@ static int gmc_v8_0_polaris_mc_load_microcode(struct amdgpu_device *adev) } static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev, - struct amdgpu_mc *mc) + struct amdgpu_gmc *mc) { u64 base = 0; @@ -434,7 +435,7 @@ static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev, base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; base <<= 24; - amdgpu_device_vram_location(adev, &adev->mc, base); + amdgpu_device_vram_location(adev, &adev->gmc, base); amdgpu_device_gart_location(adev, mc); } @@ -477,18 +478,18 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) } /* Update configuration */ WREG32(mmMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->mc.vram_start >> 12); + adev->gmc.vram_start >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->mc.vram_end >> 12); + adev->gmc.vram_end >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_DEFAULT_ADDR, adev->vram_scratch.gpu_addr >> 12); if (amdgpu_sriov_vf(adev)) { - tmp = ((adev->mc.vram_end >> 24) & 0xFFFF) << 16; - tmp |= ((adev->mc.vram_start >> 24) & 0xFFFF); + tmp = ((adev->gmc.vram_end >> 24) & 0xFFFF) << 16; + tmp |= ((adev->gmc.vram_start >> 24) & 0xFFFF); WREG32(mmMC_VM_FB_LOCATION, tmp); /* XXX double check these! */ - WREG32(mmHDP_NONSURFACE_BASE, (adev->mc.vram_start >> 8)); + WREG32(mmHDP_NONSURFACE_BASE, (adev->gmc.vram_start >> 8)); WREG32(mmHDP_NONSURFACE_INFO, (2 << 7) | (1 << 30)); WREG32(mmHDP_NONSURFACE_SIZE, 0x3FFFFFFF); } @@ -523,8 +524,8 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev) { int r; - adev->mc.vram_width = amdgpu_atombios_get_vram_width(adev); - if (!adev->mc.vram_width) { + adev->gmc.vram_width = amdgpu_atombios_get_vram_width(adev); + if (!adev->gmc.vram_width) { u32 tmp; int chansize, numchan; @@ -566,31 +567,31 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev) numchan = 16; break; } - adev->mc.vram_width = numchan * chansize; + adev->gmc.vram_width = numchan * chansize; } /* size in MB on si */ - adev->mc.mc_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; - adev->mc.real_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; + adev->gmc.mc_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; + adev->gmc.real_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; if (!(adev->flags & AMD_IS_APU)) { r = amdgpu_device_resize_fb_bar(adev); if (r) return r; } - adev->mc.aper_base = pci_resource_start(adev->pdev, 0); - adev->mc.aper_size = pci_resource_len(adev->pdev, 0); + adev->gmc.aper_base = pci_resource_start(adev->pdev, 0); + adev->gmc.aper_size = pci_resource_len(adev->pdev, 0); #ifdef CONFIG_X86_64 if (adev->flags & AMD_IS_APU) { - adev->mc.aper_base = ((u64)RREG32(mmMC_VM_FB_OFFSET)) << 22; - adev->mc.aper_size = adev->mc.real_vram_size; + adev->gmc.aper_base = ((u64)RREG32(mmMC_VM_FB_OFFSET)) << 22; + adev->gmc.aper_size = adev->gmc.real_vram_size; } #endif /* In case the PCI BAR is larger than the actual amount of vram */ - adev->mc.visible_vram_size = adev->mc.aper_size; - if (adev->mc.visible_vram_size > adev->mc.real_vram_size) - adev->mc.visible_vram_size = adev->mc.real_vram_size; + adev->gmc.visible_vram_size = adev->gmc.aper_size; + if (adev->gmc.visible_vram_size > adev->gmc.real_vram_size) + adev->gmc.visible_vram_size = adev->gmc.real_vram_size; /* set the gart size */ if (amdgpu_gart_size == -1) { @@ -599,20 +600,20 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev) case CHIP_POLARIS10: /* all engines support GPUVM */ case CHIP_POLARIS12: /* all engines support GPUVM */ default: - adev->mc.gart_size = 256ULL << 20; + adev->gmc.gart_size = 256ULL << 20; break; case CHIP_TONGA: /* UVD, VCE do not support GPUVM */ case CHIP_FIJI: /* UVD, VCE do not support GPUVM */ case CHIP_CARRIZO: /* UVD, VCE do not support GPUVM, DCE SG support */ case CHIP_STONEY: /* UVD does not support GPUVM, DCE SG support */ - adev->mc.gart_size = 1024ULL << 20; + adev->gmc.gart_size = 1024ULL << 20; break; } } else { - adev->mc.gart_size = (u64)amdgpu_gart_size << 20; + adev->gmc.gart_size = (u64)amdgpu_gart_size << 20; } - gmc_v8_0_vram_gtt_location(adev, &adev->mc); + gmc_v8_0_vram_gtt_location(adev, &adev->gmc); return 0; } @@ -625,25 +626,45 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev) */ /** - * gmc_v8_0_gart_flush_gpu_tlb - gart tlb flush callback + * gmc_v8_0_flush_gpu_tlb - gart tlb flush callback * * @adev: amdgpu_device pointer * @vmid: vm instance to flush * * Flush the TLB for the requested page table (CIK). */ -static void gmc_v8_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, +static void gmc_v8_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid) { - /* flush hdp cache */ - WREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0); - /* bits 0-15 are the VM contexts0-15 */ WREG32(mmVM_INVALIDATE_REQUEST, 1 << vmid); } +static uint64_t gmc_v8_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, + unsigned vmid, uint64_t pd_addr) +{ + uint32_t reg; + + if (vmid < 8) + reg = mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid; + else + reg = mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8; + amdgpu_ring_emit_wreg(ring, reg, pd_addr >> 12); + + /* bits 0-15 are the VM contexts0-15 */ + amdgpu_ring_emit_wreg(ring, mmVM_INVALIDATE_REQUEST, 1 << vmid); + + return pd_addr; +} + +static void gmc_v8_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid, + unsigned pasid) +{ + amdgpu_ring_emit_wreg(ring, mmIH_VMID_0_LUT + vmid, pasid); +} + /** - * gmc_v8_0_gart_set_pte_pde - update the page tables using MMIO + * gmc_v8_0_set_pte_pde - update the page tables using MMIO * * @adev: amdgpu_device pointer * @cpu_pt_addr: cpu address of the page table @@ -653,11 +674,9 @@ static void gmc_v8_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, * * Update the page tables using the CPU. */ -static int gmc_v8_0_gart_set_pte_pde(struct amdgpu_device *adev, - void *cpu_pt_addr, - uint32_t gpu_page_idx, - uint64_t addr, - uint64_t flags) +static int gmc_v8_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, + uint32_t gpu_page_idx, uint64_t addr, + uint64_t flags) { void __iomem *ptr = (void *)cpu_pt_addr; uint64_t value; @@ -751,9 +770,9 @@ static void gmc_v8_0_set_prt(struct amdgpu_device *adev, bool enable) { u32 tmp; - if (enable && !adev->mc.prt_warning) { + if (enable && !adev->gmc.prt_warning) { dev_warn(adev->dev, "Disabling VM faults because of PRT request!\n"); - adev->mc.prt_warning = true; + adev->gmc.prt_warning = true; } tmp = RREG32(mmVM_PRT_CNTL); @@ -775,7 +794,8 @@ static void gmc_v8_0_set_prt(struct amdgpu_device *adev, bool enable) if (enable) { uint32_t low = AMDGPU_VA_RESERVED_SIZE >> AMDGPU_GPU_PAGE_SHIFT; - uint32_t high = adev->vm_manager.max_pfn; + uint32_t high = adev->vm_manager.max_pfn - + (AMDGPU_VA_RESERVED_SIZE >> AMDGPU_GPU_PAGE_SHIFT); WREG32(mmVM_PRT_APERTURE0_LOW_ADDR, low); WREG32(mmVM_PRT_APERTURE1_LOW_ADDR, low); @@ -865,11 +885,11 @@ static int gmc_v8_0_gart_enable(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_L2_CNTL4, VMC_TAP_CONTEXT1_PTE_REQUEST_SNOOP, 0); WREG32(mmVM_L2_CNTL4, tmp); /* setup context0 */ - WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gart_start >> 12); - WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gart_end >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->gmc.gart_start >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->gmc.gart_end >> 12); WREG32(mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR, adev->gart.table_addr >> 12); WREG32(mmVM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, - (u32)(adev->dummy_page.addr >> 12)); + (u32)(adev->dummy_page_addr >> 12)); WREG32(mmVM_CONTEXT0_CNTL2, 0); tmp = RREG32(mmVM_CONTEXT0_CNTL); tmp = REG_SET_FIELD(tmp, VM_CONTEXT0_CNTL, ENABLE_CONTEXT, 1); @@ -899,7 +919,7 @@ static int gmc_v8_0_gart_enable(struct amdgpu_device *adev) /* enable context1-15 */ WREG32(mmVM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR, - (u32)(adev->dummy_page.addr >> 12)); + (u32)(adev->dummy_page_addr >> 12)); WREG32(mmVM_CONTEXT1_CNTL2, 4); tmp = RREG32(mmVM_CONTEXT1_CNTL); tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, ENABLE_CONTEXT, 1); @@ -919,9 +939,9 @@ static int gmc_v8_0_gart_enable(struct amdgpu_device *adev) else gmc_v8_0_set_fault_enable_default(adev, true); - gmc_v8_0_gart_flush_gpu_tlb(adev, 0); + gmc_v8_0_flush_gpu_tlb(adev, 0); DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n", - (unsigned)(adev->mc.gart_size >> 20), + (unsigned)(adev->gmc.gart_size >> 20), (unsigned long long)adev->gart.table_addr); adev->gart.ready = true; return 0; @@ -994,21 +1014,21 @@ static void gmc_v8_0_gart_fini(struct amdgpu_device *adev) * * Print human readable fault information (CIK). */ -static void gmc_v8_0_vm_decode_fault(struct amdgpu_device *adev, - u32 status, u32 addr, u32 mc_client) +static void gmc_v8_0_vm_decode_fault(struct amdgpu_device *adev, u32 status, + u32 addr, u32 mc_client, unsigned pasid) { - u32 mc_id; u32 vmid = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, VMID); u32 protections = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, PROTECTIONS); char block[5] = { mc_client >> 24, (mc_client >> 16) & 0xff, (mc_client >> 8) & 0xff, mc_client & 0xff, 0 }; + u32 mc_id; mc_id = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, MEMORY_CLIENT_ID); - dev_err(adev->dev, "VM fault (0x%02x, vmid %d) at page %u, %s from '%s' (0x%08x) (%d)\n", - protections, vmid, addr, + dev_err(adev->dev, "VM fault (0x%02x, vmid %d, pasid %d) at page %u, %s from '%s' (0x%08x) (%d)\n", + protections, vmid, pasid, addr, REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, MEMORY_CLIENT_RW) ? "write" : "read", block, mc_client, mc_id); @@ -1040,16 +1060,16 @@ static int gmc_v8_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - gmc_v8_0_set_gart_funcs(adev); + gmc_v8_0_set_gmc_funcs(adev); gmc_v8_0_set_irq_funcs(adev); - adev->mc.shared_aperture_start = 0x2000000000000000ULL; - adev->mc.shared_aperture_end = - adev->mc.shared_aperture_start + (4ULL << 30) - 1; - adev->mc.private_aperture_start = - adev->mc.shared_aperture_end + 1; - adev->mc.private_aperture_end = - adev->mc.private_aperture_start + (4ULL << 30) - 1; + adev->gmc.shared_aperture_start = 0x2000000000000000ULL; + adev->gmc.shared_aperture_end = + adev->gmc.shared_aperture_start + (4ULL << 30) - 1; + adev->gmc.private_aperture_start = + adev->gmc.shared_aperture_end + 1; + adev->gmc.private_aperture_end = + adev->gmc.private_aperture_start + (4ULL << 30) - 1; return 0; } @@ -1061,7 +1081,7 @@ static int gmc_v8_0_late_init(void *handle) amdgpu_bo_late_init(adev); if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS) - return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0); + return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); else return 0; } @@ -1080,7 +1100,7 @@ static unsigned gmc_v8_0_get_vbios_fb_size(struct amdgpu_device *adev) 4); } /* return 0 if the pre-OS buffer uses up most of vram */ - if ((adev->mc.real_vram_size - size) < (8 * 1024 * 1024)) + if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024)) return 0; return size; } @@ -1094,7 +1114,7 @@ static int gmc_v8_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (adev->flags & AMD_IS_APU) { - adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; + adev->gmc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; } else { u32 tmp; @@ -1103,14 +1123,14 @@ static int gmc_v8_0_sw_init(void *handle) else tmp = RREG32(mmMC_SEQ_MISC0); tmp &= MC_SEQ_MISC0__MT__MASK; - adev->mc.vram_type = gmc_v8_0_convert_vram_type(tmp); + adev->gmc.vram_type = gmc_v8_0_convert_vram_type(tmp); } - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->mc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->gmc.vm_fault); if (r) return r; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->mc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->gmc.vm_fault); if (r) return r; @@ -1124,7 +1144,7 @@ static int gmc_v8_0_sw_init(void *handle) * This is the max address of the GPU's * internal address space. */ - adev->mc.mc_mask = 0xffffffffffULL; /* 40 bit MC */ + adev->gmc.mc_mask = 0xffffffffffULL; /* 40 bit MC */ /* set DMA mask + need_dma32 flags. * PCIE - can handle 40-bits. @@ -1144,6 +1164,7 @@ static int gmc_v8_0_sw_init(void *handle) pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32)); pr_warn("amdgpu: No coherent DMA available\n"); } + adev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits); r = gmc_v8_0_init_microcode(adev); if (r) { @@ -1155,7 +1176,7 @@ static int gmc_v8_0_sw_init(void *handle) if (r) return r; - adev->mc.stolen_size = gmc_v8_0_get_vbios_fb_size(adev); + adev->gmc.stolen_size = gmc_v8_0_get_vbios_fb_size(adev); /* Memory manager */ r = amdgpu_bo_init(adev); @@ -1196,8 +1217,8 @@ static int gmc_v8_0_sw_fini(void *handle) amdgpu_vm_manager_fini(adev); gmc_v8_0_gart_fini(adev); amdgpu_bo_fini(adev); - release_firmware(adev->mc.fw); - adev->mc.fw = NULL; + release_firmware(adev->gmc.fw); + adev->gmc.fw = NULL; return 0; } @@ -1238,7 +1259,7 @@ static int gmc_v8_0_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - amdgpu_irq_put(adev, &adev->mc.vm_fault, 0); + amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); gmc_v8_0_gart_disable(adev); return 0; @@ -1318,10 +1339,10 @@ static bool gmc_v8_0_check_soft_reset(void *handle) SRBM_SOFT_RESET, SOFT_RESET_MC, 1); } if (srbm_soft_reset) { - adev->mc.srbm_soft_reset = srbm_soft_reset; + adev->gmc.srbm_soft_reset = srbm_soft_reset; return true; } else { - adev->mc.srbm_soft_reset = 0; + adev->gmc.srbm_soft_reset = 0; return false; } } @@ -1330,7 +1351,7 @@ static int gmc_v8_0_pre_soft_reset(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - if (!adev->mc.srbm_soft_reset) + if (!adev->gmc.srbm_soft_reset) return 0; gmc_v8_0_mc_stop(adev); @@ -1346,9 +1367,9 @@ static int gmc_v8_0_soft_reset(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; u32 srbm_soft_reset; - if (!adev->mc.srbm_soft_reset) + if (!adev->gmc.srbm_soft_reset) return 0; - srbm_soft_reset = adev->mc.srbm_soft_reset; + srbm_soft_reset = adev->gmc.srbm_soft_reset; if (srbm_soft_reset) { u32 tmp; @@ -1376,7 +1397,7 @@ static int gmc_v8_0_post_soft_reset(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - if (!adev->mc.srbm_soft_reset) + if (!adev->gmc.srbm_soft_reset) return 0; gmc_v8_0_mc_resume(adev); @@ -1457,7 +1478,8 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev, addr); dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n", status); - gmc_v8_0_vm_decode_fault(adev, status, addr, mc_client); + gmc_v8_0_vm_decode_fault(adev, status, addr, mc_client, + entry->pasid); } return 0; @@ -1689,9 +1711,11 @@ static const struct amd_ip_funcs gmc_v8_0_ip_funcs = { .get_clockgating_state = gmc_v8_0_get_clockgating_state, }; -static const struct amdgpu_gart_funcs gmc_v8_0_gart_funcs = { - .flush_gpu_tlb = gmc_v8_0_gart_flush_gpu_tlb, - .set_pte_pde = gmc_v8_0_gart_set_pte_pde, +static const struct amdgpu_gmc_funcs gmc_v8_0_gmc_funcs = { + .flush_gpu_tlb = gmc_v8_0_flush_gpu_tlb, + .emit_flush_gpu_tlb = gmc_v8_0_emit_flush_gpu_tlb, + .emit_pasid_mapping = gmc_v8_0_emit_pasid_mapping, + .set_pte_pde = gmc_v8_0_set_pte_pde, .set_prt = gmc_v8_0_set_prt, .get_vm_pte_flags = gmc_v8_0_get_vm_pte_flags, .get_vm_pde = gmc_v8_0_get_vm_pde @@ -1702,16 +1726,16 @@ static const struct amdgpu_irq_src_funcs gmc_v8_0_irq_funcs = { .process = gmc_v8_0_process_interrupt, }; -static void gmc_v8_0_set_gart_funcs(struct amdgpu_device *adev) +static void gmc_v8_0_set_gmc_funcs(struct amdgpu_device *adev) { - if (adev->gart.gart_funcs == NULL) - adev->gart.gart_funcs = &gmc_v8_0_gart_funcs; + if (adev->gmc.gmc_funcs == NULL) + adev->gmc.gmc_funcs = &gmc_v8_0_gmc_funcs; } static void gmc_v8_0_set_irq_funcs(struct amdgpu_device *adev) { - adev->mc.vm_fault.num_types = 1; - adev->mc.vm_fault.funcs = &gmc_v8_0_irq_funcs; + adev->gmc.vm_fault.num_types = 1; + adev->gmc.vm_fault.funcs = &gmc_v8_0_irq_funcs; } const struct amdgpu_ip_block_version gmc_v8_0_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 6b3f245..15ef3da 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -21,6 +21,7 @@ * */ #include <linux/firmware.h> +#include <drm/drm_cache.h> #include "amdgpu.h" #include "gmc_v9_0.h" #include "amdgpu_atomfirmware.h" @@ -33,6 +34,7 @@ #include "vega10_enum.h" #include "mmhub/mmhub_1_0_offset.h" #include "athub/athub_1_0_offset.h" +#include "oss/osssys_4_0_offset.h" #include "soc15.h" #include "soc15_common.h" @@ -270,10 +272,10 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, if (printk_ratelimit()) { dev_err(adev->dev, - "[%s] VMC page fault (src_id:%u ring:%u vmid:%u pas_id:%u)\n", + "[%s] VMC page fault (src_id:%u ring:%u vmid:%u pasid:%u)\n", entry->vmid_src ? "mmhub" : "gfxhub", entry->src_id, entry->ring_id, entry->vmid, - entry->pas_id); + entry->pasid); dev_err(adev->dev, " at page 0x%016llx from %d\n", addr, entry->client_id); if (!amdgpu_sriov_vf(adev)) @@ -292,8 +294,8 @@ static const struct amdgpu_irq_src_funcs gmc_v9_0_irq_funcs = { static void gmc_v9_0_set_irq_funcs(struct amdgpu_device *adev) { - adev->mc.vm_fault.num_types = 1; - adev->mc.vm_fault.funcs = &gmc_v9_0_irq_funcs; + adev->gmc.vm_fault.num_types = 1; + adev->gmc.vm_fault.funcs = &gmc_v9_0_irq_funcs; } static uint32_t gmc_v9_0_get_invalidate_req(unsigned int vmid) @@ -323,24 +325,21 @@ static uint32_t gmc_v9_0_get_invalidate_req(unsigned int vmid) */ /** - * gmc_v9_0_gart_flush_gpu_tlb - gart tlb flush callback + * gmc_v9_0_flush_gpu_tlb - gart tlb flush callback * * @adev: amdgpu_device pointer * @vmid: vm instance to flush * * Flush the TLB for the requested page table. */ -static void gmc_v9_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, +static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid) { /* Use register 17 for GART */ const unsigned eng = 17; unsigned i, j; - /* flush hdp cache */ - adev->nbio_funcs->hdp_flush(adev); - - spin_lock(&adev->mc.invalidate_lock); + spin_lock(&adev->gmc.invalidate_lock); for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { struct amdgpu_vmhub *hub = &adev->vmhub[i]; @@ -373,11 +372,52 @@ static void gmc_v9_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, DRM_ERROR("Timeout waiting for VM flush ACK!\n"); } - spin_unlock(&adev->mc.invalidate_lock); + spin_unlock(&adev->gmc.invalidate_lock); +} + +static uint64_t gmc_v9_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, + unsigned vmid, uint64_t pd_addr) +{ + struct amdgpu_device *adev = ring->adev; + struct amdgpu_vmhub *hub = &adev->vmhub[ring->funcs->vmhub]; + uint32_t req = gmc_v9_0_get_invalidate_req(vmid); + uint64_t flags = AMDGPU_PTE_VALID; + unsigned eng = ring->vm_inv_eng; + + amdgpu_gmc_get_vm_pde(adev, -1, &pd_addr, &flags); + pd_addr |= flags; + + amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_lo32 + (2 * vmid), + lower_32_bits(pd_addr)); + + amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_hi32 + (2 * vmid), + upper_32_bits(pd_addr)); + + amdgpu_ring_emit_wreg(ring, hub->vm_inv_eng0_req + eng, req); + + /* wait for the invalidate to complete */ + amdgpu_ring_emit_reg_wait(ring, hub->vm_inv_eng0_ack + eng, + 1 << vmid, 1 << vmid); + + return pd_addr; +} + +static void gmc_v9_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid, + unsigned pasid) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t reg; + + if (ring->funcs->vmhub == AMDGPU_GFXHUB) + reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_VMID_0_LUT) + vmid; + else + reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_VMID_0_LUT_MM) + vmid; + + amdgpu_ring_emit_wreg(ring, reg, pasid); } /** - * gmc_v9_0_gart_set_pte_pde - update the page tables using MMIO + * gmc_v9_0_set_pte_pde - update the page tables using MMIO * * @adev: amdgpu_device pointer * @cpu_pt_addr: cpu address of the page table @@ -387,11 +427,9 @@ static void gmc_v9_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, * * Update the page tables using the CPU. */ -static int gmc_v9_0_gart_set_pte_pde(struct amdgpu_device *adev, - void *cpu_pt_addr, - uint32_t gpu_page_idx, - uint64_t addr, - uint64_t flags) +static int gmc_v9_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, + uint32_t gpu_page_idx, uint64_t addr, + uint64_t flags) { void __iomem *ptr = (void *)cpu_pt_addr; uint64_t value; @@ -482,10 +520,10 @@ static void gmc_v9_0_get_vm_pde(struct amdgpu_device *adev, int level, { if (!(*flags & AMDGPU_PDE_PTE)) *addr = adev->vm_manager.vram_base_offset + *addr - - adev->mc.vram_start; + adev->gmc.vram_start; BUG_ON(*addr & 0xFFFF00000000003FULL); - if (!adev->mc.translate_further) + if (!adev->gmc.translate_further) return; if (level == AMDGPU_VM_PDB1) { @@ -501,34 +539,35 @@ static void gmc_v9_0_get_vm_pde(struct amdgpu_device *adev, int level, } } -static const struct amdgpu_gart_funcs gmc_v9_0_gart_funcs = { - .flush_gpu_tlb = gmc_v9_0_gart_flush_gpu_tlb, - .set_pte_pde = gmc_v9_0_gart_set_pte_pde, - .get_invalidate_req = gmc_v9_0_get_invalidate_req, +static const struct amdgpu_gmc_funcs gmc_v9_0_gmc_funcs = { + .flush_gpu_tlb = gmc_v9_0_flush_gpu_tlb, + .emit_flush_gpu_tlb = gmc_v9_0_emit_flush_gpu_tlb, + .emit_pasid_mapping = gmc_v9_0_emit_pasid_mapping, + .set_pte_pde = gmc_v9_0_set_pte_pde, .get_vm_pte_flags = gmc_v9_0_get_vm_pte_flags, .get_vm_pde = gmc_v9_0_get_vm_pde }; -static void gmc_v9_0_set_gart_funcs(struct amdgpu_device *adev) +static void gmc_v9_0_set_gmc_funcs(struct amdgpu_device *adev) { - if (adev->gart.gart_funcs == NULL) - adev->gart.gart_funcs = &gmc_v9_0_gart_funcs; + if (adev->gmc.gmc_funcs == NULL) + adev->gmc.gmc_funcs = &gmc_v9_0_gmc_funcs; } static int gmc_v9_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - gmc_v9_0_set_gart_funcs(adev); + gmc_v9_0_set_gmc_funcs(adev); gmc_v9_0_set_irq_funcs(adev); - adev->mc.shared_aperture_start = 0x2000000000000000ULL; - adev->mc.shared_aperture_end = - adev->mc.shared_aperture_start + (4ULL << 30) - 1; - adev->mc.private_aperture_start = - adev->mc.shared_aperture_end + 1; - adev->mc.private_aperture_end = - adev->mc.private_aperture_start + (4ULL << 30) - 1; + adev->gmc.shared_aperture_start = 0x2000000000000000ULL; + adev->gmc.shared_aperture_end = + adev->gmc.shared_aperture_start + (4ULL << 30) - 1; + adev->gmc.private_aperture_start = + adev->gmc.shared_aperture_end + 1; + adev->gmc.private_aperture_end = + adev->gmc.private_aperture_start + (4ULL << 30) - 1; return 0; } @@ -659,16 +698,16 @@ static int gmc_v9_0_late_init(void *handle) } } - return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0); + return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); } static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev, - struct amdgpu_mc *mc) + struct amdgpu_gmc *mc) { u64 base = 0; if (!amdgpu_sriov_vf(adev)) base = mmhub_v1_0_get_fb_location(adev); - amdgpu_device_vram_location(adev, &adev->mc, base); + amdgpu_device_vram_location(adev, &adev->gmc, base); amdgpu_device_gart_location(adev, mc); /* base offset of vram pages */ if (adev->flags & AMD_IS_APU) @@ -692,8 +731,9 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev) int chansize, numchan; int r; - adev->mc.vram_width = amdgpu_atomfirmware_get_vram_width(adev); - if (!adev->mc.vram_width) { + if (amdgpu_emu_mode != 1) + adev->gmc.vram_width = amdgpu_atomfirmware_get_vram_width(adev); + if (!adev->gmc.vram_width) { /* hbm memory channel size */ if (adev->flags & AMD_IS_APU) chansize = 64; @@ -733,43 +773,50 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev) numchan = 2; break; } - adev->mc.vram_width = numchan * chansize; + adev->gmc.vram_width = numchan * chansize; } /* size in MB on si */ - adev->mc.mc_vram_size = + adev->gmc.mc_vram_size = adev->nbio_funcs->get_memsize(adev) * 1024ULL * 1024ULL; - adev->mc.real_vram_size = adev->mc.mc_vram_size; + adev->gmc.real_vram_size = adev->gmc.mc_vram_size; if (!(adev->flags & AMD_IS_APU)) { r = amdgpu_device_resize_fb_bar(adev); if (r) return r; } - adev->mc.aper_base = pci_resource_start(adev->pdev, 0); - adev->mc.aper_size = pci_resource_len(adev->pdev, 0); + adev->gmc.aper_base = pci_resource_start(adev->pdev, 0); + adev->gmc.aper_size = pci_resource_len(adev->pdev, 0); +#ifdef CONFIG_X86_64 + if (adev->flags & AMD_IS_APU) { + adev->gmc.aper_base = gfxhub_v1_0_get_mc_fb_offset(adev); + adev->gmc.aper_size = adev->gmc.real_vram_size; + } +#endif /* In case the PCI BAR is larger than the actual amount of vram */ - adev->mc.visible_vram_size = adev->mc.aper_size; - if (adev->mc.visible_vram_size > adev->mc.real_vram_size) - adev->mc.visible_vram_size = adev->mc.real_vram_size; + adev->gmc.visible_vram_size = adev->gmc.aper_size; + if (adev->gmc.visible_vram_size > adev->gmc.real_vram_size) + adev->gmc.visible_vram_size = adev->gmc.real_vram_size; /* set the gart size */ if (amdgpu_gart_size == -1) { switch (adev->asic_type) { case CHIP_VEGA10: /* all engines support GPUVM */ + case CHIP_VEGA12: /* all engines support GPUVM */ default: - adev->mc.gart_size = 256ULL << 20; + adev->gmc.gart_size = 512ULL << 20; break; case CHIP_RAVEN: /* DCE SG support */ - adev->mc.gart_size = 1024ULL << 20; + adev->gmc.gart_size = 1024ULL << 20; break; } } else { - adev->mc.gart_size = (u64)amdgpu_gart_size << 20; + adev->gmc.gart_size = (u64)amdgpu_gart_size << 20; } - gmc_v9_0_vram_gtt_location(adev, &adev->mc); + gmc_v9_0_vram_gtt_location(adev, &adev->gmc); return 0; } @@ -830,7 +877,7 @@ static unsigned gmc_v9_0_get_vbios_fb_size(struct amdgpu_device *adev) } } /* return 0 if the pre-OS buffer uses up most of vram */ - if ((adev->mc.real_vram_size - size) < (8 * 1024 * 1024)) + if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024)) return 0; #endif @@ -846,23 +893,22 @@ static int gmc_v9_0_sw_init(void *handle) gfxhub_v1_0_init(adev); mmhub_v1_0_init(adev); - spin_lock_init(&adev->mc.invalidate_lock); + spin_lock_init(&adev->gmc.invalidate_lock); + adev->gmc.vram_type = amdgpu_atomfirmware_get_vram_type(adev); switch (adev->asic_type) { case CHIP_RAVEN: - adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; if (adev->rev_id == 0x0 || adev->rev_id == 0x1) { amdgpu_vm_adjust_size(adev, 256 * 1024, 9, 3, 48); } else { /* vm_size is 128TB + 512GB for legacy 3-level page support */ amdgpu_vm_adjust_size(adev, 128 * 1024 + 512, 9, 2, 48); - adev->mc.translate_further = + adev->gmc.translate_further = adev->vm_manager.num_level > 1; } break; case CHIP_VEGA10: - /* XXX Don't know how to get VRAM type yet. */ - adev->mc.vram_type = AMDGPU_VRAM_TYPE_HBM; + case CHIP_VEGA12: /* * To fulfill 4-level page support, * vm size is 256TB (48bit), maximum size of Vega10, @@ -875,10 +921,10 @@ static int gmc_v9_0_sw_init(void *handle) } /* This interrupt is VMC page fault.*/ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_VMC, 0, - &adev->mc.vm_fault); - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_UTCL2, 0, - &adev->mc.vm_fault); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VMC, 0, + &adev->gmc.vm_fault); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UTCL2, 0, + &adev->gmc.vm_fault); if (r) return r; @@ -887,7 +933,7 @@ static int gmc_v9_0_sw_init(void *handle) * This is the max address of the GPU's * internal address space. */ - adev->mc.mc_mask = 0xffffffffffffULL; /* 48 bit MC */ + adev->gmc.mc_mask = 0xffffffffffffULL; /* 48 bit MC */ /* set DMA mask + need_dma32 flags. * PCIE - can handle 44-bits. @@ -907,12 +953,13 @@ static int gmc_v9_0_sw_init(void *handle) pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32)); printk(KERN_WARNING "amdgpu: No coherent DMA available.\n"); } + adev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits); r = gmc_v9_0_mc_init(adev); if (r) return r; - adev->mc.stolen_size = gmc_v9_0_get_vbios_fb_size(adev); + adev->gmc.stolen_size = gmc_v9_0_get_vbios_fb_size(adev); /* Memory manager */ r = amdgpu_bo_init(adev); @@ -986,6 +1033,8 @@ static void gmc_v9_0_init_golden_registers(struct amdgpu_device *adev) golden_settings_athub_1_0_0, ARRAY_SIZE(golden_settings_athub_1_0_0)); break; + case CHIP_VEGA12: + break; case CHIP_RAVEN: soc15_program_register_sequence(adev, golden_settings_athub_1_0_0, @@ -1042,7 +1091,7 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev) WREG32_SOC15(HDP, 0, mmHDP_HOST_PATH_CNTL, tmp); /* After HDP is initialized, flush HDP.*/ - adev->nbio_funcs->hdp_flush(adev); + adev->nbio_funcs->hdp_flush(adev, NULL); if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS) value = false; @@ -1051,10 +1100,10 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev) gfxhub_v1_0_set_fault_enable_default(adev, value); mmhub_v1_0_set_fault_enable_default(adev, value); - gmc_v9_0_gart_flush_gpu_tlb(adev, 0); + gmc_v9_0_flush_gpu_tlb(adev, 0); DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n", - (unsigned)(adev->mc.gart_size >> 20), + (unsigned)(adev->gmc.gart_size >> 20), (unsigned long long)adev->gart.table_addr); adev->gart.ready = true; return 0; @@ -1105,7 +1154,7 @@ static int gmc_v9_0_hw_fini(void *handle) return 0; } - amdgpu_irq_put(adev, &adev->mc.vm_fault, 0); + amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); gmc_v9_0_gart_disable(adev); return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c index c4e4be3..842c4b6 100644 --- a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c
@@ -111,7 +111,7 @@ static int iceland_ih_irq_init(struct amdgpu_device *adev) iceland_ih_disable_interrupts(adev); /* setup interrupt control */ - WREG32(mmINTERRUPT_CNTL2, adev->dummy_page.addr >> 8); + WREG32(mmINTERRUPT_CNTL2, adev->dummy_page_addr >> 8); interrupt_cntl = RREG32(mmINTERRUPT_CNTL); /* INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=0 - dummy read disabled with msi, enabled without msi * INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=1 - dummy read controlled by IH_DUMMY_RD_EN @@ -260,7 +260,7 @@ static void iceland_ih_decode_iv(struct amdgpu_device *adev, entry->src_data[0] = dw[1] & 0xfffffff; entry->ring_id = dw[2] & 0xff; entry->vmid = (dw[2] >> 8) & 0xff; - entry->pas_id = (dw[2] >> 16) & 0xffff; + entry->pasid = (dw[2] >> 16) & 0xffff; /* wptr/rptr are in bytes! */ adev->irq.ih.rptr += 16;
diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c index a89ae07..ddff521 100644 --- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
@@ -42,6 +42,8 @@ #define KV_MINIMUM_ENGINE_CLOCK 800 #define SMC_RAM_END 0x40000 +static const struct amd_pm_funcs kv_dpm_funcs; + static void kv_dpm_set_irq_funcs(struct amdgpu_device *adev); static int kv_enable_nb_dpm(struct amdgpu_device *adev, bool enable); @@ -2958,6 +2960,8 @@ static int kv_dpm_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + adev->powerplay.pp_funcs = &kv_dpm_funcs; + adev->powerplay.pp_handle = adev; kv_dpm_set_irq_funcs(adev); return 0; @@ -3312,7 +3316,7 @@ static int kv_set_powergating_by_smu(void *handle, return 0; } -const struct amd_ip_funcs kv_dpm_ip_funcs = { +static const struct amd_ip_funcs kv_dpm_ip_funcs = { .name = "kv_dpm", .early_init = kv_dpm_early_init, .late_init = kv_dpm_late_init, @@ -3329,8 +3333,16 @@ const struct amd_ip_funcs kv_dpm_ip_funcs = { .set_powergating_state = kv_dpm_set_powergating_state, }; -const struct amd_pm_funcs kv_dpm_funcs = { - .get_temperature = &kv_dpm_get_temp, +const struct amdgpu_ip_block_version kv_smu_ip_block = +{ + .type = AMD_IP_BLOCK_TYPE_SMC, + .major = 1, + .minor = 0, + .rev = 0, + .funcs = &kv_dpm_ip_funcs, +}; + +static const struct amd_pm_funcs kv_dpm_funcs = { .pre_set_power_state = &kv_dpm_pre_set_power_state, .set_power_state = &kv_dpm_set_power_state, .post_set_power_state = &kv_dpm_post_set_power_state,
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index a541d0c..360b6ca 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
@@ -50,7 +50,7 @@ static void mmhub_v1_0_init_gart_pt_regs(struct amdgpu_device *adev) uint64_t value; BUG_ON(adev->gart.table_addr & (~0x0000FFFFFFFFF000ULL)); - value = adev->gart.table_addr - adev->mc.vram_start + + value = adev->gart.table_addr - adev->gmc.vram_start + adev->vm_manager.vram_base_offset; value &= 0x0000FFFFFFFFF000ULL; value |= 0x1; /* valid bit */ @@ -67,14 +67,14 @@ static void mmhub_v1_0_init_gart_aperture_regs(struct amdgpu_device *adev) mmhub_v1_0_init_gart_pt_regs(adev); WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32, - (u32)(adev->mc.gart_start >> 12)); + (u32)(adev->gmc.gart_start >> 12)); WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32, - (u32)(adev->mc.gart_start >> 44)); + (u32)(adev->gmc.gart_start >> 44)); WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32, - (u32)(adev->mc.gart_end >> 12)); + (u32)(adev->gmc.gart_end >> 12)); WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32, - (u32)(adev->mc.gart_end >> 44)); + (u32)(adev->gmc.gart_end >> 44)); } static void mmhub_v1_0_init_system_aperture_regs(struct amdgpu_device *adev) @@ -89,12 +89,12 @@ static void mmhub_v1_0_init_system_aperture_regs(struct amdgpu_device *adev) /* Program the system aperture low logical page number. */ WREG32_SOC15(MMHUB, 0, mmMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->mc.vram_start >> 18); + adev->gmc.vram_start >> 18); WREG32_SOC15(MMHUB, 0, mmMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->mc.vram_end >> 18); + adev->gmc.vram_end >> 18); /* Set default page address. */ - value = adev->vram_scratch.gpu_addr - adev->mc.vram_start + + value = adev->vram_scratch.gpu_addr - adev->gmc.vram_start + adev->vm_manager.vram_base_offset; WREG32_SOC15(MMHUB, 0, mmMC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB, (u32)(value >> 12)); @@ -103,9 +103,9 @@ static void mmhub_v1_0_init_system_aperture_regs(struct amdgpu_device *adev) /* Program "protection fault". */ WREG32_SOC15(MMHUB, 0, mmVM_L2_PROTECTION_FAULT_DEFAULT_ADDR_LO32, - (u32)(adev->dummy_page.addr >> 12)); + (u32)(adev->dummy_page_addr >> 12)); WREG32_SOC15(MMHUB, 0, mmVM_L2_PROTECTION_FAULT_DEFAULT_ADDR_HI32, - (u32)((u64)adev->dummy_page.addr >> 44)); + (u32)((u64)adev->dummy_page_addr >> 44)); tmp = RREG32_SOC15(MMHUB, 0, mmVM_L2_PROTECTION_FAULT_CNTL2); tmp = REG_SET_FIELD(tmp, VM_L2_PROTECTION_FAULT_CNTL2, @@ -155,7 +155,7 @@ static void mmhub_v1_0_init_cache_regs(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_L2_CNTL2, INVALIDATE_L2_CACHE, 1); WREG32_SOC15(MMHUB, 0, mmVM_L2_CNTL2, tmp); - if (adev->mc.translate_further) { + if (adev->gmc.translate_further) { tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 12); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, L2_CACHE_BIGK_FRAGMENT_SIZE, 9); @@ -207,7 +207,7 @@ static void mmhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) num_level = adev->vm_manager.num_level; block_size = adev->vm_manager.block_size; - if (adev->mc.translate_further) + if (adev->gmc.translate_further) num_level -= 1; else block_size -= 9; @@ -272,21 +272,21 @@ static const struct pctl_data pctl0_data[] = { {0x11, 0x6a684}, {0x19, 0xea68e}, {0x29, 0xa69e}, - {0x2b, 0x34a6c0}, - {0x61, 0x83a707}, - {0xe6, 0x8a7a4}, - {0xf0, 0x1a7b8}, - {0xf3, 0xfa7cc}, - {0x104, 0x17a7dd}, - {0x11d, 0xa7dc}, - {0x11f, 0x12a7f5}, - {0x133, 0xa808}, - {0x135, 0x12a810}, - {0x149, 0x7a82c} + {0x2b, 0x0010a6c0}, + {0x3d, 0x83a707}, + {0xc2, 0x8a7a4}, + {0xcc, 0x1a7b8}, + {0xcf, 0xfa7cc}, + {0xe0, 0x17a7dd}, + {0xf9, 0xa7dc}, + {0xfb, 0x12a7f5}, + {0x10f, 0xa808}, + {0x111, 0x12a810}, + {0x125, 0x7a82c} }; #define PCTL0_DATA_LEN (ARRAY_SIZE(pctl0_data)) -#define PCTL0_RENG_EXEC_END_PTR 0x151 +#define PCTL0_RENG_EXEC_END_PTR 0x12d #define PCTL0_STCTRL_REG_SAVE_RANGE0_BASE 0xa640 #define PCTL0_STCTRL_REG_SAVE_RANGE0_LIMIT 0xa833 @@ -385,10 +385,9 @@ void mmhub_v1_0_initialize_power_gating(struct amdgpu_device *adev) if (amdgpu_sriov_vf(adev)) return; + /****************** pctl0 **********************/ pctl0_misc = RREG32_SOC15(MMHUB, 0, mmPCTL0_MISC); pctl0_reng_execute = RREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_EXECUTE); - pctl1_misc = RREG32_SOC15(MMHUB, 0, mmPCTL1_MISC); - pctl1_reng_execute = RREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE); /* Light sleep must be disabled before writing to pctl0 registers */ pctl0_misc &= ~PCTL0_MISC__RENG_MEM_LS_ENABLE_MASK; @@ -402,12 +401,13 @@ void mmhub_v1_0_initialize_power_gating(struct amdgpu_device *adev) pctl0_data[i].data); } - /* Set the reng execute end ptr for pctl0 */ - pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute, - PCTL0_RENG_EXECUTE, - RENG_EXECUTE_END_PTR, - PCTL0_RENG_EXEC_END_PTR); - WREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_EXECUTE, pctl0_reng_execute); + /* Re-enable light sleep */ + pctl0_misc |= PCTL0_MISC__RENG_MEM_LS_ENABLE_MASK; + WREG32_SOC15(MMHUB, 0, mmPCTL0_MISC, pctl0_misc); + + /****************** pctl1 **********************/ + pctl1_misc = RREG32_SOC15(MMHUB, 0, mmPCTL1_MISC); + pctl1_reng_execute = RREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE); /* Light sleep must be disabled before writing to pctl1 registers */ pctl1_misc &= ~PCTL1_MISC__RENG_MEM_LS_ENABLE_MASK; @@ -421,20 +421,25 @@ void mmhub_v1_0_initialize_power_gating(struct amdgpu_device *adev) pctl1_data[i].data); } + /* Re-enable light sleep */ + pctl1_misc |= PCTL1_MISC__RENG_MEM_LS_ENABLE_MASK; + WREG32_SOC15(MMHUB, 0, mmPCTL1_MISC, pctl1_misc); + + mmhub_v1_0_power_gating_write_save_ranges(adev); + + /* Set the reng execute end ptr for pctl0 */ + pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute, + PCTL0_RENG_EXECUTE, + RENG_EXECUTE_END_PTR, + PCTL0_RENG_EXEC_END_PTR); + WREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_EXECUTE, pctl0_reng_execute); + /* Set the reng execute end ptr for pctl1 */ pctl1_reng_execute = REG_SET_FIELD(pctl1_reng_execute, PCTL1_RENG_EXECUTE, RENG_EXECUTE_END_PTR, PCTL1_RENG_EXEC_END_PTR); WREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE, pctl1_reng_execute); - - mmhub_v1_0_power_gating_write_save_ranges(adev); - - /* Re-enable light sleep */ - pctl0_misc |= PCTL0_MISC__RENG_MEM_LS_ENABLE_MASK; - WREG32_SOC15(MMHUB, 0, mmPCTL0_MISC, pctl0_misc); - pctl1_misc |= PCTL1_MISC__RENG_MEM_LS_ENABLE_MASK; - WREG32_SOC15(MMHUB, 0, mmPCTL1_MISC, pctl1_misc); } void mmhub_v1_0_update_power_gating(struct amdgpu_device *adev, @@ -497,9 +502,9 @@ int mmhub_v1_0_gart_enable(struct amdgpu_device *adev) * SRIOV driver need to program them */ WREG32_SOC15(MMHUB, 0, mmMC_VM_FB_LOCATION_BASE, - adev->mc.vram_start >> 24); + adev->gmc.vram_start >> 24); WREG32_SOC15(MMHUB, 0, mmMC_VM_FB_LOCATION_TOP, - adev->mc.vram_end >> 24); + adev->gmc.vram_end >> 24); } /* GART Enable. */ @@ -728,6 +733,7 @@ int mmhub_v1_0_set_clockgating(struct amdgpu_device *adev, switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: case CHIP_RAVEN: mmhub_v1_0_update_medium_grain_clock_gating(adev, state == AMD_CG_STATE_GATE ? true : false);
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index 271452d..4933486 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
@@ -33,56 +33,34 @@ static void xgpu_ai_mailbox_send_ack(struct amdgpu_device *adev) { - u32 reg; - int timeout = AI_MAILBOX_TIMEDOUT; - u32 mask = REG_FIELD_MASK(BIF_BX_PF0_MAILBOX_CONTROL, RCV_MSG_VALID); - - reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_CONTROL)); - reg = REG_SET_FIELD(reg, BIF_BX_PF0_MAILBOX_CONTROL, RCV_MSG_ACK, 1); - WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_CONTROL), reg); - - /*Wait for RCV_MSG_VALID to be 0*/ - reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_CONTROL)); - while (reg & mask) { - if (timeout <= 0) { - pr_err("RCV_MSG_VALID is not cleared\n"); - break; - } - mdelay(1); - timeout -=1; - - reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_CONTROL)); - } + WREG8(AI_MAIBOX_CONTROL_RCV_OFFSET_BYTE, 2); } static void xgpu_ai_mailbox_set_valid(struct amdgpu_device *adev, bool val) { - u32 reg; - - reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_CONTROL)); - reg = REG_SET_FIELD(reg, BIF_BX_PF0_MAILBOX_CONTROL, - TRN_MSG_VALID, val ? 1 : 0); - WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_CONTROL), - reg); + WREG8(AI_MAIBOX_CONTROL_TRN_OFFSET_BYTE, val ? 1 : 0); } +/* + * this peek_msg could *only* be called in IRQ routine becuase in IRQ routine + * RCV_MSG_VALID filed of BIF_BX_PF0_MAILBOX_CONTROL must already be set to 1 + * by host. + * + * if called no in IRQ routine, this peek_msg cannot guaranteed to return the + * correct value since it doesn't return the RCV_DW0 under the case that + * RCV_MSG_VALID is set by host. + */ +static enum idh_event xgpu_ai_mailbox_peek_msg(struct amdgpu_device *adev) +{ + return RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, + mmBIF_BX_PF0_MAILBOX_MSGBUF_RCV_DW0)); +} + + static int xgpu_ai_mailbox_rcv_msg(struct amdgpu_device *adev, enum idh_event event) { u32 reg; - u32 mask = REG_FIELD_MASK(BIF_BX_PF0_MAILBOX_CONTROL, RCV_MSG_VALID); - - if (event != IDH_FLR_NOTIFICATION_CMPL) { - reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_CONTROL)); - if (!(reg & mask)) - return -ENOENT; - } reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_MSGBUF_RCV_DW0)); @@ -94,54 +72,67 @@ static int xgpu_ai_mailbox_rcv_msg(struct amdgpu_device *adev, return 0; } +static uint8_t xgpu_ai_peek_ack(struct amdgpu_device *adev) { + return RREG8(AI_MAIBOX_CONTROL_TRN_OFFSET_BYTE) & 2; +} + static int xgpu_ai_poll_ack(struct amdgpu_device *adev) { - int r = 0, timeout = AI_MAILBOX_TIMEDOUT; - u32 mask = REG_FIELD_MASK(BIF_BX_PF0_MAILBOX_CONTROL, TRN_MSG_ACK); - u32 reg; + int timeout = AI_MAILBOX_POLL_ACK_TIMEDOUT; + u8 reg; - reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_CONTROL)); - while (!(reg & mask)) { - if (timeout <= 0) { - pr_err("Doesn't get ack from pf.\n"); - r = -ETIME; - break; - } + do { + reg = RREG8(AI_MAIBOX_CONTROL_TRN_OFFSET_BYTE); + if (reg & 2) + return 0; + mdelay(5); timeout -= 5; + } while (timeout > 1); - reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_CONTROL)); - } + pr_err("Doesn't get TRN_MSG_ACK from pf in %d msec\n", AI_MAILBOX_POLL_ACK_TIMEDOUT); - return r; + return -ETIME; } static int xgpu_ai_poll_msg(struct amdgpu_device *adev, enum idh_event event) { - int r = 0, timeout = AI_MAILBOX_TIMEDOUT; + int r, timeout = AI_MAILBOX_POLL_MSG_TIMEDOUT; - r = xgpu_ai_mailbox_rcv_msg(adev, event); - while (r) { - if (timeout <= 0) { - pr_err("Doesn't get msg:%d from pf.\n", event); - r = -ETIME; - break; - } - mdelay(5); - timeout -= 5; - + do { r = xgpu_ai_mailbox_rcv_msg(adev, event); - } + if (!r) + return 0; - return r; + msleep(10); + timeout -= 10; + } while (timeout > 1); + + pr_err("Doesn't get msg:%d from pf, error=%d\n", event, r); + + return -ETIME; } static void xgpu_ai_mailbox_trans_msg (struct amdgpu_device *adev, enum idh_request req, u32 data1, u32 data2, u32 data3) { u32 reg; int r; + uint8_t trn; + + /* IMPORTANT: + * clear TRN_MSG_VALID valid to clear host's RCV_MSG_ACK + * and with host's RCV_MSG_ACK cleared hw automatically clear host's RCV_MSG_ACK + * which lead to VF's TRN_MSG_ACK cleared, otherwise below xgpu_ai_poll_ack() + * will return immediatly + */ + do { + xgpu_ai_mailbox_set_valid(adev, false); + trn = xgpu_ai_peek_ack(adev); + if (trn) { + pr_err("trn=%x ACK should not assert! wait again !\n", trn); + msleep(1); + } + } while(trn); reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW0)); @@ -245,15 +236,36 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work) { struct amdgpu_virt *virt = container_of(work, struct amdgpu_virt, flr_work); struct amdgpu_device *adev = container_of(virt, struct amdgpu_device, virt); + int timeout = AI_MAILBOX_POLL_FLR_TIMEDOUT; + int locked; - /* wait until RCV_MSG become 3 */ - if (xgpu_ai_poll_msg(adev, IDH_FLR_NOTIFICATION_CMPL)) { - pr_err("failed to recieve FLR_CMPL\n"); - return; - } + /* block amdgpu_gpu_recover till msg FLR COMPLETE received, + * otherwise the mailbox msg will be ruined/reseted by + * the VF FLR. + * + * we can unlock the lock_reset to allow "amdgpu_job_timedout" + * to run gpu_recover() after FLR_NOTIFICATION_CMPL received + * which means host side had finished this VF's FLR. + */ + locked = mutex_trylock(&adev->lock_reset); + if (locked) + adev->in_gpu_reset = 1; - /* Trigger recovery due to world switch failure */ - amdgpu_device_gpu_recover(adev, NULL, false); + do { + if (xgpu_ai_mailbox_peek_msg(adev) == IDH_FLR_NOTIFICATION_CMPL) + goto flr_done; + + msleep(10); + timeout -= 10; + } while (timeout > 1); + +flr_done: + if (locked) + mutex_unlock(&adev->lock_reset); + + /* Trigger recovery for world switch failure if no TDR */ + if (amdgpu_lockup_timeout == 0) + amdgpu_device_gpu_recover(adev, NULL, true); } static int xgpu_ai_set_mailbox_rcv_irq(struct amdgpu_device *adev, @@ -274,24 +286,22 @@ static int xgpu_ai_mailbox_rcv_irq(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { - int r; + enum idh_event event = xgpu_ai_mailbox_peek_msg(adev); - /* trigger gpu-reset by hypervisor only if TDR disbaled */ - if (!amdgpu_gpu_recovery) { - /* see what event we get */ - r = xgpu_ai_mailbox_rcv_msg(adev, IDH_FLR_NOTIFICATION); - - /* sometimes the interrupt is delayed to inject to VM, so under such case - * the IDH_FLR_NOTIFICATION is overwritten by VF FLR from GIM side, thus - * above recieve message could be failed, we should schedule the flr_work - * anyway + switch (event) { + case IDH_FLR_NOTIFICATION: + if (amdgpu_sriov_runtime(adev)) + schedule_work(&adev->virt.flr_work); + break; + /* READY_TO_ACCESS_GPU is fetched by kernel polling, IRQ can ignore + * it byfar since that polling thread will handle it, + * other msg like flr complete is not handled here. */ - if (r) { - DRM_ERROR("FLR_NOTIFICATION is missed\n"); - xgpu_ai_mailbox_send_ack(adev); - } - - schedule_work(&adev->virt.flr_work); + case IDH_CLR_MSG_BUF: + case IDH_FLR_NOTIFICATION_CMPL: + case IDH_READY_TO_ACCESS_GPU: + default: + break; } return 0; @@ -319,11 +329,11 @@ int xgpu_ai_mailbox_add_irq_id(struct amdgpu_device *adev) { int r; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_BIF, 135, &adev->virt.rcv_irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_BIF, 135, &adev->virt.rcv_irq); if (r) return r; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_BIF, 138, &adev->virt.ack_irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_BIF, 138, &adev->virt.ack_irq); if (r) { amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0); return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h index 67e7857..b4a9cee 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h
@@ -24,7 +24,9 @@ #ifndef __MXGPU_AI_H__ #define __MXGPU_AI_H__ -#define AI_MAILBOX_TIMEDOUT 12000 +#define AI_MAILBOX_POLL_ACK_TIMEDOUT 500 +#define AI_MAILBOX_POLL_MSG_TIMEDOUT 12000 +#define AI_MAILBOX_POLL_FLR_TIMEDOUT 500 enum idh_request { IDH_REQ_GPU_INIT_ACCESS = 1, @@ -51,4 +53,7 @@ int xgpu_ai_mailbox_add_irq_id(struct amdgpu_device *adev); int xgpu_ai_mailbox_get_irq(struct amdgpu_device *adev); void xgpu_ai_mailbox_put_irq(struct amdgpu_device *adev); +#define AI_MAIBOX_CONTROL_TRN_OFFSET_BYTE SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_CONTROL) * 4 +#define AI_MAIBOX_CONTROL_RCV_OFFSET_BYTE SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_CONTROL) * 4 + 1 + #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c index d4da663d..6f9c5497 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c
@@ -53,9 +53,16 @@ static void nbio_v6_1_mc_access_enable(struct amdgpu_device *adev, bool enable) WREG32_SOC15(NBIO, 0, mmBIF_FB_EN, 0); } -static void nbio_v6_1_hdp_flush(struct amdgpu_device *adev) +static void nbio_v6_1_hdp_flush(struct amdgpu_device *adev, + struct amdgpu_ring *ring) { - WREG32_SOC15_NO_KIQ(NBIO, 0, mmBIF_BX_PF0_HDP_MEM_COHERENCY_FLUSH_CNTL, 0); + if (!ring || !ring->funcs->emit_wreg) + WREG32_SOC15_NO_KIQ(NBIO, 0, + mmBIF_BX_PF0_HDP_MEM_COHERENCY_FLUSH_CNTL, + 0); + else + amdgpu_ring_emit_wreg(ring, SOC15_REG_OFFSET( + NBIO, 0, mmBIF_BX_PF0_HDP_MEM_COHERENCY_FLUSH_CNTL), 0); } static u32 nbio_v6_1_get_memsize(struct amdgpu_device *adev) @@ -126,7 +133,7 @@ static void nbio_v6_1_ih_control(struct amdgpu_device *adev) u32 interrupt_cntl; /* setup interrupt control */ - WREG32_SOC15(NBIO, 0, mmINTERRUPT_CNTL2, adev->dummy_page.addr >> 8); + WREG32_SOC15(NBIO, 0, mmINTERRUPT_CNTL2, adev->dummy_page_addr >> 8); interrupt_cntl = RREG32_SOC15(NBIO, 0, mmINTERRUPT_CNTL); /* INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=0 - dummy read disabled with msi, enabled without msi * INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=1 - dummy read controlled by IH_DUMMY_RD_EN @@ -213,12 +220,12 @@ static u32 nbio_v6_1_get_hdp_flush_done_offset(struct amdgpu_device *adev) static u32 nbio_v6_1_get_pcie_index_offset(struct amdgpu_device *adev) { - return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_INDEX); + return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_INDEX2); } static u32 nbio_v6_1_get_pcie_data_offset(struct amdgpu_device *adev) { - return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA); + return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA2); } static const struct nbio_hdp_flush_reg nbio_v6_1_hdp_flush_reg = {
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c index 17a9131..df34dc7 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
@@ -53,9 +53,14 @@ static void nbio_v7_0_mc_access_enable(struct amdgpu_device *adev, bool enable) WREG32_SOC15(NBIO, 0, mmBIF_FB_EN, 0); } -static void nbio_v7_0_hdp_flush(struct amdgpu_device *adev) +static void nbio_v7_0_hdp_flush(struct amdgpu_device *adev, + struct amdgpu_ring *ring) { - WREG32_SOC15_NO_KIQ(NBIO, 0, mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0); + if (!ring || !ring->funcs->emit_wreg) + WREG32_SOC15_NO_KIQ(NBIO, 0, mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0); + else + amdgpu_ring_emit_wreg(ring, SOC15_REG_OFFSET( + NBIO, 0, mmHDP_MEM_COHERENCY_FLUSH_CNTL), 0); } static u32 nbio_v7_0_get_memsize(struct amdgpu_device *adev) @@ -203,7 +208,7 @@ static void nbio_v7_0_ih_control(struct amdgpu_device *adev) u32 interrupt_cntl; /* setup interrupt control */ - WREG32_SOC15(NBIO, 0, mmINTERRUPT_CNTL2, adev->dummy_page.addr >> 8); + WREG32_SOC15(NBIO, 0, mmINTERRUPT_CNTL2, adev->dummy_page_addr >> 8); interrupt_cntl = RREG32_SOC15(NBIO, 0, mmINTERRUPT_CNTL); /* INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=0 - dummy read disabled with msi, enabled without msi * INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=1 - dummy read controlled by IH_DUMMY_RD_EN
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c index f8f112d..5af4dab 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
@@ -90,7 +90,7 @@ psp_v10_0_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type * return 0; } -int psp_v10_0_init_microcode(struct psp_context *psp) +static int psp_v10_0_init_microcode(struct psp_context *psp) { struct amdgpu_device *adev = psp->adev; const char *chip_name; @@ -136,7 +136,8 @@ int psp_v10_0_init_microcode(struct psp_context *psp) return err; } -int psp_v10_0_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd_resp *cmd) +static int psp_v10_0_prep_cmd_buf(struct amdgpu_firmware_info *ucode, + struct psp_gfx_cmd_resp *cmd) { int ret; uint64_t fw_mem_mc_addr = ucode->mc_addr; @@ -155,7 +156,8 @@ int psp_v10_0_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cm return ret; } -int psp_v10_0_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) +static int psp_v10_0_ring_init(struct psp_context *psp, + enum psp_ring_type ring_type) { int ret = 0; struct psp_ring *ring; @@ -180,7 +182,8 @@ int psp_v10_0_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) return 0; } -int psp_v10_0_ring_create(struct psp_context *psp, enum psp_ring_type ring_type) +static int psp_v10_0_ring_create(struct psp_context *psp, + enum psp_ring_type ring_type) { int ret = 0; unsigned int psp_ring_reg = 0; @@ -211,7 +214,8 @@ int psp_v10_0_ring_create(struct psp_context *psp, enum psp_ring_type ring_type) return ret; } -int psp_v10_0_ring_stop(struct psp_context *psp, enum psp_ring_type ring_type) +static int psp_v10_0_ring_stop(struct psp_context *psp, + enum psp_ring_type ring_type) { int ret = 0; struct psp_ring *ring; @@ -234,7 +238,8 @@ int psp_v10_0_ring_stop(struct psp_context *psp, enum psp_ring_type ring_type) return ret; } -int psp_v10_0_ring_destroy(struct psp_context *psp, enum psp_ring_type ring_type) +static int psp_v10_0_ring_destroy(struct psp_context *psp, + enum psp_ring_type ring_type) { int ret = 0; struct psp_ring *ring = &psp->km_ring; @@ -251,10 +256,10 @@ int psp_v10_0_ring_destroy(struct psp_context *psp, enum psp_ring_type ring_type return ret; } -int psp_v10_0_cmd_submit(struct psp_context *psp, - struct amdgpu_firmware_info *ucode, - uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, - int index) +static int psp_v10_0_cmd_submit(struct psp_context *psp, + struct amdgpu_firmware_info *ucode, + uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, + int index) { unsigned int psp_write_ptr_reg = 0; struct psp_gfx_rb_frame * write_frame = psp->km_ring.ring_mem; @@ -301,9 +306,9 @@ int psp_v10_0_cmd_submit(struct psp_context *psp, static int psp_v10_0_sram_map(struct amdgpu_device *adev, - unsigned int *sram_offset, unsigned int *sram_addr_reg_offset, - unsigned int *sram_data_reg_offset, - enum AMDGPU_UCODE_ID ucode_id) + unsigned int *sram_offset, unsigned int *sram_addr_reg_offset, + unsigned int *sram_data_reg_offset, + enum AMDGPU_UCODE_ID ucode_id) { int ret = 0; @@ -386,9 +391,9 @@ psp_v10_0_sram_map(struct amdgpu_device *adev, return ret; } -bool psp_v10_0_compare_sram_data(struct psp_context *psp, - struct amdgpu_firmware_info *ucode, - enum AMDGPU_UCODE_ID ucode_type) +static bool psp_v10_0_compare_sram_data(struct psp_context *psp, + struct amdgpu_firmware_info *ucode, + enum AMDGPU_UCODE_ID ucode_type) { int err = 0; unsigned int fw_sram_reg_val = 0; @@ -422,8 +427,25 @@ bool psp_v10_0_compare_sram_data(struct psp_context *psp, } -int psp_v10_0_mode1_reset(struct psp_context *psp) +static int psp_v10_0_mode1_reset(struct psp_context *psp) { DRM_INFO("psp mode 1 reset not supported now! \n"); return -EINVAL; } + +static const struct psp_funcs psp_v10_0_funcs = { + .init_microcode = psp_v10_0_init_microcode, + .prep_cmd_buf = psp_v10_0_prep_cmd_buf, + .ring_init = psp_v10_0_ring_init, + .ring_create = psp_v10_0_ring_create, + .ring_stop = psp_v10_0_ring_stop, + .ring_destroy = psp_v10_0_ring_destroy, + .cmd_submit = psp_v10_0_cmd_submit, + .compare_sram_data = psp_v10_0_compare_sram_data, + .mode1_reset = psp_v10_0_mode1_reset, +}; + +void psp_v10_0_set_psp_funcs(struct psp_context *psp) +{ + psp->funcs = &psp_v10_0_funcs; +}
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.h b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.h index 451e830..20c2a948 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.h +++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.h
@@ -27,24 +27,6 @@ #include "amdgpu_psp.h" -extern int psp_v10_0_init_microcode(struct psp_context *psp); -extern int psp_v10_0_prep_cmd_buf(struct amdgpu_firmware_info *ucode, - struct psp_gfx_cmd_resp *cmd); -extern int psp_v10_0_ring_init(struct psp_context *psp, - enum psp_ring_type ring_type); -extern int psp_v10_0_ring_create(struct psp_context *psp, - enum psp_ring_type ring_type); -extern int psp_v10_0_ring_stop(struct psp_context *psp, - enum psp_ring_type ring_type); -extern int psp_v10_0_ring_destroy(struct psp_context *psp, - enum psp_ring_type ring_type); -extern int psp_v10_0_cmd_submit(struct psp_context *psp, - struct amdgpu_firmware_info *ucode, - uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, - int index); -extern bool psp_v10_0_compare_sram_data(struct psp_context *psp, - struct amdgpu_firmware_info *ucode, - enum AMDGPU_UCODE_ID ucode_type); +void psp_v10_0_set_psp_funcs(struct psp_context *psp); -extern int psp_v10_0_mode1_reset(struct psp_context *psp); #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c index 19bd193..196e75d 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
@@ -39,6 +39,8 @@ MODULE_FIRMWARE("amdgpu/vega10_sos.bin"); MODULE_FIRMWARE("amdgpu/vega10_asd.bin"); +MODULE_FIRMWARE("amdgpu/vega12_sos.bin"); +MODULE_FIRMWARE("amdgpu/vega12_asd.bin"); #define smnMP1_FIRMWARE_FLAGS 0x3010028 @@ -93,7 +95,7 @@ psp_v3_1_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type *t return 0; } -int psp_v3_1_init_microcode(struct psp_context *psp) +static int psp_v3_1_init_microcode(struct psp_context *psp) { struct amdgpu_device *adev = psp->adev; const char *chip_name; @@ -107,6 +109,9 @@ int psp_v3_1_init_microcode(struct psp_context *psp) case CHIP_VEGA10: chip_name = "vega10"; break; + case CHIP_VEGA12: + chip_name = "vega12"; + break; default: BUG(); } @@ -161,7 +166,7 @@ int psp_v3_1_init_microcode(struct psp_context *psp) return err; } -int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp) +static int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp) { int ret; uint32_t psp_gfxdrv_command_reg = 0; @@ -202,7 +207,7 @@ int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp) return ret; } -int psp_v3_1_bootloader_load_sos(struct psp_context *psp) +static int psp_v3_1_bootloader_load_sos(struct psp_context *psp) { int ret; unsigned int psp_gfxdrv_command_reg = 0; @@ -243,7 +248,8 @@ int psp_v3_1_bootloader_load_sos(struct psp_context *psp) return ret; } -int psp_v3_1_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd_resp *cmd) +static int psp_v3_1_prep_cmd_buf(struct amdgpu_firmware_info *ucode, + struct psp_gfx_cmd_resp *cmd) { int ret; uint64_t fw_mem_mc_addr = ucode->mc_addr; @@ -262,7 +268,8 @@ int psp_v3_1_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd return ret; } -int psp_v3_1_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) +static int psp_v3_1_ring_init(struct psp_context *psp, + enum psp_ring_type ring_type) { int ret = 0; struct psp_ring *ring; @@ -287,7 +294,8 @@ int psp_v3_1_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) return 0; } -int psp_v3_1_ring_create(struct psp_context *psp, enum psp_ring_type ring_type) +static int psp_v3_1_ring_create(struct psp_context *psp, + enum psp_ring_type ring_type) { int ret = 0; unsigned int psp_ring_reg = 0; @@ -318,7 +326,8 @@ int psp_v3_1_ring_create(struct psp_context *psp, enum psp_ring_type ring_type) return ret; } -int psp_v3_1_ring_stop(struct psp_context *psp, enum psp_ring_type ring_type) +static int psp_v3_1_ring_stop(struct psp_context *psp, + enum psp_ring_type ring_type) { int ret = 0; struct psp_ring *ring; @@ -341,7 +350,8 @@ int psp_v3_1_ring_stop(struct psp_context *psp, enum psp_ring_type ring_type) return ret; } -int psp_v3_1_ring_destroy(struct psp_context *psp, enum psp_ring_type ring_type) +static int psp_v3_1_ring_destroy(struct psp_context *psp, + enum psp_ring_type ring_type) { int ret = 0; struct psp_ring *ring = &psp->km_ring; @@ -358,10 +368,10 @@ int psp_v3_1_ring_destroy(struct psp_context *psp, enum psp_ring_type ring_type) return ret; } -int psp_v3_1_cmd_submit(struct psp_context *psp, - struct amdgpu_firmware_info *ucode, - uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, - int index) +static int psp_v3_1_cmd_submit(struct psp_context *psp, + struct amdgpu_firmware_info *ucode, + uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, + int index) { unsigned int psp_write_ptr_reg = 0; struct psp_gfx_rb_frame * write_frame = psp->km_ring.ring_mem; @@ -410,9 +420,9 @@ int psp_v3_1_cmd_submit(struct psp_context *psp, static int psp_v3_1_sram_map(struct amdgpu_device *adev, - unsigned int *sram_offset, unsigned int *sram_addr_reg_offset, - unsigned int *sram_data_reg_offset, - enum AMDGPU_UCODE_ID ucode_id) + unsigned int *sram_offset, unsigned int *sram_addr_reg_offset, + unsigned int *sram_data_reg_offset, + enum AMDGPU_UCODE_ID ucode_id) { int ret = 0; @@ -495,9 +505,9 @@ psp_v3_1_sram_map(struct amdgpu_device *adev, return ret; } -bool psp_v3_1_compare_sram_data(struct psp_context *psp, - struct amdgpu_firmware_info *ucode, - enum AMDGPU_UCODE_ID ucode_type) +static bool psp_v3_1_compare_sram_data(struct psp_context *psp, + struct amdgpu_firmware_info *ucode, + enum AMDGPU_UCODE_ID ucode_type) { int err = 0; unsigned int fw_sram_reg_val = 0; @@ -530,7 +540,7 @@ bool psp_v3_1_compare_sram_data(struct psp_context *psp, return true; } -bool psp_v3_1_smu_reload_quirk(struct psp_context *psp) +static bool psp_v3_1_smu_reload_quirk(struct psp_context *psp) { struct amdgpu_device *adev = psp->adev; uint32_t reg; @@ -541,7 +551,7 @@ bool psp_v3_1_smu_reload_quirk(struct psp_context *psp) return (reg & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) ? true : false; } -int psp_v3_1_mode1_reset(struct psp_context *psp) +static int psp_v3_1_mode1_reset(struct psp_context *psp) { int ret; uint32_t offset; @@ -574,3 +584,23 @@ int psp_v3_1_mode1_reset(struct psp_context *psp) return 0; } + +static const struct psp_funcs psp_v3_1_funcs = { + .init_microcode = psp_v3_1_init_microcode, + .bootloader_load_sysdrv = psp_v3_1_bootloader_load_sysdrv, + .bootloader_load_sos = psp_v3_1_bootloader_load_sos, + .prep_cmd_buf = psp_v3_1_prep_cmd_buf, + .ring_init = psp_v3_1_ring_init, + .ring_create = psp_v3_1_ring_create, + .ring_stop = psp_v3_1_ring_stop, + .ring_destroy = psp_v3_1_ring_destroy, + .cmd_submit = psp_v3_1_cmd_submit, + .compare_sram_data = psp_v3_1_compare_sram_data, + .smu_reload_quirk = psp_v3_1_smu_reload_quirk, + .mode1_reset = psp_v3_1_mode1_reset, +}; + +void psp_v3_1_set_psp_funcs(struct psp_context *psp) +{ + psp->funcs = &psp_v3_1_funcs; +}
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.h b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.h index b05dbad..e411e31b 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.h +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.h
@@ -32,26 +32,6 @@ enum { PSP_BINARY_ALIGNMENT = 64 }; enum { PSP_BOOTLOADER_1_MEG_ALIGNMENT = 0x100000 }; enum { PSP_BOOTLOADER_8_MEM_ALIGNMENT = 0x800000 }; -extern int psp_v3_1_init_microcode(struct psp_context *psp); -extern int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp); -extern int psp_v3_1_bootloader_load_sos(struct psp_context *psp); -extern int psp_v3_1_prep_cmd_buf(struct amdgpu_firmware_info *ucode, - struct psp_gfx_cmd_resp *cmd); -extern int psp_v3_1_ring_init(struct psp_context *psp, - enum psp_ring_type ring_type); -extern int psp_v3_1_ring_create(struct psp_context *psp, - enum psp_ring_type ring_type); -extern int psp_v3_1_ring_stop(struct psp_context *psp, - enum psp_ring_type ring_type); -extern int psp_v3_1_ring_destroy(struct psp_context *psp, - enum psp_ring_type ring_type); -extern int psp_v3_1_cmd_submit(struct psp_context *psp, - struct amdgpu_firmware_info *ucode, - uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, - int index); -extern bool psp_v3_1_compare_sram_data(struct psp_context *psp, - struct amdgpu_firmware_info *ucode, - enum AMDGPU_UCODE_ID ucode_type); -extern bool psp_v3_1_smu_reload_quirk(struct psp_context *psp); -extern int psp_v3_1_mode1_reset(struct psp_context *psp); +void psp_v3_1_set_psp_funcs(struct psp_context *psp); + #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index bd844ed..c7190c3 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
@@ -289,13 +289,6 @@ static void sdma_v2_4_ring_emit_hdp_flush(struct amdgpu_ring *ring) SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ } -static void sdma_v2_4_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, mmHDP_DEBUG0); - amdgpu_ring_write(ring, 1); -} /** * sdma_v2_4_ring_emit_fence - emit a fence on the DMA ring * @@ -346,7 +339,7 @@ static void sdma_v2_4_gfx_stop(struct amdgpu_device *adev) if ((adev->mman.buffer_funcs_ring == sdma0) || (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, false); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32(mmSDMA0_GFX_RB_CNTL + sdma_offsets[i]); @@ -491,7 +484,7 @@ static int sdma_v2_4_gfx_resume(struct amdgpu_device *adev) } if (adev->mman.buffer_funcs_ring == ring) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.real_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, true); } return 0; @@ -861,20 +854,7 @@ static void sdma_v2_4_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void sdma_v2_4_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - if (vmid < 8) { - amdgpu_ring_write(ring, (mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid)); - } else { - amdgpu_ring_write(ring, (mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8)); - } - amdgpu_ring_write(ring, pd_addr >> 12); - - /* flush TLB */ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); /* wait for flush */ amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | @@ -888,6 +868,15 @@ static void sdma_v2_4_ring_emit_vm_flush(struct amdgpu_ring *ring, SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ } +static void sdma_v2_4_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | + SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); + amdgpu_ring_write(ring, reg); + amdgpu_ring_write(ring, val); +} + static int sdma_v2_4_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -1203,9 +1192,9 @@ static const struct amdgpu_ring_funcs sdma_v2_4_ring_funcs = { .set_wptr = sdma_v2_4_ring_set_wptr, .emit_frame_size = 6 + /* sdma_v2_4_ring_emit_hdp_flush */ - 3 + /* sdma_v2_4_ring_emit_hdp_invalidate */ + 3 + /* hdp invalidate */ 6 + /* sdma_v2_4_ring_emit_pipeline_sync */ - 12 + /* sdma_v2_4_ring_emit_vm_flush */ + VI_FLUSH_GPU_TLB_NUM_WREG * 3 + 6 + /* sdma_v2_4_ring_emit_vm_flush */ 10 + 10 + 10, /* sdma_v2_4_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 7 + 6, /* sdma_v2_4_ring_emit_ib */ .emit_ib = sdma_v2_4_ring_emit_ib, @@ -1213,11 +1202,11 @@ static const struct amdgpu_ring_funcs sdma_v2_4_ring_funcs = { .emit_pipeline_sync = sdma_v2_4_ring_emit_pipeline_sync, .emit_vm_flush = sdma_v2_4_ring_emit_vm_flush, .emit_hdp_flush = sdma_v2_4_ring_emit_hdp_flush, - .emit_hdp_invalidate = sdma_v2_4_ring_emit_hdp_invalidate, .test_ring = sdma_v2_4_ring_test_ring, .test_ib = sdma_v2_4_ring_test_ib, .insert_nop = sdma_v2_4_ring_insert_nop, .pad_ib = sdma_v2_4_ring_pad_ib, + .emit_wreg = sdma_v2_4_ring_emit_wreg, }; static void sdma_v2_4_set_ring_funcs(struct amdgpu_device *adev) @@ -1316,9 +1305,6 @@ static const struct amdgpu_vm_pte_funcs sdma_v2_4_vm_pte_funcs = { .copy_pte = sdma_v2_4_vm_copy_pte, .write_pte = sdma_v2_4_vm_write_pte, - - .set_max_nums_pte_pde = 0x1fffff >> 3, - .set_pte_pde_num_dw = 10, .set_pte_pde = sdma_v2_4_vm_set_pte_pde, };
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index 1fa5911..8e8f70b 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
@@ -460,14 +460,6 @@ static void sdma_v3_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ } -static void sdma_v3_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, mmHDP_DEBUG0); - amdgpu_ring_write(ring, 1); -} - /** * sdma_v3_0_ring_emit_fence - emit a fence on the DMA ring * @@ -518,7 +510,7 @@ static void sdma_v3_0_gfx_stop(struct amdgpu_device *adev) if ((adev->mman.buffer_funcs_ring == sdma0) || (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, false); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32(mmSDMA0_GFX_RB_CNTL + sdma_offsets[i]); @@ -761,7 +753,7 @@ static int sdma_v3_0_gfx_resume(struct amdgpu_device *adev) } if (adev->mman.buffer_funcs_ring == ring) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.real_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, true); } return 0; @@ -1130,20 +1122,7 @@ static void sdma_v3_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void sdma_v3_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - if (vmid < 8) { - amdgpu_ring_write(ring, (mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid)); - } else { - amdgpu_ring_write(ring, (mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8)); - } - amdgpu_ring_write(ring, pd_addr >> 12); - - /* flush TLB */ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); /* wait for flush */ amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | @@ -1157,6 +1136,15 @@ static void sdma_v3_0_ring_emit_vm_flush(struct amdgpu_ring *ring, SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ } +static void sdma_v3_0_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | + SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); + amdgpu_ring_write(ring, reg); + amdgpu_ring_write(ring, val); +} + static int sdma_v3_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -1640,7 +1628,7 @@ static const struct amdgpu_ring_funcs sdma_v3_0_ring_funcs = { .set_wptr = sdma_v3_0_ring_set_wptr, .emit_frame_size = 6 + /* sdma_v3_0_ring_emit_hdp_flush */ - 3 + /* sdma_v3_0_ring_emit_hdp_invalidate */ + 3 + /* hdp invalidate */ 6 + /* sdma_v3_0_ring_emit_pipeline_sync */ 3 /* VI_FLUSH_GPU_TLB_NUM_WREG */ * 3 + 6 + /* sdma_v3_0_ring_emit_vm_flush */ 10 + 10 + 10, /* sdma_v3_0_ring_emit_fence x3 for user fence, vm fence */ @@ -1650,11 +1638,11 @@ static const struct amdgpu_ring_funcs sdma_v3_0_ring_funcs = { .emit_pipeline_sync = sdma_v3_0_ring_emit_pipeline_sync, .emit_vm_flush = sdma_v3_0_ring_emit_vm_flush, .emit_hdp_flush = sdma_v3_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = sdma_v3_0_ring_emit_hdp_invalidate, .test_ring = sdma_v3_0_ring_test_ring, .test_ib = sdma_v3_0_ring_test_ib, .insert_nop = sdma_v3_0_ring_insert_nop, .pad_ib = sdma_v3_0_ring_pad_ib, + .emit_wreg = sdma_v3_0_ring_emit_wreg, }; static void sdma_v3_0_set_ring_funcs(struct amdgpu_device *adev) @@ -1753,10 +1741,6 @@ static const struct amdgpu_vm_pte_funcs sdma_v3_0_vm_pte_funcs = { .copy_pte = sdma_v3_0_vm_copy_pte, .write_pte = sdma_v3_0_vm_write_pte, - - /* not 0x3fffff due to HW limitation */ - .set_max_nums_pte_pde = 0x3fffe0 >> 3, - .set_pte_pde_num_dw = 10, .set_pte_pde = sdma_v3_0_vm_set_pte_pde, };
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 036798b..399f876 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -31,8 +31,6 @@ #include "sdma0/sdma0_4_0_sh_mask.h" #include "sdma1/sdma1_4_0_offset.h" #include "sdma1/sdma1_4_0_sh_mask.h" -#include "mmhub/mmhub_1_0_offset.h" -#include "mmhub/mmhub_1_0_sh_mask.h" #include "hdp/hdp_4_0_offset.h" #include "sdma0/sdma0_4_1_default.h" @@ -42,6 +40,8 @@ MODULE_FIRMWARE("amdgpu/vega10_sdma.bin"); MODULE_FIRMWARE("amdgpu/vega10_sdma1.bin"); +MODULE_FIRMWARE("amdgpu/vega12_sdma.bin"); +MODULE_FIRMWARE("amdgpu/vega12_sdma1.bin"); MODULE_FIRMWARE("amdgpu/raven_sdma.bin"); #define SDMA0_POWER_CNTL__ON_OFF_CONDITION_HOLD_TIME_MASK 0x000000F8L @@ -86,6 +86,13 @@ static const struct soc15_reg_golden golden_settings_sdma_vg10[] = { SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG_READ, 0x0018773f, 0x00104002) }; +static const struct soc15_reg_golden golden_settings_sdma_vg12[] = { + SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG, 0x0018773f, 0x00104001), + SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG_READ, 0x0018773f, 0x00104001), + SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG, 0x0018773f, 0x00104001), + SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG_READ, 0x0018773f, 0x00104001) +}; + static const struct soc15_reg_golden golden_settings_sdma_4_1[] = { SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CHICKEN_BITS, 0xfe931f07, 0x02831d07), @@ -124,6 +131,14 @@ static void sdma_v4_0_init_golden_registers(struct amdgpu_device *adev) golden_settings_sdma_vg10, ARRAY_SIZE(golden_settings_sdma_vg10)); break; + case CHIP_VEGA12: + soc15_program_register_sequence(adev, + golden_settings_sdma_4, + ARRAY_SIZE(golden_settings_sdma_4)); + soc15_program_register_sequence(adev, + golden_settings_sdma_vg12, + ARRAY_SIZE(golden_settings_sdma_vg12)); + break; case CHIP_RAVEN: soc15_program_register_sequence(adev, golden_settings_sdma_4_1, @@ -164,6 +179,9 @@ static int sdma_v4_0_init_microcode(struct amdgpu_device *adev) case CHIP_VEGA10: chip_name = "vega10"; break; + case CHIP_VEGA12: + chip_name = "vega12"; + break; case CHIP_RAVEN: chip_name = "raven"; break; @@ -371,16 +389,6 @@ static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ } -static void sdma_v4_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - struct amdgpu_device *adev = ring->adev; - - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, SOC15_REG_OFFSET(HDP, 0, mmHDP_READ_CACHE_INVALIDATE)); - amdgpu_ring_write(ring, 1); -} - /** * sdma_v4_0_ring_emit_fence - emit a fence on the DMA ring * @@ -436,7 +444,7 @@ static void sdma_v4_0_gfx_stop(struct amdgpu_device *adev) if ((adev->mman.buffer_funcs_ring == sdma0) || (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, false); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); @@ -678,7 +686,7 @@ static int sdma_v4_0_gfx_resume(struct amdgpu_device *adev) } if (adev->mman.buffer_funcs_ring == ring) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.real_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, true); } @@ -1131,38 +1139,28 @@ static void sdma_v4_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void sdma_v4_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; - uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vmid); - uint64_t flags = AMDGPU_PTE_VALID; - unsigned eng = ring->vm_inv_eng; + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); +} - amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags); - pd_addr |= flags; - +static void sdma_v4_0_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->ctx0_ptb_addr_lo32 + vmid * 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + amdgpu_ring_write(ring, reg); + amdgpu_ring_write(ring, val); +} - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->ctx0_ptb_addr_hi32 + vmid * 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->vm_inv_eng0_req + eng); - amdgpu_ring_write(ring, req); - - /* wait for flush */ +static void sdma_v4_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg, + uint32_t val, uint32_t mask) +{ amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) | SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* equal */ - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); + amdgpu_ring_write(ring, reg << 2); amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 1 << vmid); /* reference */ - amdgpu_ring_write(ring, 1 << vmid); /* mask */ + amdgpu_ring_write(ring, val); /* reference */ + amdgpu_ring_write(ring, mask); /* mask */ amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); } @@ -1192,13 +1190,13 @@ static int sdma_v4_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* SDMA trap event */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_SDMA0, 224, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA0, 224, &adev->sdma.trap_irq); if (r) return r; /* SDMA trap event */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_SDMA1, 224, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA1, 224, &adev->sdma.trap_irq); if (r) return r; @@ -1353,7 +1351,7 @@ static int sdma_v4_0_process_trap_irq(struct amdgpu_device *adev, { DRM_DEBUG("IH: SDMA trap\n"); switch (entry->client_id) { - case AMDGPU_IH_CLIENTID_SDMA0: + case SOC15_IH_CLIENTID_SDMA0: switch (entry->ring_id) { case 0: amdgpu_fence_process(&adev->sdma.instance[0].ring); @@ -1369,7 +1367,7 @@ static int sdma_v4_0_process_trap_irq(struct amdgpu_device *adev, break; } break; - case AMDGPU_IH_CLIENTID_SDMA1: + case SOC15_IH_CLIENTID_SDMA1: switch (entry->ring_id) { case 0: amdgpu_fence_process(&adev->sdma.instance[1].ring); @@ -1419,7 +1417,7 @@ static void sdma_v4_0_update_medium_grain_clock_gating( if (def != data) WREG32(SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_CLK_CTRL), data); - if (adev->asic_type == CHIP_VEGA10) { + if (adev->sdma.num_instances > 1) { def = data = RREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_CLK_CTRL)); data &= ~(SDMA1_CLK_CTRL__SOFT_OVERRIDE7_MASK | SDMA1_CLK_CTRL__SOFT_OVERRIDE6_MASK | @@ -1447,7 +1445,7 @@ static void sdma_v4_0_update_medium_grain_clock_gating( if (def != data) WREG32(SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_CLK_CTRL), data); - if (adev->asic_type == CHIP_VEGA10) { + if (adev->sdma.num_instances > 1) { def = data = RREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_CLK_CTRL)); data |= (SDMA1_CLK_CTRL__SOFT_OVERRIDE7_MASK | SDMA1_CLK_CTRL__SOFT_OVERRIDE6_MASK | @@ -1478,7 +1476,7 @@ static void sdma_v4_0_update_medium_grain_light_sleep( WREG32(SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_POWER_CNTL), data); /* 1-not override: enable sdma1 mem light sleep */ - if (adev->asic_type == CHIP_VEGA10) { + if (adev->sdma.num_instances > 1) { def = data = RREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_POWER_CNTL)); data |= SDMA1_POWER_CNTL__MEM_POWER_OVERRIDE_MASK; if (def != data) @@ -1492,7 +1490,7 @@ static void sdma_v4_0_update_medium_grain_light_sleep( WREG32(SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_POWER_CNTL), data); /* 0-override:disable sdma1 mem light sleep */ - if (adev->asic_type == CHIP_VEGA10) { + if (adev->sdma.num_instances > 1) { def = data = RREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_POWER_CNTL)); data &= ~SDMA1_POWER_CNTL__MEM_POWER_OVERRIDE_MASK; if (def != data) @@ -1511,6 +1509,7 @@ static int sdma_v4_0_set_clockgating_state(void *handle, switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: case CHIP_RAVEN: sdma_v4_0_update_medium_grain_clock_gating(adev, state == AMD_CG_STATE_GATE ? true : false); @@ -1588,9 +1587,11 @@ static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs = { .set_wptr = sdma_v4_0_ring_set_wptr, .emit_frame_size = 6 + /* sdma_v4_0_ring_emit_hdp_flush */ - 3 + /* sdma_v4_0_ring_emit_hdp_invalidate */ + 3 + /* hdp invalidate */ 6 + /* sdma_v4_0_ring_emit_pipeline_sync */ - 18 + /* sdma_v4_0_ring_emit_vm_flush */ + /* sdma_v4_0_ring_emit_vm_flush */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 6 + 10 + 10 + 10, /* sdma_v4_0_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 7 + 6, /* sdma_v4_0_ring_emit_ib */ .emit_ib = sdma_v4_0_ring_emit_ib, @@ -1598,11 +1599,12 @@ static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs = { .emit_pipeline_sync = sdma_v4_0_ring_emit_pipeline_sync, .emit_vm_flush = sdma_v4_0_ring_emit_vm_flush, .emit_hdp_flush = sdma_v4_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = sdma_v4_0_ring_emit_hdp_invalidate, .test_ring = sdma_v4_0_ring_test_ring, .test_ib = sdma_v4_0_ring_test_ib, .insert_nop = sdma_v4_0_ring_insert_nop, .pad_ib = sdma_v4_0_ring_pad_ib, + .emit_wreg = sdma_v4_0_ring_emit_wreg, + .emit_reg_wait = sdma_v4_0_ring_emit_reg_wait, }; static void sdma_v4_0_set_ring_funcs(struct amdgpu_device *adev) @@ -1637,7 +1639,7 @@ static void sdma_v4_0_set_irq_funcs(struct amdgpu_device *adev) * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer * - * Copy GPU buffers using the DMA engine (VEGA10). + * Copy GPU buffers using the DMA engine (VEGA10/12). * Used by the amdgpu ttm implementation to move pages if * registered as the asic copy callback. */ @@ -1664,7 +1666,7 @@ static void sdma_v4_0_emit_copy_buffer(struct amdgpu_ib *ib, * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer * - * Fill GPU buffers using the DMA engine (VEGA10). + * Fill GPU buffers using the DMA engine (VEGA10/12). */ static void sdma_v4_0_emit_fill_buffer(struct amdgpu_ib *ib, uint32_t src_data, @@ -1701,9 +1703,6 @@ static const struct amdgpu_vm_pte_funcs sdma_v4_0_vm_pte_funcs = { .copy_pte = sdma_v4_0_vm_copy_pte, .write_pte = sdma_v4_0_vm_write_pte, - - .set_max_nums_pte_pde = 0x400000 >> 3, - .set_pte_pde_num_dw = 10, .set_pte_pde = sdma_v4_0_vm_set_pte_pde, };
diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c index 3598151..a675ec6 100644 --- a/drivers/gpu/drm/amd/amdgpu/si.c +++ b/drivers/gpu/drm/amd/amdgpu/si.c
@@ -32,7 +32,7 @@ #include "amdgpu_vce.h" #include "atom.h" #include "amd_pcie.h" -#include "amdgpu_powerplay.h" +#include "si_dpm.h" #include "sid.h" #include "si_ih.h" #include "gfx_v6_0.h" @@ -1231,6 +1231,27 @@ static void si_detect_hw_virtualization(struct amdgpu_device *adev) adev->virt.caps |= AMDGPU_PASSTHROUGH_MODE; } +static void si_flush_hdp(struct amdgpu_device *adev, struct amdgpu_ring *ring) +{ + if (!ring || !ring->funcs->emit_wreg) { + WREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 1); + RREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL); + } else { + amdgpu_ring_emit_wreg(ring, mmHDP_MEM_COHERENCY_FLUSH_CNTL, 1); + } +} + +static void si_invalidate_hdp(struct amdgpu_device *adev, + struct amdgpu_ring *ring) +{ + if (!ring || !ring->funcs->emit_wreg) { + WREG32(mmHDP_DEBUG0, 1); + RREG32(mmHDP_DEBUG0); + } else { + amdgpu_ring_emit_wreg(ring, mmHDP_DEBUG0, 1); + } +} + static int si_get_pcie_lanes(struct amdgpu_device *adev) { u32 link_width_cntl; @@ -1309,6 +1330,8 @@ static const struct amdgpu_asic_funcs si_asic_funcs = .get_pcie_lanes = &si_get_pcie_lanes, .set_pcie_lanes = &si_set_pcie_lanes, .get_config_memsize = &si_get_config_memsize, + .flush_hdp = &si_flush_hdp, + .invalidate_hdp = &si_invalidate_hdp, }; static uint32_t si_get_rev_id(struct amdgpu_device *adev) @@ -2027,7 +2050,7 @@ int si_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &si_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v6_0_ip_block); amdgpu_device_ip_block_add(adev, &si_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &si_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); else @@ -2041,7 +2064,7 @@ int si_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &si_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v6_0_ip_block); amdgpu_device_ip_block_add(adev, &si_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &si_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); else @@ -2055,7 +2078,7 @@ int si_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &si_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v6_0_ip_block); amdgpu_device_ip_block_add(adev, &si_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &si_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); amdgpu_device_ip_block_add(adev, &gfx_v6_0_ip_block);
diff --git a/drivers/gpu/drm/amd/amdgpu/si.h b/drivers/gpu/drm/amd/amdgpu/si.h index 5892250..06ed721 100644 --- a/drivers/gpu/drm/amd/amdgpu/si.h +++ b/drivers/gpu/drm/amd/amdgpu/si.h
@@ -24,6 +24,8 @@ #ifndef __SI_H__ #define __SI_H__ +#define SI_FLUSH_GPU_TLB_NUM_WREG 2 + void si_srbm_select(struct amdgpu_device *adev, u32 me, u32 pipe, u32 queue, u32 vmid); int si_set_ip_blocks(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/si_dma.c b/drivers/gpu/drm/amd/amdgpu/si_dma.c index 9a29c13..b75d901 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dma.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dma.c
@@ -24,6 +24,7 @@ #include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_trace.h" +#include "si.h" #include "sid.h" const u32 sdma_offsets[SDMA_MAX_INSTANCE] = @@ -74,20 +75,6 @@ static void si_dma_ring_emit_ib(struct amdgpu_ring *ring, } -static void si_dma_ring_emit_hdp_flush(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); - amdgpu_ring_write(ring, (0xf << 16) | (HDP_MEM_COHERENCY_FLUSH_CNTL)); - amdgpu_ring_write(ring, 1); -} - -static void si_dma_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); - amdgpu_ring_write(ring, (0xf << 16) | (HDP_DEBUG0)); - amdgpu_ring_write(ring, 1); -} - /** * si_dma_ring_emit_fence - emit a fence on the DMA ring * @@ -134,7 +121,7 @@ static void si_dma_stop(struct amdgpu_device *adev) WREG32(DMA_RB_CNTL + sdma_offsets[i], rb_cntl); if (adev->mman.buffer_funcs_ring == ring) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, false); ring->ready = false; } } @@ -197,7 +184,7 @@ static int si_dma_start(struct amdgpu_device *adev) } if (adev->mman.buffer_funcs_ring == ring) - amdgpu_ttm_set_active_vram_size(adev, adev->mc.real_vram_size); + amdgpu_ttm_set_buffer_funcs_status(adev, true); } return 0; @@ -475,17 +462,7 @@ static void si_dma_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void si_dma_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { - amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); - if (vmid < 8) - amdgpu_ring_write(ring, (0xf << 16) | (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid)); - else - amdgpu_ring_write(ring, (0xf << 16) | (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (vmid - 8))); - amdgpu_ring_write(ring, pd_addr >> 12); - - /* bits 0-7 are the VM contexts0-7 */ - amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); - amdgpu_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST)); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); /* wait for invalidate to complete */ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_POLL_REG_MEM, 0, 0, 0, 0)); @@ -496,6 +473,14 @@ static void si_dma_ring_emit_vm_flush(struct amdgpu_ring *ring, amdgpu_ring_write(ring, (0 << 28) | 0x20); /* func(always) | poll interval */ } +static void si_dma_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); + amdgpu_ring_write(ring, (0xf << 16) | reg); + amdgpu_ring_write(ring, val); +} + static int si_dma_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -772,22 +757,20 @@ static const struct amdgpu_ring_funcs si_dma_ring_funcs = { .get_wptr = si_dma_ring_get_wptr, .set_wptr = si_dma_ring_set_wptr, .emit_frame_size = - 3 + /* si_dma_ring_emit_hdp_flush */ - 3 + /* si_dma_ring_emit_hdp_invalidate */ + 3 + 3 + /* hdp flush / invalidate */ 6 + /* si_dma_ring_emit_pipeline_sync */ - 12 + /* si_dma_ring_emit_vm_flush */ + SI_FLUSH_GPU_TLB_NUM_WREG * 3 + 6 + /* si_dma_ring_emit_vm_flush */ 9 + 9 + 9, /* si_dma_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 7 + 3, /* si_dma_ring_emit_ib */ .emit_ib = si_dma_ring_emit_ib, .emit_fence = si_dma_ring_emit_fence, .emit_pipeline_sync = si_dma_ring_emit_pipeline_sync, .emit_vm_flush = si_dma_ring_emit_vm_flush, - .emit_hdp_flush = si_dma_ring_emit_hdp_flush, - .emit_hdp_invalidate = si_dma_ring_emit_hdp_invalidate, .test_ring = si_dma_ring_test_ring, .test_ib = si_dma_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = si_dma_ring_pad_ib, + .emit_wreg = si_dma_ring_emit_wreg, }; static void si_dma_set_ring_funcs(struct amdgpu_device *adev) @@ -891,9 +874,6 @@ static const struct amdgpu_vm_pte_funcs si_dma_vm_pte_funcs = { .copy_pte = si_dma_vm_copy_pte, .write_pte = si_dma_vm_write_pte, - - .set_max_nums_pte_pde = 0xffff8 >> 3, - .set_pte_pde_num_dw = 9, .set_pte_pde = si_dma_vm_set_pte_pde, };
diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c index aab94ea..f0889d2 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
@@ -67,6 +67,8 @@ MODULE_FIRMWARE("radeon/hainan_smc.bin"); MODULE_FIRMWARE("radeon/hainan_k_smc.bin"); MODULE_FIRMWARE("radeon/banks_k_2_smc.bin"); +static const struct amd_pm_funcs si_dpm_funcs; + union power_info { struct _ATOM_POWERPLAY_INFO info; struct _ATOM_POWERPLAY_INFO_V2 info_2; @@ -3065,7 +3067,7 @@ static bool si_dpm_vblank_too_short(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; u32 vblank_time = amdgpu_dpm_get_vblank_time(adev); /* we never hit the non-gddr5 limit so disable it */ - u32 switch_limit = adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5 ? 450 : 0; + u32 switch_limit = adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5 ? 450 : 0; if (vblank_time < switch_limit) return true; @@ -4328,7 +4330,7 @@ static u8 si_get_strobe_mode_settings(struct amdgpu_device *adev, u32 mclk) if (mclk <= pi->mclk_strobe_mode_threshold) strobe_mode = true; - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) result = si_get_mclk_frequency_ratio(mclk, strobe_mode); else result = si_get_ddr3_mclk_frequency_ratio(mclk); @@ -4915,7 +4917,7 @@ static int si_populate_smc_initial_state(struct amdgpu_device *adev, table->initialState.levels[0].bSP = cpu_to_be32(pi->dsp); table->initialState.levels[0].gen2PCIE = (u8)si_pi->boot_pcie_gen; - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { table->initialState.levels[0].strobeMode = si_get_strobe_mode_settings(adev, initial_state->performance_levels[0].mclk); @@ -5187,7 +5189,7 @@ static int si_init_smc_table(struct amdgpu_device *adev) if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_STEPVDDC) table->systemFlags |= PPSMC_SYSTEMFLAG_STEPVDDC; - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) table->systemFlags |= PPSMC_SYSTEMFLAG_GDDR5; if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_REVERT_GPIO5_POLARITY) @@ -5364,7 +5366,7 @@ static int si_populate_mclk_value(struct amdgpu_device *adev, mpll_ad_func_cntl &= ~YCLK_POST_DIV_MASK; mpll_ad_func_cntl |= YCLK_POST_DIV(mpll_param.post_div); - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { mpll_dq_func_cntl &= ~(YCLK_SEL_MASK | YCLK_POST_DIV_MASK); mpll_dq_func_cntl |= YCLK_SEL(mpll_param.yclk_sel) | YCLK_POST_DIV(mpll_param.post_div); @@ -5376,7 +5378,7 @@ static int si_populate_mclk_value(struct amdgpu_device *adev, u32 tmp; u32 reference_clock = adev->clock.mpll.reference_freq; - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) freq_nom = memory_clock * 4; else freq_nom = memory_clock * 2; @@ -5468,7 +5470,7 @@ static int si_convert_power_level_to_smc(struct amdgpu_device *adev, level->mcFlags |= SISLANDS_SMC_MC_PG_EN; } - if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { + if (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { if (pl->mclk > pi->mclk_edc_enable_threshold) level->mcFlags |= SISLANDS_SMC_MC_EDC_RD_FLAG; @@ -5839,12 +5841,12 @@ static int si_set_mc_special_registers(struct amdgpu_device *adev, table->mc_reg_table_entry[k].mc_data[j] = (temp_reg & 0xffff0000) | (table->mc_reg_table_entry[k].mc_data[i] & 0x0000ffff); - if (adev->mc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) + if (adev->gmc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) table->mc_reg_table_entry[k].mc_data[j] |= 0x100; } j++; - if (adev->mc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) { + if (adev->gmc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) { if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE) return -EINVAL; table->mc_reg_address[j].s1 = MC_PMG_AUTO_CMD; @@ -7913,6 +7915,8 @@ static int si_dpm_early_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; + adev->powerplay.pp_funcs = &si_dpm_funcs; + adev->powerplay.pp_handle = adev; si_dpm_set_irq_funcs(adev); return 0; } @@ -8013,7 +8017,7 @@ static int si_dpm_read_sensor(void *handle, int idx, } } -const struct amd_ip_funcs si_dpm_ip_funcs = { +static const struct amd_ip_funcs si_dpm_ip_funcs = { .name = "si_dpm", .early_init = si_dpm_early_init, .late_init = si_dpm_late_init, @@ -8030,8 +8034,16 @@ const struct amd_ip_funcs si_dpm_ip_funcs = { .set_powergating_state = si_dpm_set_powergating_state, }; -const struct amd_pm_funcs si_dpm_funcs = { - .get_temperature = &si_dpm_get_temp, +const struct amdgpu_ip_block_version si_smu_ip_block = +{ + .type = AMD_IP_BLOCK_TYPE_SMC, + .major = 6, + .minor = 0, + .rev = 0, + .funcs = &si_dpm_ip_funcs, +}; + +static const struct amd_pm_funcs si_dpm_funcs = { .pre_set_power_state = &si_dpm_pre_set_power_state, .set_power_state = &si_dpm_set_power_state, .post_set_power_state = &si_dpm_post_set_power_state,
diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.h b/drivers/gpu/drm/amd/amdgpu/si_dpm.h index 9fe343d..6b7d292b 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dpm.h +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.h
@@ -245,8 +245,7 @@ enum si_display_gap SI_PM_DISPLAY_GAP_IGNORE = 3, }; -extern const struct amd_ip_funcs si_dpm_ip_funcs; -extern const struct amd_pm_funcs si_dpm_funcs; +extern const struct amdgpu_ip_block_version si_smu_ip_block; struct ni_leakage_coeffients {
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index a04a033..51cf8a3 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -57,7 +57,6 @@ #include "uvd_v7_0.h" #include "vce_v4_0.h" #include "vcn_v1_0.h" -#include "amdgpu_powerplay.h" #include "dce_virtual.h" #include "mxgpu_ai.h" @@ -417,12 +416,7 @@ static int soc15_asic_reset(struct amdgpu_device *adev) pci_save_state(adev->pdev); - for (i = 0; i < AMDGPU_MAX_IP_NUM; i++) { - if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP){ - adev->ip_blocks[i].version->funcs->soft_reset((void *)adev); - break; - } - } + psp_gpu_reset(adev); pci_restore_state(adev->pdev); @@ -514,6 +508,7 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev) /* Set IP register base before any HW register access */ switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: case CHIP_RAVEN: vega10_reg_base_init(adev); break; @@ -533,13 +528,13 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev) switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: amdgpu_device_ip_block_add(adev, &vega10_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v9_0_ip_block); amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block); - if (amdgpu_fw_load_type == 2 || amdgpu_fw_load_type == -1) - amdgpu_device_ip_block_add(adev, &psp_v3_1_ip_block); + amdgpu_device_ip_block_add(adev, &psp_v3_1_ip_block); if (!amdgpu_sriov_vf(adev)) - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -558,7 +553,7 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &gmc_v9_0_ip_block); amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block); amdgpu_device_ip_block_add(adev, &psp_v10_0_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -583,6 +578,21 @@ static uint32_t soc15_get_rev_id(struct amdgpu_device *adev) return adev->nbio_funcs->get_rev_id(adev); } +static void soc15_flush_hdp(struct amdgpu_device *adev, struct amdgpu_ring *ring) +{ + adev->nbio_funcs->hdp_flush(adev, ring); +} + +static void soc15_invalidate_hdp(struct amdgpu_device *adev, + struct amdgpu_ring *ring) +{ + if (!ring || !ring->funcs->emit_wreg) + WREG32_SOC15_NO_KIQ(NBIO, 0, mmHDP_READ_CACHE_INVALIDATE, 1); + else + amdgpu_ring_emit_wreg(ring, SOC15_REG_OFFSET( + HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 1); +} + static const struct amdgpu_asic_funcs soc15_asic_funcs = { .read_disabled_bios = &soc15_read_disabled_bios, @@ -594,11 +604,12 @@ static const struct amdgpu_asic_funcs soc15_asic_funcs = .set_uvd_clocks = &soc15_set_uvd_clocks, .set_vce_clocks = &soc15_set_vce_clocks, .get_config_memsize = &soc15_get_config_memsize, + .flush_hdp = &soc15_flush_hdp, + .invalidate_hdp = &soc15_invalidate_hdp, }; static int soc15_common_early_init(void *handle) { - bool psp_enabled = false; struct amdgpu_device *adev = (struct amdgpu_device *)handle; adev->smc_rreg = NULL; @@ -616,10 +627,6 @@ static int soc15_common_early_init(void *handle) adev->asic_funcs = &soc15_asic_funcs; - if (amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_PSP) && - (amdgpu_ip_block_mask & (1 << AMD_IP_BLOCK_TYPE_PSP))) - psp_enabled = true; - adev->rev_id = soc15_get_rev_id(adev); adev->external_rev_id = 0xFF; switch (adev->asic_type) { @@ -646,6 +653,28 @@ static int soc15_common_early_init(void *handle) adev->pg_flags = 0; adev->external_rev_id = 0x1; break; + case CHIP_VEGA12: + adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG | + AMD_CG_SUPPORT_GFX_MGLS | + AMD_CG_SUPPORT_GFX_CGCG | + AMD_CG_SUPPORT_GFX_CGLS | + AMD_CG_SUPPORT_GFX_3D_CGCG | + AMD_CG_SUPPORT_GFX_3D_CGLS | + AMD_CG_SUPPORT_GFX_CP_LS | + AMD_CG_SUPPORT_MC_LS | + AMD_CG_SUPPORT_MC_MGCG | + AMD_CG_SUPPORT_SDMA_MGCG | + AMD_CG_SUPPORT_SDMA_LS | + AMD_CG_SUPPORT_BIF_MGCG | + AMD_CG_SUPPORT_BIF_LS | + AMD_CG_SUPPORT_HDP_MGCG | + AMD_CG_SUPPORT_HDP_LS | + AMD_CG_SUPPORT_ROM_MGCG | + AMD_CG_SUPPORT_VCE_MGCG | + AMD_CG_SUPPORT_UVD_MGCG; + adev->pg_flags = 0; + adev->external_rev_id = adev->rev_id + 0x14; + break; case CHIP_RAVEN: adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG | AMD_CG_SUPPORT_GFX_MGLS | @@ -680,10 +709,6 @@ static int soc15_common_early_init(void *handle) xgpu_ai_mailbox_set_irq_funcs(adev); } - adev->firmware.load_type = amdgpu_ucode_get_load_type(adev, amdgpu_fw_load_type); - - amdgpu_device_get_pcie_info(adev); - return 0; } @@ -882,6 +907,7 @@ static int soc15_common_set_clockgating_state(void *handle, switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: adev->nbio_funcs->update_medium_grain_clock_gating(adev, state == AMD_CG_STATE_GATE ? true : false); adev->nbio_funcs->update_medium_grain_light_sleep(adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.h b/drivers/gpu/drm/amd/amdgpu/soc15.h index 26b3fea..f70da8a 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.h +++ b/drivers/gpu/drm/amd/amdgpu/soc15.h
@@ -27,6 +27,9 @@ #include "nbio_v6_1.h" #include "nbio_v7_0.h" +#define SOC15_FLUSH_GPU_TLB_NUM_WREG 4 +#define SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT 1 + extern const struct amd_ip_funcs soc15_common_ip_funcs; struct soc15_reg_golden {
diff --git a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c index 5995ffc..52853d8 100644 --- a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c
@@ -107,7 +107,7 @@ static int tonga_ih_irq_init(struct amdgpu_device *adev) tonga_ih_disable_interrupts(adev); /* setup interrupt control */ - WREG32(mmINTERRUPT_CNTL2, adev->dummy_page.addr >> 8); + WREG32(mmINTERRUPT_CNTL2, adev->dummy_page_addr >> 8); interrupt_cntl = RREG32(mmINTERRUPT_CNTL); /* INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=0 - dummy read disabled with msi, enabled without msi * INTERRUPT_CNTL__IH_DUMMY_RD_OVERRIDE_MASK=1 - dummy read controlled by IH_DUMMY_RD_EN @@ -271,7 +271,7 @@ static void tonga_ih_decode_iv(struct amdgpu_device *adev, entry->src_data[0] = dw[1] & 0xfffffff; entry->ring_id = dw[2] & 0xff; entry->vmid = (dw[2] >> 8) & 0xff; - entry->pas_id = (dw[2] >> 16) & 0xffff; + entry->pasid = (dw[2] >> 16) & 0xffff; /* wptr/rptr are in bytes! */ adev->irq.ih.rptr += 16;
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c index 8ab10c2..948bb943 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -464,32 +464,6 @@ static void uvd_v4_2_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq } /** - * uvd_v4_2_ring_emit_hdp_flush - emit an hdp flush - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp flush. - */ -static void uvd_v4_2_ring_emit_hdp_flush(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET0(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0)); - amdgpu_ring_write(ring, 0); -} - -/** - * uvd_v4_2_ring_hdp_invalidate - emit an hdp invalidate - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp invalidate. - */ -static void uvd_v4_2_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET0(mmHDP_DEBUG0, 0)); - amdgpu_ring_write(ring, 1); -} - -/** * uvd_v4_2_ring_test_ring - register write test * * @ring: amdgpu_ring pointer @@ -765,14 +739,10 @@ static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = { .set_wptr = uvd_v4_2_ring_set_wptr, .parse_cs = amdgpu_uvd_ring_parse_cs, .emit_frame_size = - 2 + /* uvd_v4_2_ring_emit_hdp_flush */ - 2 + /* uvd_v4_2_ring_emit_hdp_invalidate */ 14, /* uvd_v4_2_ring_emit_fence x1 no user fence */ .emit_ib_size = 4, /* uvd_v4_2_ring_emit_ib */ .emit_ib = uvd_v4_2_ring_emit_ib, .emit_fence = uvd_v4_2_ring_emit_fence, - .emit_hdp_flush = uvd_v4_2_ring_emit_hdp_flush, - .emit_hdp_invalidate = uvd_v4_2_ring_emit_hdp_invalidate, .test_ring = uvd_v4_2_ring_test_ring, .test_ib = amdgpu_uvd_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop,
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index c1fe30c..6445d55 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -479,32 +479,6 @@ static void uvd_v5_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq } /** - * uvd_v5_0_ring_emit_hdp_flush - emit an hdp flush - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp flush. - */ -static void uvd_v5_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET0(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0)); - amdgpu_ring_write(ring, 0); -} - -/** - * uvd_v5_0_ring_hdp_invalidate - emit an hdp invalidate - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp invalidate. - */ -static void uvd_v5_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET0(mmHDP_DEBUG0, 0)); - amdgpu_ring_write(ring, 1); -} - -/** * uvd_v5_0_ring_test_ring - register write test * * @ring: amdgpu_ring pointer @@ -873,14 +847,10 @@ static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = { .set_wptr = uvd_v5_0_ring_set_wptr, .parse_cs = amdgpu_uvd_ring_parse_cs, .emit_frame_size = - 2 + /* uvd_v5_0_ring_emit_hdp_flush */ - 2 + /* uvd_v5_0_ring_emit_hdp_invalidate */ 14, /* uvd_v5_0_ring_emit_fence x1 no user fence */ .emit_ib_size = 6, /* uvd_v5_0_ring_emit_ib */ .emit_ib = uvd_v5_0_ring_emit_ib, .emit_fence = uvd_v5_0_ring_emit_fence, - .emit_hdp_flush = uvd_v5_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = uvd_v5_0_ring_emit_hdp_invalidate, .test_ring = uvd_v5_0_ring_test_ring, .test_ib = amdgpu_uvd_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop,
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index 4bc33b6..04a8754a 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -964,32 +964,6 @@ static void uvd_v6_0_enc_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, } /** - * uvd_v6_0_ring_emit_hdp_flush - emit an hdp flush - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp flush. - */ -static void uvd_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET0(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0)); - amdgpu_ring_write(ring, 0); -} - -/** - * uvd_v6_0_ring_hdp_invalidate - emit an hdp invalidate - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp invalidate. - */ -static void uvd_v6_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - amdgpu_ring_write(ring, PACKET0(mmHDP_DEBUG0, 0)); - amdgpu_ring_write(ring, 1); -} - -/** * uvd_v6_0_ring_test_ring - register write test * * @ring: amdgpu_ring pointer @@ -1072,29 +1046,21 @@ static void uvd_v6_0_enc_ring_emit_ib(struct amdgpu_ring *ring, amdgpu_ring_write(ring, ib->length_dw); } -static void uvd_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring, - unsigned vmid, uint64_t pd_addr) +static void uvd_v6_0_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) { - uint32_t reg; - - if (vmid < 8) - reg = mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vmid; - else - reg = mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vmid - 8; - amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0)); amdgpu_ring_write(ring, reg << 2); amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0)); - amdgpu_ring_write(ring, pd_addr >> 12); + amdgpu_ring_write(ring, val); amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD, 0)); amdgpu_ring_write(ring, 0x8); +} - amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0)); - amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST << 2); - amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0)); - amdgpu_ring_write(ring, 1 << vmid); - amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD, 0)); - amdgpu_ring_write(ring, 0x8); +static void uvd_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring, + unsigned vmid, uint64_t pd_addr) +{ + amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0)); amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST << 2); @@ -1140,7 +1106,7 @@ static void uvd_v6_0_enc_ring_insert_end(struct amdgpu_ring *ring) } static void uvd_v6_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring, - unsigned int vmid, uint64_t pd_addr) + unsigned int vmid, uint64_t pd_addr) { amdgpu_ring_write(ring, HEVC_ENC_CMD_UPDATE_PTB); amdgpu_ring_write(ring, vmid); @@ -1562,14 +1528,11 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = { .set_wptr = uvd_v6_0_ring_set_wptr, .parse_cs = amdgpu_uvd_ring_parse_cs, .emit_frame_size = - 6 + 6 + /* hdp flush / invalidate */ 10 + /* uvd_v6_0_ring_emit_pipeline_sync */ 14, /* uvd_v6_0_ring_emit_fence x1 no user fence */ .emit_ib_size = 8, /* uvd_v6_0_ring_emit_ib */ .emit_ib = uvd_v6_0_ring_emit_ib, .emit_fence = uvd_v6_0_ring_emit_fence, - .emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = uvd_v6_0_ring_emit_hdp_invalidate, .test_ring = uvd_v6_0_ring_test_ring, .test_ib = amdgpu_uvd_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, @@ -1588,8 +1551,7 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = { .get_wptr = uvd_v6_0_ring_get_wptr, .set_wptr = uvd_v6_0_ring_set_wptr, .emit_frame_size = - 2 + /* uvd_v6_0_ring_emit_hdp_flush */ - 2 + /* uvd_v6_0_ring_emit_hdp_invalidate */ + 6 + 6 + /* hdp flush / invalidate */ 10 + /* uvd_v6_0_ring_emit_pipeline_sync */ 3 /* VI_FLUSH_GPU_TLB_NUM_WREG */ * 6 + 8 + /* uvd_v6_0_ring_emit_vm_flush */ 14 + 14, /* uvd_v6_0_ring_emit_fence x2 vm fence */ @@ -1598,14 +1560,13 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = { .emit_fence = uvd_v6_0_ring_emit_fence, .emit_vm_flush = uvd_v6_0_ring_emit_vm_flush, .emit_pipeline_sync = uvd_v6_0_ring_emit_pipeline_sync, - .emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = uvd_v6_0_ring_emit_hdp_invalidate, .test_ring = uvd_v6_0_ring_test_ring, .test_ib = amdgpu_uvd_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_uvd_ring_begin_use, .end_use = amdgpu_uvd_ring_end_use, + .emit_wreg = uvd_v6_0_ring_emit_wreg, }; static const struct amdgpu_ring_funcs uvd_v6_0_enc_ring_vm_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 6b95f4f..eddc57f 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -25,6 +25,7 @@ #include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_uvd.h" +#include "soc15.h" #include "soc15d.h" #include "soc15_common.h" #include "mmsch_v1_0.h" @@ -389,13 +390,13 @@ static int uvd_v7_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* UVD TRAP */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_UVD, 124, &adev->uvd.irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, 124, &adev->uvd.irq); if (r) return r; /* UVD ENC TRAP */ for (i = 0; i < adev->uvd.num_enc_rings; ++i) { - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_UVD, i + 119, &adev->uvd.irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, i + 119, &adev->uvd.irq); if (r) return r; } @@ -1135,37 +1136,6 @@ static void uvd_v7_0_enc_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, } /** - * uvd_v7_0_ring_emit_hdp_flush - emit an hdp flush - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp flush. - */ -static void uvd_v7_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) -{ - struct amdgpu_device *adev = ring->adev; - - amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(NBIF, 0, - mmHDP_MEM_COHERENCY_FLUSH_CNTL), 0)); - amdgpu_ring_write(ring, 0); -} - -/** - * uvd_v7_0_ring_hdp_invalidate - emit an hdp invalidate - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp invalidate. - */ -static void uvd_v7_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - struct amdgpu_device *adev = ring->adev; - - amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 0)); - amdgpu_ring_write(ring, 1); -} - -/** * uvd_v7_0_ring_test_ring - register write test * * @ring: amdgpu_ring pointer @@ -1255,33 +1225,33 @@ static void uvd_v7_0_enc_ring_emit_ib(struct amdgpu_ring *ring, amdgpu_ring_write(ring, ib->length_dw); } -static void uvd_v7_0_vm_reg_write(struct amdgpu_ring *ring, - uint32_t data0, uint32_t data1) +static void uvd_v7_0_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) { struct amdgpu_device *adev = ring->adev; amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0)); - amdgpu_ring_write(ring, data0); + amdgpu_ring_write(ring, reg << 2); amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0)); - amdgpu_ring_write(ring, data1); + amdgpu_ring_write(ring, val); amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0)); amdgpu_ring_write(ring, 8); } -static void uvd_v7_0_vm_reg_wait(struct amdgpu_ring *ring, - uint32_t data0, uint32_t data1, uint32_t mask) +static void uvd_v7_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg, + uint32_t val, uint32_t mask) { struct amdgpu_device *adev = ring->adev; amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0)); - amdgpu_ring_write(ring, data0); + amdgpu_ring_write(ring, reg << 2); amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0)); - amdgpu_ring_write(ring, data1); + amdgpu_ring_write(ring, val); amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH8), 0)); amdgpu_ring_write(ring, mask); @@ -1294,37 +1264,15 @@ static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; - uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vmid); - uint64_t flags = AMDGPU_PTE_VALID; - unsigned eng = ring->vm_inv_eng; uint32_t data0, data1, mask; - amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags); - pd_addr |= flags; + pd_addr = amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); - data0 = (hub->ctx0_ptb_addr_hi32 + vmid * 2) << 2; - data1 = upper_32_bits(pd_addr); - uvd_v7_0_vm_reg_write(ring, data0, data1); - - data0 = (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2; - data1 = lower_32_bits(pd_addr); - uvd_v7_0_vm_reg_write(ring, data0, data1); - - data0 = (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2; + /* wait for reg writes */ + data0 = hub->ctx0_ptb_addr_lo32 + vmid * 2; data1 = lower_32_bits(pd_addr); mask = 0xffffffff; - uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); - - /* flush TLB */ - data0 = (hub->vm_inv_eng0_req + eng) << 2; - data1 = req; - uvd_v7_0_vm_reg_write(ring, data0, data1); - - /* wait for flush */ - data0 = (hub->vm_inv_eng0_ack + eng) << 2; - data1 = 1 << vmid; - mask = 1 << vmid; - uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); + uvd_v7_0_ring_emit_reg_wait(ring, data0, data1, mask); } static void uvd_v7_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count) @@ -1342,40 +1290,34 @@ static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring) amdgpu_ring_write(ring, HEVC_ENC_CMD_END); } +static void uvd_v7_0_enc_ring_emit_reg_wait(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val, + uint32_t mask) +{ + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); + amdgpu_ring_write(ring, reg << 2); + amdgpu_ring_write(ring, mask); + amdgpu_ring_write(ring, val); +} + static void uvd_v7_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring, - unsigned int vmid, uint64_t pd_addr) + unsigned int vmid, uint64_t pd_addr) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; - uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vmid); - uint64_t flags = AMDGPU_PTE_VALID; - unsigned eng = ring->vm_inv_eng; - amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags); - pd_addr |= flags; + pd_addr = amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); + /* wait for reg writes */ + uvd_v7_0_enc_ring_emit_reg_wait(ring, hub->ctx0_ptb_addr_lo32 + vmid * 2, + lower_32_bits(pd_addr), 0xffffffff); +} + +static void uvd_v7_0_enc_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_hi32 + vmid * 2) << 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2); - amdgpu_ring_write(ring, 0xffffffff); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 1 << vmid); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_ring_write(ring, reg << 2); + amdgpu_ring_write(ring, val); } #if 0 @@ -1712,22 +1654,23 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = { .get_wptr = uvd_v7_0_ring_get_wptr, .set_wptr = uvd_v7_0_ring_set_wptr, .emit_frame_size = - 2 + /* uvd_v7_0_ring_emit_hdp_flush */ - 2 + /* uvd_v7_0_ring_emit_hdp_invalidate */ - 34 + /* uvd_v7_0_ring_emit_vm_flush */ + 6 + 6 + /* hdp flush / invalidate */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 + + 8 + /* uvd_v7_0_ring_emit_vm_flush */ 14 + 14, /* uvd_v7_0_ring_emit_fence x2 vm fence */ .emit_ib_size = 8, /* uvd_v7_0_ring_emit_ib */ .emit_ib = uvd_v7_0_ring_emit_ib, .emit_fence = uvd_v7_0_ring_emit_fence, .emit_vm_flush = uvd_v7_0_ring_emit_vm_flush, - .emit_hdp_flush = uvd_v7_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = uvd_v7_0_ring_emit_hdp_invalidate, .test_ring = uvd_v7_0_ring_test_ring, .test_ib = amdgpu_uvd_ring_test_ib, .insert_nop = uvd_v7_0_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_uvd_ring_begin_use, .end_use = amdgpu_uvd_ring_end_use, + .emit_wreg = uvd_v7_0_ring_emit_wreg, + .emit_reg_wait = uvd_v7_0_ring_emit_reg_wait, }; static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = { @@ -1740,7 +1683,10 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = { .get_wptr = uvd_v7_0_enc_ring_get_wptr, .set_wptr = uvd_v7_0_enc_ring_set_wptr, .emit_frame_size = - 17 + /* uvd_v7_0_enc_ring_emit_vm_flush */ + 3 + 3 + /* hdp flush / invalidate */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 4 + + 4 + /* uvd_v7_0_enc_ring_emit_vm_flush */ 5 + 5 + /* uvd_v7_0_enc_ring_emit_fence x2 vm fence */ 1, /* uvd_v7_0_enc_ring_insert_end */ .emit_ib_size = 5, /* uvd_v7_0_enc_ring_emit_ib */ @@ -1754,6 +1700,8 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_uvd_ring_begin_use, .end_use = amdgpu_uvd_ring_end_use, + .emit_wreg = uvd_v7_0_enc_ring_emit_wreg, + .emit_reg_wait = uvd_v7_0_enc_ring_emit_reg_wait, }; static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index bb5ec09..085f0ba 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -844,7 +844,7 @@ static void vce_v3_0_ring_emit_ib(struct amdgpu_ring *ring, } static void vce_v3_0_emit_vm_flush(struct amdgpu_ring *ring, - unsigned int vmid, uint64_t pd_addr) + unsigned int vmid, uint64_t pd_addr) { amdgpu_ring_write(ring, VCE_CMD_UPDATE_PTB); amdgpu_ring_write(ring, vmid);
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c index 7cf2eef..73fd48d 100755 --- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
@@ -28,6 +28,7 @@ #include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_vce.h" +#include "soc15.h" #include "soc15d.h" #include "soc15_common.h" #include "mmsch_v1_0.h" @@ -419,7 +420,7 @@ static int vce_v4_0_sw_init(void *handle) unsigned size; int r, i; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_VCE0, 167, &adev->vce.irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCE0, 167, &adev->vce.irq); if (r) return r; @@ -964,40 +965,33 @@ static void vce_v4_0_ring_insert_end(struct amdgpu_ring *ring) amdgpu_ring_write(ring, VCE_CMD_END); } +static void vce_v4_0_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg, + uint32_t val, uint32_t mask) +{ + amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); + amdgpu_ring_write(ring, reg << 2); + amdgpu_ring_write(ring, mask); + amdgpu_ring_write(ring, val); +} + static void vce_v4_0_emit_vm_flush(struct amdgpu_ring *ring, - unsigned int vmid, uint64_t pd_addr) + unsigned int vmid, uint64_t pd_addr) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; - uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vmid); - uint64_t flags = AMDGPU_PTE_VALID; - unsigned eng = ring->vm_inv_eng; - amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags); - pd_addr |= flags; + pd_addr = amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); + /* wait for reg writes */ + vce_v4_0_emit_reg_wait(ring, hub->ctx0_ptb_addr_lo32 + vmid * 2, + lower_32_bits(pd_addr), 0xffffffff); +} + +static void vce_v4_0_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_hi32 + vmid * 2) << 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2); - amdgpu_ring_write(ring, 0xffffffff); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 1 << vmid); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_ring_write(ring, reg << 2); + amdgpu_ring_write(ring, val); } static int vce_v4_0_set_interrupt_state(struct amdgpu_device *adev, @@ -1069,7 +1063,9 @@ static const struct amdgpu_ring_funcs vce_v4_0_ring_vm_funcs = { .set_wptr = vce_v4_0_ring_set_wptr, .parse_cs = amdgpu_vce_ring_parse_cs_vm, .emit_frame_size = - 17 + /* vce_v4_0_emit_vm_flush */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 4 + + 4 + /* vce_v4_0_emit_vm_flush */ 5 + 5 + /* amdgpu_vce_ring_emit_fence x2 vm fence */ 1, /* vce_v4_0_ring_insert_end */ .emit_ib_size = 5, /* vce_v4_0_ring_emit_ib */ @@ -1083,6 +1079,8 @@ static const struct amdgpu_ring_funcs vce_v4_0_ring_vm_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_vce_ring_begin_use, .end_use = amdgpu_vce_ring_end_use, + .emit_wreg = vce_v4_0_emit_wreg, + .emit_reg_wait = vce_v4_0_emit_reg_wait, }; static void vce_v4_0_set_ring_funcs(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c index 23dbb1e..19c0649 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -25,6 +25,7 @@ #include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_vcn.h" +#include "soc15.h" #include "soc15d.h" #include "soc15_common.h" @@ -74,13 +75,13 @@ static int vcn_v1_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* VCN DEC TRAP */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_VCN, 124, &adev->vcn.irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, 124, &adev->vcn.irq); if (r) return r; /* VCN ENC TRAP */ for (i = 0; i < adev->vcn.num_enc_rings; ++i) { - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_VCN, i + 119, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, i + 119, &adev->vcn.irq); if (r) return r; @@ -831,21 +832,6 @@ static void vcn_v1_0_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 } /** - * vcn_v1_0_dec_ring_hdp_invalidate - emit an hdp invalidate - * - * @ring: amdgpu_ring pointer - * - * Emits an hdp invalidate. - */ -static void vcn_v1_0_dec_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) -{ - struct amdgpu_device *adev = ring->adev; - - amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 0)); - amdgpu_ring_write(ring, 1); -} - -/** * vcn_v1_0_dec_ring_emit_ib - execute indirect buffer * * @ring: amdgpu_ring pointer @@ -874,33 +860,18 @@ static void vcn_v1_0_dec_ring_emit_ib(struct amdgpu_ring *ring, amdgpu_ring_write(ring, ib->length_dw); } -static void vcn_v1_0_dec_vm_reg_write(struct amdgpu_ring *ring, - uint32_t data0, uint32_t data1) +static void vcn_v1_0_dec_ring_emit_reg_wait(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val, + uint32_t mask) { struct amdgpu_device *adev = ring->adev; amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0)); - amdgpu_ring_write(ring, data0); + amdgpu_ring_write(ring, reg << 2); amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0)); - amdgpu_ring_write(ring, data1); - amdgpu_ring_write(ring, - PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0)); - amdgpu_ring_write(ring, VCN_DEC_CMD_WRITE_REG << 1); -} - -static void vcn_v1_0_dec_vm_reg_wait(struct amdgpu_ring *ring, - uint32_t data0, uint32_t data1, uint32_t mask) -{ - struct amdgpu_device *adev = ring->adev; - - amdgpu_ring_write(ring, - PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0)); - amdgpu_ring_write(ring, data0); - amdgpu_ring_write(ring, - PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0)); - amdgpu_ring_write(ring, data1); + amdgpu_ring_write(ring, val); amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH8), 0)); amdgpu_ring_write(ring, mask); @@ -910,40 +881,34 @@ static void vcn_v1_0_dec_vm_reg_wait(struct amdgpu_ring *ring, } static void vcn_v1_0_dec_ring_emit_vm_flush(struct amdgpu_ring *ring, - unsigned vmid, uint64_t pd_addr) + unsigned vmid, uint64_t pd_addr) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; - uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vmid); - uint64_t flags = AMDGPU_PTE_VALID; - unsigned eng = ring->vm_inv_eng; uint32_t data0, data1, mask; - amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags); - pd_addr |= flags; + pd_addr = amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); - data0 = (hub->ctx0_ptb_addr_hi32 + vmid * 2) << 2; - data1 = upper_32_bits(pd_addr); - vcn_v1_0_dec_vm_reg_write(ring, data0, data1); - - data0 = (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2; - data1 = lower_32_bits(pd_addr); - vcn_v1_0_dec_vm_reg_write(ring, data0, data1); - - data0 = (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2; + /* wait for register write */ + data0 = hub->ctx0_ptb_addr_lo32 + vmid * 2; data1 = lower_32_bits(pd_addr); mask = 0xffffffff; - vcn_v1_0_dec_vm_reg_wait(ring, data0, data1, mask); + vcn_v1_0_dec_ring_emit_reg_wait(ring, data0, data1, mask); +} - /* flush TLB */ - data0 = (hub->vm_inv_eng0_req + eng) << 2; - data1 = req; - vcn_v1_0_dec_vm_reg_write(ring, data0, data1); +static void vcn_v1_0_dec_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ + struct amdgpu_device *adev = ring->adev; - /* wait for flush */ - data0 = (hub->vm_inv_eng0_ack + eng) << 2; - data1 = 1 << vmid; - mask = 1 << vmid; - vcn_v1_0_dec_vm_reg_wait(ring, data0, data1, mask); + amdgpu_ring_write(ring, + PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0)); + amdgpu_ring_write(ring, reg << 2); + amdgpu_ring_write(ring, + PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0)); + amdgpu_ring_write(ring, val); + amdgpu_ring_write(ring, + PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0)); + amdgpu_ring_write(ring, VCN_DEC_CMD_WRITE_REG << 1); } /** @@ -1042,43 +1007,34 @@ static void vcn_v1_0_enc_ring_emit_ib(struct amdgpu_ring *ring, amdgpu_ring_write(ring, ib->length_dw); } +static void vcn_v1_0_enc_ring_emit_reg_wait(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val, + uint32_t mask) +{ + amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WAIT); + amdgpu_ring_write(ring, reg << 2); + amdgpu_ring_write(ring, mask); + amdgpu_ring_write(ring, val); +} + static void vcn_v1_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring, - unsigned int vmid, uint64_t pd_addr) + unsigned int vmid, uint64_t pd_addr) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; - uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vmid); - uint64_t flags = AMDGPU_PTE_VALID; - unsigned eng = ring->vm_inv_eng; - amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags); - pd_addr |= flags; + pd_addr = amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); + /* wait for reg writes */ + vcn_v1_0_enc_ring_emit_reg_wait(ring, hub->ctx0_ptb_addr_lo32 + vmid * 2, + lower_32_bits(pd_addr), 0xffffffff); +} + +static void vcn_v1_0_enc_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_hi32 + vmid * 2) << 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WAIT); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vmid * 2) << 2); - amdgpu_ring_write(ring, 0xffffffff); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 1 << vmid); - amdgpu_ring_write(ring, 1 << vmid); + amdgpu_ring_write(ring, reg << 2); + amdgpu_ring_write(ring, val); } static int vcn_v1_0_set_interrupt_state(struct amdgpu_device *adev, @@ -1155,15 +1111,16 @@ static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = { .get_wptr = vcn_v1_0_dec_ring_get_wptr, .set_wptr = vcn_v1_0_dec_ring_set_wptr, .emit_frame_size = - 2 + /* vcn_v1_0_dec_ring_emit_hdp_invalidate */ - 34 + /* vcn_v1_0_dec_ring_emit_vm_flush */ + 6 + 6 + /* hdp invalidate / flush */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 + + 8 + /* vcn_v1_0_dec_ring_emit_vm_flush */ 14 + 14 + /* vcn_v1_0_dec_ring_emit_fence x2 vm fence */ 6, .emit_ib_size = 8, /* vcn_v1_0_dec_ring_emit_ib */ .emit_ib = vcn_v1_0_dec_ring_emit_ib, .emit_fence = vcn_v1_0_dec_ring_emit_fence, .emit_vm_flush = vcn_v1_0_dec_ring_emit_vm_flush, - .emit_hdp_invalidate = vcn_v1_0_dec_ring_emit_hdp_invalidate, .test_ring = amdgpu_vcn_dec_ring_test_ring, .test_ib = amdgpu_vcn_dec_ring_test_ib, .insert_nop = vcn_v1_0_ring_insert_nop, @@ -1172,6 +1129,8 @@ static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_vcn_ring_begin_use, .end_use = amdgpu_vcn_ring_end_use, + .emit_wreg = vcn_v1_0_dec_ring_emit_wreg, + .emit_reg_wait = vcn_v1_0_dec_ring_emit_reg_wait, }; static const struct amdgpu_ring_funcs vcn_v1_0_enc_ring_vm_funcs = { @@ -1184,7 +1143,9 @@ static const struct amdgpu_ring_funcs vcn_v1_0_enc_ring_vm_funcs = { .get_wptr = vcn_v1_0_enc_ring_get_wptr, .set_wptr = vcn_v1_0_enc_ring_set_wptr, .emit_frame_size = - 17 + /* vcn_v1_0_enc_ring_emit_vm_flush */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 4 + + 4 + /* vcn_v1_0_enc_ring_emit_vm_flush */ 5 + 5 + /* vcn_v1_0_enc_ring_emit_fence x2 vm fence */ 1, /* vcn_v1_0_enc_ring_insert_end */ .emit_ib_size = 5, /* vcn_v1_0_enc_ring_emit_ib */ @@ -1198,6 +1159,8 @@ static const struct amdgpu_ring_funcs vcn_v1_0_enc_ring_vm_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_vcn_ring_begin_use, .end_use = amdgpu_vcn_ring_end_use, + .emit_wreg = vcn_v1_0_enc_ring_emit_wreg, + .emit_reg_wait = vcn_v1_0_enc_ring_emit_reg_wait, }; static void vcn_v1_0_set_dec_ring_funcs(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c index ee14d78..5ae5ed2 100644 --- a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c
@@ -245,8 +245,8 @@ static bool vega10_ih_prescreen_iv(struct amdgpu_device *adev) * some faults get cleared. */ switch (dw0 & 0xff) { - case AMDGPU_IH_CLIENTID_VMC: - case AMDGPU_IH_CLIENTID_UTCL2: + case SOC15_IH_CLIENTID_VMC: + case SOC15_IH_CLIENTID_UTCL2: break; default: /* Not a VM fault */ @@ -333,7 +333,7 @@ static void vega10_ih_decode_iv(struct amdgpu_device *adev, entry->vmid_src = (dw[0] >> 31); entry->timestamp = dw[1] | ((u64)(dw[2] & 0xffff) << 32); entry->timestamp_src = dw[2] >> 31; - entry->pas_id = dw[3] & 0xffff; + entry->pasid = dw[3] & 0xffff; entry->pasid_src = dw[3] >> 31; entry->src_data[0] = dw[4]; entry->src_data[1] = dw[5];
diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c index b7bdd04..45aafca 100644 --- a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c +++ b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
@@ -24,7 +24,8 @@ #include "soc15.h" #include "soc15_common.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" int vega10_reg_base_init(struct amdgpu_device *adev) { @@ -37,6 +38,7 @@ int vega10_reg_base_init(struct amdgpu_device *adev) adev->reg_offset[ATHUB_HWIP][i] = (uint32_t *)(&(ATHUB_BASE.instance[i])); adev->reg_offset[NBIO_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i])); adev->reg_offset[MP0_HWIP][i] = (uint32_t *)(&(MP0_BASE.instance[i])); + adev->reg_offset[MP1_HWIP][i] = (uint32_t *)(&(MP1_BASE.instance[i])); adev->reg_offset[UVD_HWIP][i] = (uint32_t *)(&(UVD_BASE.instance[i])); adev->reg_offset[VCE_HWIP][i] = (uint32_t *)(&(VCE_BASE.instance[i])); adev->reg_offset[VCN_HWIP][i] = (uint32_t *)(&(VCN_BASE.instance[i])); @@ -48,7 +50,7 @@ int vega10_reg_base_init(struct amdgpu_device *adev) adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i])); adev->reg_offset[PWR_HWIP][i] = (uint32_t *)(&(PWR_BASE.instance[i])); adev->reg_offset[NBIF_HWIP][i] = (uint32_t *)(&(NBIF_BASE.instance[i])); - + adev->reg_offset[THM_HWIP][i] = (uint32_t *)(&(THM_BASE.instance[i])); } return 0; }
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index f5d714e..9ae350d 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -71,7 +71,6 @@ #include "uvd_v5_0.h" #include "uvd_v6_0.h" #include "vce_v3_0.h" -#include "amdgpu_powerplay.h" #if defined(CONFIG_DRM_AMD_ACP) #include "amdgpu_acp.h" #endif @@ -899,6 +898,27 @@ static uint32_t vi_get_rev_id(struct amdgpu_device *adev) >> PCIE_EFUSE4__STRAP_BIF_ATI_REV_ID__SHIFT; } +static void vi_flush_hdp(struct amdgpu_device *adev, struct amdgpu_ring *ring) +{ + if (!ring || !ring->funcs->emit_wreg) { + WREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 1); + RREG32(mmHDP_MEM_COHERENCY_FLUSH_CNTL); + } else { + amdgpu_ring_emit_wreg(ring, mmHDP_MEM_COHERENCY_FLUSH_CNTL, 1); + } +} + +static void vi_invalidate_hdp(struct amdgpu_device *adev, + struct amdgpu_ring *ring) +{ + if (!ring || !ring->funcs->emit_wreg) { + WREG32(mmHDP_DEBUG0, 1); + RREG32(mmHDP_DEBUG0); + } else { + amdgpu_ring_emit_wreg(ring, mmHDP_DEBUG0, 1); + } +} + static const struct amdgpu_asic_funcs vi_asic_funcs = { .read_disabled_bios = &vi_read_disabled_bios, @@ -910,6 +930,8 @@ static const struct amdgpu_asic_funcs vi_asic_funcs = .set_uvd_clocks = &vi_set_uvd_clocks, .set_vce_clocks = &vi_set_vce_clocks, .get_config_memsize = &vi_get_config_memsize, + .flush_hdp = &vi_flush_hdp, + .invalidate_hdp = &vi_invalidate_hdp, }; #define CZ_REV_BRISTOL(rev) \ @@ -917,7 +939,6 @@ static const struct amdgpu_asic_funcs vi_asic_funcs = static int vi_common_early_init(void *handle) { - bool smc_enabled = false; struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (adev->flags & AMD_IS_APU) { @@ -938,10 +959,6 @@ static int vi_common_early_init(void *handle) adev->asic_funcs = &vi_asic_funcs; - if (amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_SMC) && - (amdgpu_ip_block_mask & (1 << AMD_IP_BLOCK_TYPE_SMC))) - smc_enabled = true; - adev->rev_id = vi_get_rev_id(adev); adev->external_rev_id = 0xFF; switch (adev->asic_type) { @@ -1117,11 +1134,6 @@ static int vi_common_early_init(void *handle) xgpu_vi_mailbox_set_irq_funcs(adev); } - /* vi use smc load by default */ - adev->firmware.load_type = amdgpu_ucode_get_load_type(adev, amdgpu_fw_load_type); - - amdgpu_device_get_pcie_info(adev); - return 0; } @@ -1536,7 +1548,7 @@ int vi_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &vi_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v7_4_ip_block); amdgpu_device_ip_block_add(adev, &iceland_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); amdgpu_device_ip_block_add(adev, &gfx_v8_0_ip_block); @@ -1546,7 +1558,7 @@ int vi_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &vi_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v8_5_ip_block); amdgpu_device_ip_block_add(adev, &tonga_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -1566,7 +1578,7 @@ int vi_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &vi_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v8_0_ip_block); amdgpu_device_ip_block_add(adev, &tonga_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -1588,7 +1600,7 @@ int vi_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &vi_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v8_1_ip_block); amdgpu_device_ip_block_add(adev, &tonga_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -1606,7 +1618,7 @@ int vi_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &vi_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v8_0_ip_block); amdgpu_device_ip_block_add(adev, &cz_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -1627,7 +1639,7 @@ int vi_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &vi_common_ip_block); amdgpu_device_ip_block_add(adev, &gmc_v8_0_ip_block); amdgpu_device_ip_block_add(adev, &cz_ih_ip_block); - amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block); + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC)
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.h b/drivers/gpu/drm/amd/amdgpu/vi.h index 575d7ae..0429fe3 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.h +++ b/drivers/gpu/drm/amd/amdgpu/vi.h
@@ -24,6 +24,8 @@ #ifndef __VI_H__ #define __VI_H__ +#define VI_FLUSH_GPU_TLB_NUM_WREG 3 + void vi_srbm_select(struct amdgpu_device *adev, u32 me, u32 pipe, u32 queue, u32 vmid); int vi_set_ip_blocks(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdkfd/Kconfig b/drivers/gpu/drm/amd/amdkfd/Kconfig index bc5a2945..3858820 100644 --- a/drivers/gpu/drm/amd/amdkfd/Kconfig +++ b/drivers/gpu/drm/amd/amdkfd/Kconfig
@@ -4,6 +4,8 @@ config HSA_AMD tristate "HSA kernel driver for AMD GPU devices" - depends on DRM_AMDGPU && AMD_IOMMU_V2 && X86_64 + depends on DRM_AMDGPU && X86_64 + imply AMD_IOMMU_V2 + select MMU_NOTIFIER help Enable this if you want to use HSA features on AMD GPU devices.
diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile index a317e76f..0d02422 100644 --- a/drivers/gpu/drm/amd/amdkfd/Makefile +++ b/drivers/gpu/drm/amd/amdkfd/Makefile
@@ -37,6 +37,10 @@ kfd_interrupt.o kfd_events.o cik_event_interrupt.o \ kfd_dbgdev.o kfd_dbgmgr.o kfd_crat.o +ifneq ($(CONFIG_AMD_IOMMU_V2),) +amdkfd-y += kfd_iommu.o +endif + amdkfd-$(CONFIG_DEBUG_FS) += kfd_debugfs.o obj-$(CONFIG_HSA_AMD) += amdkfd.o
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index 0492aff..59808a3 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -24,6 +24,7 @@ #include <linux/export.h> #include <linux/err.h> #include <linux/fs.h> +#include <linux/file.h> #include <linux/sched.h> #include <linux/slab.h> #include <linux/uaccess.h> @@ -826,12 +827,155 @@ static int kfd_ioctl_get_process_apertures(struct file *filp, return 0; } +static int kfd_ioctl_get_process_apertures_new(struct file *filp, + struct kfd_process *p, void *data) +{ + struct kfd_ioctl_get_process_apertures_new_args *args = data; + struct kfd_process_device_apertures *pa; + struct kfd_process_device *pdd; + uint32_t nodes = 0; + int ret; + + dev_dbg(kfd_device, "get apertures for PASID %d", p->pasid); + + if (args->num_of_nodes == 0) { + /* Return number of nodes, so that user space can alloacate + * sufficient memory + */ + mutex_lock(&p->mutex); + + if (!kfd_has_process_device_data(p)) + goto out_unlock; + + /* Run over all pdd of the process */ + pdd = kfd_get_first_process_device_data(p); + do { + args->num_of_nodes++; + pdd = kfd_get_next_process_device_data(p, pdd); + } while (pdd); + + goto out_unlock; + } + + /* Fill in process-aperture information for all available + * nodes, but not more than args->num_of_nodes as that is + * the amount of memory allocated by user + */ + pa = kzalloc((sizeof(struct kfd_process_device_apertures) * + args->num_of_nodes), GFP_KERNEL); + if (!pa) + return -ENOMEM; + + mutex_lock(&p->mutex); + + if (!kfd_has_process_device_data(p)) { + args->num_of_nodes = 0; + kfree(pa); + goto out_unlock; + } + + /* Run over all pdd of the process */ + pdd = kfd_get_first_process_device_data(p); + do { + pa[nodes].gpu_id = pdd->dev->id; + pa[nodes].lds_base = pdd->lds_base; + pa[nodes].lds_limit = pdd->lds_limit; + pa[nodes].gpuvm_base = pdd->gpuvm_base; + pa[nodes].gpuvm_limit = pdd->gpuvm_limit; + pa[nodes].scratch_base = pdd->scratch_base; + pa[nodes].scratch_limit = pdd->scratch_limit; + + dev_dbg(kfd_device, + "gpu id %u\n", pdd->dev->id); + dev_dbg(kfd_device, + "lds_base %llX\n", pdd->lds_base); + dev_dbg(kfd_device, + "lds_limit %llX\n", pdd->lds_limit); + dev_dbg(kfd_device, + "gpuvm_base %llX\n", pdd->gpuvm_base); + dev_dbg(kfd_device, + "gpuvm_limit %llX\n", pdd->gpuvm_limit); + dev_dbg(kfd_device, + "scratch_base %llX\n", pdd->scratch_base); + dev_dbg(kfd_device, + "scratch_limit %llX\n", pdd->scratch_limit); + nodes++; + + pdd = kfd_get_next_process_device_data(p, pdd); + } while (pdd && (nodes < args->num_of_nodes)); + mutex_unlock(&p->mutex); + + args->num_of_nodes = nodes; + ret = copy_to_user( + (void __user *)args->kfd_process_device_apertures_ptr, + pa, + (nodes * sizeof(struct kfd_process_device_apertures))); + kfree(pa); + return ret ? -EFAULT : 0; + +out_unlock: + mutex_unlock(&p->mutex); + return 0; +} + static int kfd_ioctl_create_event(struct file *filp, struct kfd_process *p, void *data) { struct kfd_ioctl_create_event_args *args = data; int err; + /* For dGPUs the event page is allocated in user mode. The + * handle is passed to KFD with the first call to this IOCTL + * through the event_page_offset field. + */ + if (args->event_page_offset) { + struct kfd_dev *kfd; + struct kfd_process_device *pdd; + void *mem, *kern_addr; + uint64_t size; + + if (p->signal_page) { + pr_err("Event page is already set\n"); + return -EINVAL; + } + + kfd = kfd_device_by_id(GET_GPU_ID(args->event_page_offset)); + if (!kfd) { + pr_err("Getting device by id failed in %s\n", __func__); + return -EINVAL; + } + + mutex_lock(&p->mutex); + pdd = kfd_bind_process_to_device(kfd, p); + if (IS_ERR(pdd)) { + err = PTR_ERR(pdd); + goto out_unlock; + } + + mem = kfd_process_device_translate_handle(pdd, + GET_IDR_HANDLE(args->event_page_offset)); + if (!mem) { + pr_err("Can't find BO, offset is 0x%llx\n", + args->event_page_offset); + err = -EINVAL; + goto out_unlock; + } + mutex_unlock(&p->mutex); + + err = kfd->kfd2kgd->map_gtt_bo_to_kernel(kfd->kgd, + mem, &kern_addr, &size); + if (err) { + pr_err("Failed to map event page to kernel\n"); + return err; + } + + err = kfd_event_page_set(p, kern_addr, size); + if (err) { + pr_err("Failed to set event page\n"); + return err; + } + } + err = kfd_event_create(filp, p, args->event_type, args->auto_reset != 0, args->node_id, &args->event_id, &args->event_trigger_data, @@ -839,6 +983,10 @@ static int kfd_ioctl_create_event(struct file *filp, struct kfd_process *p, &args->event_slot_index); return err; + +out_unlock: + mutex_unlock(&p->mutex); + return err; } static int kfd_ioctl_destroy_event(struct file *filp, struct kfd_process *p, @@ -902,7 +1050,8 @@ static int kfd_ioctl_set_scratch_backing_va(struct file *filep, mutex_unlock(&p->mutex); - if (sched_policy == KFD_SCHED_POLICY_NO_HWS && pdd->qpd.vmid != 0) + if (dev->dqm->sched_policy == KFD_SCHED_POLICY_NO_HWS && + pdd->qpd.vmid != 0) dev->kfd2kgd->set_scratch_backing_va( dev->kgd, args->va_addr, pdd->qpd.vmid); @@ -955,6 +1104,371 @@ static int kfd_ioctl_get_tile_config(struct file *filep, return 0; } +static int kfd_ioctl_acquire_vm(struct file *filep, struct kfd_process *p, + void *data) +{ + struct kfd_ioctl_acquire_vm_args *args = data; + struct kfd_process_device *pdd; + struct kfd_dev *dev; + struct file *drm_file; + int ret; + + dev = kfd_device_by_id(args->gpu_id); + if (!dev) + return -EINVAL; + + drm_file = fget(args->drm_fd); + if (!drm_file) + return -EINVAL; + + mutex_lock(&p->mutex); + + pdd = kfd_get_process_device_data(dev, p); + if (!pdd) { + ret = -EINVAL; + goto err_unlock; + } + + if (pdd->drm_file) { + ret = pdd->drm_file == drm_file ? 0 : -EBUSY; + goto err_unlock; + } + + ret = kfd_process_device_init_vm(pdd, drm_file); + if (ret) + goto err_unlock; + /* On success, the PDD keeps the drm_file reference */ + mutex_unlock(&p->mutex); + + return 0; + +err_unlock: + mutex_unlock(&p->mutex); + fput(drm_file); + return ret; +} + +static bool kfd_dev_is_large_bar(struct kfd_dev *dev) +{ + struct kfd_local_mem_info mem_info; + + if (debug_largebar) { + pr_debug("Simulate large-bar allocation on non large-bar machine\n"); + return true; + } + + if (dev->device_info->needs_iommu_device) + return false; + + dev->kfd2kgd->get_local_mem_info(dev->kgd, &mem_info); + if (mem_info.local_mem_size_private == 0 && + mem_info.local_mem_size_public > 0) + return true; + return false; +} + +static int kfd_ioctl_alloc_memory_of_gpu(struct file *filep, + struct kfd_process *p, void *data) +{ + struct kfd_ioctl_alloc_memory_of_gpu_args *args = data; + struct kfd_process_device *pdd; + void *mem; + struct kfd_dev *dev; + int idr_handle; + long err; + uint64_t offset = args->mmap_offset; + uint32_t flags = args->flags; + + if (args->size == 0) + return -EINVAL; + + dev = kfd_device_by_id(args->gpu_id); + if (!dev) + return -EINVAL; + + if ((flags & KFD_IOC_ALLOC_MEM_FLAGS_PUBLIC) && + (flags & KFD_IOC_ALLOC_MEM_FLAGS_VRAM) && + !kfd_dev_is_large_bar(dev)) { + pr_err("Alloc host visible vram on small bar is not allowed\n"); + return -EINVAL; + } + + mutex_lock(&p->mutex); + + pdd = kfd_bind_process_to_device(dev, p); + if (IS_ERR(pdd)) { + err = PTR_ERR(pdd); + goto err_unlock; + } + + err = dev->kfd2kgd->alloc_memory_of_gpu( + dev->kgd, args->va_addr, args->size, + pdd->vm, (struct kgd_mem **) &mem, &offset, + flags); + + if (err) + goto err_unlock; + + idr_handle = kfd_process_device_create_obj_handle(pdd, mem); + if (idr_handle < 0) { + err = -EFAULT; + goto err_free; + } + + mutex_unlock(&p->mutex); + + args->handle = MAKE_HANDLE(args->gpu_id, idr_handle); + args->mmap_offset = offset; + + return 0; + +err_free: + dev->kfd2kgd->free_memory_of_gpu(dev->kgd, (struct kgd_mem *)mem); +err_unlock: + mutex_unlock(&p->mutex); + return err; +} + +static int kfd_ioctl_free_memory_of_gpu(struct file *filep, + struct kfd_process *p, void *data) +{ + struct kfd_ioctl_free_memory_of_gpu_args *args = data; + struct kfd_process_device *pdd; + void *mem; + struct kfd_dev *dev; + int ret; + + dev = kfd_device_by_id(GET_GPU_ID(args->handle)); + if (!dev) + return -EINVAL; + + mutex_lock(&p->mutex); + + pdd = kfd_get_process_device_data(dev, p); + if (!pdd) { + pr_err("Process device data doesn't exist\n"); + ret = -EINVAL; + goto err_unlock; + } + + mem = kfd_process_device_translate_handle( + pdd, GET_IDR_HANDLE(args->handle)); + if (!mem) { + ret = -EINVAL; + goto err_unlock; + } + + ret = dev->kfd2kgd->free_memory_of_gpu(dev->kgd, (struct kgd_mem *)mem); + + /* If freeing the buffer failed, leave the handle in place for + * clean-up during process tear-down. + */ + if (!ret) + kfd_process_device_remove_obj_handle( + pdd, GET_IDR_HANDLE(args->handle)); + +err_unlock: + mutex_unlock(&p->mutex); + return ret; +} + +static int kfd_ioctl_map_memory_to_gpu(struct file *filep, + struct kfd_process *p, void *data) +{ + struct kfd_ioctl_map_memory_to_gpu_args *args = data; + struct kfd_process_device *pdd, *peer_pdd; + void *mem; + struct kfd_dev *dev, *peer; + long err = 0; + int i; + uint32_t *devices_arr = NULL; + + dev = kfd_device_by_id(GET_GPU_ID(args->handle)); + if (!dev) + return -EINVAL; + + if (!args->n_devices) { + pr_debug("Device IDs array empty\n"); + return -EINVAL; + } + if (args->n_success > args->n_devices) { + pr_debug("n_success exceeds n_devices\n"); + return -EINVAL; + } + + devices_arr = kmalloc(args->n_devices * sizeof(*devices_arr), + GFP_KERNEL); + if (!devices_arr) + return -ENOMEM; + + err = copy_from_user(devices_arr, + (void __user *)args->device_ids_array_ptr, + args->n_devices * sizeof(*devices_arr)); + if (err != 0) { + err = -EFAULT; + goto copy_from_user_failed; + } + + mutex_lock(&p->mutex); + + pdd = kfd_bind_process_to_device(dev, p); + if (IS_ERR(pdd)) { + err = PTR_ERR(pdd); + goto bind_process_to_device_failed; + } + + mem = kfd_process_device_translate_handle(pdd, + GET_IDR_HANDLE(args->handle)); + if (!mem) { + err = -ENOMEM; + goto get_mem_obj_from_handle_failed; + } + + for (i = args->n_success; i < args->n_devices; i++) { + peer = kfd_device_by_id(devices_arr[i]); + if (!peer) { + pr_debug("Getting device by id failed for 0x%x\n", + devices_arr[i]); + err = -EINVAL; + goto get_mem_obj_from_handle_failed; + } + + peer_pdd = kfd_bind_process_to_device(peer, p); + if (IS_ERR(peer_pdd)) { + err = PTR_ERR(peer_pdd); + goto get_mem_obj_from_handle_failed; + } + err = peer->kfd2kgd->map_memory_to_gpu( + peer->kgd, (struct kgd_mem *)mem, peer_pdd->vm); + if (err) { + pr_err("Failed to map to gpu %d/%d\n", + i, args->n_devices); + goto map_memory_to_gpu_failed; + } + args->n_success = i+1; + } + + mutex_unlock(&p->mutex); + + err = dev->kfd2kgd->sync_memory(dev->kgd, (struct kgd_mem *) mem, true); + if (err) { + pr_debug("Sync memory failed, wait interrupted by user signal\n"); + goto sync_memory_failed; + } + + /* Flush TLBs after waiting for the page table updates to complete */ + for (i = 0; i < args->n_devices; i++) { + peer = kfd_device_by_id(devices_arr[i]); + if (WARN_ON_ONCE(!peer)) + continue; + peer_pdd = kfd_get_process_device_data(peer, p); + if (WARN_ON_ONCE(!peer_pdd)) + continue; + kfd_flush_tlb(peer_pdd); + } + + kfree(devices_arr); + + return err; + +bind_process_to_device_failed: +get_mem_obj_from_handle_failed: +map_memory_to_gpu_failed: + mutex_unlock(&p->mutex); +copy_from_user_failed: +sync_memory_failed: + kfree(devices_arr); + + return err; +} + +static int kfd_ioctl_unmap_memory_from_gpu(struct file *filep, + struct kfd_process *p, void *data) +{ + struct kfd_ioctl_unmap_memory_from_gpu_args *args = data; + struct kfd_process_device *pdd, *peer_pdd; + void *mem; + struct kfd_dev *dev, *peer; + long err = 0; + uint32_t *devices_arr = NULL, i; + + dev = kfd_device_by_id(GET_GPU_ID(args->handle)); + if (!dev) + return -EINVAL; + + if (!args->n_devices) { + pr_debug("Device IDs array empty\n"); + return -EINVAL; + } + if (args->n_success > args->n_devices) { + pr_debug("n_success exceeds n_devices\n"); + return -EINVAL; + } + + devices_arr = kmalloc(args->n_devices * sizeof(*devices_arr), + GFP_KERNEL); + if (!devices_arr) + return -ENOMEM; + + err = copy_from_user(devices_arr, + (void __user *)args->device_ids_array_ptr, + args->n_devices * sizeof(*devices_arr)); + if (err != 0) { + err = -EFAULT; + goto copy_from_user_failed; + } + + mutex_lock(&p->mutex); + + pdd = kfd_get_process_device_data(dev, p); + if (!pdd) { + err = -EINVAL; + goto bind_process_to_device_failed; + } + + mem = kfd_process_device_translate_handle(pdd, + GET_IDR_HANDLE(args->handle)); + if (!mem) { + err = -ENOMEM; + goto get_mem_obj_from_handle_failed; + } + + for (i = args->n_success; i < args->n_devices; i++) { + peer = kfd_device_by_id(devices_arr[i]); + if (!peer) { + err = -EINVAL; + goto get_mem_obj_from_handle_failed; + } + + peer_pdd = kfd_get_process_device_data(peer, p); + if (!peer_pdd) { + err = -ENODEV; + goto get_mem_obj_from_handle_failed; + } + err = dev->kfd2kgd->unmap_memory_to_gpu( + peer->kgd, (struct kgd_mem *)mem, peer_pdd->vm); + if (err) { + pr_err("Failed to unmap from gpu %d/%d\n", + i, args->n_devices); + goto unmap_memory_from_gpu_failed; + } + args->n_success = i+1; + } + kfree(devices_arr); + + mutex_unlock(&p->mutex); + + return 0; + +bind_process_to_device_failed: +get_mem_obj_from_handle_failed: +unmap_memory_from_gpu_failed: + mutex_unlock(&p->mutex); +copy_from_user_failed: + kfree(devices_arr); + return err; +} + #define AMDKFD_IOCTL_DEF(ioctl, _func, _flags) \ [_IOC_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, \ .cmd_drv = 0, .name = #ioctl} @@ -1017,6 +1531,25 @@ static const struct amdkfd_ioctl_desc amdkfd_ioctls[] = { AMDKFD_IOCTL_DEF(AMDKFD_IOC_SET_TRAP_HANDLER, kfd_ioctl_set_trap_handler, 0), + + AMDKFD_IOCTL_DEF(AMDKFD_IOC_GET_PROCESS_APERTURES_NEW, + kfd_ioctl_get_process_apertures_new, 0), + + AMDKFD_IOCTL_DEF(AMDKFD_IOC_ACQUIRE_VM, + kfd_ioctl_acquire_vm, 0), + + AMDKFD_IOCTL_DEF(AMDKFD_IOC_ALLOC_MEMORY_OF_GPU, + kfd_ioctl_alloc_memory_of_gpu, 0), + + AMDKFD_IOCTL_DEF(AMDKFD_IOC_FREE_MEMORY_OF_GPU, + kfd_ioctl_free_memory_of_gpu, 0), + + AMDKFD_IOCTL_DEF(AMDKFD_IOC_MAP_MEMORY_TO_GPU, + kfd_ioctl_map_memory_to_gpu, 0), + + AMDKFD_IOCTL_DEF(AMDKFD_IOC_UNMAP_MEMORY_FROM_GPU, + kfd_ioctl_unmap_memory_from_gpu, 0), + }; #define AMDKFD_CORE_IOCTL_COUNT ARRAY_SIZE(amdkfd_ioctls)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c index 2bc2816..4f126ef 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
@@ -22,10 +22,10 @@ #include <linux/pci.h> #include <linux/acpi.h> -#include <linux/amd-iommu.h> #include "kfd_crat.h" #include "kfd_priv.h" #include "kfd_topology.h" +#include "kfd_iommu.h" /* GPU Processor ID base for dGPUs for which VCRAT needs to be created. * GPU processor ID are expressed with Bit[31]=1. @@ -882,7 +882,7 @@ static int kfd_create_vcrat_image_cpu(void *pcrat_image, size_t *size) crat_table->length = sizeof(struct crat_header); status = acpi_get_table("DSDT", 0, &acpi_table); - if (status == AE_NOT_FOUND) + if (status != AE_OK) pr_warn("DSDT table not found for OEM information\n"); else { crat_table->oem_revision = acpi_table->revision; @@ -1037,15 +1037,11 @@ static int kfd_create_vcrat_image_gpu(void *pcrat_image, struct crat_subtype_generic *sub_type_hdr; struct crat_subtype_computeunit *cu; struct kfd_cu_info cu_info; - struct amd_iommu_device_info iommu_info; int avail_size = *size; uint32_t total_num_of_cu; int num_of_cache_entries = 0; int cache_mem_filled = 0; int ret = 0; - const u32 required_iommu_flags = AMD_IOMMU_DEVICE_FLAG_ATS_SUP | - AMD_IOMMU_DEVICE_FLAG_PRI_SUP | - AMD_IOMMU_DEVICE_FLAG_PASID_SUP; struct kfd_local_mem_info local_mem_info; if (!pcrat_image || avail_size < VCRAT_SIZE_FOR_GPU) @@ -1106,12 +1102,8 @@ static int kfd_create_vcrat_image_gpu(void *pcrat_image, /* Check if this node supports IOMMU. During parsing this flag will * translate to HSA_CAP_ATS_PRESENT */ - iommu_info.flags = 0; - if (amd_iommu_device_info(kdev->pdev, &iommu_info) == 0) { - if ((iommu_info.flags & required_iommu_flags) == - required_iommu_flags) - cu->hsa_capability |= CRAT_CU_FLAGS_IOMMU_PRESENT; - } + if (!kfd_iommu_check_device(kdev)) + cu->hsa_capability |= CRAT_CU_FLAGS_IOMMU_PRESENT; crat_table->length += sub_type_hdr->length; crat_table->total_entries++; @@ -1125,6 +1117,9 @@ static int kfd_create_vcrat_image_gpu(void *pcrat_image, sub_type_hdr = (typeof(sub_type_hdr))((char *)sub_type_hdr + sub_type_hdr->length); + if (debug_largebar) + local_mem_info.local_mem_size_private = 0; + if (local_mem_info.local_mem_size_private == 0) ret = kfd_fill_gpu_memory_affinity(&avail_size, kdev, HSA_MEM_HEAP_TYPE_FB_PUBLIC,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_dbgmgr.c b/drivers/gpu/drm/amd/amdkfd/kfd_dbgmgr.c index 3da25f7b..9d4af96 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgmgr.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgmgr.c
@@ -33,6 +33,7 @@ #include "kfd_pm4_headers_diq.h" #include "kfd_dbgmgr.h" #include "kfd_dbgdev.h" +#include "kfd_device_queue_manager.h" static DEFINE_MUTEX(kfd_dbgmgr_mutex); @@ -83,7 +84,7 @@ bool kfd_dbgmgr_create(struct kfd_dbgmgr **ppmgr, struct kfd_dev *pdev) } /* get actual type of DBGDevice cpsch or not */ - if (sched_policy == KFD_SCHED_POLICY_NO_HWS) + if (pdev->dqm->sched_policy == KFD_SCHED_POLICY_NO_HWS) type = DBGDEV_TYPE_NODIQ; kfd_dbgdev_init(new_buff->dbgdev, pdev, type);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index a8fa33a..334669996 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -20,7 +20,9 @@ * OTHER DEALINGS IN THE SOFTWARE. */ +#if defined(CONFIG_AMD_IOMMU_V2_MODULE) || defined(CONFIG_AMD_IOMMU_V2) #include <linux/amd-iommu.h> +#endif #include <linux/bsearch.h> #include <linux/pci.h> #include <linux/slab.h> @@ -28,9 +30,12 @@ #include "kfd_device_queue_manager.h" #include "kfd_pm4_headers_vi.h" #include "cwsr_trap_handler_gfx8.asm" +#include "kfd_iommu.h" #define MQD_SIZE_ALIGNED 768 +static atomic_t kfd_device_suspended = ATOMIC_INIT(0); +#ifdef KFD_SUPPORT_IOMMU_V2 static const struct kfd_device_info kaveri_device_info = { .asic_family = CHIP_KAVERI, .max_pasid_bits = 16, @@ -41,6 +46,8 @@ static const struct kfd_device_info kaveri_device_info = { .num_of_watch_points = 4, .mqd_size_aligned = MQD_SIZE_ALIGNED, .supports_cwsr = false, + .needs_iommu_device = true, + .needs_pci_atomics = false, }; static const struct kfd_device_info carrizo_device_info = { @@ -53,15 +60,125 @@ static const struct kfd_device_info carrizo_device_info = { .num_of_watch_points = 4, .mqd_size_aligned = MQD_SIZE_ALIGNED, .supports_cwsr = true, + .needs_iommu_device = true, + .needs_pci_atomics = false, }; +#endif + +static const struct kfd_device_info hawaii_device_info = { + .asic_family = CHIP_HAWAII, + .max_pasid_bits = 16, + /* max num of queues for KV.TODO should be a dynamic value */ + .max_no_of_hqd = 24, + .ih_ring_entry_size = 4 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_cik, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = false, + .needs_iommu_device = false, + .needs_pci_atomics = false, +}; + +static const struct kfd_device_info tonga_device_info = { + .asic_family = CHIP_TONGA, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .ih_ring_entry_size = 4 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_cik, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = false, + .needs_iommu_device = false, + .needs_pci_atomics = true, +}; + +static const struct kfd_device_info tonga_vf_device_info = { + .asic_family = CHIP_TONGA, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .ih_ring_entry_size = 4 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_cik, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = false, + .needs_iommu_device = false, + .needs_pci_atomics = false, +}; + +static const struct kfd_device_info fiji_device_info = { + .asic_family = CHIP_FIJI, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .ih_ring_entry_size = 4 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_cik, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = true, + .needs_iommu_device = false, + .needs_pci_atomics = true, +}; + +static const struct kfd_device_info fiji_vf_device_info = { + .asic_family = CHIP_FIJI, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .ih_ring_entry_size = 4 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_cik, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = true, + .needs_iommu_device = false, + .needs_pci_atomics = false, +}; + + +static const struct kfd_device_info polaris10_device_info = { + .asic_family = CHIP_POLARIS10, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .ih_ring_entry_size = 4 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_cik, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = true, + .needs_iommu_device = false, + .needs_pci_atomics = true, +}; + +static const struct kfd_device_info polaris10_vf_device_info = { + .asic_family = CHIP_POLARIS10, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .ih_ring_entry_size = 4 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_cik, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = true, + .needs_iommu_device = false, + .needs_pci_atomics = false, +}; + +static const struct kfd_device_info polaris11_device_info = { + .asic_family = CHIP_POLARIS11, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .ih_ring_entry_size = 4 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_cik, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = true, + .needs_iommu_device = false, + .needs_pci_atomics = true, +}; + struct kfd_deviceid { unsigned short did; const struct kfd_device_info *device_info; }; -/* Please keep this sorted by increasing device id. */ static const struct kfd_deviceid supported_devices[] = { +#ifdef KFD_SUPPORT_IOMMU_V2 { 0x1304, &kaveri_device_info }, /* Kaveri */ { 0x1305, &kaveri_device_info }, /* Kaveri */ { 0x1306, &kaveri_device_info }, /* Kaveri */ @@ -88,7 +205,51 @@ static const struct kfd_deviceid supported_devices[] = { { 0x9874, &carrizo_device_info }, /* Carrizo */ { 0x9875, &carrizo_device_info }, /* Carrizo */ { 0x9876, &carrizo_device_info }, /* Carrizo */ - { 0x9877, &carrizo_device_info } /* Carrizo */ + { 0x9877, &carrizo_device_info }, /* Carrizo */ +#endif + { 0x67A0, &hawaii_device_info }, /* Hawaii */ + { 0x67A1, &hawaii_device_info }, /* Hawaii */ + { 0x67A2, &hawaii_device_info }, /* Hawaii */ + { 0x67A8, &hawaii_device_info }, /* Hawaii */ + { 0x67A9, &hawaii_device_info }, /* Hawaii */ + { 0x67AA, &hawaii_device_info }, /* Hawaii */ + { 0x67B0, &hawaii_device_info }, /* Hawaii */ + { 0x67B1, &hawaii_device_info }, /* Hawaii */ + { 0x67B8, &hawaii_device_info }, /* Hawaii */ + { 0x67B9, &hawaii_device_info }, /* Hawaii */ + { 0x67BA, &hawaii_device_info }, /* Hawaii */ + { 0x67BE, &hawaii_device_info }, /* Hawaii */ + { 0x6920, &tonga_device_info }, /* Tonga */ + { 0x6921, &tonga_device_info }, /* Tonga */ + { 0x6928, &tonga_device_info }, /* Tonga */ + { 0x6929, &tonga_device_info }, /* Tonga */ + { 0x692B, &tonga_device_info }, /* Tonga */ + { 0x692F, &tonga_vf_device_info }, /* Tonga vf */ + { 0x6938, &tonga_device_info }, /* Tonga */ + { 0x6939, &tonga_device_info }, /* Tonga */ + { 0x7300, &fiji_device_info }, /* Fiji */ + { 0x730F, &fiji_vf_device_info }, /* Fiji vf*/ + { 0x67C0, &polaris10_device_info }, /* Polaris10 */ + { 0x67C1, &polaris10_device_info }, /* Polaris10 */ + { 0x67C2, &polaris10_device_info }, /* Polaris10 */ + { 0x67C4, &polaris10_device_info }, /* Polaris10 */ + { 0x67C7, &polaris10_device_info }, /* Polaris10 */ + { 0x67C8, &polaris10_device_info }, /* Polaris10 */ + { 0x67C9, &polaris10_device_info }, /* Polaris10 */ + { 0x67CA, &polaris10_device_info }, /* Polaris10 */ + { 0x67CC, &polaris10_device_info }, /* Polaris10 */ + { 0x67CF, &polaris10_device_info }, /* Polaris10 */ + { 0x67D0, &polaris10_vf_device_info }, /* Polaris10 vf*/ + { 0x67DF, &polaris10_device_info }, /* Polaris10 */ + { 0x67E0, &polaris11_device_info }, /* Polaris11 */ + { 0x67E1, &polaris11_device_info }, /* Polaris11 */ + { 0x67E3, &polaris11_device_info }, /* Polaris11 */ + { 0x67E7, &polaris11_device_info }, /* Polaris11 */ + { 0x67E8, &polaris11_device_info }, /* Polaris11 */ + { 0x67E9, &polaris11_device_info }, /* Polaris11 */ + { 0x67EB, &polaris11_device_info }, /* Polaris11 */ + { 0x67EF, &polaris11_device_info }, /* Polaris11 */ + { 0x67FF, &polaris11_device_info }, /* Polaris11 */ }; static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size, @@ -127,6 +288,21 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, return NULL; } + if (device_info->needs_pci_atomics) { + /* Allow BIF to recode atomics to PCIe 3.0 + * AtomicOps. 32 and 64-bit requests are possible and + * must be supported. + */ + if (pci_enable_atomic_ops_to_root(pdev, + PCI_EXP_DEVCAP2_ATOMIC_COMP32 | + PCI_EXP_DEVCAP2_ATOMIC_COMP64) < 0) { + dev_info(kfd_device, + "skipped device %x:%x, PCI rejects atomics", + pdev->vendor, pdev->device); + return NULL; + } + } + kfd = kzalloc(sizeof(*kfd), GFP_KERNEL); if (!kfd) return NULL; @@ -144,77 +320,6 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, return kfd; } -static bool device_iommu_pasid_init(struct kfd_dev *kfd) -{ - const u32 required_iommu_flags = AMD_IOMMU_DEVICE_FLAG_ATS_SUP | - AMD_IOMMU_DEVICE_FLAG_PRI_SUP | - AMD_IOMMU_DEVICE_FLAG_PASID_SUP; - - struct amd_iommu_device_info iommu_info; - unsigned int pasid_limit; - int err; - - err = amd_iommu_device_info(kfd->pdev, &iommu_info); - if (err < 0) { - dev_err(kfd_device, - "error getting iommu info. is the iommu enabled?\n"); - return false; - } - - if ((iommu_info.flags & required_iommu_flags) != required_iommu_flags) { - dev_err(kfd_device, "error required iommu flags ats %i, pri %i, pasid %i\n", - (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_ATS_SUP) != 0, - (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_PRI_SUP) != 0, - (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_PASID_SUP) - != 0); - return false; - } - - pasid_limit = min_t(unsigned int, - (unsigned int)(1 << kfd->device_info->max_pasid_bits), - iommu_info.max_pasids); - - if (!kfd_set_pasid_limit(pasid_limit)) { - dev_err(kfd_device, "error setting pasid limit\n"); - return false; - } - - return true; -} - -static void iommu_pasid_shutdown_callback(struct pci_dev *pdev, int pasid) -{ - struct kfd_dev *dev = kfd_device_by_pci_dev(pdev); - - if (dev) - kfd_process_iommu_unbind_callback(dev, pasid); -} - -/* - * This function called by IOMMU driver on PPR failure - */ -static int iommu_invalid_ppr_cb(struct pci_dev *pdev, int pasid, - unsigned long address, u16 flags) -{ - struct kfd_dev *dev; - - dev_warn(kfd_device, - "Invalid PPR device %x:%x.%x pasid %d address 0x%lX flags 0x%X", - PCI_BUS_NUM(pdev->devfn), - PCI_SLOT(pdev->devfn), - PCI_FUNC(pdev->devfn), - pasid, - address, - flags); - - dev = kfd_device_by_pci_dev(pdev); - if (!WARN_ON(!dev)) - kfd_signal_iommu_event(dev, pasid, address, - flags & PPR_FAULT_WRITE, flags & PPR_FAULT_EXEC); - - return AMD_IOMMU_INV_PRI_RSP_INVALID; -} - static void kfd_cwsr_init(struct kfd_dev *kfd) { if (cwsr_enable && kfd->device_info->supports_cwsr) { @@ -304,11 +409,9 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd, goto device_queue_manager_error; } - if (!device_iommu_pasid_init(kfd)) { - dev_err(kfd_device, - "Error initializing iommuv2 for device %x:%x\n", - kfd->pdev->vendor, kfd->pdev->device); - goto device_iommu_pasid_error; + if (kfd_iommu_device_init(kfd)) { + dev_err(kfd_device, "Error initializing iommuv2\n"); + goto device_iommu_error; } kfd_cwsr_init(kfd); @@ -323,12 +426,12 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd, kfd->pdev->device); pr_debug("Starting kfd with the following scheduling policy %d\n", - sched_policy); + kfd->dqm->sched_policy); goto out; kfd_resume_error: -device_iommu_pasid_error: +device_iommu_error: device_queue_manager_uninit(kfd->dqm); device_queue_manager_error: kfd_interrupt_exit(kfd); @@ -367,40 +470,45 @@ void kgd2kfd_suspend(struct kfd_dev *kfd) if (!kfd->init_complete) return; + /* For first KFD device suspend all the KFD processes */ + if (atomic_inc_return(&kfd_device_suspended) == 1) + kfd_suspend_all_processes(); + kfd->dqm->ops.stop(kfd->dqm); - kfd_unbind_processes_from_device(kfd); - - amd_iommu_set_invalidate_ctx_cb(kfd->pdev, NULL); - amd_iommu_set_invalid_ppr_cb(kfd->pdev, NULL); - amd_iommu_free_device(kfd->pdev); + kfd_iommu_suspend(kfd); } int kgd2kfd_resume(struct kfd_dev *kfd) { + int ret, count; + if (!kfd->init_complete) return 0; - return kfd_resume(kfd); + ret = kfd_resume(kfd); + if (ret) + return ret; + count = atomic_dec_return(&kfd_device_suspended); + WARN_ONCE(count < 0, "KFD suspend / resume ref. error"); + if (count == 0) + ret = kfd_resume_all_processes(); + + return ret; } static int kfd_resume(struct kfd_dev *kfd) { int err = 0; - unsigned int pasid_limit = kfd_get_pasid_limit(); - err = amd_iommu_init_device(kfd->pdev, pasid_limit); - if (err) - return -ENXIO; - amd_iommu_set_invalidate_ctx_cb(kfd->pdev, - iommu_pasid_shutdown_callback); - amd_iommu_set_invalid_ppr_cb(kfd->pdev, - iommu_invalid_ppr_cb); - - err = kfd_bind_processes_to_device(kfd); - if (err) - goto processes_bind_error; + err = kfd_iommu_resume(kfd); + if (err) { + dev_err(kfd_device, + "Failed to resume IOMMU for device %x:%x\n", + kfd->pdev->vendor, kfd->pdev->device); + return err; + } err = kfd->dqm->ops.start(kfd->dqm); if (err) { @@ -413,9 +521,7 @@ static int kfd_resume(struct kfd_dev *kfd) return err; dqm_start_error: -processes_bind_error: - amd_iommu_free_device(kfd->pdev); - + kfd_iommu_suspend(kfd); return err; } @@ -435,6 +541,54 @@ void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry) spin_unlock(&kfd->interrupt_lock); } +/** kgd2kfd_schedule_evict_and_restore_process - Schedules work queue that will + * prepare for safe eviction of KFD BOs that belong to the specified + * process. + * + * @mm: mm_struct that identifies the specified KFD process + * @fence: eviction fence attached to KFD process BOs + * + */ +int kgd2kfd_schedule_evict_and_restore_process(struct mm_struct *mm, + struct dma_fence *fence) +{ + struct kfd_process *p; + unsigned long active_time; + unsigned long delay_jiffies = msecs_to_jiffies(PROCESS_ACTIVE_TIME_MS); + + if (!fence) + return -EINVAL; + + if (dma_fence_is_signaled(fence)) + return 0; + + p = kfd_lookup_process_by_mm(mm); + if (!p) + return -ENODEV; + + if (fence->seqno == p->last_eviction_seqno) + goto out; + + p->last_eviction_seqno = fence->seqno; + + /* Avoid KFD process starvation. Wait for at least + * PROCESS_ACTIVE_TIME_MS before evicting the process again + */ + active_time = get_jiffies_64() - p->last_restore_timestamp; + if (delay_jiffies > active_time) + delay_jiffies -= active_time; + else + delay_jiffies = 0; + + /* During process initialization eviction_work.dwork is initialized + * to kfd_evict_bo_worker + */ + schedule_delayed_work(&p->eviction_work, delay_jiffies); +out: + kfd_unref_process(p); + return 0; +} + static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size, unsigned int chunk_size) {
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 1bd5f26..d55d29d 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -21,10 +21,11 @@ * */ +#include <linux/ratelimit.h> +#include <linux/printk.h> #include <linux/slab.h> #include <linux/list.h> #include <linux/types.h> -#include <linux/printk.h> #include <linux/bitops.h> #include <linux/sched.h> #include "kfd_priv.h" @@ -118,9 +119,8 @@ static int allocate_vmid(struct device_queue_manager *dqm, if (dqm->vmid_bitmap == 0) return -ENOMEM; - bit = find_first_bit((unsigned long *)&dqm->vmid_bitmap, - dqm->dev->vm_info.vmid_num_kfd); - clear_bit(bit, (unsigned long *)&dqm->vmid_bitmap); + bit = ffs(dqm->vmid_bitmap) - 1; + dqm->vmid_bitmap &= ~(1 << bit); allocated_vmid = bit + dqm->dev->vm_info.first_vmid_kfd; pr_debug("vmid allocation %d\n", allocated_vmid); @@ -130,19 +130,49 @@ static int allocate_vmid(struct device_queue_manager *dqm, set_pasid_vmid_mapping(dqm, q->process->pasid, q->properties.vmid); program_sh_mem_settings(dqm, qpd); + /* qpd->page_table_base is set earlier when register_process() + * is called, i.e. when the first queue is created. + */ + dqm->dev->kfd2kgd->set_vm_context_page_table_base(dqm->dev->kgd, + qpd->vmid, + qpd->page_table_base); + /* invalidate the VM context after pasid and vmid mapping is set up */ + kfd_flush_tlb(qpd_to_pdd(qpd)); + return 0; } +static int flush_texture_cache_nocpsch(struct kfd_dev *kdev, + struct qcm_process_device *qpd) +{ + uint32_t len; + + if (!qpd->ib_kaddr) + return -ENOMEM; + + len = pm_create_release_mem(qpd->ib_base, (uint32_t *)qpd->ib_kaddr); + + return kdev->kfd2kgd->submit_ib(kdev->kgd, KGD_ENGINE_MEC1, qpd->vmid, + qpd->ib_base, (uint32_t *)qpd->ib_kaddr, len); +} + static void deallocate_vmid(struct device_queue_manager *dqm, struct qcm_process_device *qpd, struct queue *q) { int bit = qpd->vmid - dqm->dev->vm_info.first_vmid_kfd; + /* On GFX v7, CP doesn't flush TC at dequeue */ + if (q->device->device_info->asic_family == CHIP_HAWAII) + if (flush_texture_cache_nocpsch(q->device, qpd)) + pr_err("Failed to flush TC\n"); + + kfd_flush_tlb(qpd_to_pdd(qpd)); + /* Release the vmid mapping */ set_pasid_vmid_mapping(dqm, 0, qpd->vmid); - set_bit(bit, (unsigned long *)&dqm->vmid_bitmap); + dqm->vmid_bitmap |= (1 << bit); qpd->vmid = 0; q->properties.vmid = 0; } @@ -170,6 +200,14 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm, goto out_unlock; } q->properties.vmid = qpd->vmid; + /* + * Eviction state logic: we only mark active queues as evicted + * to avoid the overhead of restoring inactive queues later + */ + if (qpd->evicted) + q->properties.is_evicted = (q->properties.queue_size > 0 && + q->properties.queue_percent > 0 && + q->properties.queue_address != 0); q->properties.tba_addr = qpd->tba_addr; q->properties.tma_addr = qpd->tma_addr; @@ -223,12 +261,8 @@ static int allocate_hqd(struct device_queue_manager *dqm, struct queue *q) continue; if (dqm->allocated_queues[pipe] != 0) { - bit = find_first_bit( - (unsigned long *)&dqm->allocated_queues[pipe], - get_queues_per_pipe(dqm)); - - clear_bit(bit, - (unsigned long *)&dqm->allocated_queues[pipe]); + bit = ffs(dqm->allocated_queues[pipe]) - 1; + dqm->allocated_queues[pipe] &= ~(1 << bit); q->pipe = pipe; q->queue = bit; set = true; @@ -249,7 +283,7 @@ static int allocate_hqd(struct device_queue_manager *dqm, struct queue *q) static inline void deallocate_hqd(struct device_queue_manager *dqm, struct queue *q) { - set_bit(q->queue, (unsigned long *)&dqm->allocated_queues[q->pipe]); + dqm->allocated_queues[q->pipe] |= (1 << q->queue); } static int create_compute_queue_nocpsch(struct device_queue_manager *dqm, @@ -371,21 +405,35 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q) { int retval; struct mqd_manager *mqd; + struct kfd_process_device *pdd; bool prev_active = false; mutex_lock(&dqm->lock); + pdd = kfd_get_process_device_data(q->device, q->process); + if (!pdd) { + retval = -ENODEV; + goto out_unlock; + } mqd = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); if (!mqd) { retval = -ENOMEM; goto out_unlock; } + /* + * Eviction state logic: we only mark active queues as evicted + * to avoid the overhead of restoring inactive queues later + */ + if (pdd->qpd.evicted) + q->properties.is_evicted = (q->properties.queue_size > 0 && + q->properties.queue_percent > 0 && + q->properties.queue_address != 0); /* Save previous activity state for counters */ prev_active = q->properties.is_active; /* Make sure the queue is unmapped before updating the MQD */ - if (sched_policy != KFD_SCHED_POLICY_NO_HWS) { + if (dqm->sched_policy != KFD_SCHED_POLICY_NO_HWS) { retval = unmap_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); if (retval) { @@ -417,7 +465,7 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q) else if (!q->properties.is_active && prev_active) dqm->queue_count--; - if (sched_policy != KFD_SCHED_POLICY_NO_HWS) + if (dqm->sched_policy != KFD_SCHED_POLICY_NO_HWS) retval = map_queues_cpsch(dqm); else if (q->properties.is_active && (q->properties.type == KFD_QUEUE_TYPE_COMPUTE || @@ -451,10 +499,193 @@ static struct mqd_manager *get_mqd_manager( return mqd; } +static int evict_process_queues_nocpsch(struct device_queue_manager *dqm, + struct qcm_process_device *qpd) +{ + struct queue *q; + struct mqd_manager *mqd; + struct kfd_process_device *pdd; + int retval = 0; + + mutex_lock(&dqm->lock); + if (qpd->evicted++ > 0) /* already evicted, do nothing */ + goto out; + + pdd = qpd_to_pdd(qpd); + pr_info_ratelimited("Evicting PASID %u queues\n", + pdd->process->pasid); + + /* unactivate all active queues on the qpd */ + list_for_each_entry(q, &qpd->queues_list, list) { + if (!q->properties.is_active) + continue; + mqd = dqm->ops.get_mqd_manager(dqm, + get_mqd_type_from_queue_type(q->properties.type)); + if (!mqd) { /* should not be here */ + pr_err("Cannot evict queue, mqd mgr is NULL\n"); + retval = -ENOMEM; + goto out; + } + q->properties.is_evicted = true; + q->properties.is_active = false; + retval = mqd->destroy_mqd(mqd, q->mqd, + KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN, + KFD_UNMAP_LATENCY_MS, q->pipe, q->queue); + if (retval) + goto out; + dqm->queue_count--; + } + +out: + mutex_unlock(&dqm->lock); + return retval; +} + +static int evict_process_queues_cpsch(struct device_queue_manager *dqm, + struct qcm_process_device *qpd) +{ + struct queue *q; + struct kfd_process_device *pdd; + int retval = 0; + + mutex_lock(&dqm->lock); + if (qpd->evicted++ > 0) /* already evicted, do nothing */ + goto out; + + pdd = qpd_to_pdd(qpd); + pr_info_ratelimited("Evicting PASID %u queues\n", + pdd->process->pasid); + + /* unactivate all active queues on the qpd */ + list_for_each_entry(q, &qpd->queues_list, list) { + if (!q->properties.is_active) + continue; + q->properties.is_evicted = true; + q->properties.is_active = false; + dqm->queue_count--; + } + retval = execute_queues_cpsch(dqm, + qpd->is_debug ? + KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES : + KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); + +out: + mutex_unlock(&dqm->lock); + return retval; +} + +static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, + struct qcm_process_device *qpd) +{ + struct queue *q; + struct mqd_manager *mqd; + struct kfd_process_device *pdd; + uint32_t pd_base; + int retval = 0; + + pdd = qpd_to_pdd(qpd); + /* Retrieve PD base */ + pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); + + mutex_lock(&dqm->lock); + if (WARN_ON_ONCE(!qpd->evicted)) /* already restored, do nothing */ + goto out; + if (qpd->evicted > 1) { /* ref count still > 0, decrement & quit */ + qpd->evicted--; + goto out; + } + + pr_info_ratelimited("Restoring PASID %u queues\n", + pdd->process->pasid); + + /* Update PD Base in QPD */ + qpd->page_table_base = pd_base; + pr_debug("Updated PD address to 0x%08x\n", pd_base); + + if (!list_empty(&qpd->queues_list)) { + dqm->dev->kfd2kgd->set_vm_context_page_table_base( + dqm->dev->kgd, + qpd->vmid, + qpd->page_table_base); + kfd_flush_tlb(pdd); + } + + /* activate all active queues on the qpd */ + list_for_each_entry(q, &qpd->queues_list, list) { + if (!q->properties.is_evicted) + continue; + mqd = dqm->ops.get_mqd_manager(dqm, + get_mqd_type_from_queue_type(q->properties.type)); + if (!mqd) { /* should not be here */ + pr_err("Cannot restore queue, mqd mgr is NULL\n"); + retval = -ENOMEM; + goto out; + } + q->properties.is_evicted = false; + q->properties.is_active = true; + retval = mqd->load_mqd(mqd, q->mqd, q->pipe, + q->queue, &q->properties, + q->process->mm); + if (retval) + goto out; + dqm->queue_count++; + } + qpd->evicted = 0; +out: + mutex_unlock(&dqm->lock); + return retval; +} + +static int restore_process_queues_cpsch(struct device_queue_manager *dqm, + struct qcm_process_device *qpd) +{ + struct queue *q; + struct kfd_process_device *pdd; + uint32_t pd_base; + int retval = 0; + + pdd = qpd_to_pdd(qpd); + /* Retrieve PD base */ + pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); + + mutex_lock(&dqm->lock); + if (WARN_ON_ONCE(!qpd->evicted)) /* already restored, do nothing */ + goto out; + if (qpd->evicted > 1) { /* ref count still > 0, decrement & quit */ + qpd->evicted--; + goto out; + } + + pr_info_ratelimited("Restoring PASID %u queues\n", + pdd->process->pasid); + + /* Update PD Base in QPD */ + qpd->page_table_base = pd_base; + pr_debug("Updated PD address to 0x%08x\n", pd_base); + + /* activate all active queues on the qpd */ + list_for_each_entry(q, &qpd->queues_list, list) { + if (!q->properties.is_evicted) + continue; + q->properties.is_evicted = false; + q->properties.is_active = true; + dqm->queue_count++; + } + retval = execute_queues_cpsch(dqm, + KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); + if (!retval) + qpd->evicted = 0; +out: + mutex_unlock(&dqm->lock); + return retval; +} + static int register_process(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { struct device_process_node *n; + struct kfd_process_device *pdd; + uint32_t pd_base; int retval; n = kzalloc(sizeof(*n), GFP_KERNEL); @@ -463,9 +694,16 @@ static int register_process(struct device_queue_manager *dqm, n->qpd = qpd; + pdd = qpd_to_pdd(qpd); + /* Retrieve PD base */ + pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); + mutex_lock(&dqm->lock); list_add(&n->list, &dqm->queues); + /* Update PD Base in QPD */ + qpd->page_table_base = pd_base; + retval = dqm->asic_ops.update_qpd(dqm, qpd); dqm->processes_count++; @@ -573,11 +811,12 @@ static void uninitialize(struct device_queue_manager *dqm) static int start_nocpsch(struct device_queue_manager *dqm) { init_interrupts(dqm); - return 0; + return pm_init(&dqm->packets, dqm); } static int stop_nocpsch(struct device_queue_manager *dqm) { + pm_uninit(&dqm->packets); return 0; } @@ -589,10 +828,8 @@ static int allocate_sdma_queue(struct device_queue_manager *dqm, if (dqm->sdma_bitmap == 0) return -ENOMEM; - bit = find_first_bit((unsigned long *)&dqm->sdma_bitmap, - CIK_SDMA_QUEUES); - - clear_bit(bit, (unsigned long *)&dqm->sdma_bitmap); + bit = ffs(dqm->sdma_bitmap) - 1; + dqm->sdma_bitmap &= ~(1 << bit); *sdma_queue_id = bit; return 0; @@ -603,7 +840,7 @@ static void deallocate_sdma_queue(struct device_queue_manager *dqm, { if (sdma_queue_id >= CIK_SDMA_QUEUES) return; - set_bit(sdma_queue_id, (unsigned long *)&dqm->sdma_bitmap); + dqm->sdma_bitmap |= (1 << sdma_queue_id); } static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, @@ -840,6 +1077,14 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, retval = -ENOMEM; goto out_deallocate_sdma_queue; } + /* + * Eviction state logic: we only mark active queues as evicted + * to avoid the overhead of restoring inactive queues later + */ + if (qpd->evicted) + q->properties.is_evicted = (q->properties.queue_size > 0 && + q->properties.queue_percent > 0 && + q->properties.queue_address != 0); dqm->asic_ops.init_sdma_vm(dqm, q, qpd); @@ -1103,7 +1348,7 @@ static bool set_cache_memory_policy(struct device_queue_manager *dqm, alternate_aperture_base, alternate_aperture_size); - if ((sched_policy == KFD_SCHED_POLICY_NO_HWS) && (qpd->vmid != 0)) + if ((dqm->sched_policy == KFD_SCHED_POLICY_NO_HWS) && (qpd->vmid != 0)) program_sh_mem_settings(dqm, qpd); pr_debug("sh_mem_config: 0x%x, ape1_base: 0x%x, ape1_limit: 0x%x\n", @@ -1250,8 +1495,24 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev) if (!dqm) return NULL; + switch (dev->device_info->asic_family) { + /* HWS is not available on Hawaii. */ + case CHIP_HAWAII: + /* HWS depends on CWSR for timely dequeue. CWSR is not + * available on Tonga. + * + * FIXME: This argument also applies to Kaveri. + */ + case CHIP_TONGA: + dqm->sched_policy = KFD_SCHED_POLICY_NO_HWS; + break; + default: + dqm->sched_policy = sched_policy; + break; + } + dqm->dev = dev; - switch (sched_policy) { + switch (dqm->sched_policy) { case KFD_SCHED_POLICY_HWS: case KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION: /* initialize dqm for cp scheduling */ @@ -1270,6 +1531,8 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev) dqm->ops.set_cache_memory_policy = set_cache_memory_policy; dqm->ops.set_trap_handler = set_trap_handler; dqm->ops.process_termination = process_termination_cpsch; + dqm->ops.evict_process_queues = evict_process_queues_cpsch; + dqm->ops.restore_process_queues = restore_process_queues_cpsch; break; case KFD_SCHED_POLICY_NO_HWS: /* initialize dqm for no cp scheduling */ @@ -1286,9 +1549,12 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev) dqm->ops.set_cache_memory_policy = set_cache_memory_policy; dqm->ops.set_trap_handler = set_trap_handler; dqm->ops.process_termination = process_termination_nocpsch; + dqm->ops.evict_process_queues = evict_process_queues_nocpsch; + dqm->ops.restore_process_queues = + restore_process_queues_nocpsch; break; default: - pr_err("Invalid scheduling policy %d\n", sched_policy); + pr_err("Invalid scheduling policy %d\n", dqm->sched_policy); goto out_free; } @@ -1300,6 +1566,17 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev) case CHIP_KAVERI: device_queue_manager_init_cik(&dqm->asic_ops); break; + + case CHIP_HAWAII: + device_queue_manager_init_cik_hawaii(&dqm->asic_ops); + break; + + case CHIP_TONGA: + case CHIP_FIJI: + case CHIP_POLARIS10: + case CHIP_POLARIS11: + device_queue_manager_init_vi_tonga(&dqm->asic_ops); + break; default: WARN(1, "Unexpected ASIC family %u", dev->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h index c61b693b..412beff 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
@@ -79,6 +79,10 @@ struct device_process_node { * * @process_termination: Clears all process queues belongs to that device. * + * @evict_process_queues: Evict all active queues of a process + * + * @restore_process_queues: Restore all evicted queues queues of a process + * */ struct device_queue_manager_ops { @@ -129,6 +133,11 @@ struct device_queue_manager_ops { int (*process_termination)(struct device_queue_manager *dqm, struct qcm_process_device *qpd); + + int (*evict_process_queues)(struct device_queue_manager *dqm, + struct qcm_process_device *qpd); + int (*restore_process_queues)(struct device_queue_manager *dqm, + struct qcm_process_device *qpd); }; struct device_queue_manager_asic_ops { @@ -180,12 +189,17 @@ struct device_queue_manager { unsigned int *fence_addr; struct kfd_mem_obj *fence_mem; bool active_runlist; + int sched_policy; }; void device_queue_manager_init_cik( struct device_queue_manager_asic_ops *asic_ops); +void device_queue_manager_init_cik_hawaii( + struct device_queue_manager_asic_ops *asic_ops); void device_queue_manager_init_vi( struct device_queue_manager_asic_ops *asic_ops); +void device_queue_manager_init_vi_tonga( + struct device_queue_manager_asic_ops *asic_ops); void program_sh_mem_settings(struct device_queue_manager *dqm, struct qcm_process_device *qpd); unsigned int get_queues_num(struct device_queue_manager *dqm);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_cik.c index 28e48c9..aed4c21 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_cik.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_cik.c
@@ -34,8 +34,13 @@ static bool set_cache_memory_policy_cik(struct device_queue_manager *dqm, uint64_t alternate_aperture_size); static int update_qpd_cik(struct device_queue_manager *dqm, struct qcm_process_device *qpd); +static int update_qpd_cik_hawaii(struct device_queue_manager *dqm, + struct qcm_process_device *qpd); static void init_sdma_vm(struct device_queue_manager *dqm, struct queue *q, struct qcm_process_device *qpd); +static void init_sdma_vm_hawaii(struct device_queue_manager *dqm, + struct queue *q, + struct qcm_process_device *qpd); void device_queue_manager_init_cik( struct device_queue_manager_asic_ops *asic_ops) @@ -45,6 +50,14 @@ void device_queue_manager_init_cik( asic_ops->init_sdma_vm = init_sdma_vm; } +void device_queue_manager_init_cik_hawaii( + struct device_queue_manager_asic_ops *asic_ops) +{ + asic_ops->set_cache_memory_policy = set_cache_memory_policy_cik; + asic_ops->update_qpd = update_qpd_cik_hawaii; + asic_ops->init_sdma_vm = init_sdma_vm_hawaii; +} + static uint32_t compute_sh_mem_bases_64bit(unsigned int top_address_nybble) { /* In 64-bit mode, we can only control the top 3 bits of the LDS, @@ -132,6 +145,36 @@ static int update_qpd_cik(struct device_queue_manager *dqm, return 0; } +static int update_qpd_cik_hawaii(struct device_queue_manager *dqm, + struct qcm_process_device *qpd) +{ + struct kfd_process_device *pdd; + unsigned int temp; + + pdd = qpd_to_pdd(qpd); + + /* check if sh_mem_config register already configured */ + if (qpd->sh_mem_config == 0) { + qpd->sh_mem_config = + ALIGNMENT_MODE(SH_MEM_ALIGNMENT_MODE_UNALIGNED) | + DEFAULT_MTYPE(MTYPE_NONCACHED) | + APE1_MTYPE(MTYPE_NONCACHED); + qpd->sh_mem_ape1_limit = 0; + qpd->sh_mem_ape1_base = 0; + } + + /* On dGPU we're always in GPUVM64 addressing mode with 64-bit + * aperture addresses. + */ + temp = get_sh_mem_bases_nybble_64(pdd); + qpd->sh_mem_bases = compute_sh_mem_bases_64bit(temp); + + pr_debug("is32bit process: %d sh_mem_bases nybble: 0x%X and register 0x%X\n", + qpd->pqm->process->is_32bit_user_mode, temp, qpd->sh_mem_bases); + + return 0; +} + static void init_sdma_vm(struct device_queue_manager *dqm, struct queue *q, struct qcm_process_device *qpd) { @@ -147,3 +190,16 @@ static void init_sdma_vm(struct device_queue_manager *dqm, struct queue *q, q->properties.sdma_vm_addr = value; } + +static void init_sdma_vm_hawaii(struct device_queue_manager *dqm, + struct queue *q, + struct qcm_process_device *qpd) +{ + /* On dGPU we're always in GPUVM64 addressing mode with 64-bit + * aperture addresses. + */ + q->properties.sdma_vm_addr = + ((get_sh_mem_bases_nybble_64(qpd_to_pdd(qpd))) << + SDMA0_RLC0_VIRTUAL_ADDR__SHARED_BASE__SHIFT) & + SDMA0_RLC0_VIRTUAL_ADDR__SHARED_BASE_MASK; +}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_vi.c index 2fbce57..fd60a11 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_vi.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_vi.c
@@ -33,10 +33,21 @@ static bool set_cache_memory_policy_vi(struct device_queue_manager *dqm, enum cache_policy alternate_policy, void __user *alternate_aperture_base, uint64_t alternate_aperture_size); +static bool set_cache_memory_policy_vi_tonga(struct device_queue_manager *dqm, + struct qcm_process_device *qpd, + enum cache_policy default_policy, + enum cache_policy alternate_policy, + void __user *alternate_aperture_base, + uint64_t alternate_aperture_size); static int update_qpd_vi(struct device_queue_manager *dqm, struct qcm_process_device *qpd); +static int update_qpd_vi_tonga(struct device_queue_manager *dqm, + struct qcm_process_device *qpd); static void init_sdma_vm(struct device_queue_manager *dqm, struct queue *q, struct qcm_process_device *qpd); +static void init_sdma_vm_tonga(struct device_queue_manager *dqm, + struct queue *q, + struct qcm_process_device *qpd); void device_queue_manager_init_vi( struct device_queue_manager_asic_ops *asic_ops) @@ -46,6 +57,14 @@ void device_queue_manager_init_vi( asic_ops->init_sdma_vm = init_sdma_vm; } +void device_queue_manager_init_vi_tonga( + struct device_queue_manager_asic_ops *asic_ops) +{ + asic_ops->set_cache_memory_policy = set_cache_memory_policy_vi_tonga; + asic_ops->update_qpd = update_qpd_vi_tonga; + asic_ops->init_sdma_vm = init_sdma_vm_tonga; +} + static uint32_t compute_sh_mem_bases_64bit(unsigned int top_address_nybble) { /* In 64-bit mode, we can only control the top 3 bits of the LDS, @@ -103,6 +122,33 @@ static bool set_cache_memory_policy_vi(struct device_queue_manager *dqm, return true; } +static bool set_cache_memory_policy_vi_tonga(struct device_queue_manager *dqm, + struct qcm_process_device *qpd, + enum cache_policy default_policy, + enum cache_policy alternate_policy, + void __user *alternate_aperture_base, + uint64_t alternate_aperture_size) +{ + uint32_t default_mtype; + uint32_t ape1_mtype; + + default_mtype = (default_policy == cache_policy_coherent) ? + MTYPE_UC : + MTYPE_NC; + + ape1_mtype = (alternate_policy == cache_policy_coherent) ? + MTYPE_UC : + MTYPE_NC; + + qpd->sh_mem_config = + SH_MEM_ALIGNMENT_MODE_UNALIGNED << + SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT | + default_mtype << SH_MEM_CONFIG__DEFAULT_MTYPE__SHIFT | + ape1_mtype << SH_MEM_CONFIG__APE1_MTYPE__SHIFT; + + return true; +} + static int update_qpd_vi(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { @@ -144,6 +190,40 @@ static int update_qpd_vi(struct device_queue_manager *dqm, return 0; } +static int update_qpd_vi_tonga(struct device_queue_manager *dqm, + struct qcm_process_device *qpd) +{ + struct kfd_process_device *pdd; + unsigned int temp; + + pdd = qpd_to_pdd(qpd); + + /* check if sh_mem_config register already configured */ + if (qpd->sh_mem_config == 0) { + qpd->sh_mem_config = + SH_MEM_ALIGNMENT_MODE_UNALIGNED << + SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT | + MTYPE_UC << + SH_MEM_CONFIG__DEFAULT_MTYPE__SHIFT | + MTYPE_UC << + SH_MEM_CONFIG__APE1_MTYPE__SHIFT; + + qpd->sh_mem_ape1_limit = 0; + qpd->sh_mem_ape1_base = 0; + } + + /* On dGPU we're always in GPUVM64 addressing mode with 64-bit + * aperture addresses. + */ + temp = get_sh_mem_bases_nybble_64(pdd); + qpd->sh_mem_bases = compute_sh_mem_bases_64bit(temp); + + pr_debug("sh_mem_bases nybble: 0x%X and register 0x%X\n", + temp, qpd->sh_mem_bases); + + return 0; +} + static void init_sdma_vm(struct device_queue_manager *dqm, struct queue *q, struct qcm_process_device *qpd) { @@ -159,3 +239,16 @@ static void init_sdma_vm(struct device_queue_manager *dqm, struct queue *q, q->properties.sdma_vm_addr = value; } + +static void init_sdma_vm_tonga(struct device_queue_manager *dqm, + struct queue *q, + struct qcm_process_device *qpd) +{ + /* On dGPU we're always in GPUVM64 addressing mode with 64-bit + * aperture addresses. + */ + q->properties.sdma_vm_addr = + ((get_sh_mem_bases_nybble_64(qpd_to_pdd(qpd))) << + SDMA0_RLC0_VIRTUAL_ADDR__SHARED_BASE__SHIFT) & + SDMA0_RLC0_VIRTUAL_ADDR__SHARED_BASE_MASK; +}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index 93aae5c..4890a90 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
@@ -30,6 +30,7 @@ #include <linux/memory.h> #include "kfd_priv.h" #include "kfd_events.h" +#include "kfd_iommu.h" #include <linux/device.h> /* @@ -51,6 +52,7 @@ struct kfd_event_waiter { struct kfd_signal_page { uint64_t *kernel_address; uint64_t __user *user_address; + bool need_to_free_pages; }; @@ -78,6 +80,7 @@ static struct kfd_signal_page *allocate_signal_page(struct kfd_process *p) KFD_SIGNAL_EVENT_LIMIT * 8); page->kernel_address = backing_store; + page->need_to_free_pages = true; pr_debug("Allocated new event signal page at %p, for process %p\n", page, p); @@ -268,8 +271,9 @@ static void shutdown_signal_page(struct kfd_process *p) struct kfd_signal_page *page = p->signal_page; if (page) { - free_pages((unsigned long)page->kernel_address, - get_order(KFD_SIGNAL_EVENT_LIMIT * 8)); + if (page->need_to_free_pages) + free_pages((unsigned long)page->kernel_address, + get_order(KFD_SIGNAL_EVENT_LIMIT * 8)); kfree(page); } } @@ -291,6 +295,30 @@ static bool event_can_be_cpu_signaled(const struct kfd_event *ev) return ev->type == KFD_EVENT_TYPE_SIGNAL; } +int kfd_event_page_set(struct kfd_process *p, void *kernel_address, + uint64_t size) +{ + struct kfd_signal_page *page; + + if (p->signal_page) + return -EBUSY; + + page = kzalloc(sizeof(*page), GFP_KERNEL); + if (!page) + return -ENOMEM; + + /* Initialize all events to unsignaled */ + memset(kernel_address, (uint8_t) UNSIGNALED_EVENT_SLOT, + KFD_SIGNAL_EVENT_LIMIT * 8); + + page->kernel_address = kernel_address; + + p->signal_page = page; + p->signal_mapped_size = size; + + return 0; +} + int kfd_event_create(struct file *devkfd, struct kfd_process *p, uint32_t event_type, bool auto_reset, uint32_t node_id, uint32_t *event_id, uint32_t *event_trigger_data, @@ -837,6 +865,7 @@ static void lookup_events_by_type_and_signal(struct kfd_process *p, } } +#ifdef KFD_SUPPORT_IOMMU_V2 void kfd_signal_iommu_event(struct kfd_dev *dev, unsigned int pasid, unsigned long address, bool is_write_requested, bool is_execute_requested) @@ -905,6 +934,7 @@ void kfd_signal_iommu_event(struct kfd_dev *dev, unsigned int pasid, mutex_unlock(&p->event_mutex); kfd_unref_process(p); } +#endif /* KFD_SUPPORT_IOMMU_V2 */ void kfd_signal_hw_exception_event(unsigned int pasid) {
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c index 7377513..66852de4 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
@@ -278,21 +278,28 @@ #define MAKE_GPUVM_APP_BASE(gpu_num) \ (((uint64_t)(gpu_num) << 61) + 0x1000000000000L) -#define MAKE_GPUVM_APP_LIMIT(base) \ - (((uint64_t)(base) & \ - 0xFFFFFF0000000000UL) | 0xFFFFFFFFFFL) +#define MAKE_GPUVM_APP_LIMIT(base, size) \ + (((uint64_t)(base) & 0xFFFFFF0000000000UL) + (size) - 1) -#define MAKE_SCRATCH_APP_BASE(gpu_num) \ - (((uint64_t)(gpu_num) << 61) + 0x100000000L) +#define MAKE_SCRATCH_APP_BASE() \ + (((uint64_t)(0x1UL) << 61) + 0x100000000L) #define MAKE_SCRATCH_APP_LIMIT(base) \ (((uint64_t)base & 0xFFFFFFFF00000000UL) | 0xFFFFFFFF) -#define MAKE_LDS_APP_BASE(gpu_num) \ - (((uint64_t)(gpu_num) << 61) + 0x0) +#define MAKE_LDS_APP_BASE() \ + (((uint64_t)(0x1UL) << 61) + 0x0) #define MAKE_LDS_APP_LIMIT(base) \ (((uint64_t)(base) & 0xFFFFFFFF00000000UL) | 0xFFFFFFFF) +/* User mode manages most of the SVM aperture address space. The low + * 16MB are reserved for kernel use (CWSR trap handler and kernel IB + * for now). + */ +#define SVM_USER_BASE 0x1000000ull +#define SVM_CWSR_BASE (SVM_USER_BASE - KFD_CWSR_TBA_TMA_SIZE) +#define SVM_IB_BASE (SVM_CWSR_BASE - PAGE_SIZE) + int kfd_init_apertures(struct kfd_process *process) { uint8_t id = 0; @@ -314,7 +321,7 @@ int kfd_init_apertures(struct kfd_process *process) return -1; } /* - * For 64 bit process aperture will be statically reserved in + * For 64 bit process apertures will be statically reserved in * the x86_64 non canonical process address space * amdkfd doesn't currently support apertures for 32 bit process */ @@ -323,23 +330,35 @@ int kfd_init_apertures(struct kfd_process *process) pdd->gpuvm_base = pdd->gpuvm_limit = 0; pdd->scratch_base = pdd->scratch_limit = 0; } else { - /* - * node id couldn't be 0 - the three MSB bits of - * aperture shoudn't be 0 + /* Same LDS and scratch apertures can be used + * on all GPUs. This allows using more dGPUs + * than placement options for apertures. */ - pdd->lds_base = MAKE_LDS_APP_BASE(id + 1); - + pdd->lds_base = MAKE_LDS_APP_BASE(); pdd->lds_limit = MAKE_LDS_APP_LIMIT(pdd->lds_base); - pdd->gpuvm_base = MAKE_GPUVM_APP_BASE(id + 1); - - pdd->gpuvm_limit = - MAKE_GPUVM_APP_LIMIT(pdd->gpuvm_base); - - pdd->scratch_base = MAKE_SCRATCH_APP_BASE(id + 1); - + pdd->scratch_base = MAKE_SCRATCH_APP_BASE(); pdd->scratch_limit = MAKE_SCRATCH_APP_LIMIT(pdd->scratch_base); + + if (dev->device_info->needs_iommu_device) { + /* APUs: GPUVM aperture in + * non-canonical address space + */ + pdd->gpuvm_base = MAKE_GPUVM_APP_BASE(id + 1); + pdd->gpuvm_limit = MAKE_GPUVM_APP_LIMIT( + pdd->gpuvm_base, + dev->shared_resources.gpuvm_size); + } else { + /* dGPUs: SVM aperture starting at 0 + * with small reserved space for kernel + */ + pdd->gpuvm_base = SVM_USER_BASE; + pdd->gpuvm_limit = + dev->shared_resources.gpuvm_size - 1; + pdd->qpd.cwsr_base = SVM_CWSR_BASE; + pdd->qpd.ib_base = SVM_IB_BASE; + } } dev_dbg(kfd_device, "node id %u\n", id);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c new file mode 100644 index 0000000..c718179 --- /dev/null +++ b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c
@@ -0,0 +1,357 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include <linux/printk.h> +#include <linux/device.h> +#include <linux/slab.h> +#include <linux/pci.h> +#include <linux/amd-iommu.h> +#include "kfd_priv.h" +#include "kfd_dbgmgr.h" +#include "kfd_topology.h" +#include "kfd_iommu.h" + +static const u32 required_iommu_flags = AMD_IOMMU_DEVICE_FLAG_ATS_SUP | + AMD_IOMMU_DEVICE_FLAG_PRI_SUP | + AMD_IOMMU_DEVICE_FLAG_PASID_SUP; + +/** kfd_iommu_check_device - Check whether IOMMU is available for device + */ +int kfd_iommu_check_device(struct kfd_dev *kfd) +{ + struct amd_iommu_device_info iommu_info; + int err; + + if (!kfd->device_info->needs_iommu_device) + return -ENODEV; + + iommu_info.flags = 0; + err = amd_iommu_device_info(kfd->pdev, &iommu_info); + if (err) + return err; + + if ((iommu_info.flags & required_iommu_flags) != required_iommu_flags) + return -ENODEV; + + return 0; +} + +/** kfd_iommu_device_init - Initialize IOMMU for device + */ +int kfd_iommu_device_init(struct kfd_dev *kfd) +{ + struct amd_iommu_device_info iommu_info; + unsigned int pasid_limit; + int err; + + if (!kfd->device_info->needs_iommu_device) + return 0; + + iommu_info.flags = 0; + err = amd_iommu_device_info(kfd->pdev, &iommu_info); + if (err < 0) { + dev_err(kfd_device, + "error getting iommu info. is the iommu enabled?\n"); + return -ENODEV; + } + + if ((iommu_info.flags & required_iommu_flags) != required_iommu_flags) { + dev_err(kfd_device, + "error required iommu flags ats %i, pri %i, pasid %i\n", + (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_ATS_SUP) != 0, + (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_PRI_SUP) != 0, + (iommu_info.flags & AMD_IOMMU_DEVICE_FLAG_PASID_SUP) + != 0); + return -ENODEV; + } + + pasid_limit = min_t(unsigned int, + (unsigned int)(1 << kfd->device_info->max_pasid_bits), + iommu_info.max_pasids); + + if (!kfd_set_pasid_limit(pasid_limit)) { + dev_err(kfd_device, "error setting pasid limit\n"); + return -EBUSY; + } + + return 0; +} + +/** kfd_iommu_bind_process_to_device - Have the IOMMU bind a process + * + * Binds the given process to the given device using its PASID. This + * enables IOMMUv2 address translation for the process on the device. + * + * This function assumes that the process mutex is held. + */ +int kfd_iommu_bind_process_to_device(struct kfd_process_device *pdd) +{ + struct kfd_dev *dev = pdd->dev; + struct kfd_process *p = pdd->process; + int err; + + if (!dev->device_info->needs_iommu_device || pdd->bound == PDD_BOUND) + return 0; + + if (unlikely(pdd->bound == PDD_BOUND_SUSPENDED)) { + pr_err("Binding PDD_BOUND_SUSPENDED pdd is unexpected!\n"); + return -EINVAL; + } + + err = amd_iommu_bind_pasid(dev->pdev, p->pasid, p->lead_thread); + if (!err) + pdd->bound = PDD_BOUND; + + return err; +} + +/** kfd_iommu_unbind_process - Unbind process from all devices + * + * This removes all IOMMU device bindings of the process. To be used + * before process termination. + */ +void kfd_iommu_unbind_process(struct kfd_process *p) +{ + struct kfd_process_device *pdd; + + list_for_each_entry(pdd, &p->per_device_data, per_device_list) + if (pdd->bound == PDD_BOUND) + amd_iommu_unbind_pasid(pdd->dev->pdev, p->pasid); +} + +/* Callback for process shutdown invoked by the IOMMU driver */ +static void iommu_pasid_shutdown_callback(struct pci_dev *pdev, int pasid) +{ + struct kfd_dev *dev = kfd_device_by_pci_dev(pdev); + struct kfd_process *p; + struct kfd_process_device *pdd; + + if (!dev) + return; + + /* + * Look for the process that matches the pasid. If there is no such + * process, we either released it in amdkfd's own notifier, or there + * is a bug. Unfortunately, there is no way to tell... + */ + p = kfd_lookup_process_by_pasid(pasid); + if (!p) + return; + + pr_debug("Unbinding process %d from IOMMU\n", pasid); + + mutex_lock(kfd_get_dbgmgr_mutex()); + + if (dev->dbgmgr && dev->dbgmgr->pasid == p->pasid) { + if (!kfd_dbgmgr_unregister(dev->dbgmgr, p)) { + kfd_dbgmgr_destroy(dev->dbgmgr); + dev->dbgmgr = NULL; + } + } + + mutex_unlock(kfd_get_dbgmgr_mutex()); + + mutex_lock(&p->mutex); + + pdd = kfd_get_process_device_data(dev, p); + if (pdd) + /* For GPU relying on IOMMU, we need to dequeue here + * when PASID is still bound. + */ + kfd_process_dequeue_from_device(pdd); + + mutex_unlock(&p->mutex); + + kfd_unref_process(p); +} + +/* This function called by IOMMU driver on PPR failure */ +static int iommu_invalid_ppr_cb(struct pci_dev *pdev, int pasid, + unsigned long address, u16 flags) +{ + struct kfd_dev *dev; + + dev_warn(kfd_device, + "Invalid PPR device %x:%x.%x pasid %d address 0x%lX flags 0x%X", + PCI_BUS_NUM(pdev->devfn), + PCI_SLOT(pdev->devfn), + PCI_FUNC(pdev->devfn), + pasid, + address, + flags); + + dev = kfd_device_by_pci_dev(pdev); + if (!WARN_ON(!dev)) + kfd_signal_iommu_event(dev, pasid, address, + flags & PPR_FAULT_WRITE, flags & PPR_FAULT_EXEC); + + return AMD_IOMMU_INV_PRI_RSP_INVALID; +} + +/* + * Bind processes do the device that have been temporarily unbound + * (PDD_BOUND_SUSPENDED) in kfd_unbind_processes_from_device. + */ +static int kfd_bind_processes_to_device(struct kfd_dev *kfd) +{ + struct kfd_process_device *pdd; + struct kfd_process *p; + unsigned int temp; + int err = 0; + + int idx = srcu_read_lock(&kfd_processes_srcu); + + hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { + mutex_lock(&p->mutex); + pdd = kfd_get_process_device_data(kfd, p); + + if (WARN_ON(!pdd) || pdd->bound != PDD_BOUND_SUSPENDED) { + mutex_unlock(&p->mutex); + continue; + } + + err = amd_iommu_bind_pasid(kfd->pdev, p->pasid, + p->lead_thread); + if (err < 0) { + pr_err("Unexpected pasid %d binding failure\n", + p->pasid); + mutex_unlock(&p->mutex); + break; + } + + pdd->bound = PDD_BOUND; + mutex_unlock(&p->mutex); + } + + srcu_read_unlock(&kfd_processes_srcu, idx); + + return err; +} + +/* + * Mark currently bound processes as PDD_BOUND_SUSPENDED. These + * processes will be restored to PDD_BOUND state in + * kfd_bind_processes_to_device. + */ +static void kfd_unbind_processes_from_device(struct kfd_dev *kfd) +{ + struct kfd_process_device *pdd; + struct kfd_process *p; + unsigned int temp; + + int idx = srcu_read_lock(&kfd_processes_srcu); + + hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { + mutex_lock(&p->mutex); + pdd = kfd_get_process_device_data(kfd, p); + + if (WARN_ON(!pdd)) { + mutex_unlock(&p->mutex); + continue; + } + + if (pdd->bound == PDD_BOUND) + pdd->bound = PDD_BOUND_SUSPENDED; + mutex_unlock(&p->mutex); + } + + srcu_read_unlock(&kfd_processes_srcu, idx); +} + +/** kfd_iommu_suspend - Prepare IOMMU for suspend + * + * This unbinds processes from the device and disables the IOMMU for + * the device. + */ +void kfd_iommu_suspend(struct kfd_dev *kfd) +{ + if (!kfd->device_info->needs_iommu_device) + return; + + kfd_unbind_processes_from_device(kfd); + + amd_iommu_set_invalidate_ctx_cb(kfd->pdev, NULL); + amd_iommu_set_invalid_ppr_cb(kfd->pdev, NULL); + amd_iommu_free_device(kfd->pdev); +} + +/** kfd_iommu_resume - Restore IOMMU after resume + * + * This reinitializes the IOMMU for the device and re-binds previously + * suspended processes to the device. + */ +int kfd_iommu_resume(struct kfd_dev *kfd) +{ + unsigned int pasid_limit; + int err; + + if (!kfd->device_info->needs_iommu_device) + return 0; + + pasid_limit = kfd_get_pasid_limit(); + + err = amd_iommu_init_device(kfd->pdev, pasid_limit); + if (err) + return -ENXIO; + + amd_iommu_set_invalidate_ctx_cb(kfd->pdev, + iommu_pasid_shutdown_callback); + amd_iommu_set_invalid_ppr_cb(kfd->pdev, + iommu_invalid_ppr_cb); + + err = kfd_bind_processes_to_device(kfd); + if (err) { + amd_iommu_set_invalidate_ctx_cb(kfd->pdev, NULL); + amd_iommu_set_invalid_ppr_cb(kfd->pdev, NULL); + amd_iommu_free_device(kfd->pdev); + return err; + } + + return 0; +} + +extern bool amd_iommu_pc_supported(void); +extern u8 amd_iommu_pc_get_max_banks(u16 devid); +extern u8 amd_iommu_pc_get_max_counters(u16 devid); + +/** kfd_iommu_add_perf_counters - Add IOMMU performance counters to topology + */ +int kfd_iommu_add_perf_counters(struct kfd_topology_device *kdev) +{ + struct kfd_perf_properties *props; + + if (!(kdev->node_props.capability & HSA_CAP_ATS_PRESENT)) + return 0; + + if (!amd_iommu_pc_supported()) + return 0; + + props = kfd_alloc_struct(props); + if (!props) + return -ENOMEM; + strcpy(props->block_name, "iommu"); + props->max_concurrent = amd_iommu_pc_get_max_banks(0) * + amd_iommu_pc_get_max_counters(0); /* assume one iommu */ + list_add_tail(&props->list, &kdev->perf_props); + + return 0; +}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.h b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.h new file mode 100644 index 0000000..dd23d9f --- /dev/null +++ b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.h
@@ -0,0 +1,78 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#ifndef __KFD_IOMMU_H__ +#define __KFD_IOMMU_H__ + +#if defined(CONFIG_AMD_IOMMU_V2_MODULE) || defined(CONFIG_AMD_IOMMU_V2) + +#define KFD_SUPPORT_IOMMU_V2 + +int kfd_iommu_check_device(struct kfd_dev *kfd); +int kfd_iommu_device_init(struct kfd_dev *kfd); + +int kfd_iommu_bind_process_to_device(struct kfd_process_device *pdd); +void kfd_iommu_unbind_process(struct kfd_process *p); + +void kfd_iommu_suspend(struct kfd_dev *kfd); +int kfd_iommu_resume(struct kfd_dev *kfd); + +int kfd_iommu_add_perf_counters(struct kfd_topology_device *kdev); + +#else + +static inline int kfd_iommu_check_device(struct kfd_dev *kfd) +{ + return -ENODEV; +} +static inline int kfd_iommu_device_init(struct kfd_dev *kfd) +{ + return 0; +} + +static inline int kfd_iommu_bind_process_to_device( + struct kfd_process_device *pdd) +{ + return 0; +} +static inline void kfd_iommu_unbind_process(struct kfd_process *p) +{ + /* empty */ +} + +static inline void kfd_iommu_suspend(struct kfd_dev *kfd) +{ + /* empty */ +} +static inline int kfd_iommu_resume(struct kfd_dev *kfd) +{ + return 0; +} + +static inline int kfd_iommu_add_perf_counters(struct kfd_topology_device *kdev) +{ + return 0; +} + +#endif /* defined(CONFIG_AMD_IOMMU_V2) */ + +#endif /* __KFD_IOMMU_H__ */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c index 5dc6567..69f4964 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
@@ -297,10 +297,15 @@ struct kernel_queue *kernel_queue_init(struct kfd_dev *dev, switch (dev->device_info->asic_family) { case CHIP_CARRIZO: + case CHIP_TONGA: + case CHIP_FIJI: + case CHIP_POLARIS10: + case CHIP_POLARIS11: kernel_queue_init_vi(&kq->ops_asic_specific); break; case CHIP_KAVERI: + case CHIP_HAWAII: kernel_queue_init_cik(&kq->ops_asic_specific); break; default:
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c index 3ac72be..e0c07d2 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_module.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
@@ -43,6 +43,8 @@ static const struct kgd2kfd_calls kgd2kfd = { .interrupt = kgd2kfd_interrupt, .suspend = kgd2kfd_suspend, .resume = kgd2kfd_resume, + .schedule_evict_and_restore_process = + kgd2kfd_schedule_evict_and_restore_process, }; int sched_policy = KFD_SCHED_POLICY_HWS; @@ -69,6 +71,11 @@ module_param(send_sigterm, int, 0444); MODULE_PARM_DESC(send_sigterm, "Send sigterm to HSA process on unhandled exception (0 = disable, 1 = enable)"); +int debug_largebar; +module_param(debug_largebar, int, 0444); +MODULE_PARM_DESC(debug_largebar, + "Debug large-bar flag used to simulate large-bar capability on non-large bar machine (0 = disable, 1 = enable)"); + int ignore_crat; module_param(ignore_crat, int, 0444); MODULE_PARM_DESC(ignore_crat, @@ -126,7 +133,9 @@ static int __init kfd_module_init(void) if (err < 0) goto err_topology; - kfd_process_create_wq(); + err = kfd_process_create_wq(); + if (err < 0) + goto err_create_wq; kfd_debugfs_init(); @@ -136,6 +145,8 @@ static int __init kfd_module_init(void) return 0; +err_create_wq: + kfd_topology_shutdown(); err_topology: kfd_chardev_exit(); err_ioctl:
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c index dfd260e..ee7061e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
@@ -29,8 +29,15 @@ struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type, switch (dev->device_info->asic_family) { case CHIP_KAVERI: return mqd_manager_init_cik(type, dev); + case CHIP_HAWAII: + return mqd_manager_init_cik_hawaii(type, dev); case CHIP_CARRIZO: return mqd_manager_init_vi(type, dev); + case CHIP_TONGA: + case CHIP_FIJI: + case CHIP_POLARIS10: + case CHIP_POLARIS11: + return mqd_manager_init_vi_tonga(type, dev); default: WARN(1, "Unexpected ASIC family %u", dev->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c index f8ef4a0..c00c325 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
@@ -170,14 +170,19 @@ static int load_mqd_sdma(struct mqd_manager *mm, void *mqd, mms); } -static int update_mqd(struct mqd_manager *mm, void *mqd, - struct queue_properties *q) +static int __update_mqd(struct mqd_manager *mm, void *mqd, + struct queue_properties *q, unsigned int atc_bit) { struct cik_mqd *m; m = get_mqd(mqd); m->cp_hqd_pq_control = DEFAULT_RPTR_BLOCK_SIZE | - DEFAULT_MIN_AVAIL_SIZE | PQ_ATC_EN; + DEFAULT_MIN_AVAIL_SIZE; + m->cp_hqd_ib_control = DEFAULT_MIN_IB_AVAIL_SIZE; + if (atc_bit) { + m->cp_hqd_pq_control |= PQ_ATC_EN; + m->cp_hqd_ib_control |= IB_ATC_EN; + } /* * Calculating queue size which is log base 2 of actual queue size -1 @@ -197,11 +202,24 @@ static int update_mqd(struct mqd_manager *mm, void *mqd, q->is_active = (q->queue_size > 0 && q->queue_address != 0 && - q->queue_percent > 0); + q->queue_percent > 0 && + !q->is_evicted); return 0; } +static int update_mqd(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + return __update_mqd(mm, mqd, q, 1); +} + +static int update_mqd_hawaii(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + return __update_mqd(mm, mqd, q, 0); +} + static int update_mqd_sdma(struct mqd_manager *mm, void *mqd, struct queue_properties *q) { @@ -228,7 +246,8 @@ static int update_mqd_sdma(struct mqd_manager *mm, void *mqd, q->is_active = (q->queue_size > 0 && q->queue_address != 0 && - q->queue_percent > 0); + q->queue_percent > 0 && + !q->is_evicted); return 0; } @@ -360,7 +379,8 @@ static int update_mqd_hiq(struct mqd_manager *mm, void *mqd, q->is_active = (q->queue_size > 0 && q->queue_address != 0 && - q->queue_percent > 0); + q->queue_percent > 0 && + !q->is_evicted); return 0; } @@ -441,3 +461,15 @@ struct mqd_manager *mqd_manager_init_cik(enum KFD_MQD_TYPE type, return mqd; } +struct mqd_manager *mqd_manager_init_cik_hawaii(enum KFD_MQD_TYPE type, + struct kfd_dev *dev) +{ + struct mqd_manager *mqd; + + mqd = mqd_manager_init_cik(type, dev); + if (!mqd) + return NULL; + if ((type == KFD_MQD_TYPE_CP) || (type == KFD_MQD_TYPE_COMPUTE)) + mqd->update_mqd = update_mqd_hawaii; + return mqd; +}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c index 971aec0..89e4242 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
@@ -151,6 +151,8 @@ static int __update_mqd(struct mqd_manager *mm, void *mqd, m->cp_hqd_pq_rptr_report_addr_lo = lower_32_bits((uint64_t)q->read_ptr); m->cp_hqd_pq_rptr_report_addr_hi = upper_32_bits((uint64_t)q->read_ptr); + m->cp_hqd_pq_wptr_poll_addr_lo = lower_32_bits((uint64_t)q->write_ptr); + m->cp_hqd_pq_wptr_poll_addr_hi = upper_32_bits((uint64_t)q->write_ptr); m->cp_hqd_pq_doorbell_control = q->doorbell_off << @@ -196,7 +198,8 @@ static int __update_mqd(struct mqd_manager *mm, void *mqd, q->is_active = (q->queue_size > 0 && q->queue_address != 0 && - q->queue_percent > 0); + q->queue_percent > 0 && + !q->is_evicted); return 0; } @@ -208,6 +211,12 @@ static int update_mqd(struct mqd_manager *mm, void *mqd, return __update_mqd(mm, mqd, q, MTYPE_CC, 1); } +static int update_mqd_tonga(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + return __update_mqd(mm, mqd, q, MTYPE_UC, 0); +} + static int destroy_mqd(struct mqd_manager *mm, void *mqd, enum kfd_preempt_type type, unsigned int timeout, uint32_t pipe_id, @@ -334,7 +343,8 @@ static int update_mqd_sdma(struct mqd_manager *mm, void *mqd, q->is_active = (q->queue_size > 0 && q->queue_address != 0 && - q->queue_percent > 0); + q->queue_percent > 0 && + !q->is_evicted); return 0; } @@ -432,3 +442,16 @@ struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type, return mqd; } + +struct mqd_manager *mqd_manager_init_vi_tonga(enum KFD_MQD_TYPE type, + struct kfd_dev *dev) +{ + struct mqd_manager *mqd; + + mqd = mqd_manager_init_vi(type, dev); + if (!mqd) + return NULL; + if ((type == KFD_MQD_TYPE_CP) || (type == KFD_MQD_TYPE_COMPUTE)) + mqd->update_mqd = update_mqd_tonga; + return mqd; +}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c index 0c3bc00..89ba4c6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
@@ -355,6 +355,43 @@ static int pm_create_runlist_ib(struct packet_manager *pm, return retval; } +/* pm_create_release_mem - Create a RELEASE_MEM packet and return the size + * of this packet + * @gpu_addr - GPU address of the packet. It's a virtual address. + * @buffer - buffer to fill up with the packet. It's a CPU kernel pointer + * Return - length of the packet + */ +uint32_t pm_create_release_mem(uint64_t gpu_addr, uint32_t *buffer) +{ + struct pm4_mec_release_mem *packet; + + WARN_ON(!buffer); + + packet = (struct pm4_mec_release_mem *)buffer; + memset(buffer, 0, sizeof(*packet)); + + packet->header.u32All = build_pm4_header(IT_RELEASE_MEM, + sizeof(*packet)); + + packet->bitfields2.event_type = CACHE_FLUSH_AND_INV_TS_EVENT; + packet->bitfields2.event_index = event_index___release_mem__end_of_pipe; + packet->bitfields2.tcl1_action_ena = 1; + packet->bitfields2.tc_action_ena = 1; + packet->bitfields2.cache_policy = cache_policy___release_mem__lru; + packet->bitfields2.atc = 0; + + packet->bitfields3.data_sel = data_sel___release_mem__send_32_bit_low; + packet->bitfields3.int_sel = + int_sel___release_mem__send_interrupt_after_write_confirm; + + packet->bitfields4.address_lo_32b = (gpu_addr & 0xffffffff) >> 2; + packet->address_hi = upper_32_bits(gpu_addr); + + packet->data_lo = 0; + + return sizeof(*packet) / sizeof(unsigned int); +} + int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm) { pm->dqm = dqm;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 0bedcf9..96a9cc0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -105,6 +105,12 @@ extern int cwsr_enable; extern int send_sigterm; /* + * This kernel module is used to simulate large bar machine on non-large bar + * enabled machines. + */ +extern int debug_largebar; + +/* * Ignore CRAT table during KFD initialization, can be used to work around * broken CRAT tables on some AMD systems */ @@ -158,6 +164,8 @@ struct kfd_device_info { uint8_t num_of_watch_points; uint16_t mqd_size_aligned; bool supports_cwsr; + bool needs_iommu_device; + bool needs_pci_atomics; }; struct kfd_mem_obj { @@ -333,7 +341,11 @@ enum kfd_queue_format { * @is_interop: Defines if this is a interop queue. Interop queue means that * the queue can access both graphics and compute resources. * - * @is_active: Defines if the queue is active or not. + * @is_evicted: Defines if the queue is evicted. Only active queues + * are evicted, rendering them inactive. + * + * @is_active: Defines if the queue is active or not. @is_active and + * @is_evicted are protected by the DQM lock. * * @vmid: If the scheduling mode is no cp scheduling the field defines the vmid * of the queue. @@ -355,6 +367,7 @@ struct queue_properties { uint32_t __iomem *doorbell_ptr; uint32_t doorbell_off; bool is_interop; + bool is_evicted; bool is_active; /* Not relevant for user mode queues in cp scheduling */ unsigned int vmid; @@ -458,6 +471,7 @@ struct qcm_process_device { unsigned int queue_count; unsigned int vmid; bool is_debug; + unsigned int evicted; /* eviction counter, 0=active */ /* This flag tells if we should reset all wavefronts on * process termination @@ -480,10 +494,34 @@ struct qcm_process_device { /* CWSR memory */ void *cwsr_kaddr; + uint64_t cwsr_base; uint64_t tba_addr; uint64_t tma_addr; + + /* IB memory */ + uint64_t ib_base; + void *ib_kaddr; }; +/* KFD Memory Eviction */ + +/* Approx. wait time before attempting to restore evicted BOs */ +#define PROCESS_RESTORE_TIME_MS 100 +/* Approx. back off time if restore fails due to lack of memory */ +#define PROCESS_BACK_OFF_TIME_MS 100 +/* Approx. time before evicting the process again */ +#define PROCESS_ACTIVE_TIME_MS 10 + +int kgd2kfd_schedule_evict_and_restore_process(struct mm_struct *mm, + struct dma_fence *fence); + +/* 8 byte handle containing GPU ID in the most significant 4 bytes and + * idr_handle in the least significant 4 bytes + */ +#define MAKE_HANDLE(gpu_id, idr_handle) \ + (((uint64_t)(gpu_id) << 32) + idr_handle) +#define GET_GPU_ID(handle) (handle >> 32) +#define GET_IDR_HANDLE(handle) (handle & 0xFFFFFFFF) enum kfd_pdd_bound { PDD_UNBOUND = 0, @@ -516,8 +554,12 @@ struct kfd_process_device { uint64_t scratch_base; uint64_t scratch_limit; - /* Is this process/pasid bound to this device? (amd_iommu_bind_pasid) */ - enum kfd_pdd_bound bound; + /* VM context for GPUVM allocations */ + struct file *drm_file; + void *vm; + + /* GPUVM allocations storage */ + struct idr alloc_idr; /* Flag used to tell the pdd has dequeued from the dqm. * This is used to prevent dev->dqm->ops.process_termination() from @@ -525,6 +567,9 @@ struct kfd_process_device { * function. */ bool already_dequeued; + + /* Is this process/pasid bound to this device? (amd_iommu_bind_pasid) */ + enum kfd_pdd_bound bound; }; #define qpd_to_pdd(x) container_of(x, struct kfd_process_device, qpd) @@ -587,8 +632,30 @@ struct kfd_process { size_t signal_mapped_size; size_t signal_event_count; bool signal_event_limit_reached; + + /* Information used for memory eviction */ + void *kgd_process_info; + /* Eviction fence that is attached to all the BOs of this process. The + * fence will be triggered during eviction and new one will be created + * during restore + */ + struct dma_fence *ef; + + /* Work items for evicting and restoring BOs */ + struct delayed_work eviction_work; + struct delayed_work restore_work; + /* seqno of the last scheduled eviction */ + unsigned int last_eviction_seqno; + /* Approx. the last timestamp (in jiffies) when the process was + * restored after an eviction + */ + unsigned long last_restore_timestamp; }; +#define KFD_PROCESS_TABLE_SIZE 5 /* bits: 32 entries */ +extern DECLARE_HASHTABLE(kfd_processes_table, KFD_PROCESS_TABLE_SIZE); +extern struct srcu_struct kfd_processes_srcu; + /** * Ioctl function type. * @@ -607,18 +674,20 @@ struct amdkfd_ioctl_desc { const char *name; }; -void kfd_process_create_wq(void); +int kfd_process_create_wq(void); void kfd_process_destroy_wq(void); struct kfd_process *kfd_create_process(struct file *filep); struct kfd_process *kfd_get_process(const struct task_struct *); struct kfd_process *kfd_lookup_process_by_pasid(unsigned int pasid); +struct kfd_process *kfd_lookup_process_by_mm(const struct mm_struct *mm); void kfd_unref_process(struct kfd_process *p); +void kfd_suspend_all_processes(void); +int kfd_resume_all_processes(void); +int kfd_process_device_init_vm(struct kfd_process_device *pdd, + struct file *drm_file); struct kfd_process_device *kfd_bind_process_to_device(struct kfd_dev *dev, struct kfd_process *p); -int kfd_bind_processes_to_device(struct kfd_dev *dev); -void kfd_unbind_processes_from_device(struct kfd_dev *dev); -void kfd_process_iommu_unbind_callback(struct kfd_dev *dev, unsigned int pasid); struct kfd_process_device *kfd_get_process_device_data(struct kfd_dev *dev, struct kfd_process *p); struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev, @@ -627,6 +696,14 @@ struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev, int kfd_reserved_mem_mmap(struct kfd_process *process, struct vm_area_struct *vma); +/* KFD process API for creating and translating handles */ +int kfd_process_device_create_obj_handle(struct kfd_process_device *pdd, + void *mem); +void *kfd_process_device_translate_handle(struct kfd_process_device *p, + int handle); +void kfd_process_device_remove_obj_handle(struct kfd_process_device *pdd, + int handle); + /* Process device data iterator */ struct kfd_process_device *kfd_get_first_process_device_data( struct kfd_process *p); @@ -705,8 +782,12 @@ struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type, struct kfd_dev *dev); struct mqd_manager *mqd_manager_init_cik(enum KFD_MQD_TYPE type, struct kfd_dev *dev); +struct mqd_manager *mqd_manager_init_cik_hawaii(enum KFD_MQD_TYPE type, + struct kfd_dev *dev); struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type, struct kfd_dev *dev); +struct mqd_manager *mqd_manager_init_vi_tonga(enum KFD_MQD_TYPE type, + struct kfd_dev *dev); struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev); void device_queue_manager_uninit(struct device_queue_manager *dqm); struct kernel_queue *kernel_queue_init(struct kfd_dev *dev, @@ -768,6 +849,8 @@ int pm_send_unmap_queue(struct packet_manager *pm, enum kfd_queue_type type, void pm_release_ib(struct packet_manager *pm); +uint32_t pm_create_release_mem(uint64_t gpu_addr, uint32_t *buffer); + uint64_t kfd_get_number_elems(struct kfd_dev *kfd); /* Events */ @@ -789,12 +872,16 @@ void kfd_signal_iommu_event(struct kfd_dev *dev, void kfd_signal_hw_exception_event(unsigned int pasid); int kfd_set_event(struct kfd_process *p, uint32_t event_id); int kfd_reset_event(struct kfd_process *p, uint32_t event_id); +int kfd_event_page_set(struct kfd_process *p, void *kernel_address, + uint64_t size); int kfd_event_create(struct file *devkfd, struct kfd_process *p, uint32_t event_type, bool auto_reset, uint32_t node_id, uint32_t *event_id, uint32_t *event_trigger_data, uint64_t *event_page_offset, uint32_t *event_slot_index); int kfd_event_destroy(struct kfd_process *p, uint32_t event_id); +void kfd_flush_tlb(struct kfd_process_device *pdd); + int dbgdev_wave_reset_wavefronts(struct kfd_dev *dev, struct kfd_process *p); /* Debugfs */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 23541c6..59a8352 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -30,35 +30,57 @@ #include <linux/notifier.h> #include <linux/compat.h> #include <linux/mman.h> +#include <linux/file.h> struct mm_struct; #include "kfd_priv.h" +#include "kfd_device_queue_manager.h" #include "kfd_dbgmgr.h" +#include "kfd_iommu.h" /* * List of struct kfd_process (field kfd_process). * Unique/indexed by mm_struct* */ -#define KFD_PROCESS_TABLE_SIZE 5 /* bits: 32 entries */ -static DEFINE_HASHTABLE(kfd_processes_table, KFD_PROCESS_TABLE_SIZE); +DEFINE_HASHTABLE(kfd_processes_table, KFD_PROCESS_TABLE_SIZE); static DEFINE_MUTEX(kfd_processes_mutex); -DEFINE_STATIC_SRCU(kfd_processes_srcu); +DEFINE_SRCU(kfd_processes_srcu); +/* For process termination handling */ static struct workqueue_struct *kfd_process_wq; +/* Ordered, single-threaded workqueue for restoring evicted + * processes. Restoring multiple processes concurrently under memory + * pressure can lead to processes blocking each other from validating + * their BOs and result in a live-lock situation where processes + * remain evicted indefinitely. + */ +static struct workqueue_struct *kfd_restore_wq; + static struct kfd_process *find_process(const struct task_struct *thread); static void kfd_process_ref_release(struct kref *ref); static struct kfd_process *create_process(const struct task_struct *thread, struct file *filep); -static int kfd_process_init_cwsr(struct kfd_process *p, struct file *filep); + +static void evict_process_worker(struct work_struct *work); +static void restore_process_worker(struct work_struct *work); -void kfd_process_create_wq(void) +int kfd_process_create_wq(void) { if (!kfd_process_wq) kfd_process_wq = alloc_workqueue("kfd_process_wq", 0, 0); + if (!kfd_restore_wq) + kfd_restore_wq = alloc_ordered_workqueue("kfd_restore_wq", 0); + + if (!kfd_process_wq || !kfd_restore_wq) { + kfd_process_destroy_wq(); + return -ENOMEM; + } + + return 0; } void kfd_process_destroy_wq(void) @@ -67,6 +89,116 @@ void kfd_process_destroy_wq(void) destroy_workqueue(kfd_process_wq); kfd_process_wq = NULL; } + if (kfd_restore_wq) { + destroy_workqueue(kfd_restore_wq); + kfd_restore_wq = NULL; + } +} + +static void kfd_process_free_gpuvm(struct kgd_mem *mem, + struct kfd_process_device *pdd) +{ + struct kfd_dev *dev = pdd->dev; + + dev->kfd2kgd->unmap_memory_to_gpu(dev->kgd, mem, pdd->vm); + dev->kfd2kgd->free_memory_of_gpu(dev->kgd, mem); +} + +/* kfd_process_alloc_gpuvm - Allocate GPU VM for the KFD process + * This function should be only called right after the process + * is created and when kfd_processes_mutex is still being held + * to avoid concurrency. Because of that exclusiveness, we do + * not need to take p->mutex. + */ +static int kfd_process_alloc_gpuvm(struct kfd_process_device *pdd, + uint64_t gpu_va, uint32_t size, + uint32_t flags, void **kptr) +{ + struct kfd_dev *kdev = pdd->dev; + struct kgd_mem *mem = NULL; + int handle; + int err; + + err = kdev->kfd2kgd->alloc_memory_of_gpu(kdev->kgd, gpu_va, size, + pdd->vm, &mem, NULL, flags); + if (err) + goto err_alloc_mem; + + err = kdev->kfd2kgd->map_memory_to_gpu(kdev->kgd, mem, pdd->vm); + if (err) + goto err_map_mem; + + err = kdev->kfd2kgd->sync_memory(kdev->kgd, mem, true); + if (err) { + pr_debug("Sync memory failed, wait interrupted by user signal\n"); + goto sync_memory_failed; + } + + /* Create an obj handle so kfd_process_device_remove_obj_handle + * will take care of the bo removal when the process finishes. + * We do not need to take p->mutex, because the process is just + * created and the ioctls have not had the chance to run. + */ + handle = kfd_process_device_create_obj_handle(pdd, mem); + + if (handle < 0) { + err = handle; + goto free_gpuvm; + } + + if (kptr) { + err = kdev->kfd2kgd->map_gtt_bo_to_kernel(kdev->kgd, + (struct kgd_mem *)mem, kptr, NULL); + if (err) { + pr_debug("Map GTT BO to kernel failed\n"); + goto free_obj_handle; + } + } + + return err; + +free_obj_handle: + kfd_process_device_remove_obj_handle(pdd, handle); +free_gpuvm: +sync_memory_failed: + kfd_process_free_gpuvm(mem, pdd); + return err; + +err_map_mem: + kdev->kfd2kgd->free_memory_of_gpu(kdev->kgd, mem); +err_alloc_mem: + *kptr = NULL; + return err; +} + +/* kfd_process_device_reserve_ib_mem - Reserve memory inside the + * process for IB usage The memory reserved is for KFD to submit + * IB to AMDGPU from kernel. If the memory is reserved + * successfully, ib_kaddr will have the CPU/kernel + * address. Check ib_kaddr before accessing the memory. + */ +static int kfd_process_device_reserve_ib_mem(struct kfd_process_device *pdd) +{ + struct qcm_process_device *qpd = &pdd->qpd; + uint32_t flags = ALLOC_MEM_FLAGS_GTT | + ALLOC_MEM_FLAGS_NO_SUBSTITUTE | + ALLOC_MEM_FLAGS_WRITABLE | + ALLOC_MEM_FLAGS_EXECUTABLE; + void *kaddr; + int ret; + + if (qpd->ib_kaddr || !qpd->ib_base) + return 0; + + /* ib_base is only set for dGPU */ + ret = kfd_process_alloc_gpuvm(pdd, qpd->ib_base, PAGE_SIZE, flags, + &kaddr); + if (ret) + return ret; + + qpd->ib_kaddr = kaddr; + + return 0; } struct kfd_process *kfd_create_process(struct file *filep) @@ -147,6 +279,40 @@ void kfd_unref_process(struct kfd_process *p) kref_put(&p->ref, kfd_process_ref_release); } +static void kfd_process_device_free_bos(struct kfd_process_device *pdd) +{ + struct kfd_process *p = pdd->process; + void *mem; + int id; + + /* + * Remove all handles from idr and release appropriate + * local memory object + */ + idr_for_each_entry(&pdd->alloc_idr, mem, id) { + struct kfd_process_device *peer_pdd; + + list_for_each_entry(peer_pdd, &p->per_device_data, + per_device_list) { + if (!peer_pdd->vm) + continue; + peer_pdd->dev->kfd2kgd->unmap_memory_to_gpu( + peer_pdd->dev->kgd, mem, peer_pdd->vm); + } + + pdd->dev->kfd2kgd->free_memory_of_gpu(pdd->dev->kgd, mem); + kfd_process_device_remove_obj_handle(pdd, id); + } +} + +static void kfd_process_free_outstanding_kfd_bos(struct kfd_process *p) +{ + struct kfd_process_device *pdd; + + list_for_each_entry(pdd, &p->per_device_data, per_device_list) + kfd_process_device_free_bos(pdd); +} + static void kfd_process_destroy_pdds(struct kfd_process *p) { struct kfd_process_device *pdd, *temp; @@ -156,12 +322,20 @@ static void kfd_process_destroy_pdds(struct kfd_process *p) pr_debug("Releasing pdd (topology id %d) for process (pasid %d)\n", pdd->dev->id, p->pasid); + if (pdd->drm_file) + fput(pdd->drm_file); + else if (pdd->vm) + pdd->dev->kfd2kgd->destroy_process_vm( + pdd->dev->kgd, pdd->vm); + list_del(&pdd->per_device_list); - if (pdd->qpd.cwsr_kaddr) + if (pdd->qpd.cwsr_kaddr && !pdd->qpd.cwsr_base) free_pages((unsigned long)pdd->qpd.cwsr_kaddr, get_order(KFD_CWSR_TBA_TMA_SIZE)); + idr_destroy(&pdd->alloc_idr); + kfree(pdd); } } @@ -175,16 +349,13 @@ static void kfd_process_wq_release(struct work_struct *work) { struct kfd_process *p = container_of(work, struct kfd_process, release_work); - struct kfd_process_device *pdd; - pr_debug("Releasing process (pasid %d) in workqueue\n", p->pasid); + kfd_iommu_unbind_process(p); - list_for_each_entry(pdd, &p->per_device_data, per_device_list) { - if (pdd->bound == PDD_BOUND) - amd_iommu_unbind_pasid(pdd->dev->pdev, p->pasid); - } + kfd_process_free_outstanding_kfd_bos(p); kfd_process_destroy_pdds(p); + dma_fence_put(p->ef); kfd_event_free_process(p); @@ -232,6 +403,9 @@ static void kfd_process_notifier_release(struct mmu_notifier *mn, mutex_unlock(&kfd_processes_mutex); synchronize_srcu(&kfd_processes_srcu); + cancel_delayed_work_sync(&p->eviction_work); + cancel_delayed_work_sync(&p->restore_work); + mutex_lock(&p->mutex); /* Iterate over all process device data structures and if the @@ -267,18 +441,18 @@ static const struct mmu_notifier_ops kfd_process_mmu_notifier_ops = { .release = kfd_process_notifier_release, }; -static int kfd_process_init_cwsr(struct kfd_process *p, struct file *filep) +static int kfd_process_init_cwsr_apu(struct kfd_process *p, struct file *filep) { unsigned long offset; - struct kfd_process_device *pdd = NULL; - struct kfd_dev *dev = NULL; - struct qcm_process_device *qpd = NULL; + struct kfd_process_device *pdd; list_for_each_entry(pdd, &p->per_device_data, per_device_list) { - dev = pdd->dev; - qpd = &pdd->qpd; - if (!dev->cwsr_enabled || qpd->cwsr_kaddr) + struct kfd_dev *dev = pdd->dev; + struct qcm_process_device *qpd = &pdd->qpd; + + if (!dev->cwsr_enabled || qpd->cwsr_kaddr || qpd->cwsr_base) continue; + offset = (dev->id | KFD_MMAP_RESERVED_MEM_MASK) << PAGE_SHIFT; qpd->tba_addr = (int64_t)vm_mmap(filep, 0, KFD_CWSR_TBA_TMA_SIZE, PROT_READ | PROT_EXEC, @@ -303,6 +477,36 @@ static int kfd_process_init_cwsr(struct kfd_process *p, struct file *filep) return 0; } +static int kfd_process_device_init_cwsr_dgpu(struct kfd_process_device *pdd) +{ + struct kfd_dev *dev = pdd->dev; + struct qcm_process_device *qpd = &pdd->qpd; + uint32_t flags = ALLOC_MEM_FLAGS_GTT | + ALLOC_MEM_FLAGS_NO_SUBSTITUTE | ALLOC_MEM_FLAGS_EXECUTABLE; + void *kaddr; + int ret; + + if (!dev->cwsr_enabled || qpd->cwsr_kaddr || !qpd->cwsr_base) + return 0; + + /* cwsr_base is only set for dGPU */ + ret = kfd_process_alloc_gpuvm(pdd, qpd->cwsr_base, + KFD_CWSR_TBA_TMA_SIZE, flags, &kaddr); + if (ret) + return ret; + + qpd->cwsr_kaddr = kaddr; + qpd->tba_addr = qpd->cwsr_base; + + memcpy(qpd->cwsr_kaddr, dev->cwsr_isa, dev->cwsr_isa_size); + + qpd->tma_addr = qpd->tba_addr + KFD_CWSR_TMA_OFFSET; + pr_debug("set tba :0x%llx, tma:0x%llx, cwsr_kaddr:%p for pqm.\n", + qpd->tba_addr, qpd->tma_addr, qpd->cwsr_kaddr); + + return 0; +} + static struct kfd_process *create_process(const struct task_struct *thread, struct file *filep) { @@ -353,13 +557,18 @@ static struct kfd_process *create_process(const struct task_struct *thread, if (err != 0) goto err_init_apertures; - err = kfd_process_init_cwsr(process, filep); + INIT_DELAYED_WORK(&process->eviction_work, evict_process_worker); + INIT_DELAYED_WORK(&process->restore_work, restore_process_worker); + process->last_restore_timestamp = get_jiffies_64(); + + err = kfd_process_init_cwsr_apu(process, filep); if (err) goto err_init_cwsr; return process; err_init_cwsr: + kfd_process_free_outstanding_kfd_bos(process); kfd_process_destroy_pdds(process); err_init_apertures: pqm_uninit(&process->pqm); @@ -404,14 +613,78 @@ struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev, INIT_LIST_HEAD(&pdd->qpd.priv_queue_list); pdd->qpd.dqm = dev->dqm; pdd->qpd.pqm = &p->pqm; + pdd->qpd.evicted = 0; pdd->process = p; pdd->bound = PDD_UNBOUND; pdd->already_dequeued = false; list_add(&pdd->per_device_list, &p->per_device_data); + /* Init idr used for memory handle translation */ + idr_init(&pdd->alloc_idr); + return pdd; } +/** + * kfd_process_device_init_vm - Initialize a VM for a process-device + * + * @pdd: The process-device + * @drm_file: Optional pointer to a DRM file descriptor + * + * If @drm_file is specified, it will be used to acquire the VM from + * that file descriptor. If successful, the @pdd takes ownership of + * the file descriptor. + * + * If @drm_file is NULL, a new VM is created. + * + * Returns 0 on success, -errno on failure. + */ +int kfd_process_device_init_vm(struct kfd_process_device *pdd, + struct file *drm_file) +{ + struct kfd_process *p; + struct kfd_dev *dev; + int ret; + + if (pdd->vm) + return drm_file ? -EBUSY : 0; + + p = pdd->process; + dev = pdd->dev; + + if (drm_file) + ret = dev->kfd2kgd->acquire_process_vm( + dev->kgd, drm_file, + &pdd->vm, &p->kgd_process_info, &p->ef); + else + ret = dev->kfd2kgd->create_process_vm( + dev->kgd, &pdd->vm, &p->kgd_process_info, &p->ef); + if (ret) { + pr_err("Failed to create process VM object\n"); + return ret; + } + + ret = kfd_process_device_reserve_ib_mem(pdd); + if (ret) + goto err_reserve_ib_mem; + ret = kfd_process_device_init_cwsr_dgpu(pdd); + if (ret) + goto err_init_cwsr; + + pdd->drm_file = drm_file; + + return 0; + +err_init_cwsr: +err_reserve_ib_mem: + kfd_process_device_free_bos(pdd); + if (!drm_file) + dev->kfd2kgd->destroy_process_vm(dev->kgd, pdd->vm); + pdd->vm = NULL; + + return ret; +} + /* * Direct the IOMMU to bind the process (specifically the pasid->mm) * to the device. @@ -431,133 +704,17 @@ struct kfd_process_device *kfd_bind_process_to_device(struct kfd_dev *dev, return ERR_PTR(-ENOMEM); } - if (pdd->bound == PDD_BOUND) { - return pdd; - } else if (unlikely(pdd->bound == PDD_BOUND_SUSPENDED)) { - pr_err("Binding PDD_BOUND_SUSPENDED pdd is unexpected!\n"); - return ERR_PTR(-EINVAL); - } - - err = amd_iommu_bind_pasid(dev->pdev, p->pasid, p->lead_thread); - if (err < 0) + err = kfd_iommu_bind_process_to_device(pdd); + if (err) return ERR_PTR(err); - pdd->bound = PDD_BOUND; + err = kfd_process_device_init_vm(pdd, NULL); + if (err) + return ERR_PTR(err); return pdd; } -/* - * Bind processes do the device that have been temporarily unbound - * (PDD_BOUND_SUSPENDED) in kfd_unbind_processes_from_device. - */ -int kfd_bind_processes_to_device(struct kfd_dev *dev) -{ - struct kfd_process_device *pdd; - struct kfd_process *p; - unsigned int temp; - int err = 0; - - int idx = srcu_read_lock(&kfd_processes_srcu); - - hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { - mutex_lock(&p->mutex); - pdd = kfd_get_process_device_data(dev, p); - - if (WARN_ON(!pdd) || pdd->bound != PDD_BOUND_SUSPENDED) { - mutex_unlock(&p->mutex); - continue; - } - - err = amd_iommu_bind_pasid(dev->pdev, p->pasid, - p->lead_thread); - if (err < 0) { - pr_err("Unexpected pasid %d binding failure\n", - p->pasid); - mutex_unlock(&p->mutex); - break; - } - - pdd->bound = PDD_BOUND; - mutex_unlock(&p->mutex); - } - - srcu_read_unlock(&kfd_processes_srcu, idx); - - return err; -} - -/* - * Mark currently bound processes as PDD_BOUND_SUSPENDED. These - * processes will be restored to PDD_BOUND state in - * kfd_bind_processes_to_device. - */ -void kfd_unbind_processes_from_device(struct kfd_dev *dev) -{ - struct kfd_process_device *pdd; - struct kfd_process *p; - unsigned int temp; - - int idx = srcu_read_lock(&kfd_processes_srcu); - - hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { - mutex_lock(&p->mutex); - pdd = kfd_get_process_device_data(dev, p); - - if (WARN_ON(!pdd)) { - mutex_unlock(&p->mutex); - continue; - } - - if (pdd->bound == PDD_BOUND) - pdd->bound = PDD_BOUND_SUSPENDED; - mutex_unlock(&p->mutex); - } - - srcu_read_unlock(&kfd_processes_srcu, idx); -} - -void kfd_process_iommu_unbind_callback(struct kfd_dev *dev, unsigned int pasid) -{ - struct kfd_process *p; - struct kfd_process_device *pdd; - - /* - * Look for the process that matches the pasid. If there is no such - * process, we either released it in amdkfd's own notifier, or there - * is a bug. Unfortunately, there is no way to tell... - */ - p = kfd_lookup_process_by_pasid(pasid); - if (!p) - return; - - pr_debug("Unbinding process %d from IOMMU\n", pasid); - - mutex_lock(kfd_get_dbgmgr_mutex()); - - if (dev->dbgmgr && dev->dbgmgr->pasid == p->pasid) { - if (!kfd_dbgmgr_unregister(dev->dbgmgr, p)) { - kfd_dbgmgr_destroy(dev->dbgmgr); - dev->dbgmgr = NULL; - } - } - - mutex_unlock(kfd_get_dbgmgr_mutex()); - - mutex_lock(&p->mutex); - - pdd = kfd_get_process_device_data(dev, p); - if (pdd) - /* For GPU relying on IOMMU, we need to dequeue here - * when PASID is still bound. - */ - kfd_process_dequeue_from_device(pdd); - - mutex_unlock(&p->mutex); - - kfd_unref_process(p); -} - struct kfd_process_device *kfd_get_first_process_device_data( struct kfd_process *p) { @@ -580,6 +737,37 @@ bool kfd_has_process_device_data(struct kfd_process *p) return !(list_empty(&p->per_device_data)); } +/* Create specific handle mapped to mem from process local memory idr + * Assumes that the process lock is held. + */ +int kfd_process_device_create_obj_handle(struct kfd_process_device *pdd, + void *mem) +{ + return idr_alloc(&pdd->alloc_idr, mem, 0, 0, GFP_KERNEL); +} + +/* Translate specific handle from process local memory idr + * Assumes that the process lock is held. + */ +void *kfd_process_device_translate_handle(struct kfd_process_device *pdd, + int handle) +{ + if (handle < 0) + return NULL; + + return idr_find(&pdd->alloc_idr, handle); +} + +/* Remove specific handle from process local memory idr + * Assumes that the process lock is held. + */ +void kfd_process_device_remove_obj_handle(struct kfd_process_device *pdd, + int handle) +{ + if (handle >= 0) + idr_remove(&pdd->alloc_idr, handle); +} + /* This increments the process->ref counter. */ struct kfd_process *kfd_lookup_process_by_pasid(unsigned int pasid) { @@ -601,6 +789,208 @@ struct kfd_process *kfd_lookup_process_by_pasid(unsigned int pasid) return ret_p; } +/* This increments the process->ref counter. */ +struct kfd_process *kfd_lookup_process_by_mm(const struct mm_struct *mm) +{ + struct kfd_process *p; + + int idx = srcu_read_lock(&kfd_processes_srcu); + + p = find_process_by_mm(mm); + if (p) + kref_get(&p->ref); + + srcu_read_unlock(&kfd_processes_srcu, idx); + + return p; +} + +/* process_evict_queues - Evict all user queues of a process + * + * Eviction is reference-counted per process-device. This means multiple + * evictions from different sources can be nested safely. + */ +static int process_evict_queues(struct kfd_process *p) +{ + struct kfd_process_device *pdd; + int r = 0; + unsigned int n_evicted = 0; + + list_for_each_entry(pdd, &p->per_device_data, per_device_list) { + r = pdd->dev->dqm->ops.evict_process_queues(pdd->dev->dqm, + &pdd->qpd); + if (r) { + pr_err("Failed to evict process queues\n"); + goto fail; + } + n_evicted++; + } + + return r; + +fail: + /* To keep state consistent, roll back partial eviction by + * restoring queues + */ + list_for_each_entry(pdd, &p->per_device_data, per_device_list) { + if (n_evicted == 0) + break; + if (pdd->dev->dqm->ops.restore_process_queues(pdd->dev->dqm, + &pdd->qpd)) + pr_err("Failed to restore queues\n"); + + n_evicted--; + } + + return r; +} + +/* process_restore_queues - Restore all user queues of a process */ +static int process_restore_queues(struct kfd_process *p) +{ + struct kfd_process_device *pdd; + int r, ret = 0; + + list_for_each_entry(pdd, &p->per_device_data, per_device_list) { + r = pdd->dev->dqm->ops.restore_process_queues(pdd->dev->dqm, + &pdd->qpd); + if (r) { + pr_err("Failed to restore process queues\n"); + if (!ret) + ret = r; + } + } + + return ret; +} + +static void evict_process_worker(struct work_struct *work) +{ + int ret; + struct kfd_process *p; + struct delayed_work *dwork; + + dwork = to_delayed_work(work); + + /* Process termination destroys this worker thread. So during the + * lifetime of this thread, kfd_process p will be valid + */ + p = container_of(dwork, struct kfd_process, eviction_work); + WARN_ONCE(p->last_eviction_seqno != p->ef->seqno, + "Eviction fence mismatch\n"); + + /* Narrow window of overlap between restore and evict work + * item is possible. Once amdgpu_amdkfd_gpuvm_restore_process_bos + * unreserves KFD BOs, it is possible to evicted again. But + * restore has few more steps of finish. So lets wait for any + * previous restore work to complete + */ + flush_delayed_work(&p->restore_work); + + pr_debug("Started evicting pasid %d\n", p->pasid); + ret = process_evict_queues(p); + if (!ret) { + dma_fence_signal(p->ef); + dma_fence_put(p->ef); + p->ef = NULL; + queue_delayed_work(kfd_restore_wq, &p->restore_work, + msecs_to_jiffies(PROCESS_RESTORE_TIME_MS)); + + pr_debug("Finished evicting pasid %d\n", p->pasid); + } else + pr_err("Failed to evict queues of pasid %d\n", p->pasid); +} + +static void restore_process_worker(struct work_struct *work) +{ + struct delayed_work *dwork; + struct kfd_process *p; + struct kfd_process_device *pdd; + int ret = 0; + + dwork = to_delayed_work(work); + + /* Process termination destroys this worker thread. So during the + * lifetime of this thread, kfd_process p will be valid + */ + p = container_of(dwork, struct kfd_process, restore_work); + + /* Call restore_process_bos on the first KGD device. This function + * takes care of restoring the whole process including other devices. + * Restore can fail if enough memory is not available. If so, + * reschedule again. + */ + pdd = list_first_entry(&p->per_device_data, + struct kfd_process_device, + per_device_list); + + pr_debug("Started restoring pasid %d\n", p->pasid); + + /* Setting last_restore_timestamp before successful restoration. + * Otherwise this would have to be set by KGD (restore_process_bos) + * before KFD BOs are unreserved. If not, the process can be evicted + * again before the timestamp is set. + * If restore fails, the timestamp will be set again in the next + * attempt. This would mean that the minimum GPU quanta would be + * PROCESS_ACTIVE_TIME_MS - (time to execute the following two + * functions) + */ + + p->last_restore_timestamp = get_jiffies_64(); + ret = pdd->dev->kfd2kgd->restore_process_bos(p->kgd_process_info, + &p->ef); + if (ret) { + pr_debug("Failed to restore BOs of pasid %d, retry after %d ms\n", + p->pasid, PROCESS_BACK_OFF_TIME_MS); + ret = queue_delayed_work(kfd_restore_wq, &p->restore_work, + msecs_to_jiffies(PROCESS_BACK_OFF_TIME_MS)); + WARN(!ret, "reschedule restore work failed\n"); + return; + } + + ret = process_restore_queues(p); + if (!ret) + pr_debug("Finished restoring pasid %d\n", p->pasid); + else + pr_err("Failed to restore queues of pasid %d\n", p->pasid); +} + +void kfd_suspend_all_processes(void) +{ + struct kfd_process *p; + unsigned int temp; + int idx = srcu_read_lock(&kfd_processes_srcu); + + hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { + cancel_delayed_work_sync(&p->eviction_work); + cancel_delayed_work_sync(&p->restore_work); + + if (process_evict_queues(p)) + pr_err("Failed to suspend process %d\n", p->pasid); + dma_fence_signal(p->ef); + dma_fence_put(p->ef); + p->ef = NULL; + } + srcu_read_unlock(&kfd_processes_srcu, idx); +} + +int kfd_resume_all_processes(void) +{ + struct kfd_process *p; + unsigned int temp; + int ret = 0, idx = srcu_read_lock(&kfd_processes_srcu); + + hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { + if (!queue_delayed_work(kfd_restore_wq, &p->restore_work, 0)) { + pr_err("Restore process %d failed during resume\n", + p->pasid); + ret = -EFAULT; + } + } + srcu_read_unlock(&kfd_processes_srcu, idx); + return ret; +} + int kfd_reserved_mem_mmap(struct kfd_process *process, struct vm_area_struct *vma) { @@ -635,6 +1025,22 @@ int kfd_reserved_mem_mmap(struct kfd_process *process, KFD_CWSR_TBA_TMA_SIZE, vma->vm_page_prot); } +void kfd_flush_tlb(struct kfd_process_device *pdd) +{ + struct kfd_dev *dev = pdd->dev; + const struct kfd2kgd_calls *f2g = dev->kfd2kgd; + + if (dev->dqm->sched_policy == KFD_SCHED_POLICY_NO_HWS) { + /* Nothing to flush until a VMID is assigned, which + * only happens when the first queue is created. + */ + if (pdd->qpd.vmid) + f2g->invalidate_tlbs_vmid(dev->kgd, pdd->qpd.vmid); + } else { + f2g->invalidate_tlbs(dev->kgd, pdd->process->pasid); + } +} + #if defined(CONFIG_DEBUG_FS) int kfd_debugfs_mqds_by_process(struct seq_file *m, void *data)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c index 8763806..7817e32 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -208,7 +208,8 @@ int pqm_create_queue(struct process_queue_manager *pqm, case KFD_QUEUE_TYPE_COMPUTE: /* check if there is over subscription */ - if ((sched_policy == KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION) && + if ((dev->dqm->sched_policy == + KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION) && ((dev->dqm->processes_count >= dev->vm_info.vmid_num_kfd) || (dev->dqm->queue_count >= get_queues_num(dev->dqm)))) { pr_err("Over-subscription is not allowed in radeon_kfd.sched_policy == 1\n");
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index c6a7609..ac28abc 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -35,6 +35,7 @@ #include "kfd_crat.h" #include "kfd_topology.h" #include "kfd_device_queue_manager.h" +#include "kfd_iommu.h" /* topology_device_list - Master list of all topology devices */ static struct list_head topology_device_list; @@ -440,6 +441,8 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr, dev->node_props.device_id); sysfs_show_32bit_prop(buffer, "location_id", dev->node_props.location_id); + sysfs_show_32bit_prop(buffer, "drm_render_minor", + dev->node_props.drm_render_minor); if (dev->gpu) { log_max_watch_addr = @@ -677,7 +680,7 @@ static int kfd_build_sysfs_node_entry(struct kfd_topology_device *dev, } /* All hardware blocks have the same number of attributes. */ - num_attrs = sizeof(perf_attr_iommu)/sizeof(struct kfd_perf_attr); + num_attrs = ARRAY_SIZE(perf_attr_iommu); list_for_each_entry(perf, &dev->perf_props, list) { perf->attr_group = kzalloc(sizeof(struct kfd_perf_attr) * num_attrs + sizeof(struct attribute_group), @@ -875,19 +878,8 @@ static void find_system_memory(const struct dmi_header *dm, */ static int kfd_add_perf_to_topology(struct kfd_topology_device *kdev) { - struct kfd_perf_properties *props; - - if (amd_iommu_pc_supported()) { - props = kfd_alloc_struct(props); - if (!props) - return -ENOMEM; - strcpy(props->block_name, "iommu"); - props->max_concurrent = amd_iommu_pc_get_max_banks(0) * - amd_iommu_pc_get_max_counters(0); /* assume one iommu */ - list_add_tail(&props->list, &kdev->perf_props); - } - - return 0; + /* These are the only counters supported so far */ + return kfd_iommu_add_perf_counters(kdev); } /* kfd_add_non_crat_information - Add information that is not currently @@ -1224,6 +1216,8 @@ int kfd_topology_add_device(struct kfd_dev *gpu) dev->gpu->kfd2kgd->get_max_engine_clock_in_mhz(dev->gpu->kgd); dev->node_props.max_engine_clk_ccompute = cpufreq_quick_get_max(0) / 1000; + dev->node_props.drm_render_minor = + gpu->shared_resources.drm_render_minor; kfd_fill_mem_clk_max_info(dev); kfd_fill_iolink_non_crat_info(dev);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h index 53fca1f..eb54cfc 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
@@ -25,7 +25,7 @@ #include <linux/types.h> #include <linux/list.h> -#include "kfd_priv.h" +#include "kfd_crat.h" #define KFD_TOPOLOGY_PUBLIC_NAME_SIZE 128 @@ -71,6 +71,7 @@ struct kfd_node_properties { uint32_t location_id; uint32_t max_engine_clk_fcompute; uint32_t max_engine_clk_ccompute; + int32_t drm_render_minor; uint16_t marketing_name[KFD_TOPOLOGY_PUBLIC_NAME_SIZE]; }; @@ -183,8 +184,4 @@ struct kfd_topology_device *kfd_create_topology_device( struct list_head *device_list); void kfd_release_topology_device_list(struct list_head *device_list); -extern bool amd_iommu_pc_supported(void); -extern u8 amd_iommu_pc_get_max_banks(u16 devid); -extern u8 amd_iommu_pc_get_max_counters(u16 devid); - #endif /* __KFD_TOPOLOGY_H__ */
diff --git a/drivers/gpu/drm/amd/display/Makefile b/drivers/gpu/drm/amd/display/Makefile index c27c81c..a2c5be4 100644 --- a/drivers/gpu/drm/amd/display/Makefile +++ b/drivers/gpu/drm/amd/display/Makefile
@@ -26,17 +26,16 @@ AMDDALPATH = $(RELATIVE_AMD_DISPLAY_PATH) -subdir-ccflags-y += -I$(AMDDALPATH)/ -I$(AMDDALPATH)/include - subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/inc/ subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/inc/hw subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/inc subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/freesync +subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/color #TODO: remove when Timing Sync feature is complete subdir-ccflags-y += -DBUILD_FEATURE_TIMING_SYNC=0 -DAL_LIBS = amdgpu_dm dc modules/freesync +DAL_LIBS = amdgpu_dm dc modules/freesync modules/color AMD_DAL = $(addsuffix /Makefile, $(addprefix $(FULL_AMD_DISPLAY_PATH)/,$(DAL_LIBS)))
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile index dfb6dba..589c60e 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
@@ -25,7 +25,7 @@ -AMDGPUDM = amdgpu_dm.o amdgpu_dm_irq.o amdgpu_dm_mst_types.o +AMDGPUDM = amdgpu_dm.o amdgpu_dm_irq.o amdgpu_dm_mst_types.o amdgpu_dm_color.o ifneq ($(CONFIG_DRM_AMD_DC),) AMDGPUDM += amdgpu_dm_services.o amdgpu_dm_helpers.o
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 84cf6ea..3bc2706 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -64,7 +64,8 @@ #include "dcn/dcn_1_0_offset.h" #include "dcn/dcn_1_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #include "soc15_common.h" #endif @@ -405,12 +406,6 @@ static int amdgpu_dm_init(struct amdgpu_device *adev) /* Zero all the fields */ memset(&init_data, 0, sizeof(init_data)); - /* initialize DAL's lock (for SYNC context use) */ - spin_lock_init(&adev->dm.dal_lock); - - /* initialize DAL's mutex */ - mutex_init(&adev->dm.dal_mutex); - if(amdgpu_dm_irq_init(adev)) { DRM_ERROR("amdgpu: failed to initialize DM IRQ support.\n"); goto error; @@ -421,7 +416,7 @@ static int amdgpu_dm_init(struct amdgpu_device *adev) init_data.asic_id.pci_revision_id = adev->rev_id; init_data.asic_id.hw_internal_rev = adev->external_rev_id; - init_data.asic_id.vram_width = adev->mc.vram_width; + init_data.asic_id.vram_width = adev->gmc.vram_width; /* TODO: initialize init_data.asic_id.vram_type here!!!! */ init_data.asic_id.atombios_base_address = adev->mode_info.atom_context->bios; @@ -474,6 +469,8 @@ static int amdgpu_dm_init(struct amdgpu_device *adev) DRM_DEBUG_DRIVER("amdgpu: freesync_module init done %p.\n", adev->dm.freesync_module); + amdgpu_dm_init_color_mod(); + if (amdgpu_dm_initialize_drm_device(adev)) { DRM_ERROR( "amdgpu: failed to initialize sw for display support.\n"); @@ -820,7 +817,7 @@ dm_atomic_state_alloc_free(struct drm_atomic_state *state) } static const struct drm_mode_config_funcs amdgpu_dm_mode_funcs = { - .fb_create = amdgpu_user_framebuffer_create, + .fb_create = amdgpu_display_user_framebuffer_create, .output_poll_changed = drm_fb_helper_output_poll_changed, .atomic_check = amdgpu_dm_atomic_check, .atomic_commit = amdgpu_dm_atomic_commit, @@ -1136,8 +1133,9 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev) unsigned client_id = AMDGPU_IH_CLIENTID_LEGACY; if (adev->asic_type == CHIP_VEGA10 || + adev->asic_type == CHIP_VEGA12 || adev->asic_type == CHIP_RAVEN) - client_id = AMDGPU_IH_CLIENTID_DCE; + client_id = SOC15_IH_CLIENTID_DCE; int_params.requested_polarity = INTERRUPT_POLARITY_DEFAULT; int_params.current_polarity = INTERRUPT_POLARITY_DEFAULT; @@ -1237,7 +1235,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev) for (i = DCN_1_0__SRCID__DC_D1_OTG_VSTARTUP; i <= DCN_1_0__SRCID__DC_D1_OTG_VSTARTUP + adev->mode_info.num_crtc - 1; i++) { - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_DCE, i, &adev->crtc_irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, i, &adev->crtc_irq); if (r) { DRM_ERROR("Failed to add crtc irq id!\n"); @@ -1261,7 +1259,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev) for (i = DCN_1_0__SRCID__HUBP0_FLIP_INTERRUPT; i <= DCN_1_0__SRCID__HUBP0_FLIP_INTERRUPT + adev->mode_info.num_crtc - 1; i++) { - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_DCE, i, &adev->pageflip_irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, i, &adev->pageflip_irq); if (r) { DRM_ERROR("Failed to add page flip irq id!\n"); return r; @@ -1282,7 +1280,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev) } /* HPD */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_DCE, DCN_1_0__SRCID__DC_HPD1_INT, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, DCN_1_0__SRCID__DC_HPD1_INT, &adev->hpd_irq); if (r) { DRM_ERROR("Failed to add hpd irq id!\n"); @@ -1312,9 +1310,9 @@ static int amdgpu_dm_mode_config_init(struct amdgpu_device *adev) /* indicate support of immediate flip */ adev->ddev->mode_config.async_page_flip = true; - adev->ddev->mode_config.fb_base = adev->mc.aper_base; + adev->ddev->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_modeset_create_props(adev); + r = amdgpu_display_modeset_create_props(adev); if (r) return r; @@ -1415,6 +1413,28 @@ static int initialize_plane(struct amdgpu_display_manager *dm, return ret; } + +static void register_backlight_device(struct amdgpu_display_manager *dm, + struct dc_link *link) +{ +#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\ + defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) + + if ((link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) && + link->type != dc_connection_none) { + /* Event if registration failed, we should continue with + * DM initialization because not having a backlight control + * is better then a black screen. + */ + amdgpu_dm_register_backlight_device(dm); + + if (dm->backlight_dev) + dm->backlight_link = link; + } +#endif +} + + /* In this architecture, the association * connector -> encoder -> crtc * id not really requried. The crtc and connector will hold the @@ -1459,6 +1479,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) /* loops over all connectors on the board */ for (i = 0; i < link_cnt; i++) { + struct dc_link *link = NULL; if (i > AMDGPU_DM_MAX_DISPLAY_INDEX) { DRM_ERROR( @@ -1485,9 +1506,14 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) goto fail; } - if (dc_link_detect(dc_get_link_at_index(dm->dc, i), - DETECT_REASON_BOOT)) + link = dc_get_link_at_index(dm->dc, i); + + if (dc_link_detect(link, DETECT_REASON_BOOT)) { amdgpu_dm_update_connector_after_detect(aconnector); + register_backlight_device(dm, link); + } + + } /* Software is initialized. Now we can register interrupt handlers. */ @@ -1505,6 +1531,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) case CHIP_POLARIS10: case CHIP_POLARIS12: case CHIP_VEGA10: + case CHIP_VEGA12: if (dce110_register_irq_handlers(dm->adev)) { DRM_ERROR("DM: Failed to initialize IRQ\n"); goto fail; @@ -1598,7 +1625,6 @@ static int amdgpu_notify_freesync(struct drm_device *dev, void *data, static const struct amdgpu_display_funcs dm_display_funcs = { .bandwidth_update = dm_bandwidth_update, /* called unconditionally */ .vblank_get_counter = dm_vblank_get_counter,/* called unconditionally */ - .vblank_wait = NULL, .backlight_set_level = dm_set_backlight_level,/* called unconditionally */ .backlight_get_level = @@ -1703,6 +1729,7 @@ static int dm_early_init(void *handle) adev->mode_info.plane_type = dm_plane_type_default; break; case CHIP_VEGA10: + case CHIP_VEGA12: adev->mode_info.num_crtc = 6; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 6; @@ -1950,6 +1977,7 @@ static int fill_plane_attributes_from_fb(struct amdgpu_device *adev, AMDGPU_TILING_GET(tiling_flags, PIPE_CONFIG); if (adev->asic_type == CHIP_VEGA10 || + adev->asic_type == CHIP_VEGA12 || adev->asic_type == CHIP_RAVEN) { /* Fill GFX9 params */ plane_state->tiling_info.gfx9.num_pipes = @@ -1982,32 +2010,6 @@ static int fill_plane_attributes_from_fb(struct amdgpu_device *adev, } -static void fill_gamma_from_crtc_state(const struct drm_crtc_state *crtc_state, - struct dc_plane_state *plane_state) -{ - int i; - struct dc_gamma *gamma; - struct drm_color_lut *lut = - (struct drm_color_lut *) crtc_state->gamma_lut->data; - - gamma = dc_create_gamma(); - - if (gamma == NULL) { - WARN_ON(1); - return; - } - - gamma->type = GAMMA_RGB_256; - gamma->num_entries = GAMMA_RGB_256_ENTRIES; - for (i = 0; i < GAMMA_RGB_256_ENTRIES; i++) { - gamma->entries.red[i] = dal_fixed31_32_from_int(lut[i].red); - gamma->entries.green[i] = dal_fixed31_32_from_int(lut[i].green); - gamma->entries.blue[i] = dal_fixed31_32_from_int(lut[i].blue); - } - - plane_state->gamma_correction = gamma; -} - static int fill_plane_attributes(struct amdgpu_device *adev, struct dc_plane_state *dc_plane_state, struct drm_plane_state *plane_state, @@ -2035,14 +2037,17 @@ static int fill_plane_attributes(struct amdgpu_device *adev, if (input_tf == NULL) return -ENOMEM; - input_tf->type = TF_TYPE_PREDEFINED; - input_tf->tf = TRANSFER_FUNCTION_SRGB; - dc_plane_state->in_transfer_func = input_tf; - /* In case of gamma set, update gamma value */ - if (crtc_state->gamma_lut) - fill_gamma_from_crtc_state(crtc_state, dc_plane_state); + /* + * Always set input transfer function, since plane state is refreshed + * every time. + */ + ret = amdgpu_dm_set_degamma_lut(crtc_state, dc_plane_state); + if (ret) { + dc_transfer_func_release(dc_plane_state->in_transfer_func); + dc_plane_state->in_transfer_func = NULL; + } return ret; } @@ -2376,7 +2381,7 @@ static void set_master_stream(struct dc_stream_state *stream_set[], } } for (j = 0; j < stream_count; j++) { - if (stream_set[j] && j != master_stream) + if (stream_set[j]) stream_set[j]->triggered_crtc_reset.event_source = stream_set[master_stream]; } } @@ -2703,7 +2708,8 @@ static void amdgpu_dm_connector_destroy(struct drm_connector *connector) #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\ defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) - if (link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) { + if ((link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) && + link->type != dc_connection_none) { amdgpu_dm_register_backlight_device(dm); if (dm->backlight_dev) { @@ -2853,6 +2859,7 @@ int amdgpu_dm_connector_mode_valid(struct drm_connector *connector, /* TODO: Unhardcode stream count */ struct dc_stream_state *stream; struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector); + enum dc_status dc_result = DC_OK; if ((mode->flags & DRM_MODE_FLAG_INTERLACE) || (mode->flags & DRM_MODE_FLAG_DBLSCAN)) @@ -2885,8 +2892,16 @@ int amdgpu_dm_connector_mode_valid(struct drm_connector *connector, stream->src.height = mode->vdisplay; stream->dst = stream->src; - if (dc_validate_stream(adev->dm.dc, stream) == DC_OK) + dc_result = dc_validate_stream(adev->dm.dc, stream); + + if (dc_result == DC_OK) result = MODE_OK; + else + DRM_DEBUG_KMS("Mode %dx%d (clk %d) failed DC validation with error %d\n", + mode->vdisplay, + mode->hdisplay, + mode->clock, + dc_result); dc_stream_release(stream); @@ -3260,7 +3275,9 @@ static int amdgpu_dm_crtc_init(struct amdgpu_display_manager *dm, acrtc->base.enabled = false; dm->adev->mode_info.crtcs[crtc_index] = acrtc; - drm_mode_crtc_set_gamma_size(&acrtc->base, 256); + drm_crtc_enable_color_mgmt(&acrtc->base, MAX_COLOR_LUT_ENTRIES, + true, MAX_COLOR_LUT_ENTRIES); + drm_mode_crtc_set_gamma_size(&acrtc->base, MAX_COLOR_LEGACY_LUT_ENTRIES); return 0; @@ -3568,6 +3585,7 @@ create_i2c(struct ddc_service *ddc_service, return i2c; } + /* Note: this function assumes that dc_link_detect() was called for the * dc_link which will be represented by this aconnector. */ @@ -3644,28 +3662,6 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm, || connector_type == DRM_MODE_CONNECTOR_eDP) amdgpu_dm_initialize_dp_connector(dm, aconnector); -#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\ - defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) - - /* NOTE: this currently will create backlight device even if a panel - * is not connected to the eDP/LVDS connector. - * - * This is less than ideal but we don't have sink information at this - * stage since detection happens after. We can't do detection earlier - * since MST detection needs connectors to be created first. - */ - if (link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) { - /* Event if registration failed, we should continue with - * DM initialization because not having a backlight control - * is better then a black screen. - */ - amdgpu_dm_register_backlight_device(dm); - - if (dm->backlight_dev) - dm->backlight_link = link; - } -#endif - out_free: if (res) { kfree(i2c); @@ -3726,7 +3722,7 @@ static void manage_dm_interrupts(struct amdgpu_device *adev, * constant is the same as PFLIP */ int irq_type = - amdgpu_crtc_idx_to_irq_type( + amdgpu_display_crtc_idx_to_irq_type( adev, acrtc->crtc_id); @@ -3921,7 +3917,7 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc, /* Prepare wait for target vblank early - before the fence-waits */ - target_vblank = target - drm_crtc_vblank_count(crtc) + + target_vblank = target - (uint32_t)drm_crtc_vblank_count(crtc) + amdgpu_get_vblank_counter_kms(crtc->dev, acrtc->crtc_id); /* TODO This might fail and hence better not used, wait @@ -3945,9 +3941,9 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc, * targeted by the flip */ while ((acrtc->enabled && - (amdgpu_get_crtc_scanoutpos(adev->ddev, acrtc->crtc_id, 0, - &vpos, &hpos, NULL, NULL, - &crtc->hwmode) + (amdgpu_display_get_crtc_scanoutpos(adev->ddev, acrtc->crtc_id, + 0, &vpos, &hpos, NULL, + NULL, &crtc->hwmode) & (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK)) == (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK) && (int)(target_vblank - @@ -4065,7 +4061,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, amdgpu_dm_do_flip( crtc, fb, - drm_crtc_vblank_count(crtc) + *wait_for_vblank, + (uint32_t)drm_crtc_vblank_count(crtc) + *wait_for_vblank, dm_state->context); } @@ -4565,8 +4561,8 @@ static int dm_update_crtcs_state(struct dc *dc, for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) { struct amdgpu_crtc *acrtc = NULL; struct amdgpu_dm_connector *aconnector = NULL; - struct drm_connector_state *new_con_state = NULL; - struct dm_connector_state *dm_conn_state = NULL; + struct drm_connector_state *drm_new_conn_state = NULL, *drm_old_conn_state = NULL; + struct dm_connector_state *dm_new_conn_state = NULL, *dm_old_conn_state = NULL; struct drm_plane_state *new_plane_state = NULL; new_stream = NULL; @@ -4587,19 +4583,23 @@ static int dm_update_crtcs_state(struct dc *dc, /* TODO This hack should go away */ if (aconnector && enable) { // Make sure fake sink is created in plug-in scenario - new_con_state = drm_atomic_get_connector_state(state, + drm_new_conn_state = drm_atomic_get_new_connector_state(state, &aconnector->base); + drm_old_conn_state = drm_atomic_get_old_connector_state(state, + &aconnector->base); - if (IS_ERR(new_con_state)) { - ret = PTR_ERR_OR_ZERO(new_con_state); + + if (IS_ERR(drm_new_conn_state)) { + ret = PTR_ERR_OR_ZERO(drm_new_conn_state); break; } - dm_conn_state = to_dm_connector_state(new_con_state); + dm_new_conn_state = to_dm_connector_state(drm_new_conn_state); + dm_old_conn_state = to_dm_connector_state(drm_old_conn_state); new_stream = create_stream_for_sink(aconnector, &new_crtc_state->mode, - dm_conn_state); + dm_new_conn_state); /* * we can have no stream on ACTION_SET if a display @@ -4700,6 +4700,40 @@ static int dm_update_crtcs_state(struct dc *dc, /* Release extra reference */ if (new_stream) dc_stream_release(new_stream); + + /* + * We want to do dc stream updates that do not require a + * full modeset below. + */ + if (!(enable && aconnector && new_crtc_state->enable && + new_crtc_state->active)) + continue; + /* + * Given above conditions, the dc state cannot be NULL because: + * 1. We're in the process of enabling CRTCs (just been added + * to the dc context, or already is on the context) + * 2. Has a valid connector attached, and + * 3. Is currently active and enabled. + * => The dc stream state currently exists. + */ + BUG_ON(dm_new_crtc_state->stream == NULL); + + /* Scaling or underscan settings */ + if (is_scaling_state_different(dm_old_conn_state, dm_new_conn_state)) + update_stream_scaling_settings( + &new_crtc_state->mode, dm_new_conn_state, dm_new_crtc_state->stream); + + /* + * Color management settings. We also update color properties + * when a modeset is needed, to ensure it gets reprogrammed. + */ + if (dm_new_crtc_state->base.color_mgmt_changed || + drm_atomic_crtc_needs_modeset(new_crtc_state)) { + ret = amdgpu_dm_set_regamma_lut(dm_new_crtc_state); + if (ret) + goto fail; + amdgpu_dm_set_ctm(dm_new_crtc_state); + } } return ret; @@ -4796,10 +4830,8 @@ static int dm_update_planes_state(struct dc *dc, WARN_ON(dm_new_plane_state->dc_state); dc_new_plane_state = dc_create_plane_state(dc); - if (!dc_new_plane_state) { - ret = -EINVAL; - return ret; - } + if (!dc_new_plane_state) + return -ENOMEM; DRM_DEBUG_DRIVER("Enabling DRM plane: %d on DRM crtc %d\n", plane->base.id, new_plane_crtc->base.id); @@ -4828,8 +4860,7 @@ static int dm_update_planes_state(struct dc *dc, dm_state->context)) { dc_plane_state_release(dc_new_plane_state); - ret = -EINVAL; - return ret; + return -EINVAL; } dm_new_plane_state->dc_state = dc_new_plane_state;
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h index 93b840c..e78e1aa 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -108,8 +108,6 @@ struct amdgpu_display_manager { struct dal *dal; struct dc *dc; struct cgs_device *cgs_device; - /* lock to be used when DAL is called from SYNC IRQ context */ - spinlock_t dal_lock; struct amdgpu_device *adev; /*AMD base driver*/ struct drm_device *ddev; /*DRM base driver*/ @@ -142,17 +140,6 @@ struct amdgpu_display_manager { /* this spin lock synchronizes access to 'irq_handler_list_table' */ spinlock_t irq_handler_list_table_lock; - /* Timer-related data. */ - struct list_head timer_handler_list; - struct workqueue_struct *timer_workqueue; - - /* Use dal_mutex for any activity which is NOT syncronized by - * DRM mode setting locks. - * For example: amdgpu_dm_hpd_low_irq() calls into DAL *without* - * DRM mode setting locks being acquired. This is where dal_mutex - * is acquired before calling into DAL. */ - struct mutex dal_mutex; - struct backlight_device *backlight_dev; const struct dc_link *backlight_link; @@ -234,7 +221,7 @@ struct dm_crtc_state { struct drm_crtc_state base; struct dc_stream_state *stream; - bool crc_first_skipped; + int crc_skip_count; bool crc_enabled; }; @@ -301,9 +288,19 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name, void amdgpu_dm_crtc_handle_crc_irq(struct drm_crtc *crtc); #else #define amdgpu_dm_crtc_set_crc_source NULL -void amdgpu_dm_crtc_handle_crc_irq(struct drm_crtc *crtc) {} +#define amdgpu_dm_crtc_handle_crc_irq(x) #endif +#define MAX_COLOR_LUT_ENTRIES 4096 +/* Legacy gamm LUT users such as X doesn't like large LUT sizes */ +#define MAX_COLOR_LEGACY_LUT_ENTRIES 256 + +void amdgpu_dm_init_color_mod(void); +int amdgpu_dm_set_degamma_lut(struct drm_crtc_state *crtc_state, + struct dc_plane_state *dc_plane_state); +void amdgpu_dm_set_ctm(struct dm_crtc_state *crtc); +int amdgpu_dm_set_regamma_lut(struct dm_crtc_state *crtc); + extern const struct drm_encoder_helper_funcs amdgpu_dm_encoder_helper_funcs; #endif /* __AMDGPU_DM_H__ */
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c new file mode 100644 index 0000000..25f064c --- /dev/null +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -0,0 +1,267 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#include "amdgpu_mode.h" +#include "amdgpu_dm.h" +#include "modules/color/color_gamma.h" + +#define MAX_DRM_LUT_VALUE 0xFFFF + +/* + * Initialize the color module. + * + * We're not using the full color module, only certain components. + * Only call setup functions for components that we need. + */ +void amdgpu_dm_init_color_mod(void) +{ + setup_x_points_distribution(); +} + + +/* + * Return true if the given lut is a linear mapping of values, i.e. it acts + * like a bypass LUT. + * + * It is considered linear if the lut represents: + * f(a) = (0xFF00/MAX_COLOR_LUT_ENTRIES-1)a; for integer a in + * [0, MAX_COLOR_LUT_ENTRIES) + */ +static bool __is_lut_linear(struct drm_color_lut *lut, uint32_t size) +{ + int i; + uint32_t expected; + int delta; + + for (i = 0; i < size; i++) { + /* All color values should equal */ + if ((lut[i].red != lut[i].green) || (lut[i].green != lut[i].blue)) + return false; + + expected = i * MAX_DRM_LUT_VALUE / (size-1); + + /* Allow a +/-1 error. */ + delta = lut[i].red - expected; + if (delta < -1 || 1 < delta) + return false; + } + return true; +} + +/** + * Convert the drm_color_lut to dc_gamma. The conversion depends on the size + * of the lut - whether or not it's legacy. + */ +static void __drm_lut_to_dc_gamma(struct drm_color_lut *lut, + struct dc_gamma *gamma, + bool is_legacy) +{ + uint32_t r, g, b; + int i; + + if (is_legacy) { + for (i = 0; i < MAX_COLOR_LEGACY_LUT_ENTRIES; i++) { + r = drm_color_lut_extract(lut[i].red, 16); + g = drm_color_lut_extract(lut[i].green, 16); + b = drm_color_lut_extract(lut[i].blue, 16); + + gamma->entries.red[i] = dal_fixed31_32_from_int(r); + gamma->entries.green[i] = dal_fixed31_32_from_int(g); + gamma->entries.blue[i] = dal_fixed31_32_from_int(b); + } + return; + } + + /* else */ + for (i = 0; i < MAX_COLOR_LUT_ENTRIES; i++) { + r = drm_color_lut_extract(lut[i].red, 16); + g = drm_color_lut_extract(lut[i].green, 16); + b = drm_color_lut_extract(lut[i].blue, 16); + + gamma->entries.red[i] = dal_fixed31_32_from_fraction(r, MAX_DRM_LUT_VALUE); + gamma->entries.green[i] = dal_fixed31_32_from_fraction(g, MAX_DRM_LUT_VALUE); + gamma->entries.blue[i] = dal_fixed31_32_from_fraction(b, MAX_DRM_LUT_VALUE); + } +} + +/** + * amdgpu_dm_set_regamma_lut: Set regamma lut for the given CRTC. + * @crtc: amdgpu_dm crtc state + * + * Update the underlying dc_stream_state's output transfer function (OTF) in + * preparation for hardware commit. If no lut is specified by user, we default + * to SRGB. + * + * RETURNS: + * 0 on success, -ENOMEM if memory cannot be allocated to calculate the OTF. + */ +int amdgpu_dm_set_regamma_lut(struct dm_crtc_state *crtc) +{ + struct drm_property_blob *blob = crtc->base.gamma_lut; + struct dc_stream_state *stream = crtc->stream; + struct drm_color_lut *lut; + uint32_t lut_size; + struct dc_gamma *gamma; + enum dc_transfer_func_type old_type = stream->out_transfer_func->type; + + bool ret; + + if (!blob) { + /* By default, use the SRGB predefined curve.*/ + stream->out_transfer_func->type = TF_TYPE_PREDEFINED; + stream->out_transfer_func->tf = TRANSFER_FUNCTION_SRGB; + return 0; + } + + lut = (struct drm_color_lut *)blob->data; + lut_size = blob->length / sizeof(struct drm_color_lut); + + gamma = dc_create_gamma(); + if (!gamma) + return -ENOMEM; + + gamma->num_entries = lut_size; + if (gamma->num_entries == MAX_COLOR_LEGACY_LUT_ENTRIES) + gamma->type = GAMMA_RGB_256; + else if (gamma->num_entries == MAX_COLOR_LUT_ENTRIES) + gamma->type = GAMMA_CS_TFM_1D; + else { + /* Invalid lut size */ + dc_gamma_release(&gamma); + return -EINVAL; + } + + /* Convert drm_lut into dc_gamma */ + __drm_lut_to_dc_gamma(lut, gamma, gamma->type == GAMMA_RGB_256); + + /* Call color module to translate into something DC understands. Namely + * a transfer function. + */ + stream->out_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; + ret = mod_color_calculate_regamma_params(stream->out_transfer_func, + gamma, true); + dc_gamma_release(&gamma); + if (!ret) { + stream->out_transfer_func->type = old_type; + DRM_ERROR("Out of memory when calculating regamma params\n"); + return -ENOMEM; + } + + return 0; +} + +/** + * amdgpu_dm_set_ctm: Set the color transform matrix for the given CRTC. + * @crtc: amdgpu_dm crtc state + * + * Update the underlying dc_stream_state's gamut remap matrix in preparation + * for hardware commit. If no matrix is specified by user, gamut remap will be + * disabled. + */ +void amdgpu_dm_set_ctm(struct dm_crtc_state *crtc) +{ + + struct drm_property_blob *blob = crtc->base.ctm; + struct dc_stream_state *stream = crtc->stream; + struct drm_color_ctm *ctm; + int64_t val; + int i; + + if (!blob) { + stream->gamut_remap_matrix.enable_remap = false; + return; + } + + stream->gamut_remap_matrix.enable_remap = true; + ctm = (struct drm_color_ctm *)blob->data; + /* + * DRM gives a 3x3 matrix, but DC wants 3x4. Assuming we're operating + * with homogeneous coordinates, augment the matrix with 0's. + * + * The format provided is S31.32, using signed-magnitude representation. + * Our fixed31_32 is also S31.32, but is using 2's complement. We have + * to convert from signed-magnitude to 2's complement. + */ + for (i = 0; i < 12; i++) { + /* Skip 4th element */ + if (i % 4 == 3) { + stream->gamut_remap_matrix.matrix[i] = dal_fixed31_32_zero; + continue; + } + + /* gamut_remap_matrix[i] = ctm[i - floor(i/4)] */ + val = ctm->matrix[i - (i/4)]; + /* If negative, convert to 2's complement. */ + if (val & (1ULL << 63)) + val = -(val & ~(1ULL << 63)); + + stream->gamut_remap_matrix.matrix[i].value = val; + } +} + + +/** + * amdgpu_dm_set_degamma_lut: Set degamma lut for the given CRTC. + * @crtc: amdgpu_dm crtc state + * + * Update the underlying dc_stream_state's input transfer function (ITF) in + * preparation for hardware commit. If no lut is specified by user, we default + * to SRGB degamma. + * + * Currently, we only support degamma bypass, or preprogrammed SRGB degamma. + * Programmable degamma is not supported, and an attempt to do so will return + * -EINVAL. + * + * RETURNS: + * 0 on success, -EINVAL if custom degamma curve is given. + */ +int amdgpu_dm_set_degamma_lut(struct drm_crtc_state *crtc_state, + struct dc_plane_state *dc_plane_state) +{ + struct drm_property_blob *blob = crtc_state->degamma_lut; + struct drm_color_lut *lut; + + if (!blob) { + /* Default to SRGB */ + dc_plane_state->in_transfer_func->type = TF_TYPE_PREDEFINED; + dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_SRGB; + return 0; + } + + lut = (struct drm_color_lut *)blob->data; + if (__is_lut_linear(lut, MAX_COLOR_LUT_ENTRIES)) { + dc_plane_state->in_transfer_func->type = TF_TYPE_BYPASS; + dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; + return 0; + } + + /* Otherwise, assume SRGB, since programmable degamma is not + * supported. + */ + dc_plane_state->in_transfer_func->type = TF_TYPE_PREDEFINED; + dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_SRGB; + return -EINVAL; +} +
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c index 4caa2f8..9bfb040 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
@@ -60,25 +60,32 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name, return -EINVAL; } + /* When enabling CRC, we should also disable dithering. */ if (source == AMDGPU_DM_PIPE_CRC_SOURCE_AUTO) { if (dc_stream_configure_crc(stream_state->ctx->dc, stream_state, - true, true)) + true, true)) { crtc_state->crc_enabled = true; + dc_stream_set_dither_option(stream_state, + DITHER_OPTION_TRUN8); + } else return -EINVAL; } else { if (dc_stream_configure_crc(stream_state->ctx->dc, stream_state, - false, false)) + false, false)) { crtc_state->crc_enabled = false; + dc_stream_set_dither_option(stream_state, + DITHER_OPTION_DEFAULT); + } else return -EINVAL; } *values_cnt = 3; - /* Reset crc_skipped flag on dm state */ - crtc_state->crc_first_skipped = false; + /* Reset crc_skipped on dm state */ + crtc_state->crc_skip_count = 0; return 0; } @@ -108,11 +115,11 @@ void amdgpu_dm_crtc_handle_crc_irq(struct drm_crtc *crtc) /* * Since flipping and crc enablement happen asynchronously, we - more * often than not - will be returning an 'uncooked' crc on first frame. - * Probably because hw isn't ready yet. Simply skip the first crc - * value. + * Probably because hw isn't ready yet. For added security, skip the + * first two CRC values. */ - if (!crtc_state->crc_first_skipped) { - crtc_state->crc_first_skipped = true; + if (crtc_state->crc_skip_count < 2) { + crtc_state->crc_skip_count += 1; return; }
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c index 6f90287..174424a 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -258,6 +258,15 @@ bool dm_helpers_dp_mst_write_payload_allocation_table( return true; } + +/* + * Clear payload allocation table before enable MST DP link. + */ +void dm_helpers_dp_mst_clear_payload_allocation_table( + struct dc_context *ctx, + const struct dc_link *link) +{} + /* * Polls for ACT (allocation change trigger) handled and sends * ALLOCATE_PAYLOAD message. @@ -524,3 +533,8 @@ enum dc_edid_status dm_helpers_read_local_edid( return edid_status; } + +void dm_set_dcn_clocks(struct dc_context *ctx, struct dc_clocks *clks) +{ + /* TODO: something */ +}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c index 4220550..4be21bf5 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
@@ -51,11 +51,6 @@ struct amdgpu_dm_irq_handler_data { enum dc_irq_source irq_source; }; -struct amdgpu_dm_timer_handler_data { - struct handler_common_data hcd; - struct delayed_work d_work; -}; - #define DM_IRQ_TABLE_LOCK(adev, flags) \ spin_lock_irqsave(&adev->dm.irq_handler_list_table_lock, flags) @@ -169,62 +164,6 @@ static struct list_head *remove_irq_handler(struct amdgpu_device *adev, return hnd_list; } -/* If 'handler_in == NULL' then remove ALL handlers. */ -static void remove_timer_handler(struct amdgpu_device *adev, - struct amdgpu_dm_timer_handler_data *handler_in) -{ - struct amdgpu_dm_timer_handler_data *handler_temp; - struct list_head *handler_list; - struct list_head *entry, *tmp; - unsigned long irq_table_flags; - bool handler_removed = false; - - DM_IRQ_TABLE_LOCK(adev, irq_table_flags); - - handler_list = &adev->dm.timer_handler_list; - - list_for_each_safe(entry, tmp, handler_list) { - /* Note that list_for_each_safe() guarantees that - * handler_temp is NOT null. */ - handler_temp = list_entry(entry, - struct amdgpu_dm_timer_handler_data, hcd.list); - - if (handler_in == NULL || handler_in == handler_temp) { - list_del(&handler_temp->hcd.list); - DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags); - - DRM_DEBUG_KMS("DM_IRQ: removing timer handler: %p\n", - handler_temp); - - if (handler_in == NULL) { - /* Since it is still in the queue, it must - * be cancelled. */ - cancel_delayed_work_sync(&handler_temp->d_work); - } - - kfree(handler_temp); - handler_removed = true; - - DM_IRQ_TABLE_LOCK(adev, irq_table_flags); - } - - /* Remove ALL handlers. */ - if (handler_in == NULL) - continue; - - /* Remove a SPECIFIC handler. - * Found our handler - we can stop here. */ - if (handler_in == handler_temp) - break; - } - - DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags); - - if (handler_in != NULL && handler_removed == false) - DRM_ERROR("DM_IRQ: handler: %p is not in the list!\n", - handler_in); -} - static bool validate_irq_registration_params(struct dc_interrupt_params *int_params, void (*ih)(void *)) @@ -382,16 +321,6 @@ int amdgpu_dm_irq_init(struct amdgpu_device *adev) INIT_LIST_HEAD(&adev->dm.irq_handler_list_high_tab[src]); } - INIT_LIST_HEAD(&adev->dm.timer_handler_list); - - /* allocate and initialize the workqueue for DM timer */ - adev->dm.timer_workqueue = create_singlethread_workqueue( - "dm_timer_queue"); - if (adev->dm.timer_workqueue == NULL) { - DRM_ERROR("DM_IRQ: unable to create timer queue!\n"); - return -1; - } - return 0; } @@ -400,21 +329,17 @@ void amdgpu_dm_irq_fini(struct amdgpu_device *adev) { int src; struct irq_list_head *lh; + unsigned long irq_table_flags; DRM_DEBUG_KMS("DM_IRQ: releasing resources.\n"); - for (src = 0; src < DAL_IRQ_SOURCES_NUMBER; src++) { - + DM_IRQ_TABLE_LOCK(adev, irq_table_flags); /* The handler was removed from the table, * it means it is safe to flush all the 'work' * (because no code can schedule a new one). */ lh = &adev->dm.irq_handler_list_low_tab[src]; + DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags); flush_work(&lh->work); } - - /* Cancel ALL timers and release handlers (if any). */ - remove_timer_handler(adev, NULL); - /* Release the queue itself. */ - destroy_workqueue(adev->dm.timer_workqueue); } int amdgpu_dm_irq_suspend(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index 79a2373..94e06a52 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -88,6 +88,9 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, if (WARN_ON(msg->size > 16)) return -E2BIG; + if (WARN_ON(msg->size > 16)) + return -E2BIG; + switch (msg->request & ~DP_AUX_I2C_MOT) { case DP_AUX_NATIVE_READ: res = dal_ddc_service_read_dpcd_data( @@ -165,6 +168,11 @@ dm_dp_mst_connector_destroy(struct drm_connector *connector) struct amdgpu_dm_connector *amdgpu_dm_connector = to_amdgpu_dm_connector(connector); struct amdgpu_encoder *amdgpu_encoder = amdgpu_dm_connector->mst_encoder; + if (amdgpu_dm_connector->edid) { + kfree(amdgpu_dm_connector->edid); + amdgpu_dm_connector->edid = NULL; + } + drm_encoder_cleanup(&amdgpu_encoder->base); kfree(amdgpu_encoder); drm_connector_cleanup(connector); @@ -182,41 +190,25 @@ static const struct drm_connector_funcs dm_dp_mst_connector_funcs = { .atomic_get_property = amdgpu_dm_connector_atomic_get_property }; -static int dm_connector_update_modes(struct drm_connector *connector, - struct edid *edid) -{ - int ret; - - ret = drm_add_edid_modes(connector, edid); - - return ret; -} - void dm_dp_mst_dc_sink_create(struct drm_connector *connector) { struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector); - struct edid *edid; struct dc_sink *dc_sink; struct dc_sink_init_data init_params = { .link = aconnector->dc_link, .sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST }; + /* FIXME none of this is safe. we shouldn't touch aconnector here in + * atomic_check + */ + /* * TODO: Need to further figure out why ddc.algo is NULL while MST port exists */ if (!aconnector->port || !aconnector->port->aux.ddc.algo) return; - edid = drm_dp_mst_get_edid(connector, &aconnector->mst_port->mst_mgr, aconnector->port); - - if (!edid) { - drm_mode_connector_update_edid_property( - &aconnector->base, - NULL); - return; - } - - aconnector->edid = edid; + ASSERT(aconnector->edid); dc_sink = dc_link_add_remote_sink( aconnector->dc_link, @@ -229,9 +221,6 @@ void dm_dp_mst_dc_sink_create(struct drm_connector *connector) amdgpu_dm_add_sink_to_freesync_module( connector, aconnector->edid); - - drm_mode_connector_update_edid_property( - &aconnector->base, aconnector->edid); } static int dm_dp_mst_get_modes(struct drm_connector *connector) @@ -240,14 +229,10 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) int ret = 0; if (!aconnector) - return dm_connector_update_modes(connector, NULL); + return drm_add_edid_modes(connector, NULL); if (!aconnector->edid) { struct edid *edid; - struct dc_sink *dc_sink; - struct dc_sink_init_data init_params = { - .link = aconnector->dc_link, - .sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST }; edid = drm_dp_mst_get_edid(connector, &aconnector->mst_port->mst_mgr, aconnector->port); if (!edid) { @@ -258,11 +243,17 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) } aconnector->edid = edid; + } + if (!aconnector->dc_sink) { + struct dc_sink *dc_sink; + struct dc_sink_init_data init_params = { + .link = aconnector->dc_link, + .sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST }; dc_sink = dc_link_add_remote_sink( aconnector->dc_link, - (uint8_t *)edid, - (edid->extensions + 1) * EDID_LENGTH, + (uint8_t *)aconnector->edid, + (aconnector->edid->extensions + 1) * EDID_LENGTH, &init_params); dc_sink->priv = aconnector; @@ -270,13 +261,13 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) if (aconnector->dc_sink) amdgpu_dm_add_sink_to_freesync_module( - connector, edid); - - drm_mode_connector_update_edid_property( - &aconnector->base, edid); + connector, aconnector->edid); } - ret = dm_connector_update_modes(connector, aconnector->edid); + drm_mode_connector_update_edid_property( + &aconnector->base, aconnector->edid); + + ret = drm_add_edid_modes(connector, aconnector->edid); return ret; } @@ -438,14 +429,6 @@ static void dm_dp_destroy_mst_connector(struct drm_dp_mst_topology_mgr *mgr, dc_sink_release(aconnector->dc_sink); aconnector->dc_sink = NULL; } - if (aconnector->edid) { - kfree(aconnector->edid); - aconnector->edid = NULL; - } - - drm_mode_connector_update_edid_property( - &aconnector->base, - NULL); aconnector->mst_connected = false; }
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c index 56e5492..89342b4 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
@@ -71,15 +71,6 @@ bool dm_read_persistent_data(struct dc_context *ctx, /**** power component interfaces ****/ -bool dm_pp_pre_dce_clock_change( - struct dc_context *ctx, - struct dm_pp_gpu_clock_range *requested_state, - struct dm_pp_gpu_clock_range *actual_state) -{ - /*TODO*/ - return false; -} - bool dm_pp_apply_display_requirements( const struct dc_context *ctx, const struct dm_pp_display_configuration *pp_display_cfg) @@ -151,30 +142,6 @@ bool dm_pp_apply_display_requirements( return true; } -bool dc_service_get_system_clocks_range( - const struct dc_context *ctx, - struct dm_pp_gpu_clock_range *sys_clks) -{ - struct amdgpu_device *adev = ctx->driver_context; - - /* Default values, in case PPLib is not compiled-in. */ - sys_clks->mclk.max_khz = 800000; - sys_clks->mclk.min_khz = 800000; - - sys_clks->sclk.max_khz = 600000; - sys_clks->sclk.min_khz = 300000; - - if (adev->pm.dpm_enabled) { - sys_clks->mclk.max_khz = amdgpu_dpm_get_mclk(adev, false); - sys_clks->mclk.min_khz = amdgpu_dpm_get_mclk(adev, true); - - sys_clks->sclk.max_khz = amdgpu_dpm_get_sclk(adev, false); - sys_clks->sclk.min_khz = amdgpu_dpm_get_sclk(adev, true); - } - - return true; -} - static void get_default_clock_levels( enum dm_pp_clock_type clk_type, struct dm_pp_clock_levels *clks)
diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c index 011a97f..8a9bba8 100644 --- a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c +++ b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
@@ -593,3 +593,12 @@ uint32_t dal_fixed31_32_clamp_u0d10( { return clamp_ux_dy(arg.value, 0, 10, 1); } + +int32_t dal_fixed31_32_s4d19( + struct fixed31_32 arg) +{ + if (arg.value < 0) + return -(int32_t)ux_dy(dal_fixed31_32_abs(arg).value, 4, 19); + else + return ux_dy(arg.value, 4, 19); +}
diff --git a/drivers/gpu/drm/amd/display/dc/basics/logger.c b/drivers/gpu/drm/amd/display/dc/basics/logger.c index 180a9d6..31bee054 100644 --- a/drivers/gpu/drm/amd/display/dc/basics/logger.c +++ b/drivers/gpu/drm/amd/display/dc/basics/logger.c
@@ -60,7 +60,8 @@ static const struct dc_log_type_info log_type_info_tbl[] = { {LOG_EVENT_LINK_LOSS, "LinkLoss"}, {LOG_EVENT_UNDERFLOW, "Underflow"}, {LOG_IF_TRACE, "InterfaceTrace"}, - {LOG_DTN, "DTN"} + {LOG_DTN, "DTN"}, + {LOG_PROFILING, "Profiling"} };
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c index c00e405..c7f0b27e4 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
@@ -49,6 +49,9 @@ #define LAST_RECORD_TYPE 0xff +#define DC_LOGGER \ + bp->base.ctx->logger + /* GUID to validate external display connection info table (aka OPM module) */ static const uint8_t ext_display_connection_guid[NUMBER_OF_UCHAR_FOR_GUID] = { 0x91, 0x6E, 0x57, 0x09, @@ -3079,8 +3082,7 @@ static enum bp_result patch_bios_image_from_ext_display_connection_info( opm_object, &ext_display_connection_info_tbl) != BP_RESULT_OK) { - dm_logger_write(bp->base.ctx->logger, LOG_WARNING, - "%s: Failed to read Connection Info Table", __func__); + DC_LOG_WARNING("%s: Failed to read Connection Info Table", __func__); return BP_RESULT_UNSUPPORTED; } @@ -3795,14 +3797,11 @@ static const struct dc_vbios_funcs vbios_funcs = { .get_gpio_pin_info = bios_parser_get_gpio_pin_info, - .get_embedded_panel_info = bios_parser_get_embedded_panel_info, - - .get_gpio_pin_info = bios_parser_get_gpio_pin_info, - .get_encoder_cap_info = bios_parser_get_encoder_cap_info, /* bios scratch register communication */ .is_accelerated_mode = bios_is_accelerated_mode, + .get_vga_enabled_displays = bios_get_vga_enabled_displays, .set_scratch_critical_state = bios_parser_set_scratch_critical_state,
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c index 1ee1717..10a5807 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -44,7 +44,7 @@ #include "bios_parser_common.h" #define LAST_RECORD_TYPE 0xff - +#define SMU9_SYSPLL0_ID 0 struct i2c_id_config_access { uint8_t bfI2C_LineMux:4; @@ -70,6 +70,10 @@ static enum bp_result get_firmware_info_v3_1( struct bios_parser *bp, struct dc_firmware_info *info); +static enum bp_result get_firmware_info_v3_2( + struct bios_parser *bp, + struct dc_firmware_info *info); + static struct atom_hpd_int_record *get_hpd_record(struct bios_parser *bp, struct atom_display_object_path_v2 *object); @@ -1220,7 +1224,7 @@ static unsigned int bios_parser_get_smu_clock_info( if (!bp->cmd_tbl.get_smu_clock_info) return BP_RESULT_FAILURE; - return bp->cmd_tbl.get_smu_clock_info(bp); + return bp->cmd_tbl.get_smu_clock_info(bp, 0); } static enum bp_result bios_parser_program_crtc_timing( @@ -1280,6 +1284,12 @@ static bool bios_parser_is_accelerated_mode( return bios_is_accelerated_mode(dcb); } +static uint32_t bios_parser_get_vga_enabled_displays( + struct dc_bios *bios) +{ + return bios_get_vga_enabled_displays(bios); +} + /** * bios_parser_set_scratch_critical_state @@ -1317,6 +1327,9 @@ static enum bp_result bios_parser_get_firmware_info( case 1: result = get_firmware_info_v3_1(bp, info); break; + case 2: + result = get_firmware_info_v3_2(bp, info); + break; default: break; } @@ -1370,7 +1383,85 @@ static enum bp_result get_firmware_info_v3_1( if (bp->cmd_tbl.get_smu_clock_info != NULL) { /* VBIOS gives in 10KHz */ info->smu_gpu_pll_output_freq = - bp->cmd_tbl.get_smu_clock_info(bp) * 10; + bp->cmd_tbl.get_smu_clock_info(bp, SMU9_SYSPLL0_ID) * 10; + } + + return BP_RESULT_OK; +} + +static enum bp_result get_firmware_info_v3_2( + struct bios_parser *bp, + struct dc_firmware_info *info) +{ + struct atom_firmware_info_v3_2 *firmware_info; + struct atom_display_controller_info_v4_1 *dce_info = NULL; + struct atom_common_table_header *header; + struct atom_data_revision revision; + struct atom_smu_info_v3_2 *smu_info_v3_2 = NULL; + struct atom_smu_info_v3_3 *smu_info_v3_3 = NULL; + + if (!info) + return BP_RESULT_BADINPUT; + + firmware_info = GET_IMAGE(struct atom_firmware_info_v3_2, + DATA_TABLES(firmwareinfo)); + + dce_info = GET_IMAGE(struct atom_display_controller_info_v4_1, + DATA_TABLES(dce_info)); + + if (!firmware_info || !dce_info) + return BP_RESULT_BADBIOSTABLE; + + memset(info, 0, sizeof(*info)); + + header = GET_IMAGE(struct atom_common_table_header, + DATA_TABLES(smu_info)); + get_atom_data_table_revision(header, &revision); + + if (revision.minor == 2) { + /* Vega12 */ + smu_info_v3_2 = GET_IMAGE(struct atom_smu_info_v3_2, + DATA_TABLES(smu_info)); + + if (!smu_info_v3_2) + return BP_RESULT_BADBIOSTABLE; + + info->default_engine_clk = smu_info_v3_2->bootup_dcefclk_10khz * 10; + } else if (revision.minor == 3) { + /* Vega20 */ + smu_info_v3_3 = GET_IMAGE(struct atom_smu_info_v3_3, + DATA_TABLES(smu_info)); + + if (!smu_info_v3_3) + return BP_RESULT_BADBIOSTABLE; + + info->default_engine_clk = smu_info_v3_3->bootup_dcefclk_10khz * 10; + } + + // We need to convert from 10KHz units into KHz units. + info->default_memory_clk = firmware_info->bootup_mclk_in10khz * 10; + + /* 27MHz for Vega10 & Vega12; 100MHz for Vega20 */ + info->pll_info.crystal_frequency = dce_info->dce_refclk_10khz * 10; + /* Hardcode frequency if BIOS gives no DCE Ref Clk */ + if (info->pll_info.crystal_frequency == 0) { + if (revision.minor == 2) + info->pll_info.crystal_frequency = 27000; + else if (revision.minor == 3) + info->pll_info.crystal_frequency = 100000; + } + /*dp_phy_ref_clk is not correct for atom_display_controller_info_v4_2, but we don't use it*/ + info->dp_phy_ref_clk = dce_info->dpphy_refclk_10khz * 10; + info->i2c_engine_ref_clk = dce_info->i2c_engine_refclk_10khz * 10; + + /* Get GPU PLL VCO Clock */ + if (bp->cmd_tbl.get_smu_clock_info != NULL) { + if (revision.minor == 2) + info->smu_gpu_pll_output_freq = + bp->cmd_tbl.get_smu_clock_info(bp, SMU9_SYSPLL0_ID) * 10; + else if (revision.minor == 3) + info->smu_gpu_pll_output_freq = + bp->cmd_tbl.get_smu_clock_info(bp, SMU11_SYSPLL3_0_ID) * 10; } return BP_RESULT_OK; @@ -1800,6 +1891,7 @@ static const struct dc_vbios_funcs vbios_funcs = { .is_accelerated_mode = bios_parser_is_accelerated_mode, + .get_vga_enabled_displays = bios_parser_get_vga_enabled_displays, .set_scratch_critical_state = bios_parser_set_scratch_critical_state,
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.c index 5c9e510..d458947 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.c
@@ -78,5 +78,13 @@ void bios_set_scratch_critical_state( REG_UPDATE(BIOS_SCRATCH_6, S6_CRITICAL_STATE, critial_state); } +uint32_t bios_get_vga_enabled_displays( + struct dc_bios *bios) +{ + uint32_t active_disp = 1; + if (bios->regs->BIOS_SCRATCH_3) /*follow up with other asic, todo*/ + active_disp = REG_READ(BIOS_SCRATCH_3) & 0XFFFF; + return active_disp; +}
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.h b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.h index c0047ef..75a29e6 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.h +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.h
@@ -34,6 +34,7 @@ uint8_t *bios_get_image(struct dc_bios *bp, uint32_t offset, bool bios_is_accelerated_mode(struct dc_bios *bios); void bios_set_scratch_acc_mode_change(struct dc_bios *bios); void bios_set_scratch_critical_state(struct dc_bios *bios, bool state); +uint32_t bios_get_vga_enabled_displays(struct dc_bios *bios); #define GET_IMAGE(type, offset) ((type *) bios_get_image(&bp->base, offset, sizeof(type)))
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c index fea5e83..3f63f71 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
@@ -34,6 +34,8 @@ #include "command_table_helper2.h" #include "bios_parser_helper.h" #include "bios_parser_types_internal2.h" +#define DC_LOGGER \ + bp->base.ctx->logger #define GET_INDEX_INTO_MASTER_TABLE(MasterOrData, FieldName)\ (((char *)(&((\ @@ -239,8 +241,7 @@ static enum bp_result transmitter_control_v1_6( if (cntl->action == TRANSMITTER_CONTROL_ENABLE || cntl->action == TRANSMITTER_CONTROL_ACTIAVATE || cntl->action == TRANSMITTER_CONTROL_DEACTIVATE) { - dm_logger_write(bp->base.ctx->logger, LOG_BIOS,\ - "%s:ps.param.symclk_10khz = %d\n",\ + DC_LOG_BIOS("%s:ps.param.symclk_10khz = %d\n",\ __func__, ps.param.symclk_10khz); } @@ -331,8 +332,7 @@ static enum bp_result set_pixel_clock_v7( (uint8_t) bp->cmd_helper-> transmitter_color_depth_to_atom( bp_params->color_depth); - dm_logger_write(bp->base.ctx->logger, LOG_BIOS,\ - "%s:program display clock = %d"\ + DC_LOG_BIOS("%s:program display clock = %d"\ "colorDepth = %d\n", __func__,\ bp_params->target_pixel_clock, bp_params->color_depth); @@ -772,8 +772,7 @@ static enum bp_result set_dce_clock_v2_1( */ params.param.dceclk_10khz = cpu_to_le32( bp_params->target_clock_frequency / 10); - dm_logger_write(bp->base.ctx->logger, LOG_BIOS, - "%s:target_clock_frequency = %d"\ + DC_LOG_BIOS("%s:target_clock_frequency = %d"\ "clock_type = %d \n", __func__,\ bp_params->target_clock_frequency,\ bp_params->clock_type); @@ -797,7 +796,7 @@ static enum bp_result set_dce_clock_v2_1( ****************************************************************************** *****************************************************************************/ -static unsigned int get_smu_clock_info_v3_1(struct bios_parser *bp); +static unsigned int get_smu_clock_info_v3_1(struct bios_parser *bp, uint8_t id); static void init_get_smu_clock_info(struct bios_parser *bp) { @@ -806,12 +805,13 @@ static void init_get_smu_clock_info(struct bios_parser *bp) } -static unsigned int get_smu_clock_info_v3_1(struct bios_parser *bp) +static unsigned int get_smu_clock_info_v3_1(struct bios_parser *bp, uint8_t id) { struct atom_get_smu_clock_info_parameters_v3_1 smu_input = {0}; struct atom_get_smu_clock_info_output_parameters_v3_1 smu_output; smu_input.command = GET_SMU_CLOCK_INFO_V3_1_GET_PLLVCO_FREQ; + smu_input.syspll_id = id; /* Get Specific Clock */ if (EXEC_BIOS_CMD_TABLE(getsmuclockinfo, smu_input)) {
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table2.h b/drivers/gpu/drm/amd/display/dc/bios/command_table2.h index 59061b8..ec1c0c9 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table2.h +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table2.h
@@ -96,7 +96,7 @@ struct cmd_tbl { struct bios_parser *bp, struct bp_set_dce_clock_parameters *bp_params); unsigned int (*get_smu_clock_info)( - struct bios_parser *bp); + struct bios_parser *bp, uint8_t id); };
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h index 1fab634..4c3789df 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h
@@ -29,38 +29,7 @@ #include "dce80/command_table_helper_dce80.h" #include "dce110/command_table_helper_dce110.h" #include "dce112/command_table_helper_dce112.h" - -struct command_table_helper { - bool (*controller_id_to_atom)(enum controller_id id, uint8_t *atom_id); - uint8_t (*encoder_action_to_atom)( - enum bp_encoder_control_action action); - uint32_t (*encoder_mode_bp_to_atom)(enum signal_type s, - bool enable_dp_audio); - bool (*engine_bp_to_atom)(enum engine_id engine_id, - uint32_t *atom_engine_id); - void (*assign_control_parameter)( - const struct command_table_helper *h, - struct bp_encoder_control *control, - DIG_ENCODER_CONTROL_PARAMETERS_V2 *ctrl_param); - bool (*clock_source_id_to_atom)(enum clock_source_id id, - uint32_t *atom_pll_id); - bool (*clock_source_id_to_ref_clk_src)( - enum clock_source_id id, - uint32_t *ref_clk_src_id); - uint8_t (*transmitter_bp_to_atom)(enum transmitter t); - uint8_t (*encoder_id_to_atom)(enum encoder_id id); - uint8_t (*clock_source_id_to_atom_phy_clk_src_id)( - enum clock_source_id id); - uint8_t (*signal_type_to_atom_dig_mode)(enum signal_type s); - uint8_t (*hpd_sel_to_atom)(enum hpd_source_id id); - uint8_t (*dig_encoder_sel_to_atom)(enum engine_id engine_id); - uint8_t (*phy_id_to_atom)(enum transmitter t); - uint8_t (*disp_power_gating_action_to_atom)( - enum bp_pipe_control_action action); - bool (*dc_clock_type_to_atom)(enum bp_dce_clock_type id, - uint32_t *atom_clock_type); - uint8_t (*transmitter_color_depth_to_atom)(enum transmitter_color_depth id); -}; +#include "command_table_helper_struct.h" bool dal_bios_parser_init_cmd_tbl_helper(const struct command_table_helper **h, enum dce_version dce);
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.h b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.h index 9f587c91..785fcb2 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.h +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.h
@@ -29,35 +29,7 @@ #include "dce80/command_table_helper_dce80.h" #include "dce110/command_table_helper_dce110.h" #include "dce112/command_table_helper2_dce112.h" - -struct command_table_helper { - bool (*controller_id_to_atom)(enum controller_id id, uint8_t *atom_id); - uint8_t (*encoder_action_to_atom)( - enum bp_encoder_control_action action); - uint32_t (*encoder_mode_bp_to_atom)(enum signal_type s, - bool enable_dp_audio); - bool (*engine_bp_to_atom)(enum engine_id engine_id, - uint32_t *atom_engine_id); - bool (*clock_source_id_to_atom)(enum clock_source_id id, - uint32_t *atom_pll_id); - bool (*clock_source_id_to_ref_clk_src)( - enum clock_source_id id, - uint32_t *ref_clk_src_id); - uint8_t (*transmitter_bp_to_atom)(enum transmitter t); - uint8_t (*encoder_id_to_atom)(enum encoder_id id); - uint8_t (*clock_source_id_to_atom_phy_clk_src_id)( - enum clock_source_id id); - uint8_t (*signal_type_to_atom_dig_mode)(enum signal_type s); - uint8_t (*hpd_sel_to_atom)(enum hpd_source_id id); - uint8_t (*dig_encoder_sel_to_atom)(enum engine_id engine_id); - uint8_t (*phy_id_to_atom)(enum transmitter t); - uint8_t (*disp_power_gating_action_to_atom)( - enum bp_pipe_control_action action); - bool (*dc_clock_type_to_atom)(enum bp_dce_clock_type id, - uint32_t *atom_clock_type); - uint8_t (*transmitter_color_depth_to_atom)( - enum transmitter_color_depth id); -}; +#include "command_table_helper_struct.h" bool dal_bios_parser_init_cmd_tbl_helper2(const struct command_table_helper **h, enum dce_version dce);
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper_struct.h b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper_struct.h new file mode 100644 index 0000000..1f2c0a3 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper_struct.h
@@ -0,0 +1,66 @@ +/* + * Copyright 2012-15 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#ifndef __DAL_COMMAND_TABLE_HELPER_STRUCT_H__ +#define __DAL_COMMAND_TABLE_HELPER_STRUCT_H__ + +#include "dce80/command_table_helper_dce80.h" +#include "dce110/command_table_helper_dce110.h" +#include "dce112/command_table_helper_dce112.h" + +struct _DIG_ENCODER_CONTROL_PARAMETERS_V2; +struct command_table_helper { + bool (*controller_id_to_atom)(enum controller_id id, uint8_t *atom_id); + uint8_t (*encoder_action_to_atom)( + enum bp_encoder_control_action action); + uint32_t (*encoder_mode_bp_to_atom)(enum signal_type s, + bool enable_dp_audio); + bool (*engine_bp_to_atom)(enum engine_id engine_id, + uint32_t *atom_engine_id); + void (*assign_control_parameter)( + const struct command_table_helper *h, + struct bp_encoder_control *control, + struct _DIG_ENCODER_CONTROL_PARAMETERS_V2 *ctrl_param); + bool (*clock_source_id_to_atom)(enum clock_source_id id, + uint32_t *atom_pll_id); + bool (*clock_source_id_to_ref_clk_src)( + enum clock_source_id id, + uint32_t *ref_clk_src_id); + uint8_t (*transmitter_bp_to_atom)(enum transmitter t); + uint8_t (*encoder_id_to_atom)(enum encoder_id id); + uint8_t (*clock_source_id_to_atom_phy_clk_src_id)( + enum clock_source_id id); + uint8_t (*signal_type_to_atom_dig_mode)(enum signal_type s); + uint8_t (*hpd_sel_to_atom)(enum hpd_source_id id); + uint8_t (*dig_encoder_sel_to_atom)(enum engine_id engine_id); + uint8_t (*phy_id_to_atom)(enum transmitter t); + uint8_t (*disp_power_gating_action_to_atom)( + enum bp_pipe_control_action action); + bool (*dc_clock_type_to_atom)(enum bp_dce_clock_type id, + uint32_t *atom_clock_type); + uint8_t (*transmitter_color_depth_to_atom)(enum transmitter_color_depth id); +}; + +#endif
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c index 6d38b8f..0cbab81 100644 --- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
@@ -85,7 +85,6 @@ static void calculate_bandwidth( const uint32_t s_mid5 = 5; const uint32_t s_mid6 = 6; const uint32_t s_high = 7; - const uint32_t bus_efficiency = 1; const uint32_t dmif_chunk_buff_margin = 1; uint32_t max_chunks_fbc_mode; @@ -592,7 +591,12 @@ static void calculate_bandwidth( /* 1 = use channel 0 and 1*/ /* 2 = use channel 0,1,2,3*/ if ((fbc_enabled == 1 && lpt_enabled == 1)) { - data->dram_efficiency = bw_int_to_fixed(1); + if (vbios->memory_type == bw_def_hbm) + data->dram_efficiency = bw_frc_to_fixed(5, 10); + else + data->dram_efficiency = bw_int_to_fixed(1); + + if (dceip->low_power_tiling_mode == 0) { data->number_of_dram_channels = 1; } @@ -607,7 +611,10 @@ static void calculate_bandwidth( } } else { - data->dram_efficiency = bw_frc_to_fixed(8, 10); + if (vbios->memory_type == bw_def_hbm) + data->dram_efficiency = bw_frc_to_fixed(5, 10); + else + data->dram_efficiency = bw_frc_to_fixed(8, 10); } /*memory request size and latency hiding:*/ /*request size is normally 64 byte, 2-line interleaved, with full latency hiding*/ @@ -1171,9 +1178,9 @@ static void calculate_bandwidth( } for (i = 0; i <= 2; i++) { for (j = 0; j <= 7; j++) { - data->dmif_burst_time[i][j] = bw_max3(data->dmif_total_page_close_open_time, bw_div(data->total_display_reads_required_dram_access_data, (bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[i]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels)))), bw_div(data->total_display_reads_required_data, (bw_mul(bw_mul(sclk[j], vbios->data_return_bus_width), bw_int_to_fixed(bus_efficiency))))); + data->dmif_burst_time[i][j] = bw_max3(data->dmif_total_page_close_open_time, bw_div(data->total_display_reads_required_dram_access_data, (bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[i]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels)))), bw_div(data->total_display_reads_required_data, (bw_mul(bw_mul(sclk[j], vbios->data_return_bus_width), bw_frc_to_fixed(dceip->percent_of_ideal_port_bw_received_after_urgent_latency, 100))))); if (data->d1_display_write_back_dwb_enable == 1) { - data->mcifwr_burst_time[i][j] = bw_max3(data->mcifwr_total_page_close_open_time, bw_div(data->total_display_writes_required_dram_access_data, (bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[i]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_wrchannels)))), bw_div(data->total_display_writes_required_data, (bw_mul(bw_mul(sclk[j], vbios->data_return_bus_width), bw_int_to_fixed(bus_efficiency))))); + data->mcifwr_burst_time[i][j] = bw_max3(data->mcifwr_total_page_close_open_time, bw_div(data->total_display_writes_required_dram_access_data, (bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[i]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_wrchannels)))), bw_div(data->total_display_writes_required_data, (bw_mul(sclk[j], vbios->data_return_bus_width)))); } } } @@ -1258,6 +1265,16 @@ static void calculate_bandwidth( /* / (dispclk - display bw)*/ /*the minimum latency hiding is the minimum for all pipes of one screen line time, plus one more line time if doing lb prefetch, plus the dmif data buffer size equivalent in time, minus the urgent latency.*/ /*the minimum latency hiding is further limited by the cursor. the cursor latency hiding is the number of lines of the cursor buffer, minus one if the downscaling is less than two, or minus three if it is more*/ + + /*initialize variables*/ + number_of_displays_enabled = 0; + number_of_displays_enabled_with_margin = 0; + for (k = 0; k <= maximum_number_of_surfaces - 1; k++) { + if (data->enable[k]) { + number_of_displays_enabled = number_of_displays_enabled + 1; + } + data->display_pstate_change_enable[k] = 0; + } for (i = 0; i <= maximum_number_of_surfaces - 1; i++) { if (data->enable[i]) { if ((bw_equ(dceip->stutter_and_dram_clock_state_change_gated_before_cursor, bw_int_to_fixed(0)) && bw_mtn(data->cursor_width_pixels[i], bw_int_to_fixed(0)))) { @@ -1276,7 +1293,10 @@ static void calculate_bandwidth( for (i = 0; i <= maximum_number_of_surfaces - 1; i++) { if (data->enable[i]) { if (dceip->graphics_lb_nodownscaling_multi_line_prefetching == 1 && (bw_equ(data->vsr[i], bw_int_to_fixed(1)) || (bw_leq(data->vsr[i], bw_frc_to_fixed(8, 10)) && bw_leq(data->v_taps[i], bw_int_to_fixed(2)) && data->lb_bpc[i] == 8)) && surface_type[i] == bw_def_graphics) { - data->minimum_latency_hiding[i] = bw_sub(bw_div(bw_mul((bw_div((bw_add(bw_sub(data->lb_partitions[i], bw_int_to_fixed(1)), bw_div(bw_div(data->data_buffer_size[i], bw_int_to_fixed(data->bytes_per_pixel[i])), data->source_width_pixels[i]))), data->vsr[i])), data->h_total[i]), data->pixel_rate[i]), data->total_dmifmc_urgent_latency); + if (number_of_displays_enabled > 2) + data->minimum_latency_hiding[i] = bw_sub(bw_div(bw_mul((bw_div((bw_add(bw_sub(data->lb_partitions[i], bw_int_to_fixed(2)), bw_div(bw_div(data->data_buffer_size[i], bw_int_to_fixed(data->bytes_per_pixel[i])), data->source_width_pixels[i]))), data->vsr[i])), data->h_total[i]), data->pixel_rate[i]), data->total_dmifmc_urgent_latency); + else + data->minimum_latency_hiding[i] = bw_sub(bw_div(bw_mul((bw_div((bw_add(bw_sub(data->lb_partitions[i], bw_int_to_fixed(1)), bw_div(bw_div(data->data_buffer_size[i], bw_int_to_fixed(data->bytes_per_pixel[i])), data->source_width_pixels[i]))), data->vsr[i])), data->h_total[i]), data->pixel_rate[i]), data->total_dmifmc_urgent_latency); } else { data->minimum_latency_hiding[i] = bw_sub(bw_div(bw_mul((bw_div((bw_add(bw_int_to_fixed(1 + data->line_buffer_prefetch[i]), bw_div(bw_div(data->data_buffer_size[i], bw_int_to_fixed(data->bytes_per_pixel[i])), data->source_width_pixels[i]))), data->vsr[i])), data->h_total[i]), data->pixel_rate[i]), data->total_dmifmc_urgent_latency); @@ -1338,24 +1358,15 @@ static void calculate_bandwidth( for (i = 0; i <= maximum_number_of_surfaces - 1; i++) { if (data->enable[i]) { if (dceip->graphics_lb_nodownscaling_multi_line_prefetching == 1) { - data->maximum_latency_hiding[i] = bw_add(data->minimum_latency_hiding[i], bw_mul(bw_frc_to_fixed(8, 10), data->total_dmifmc_urgent_latency)); + data->maximum_latency_hiding[i] = bw_add(data->minimum_latency_hiding[i], bw_mul(bw_frc_to_fixed(5, 10), data->total_dmifmc_urgent_latency)); } else { /*maximum_latency_hiding(i) = minimum_latency_hiding(i) + 1 / vsr(i) * h_total(i) / pixel_rate(i) + 0.5 * total_dmifmc_urgent_latency*/ - data->maximum_latency_hiding[i] = bw_add(data->minimum_latency_hiding[i], bw_mul(bw_frc_to_fixed(8, 10), data->total_dmifmc_urgent_latency)); + data->maximum_latency_hiding[i] = bw_add(data->minimum_latency_hiding[i], bw_mul(bw_frc_to_fixed(5, 10), data->total_dmifmc_urgent_latency)); } data->maximum_latency_hiding_with_cursor[i] = bw_min2(data->maximum_latency_hiding[i], data->cursor_latency_hiding[i]); } } - /*initialize variables*/ - number_of_displays_enabled = 0; - number_of_displays_enabled_with_margin = 0; - for (k = 0; k <= maximum_number_of_surfaces - 1; k++) { - if (data->enable[k]) { - number_of_displays_enabled = number_of_displays_enabled + 1; - } - data->display_pstate_change_enable[k] = 0; - } for (i = 0; i <= 2; i++) { for (j = 0; j <= 7; j++) { data->min_dram_speed_change_margin[i][j] = bw_int_to_fixed(9999); @@ -1370,10 +1381,11 @@ static void calculate_bandwidth( /*determine the minimum dram clock change margin for each set of clock frequencies*/ data->min_dram_speed_change_margin[i][j] = bw_min2(data->min_dram_speed_change_margin[i][j], data->dram_speed_change_margin); /*compute the maximum clock frequuency required for the dram clock change at each set of clock frequencies*/ - data->dispclk_required_for_dram_speed_change[i][j] = bw_max3(data->dispclk_required_for_dram_speed_change[i][j], bw_div(bw_div(bw_mul(data->src_pixels_for_first_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]))), bw_div(bw_div(bw_mul(data->src_pixels_for_last_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_add(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->active_time[k])))); - if ((bw_ltn(data->dispclk_required_for_dram_speed_change[i][j], vbios->high_voltage_max_dispclk))) { + data->dispclk_required_for_dram_speed_change_pipe[i][j] = bw_max2(bw_div(bw_div(bw_mul(data->src_pixels_for_first_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]))), bw_div(bw_div(bw_mul(data->src_pixels_for_last_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_add(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->active_time[k])))); + if ((bw_ltn(data->dispclk_required_for_dram_speed_change_pipe[i][j], vbios->high_voltage_max_dispclk))) { data->display_pstate_change_enable[k] = 1; data->num_displays_with_margin[i][j] = data->num_displays_with_margin[i][j] + 1; + data->dispclk_required_for_dram_speed_change[i][j] = bw_max2(data->dispclk_required_for_dram_speed_change[i][j], data->dispclk_required_for_dram_speed_change_pipe[i][j]); } } } @@ -1383,10 +1395,11 @@ static void calculate_bandwidth( /*determine the minimum dram clock change margin for each display pipe*/ data->min_dram_speed_change_margin[i][j] = bw_min2(data->min_dram_speed_change_margin[i][j], data->dram_speed_change_margin); /*compute the maximum clock frequuency required for the dram clock change at each set of clock frequencies*/ - data->dispclk_required_for_dram_speed_change[i][j] = bw_max3(data->dispclk_required_for_dram_speed_change[i][j], bw_div(bw_div(bw_mul(data->src_pixels_for_first_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_sub(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->mcifwr_burst_time[i][j]))), bw_div(bw_div(bw_mul(data->src_pixels_for_last_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_add(bw_sub(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->mcifwr_burst_time[i][j]), data->active_time[k])))); - if ((bw_ltn(data->dispclk_required_for_dram_speed_change[i][j], vbios->high_voltage_max_dispclk))) { + data->dispclk_required_for_dram_speed_change_pipe[i][j] = bw_max2(bw_div(bw_div(bw_mul(data->src_pixels_for_first_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_sub(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->mcifwr_burst_time[i][j]))), bw_div(bw_div(bw_mul(data->src_pixels_for_last_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_add(bw_sub(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->mcifwr_burst_time[i][j]), data->active_time[k])))); + if ((bw_ltn(data->dispclk_required_for_dram_speed_change_pipe[i][j], vbios->high_voltage_max_dispclk))) { data->display_pstate_change_enable[k] = 1; data->num_displays_with_margin[i][j] = data->num_displays_with_margin[i][j] + 1; + data->dispclk_required_for_dram_speed_change[i][j] = bw_max2(data->dispclk_required_for_dram_speed_change[i][j], data->dispclk_required_for_dram_speed_change_pipe[i][j]); } } } @@ -1420,7 +1433,7 @@ static void calculate_bandwidth( data->displays_with_same_mode[i] = bw_int_to_fixed(0); if (data->enable[i] == 1 && data->display_pstate_change_enable[i] == 0 && bw_mtn(data->v_blank_dram_speed_change_margin[i], bw_int_to_fixed(0))) { for (j = 0; j <= maximum_number_of_surfaces - 1; j++) { - if ((data->enable[j] == 1 && bw_equ(data->source_width_rounded_up_to_chunks[i], data->source_width_rounded_up_to_chunks[j]) && bw_equ(data->source_height_rounded_up_to_chunks[i], data->source_height_rounded_up_to_chunks[j]) && bw_equ(data->vsr[i], data->vsr[j]) && bw_equ(data->hsr[i], data->hsr[j]) && bw_equ(data->pixel_rate[i], data->pixel_rate[j]))) { + if ((i == j || data->display_synchronization_enabled) && (data->enable[j] == 1 && bw_equ(data->source_width_rounded_up_to_chunks[i], data->source_width_rounded_up_to_chunks[j]) && bw_equ(data->source_height_rounded_up_to_chunks[i], data->source_height_rounded_up_to_chunks[j]) && bw_equ(data->vsr[i], data->vsr[j]) && bw_equ(data->hsr[i], data->hsr[j]) && bw_equ(data->pixel_rate[i], data->pixel_rate[j]))) { data->displays_with_same_mode[i] = bw_add(data->displays_with_same_mode[i], bw_int_to_fixed(1)); } } @@ -1435,7 +1448,7 @@ static void calculate_bandwidth( /*aligned displays with the same timing.*/ /*the display(s) with the negative margin can be switched in the v_blank region while the other*/ /*displays are in v_blank or v_active.*/ - if ((number_of_displays_enabled_with_margin + number_of_aligned_displays_with_no_margin == number_of_displays_enabled && bw_mtn(data->min_dram_speed_change_margin[high][s_high], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[high][s_high], bw_int_to_fixed(9999)) && bw_ltn(data->dispclk_required_for_dram_speed_change[high][s_high], vbios->high_voltage_max_dispclk))) { + if (number_of_displays_enabled_with_margin > 0 && (number_of_displays_enabled_with_margin + number_of_aligned_displays_with_no_margin) == number_of_displays_enabled && bw_mtn(data->min_dram_speed_change_margin[high][s_high], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[high][s_high], bw_int_to_fixed(9999)) && bw_ltn(data->dispclk_required_for_dram_speed_change[high][s_high], vbios->high_voltage_max_dispclk)) { data->nbp_state_change_enable = bw_def_yes; } else { @@ -1448,6 +1461,25 @@ static void calculate_bandwidth( else { nbp_state_change_enable_blank = bw_def_no; } + + /*average bandwidth*/ + /*the average bandwidth with no compression is the vertical active time is the source width times the bytes per pixel divided by the line time, multiplied by the vertical scale ratio and the ratio of bytes per request divided by the useful bytes per request.*/ + /*the average bandwidth with compression is the same, divided by the compression ratio*/ + for (i = 0; i <= maximum_number_of_surfaces - 1; i++) { + if (data->enable[i]) { + data->average_bandwidth_no_compression[i] = bw_div(bw_mul(bw_mul(bw_div(bw_mul(data->source_width_rounded_up_to_chunks[i], bw_int_to_fixed(data->bytes_per_pixel[i])), (bw_div(data->h_total[i], data->pixel_rate[i]))), data->vsr[i]), data->bytes_per_request[i]), data->useful_bytes_per_request[i]); + data->average_bandwidth[i] = bw_div(data->average_bandwidth_no_compression[i], data->compression_rate[i]); + } + } + data->total_average_bandwidth_no_compression = bw_int_to_fixed(0); + data->total_average_bandwidth = bw_int_to_fixed(0); + for (i = 0; i <= maximum_number_of_surfaces - 1; i++) { + if (data->enable[i]) { + data->total_average_bandwidth_no_compression = bw_add(data->total_average_bandwidth_no_compression, data->average_bandwidth_no_compression[i]); + data->total_average_bandwidth = bw_add(data->total_average_bandwidth, data->average_bandwidth[i]); + } + } + /*required yclk(pclk)*/ /*yclk requirement only makes sense if the dmif and mcifwr data total page close-open time is less than the time for data transfer and the total pte requests fit in the scatter-gather saw queque size*/ /*if that is the case, the yclk requirement is the maximum of the ones required by dmif and mcifwr, and the high/low yclk(pclk) is chosen accordingly*/ @@ -1497,17 +1529,20 @@ static void calculate_bandwidth( } else { data->required_dram_bandwidth_gbyte_per_second = bw_div(bw_max2(data->dmif_required_dram_bandwidth, data->mcifwr_required_dram_bandwidth), bw_int_to_fixed(1000)); - if (bw_ltn(bw_mul(data->required_dram_bandwidth_gbyte_per_second, bw_int_to_fixed(1000)), bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[low]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels))) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[low][s_high], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[low][s_high], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[low][s_high], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[low][s_high], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[low][s_high], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[low][s_high], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[low][s_high], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[low][s_high], vbios->high_voltage_max_dispclk) && data->num_displays_with_margin[low][s_high] == number_of_displays_enabled_with_margin))) { + if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation, 100),yclk[low]),bw_div(bw_int_to_fixed(vbios->dram_channel_width_in_bits),bw_int_to_fixed(8))),bw_int_to_fixed(vbios->number_of_dram_channels))) + && bw_ltn(bw_mul(data->required_dram_bandwidth_gbyte_per_second, bw_int_to_fixed(1000)), bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[low]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels))) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[low][s_high], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[low][s_high], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[low][s_high], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[low][s_high], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[low][s_high], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[low][s_high], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[low][s_high], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[low][s_high], vbios->high_voltage_max_dispclk) && data->num_displays_with_margin[low][s_high] == number_of_displays_enabled_with_margin))) { yclk_message = bw_fixed_to_int(vbios->low_yclk); data->y_clk_level = low; data->dram_bandwidth = bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[low]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels)); } - else if (bw_ltn(bw_mul(data->required_dram_bandwidth_gbyte_per_second, bw_int_to_fixed(1000)), bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[mid]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels))) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[mid][s_high], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[mid][s_high], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[mid][s_high], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[mid][s_high], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[mid][s_high], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[mid][s_high], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[mid][s_high], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[mid][s_high], vbios->high_voltage_max_dispclk) && data->num_displays_with_margin[mid][s_high] == number_of_displays_enabled_with_margin))) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation, 100),yclk[mid]),bw_div(bw_int_to_fixed(vbios->dram_channel_width_in_bits),bw_int_to_fixed(8))),bw_int_to_fixed(vbios->number_of_dram_channels))) + && bw_ltn(bw_mul(data->required_dram_bandwidth_gbyte_per_second, bw_int_to_fixed(1000)), bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[mid]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels))) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[mid][s_high], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[mid][s_high], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[mid][s_high], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[mid][s_high], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[mid][s_high], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[mid][s_high], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[mid][s_high], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[mid][s_high], vbios->high_voltage_max_dispclk) && data->num_displays_with_margin[mid][s_high] == number_of_displays_enabled_with_margin))) { yclk_message = bw_fixed_to_int(vbios->mid_yclk); data->y_clk_level = mid; data->dram_bandwidth = bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[mid]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels)); } - else if (bw_ltn(bw_mul(data->required_dram_bandwidth_gbyte_per_second, bw_int_to_fixed(1000)), bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[high]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels)))) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation, 100),yclk[high]),bw_div(bw_int_to_fixed(vbios->dram_channel_width_in_bits),bw_int_to_fixed(8))),bw_int_to_fixed(vbios->number_of_dram_channels))) + && bw_ltn(bw_mul(data->required_dram_bandwidth_gbyte_per_second, bw_int_to_fixed(1000)), bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[high]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels)))) { yclk_message = bw_fixed_to_int(vbios->high_yclk); data->y_clk_level = high; data->dram_bandwidth = bw_mul(bw_div(bw_mul(bw_mul(data->dram_efficiency, yclk[high]), bw_int_to_fixed(vbios->dram_channel_width_in_bits)), bw_int_to_fixed(8)), bw_int_to_fixed(data->number_of_dram_channels)); @@ -1523,8 +1558,8 @@ static void calculate_bandwidth( /*if that is the case, the sclk requirement is the maximum of the ones required by dmif and mcifwr, and the high/mid/low sclk is chosen accordingly, unless that choice results in foresaking dram speed/nb p-state change.*/ /*the dmif and mcifwr sclk required is the one that allows the transfer of all pipe's data buffer size through the sclk bus in the time for data transfer*/ /*for dmif, pte and cursor requests have to be included.*/ - data->dmif_required_sclk = bw_div(bw_div(data->total_display_reads_required_data, data->display_reads_time_for_data_transfer), (bw_mul(vbios->data_return_bus_width, bw_int_to_fixed(bus_efficiency)))); - data->mcifwr_required_sclk = bw_div(bw_div(data->total_display_writes_required_data, data->display_writes_time_for_data_transfer), (bw_mul(vbios->data_return_bus_width, bw_int_to_fixed(bus_efficiency)))); + data->dmif_required_sclk = bw_div(bw_div(data->total_display_reads_required_data, data->display_reads_time_for_data_transfer), (bw_mul(vbios->data_return_bus_width, bw_frc_to_fixed(dceip->percent_of_ideal_port_bw_received_after_urgent_latency, 100)))); + data->mcifwr_required_sclk = bw_div(bw_div(data->total_display_writes_required_data, data->display_writes_time_for_data_transfer), vbios->data_return_bus_width); if (bw_mtn(data->scatter_gather_total_pte_requests, dceip->maximum_total_outstanding_pte_requests_allowed_by_saw)) { data->required_sclk = bw_int_to_fixed(9999); sclk_message = bw_def_exceeded_allowed_outstanding_pte_req_queue_size; @@ -1537,42 +1572,56 @@ static void calculate_bandwidth( } else { data->required_sclk = bw_max2(data->dmif_required_sclk, data->mcifwr_required_sclk); - if (bw_ltn(data->required_sclk, sclk[s_low]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_low], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_low], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_low], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_low], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_low], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_low], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_low], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_low], vbios->low_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_low] == number_of_displays_enabled_with_margin))) { + if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[low]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_low]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_low], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_low], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_low], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_low], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_low], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_low], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_low], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_low], vbios->low_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_low] == number_of_displays_enabled_with_margin))) { sclk_message = bw_def_low; data->sclk_level = s_low; data->required_sclk = vbios->low_sclk; } - else if (bw_ltn(data->required_sclk, sclk[s_mid1]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid1], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid1], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid1], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid1], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid1], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid1], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid1], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid1], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid1] == number_of_displays_enabled_with_margin))) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[mid]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_mid1]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid1], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid1], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid1], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid1], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid1], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid1], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid1], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid1], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid1] == number_of_displays_enabled_with_margin))) { sclk_message = bw_def_mid; data->sclk_level = s_mid1; data->required_sclk = vbios->mid1_sclk; } - else if (bw_ltn(data->required_sclk, sclk[s_mid2]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid2], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid2], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid2], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid2], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid2], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid2], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid2], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid2], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid2] == number_of_displays_enabled_with_margin))) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[s_mid2]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_mid2]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid2], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid2], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid2], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid2], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid2], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid2], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid2], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid2], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid2] == number_of_displays_enabled_with_margin))) { sclk_message = bw_def_mid; data->sclk_level = s_mid2; data->required_sclk = vbios->mid2_sclk; } - else if (bw_ltn(data->required_sclk, sclk[s_mid3]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid3], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid3], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid3], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid3], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid3], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid3], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid3], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid3], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid3] == number_of_displays_enabled_with_margin))) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[s_mid3]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_mid3]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid3], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid3], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid3], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid3], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid3], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid3], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid3], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid3], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid3] == number_of_displays_enabled_with_margin))) { sclk_message = bw_def_mid; data->sclk_level = s_mid3; data->required_sclk = vbios->mid3_sclk; } - else if (bw_ltn(data->required_sclk, sclk[s_mid4]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid4], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid4], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid4], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid4], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid4], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid4], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid4], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid4], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid4] == number_of_displays_enabled_with_margin))) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[s_mid4]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_mid4]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid4], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid4], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid4], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid4], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid4], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid4], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid4], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid4], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid4] == number_of_displays_enabled_with_margin))) { sclk_message = bw_def_mid; data->sclk_level = s_mid4; data->required_sclk = vbios->mid4_sclk; } - else if (bw_ltn(data->required_sclk, sclk[s_mid5]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid5], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid5], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid5], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid5], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid5], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid5], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid5], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid5], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid5] == number_of_displays_enabled_with_margin))) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[s_mid5]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_mid5]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid5], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid5], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid5], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid5], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid5], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid5], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid5], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid5], vbios->mid_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid5] == number_of_displays_enabled_with_margin))) { sclk_message = bw_def_mid; data->sclk_level = s_mid5; data->required_sclk = vbios->mid5_sclk; } - else if (bw_ltn(data->required_sclk, sclk[s_mid6]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid6], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid6], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid6], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid6], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid6], vbios->high_voltage_max_dispclk))) && (data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid6], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid6], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid6], vbios->high_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid6] == number_of_displays_enabled_with_margin))) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[s_mid6]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_mid6]) && (data->cpup_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid6], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid6], vbios->high_voltage_max_dispclk))) && (data->cpuc_state_change_enable == bw_def_no || (bw_mtn(data->blackout_duration_margin[data->y_clk_level][s_mid6], bw_int_to_fixed(0)) && bw_ltn(data->dispclk_required_for_blackout_duration[data->y_clk_level][s_mid6], vbios->high_voltage_max_dispclk) && bw_ltn(data->dispclk_required_for_blackout_recovery[data->y_clk_level][s_mid6], vbios->high_voltage_max_dispclk))) && (!data->increase_voltage_to_support_mclk_switch || data->nbp_state_change_enable == bw_def_no || (bw_mtn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid6], bw_int_to_fixed(0)) && bw_ltn(data->min_dram_speed_change_margin[data->y_clk_level][s_mid6], bw_int_to_fixed(9999)) && bw_leq(data->dispclk_required_for_dram_speed_change[data->y_clk_level][s_mid6], vbios->high_voltage_max_dispclk) && data->num_displays_with_margin[data->y_clk_level][s_mid6] == number_of_displays_enabled_with_margin))) { sclk_message = bw_def_mid; data->sclk_level = s_mid6; data->required_sclk = vbios->mid6_sclk; } - else if (bw_ltn(data->required_sclk, sclk[s_high])) { + else if (bw_ltn(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[s_high]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_high])) { + sclk_message = bw_def_high; + data->sclk_level = s_high; + data->required_sclk = vbios->high_sclk; + } + else if (bw_meq(data->total_average_bandwidth_no_compression, bw_mul(bw_mul(bw_frc_to_fixed(dceip->max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation, 100),sclk[s_high]),vbios->data_return_bus_width)) + && bw_ltn(data->required_sclk, sclk[s_high])) { sclk_message = bw_def_high; data->sclk_level = s_high; data->required_sclk = vbios->high_sclk; @@ -1681,7 +1730,7 @@ static void calculate_bandwidth( data->total_dispclk_required_with_ramping_with_request_bandwidth = bw_max2(data->total_dispclk_required_with_ramping_with_request_bandwidth, data->dispclk_required_for_blackout_duration[data->y_clk_level][data->sclk_level]); data->total_dispclk_required_without_ramping_with_request_bandwidth = bw_max2(data->total_dispclk_required_without_ramping_with_request_bandwidth, data->dispclk_required_for_blackout_duration[data->y_clk_level][data->sclk_level]); } - if (data->nbp_state_change_enable == bw_def_yes) { + if (data->nbp_state_change_enable == bw_def_yes && data->increase_voltage_to_support_mclk_switch) { data->total_dispclk_required_with_ramping_with_request_bandwidth = bw_max2(data->total_dispclk_required_with_ramping_with_request_bandwidth, data->dispclk_required_for_dram_speed_change[data->y_clk_level][data->sclk_level]); data->total_dispclk_required_without_ramping_with_request_bandwidth = bw_max2(data->total_dispclk_required_without_ramping_with_request_bandwidth, data->dispclk_required_for_dram_speed_change[data->y_clk_level][data->sclk_level]); } @@ -1861,23 +1910,6 @@ static void calculate_bandwidth( else { data->mcifwrdram_access_efficiency = bw_int_to_fixed(0); } - /*average bandwidth*/ - /*the average bandwidth with no compression is the vertical active time is the source width times the bytes per pixel divided by the line time, multiplied by the vertical scale ratio and the ratio of bytes per request divided by the useful bytes per request.*/ - /*the average bandwidth with compression is the same, divided by the compression ratio*/ - for (i = 0; i <= maximum_number_of_surfaces - 1; i++) { - if (data->enable[i]) { - data->average_bandwidth_no_compression[i] = bw_div(bw_mul(bw_mul(bw_div(bw_mul(data->source_width_rounded_up_to_chunks[i], bw_int_to_fixed(data->bytes_per_pixel[i])), (bw_div(data->h_total[i], data->pixel_rate[i]))), data->vsr[i]), data->bytes_per_request[i]), data->useful_bytes_per_request[i]); - data->average_bandwidth[i] = bw_div(data->average_bandwidth_no_compression[i], data->compression_rate[i]); - } - } - data->total_average_bandwidth_no_compression = bw_int_to_fixed(0); - data->total_average_bandwidth = bw_int_to_fixed(0); - for (i = 0; i <= maximum_number_of_surfaces - 1; i++) { - if (data->enable[i]) { - data->total_average_bandwidth_no_compression = bw_add(data->total_average_bandwidth_no_compression, data->average_bandwidth_no_compression[i]); - data->total_average_bandwidth = bw_add(data->total_average_bandwidth, data->average_bandwidth[i]); - } - } /*stutter efficiency*/ /*the stutter efficiency is the frame-average time in self-refresh divided by the frame-average stutter cycle duration. only applies if the display write-back is not enabled.*/ /*the frame-average stutter cycle used is the minimum for all pipes of the frame-average data buffer size in time, times the compression rate*/ @@ -1905,7 +1937,7 @@ static void calculate_bandwidth( data->total_stutter_dmif_buffer_size = bw_fixed_to_int(bw_add(data->stutter_dmif_buffer_size[i], bw_int_to_fixed(data->total_stutter_dmif_buffer_size))); } } - data->stutter_burst_time = bw_div(bw_int_to_fixed(data->total_stutter_dmif_buffer_size), bw_min2(bw_mul(data->dram_bandwidth, data->dmifdram_access_efficiency), bw_mul(sclk[data->sclk_level], bw_int_to_fixed(32)))); + data->stutter_burst_time = bw_div(bw_int_to_fixed(data->total_stutter_dmif_buffer_size), bw_mul(sclk[data->sclk_level], vbios->data_return_bus_width)); data->num_stutter_bursts = data->total_bytes_requested / data->min_stutter_dmif_buffer_size; data->total_stutter_cycle_duration = bw_add(bw_add(data->min_stutter_refresh_duration, vbios->stutter_self_refresh_exit_latency), data->stutter_burst_time); data->time_in_self_refresh = data->min_stutter_refresh_duration; @@ -1957,7 +1989,7 @@ static void calculate_bandwidth( for (i = 1; i <= 5; i++) { data->display_reads_time_for_data_transfer_and_urgent_latency = bw_sub(data->min_read_buffer_size_in_time, bw_mul(data->total_dmifmc_urgent_trips, bw_int_to_fixed(i))); if (pipe_check == bw_def_ok && (bw_mtn(data->display_reads_time_for_data_transfer_and_urgent_latency, data->dmif_total_page_close_open_time))) { - data->dmif_required_sclk_for_urgent_latency[i] = bw_div(bw_div(data->total_display_reads_required_data, data->display_reads_time_for_data_transfer_and_urgent_latency), (bw_mul(vbios->data_return_bus_width, bw_int_to_fixed(bus_efficiency)))); + data->dmif_required_sclk_for_urgent_latency[i] = bw_div(bw_div(data->total_display_reads_required_data, data->display_reads_time_for_data_transfer_and_urgent_latency), (bw_mul(vbios->data_return_bus_width, bw_frc_to_fixed(dceip->percent_of_ideal_port_bw_received_after_urgent_latency, 100)))); } else { data->dmif_required_sclk_for_urgent_latency[i] = bw_int_to_fixed(bw_def_na); @@ -2036,6 +2068,9 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip, vbios.blackout_duration = bw_int_to_fixed(0); /* us */ vbios.maximum_blackout_recovery_time = bw_int_to_fixed(0); + dceip.max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation = 100; + dceip.max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation = 100; + dceip.percent_of_ideal_port_bw_received_after_urgent_latency = 100; dceip.large_cursor = false; dceip.dmif_request_buffer_size = bw_int_to_fixed(768); dceip.dmif_pipe_en_fbc_chunk_tracker = false; @@ -2146,6 +2181,9 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip, vbios.blackout_duration = bw_int_to_fixed(0); /* us */ vbios.maximum_blackout_recovery_time = bw_int_to_fixed(0); + dceip.max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation = 100; + dceip.max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation = 100; + dceip.percent_of_ideal_port_bw_received_after_urgent_latency = 100; dceip.large_cursor = false; dceip.dmif_request_buffer_size = bw_int_to_fixed(768); dceip.dmif_pipe_en_fbc_chunk_tracker = false; @@ -2259,6 +2297,9 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip, vbios.blackout_duration = bw_int_to_fixed(0); /* us */ vbios.maximum_blackout_recovery_time = bw_int_to_fixed(0); + dceip.max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation = 100; + dceip.max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation = 100; + dceip.percent_of_ideal_port_bw_received_after_urgent_latency = 100; dceip.large_cursor = false; dceip.dmif_request_buffer_size = bw_int_to_fixed(768); dceip.dmif_pipe_en_fbc_chunk_tracker = false; @@ -2369,6 +2410,9 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip, vbios.blackout_duration = bw_int_to_fixed(0); /* us */ vbios.maximum_blackout_recovery_time = bw_int_to_fixed(0); + dceip.max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation = 100; + dceip.max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation = 100; + dceip.percent_of_ideal_port_bw_received_after_urgent_latency = 100; dceip.large_cursor = false; dceip.dmif_request_buffer_size = bw_int_to_fixed(768); dceip.dmif_pipe_en_fbc_chunk_tracker = false; @@ -2479,6 +2523,9 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip, vbios.blackout_duration = bw_int_to_fixed(0); /* us */ vbios.maximum_blackout_recovery_time = bw_int_to_fixed(0); + dceip.max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation = 100; + dceip.max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation = 100; + dceip.percent_of_ideal_port_bw_received_after_urgent_latency = 100; dceip.large_cursor = false; dceip.dmif_request_buffer_size = bw_int_to_fixed(2304); dceip.dmif_pipe_en_fbc_chunk_tracker = true; @@ -2597,6 +2644,7 @@ static void populate_initial_data( data->graphics_tiling_mode = bw_def_tiled; data->underlay_micro_tile_mode = bw_def_display_micro_tiling; data->graphics_micro_tile_mode = bw_def_display_micro_tiling; + data->increase_voltage_to_support_mclk_switch = true; /* Pipes with underlay first */ for (i = 0; i < pipe_count; i++) {
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c index 331891c..4bb43a3 100644 --- a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
@@ -33,6 +33,8 @@ #include "dcn10/dcn10_resource.h" #include "dcn_calc_math.h" +#define DC_LOGGER \ + dc->ctx->logger /* * NOTE: * This file is gcc-parseable HW gospel, coming straight from HW engineers. @@ -486,6 +488,7 @@ static void split_stream_across_pipes( secondary_pipe->plane_res.ipp = pool->ipps[secondary_pipe->pipe_idx]; secondary_pipe->plane_res.xfm = pool->transforms[secondary_pipe->pipe_idx]; secondary_pipe->plane_res.dpp = pool->dpps[secondary_pipe->pipe_idx]; + secondary_pipe->plane_res.mpcc_inst = pool->dpps[secondary_pipe->pipe_idx]->inst; if (primary_pipe->bottom_pipe) { ASSERT(primary_pipe->bottom_pipe != secondary_pipe); secondary_pipe->bottom_pipe = primary_pipe->bottom_pipe; @@ -625,7 +628,7 @@ static bool dcn_bw_apply_registry_override(struct dc *dc) return updated; } -void hack_disable_optional_pipe_split(struct dcn_bw_internal_vars *v) +static void hack_disable_optional_pipe_split(struct dcn_bw_internal_vars *v) { /* * disable optional pipe split by lower dispclk bounding box @@ -634,7 +637,7 @@ void hack_disable_optional_pipe_split(struct dcn_bw_internal_vars *v) v->max_dispclk[0] = v->max_dppclk_vmin0p65; } -void hack_force_pipe_split(struct dcn_bw_internal_vars *v, +static void hack_force_pipe_split(struct dcn_bw_internal_vars *v, unsigned int pixel_rate_khz) { float pixel_rate_mhz = pixel_rate_khz / 1000; @@ -647,25 +650,20 @@ void hack_force_pipe_split(struct dcn_bw_internal_vars *v, v->max_dppclk[0] = pixel_rate_mhz; } -void hack_bounding_box(struct dcn_bw_internal_vars *v, +static void hack_bounding_box(struct dcn_bw_internal_vars *v, struct dc_debug *dbg, struct dc_state *context) { - if (dbg->pipe_split_policy == MPC_SPLIT_AVOID) { + if (dbg->pipe_split_policy == MPC_SPLIT_AVOID) hack_disable_optional_pipe_split(v); - } if (dbg->pipe_split_policy == MPC_SPLIT_AVOID_MULT_DISP && - context->stream_count >= 2) { + context->stream_count >= 2) hack_disable_optional_pipe_split(v); - } if (context->stream_count == 1 && - dbg->force_single_disp_pipe_split) { - struct dc_stream_state *stream0 = context->streams[0]; - - hack_force_pipe_split(v, stream0->timing.pix_clk_khz); - } + dbg->force_single_disp_pipe_split) + hack_force_pipe_split(v, context->streams[0]->timing.pix_clk_khz); } bool dcn_validate_bandwidth( @@ -799,23 +797,10 @@ bool dcn_validate_bandwidth( v->phyclk_per_state[2] = v->phyclkv_nom0p8; v->phyclk_per_state[1] = v->phyclkv_mid0p72; v->phyclk_per_state[0] = v->phyclkv_min0p65; - - hack_bounding_box(v, &dc->debug, context); - - if (v->voltage_override == dcn_bw_v_max0p9) { - v->voltage_override_level = number_of_states - 1; - } else if (v->voltage_override == dcn_bw_v_nom0p8) { - v->voltage_override_level = number_of_states - 2; - } else if (v->voltage_override == dcn_bw_v_mid0p72) { - v->voltage_override_level = number_of_states - 3; - } else { - v->voltage_override_level = 0; - } v->synchronized_vblank = dcn_bw_no; v->ta_pscalculation = dcn_bw_override; v->allow_different_hratio_vratio = dcn_bw_yes; - for (i = 0, input_idx = 0; i < pool->pipe_count; i++) { struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; @@ -948,8 +933,19 @@ bool dcn_validate_bandwidth( v->number_of_active_planes = input_idx; scaler_settings_calculation(v); + + hack_bounding_box(v, &dc->debug, context); + mode_support_and_system_configuration(v); + /* Unhack dppclk: dont bother with trying to pipe split if we cannot maintain dpm0 */ + if (v->voltage_level != 0 + && context->stream_count == 1 + && dc->debug.force_single_disp_pipe_split) { + v->max_dppclk[0] = v->max_dppclk_vmin0p65; + mode_support_and_system_configuration(v); + } + if (v->voltage_level == 0 && (dc->debug.sr_exit_time_dpm0_ns || dc->debug.sr_enter_plus_exit_time_dpm0_ns)) { @@ -987,8 +983,6 @@ bool dcn_validate_bandwidth( context->bw.dcn.calc_clk.fclk_khz = (int)(bw_consumed * 1000000 / 32); } - context->bw.dcn.calc_clk.dram_ccm_us = (int)(v->dram_clock_change_margin); - context->bw.dcn.calc_clk.min_active_dram_ccm_us = (int)(v->min_active_dram_clock_change_margin); context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz = (int)(v->dcf_clk_deep_sleep * 1000); context->bw.dcn.calc_clk.dcfclk_khz = (int)(v->dcfclk * 1000); @@ -1002,7 +996,26 @@ bool dcn_validate_bandwidth( dc->debug.min_disp_clk_khz; } - context->bw.dcn.calc_clk.dppclk_div = (int)(v->dispclk_dppclk_ratio) == 2; + context->bw.dcn.calc_clk.dppclk_khz = context->bw.dcn.calc_clk.dispclk_khz / v->dispclk_dppclk_ratio; + + switch (v->voltage_level) { + case 0: + context->bw.dcn.calc_clk.max_supported_dppclk_khz = + (int)(dc->dcn_soc->max_dppclk_vmin0p65 * 1000); + break; + case 1: + context->bw.dcn.calc_clk.max_supported_dppclk_khz = + (int)(dc->dcn_soc->max_dppclk_vmid0p72 * 1000); + break; + case 2: + context->bw.dcn.calc_clk.max_supported_dppclk_khz = + (int)(dc->dcn_soc->max_dppclk_vnom0p8 * 1000); + break; + default: + context->bw.dcn.calc_clk.max_supported_dppclk_khz = + (int)(dc->dcn_soc->max_dppclk_vmax0p9 * 1000); + break; + } for (i = 0, input_idx = 0; i < pool->pipe_count; i++) { struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; @@ -1248,8 +1261,7 @@ unsigned int dcn_find_dcfclk_suits_all( else dcf_clk = dc->dcn_soc->dcfclkv_min0p65*1000; - dm_logger_write(dc->ctx->logger, LOG_BANDWIDTH_CALCS, - "\tdcf_clk for voltage = %d\n", dcf_clk); + DC_LOG_BANDWIDTH_CALCS("\tdcf_clk for voltage = %d\n", dcf_clk); return dcf_clk; } @@ -1447,8 +1459,7 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) void dcn_bw_sync_calcs_and_dml(struct dc *dc) { kernel_fpu_begin(); - dm_logger_write(dc->ctx->logger, LOG_BANDWIDTH_CALCS, - "sr_exit_time: %d ns\n" + DC_LOG_BANDWIDTH_CALCS("sr_exit_time: %d ns\n" "sr_enter_plus_exit_time: %d ns\n" "urgent_latency: %d ns\n" "write_back_latency: %d ns\n" @@ -1516,8 +1527,7 @@ void dcn_bw_sync_calcs_and_dml(struct dc *dc) dc->dcn_soc->vmm_page_size, dc->dcn_soc->dram_clock_change_latency * 1000, dc->dcn_soc->return_bus_width); - dm_logger_write(dc->ctx->logger, LOG_BANDWIDTH_CALCS, - "rob_buffer_size_in_kbyte: %d\n" + DC_LOG_BANDWIDTH_CALCS("rob_buffer_size_in_kbyte: %d\n" "det_buffer_size_in_kbyte: %d\n" "dpp_output_buffer_pixels: %d\n" "opp_output_buffer_lines: %d\n"
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 011c193..9c450d2 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -39,8 +39,10 @@ #include "bios_parser_interface.h" #include "include/irq_service_interface.h" #include "transform.h" +#include "dmcu.h" #include "dpp.h" #include "timing_generator.h" +#include "abm.h" #include "virtual/virtual_link_encoder.h" #include "link_hwss.h" @@ -50,6 +52,8 @@ #include "dm_helpers.h" #include "mem_input.h" #include "hubp.h" +#define DC_LOGGER \ + dc->ctx->logger /******************************************************************************* @@ -263,7 +267,7 @@ bool dc_stream_configure_crc(struct dc *dc, struct dc_stream_state *stream, /* Only call if supported */ if (tg->funcs->configure_crc) return tg->funcs->configure_crc(tg, ¶m); - dm_logger_write(dc->ctx->logger, LOG_WARNING, "CRC capture not supported."); + DC_LOG_WARNING("CRC capture not supported."); return false; } @@ -296,10 +300,49 @@ bool dc_stream_get_crc(struct dc *dc, struct dc_stream_state *stream, if (tg->funcs->get_crc) return tg->funcs->get_crc(tg, r_cr, g_y, b_cb); - dm_logger_write(dc->ctx->logger, LOG_WARNING, "CRC capture not supported."); + DC_LOG_WARNING("CRC capture not supported."); return false; } +void dc_stream_set_dither_option(struct dc_stream_state *stream, + enum dc_dither_option option) +{ + struct bit_depth_reduction_params params; + struct dc_link *link = stream->status.link; + struct pipe_ctx *pipes = NULL; + int i; + + for (i = 0; i < MAX_PIPES; i++) { + if (link->dc->current_state->res_ctx.pipe_ctx[i].stream == + stream) { + pipes = &link->dc->current_state->res_ctx.pipe_ctx[i]; + break; + } + } + + if (!pipes) + return; + if (option > DITHER_OPTION_MAX) + return; + + stream->dither_option = option; + + memset(¶ms, 0, sizeof(params)); + resource_build_bit_depth_reduction_params(stream, ¶ms); + stream->bit_depth_params = params; + + if (pipes->plane_res.xfm && + pipes->plane_res.xfm->funcs->transform_set_pixel_storage_depth) { + pipes->plane_res.xfm->funcs->transform_set_pixel_storage_depth( + pipes->plane_res.xfm, + pipes->plane_res.scl_data.lb_params.depth, + &stream->bit_depth_params); + } + + pipes->stream_res.opp->funcs-> + opp_program_bit_depth_reduction(pipes->stream_res.opp, ¶ms); +} + void dc_stream_set_static_screen_events(struct dc *dc, struct dc_stream_state **streams, int num_streams, @@ -510,9 +553,6 @@ static bool construct(struct dc *dc, dc_version = resource_parse_asic_id(init_params->asic_id); dc_ctx->dce_version = dc_version; -#if defined(CONFIG_DRM_AMD_DC_FBC) - dc->ctx->fbc_gpu_addr = init_params->fbc_gpu_addr; -#endif /* Resource should construct all asic specific resources. * This should be the only place where we need to parse the asic id */ @@ -638,10 +678,15 @@ struct dc *dc_create(const struct dc_init_data *init_params) dc->caps.max_audios = dc->res_pool->audio_count; dc->caps.linear_pitch_alignment = 64; + /* Populate versioning information */ + dc->versions.dc_ver = DC_VER; + + if (dc->res_pool->dmcu != NULL) + dc->versions.dmcu_version = dc->res_pool->dmcu->dmcu_version; + dc->config = init_params->flags; - dm_logger_write(dc->ctx->logger, LOG_DC, - "Display Core initialized\n"); + DC_LOG_DC("Display Core initialized\n"); @@ -673,11 +718,13 @@ static void enable_timing_multisync( if (!ctx->res_ctx.pipe_ctx[i].stream || !ctx->res_ctx.pipe_ctx[i].stream->triggered_crtc_reset.enabled) continue; + if (ctx->res_ctx.pipe_ctx[i].stream == ctx->res_ctx.pipe_ctx[i].stream->triggered_crtc_reset.event_source) + continue; multisync_pipes[multisync_count] = &ctx->res_ctx.pipe_ctx[i]; multisync_count++; } - if (multisync_count > 1) { + if (multisync_count > 0) { dc->hwss.enable_per_frame_crtc_position_reset( dc, multisync_count, multisync_pipes); } @@ -799,7 +846,6 @@ bool dc_enable_stereo( return ret; } - /* * Applies given context to HW and copy it into current context. * It's up to the user to release the src context afterwards. @@ -818,7 +864,9 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c dc_streams[i] = context->streams[i]; if (!dcb->funcs->is_accelerated_mode(dcb)) - dc->hwss.enable_accelerated_mode(dc); + dc->hwss.enable_accelerated_mode(dc, context); + + dc->hwss.set_bandwidth(dc, context, false); /* re-program planes for existing stream, in case we need to * free up plane resource for later use @@ -888,6 +936,9 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c dc_enable_stereo(dc, context, dc_streams, context->stream_count); + /* pplib is notified if disp_num changed */ + dc->hwss.set_bandwidth(dc, context, true); + dc_release_state(dc->current_state); dc->current_state = context; @@ -907,7 +958,7 @@ bool dc_commit_state(struct dc *dc, struct dc_state *context) if (false == context_changed(dc, context)) return DC_OK; - dm_logger_write(dc->ctx->logger, LOG_DC, "%s: %d streams\n", + DC_LOG_DC("%s: %d streams\n", __func__, context->stream_count); for (i = 0; i < context->stream_count; i++) { @@ -1128,6 +1179,9 @@ static enum surface_update_type get_plane_info_update_type(const struct dc_surfa if (u->plane_info->rotation != u->surface->rotation) update_flags->bits.rotation_change = 1; + if (u->plane_info->format != u->surface->format) + update_flags->bits.pixel_format_change = 1; + if (u->plane_info->stereo_format != u->surface->stereo_format) update_flags->bits.stereo_format_change = 1; @@ -1164,9 +1218,11 @@ static enum surface_update_type get_plane_info_update_type(const struct dc_surfa if (update_flags->bits.rotation_change || update_flags->bits.stereo_format_change + || update_flags->bits.pixel_format_change || update_flags->bits.gamma_change || update_flags->bits.bpp_change - || update_flags->bits.bandwidth_change) + || update_flags->bits.bandwidth_change + || update_flags->bits.output_tf_change) return UPDATE_TYPE_FULL; return UPDATE_TYPE_MED; @@ -1336,6 +1392,7 @@ static void commit_planes_for_stream(struct dc *dc, struct dc_state *context) { int i, j; + struct pipe_ctx *top_pipe_to_program = NULL; if (update_type == UPDATE_TYPE_FULL) { dc->hwss.set_bandwidth(dc, context, false); @@ -1355,39 +1412,64 @@ static void commit_planes_for_stream(struct dc *dc, for (j = 0; j < dc->res_pool->pipe_count; j++) { struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j]; - if (update_type == UPDATE_TYPE_FAST || !pipe_ctx->plane_state) - continue; - if (!pipe_ctx->top_pipe && - pipe_ctx->stream && - pipe_ctx->stream == stream) { - struct dc_stream_status *stream_status = + pipe_ctx->stream && + pipe_ctx->stream == stream) { + struct dc_stream_status *stream_status = NULL; + + top_pipe_to_program = pipe_ctx; + + if (update_type == UPDATE_TYPE_FAST || !pipe_ctx->plane_state) + continue; + + stream_status = stream_get_status(context, pipe_ctx->stream); dc->hwss.apply_ctx_for_surface( dc, pipe_ctx->stream, stream_status->plane_count, context); + + if (stream_update && stream_update->abm_level && pipe_ctx->stream_res.abm) { + if (pipe_ctx->stream_res.tg->funcs->is_blanked) { + // if otg funcs defined check if blanked before programming + if (!pipe_ctx->stream_res.tg->funcs->is_blanked(pipe_ctx->stream_res.tg)) + pipe_ctx->stream_res.abm->funcs->set_abm_level( + pipe_ctx->stream_res.abm, stream->abm_level); + } else + pipe_ctx->stream_res.abm->funcs->set_abm_level( + pipe_ctx->stream_res.abm, stream->abm_level); + } } } if (update_type == UPDATE_TYPE_FULL) context_timing_trace(dc, &context->res_ctx); - /* Perform requested Updates */ - for (i = 0; i < surface_count; i++) { - struct dc_plane_state *plane_state = srf_updates[i].surface; + /* Lock the top pipe while updating plane addrs, since freesync requires + * plane addr update event triggers to be synchronized. + * top_pipe_to_program is expected to never be NULL + */ + if (update_type == UPDATE_TYPE_FAST) { + dc->hwss.pipe_control_lock(dc, top_pipe_to_program, true); - for (j = 0; j < dc->res_pool->pipe_count; j++) { - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j]; + /* Perform requested Updates */ + for (i = 0; i < surface_count; i++) { + struct dc_plane_state *plane_state = srf_updates[i].surface; - if (pipe_ctx->stream != stream) - continue; + for (j = 0; j < dc->res_pool->pipe_count; j++) { + struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j]; - if (pipe_ctx->plane_state != plane_state) - continue; + if (pipe_ctx->stream != stream) + continue; - if (update_type == UPDATE_TYPE_FAST && srf_updates[i].flip_addr) + if (pipe_ctx->plane_state != plane_state) + continue; + + if (srf_updates[i].flip_addr) dc->hwss.update_plane_addr(dc, pipe_ctx); + } } + + dc->hwss.pipe_control_lock(dc, top_pipe_to_program, false); } if (stream && stream_update && update_type > UPDATE_TYPE_FAST) @@ -1640,12 +1722,17 @@ struct dc_sink *dc_link_add_remote_sink( &dc_sink->dc_edid, &dc_sink->edid_caps); - if (edid_status != EDID_OK) - goto fail; + /* + * Treat device as no EDID device if EDID + * parsing fails + */ + if (edid_status != EDID_OK) { + dc_sink->dc_edid.length = 0; + dm_error("Bad EDID, status%d!\n", edid_status); + } return dc_sink; -fail: - dc_link_remove_remote_sink(link, dc_sink); + fail_add_sink: dc_sink_release(dc_sink); return NULL;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c index 1babac0..5a552cb3 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
@@ -36,26 +36,22 @@ #include "hw_sequencer.h" #include "resource.h" +#define DC_LOGGER \ + logger #define SURFACE_TRACE(...) do {\ if (dc->debug.surface_trace) \ - dm_logger_write(logger, \ - LOG_IF_TRACE, \ - ##__VA_ARGS__); \ + DC_LOG_IF_TRACE(__VA_ARGS__); \ } while (0) #define TIMING_TRACE(...) do {\ if (dc->debug.timing_trace) \ - dm_logger_write(logger, \ - LOG_SYNC, \ - ##__VA_ARGS__); \ + DC_LOG_SYNC(__VA_ARGS__); \ } while (0) #define CLOCK_TRACE(...) do {\ if (dc->debug.clock_trace) \ - dm_logger_write(logger, \ - LOG_BANDWIDTH_CALCS, \ - ##__VA_ARGS__); \ + DC_LOG_BANDWIDTH_CALCS(__VA_ARGS__); \ } while (0) void pre_surface_trace( @@ -361,25 +357,20 @@ void context_clock_trace( struct dc *core_dc = dc; struct dal_logger *logger = core_dc->ctx->logger; - CLOCK_TRACE("Current: dispclk_khz:%d dppclk_div:%d dcfclk_khz:%d\n" - "dcfclk_deep_sleep_khz:%d fclk_khz:%d\n" - "dram_ccm_us:%d min_active_dram_ccm_us:%d\n", + CLOCK_TRACE("Current: dispclk_khz:%d max_dppclk_khz:%d dcfclk_khz:%d\n" + "dcfclk_deep_sleep_khz:%d fclk_khz:%d socclk_khz:%d\n", context->bw.dcn.calc_clk.dispclk_khz, - context->bw.dcn.calc_clk.dppclk_div, + context->bw.dcn.calc_clk.dppclk_khz, context->bw.dcn.calc_clk.dcfclk_khz, context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz, context->bw.dcn.calc_clk.fclk_khz, - context->bw.dcn.calc_clk.dram_ccm_us, - context->bw.dcn.calc_clk.min_active_dram_ccm_us); - CLOCK_TRACE("Calculated: dispclk_khz:%d dppclk_div:%d dcfclk_khz:%d\n" - "dcfclk_deep_sleep_khz:%d fclk_khz:%d\n" - "dram_ccm_us:%d min_active_dram_ccm_us:%d\n", + context->bw.dcn.calc_clk.socclk_khz); + CLOCK_TRACE("Calculated: dispclk_khz:%d max_dppclk_khz:%d dcfclk_khz:%d\n" + "dcfclk_deep_sleep_khz:%d fclk_khz:%d socclk_khz:%d\n", context->bw.dcn.calc_clk.dispclk_khz, - context->bw.dcn.calc_clk.dppclk_div, + context->bw.dcn.calc_clk.dppclk_khz, context->bw.dcn.calc_clk.dcfclk_khz, context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz, - context->bw.dcn.calc_clk.fclk_khz, - context->bw.dcn.calc_clk.dram_ccm_us, - context->bw.dcn.calc_clk.min_active_dram_ccm_us); + context->bw.dcn.calc_clk.fclk_khz); #endif }
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index dbc8b52..6d1c498 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -45,9 +45,11 @@ #include "dce/dce_11_0_d.h" #include "dce/dce_11_0_enum.h" #include "dce/dce_11_0_sh_mask.h" +#define DC_LOGGER \ + dc_ctx->logger #define LINK_INFO(...) \ - dm_logger_write(dc_ctx->logger, LOG_HW_HOTPLUG, \ + DC_LOG_HW_HOTPLUG( \ __VA_ARGS__) /******************************************************************************* @@ -126,6 +128,8 @@ static bool program_hpd_filter( int delay_on_connect_in_ms = 0; int delay_on_disconnect_in_ms = 0; + if (link->is_hpd_filter_disabled) + return false; /* Verify feature is supported */ switch (link->connector_signal) { case SIGNAL_TYPE_DVI_SINGLE_LINK: @@ -464,7 +468,7 @@ static void link_disconnect_sink(struct dc_link *link) link->dpcd_sink_count = 0; } -static void detect_dp( +static bool detect_dp( struct dc_link *link, struct display_sink_capability *sink_caps, bool *converter_disable_audio, @@ -478,7 +482,8 @@ static void detect_dp( if (sink_caps->transaction_type == DDC_TRANSACTION_TYPE_I2C_OVER_AUX) { sink_caps->signal = SIGNAL_TYPE_DISPLAY_PORT; - detect_dp_sink_caps(link); + if (!detect_dp_sink_caps(link)) + return false; if (is_mst_supported(link)) { sink_caps->signal = SIGNAL_TYPE_DISPLAY_PORT_MST; @@ -529,7 +534,7 @@ static void detect_dp( * active dongle unplug processing for short irq */ link_disconnect_sink(link); - return; + return true; } if (link->dpcd_caps.dongle_type != DISPLAY_DONGLE_DP_HDMI_CONVERTER) @@ -541,6 +546,8 @@ static void detect_dp( sink_caps, audio_support); } + + return true; } bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) @@ -604,11 +611,12 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) } case SIGNAL_TYPE_DISPLAY_PORT: { - detect_dp( + if (!detect_dp( link, &sink_caps, &converter_disable_audio, - aud_support, reason); + aud_support, reason)) + return false; /* Active dongle downstream unplug */ if (link->type == dc_connection_active_dongle @@ -671,12 +679,10 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) switch (edid_status) { case EDID_BAD_CHECKSUM: - dm_logger_write(link->ctx->logger, LOG_ERROR, - "EDID checksum invalid.\n"); + DC_LOG_ERROR("EDID checksum invalid.\n"); break; case EDID_NO_RESPONSE: - dm_logger_write(link->ctx->logger, LOG_ERROR, - "No EDID read.\n"); + DC_LOG_ERROR("No EDID read.\n"); default: break; } @@ -706,8 +712,7 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) "%s: [Block %d] ", sink->edid_caps.display_name, i); } - dm_logger_write(link->ctx->logger, LOG_DETECTION_EDID_PARSER, - "%s: " + DC_LOG_DETECTION_EDID_PARSER("%s: " "manufacturer_id = %X, " "product_id = %X, " "serial_number = %X, " @@ -727,8 +732,7 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) sink->edid_caps.audio_mode_count); for (i = 0; i < sink->edid_caps.audio_mode_count; i++) { - dm_logger_write(link->ctx->logger, LOG_DETECTION_EDID_PARSER, - "%s: mode number = %d, " + DC_LOG_DETECTION_EDID_PARSER("%s: mode number = %d, " "format_code = %d, " "channel_count = %d, " "sample_rate = %d, " @@ -978,8 +982,7 @@ static bool construct( } break; default: - dm_logger_write(dc_ctx->logger, LOG_WARNING, - "Unsupported Connector type:%d!\n", link->link_id.id); + DC_LOG_WARNING("Unsupported Connector type:%d!\n", link->link_id.id); goto create_fail; } @@ -1132,7 +1135,7 @@ static void dpcd_configure_panel_mode( { union dpcd_edp_config edp_config_set; bool panel_mode_edp = false; - + struct dc_context *dc_ctx = link->ctx; memset(&edp_config_set, '\0', sizeof(union dpcd_edp_config)); if (DP_PANEL_MODE_DEFAULT != panel_mode) { @@ -1169,8 +1172,7 @@ static void dpcd_configure_panel_mode( ASSERT(result == DDC_RESULT_SUCESSFULL); } } - dm_logger_write(link->ctx->logger, LOG_DETECTION_DP_CAPS, - "Link: %d eDP panel mode supported: %d " + DC_LOG_DETECTION_DP_CAPS("Link: %d eDP panel mode supported: %d " "eDP panel mode enabled: %d \n", link->link_index, link->dpcd_caps.panel_mode_edp, @@ -1246,6 +1248,12 @@ static enum dc_status enable_link_dp( pipe_ctx->clock_source->id, &link_settings); + if (stream->sink->edid_caps.panel_patch.dppowerup_delay > 0) { + int delay_dp_power_up_in_ms = stream->sink->edid_caps.panel_patch.dppowerup_delay; + + msleep(delay_dp_power_up_in_ms); + } + panel_mode = dp_get_panel_mode(link); dpcd_configure_panel_mode(link, panel_mode); @@ -1277,13 +1285,12 @@ static enum dc_status enable_link_edp( enum dc_status status; struct dc_stream_state *stream = pipe_ctx->stream; struct dc_link *link = stream->sink->link; - + /*in case it is not on*/ link->dc->hwss.edp_power_control(link, true); link->dc->hwss.edp_wait_for_hpd_ready(link, true); status = enable_link_dp(state, pipe_ctx); - link->dc->hwss.edp_backlight_control(link, true); return status; } @@ -1300,6 +1307,9 @@ static enum dc_status enable_link_dp_mst( if (link->cur_link_settings.lane_count != LANE_COUNT_UNKNOWN) return DC_OK; + /* clear payload table */ + dm_helpers_dp_mst_clear_payload_allocation_table(link->ctx, link); + /* set the sink to MST mode before enabling the link */ dp_enable_mst_on_sink(link, true); @@ -1785,9 +1795,21 @@ static enum dc_status enable_link( } if (pipe_ctx->stream_res.audio && status == DC_OK) { + struct dc *core_dc = pipe_ctx->stream->ctx->dc; /* notify audio driver for audio modes of monitor */ + struct pp_smu_funcs_rv *pp_smu = core_dc->res_pool->pp_smu; + unsigned int i, num_audio = 1; + for (i = 0; i < MAX_PIPES; i++) { + /*current_state not updated yet*/ + if (core_dc->current_state->res_ctx.pipe_ctx[i].stream_res.audio != NULL) + num_audio++; + } + pipe_ctx->stream_res.audio->funcs->az_enable(pipe_ctx->stream_res.audio); + if (num_audio == 1 && pp_smu != NULL && pp_smu->set_pme_wa_enable != NULL) + /*this is the first audio. apply the PME w/a in order to wake AZ from D3*/ + pp_smu->set_pme_wa_enable(&pp_smu->pp_smu); /* un-mute audio */ /* TODO: audio should be per stream rather than per link */ pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control( @@ -1928,6 +1950,7 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level, struct dc *core_dc = link->ctx->dc; struct abm *abm = core_dc->res_pool->abm; struct dmcu *dmcu = core_dc->res_pool->dmcu; + struct dc_context *dc_ctx = link->ctx; unsigned int controller_id = 0; bool use_smooth_brightness = true; int i; @@ -1937,10 +1960,16 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level, (abm->funcs->set_backlight_level == NULL)) return false; + if (stream) { + if (stream->bl_pwm_level == 0) + frame_ramp = 0; + + ((struct dc_stream_state *)stream)->bl_pwm_level = level; + } + use_smooth_brightness = dmcu->funcs->is_dmcu_initialized(dmcu); - dm_logger_write(link->ctx->logger, LOG_BACKLIGHT, - "New Backlight level: %d (0x%X)\n", level, level); + DC_LOG_BACKLIGHT("New Backlight level: %d (0x%X)\n", level, level); if (dc_is_embedded_signal(link->connector_signal)) { if (stream != NULL) { @@ -1968,6 +1997,19 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level, return true; } +bool dc_link_set_abm_disable(const struct dc_link *link) +{ + struct dc *core_dc = link->ctx->dc; + struct abm *abm = core_dc->res_pool->abm; + + if ((abm == NULL) || (abm->funcs->set_backlight_level == NULL)) + return false; + + abm->funcs->set_abm_immediate_disable(abm); + + return true; +} + bool dc_link_set_psr_enable(const struct dc_link *link, bool enable, bool wait) { struct dc *core_dc = link->ctx->dc; @@ -2107,6 +2149,7 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx) struct fixed31_32 avg_time_slots_per_mtp; struct fixed31_32 pbn; struct fixed31_32 pbn_per_slot; + struct dc_context *dc_ctx = link->ctx; uint8_t i; /* enable_link_dp_mst already check link->enabled_stream_count @@ -2124,21 +2167,18 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx) link, pipe_ctx->stream_res.stream_enc, &proposed_table); } else - dm_logger_write(link->ctx->logger, LOG_WARNING, - "Failed to update" + DC_LOG_WARNING("Failed to update" "MST allocation table for" "pipe idx:%d\n", pipe_ctx->pipe_idx); - dm_logger_write(link->ctx->logger, LOG_MST, - "%s " + DC_LOG_MST("%s " "stream_count: %d: \n ", __func__, link->mst_stream_alloc_table.stream_count); for (i = 0; i < MAX_CONTROLLER_NUM; i++) { - dm_logger_write(link->ctx->logger, LOG_MST, - "stream_enc[%d]: 0x%x " + DC_LOG_MST("stream_enc[%d]: 0x%x " "stream[%d].vcp_id: %d " "stream[%d].slot_count: %d\n", i, @@ -2189,6 +2229,7 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx) struct fixed31_32 avg_time_slots_per_mtp = dal_fixed31_32_from_int(0); uint8_t i; bool mst_mode = (link->type == dc_connection_mst_branch); + struct dc_context *dc_ctx = link->ctx; /* deallocate_mst_payload is called before disable link. When mode or * disable/enable monitor, new stream is created which is not in link @@ -2214,23 +2255,20 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx) link, pipe_ctx->stream_res.stream_enc, &proposed_table); } else { - dm_logger_write(link->ctx->logger, LOG_WARNING, - "Failed to update" + DC_LOG_WARNING("Failed to update" "MST allocation table for" "pipe idx:%d\n", pipe_ctx->pipe_idx); } } - dm_logger_write(link->ctx->logger, LOG_MST, - "%s" + DC_LOG_MST("%s" "stream_count: %d: ", __func__, link->mst_stream_alloc_table.stream_count); for (i = 0; i < MAX_CONTROLLER_NUM; i++) { - dm_logger_write(link->ctx->logger, LOG_MST, - "stream_enc[%d]: 0x%x " + DC_LOG_MST("stream_enc[%d]: 0x%x " "stream[%d].vcp_id: %d " "stream[%d].slot_count: %d\n", i, @@ -2264,12 +2302,24 @@ void core_link_enable_stream( struct pipe_ctx *pipe_ctx) { struct dc *core_dc = pipe_ctx->stream->ctx->dc; + struct dc_context *dc_ctx = pipe_ctx->stream->ctx; + enum dc_status status; - enum dc_status status = enable_link(state, pipe_ctx); + /* eDP lit up by bios already, no need to enable again. */ + if (pipe_ctx->stream->signal == SIGNAL_TYPE_EDP && + core_dc->apply_edp_fast_boot_optimization) { + core_dc->apply_edp_fast_boot_optimization = false; + pipe_ctx->stream->dpms_off = false; + return; + } + + if (pipe_ctx->stream->dpms_off) + return; + + status = enable_link(state, pipe_ctx); if (status != DC_OK) { - dm_logger_write(pipe_ctx->stream->ctx->logger, - LOG_WARNING, "enabling link %u failed: %d\n", + DC_LOG_WARNING("enabling link %u failed: %d\n", pipe_ctx->stream->sink->link->link_index, status); @@ -2295,9 +2345,8 @@ void core_link_enable_stream( if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) allocate_mst_payload(pipe_ctx); - if (dc_is_dp_signal(pipe_ctx->stream->signal)) - core_dc->hwss.unblank_stream(pipe_ctx, - &pipe_ctx->stream->sink->link->cur_link_settings); + core_dc->hwss.unblank_stream(pipe_ctx, + &pipe_ctx->stream->sink->link->cur_link_settings); } void core_link_disable_stream(struct pipe_ctx *pipe_ctx, int option) @@ -2307,8 +2356,7 @@ void core_link_disable_stream(struct pipe_ctx *pipe_ctx, int option) if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) deallocate_mst_payload(pipe_ctx); - if (pipe_ctx->stream->signal == SIGNAL_TYPE_EDP) - core_dc->hwss.edp_backlight_control(pipe_ctx->stream->sink->link, false); + core_dc->hwss.blank_stream(pipe_ctx); core_dc->hwss.disable_stream(pipe_ctx, option); @@ -2325,3 +2373,36 @@ void core_link_set_avmute(struct pipe_ctx *pipe_ctx, bool enable) core_dc->hwss.set_avmute(pipe_ctx, enable); } +void dc_link_enable_hpd_filter(struct dc_link *link, bool enable) +{ + struct gpio *hpd; + + if (enable) { + link->is_hpd_filter_disabled = false; + program_hpd_filter(link); + } else { + link->is_hpd_filter_disabled = true; + /* Obtain HPD handle */ + hpd = get_hpd_gpio(link->ctx->dc_bios, link->link_id, link->ctx->gpio_service); + + if (!hpd) + return; + + /* Setup HPD filtering */ + if (dal_gpio_open(hpd, GPIO_MODE_INTERRUPT) == GPIO_RESULT_OK) { + struct gpio_hpd_config config; + + config.delay_on_connect = 0; + config.delay_on_disconnect = 0; + + dal_irq_setup_hpd_filter(hpd, &config); + + dal_gpio_close(hpd); + } else { + ASSERT_CRITICAL(false); + } + /* Release HPD handle */ + dal_gpio_destroy_irq(&hpd); + } +} +
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index 8458c79..5fce302 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -11,6 +11,8 @@ #include "dpcd_defs.h" #include "resource.h" +#define DC_LOGGER \ + link->ctx->logger /* maximum pre emphasis level allowed for each voltage swing level*/ static const enum dc_pre_emphasis voltage_swing_to_pre_emphasis[] = { @@ -63,8 +65,7 @@ static void wait_for_training_aux_rd_interval( udelay(default_wait_in_micro_secs); - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s:\n wait = %d\n", + DC_LOG_HW_LINK_TRAINING("%s:\n wait = %d\n", __func__, default_wait_in_micro_secs); } @@ -79,8 +80,7 @@ static void dpcd_set_training_pattern( &dpcd_pattern.raw, 1); - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s\n %x pattern = %x\n", + DC_LOG_HW_LINK_TRAINING("%s\n %x pattern = %x\n", __func__, DP_TRAINING_PATTERN_SET, dpcd_pattern.v1_4.TRAINING_PATTERN_SET); @@ -116,8 +116,7 @@ static void dpcd_set_link_settings( core_link_write_dpcd(link, DP_DOWNSPREAD_CTRL, &downspread.raw, sizeof(downspread)); - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s\n %x rate = %x\n %x lane = %x\n %x spread = %x\n", + DC_LOG_HW_LINK_TRAINING("%s\n %x rate = %x\n %x lane = %x\n %x spread = %x\n", __func__, DP_LINK_BW_SET, lt_settings->link_settings.link_rate, @@ -151,8 +150,7 @@ static enum dpcd_training_patterns break; default: ASSERT(0); - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s: Invalid HW Training pattern: %d\n", + DC_LOG_HW_LINK_TRAINING("%s: Invalid HW Training pattern: %d\n", __func__, pattern); break; } @@ -184,8 +182,7 @@ static void dpcd_set_lt_pattern_and_lane_settings( dpcd_lt_buffer[DP_TRAINING_PATTERN_SET - dpcd_base_lt_offset] = dpcd_pattern.raw; - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s\n %x pattern = %x\n", + DC_LOG_HW_LINK_TRAINING("%s\n %x pattern = %x\n", __func__, DP_TRAINING_PATTERN_SET, dpcd_pattern.v1_4.TRAINING_PATTERN_SET); @@ -219,8 +216,7 @@ static void dpcd_set_lt_pattern_and_lane_settings( dpcd_lane, size_in_bytes); - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s:\n %x VS set = %x PE set = %x max VS Reached = %x max PE Reached = %x\n", + DC_LOG_HW_LINK_TRAINING("%s:\n %x VS set = %x PE set = %x max VS Reached = %x max PE Reached = %x\n", __func__, DP_TRAINING_LANE0_SET, dpcd_lane[0].bits.VOLTAGE_SWING_SET, @@ -456,14 +452,12 @@ static void get_lane_status_and_drive_settings( ln_status_updated->raw = dpcd_buf[2]; - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s:\n%x Lane01Status = %x\n %x Lane23Status = %x\n ", + DC_LOG_HW_LINK_TRAINING("%s:\n%x Lane01Status = %x\n %x Lane23Status = %x\n ", __func__, DP_LANE0_1_STATUS, dpcd_buf[0], DP_LANE2_3_STATUS, dpcd_buf[1]); - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s:\n %x Lane01AdjustRequest = %x\n %x Lane23AdjustRequest = %x\n", + DC_LOG_HW_LINK_TRAINING("%s:\n %x Lane01AdjustRequest = %x\n %x Lane23AdjustRequest = %x\n", __func__, DP_ADJUST_REQUEST_LANE0_1, dpcd_buf[4], @@ -556,8 +550,7 @@ static void dpcd_set_lane_settings( } */ - dm_logger_write(link->ctx->logger, LOG_HW_LINK_TRAINING, - "%s\n %x VS set = %x PE set = %x max VS Reached = %x max PE Reached = %x\n", + DC_LOG_HW_LINK_TRAINING("%s\n %x VS set = %x PE set = %x max VS Reached = %x max PE Reached = %x\n", __func__, DP_TRAINING_LANE0_SET, dpcd_lane[0].bits.VOLTAGE_SWING_SET, @@ -669,16 +662,14 @@ static bool perform_post_lt_adj_req_sequence( } if (!req_drv_setting_changed) { - dm_logger_write(link->ctx->logger, LOG_WARNING, - "%s: Post Link Training Adjust Request Timed out\n", + DC_LOG_WARNING("%s: Post Link Training Adjust Request Timed out\n", __func__); ASSERT(0); return true; } } - dm_logger_write(link->ctx->logger, LOG_WARNING, - "%s: Post Link Training Adjust Request limit reached\n", + DC_LOG_WARNING("%s: Post Link Training Adjust Request limit reached\n", __func__); ASSERT(0); @@ -709,6 +700,22 @@ static enum hw_dp_training_pattern get_supported_tp(struct dc_link *link) return HW_DP_TRAINING_PATTERN_2; } +static enum link_training_result get_cr_failure(enum dc_lane_count ln_count, + union lane_status *dpcd_lane_status) +{ + enum link_training_result result = LINK_TRAINING_SUCCESS; + + if (ln_count >= LANE_COUNT_ONE && !dpcd_lane_status[0].bits.CR_DONE_0) + result = LINK_TRAINING_CR_FAIL_LANE0; + else if (ln_count >= LANE_COUNT_TWO && !dpcd_lane_status[1].bits.CR_DONE_0) + result = LINK_TRAINING_CR_FAIL_LANE1; + else if (ln_count >= LANE_COUNT_FOUR && !dpcd_lane_status[2].bits.CR_DONE_0) + result = LINK_TRAINING_CR_FAIL_LANE23; + else if (ln_count >= LANE_COUNT_FOUR && !dpcd_lane_status[3].bits.CR_DONE_0) + result = LINK_TRAINING_CR_FAIL_LANE23; + return result; +} + static enum link_training_result perform_channel_equalization_sequence( struct dc_link *link, struct link_training_settings *lt_settings) @@ -718,7 +725,7 @@ static enum link_training_result perform_channel_equalization_sequence( uint32_t retries_ch_eq; enum dc_lane_count lane_count = lt_settings->link_settings.lane_count; union lane_align_status_updated dpcd_lane_status_updated = {{0}}; - union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = {{{0}}};; + union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = {{{0}}}; hw_tr_pattern = get_supported_tp(link); @@ -771,7 +778,7 @@ static enum link_training_result perform_channel_equalization_sequence( } -static bool perform_clock_recovery_sequence( +static enum link_training_result perform_clock_recovery_sequence( struct dc_link *link, struct link_training_settings *lt_settings) { @@ -846,11 +853,11 @@ static bool perform_clock_recovery_sequence( /* 5. check CR done*/ if (is_cr_done(lane_count, dpcd_lane_status)) - return true; + return LINK_TRAINING_SUCCESS; /* 6. max VS reached*/ if (is_max_vs_reached(lt_settings)) - return false; + break; /* 7. same voltage*/ /* Note: VS same for all lanes, @@ -869,20 +876,19 @@ static bool perform_clock_recovery_sequence( if (retry_count >= LINK_TRAINING_MAX_CR_RETRY) { ASSERT(0); - dm_logger_write(link->ctx->logger, LOG_ERROR, - "%s: Link Training Error, could not get CR after %d tries. Possibly voltage swing issue", + DC_LOG_ERROR("%s: Link Training Error, could not get CR after %d tries. Possibly voltage swing issue", __func__, LINK_TRAINING_MAX_CR_RETRY); } - return false; + return get_cr_failure(lane_count, dpcd_lane_status); } -static inline bool perform_link_training_int( +static inline enum link_training_result perform_link_training_int( struct dc_link *link, struct link_training_settings *lt_settings, - bool status) + enum link_training_result status) { union lane_count_set lane_count_set = { {0} }; union dpcd_training_pattern dpcd_pattern = { {0} }; @@ -903,9 +909,9 @@ static inline bool perform_link_training_int( get_supported_tp(link) == HW_DP_TRAINING_PATTERN_4) return status; - if (status && + if (status == LINK_TRAINING_SUCCESS && perform_post_lt_adj_req_sequence(link, lt_settings) == false) - status = false; + status = LINK_TRAINING_LQA_FAIL; lane_count_set.bits.LANE_COUNT_SET = lt_settings->link_settings.lane_count; lane_count_set.bits.ENHANCED_FRAMING = 1; @@ -928,6 +934,8 @@ enum link_training_result dc_link_dp_perform_link_training( enum link_training_result status = LINK_TRAINING_SUCCESS; char *link_rate = "Unknown"; + char *lt_result = "Unknown"; + struct link_training_settings lt_settings; memset(<_settings, '\0', sizeof(lt_settings)); @@ -951,22 +959,16 @@ enum link_training_result dc_link_dp_perform_link_training( /* 2. perform link training (set link training done * to false is done as well)*/ - if (!perform_clock_recovery_sequence(link, <_settings)) { - status = LINK_TRAINING_CR_FAIL; - } else { + status = perform_clock_recovery_sequence(link, <_settings); + if (status == LINK_TRAINING_SUCCESS) { status = perform_channel_equalization_sequence(link, <_settings); } if ((status == LINK_TRAINING_SUCCESS) || !skip_video_pattern) { - if (!perform_link_training_int(link, + status = perform_link_training_int(link, <_settings, - status == LINK_TRAINING_SUCCESS)) { - /* the next link training setting in this case - * would be the same as CR failure case. - */ - status = LINK_TRAINING_CR_FAIL; - } + status); } /* 6. print status message*/ @@ -991,13 +993,37 @@ enum link_training_result dc_link_dp_perform_link_training( break; } + switch (status) { + case LINK_TRAINING_SUCCESS: + lt_result = "pass"; + break; + case LINK_TRAINING_CR_FAIL_LANE0: + lt_result = "CR failed lane0"; + break; + case LINK_TRAINING_CR_FAIL_LANE1: + lt_result = "CR failed lane1"; + break; + case LINK_TRAINING_CR_FAIL_LANE23: + lt_result = "CR failed lane23"; + break; + case LINK_TRAINING_EQ_FAIL_CR: + lt_result = "CR failed in EQ"; + break; + case LINK_TRAINING_EQ_FAIL_EQ: + lt_result = "EQ failed"; + break; + case LINK_TRAINING_LQA_FAIL: + lt_result = "LQA failed"; + break; + default: + break; + } + /* Connectivity log: link training */ CONN_MSG_LT(link, "%sx%d %s VS=%d, PE=%d", link_rate, lt_settings.link_settings.lane_count, - (status == LINK_TRAINING_SUCCESS) ? "pass" : - ((status == LINK_TRAINING_CR_FAIL) ? "CR failed" : - "EQ failed"), + lt_result, lt_settings.lane_settings[0].VOLTAGE_SWING, lt_settings.lane_settings[0].PRE_EMPHASIS); @@ -1115,6 +1141,7 @@ bool dp_hbr_verify_link_cap( dp_cs_id, cur); + if (skip_link_training) success = true; else { @@ -1279,7 +1306,10 @@ static bool decide_fallback_link_setting( return false; switch (training_result) { - case LINK_TRAINING_CR_FAIL: + case LINK_TRAINING_CR_FAIL_LANE0: + case LINK_TRAINING_CR_FAIL_LANE1: + case LINK_TRAINING_CR_FAIL_LANE23: + case LINK_TRAINING_LQA_FAIL: { if (!reached_minimum_link_rate (current_link_setting->link_rate)) { @@ -1290,8 +1320,18 @@ static bool decide_fallback_link_setting( (current_link_setting->lane_count)) { current_link_setting->link_rate = initial_link_settings.link_rate; - current_link_setting->lane_count = - reduce_lane_count( + if (training_result == LINK_TRAINING_CR_FAIL_LANE0) + return false; + else if (training_result == LINK_TRAINING_CR_FAIL_LANE1) + current_link_setting->lane_count = + LANE_COUNT_ONE; + else if (training_result == + LINK_TRAINING_CR_FAIL_LANE23) + current_link_setting->lane_count = + LANE_COUNT_TWO; + else + current_link_setting->lane_count = + reduce_lane_count( current_link_setting->lane_count); } else { return false; @@ -1556,8 +1596,7 @@ static bool hpd_rx_irq_check_link_loss_status( if (sink_status_changed || !hpd_irq_dpcd_data->bytes.lane_status_updated.bits.INTERLANE_ALIGN_DONE) { - dm_logger_write(link->ctx->logger, LOG_HW_HPD_IRQ, - "%s: Link Status changed.\n", __func__); + DC_LOG_HW_HPD_IRQ("%s: Link Status changed.\n", __func__); return_code = true; @@ -1570,8 +1609,7 @@ static bool hpd_rx_irq_check_link_loss_status( sizeof(irq_reg_rx_power_state)); if (dpcd_result != DC_OK) { - dm_logger_write(link->ctx->logger, LOG_HW_HPD_IRQ, - "%s: DPCD read failed to obtain power state.\n", + DC_LOG_HW_HPD_IRQ("%s: DPCD read failed to obtain power state.\n", __func__); } else { if (irq_reg_rx_power_state != DP_SET_POWER_D0) @@ -1930,8 +1968,7 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd * PSR and device auto test, refer to function handle_sst_hpd_irq * in DAL2.1*/ - dm_logger_write(link->ctx->logger, LOG_HW_HPD_IRQ, - "%s: Got short pulse HPD on link %d\n", + DC_LOG_HW_HPD_IRQ("%s: Got short pulse HPD on link %d\n", __func__, link->link_index); @@ -1945,8 +1982,7 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd *out_hpd_irq_dpcd_data = hpd_irq_dpcd_data; if (result != DC_OK) { - dm_logger_write(link->ctx->logger, LOG_HW_HPD_IRQ, - "%s: DPCD read failed to obtain irq data\n", + DC_LOG_HW_HPD_IRQ("%s: DPCD read failed to obtain irq data\n", __func__); return false; } @@ -1964,8 +2000,7 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd } if (!allow_hpd_rx_irq(link)) { - dm_logger_write(link->ctx->logger, LOG_HW_HPD_IRQ, - "%s: skipping HPD handling on %d\n", + DC_LOG_HW_HPD_IRQ("%s: skipping HPD handling on %d\n", __func__, link->link_index); return false; } @@ -2233,13 +2268,14 @@ static void dp_wa_power_up_0010FA(struct dc_link *link, uint8_t *dpcd_data, link->wa_flags.dp_keep_receiver_powered = false; } -static void retrieve_link_cap(struct dc_link *link) +static bool retrieve_link_cap(struct dc_link *link) { uint8_t dpcd_data[DP_TRAINING_AUX_RD_INTERVAL - DP_DPCD_REV + 1]; union down_stream_port_count down_strm_port_count; union edp_configuration_cap edp_config_cap; union dp_downstream_port_present ds_port = { 0 }; + enum dc_status status = DC_ERROR_UNEXPECTED; memset(dpcd_data, '\0', sizeof(dpcd_data)); memset(&down_strm_port_count, @@ -2247,11 +2283,16 @@ static void retrieve_link_cap(struct dc_link *link) memset(&edp_config_cap, '\0', sizeof(union edp_configuration_cap)); - core_link_read_dpcd( - link, - DP_DPCD_REV, - dpcd_data, - sizeof(dpcd_data)); + status = core_link_read_dpcd( + link, + DP_DPCD_REV, + dpcd_data, + sizeof(dpcd_data)); + + if (status != DC_OK) { + dm_error("%s: Read dpcd data failed.\n", __func__); + return false; + } { union training_aux_rd_interval aux_rd_interval; @@ -2313,11 +2354,13 @@ static void retrieve_link_cap(struct dc_link *link) /* Connectivity log: detection */ CONN_DATA_DETECT(link, dpcd_data, sizeof(dpcd_data), "Rx Caps: "); + + return true; } -void detect_dp_sink_caps(struct dc_link *link) +bool detect_dp_sink_caps(struct dc_link *link) { - retrieve_link_cap(link); + return retrieve_link_cap(link); /* dc init_hw has power encoder using default * signal for connector. For native DP, no
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c index 2096f2a..7c866a7 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
@@ -102,7 +102,7 @@ void dp_enable_link_phy( dp_receiver_power_ctrl(link, true); } -static bool edp_receiver_ready_T9(struct dc_link *link) +bool edp_receiver_ready_T9(struct dc_link *link) { unsigned int tries = 0; unsigned char sinkstatus = 0; @@ -123,6 +123,28 @@ static bool edp_receiver_ready_T9(struct dc_link *link) } while (++tries < 50); return result; } +bool edp_receiver_ready_T7(struct dc_link *link) +{ + unsigned int tries = 0; + unsigned char sinkstatus = 0; + unsigned char edpRev = 0; + enum dc_status result = DC_OK; + + result = core_link_read_dpcd(link, DP_EDP_DPCD_REV, &edpRev, sizeof(edpRev)); + if (result == DC_OK && edpRev < DP_EDP_12) + return true; + /* start from eDP version 1.2, SINK_STAUS indicate the sink is ready.*/ + do { + sinkstatus = 0; + result = core_link_read_dpcd(link, DP_SINK_STATUS, &sinkstatus, sizeof(sinkstatus)); + if (sinkstatus == 1) + break; + if (result != DC_OK) + break; + udelay(25); //MAx T7 is 50ms + } while (++tries < 300); + return result; +} void dp_disable_link_phy(struct dc_link *link, enum signal_type signal) { @@ -130,7 +152,6 @@ void dp_disable_link_phy(struct dc_link *link, enum signal_type signal) dp_receiver_power_ctrl(link, false); if (signal == SIGNAL_TYPE_EDP) { - edp_receiver_ready_T9(link); link->link_enc->funcs->disable_output(link->link_enc, signal); link->dc->hwss.edp_power_control(link, false); } else @@ -258,6 +279,7 @@ void dp_retrain_link_dp_test(struct dc_link *link, for (i = 0; i < MAX_PIPES; i++) { if (pipes[i].stream != NULL && + !pipes[i].top_pipe && pipes[i].stream->sink != NULL && pipes[i].stream->sink->link != NULL && pipes[i].stream_res.stream_enc != NULL &&
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c index 4d07ffe..ba3487e 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -35,6 +35,7 @@ #include "core_types.h" #include "set_mode_types.h" #include "virtual/virtual_stream_encoder.h" +#include "dpcd_defs.h" #include "dce80/dce80_resource.h" #include "dce100/dce100_resource.h" @@ -44,7 +45,8 @@ #include "dcn10/dcn10_resource.h" #endif #include "dce120/dce120_resource.h" - +#define DC_LOGGER \ + ctx->logger enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id) { enum dce_version dc_version = DCE_VERSION_UNKNOWN; @@ -696,7 +698,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r /* Adjust for viewport end clip-off */ - if ((data->viewport.x + data->viewport.width) < (src.x + src.width)) { + if ((data->viewport.x + data->viewport.width) < (src.x + src.width) && !flip_horz_scan_dir) { int vp_clip = src.x + src.width - data->viewport.width - data->viewport.x; int int_part = dal_fixed31_32_floor( dal_fixed31_32_sub(data->inits.h, data->ratios.horz)); @@ -704,7 +706,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r int_part = int_part > 0 ? int_part : 0; data->viewport.width += int_part < vp_clip ? int_part : vp_clip; } - if ((data->viewport.y + data->viewport.height) < (src.y + src.height)) { + if ((data->viewport.y + data->viewport.height) < (src.y + src.height) && !flip_vert_scan_dir) { int vp_clip = src.y + src.height - data->viewport.height - data->viewport.y; int int_part = dal_fixed31_32_floor( dal_fixed31_32_sub(data->inits.v, data->ratios.vert)); @@ -712,7 +714,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r int_part = int_part > 0 ? int_part : 0; data->viewport.height += int_part < vp_clip ? int_part : vp_clip; } - if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div) { + if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div && !flip_horz_scan_dir) { int vp_clip = (src.x + src.width) / vpc_div - data->viewport_c.width - data->viewport_c.x; int int_part = dal_fixed31_32_floor( @@ -721,7 +723,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r int_part = int_part > 0 ? int_part : 0; data->viewport_c.width += int_part < vp_clip ? int_part : vp_clip; } - if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div) { + if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div && !flip_vert_scan_dir) { int vp_clip = (src.y + src.height) / vpc_div - data->viewport_c.height - data->viewport_c.y; int int_part = dal_fixed31_32_floor( @@ -833,7 +835,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) struct dc_crtc_timing *timing = &pipe_ctx->stream->timing; struct view recout_skip = { 0 }; bool res = false; - + struct dc_context *ctx = pipe_ctx->stream->ctx; /* Important: scaling ratio calculation requires pixel format, * lb depth calculation requires recout and taps require scaling ratios. * Inits require viewport, taps, ratios and recout of split pipe @@ -892,7 +894,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) /* May need to re-check lb size after this in some obscure scenario */ calculate_inits_and_adj_vp(pipe_ctx, &recout_skip); - dm_logger_write(pipe_ctx->stream->ctx->logger, LOG_SCALER, + DC_LOG_SCALER( "%s: Viewport:\nheight:%d width:%d x:%d " "y:%d\n dst_rect:\nheight:%d width:%d x:%d " "y:%d\n", @@ -1054,6 +1056,7 @@ static int acquire_first_split_pipe( pipe_ctx->plane_res.ipp = pool->ipps[i]; pipe_ctx->plane_res.dpp = pool->dpps[i]; pipe_ctx->stream_res.opp = pool->opps[i]; + pipe_ctx->plane_res.mpcc_inst = pool->dpps[i]->inst; pipe_ctx->pipe_idx = i; pipe_ctx->stream = stream; @@ -1121,6 +1124,7 @@ bool dc_add_plane_to_context( ASSERT(tail_pipe); free_pipe->stream_res.tg = tail_pipe->stream_res.tg; + free_pipe->stream_res.abm = tail_pipe->stream_res.abm; free_pipe->stream_res.opp = tail_pipe->stream_res.opp; free_pipe->stream_res.stream_enc = tail_pipe->stream_res.stream_enc; free_pipe->stream_res.audio = tail_pipe->stream_res.audio; @@ -1406,6 +1410,8 @@ static int acquire_first_free_pipe( pipe_ctx->plane_res.xfm = pool->transforms[i]; pipe_ctx->plane_res.dpp = pool->dpps[i]; pipe_ctx->stream_res.opp = pool->opps[i]; + if (pool->dpps[i]) + pipe_ctx->plane_res.mpcc_inst = pool->dpps[i]->inst; pipe_ctx->pipe_idx = i; @@ -1552,6 +1558,9 @@ enum dc_status dc_remove_stream_from_ctx( dc->res_pool, del_pipe->clock_source); + if (dc->res_pool->funcs->remove_stream_from_ctx) + dc->res_pool->funcs->remove_stream_from_ctx(dc, new_ctx, stream); + memset(del_pipe, 0, sizeof(*del_pipe)); break; @@ -1728,6 +1737,10 @@ enum dc_status resource_map_pool_resources( pipe_ctx->stream_res.audio, true); } + /* Add ABM to the resource if on EDP */ + if (pipe_ctx->stream && dc_is_embedded_signal(pipe_ctx->stream->signal)) + pipe_ctx->stream_res.abm = pool->abm; + for (i = 0; i < context->stream_count; i++) if (context->streams[i] == stream) { context->stream_status[i].primary_otg_inst = pipe_ctx->stream_res.tg->inst; @@ -2428,7 +2441,8 @@ static void set_vsc_info_packet( unsigned int vscPacketRevision = 0; unsigned int i; - if (stream->sink->link->psr_enabled) { + /*VSC packet set to 2 when DP revision >= 1.2*/ + if (stream->psr_version != 0) { vscPacketRevision = 2; }
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c index cd58197..ce0747e 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
@@ -198,8 +198,7 @@ bool dc_stream_set_cursor_attributes( for (i = 0; i < MAX_PIPES; i++) { struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i]; - if (pipe_ctx->stream != stream || (!pipe_ctx->plane_res.xfm && - !pipe_ctx->plane_res.dpp) || !pipe_ctx->plane_res.ipp) + if (pipe_ctx->stream != stream) continue; if (pipe_ctx->top_pipe && pipe_ctx->plane_state != pipe_ctx->top_pipe->plane_state) continue;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c index ade5b8e..132eef3 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
@@ -66,8 +66,8 @@ struct dc_plane_state *dc_create_plane_state(struct dc *dc) { struct dc *core_dc = dc; - struct dc_plane_state *plane_state = kzalloc(sizeof(*plane_state), - GFP_KERNEL); + struct dc_plane_state *plane_state = kvzalloc(sizeof(*plane_state), + GFP_KERNEL); if (NULL == plane_state) return NULL; @@ -120,7 +120,7 @@ static void dc_plane_state_free(struct kref *kref) { struct dc_plane_state *plane_state = container_of(kref, struct dc_plane_state, refcount); destruct(plane_state); - kfree(plane_state); + kvfree(plane_state); } void dc_plane_state_release(struct dc_plane_state *plane_state) @@ -136,7 +136,7 @@ void dc_gamma_retain(struct dc_gamma *gamma) static void dc_gamma_free(struct kref *kref) { struct dc_gamma *gamma = container_of(kref, struct dc_gamma, refcount); - kfree(gamma); + kvfree(gamma); } void dc_gamma_release(struct dc_gamma **gamma) @@ -147,7 +147,7 @@ void dc_gamma_release(struct dc_gamma **gamma) struct dc_gamma *dc_create_gamma(void) { - struct dc_gamma *gamma = kzalloc(sizeof(*gamma), GFP_KERNEL); + struct dc_gamma *gamma = kvzalloc(sizeof(*gamma), GFP_KERNEL); if (gamma == NULL) goto alloc_fail; @@ -167,7 +167,7 @@ void dc_transfer_func_retain(struct dc_transfer_func *tf) static void dc_transfer_func_free(struct kref *kref) { struct dc_transfer_func *tf = container_of(kref, struct dc_transfer_func, refcount); - kfree(tf); + kvfree(tf); } void dc_transfer_func_release(struct dc_transfer_func *tf) @@ -177,7 +177,7 @@ void dc_transfer_func_release(struct dc_transfer_func *tf) struct dc_transfer_func *dc_create_transfer_func(void) { - struct dc_transfer_func *tf = kzalloc(sizeof(*tf), GFP_KERNEL); + struct dc_transfer_func *tf = kvzalloc(sizeof(*tf), GFP_KERNEL); if (tf == NULL) goto alloc_fail;
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index ba1e507..99546a7 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -38,7 +38,7 @@ #include "inc/compressor.h" #include "dml/display_mode_lib.h" -#define DC_VER "3.1.27" +#define DC_VER "3.1.38" #define MAX_SURFACES 3 #define MAX_STREAMS 6 @@ -48,6 +48,18 @@ /******************************************************************************* * Display Core Interfaces ******************************************************************************/ +struct dmcu_version { + unsigned int date; + unsigned int month; + unsigned int year; + unsigned int interface_version; +}; + +struct dc_versions { + const char *dc_ver; + struct dmcu_version dmcu_version; +}; + struct dc_caps { uint32_t max_streams; uint32_t max_links; @@ -97,6 +109,7 @@ struct dc_surface_dcc_cap { }; struct dc_static_screen_events { + bool force_trigger; bool cursor_update; bool surface_update; bool overlay_update; @@ -173,6 +186,16 @@ enum wm_report_mode { WM_REPORT_OVERRIDE = 1, }; +struct dc_clocks { + int dispclk_khz; + int max_supported_dppclk_khz; + int dppclk_khz; + int dcfclk_khz; + int socclk_khz; + int dcfclk_deep_sleep_khz; + int fclk_khz; +}; + struct dc_debug { bool surface_visual_confirm; bool sanity_checks; @@ -214,11 +237,15 @@ struct dc_debug { bool disable_stereo_support; bool vsr_support; bool performance_trace; + bool az_endpoint_mute_only; + bool always_use_regamma; + bool p010_mpo_support; }; struct dc_state; struct resource_pool; struct dce_hwseq; struct dc { + struct dc_versions versions; struct dc_caps caps; struct dc_cap_funcs cap_funcs; struct dc_config config; @@ -255,6 +282,8 @@ struct dc { bool optimized_required; + bool apply_edp_fast_boot_optimization; + /* FBC compressor */ #if defined(CONFIG_DRM_AMD_DC_FBC) struct compressor *fbc_compressor; @@ -291,9 +320,6 @@ struct dc_init_data { struct dc_config flags; uint32_t log_mask; -#if defined(CONFIG_DRM_AMD_DC_FBC) - uint64_t fbc_gpu_addr; -#endif }; struct dc *dc_create(const struct dc_init_data *init_params); @@ -372,6 +398,8 @@ struct dc_transfer_func { struct dc_transfer_func_distributed_points tf_pts; enum dc_transfer_func_type type; enum dc_transfer_func_predefined tf; + /* FP16 1.0 reference level in nits, default is 80 nits, only for PQ*/ + uint32_t sdr_ref_white_level; struct dc_context *ctx; }; @@ -402,6 +430,8 @@ union surface_update_flags { uint32_t position_change:1; uint32_t in_transfer_func_change:1; uint32_t input_csc_change:1; + uint32_t output_tf_change:1; + uint32_t pixel_format_change:1; /* Full updates */ uint32_t new_plane:1; @@ -418,6 +448,7 @@ union surface_update_flags { struct dc_plane_state { struct dc_plane_address address; + struct dc_plane_flip_time time; struct scaling_taps scaling_quality; struct rect src_rect; struct rect dst_rect; @@ -433,6 +464,7 @@ struct dc_plane_state { struct dc_bias_and_scale *bias_and_scale; struct csc_transform input_csc_color_matrix; struct fixed31_32 coeff_reduction_factor; + uint32_t sdr_white_level; // TODO: No longer used, remove struct dc_hdr_static_metadata hdr_static_ctx; @@ -469,6 +501,7 @@ struct dc_plane_info { enum plane_stereo_format stereo_format; enum dc_color_space color_space; enum color_transfer_func input_tf; + unsigned int sdr_white_level; bool horizontal_mirror; bool visible; bool per_pixel_alpha; @@ -493,10 +526,8 @@ struct dc_surface_update { /* following updates require alloc/sleep/spin that is not isr safe, * null means no updates */ - /* gamma TO BE REMOVED */ struct dc_gamma *gamma; enum color_transfer_func color_input_tf; - enum color_transfer_func color_output_tf; struct dc_transfer_func *in_transfer_func; struct csc_transform *input_csc_color_matrix; @@ -528,6 +559,7 @@ struct dc_transfer_func *dc_create_transfer_func(void); */ struct dc_flip_addrs { struct dc_plane_address address; + unsigned int flip_timestamp_in_us; bool flip_immediate; /* TODO: add flip duration for FreeSync */ };
diff --git a/drivers/gpu/drm/amd/display/dc/dc_bios_types.h b/drivers/gpu/drm/amd/display/dc/dc_bios_types.h index 273d80a..d9b84ec 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_bios_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_bios_types.h
@@ -111,6 +111,8 @@ struct dc_vbios_funcs { struct dc_bios *bios); bool (*is_accelerated_mode)( struct dc_bios *bios); + uint32_t (*get_vga_enabled_displays)( + struct dc_bios *bios); void (*get_bios_event_info)( struct dc_bios *bios, struct bios_event_info *info); @@ -199,6 +201,7 @@ struct dc_vbios_funcs { }; struct bios_registers { + uint32_t BIOS_SCRATCH_3; uint32_t BIOS_SCRATCH_6; };
diff --git a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h index 03029f7..b83a7dc 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
@@ -413,12 +413,14 @@ struct dc_cursor_mi_param { enum { GAMMA_RGB_256_ENTRIES = 256, GAMMA_RGB_FLOAT_1024_ENTRIES = 1024, - GAMMA_MAX_ENTRIES = 1024 + GAMMA_CS_TFM_1D_ENTRIES = 4096, + GAMMA_MAX_ENTRIES = 4096 }; enum dc_gamma_type { GAMMA_RGB_256 = 1, - GAMMA_RGB_FLOAT_1024 = 2 + GAMMA_RGB_FLOAT_1024 = 2, + GAMMA_CS_TFM_1D = 3, }; struct dc_gamma { @@ -434,6 +436,8 @@ struct dc_gamma { /* private to DC core */ struct dc_context *ctx; + + bool is_identity; }; /* Used by both ipp amd opp functions*/ @@ -688,8 +692,18 @@ struct crtc_trigger_info { enum trigger_delay delay; }; -struct dc_crtc_timing { +enum vrr_state { + VRR_STATE_OFF = 0, + VRR_STATE_VARIABLE, + VRR_STATE_FIXED, +}; +struct dc_crtc_timing_adjust { + uint32_t v_total_min; + uint32_t v_total_max; +}; + +struct dc_crtc_timing { uint32_t h_total; uint32_t h_border_left; uint32_t h_addressable;
diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h index 6aacb0a..9fd7916 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_link.h +++ b/drivers/gpu/drm/amd/display/dc/dc_link.h
@@ -64,6 +64,8 @@ struct dc_link { enum signal_type connector_signal; enum dc_irq_source irq_source_hpd; enum dc_irq_source irq_source_hpd_rx;/* aka DP Short Pulse */ + bool is_hpd_filter_disabled; + /* caps is the same as reported_link_cap. link_traing use * reported_link_cap. Will clean up. TODO */ @@ -130,6 +132,8 @@ static inline struct dc_link *dc_get_link_at_index(struct dc *dc, uint32_t link_ bool dc_link_set_backlight_level(const struct dc_link *dc_link, uint32_t level, uint32_t frame_ramp, const struct dc_stream_state *stream); +bool dc_link_set_abm_disable(const struct dc_link *dc_link); + bool dc_link_set_psr_enable(const struct dc_link *dc_link, bool enable, bool wait); bool dc_link_get_psr_state(const struct dc_link *dc_link, uint32_t *psr_state); @@ -195,6 +199,8 @@ bool dc_link_dp_set_test_pattern( const unsigned char *p_custom_pattern, unsigned int cust_pattern_size); +void dc_link_enable_hpd_filter(struct dc_link *link, bool enable); + /* * DPCD access interfaces */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h index 7c3a577..d017df5 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stream.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -48,6 +48,8 @@ struct dc_stream_status { struct dc_stream_state { struct dc_sink *sink; struct dc_crtc_timing timing; + struct dc_crtc_timing_adjust timing_adjust; + struct vrr_params vrr_params; struct rect src; /* composition area */ struct rect dst; /* stream addressable area */ @@ -65,13 +67,19 @@ struct dc_stream_state { enum dc_dither_option dither_option; enum view_3d_format view_format; + enum color_transfer_func output_tf; bool ignore_msa_timing_param; /* TODO: custom INFO packets */ /* TODO: ABM info (DMCU) */ - /* TODO: PSR info */ + /* PSR info */ + unsigned char psr_version; /* TODO: CEA VIC */ + /* DMCU info */ + unsigned int abm_level; + unsigned int bl_pwm_level; + /* from core_stream struct */ struct dc_context *ctx; @@ -103,6 +111,8 @@ struct dc_stream_update { struct rect dst; struct dc_transfer_func *out_transfer_func; struct dc_hdr_static_metadata *hdr_static_metadata; + enum color_transfer_func color_output_tf; + unsigned int *abm_level; }; bool dc_is_stream_unchanged( @@ -152,7 +162,7 @@ struct dc_stream_state *dc_get_stream_at_index(struct dc *dc, uint8_t i); uint32_t dc_stream_get_vblank_counter(const struct dc_stream_state *stream); /* TODO: Return parsed values rather than direct register read - * This has a dependency on the caller (amdgpu_get_crtc_scanoutpos) + * This has a dependency on the caller (amdgpu_display_get_crtc_scanoutpos) * being refactored properly to be dce-specific */ bool dc_stream_get_scanoutpos(const struct dc_stream_state *stream, @@ -285,6 +295,9 @@ void dc_stream_set_static_screen_events(struct dc *dc, int num_streams, const struct dc_static_screen_events *events); +void dc_stream_set_dither_option(struct dc_stream_state *stream, + enum dc_dither_option option); + bool dc_stream_adjust_vmin_vmax(struct dc *dc, struct dc_stream_state **stream,
diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h index 9faddfa..9441305 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_types.h
@@ -193,6 +193,10 @@ union display_content_support { } bits; }; +struct dc_panel_patch { + unsigned int dppowerup_delay; +}; + struct dc_edid_caps { /* sink identification */ uint16_t manufacturer_id; @@ -219,6 +223,8 @@ struct dc_edid_caps { bool edid_hdmi; bool hdr_supported; + + struct dc_panel_patch panel_patch; }; struct view { @@ -515,6 +521,24 @@ struct audio_info { struct audio_mode modes[DC_MAX_AUDIO_DESC_COUNT]; }; +struct vrr_params { + enum vrr_state state; + uint32_t window_min; + uint32_t window_max; + uint32_t inserted_frame_duration_in_us; + uint32_t frames_to_insert; + uint32_t frame_counter; +}; + +#define DC_PLANE_UPDATE_TIMES_MAX 10 + +struct dc_plane_flip_time { + unsigned int time_elapsed_in_us[DC_PLANE_UPDATE_TIMES_MAX]; + unsigned int index; + unsigned int prev_update_time_in_us; +}; + +// Will combine with vrr_params at some point. struct freesync_context { bool supported; bool enabled;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c index 7a57a63..c83026fb 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
@@ -42,6 +42,8 @@ #define FN(reg_name, field_name) \ abm_dce->abm_shift->field_name, abm_dce->abm_mask->field_name +#define DC_LOGGER \ + abm->ctx->logger #define CTX \ abm_dce->base.ctx @@ -324,6 +326,15 @@ static bool dce_abm_immediate_disable(struct abm *abm) /* notifyDMCUMsg */ REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1); + abm->stored_backlight_registers.BL_PWM_CNTL = + REG_READ(BL_PWM_CNTL); + abm->stored_backlight_registers.BL_PWM_CNTL2 = + REG_READ(BL_PWM_CNTL2); + abm->stored_backlight_registers.BL_PWM_PERIOD_CNTL = + REG_READ(BL_PWM_PERIOD_CNTL); + + REG_GET(LVTMA_PWRSEQ_REF_DIV, BL_PWM_REF_DIV, + &abm->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); return true; } @@ -395,8 +406,7 @@ static bool dce_abm_set_backlight_level( { struct dce_abm *abm_dce = TO_DCE_ABM(abm); - dm_logger_write(abm->ctx->logger, LOG_BACKLIGHT, - "New Backlight level: %d (0x%X)\n", + DC_LOG_BACKLIGHT("New Backlight level: %d (0x%X)\n", backlight_level, backlight_level); /* If DMCU is in reset state, DMCU is uninitialized */
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c index 0df9ecb..6d5cdcd 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
@@ -33,6 +33,8 @@ #define CTX \ aud->base.ctx +#define DC_LOGGER \ + aud->base.ctx->logger #define REG(reg)\ (aud->regs->reg) @@ -63,8 +65,7 @@ static void write_indirect_azalia_reg(struct audio *audio, REG_SET(AZALIA_F0_CODEC_ENDPOINT_DATA, 0, AZALIA_ENDPOINT_REG_DATA, reg_data); - dm_logger_write(CTX->logger, LOG_HW_AUDIO, - "AUDIO:write_indirect_azalia_reg: index: %u data: %u\n", + DC_LOG_HW_AUDIO("AUDIO:write_indirect_azalia_reg: index: %u data: %u\n", reg_index, reg_data); } @@ -81,8 +82,7 @@ static uint32_t read_indirect_azalia_reg(struct audio *audio, uint32_t reg_index /* AZALIA_F0_CODEC_ENDPOINT_DATA endpoint data */ value = REG_READ(AZALIA_F0_CODEC_ENDPOINT_DATA); - dm_logger_write(CTX->logger, LOG_HW_AUDIO, - "AUDIO:read_indirect_azalia_reg: index: %u data: %u\n", + DC_LOG_HW_AUDIO("AUDIO:read_indirect_azalia_reg: index: %u data: %u\n", reg_index, value); return value; @@ -359,10 +359,12 @@ void dce_aud_az_enable(struct audio *audio) AUDIO_ENABLED); AZ_REG_WRITE(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, value); - value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL); + set_reg_field_value(value, 0, + AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, + CLOCK_GATING_DISABLE); + AZ_REG_WRITE(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, value); - dm_logger_write(CTX->logger, LOG_HW_AUDIO, - "\n\t========= AUDIO:dce_aud_az_enable: index: %u data: 0x%x\n", + DC_LOG_HW_AUDIO("\n\t========= AUDIO:dce_aud_az_enable: index: %u data: 0x%x\n", audio->inst, value); } @@ -372,6 +374,10 @@ void dce_aud_az_disable(struct audio *audio) struct dce_audio *aud = DCE_AUD(audio); value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL); + set_reg_field_value(value, 1, + AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, + CLOCK_GATING_DISABLE); + AZ_REG_WRITE(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, value); set_reg_field_value(value, 0, AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, @@ -383,8 +389,7 @@ void dce_aud_az_disable(struct audio *audio) CLOCK_GATING_DISABLE); AZ_REG_WRITE(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, value); value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL); - dm_logger_write(CTX->logger, LOG_HW_AUDIO, - "\n\t========= AUDIO:dce_aud_az_disable: index: %u data: 0x%x\n", + DC_LOG_HW_AUDIO("\n\t========= AUDIO:dce_aud_az_disable: index: %u data: 0x%x\n", audio->inst, value); } @@ -716,6 +721,11 @@ void dce_aud_az_configure( DESCRIPTION17); AZ_REG_WRITE(AZALIA_F0_CODEC_PIN_CONTROL_SINK_INFO8, value); + value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL); + set_reg_field_value(value, 0, + AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, + CLOCK_GATING_DISABLE); + AZ_REG_WRITE(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, value); } /* @@ -783,8 +793,7 @@ void dce_aud_wall_dto_setup( crtc_info->calculated_pixel_clock, &clock_info); - dm_logger_write(audio->ctx->logger, LOG_HW_AUDIO,\ - "\n%s:Input::requested_pixel_clock = %d"\ + DC_LOG_HW_AUDIO("\n%s:Input::requested_pixel_clock = %d"\ "calculated_pixel_clock =%d\n"\ "audio_dto_module = %d audio_dto_phase =%d \n\n", __func__,\ crtc_info->requested_pixel_clock,\ @@ -897,6 +906,10 @@ void dce_aud_hw_init( REG_UPDATE_2(AZALIA_F0_CODEC_FUNCTION_PARAMETER_POWER_STATES, CLKSTOP, 1, EPSS, 1); + set_reg_field_value(value, 0, + AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, + CLOCK_GATING_DISABLE); + AZ_REG_WRITE(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, value); } static const struct audio_funcs funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c index 31280d2..0aa2cda 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
@@ -41,7 +41,8 @@ #define CTX \ clk_src->base.ctx - +#define DC_LOGGER \ + calc_pll_cs->ctx->logger #undef FN #define FN(reg_name, field_name) \ clk_src->cs_shift->field_name, clk_src->cs_mask->field_name @@ -288,7 +289,7 @@ static uint32_t calculate_pixel_clock_pll_dividers( uint32_t max_ref_divider; if (pll_settings->adjusted_pix_clk == 0) { - dm_logger_write(calc_pll_cs->ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "%s Bad requested pixel clock", __func__); return MAX_PLL_CALC_ERROR; } @@ -349,13 +350,13 @@ static uint32_t calculate_pixel_clock_pll_dividers( * ## SVS Wed 15 Jul 2009 */ if (min_post_divider > max_post_divider) { - dm_logger_write(calc_pll_cs->ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "%s Post divider range is invalid", __func__); return MAX_PLL_CALC_ERROR; } if (min_ref_divider > max_ref_divider) { - dm_logger_write(calc_pll_cs->ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "%s Reference divider range is invalid", __func__); return MAX_PLL_CALC_ERROR; } @@ -466,7 +467,7 @@ static uint32_t dce110_get_pix_clk_dividers_helper ( { uint32_t field = 0; uint32_t pll_calc_error = MAX_PLL_CALC_ERROR; - + struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll; /* Check if reference clock is external (not pcie/xtalin) * HW Dce80 spec: * 00 - PCIE_REFCLK, 01 - XTALIN, 02 - GENERICA, 03 - GENERICB @@ -493,7 +494,7 @@ static uint32_t dce110_get_pix_clk_dividers_helper ( if (!pll_adjust_pix_clk(clk_src, pix_clk_params, pll_settings)) { /* Should never happen, ASSERT and fill up values to be able * to continue. */ - dm_logger_write(clk_src->base.ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "%s: Failed to adjust pixel clock!!", __func__); pll_settings->actual_pix_clk = pix_clk_params->requested_pix_clk; @@ -556,11 +557,12 @@ static uint32_t dce110_get_pix_clk_dividers( struct pll_settings *pll_settings) { struct dce110_clk_src *clk_src = TO_DCE110_CLK_SRC(cs); + struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll; uint32_t pll_calc_error = MAX_PLL_CALC_ERROR; if (pix_clk_params == NULL || pll_settings == NULL || pix_clk_params->requested_pix_clk == 0) { - dm_logger_write(clk_src->base.ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "%s: Invalid parameters!!\n", __func__); return pll_calc_error; } @@ -908,19 +910,9 @@ static bool dce110_program_pix_clk( #if defined(CONFIG_DRM_AMD_DC_DCN1_0) if (IS_FPGA_MAXIMUS_DC(clock_source->ctx->dce_environment)) { unsigned int inst = pix_clk_params->controller_id - CONTROLLER_ID_D0; - unsigned dp_dto_ref_kHz = 600000; - /* DPREF clock from FPGA TODO: Does FPGA have this value? */ + unsigned dp_dto_ref_kHz = 700000; unsigned clock_kHz = pll_settings->actual_pix_clk; - /* For faster simulation, if mode pixe clock less than 290MHz, - * pixel clock can be hard coded to 290Mhz. For 4K mode, pixel clock - * is greater than 500Mhz, need real pixel clock - * clock_kHz = 290000; - */ - /* TODO: un-hardcode when we can set display clock properly*/ - /*clock_kHz = pix_clk_params->requested_pix_clk;*/ - clock_kHz = 290000; - /* Set DTO values: phase = target clock, modulo = reference clock */ REG_WRITE(PHASE[inst], clock_kHz); REG_WRITE(MODULO[inst], dp_dto_ref_kHz); @@ -1062,14 +1054,14 @@ static void get_ss_info_from_atombios( struct spread_spectrum_info *ss_info_cur; struct spread_spectrum_data *ss_data_cur; uint32_t i; - + struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll; if (ss_entries_num == NULL) { - dm_logger_write(clk_src->base.ctx->logger, LOG_SYNC, + DC_LOG_SYNC( "Invalid entry !!!\n"); return; } if (spread_spectrum_data == NULL) { - dm_logger_write(clk_src->base.ctx->logger, LOG_SYNC, + DC_LOG_SYNC( "Invalid array pointer!!!\n"); return; } @@ -1114,7 +1106,7 @@ static void get_ss_info_from_atombios( ++i, ++ss_info_cur, ++ss_data_cur) { if (ss_info_cur->type.STEP_AND_DELAY_INFO != false) { - dm_logger_write(clk_src->base.ctx->logger, LOG_SYNC, + DC_LOG_SYNC( "Invalid ATOMBIOS SS Table!!!\n"); goto out_free_data; } @@ -1124,9 +1116,9 @@ static void get_ss_info_from_atombios( if (as_signal == AS_SIGNAL_TYPE_HDMI && ss_info_cur->spread_spectrum_percentage > 6){ /* invalid input, do nothing */ - dm_logger_write(clk_src->base.ctx->logger, LOG_SYNC, + DC_LOG_SYNC( "Invalid SS percentage "); - dm_logger_write(clk_src->base.ctx->logger, LOG_SYNC, + DC_LOG_SYNC( "for HDMI in ATOMBIOS info Table!!!\n"); continue; } @@ -1238,12 +1230,12 @@ static bool calc_pll_max_vco_construct( if (init_data->num_fract_fb_divider_decimal_point == 0 || init_data->num_fract_fb_divider_decimal_point_precision > init_data->num_fract_fb_divider_decimal_point) { - dm_logger_write(calc_pll_cs->ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "The dec point num or precision is incorrect!"); return false; } if (init_data->num_fract_fb_divider_decimal_point_precision == 0) { - dm_logger_write(calc_pll_cs->ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "Incorrect fract feedback divider precision num!"); return false; }
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c index 88538b4..0bf9b7f 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
@@ -34,6 +34,7 @@ #include "dcn_calcs.h" #endif #include "core_types.h" +#include "dc_types.h" #define TO_DCE_CLOCKS(clocks)\ @@ -48,6 +49,8 @@ #define CTX \ clk_dce->base.ctx +#define DC_LOGGER \ + clk->ctx->logger /* Max clock values for each state indexed by "enum clocks_state": */ static const struct state_dependent_clocks dce80_max_clks_by_state[] = { @@ -291,8 +294,10 @@ static enum dm_pp_clocks_state dce_get_required_clocks_state( low_req_clk = i + 1; if (low_req_clk > clk->max_clks_state) { - dm_logger_write(clk->ctx->logger, LOG_WARNING, - "%s: clocks unsupported", __func__); + DC_LOG_WARNING("%s: clocks unsupported disp_clk %d pix_clk %d", + __func__, + req_clocks->display_clk_khz, + req_clocks->pixel_clk_khz); low_req_clk = DM_PP_CLOCKS_STATE_INVALID; } @@ -308,8 +313,7 @@ static bool dce_clock_set_min_clocks_state( if (clocks_state > clk->max_clks_state) { /*Requested state exceeds max supported state.*/ - dm_logger_write(clk->ctx->logger, LOG_WARNING, - "Requested state exceeds max supported state"); + DC_LOG_WARNING("Requested state exceeds max supported state"); return false; } else if (clocks_state == clk->cur_min_clks_state) { /*if we're trying to set the same state, we can just return @@ -417,9 +421,12 @@ static int dce112_set_clock( bp->funcs->set_dce_clock(bp, &dce_clk_params); - if (clk_dce->dfs_bypass_disp_clk != actual_clock) - dmcu->funcs->set_psr_wait_loop(dmcu, - actual_clock / 1000 / 7); + if (!IS_FPGA_MAXIMUS_DC(core_dc->ctx->dce_environment)) { + if (clk_dce->dfs_bypass_disp_clk != actual_clock) + dmcu->funcs->set_psr_wait_loop(dmcu, + actual_clock / 1000 / 7); + } + clk_dce->dfs_bypass_disp_clk = actual_clock; return actual_clock; } @@ -477,7 +484,7 @@ static void dce_clock_read_integrated_info(struct dce_disp_clk *clk_dce) info.disp_clk_voltage[i].max_supported_clk; } - if (!debug->disable_dfs_bypass) + if (!debug->disable_dfs_bypass && bp->integrated_info) if (bp->integrated_info->gpu_cap_info & DFS_BYPASS_ENABLE) clk_dce->dfs_bypass_enabled = true;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c index f663adb..2ee3d9b 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
@@ -360,7 +360,7 @@ static void dcn10_get_dmcu_version(struct dmcu *dmcu) dmcu->dmcu_version.year = ((REG_READ(DMCU_IRAM_RD_DATA) << 8) | REG_READ(DMCU_IRAM_RD_DATA)); dmcu->dmcu_version.month = REG_READ(DMCU_IRAM_RD_DATA); - dmcu->dmcu_version.day = REG_READ(DMCU_IRAM_RD_DATA); + dmcu->dmcu_version.date = REG_READ(DMCU_IRAM_RD_DATA); /* Disable write access to IRAM to allow dynamic sleep state */ REG_UPDATE_2(DMCU_RAM_ACCESS_CTRL, @@ -521,6 +521,9 @@ static void dcn10_dmcu_set_psr_enable(struct dmcu *dmcu, bool enable, bool wait) if (dmcu->dmcu_state != DMCU_RUNNING) return; + dcn10_get_dmcu_psr_state(dmcu, &psr_state); + if (psr_state == 0 && !enable) + return; /* waitDMCUReadyForCmd */ REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, dmcu_wait_reg_ready_interval,
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.h b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.h index 1d4546f..c24c0e5 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.h
@@ -46,6 +46,23 @@ SR(SMU_INTERRUPT_CONTROL), \ SR(DC_DMCU_SCRATCH) +#define DMCU_DCE80_REG_LIST() \ + SR(DMCU_CTRL), \ + SR(DMCU_STATUS), \ + SR(DMCU_RAM_ACCESS_CTRL), \ + SR(DMCU_IRAM_WR_CTRL), \ + SR(DMCU_IRAM_WR_DATA), \ + SR(MASTER_COMM_DATA_REG1), \ + SR(MASTER_COMM_DATA_REG2), \ + SR(MASTER_COMM_DATA_REG3), \ + SR(MASTER_COMM_CMD_REG), \ + SR(MASTER_COMM_CNTL_REG), \ + SR(DMCU_IRAM_RD_CTRL), \ + SR(DMCU_IRAM_RD_DATA), \ + SR(DMCU_INTERRUPT_TO_UC_EN_MASK), \ + SR(SMU_INTERRUPT_CONTROL), \ + SR(DC_DMCU_SCRATCH) + #define DMCU_DCE110_COMMON_REG_LIST() \ DMCU_COMMON_REG_LIST_DCE_BASE(), \ SR(DCI_MEM_PWR_STATUS) @@ -83,6 +100,24 @@ STATIC_SCREEN4_INT_TO_UC_EN, mask_sh), \ DMCU_SF(SMU_INTERRUPT_CONTROL, DC_SMU_INT_ENABLE, mask_sh) +#define DMCU_MASK_SH_LIST_DCE80(mask_sh) \ + DMCU_SF(DMCU_CTRL, \ + DMCU_ENABLE, mask_sh), \ + DMCU_SF(DMCU_STATUS, \ + UC_IN_STOP_MODE, mask_sh), \ + DMCU_SF(DMCU_STATUS, \ + UC_IN_RESET, mask_sh), \ + DMCU_SF(DMCU_RAM_ACCESS_CTRL, \ + IRAM_HOST_ACCESS_EN, mask_sh), \ + DMCU_SF(DMCU_RAM_ACCESS_CTRL, \ + IRAM_WR_ADDR_AUTO_INC, mask_sh), \ + DMCU_SF(DMCU_RAM_ACCESS_CTRL, \ + IRAM_RD_ADDR_AUTO_INC, mask_sh), \ + DMCU_SF(MASTER_COMM_CMD_REG, \ + MASTER_COMM_CMD_REG_BYTE0, mask_sh), \ + DMCU_SF(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, mask_sh), \ + DMCU_SF(SMU_INTERRUPT_CONTROL, DC_SMU_INT_ENABLE, mask_sh) + #define DMCU_MASK_SH_LIST_DCE110(mask_sh) \ DMCU_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(mask_sh), \ DMCU_SF(DCI_MEM_PWR_STATUS, \
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c index dba6958..0275d6d 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c
@@ -57,7 +57,7 @@ void dce_pipe_control_lock(struct dc *dc, pipe->stream_res.tg->funcs->is_blanked(pipe->stream_res.tg)) return; - val = REG_GET_4(BLND_V_UPDATE_LOCK[pipe->pipe_idx], + val = REG_GET_4(BLND_V_UPDATE_LOCK[pipe->stream_res.tg->inst], BLND_DCP_GRPH_V_UPDATE_LOCK, &dcp_grph, BLND_SCL_V_UPDATE_LOCK, &scl, BLND_BLND_V_UPDATE_LOCK, &blnd, @@ -68,19 +68,19 @@ void dce_pipe_control_lock(struct dc *dc, blnd = lock_val; update_lock_mode = lock_val; - REG_SET_2(BLND_V_UPDATE_LOCK[pipe->pipe_idx], val, + REG_SET_2(BLND_V_UPDATE_LOCK[pipe->stream_res.tg->inst], val, BLND_DCP_GRPH_V_UPDATE_LOCK, dcp_grph, BLND_SCL_V_UPDATE_LOCK, scl); if (hws->masks->BLND_BLND_V_UPDATE_LOCK != 0) - REG_SET_2(BLND_V_UPDATE_LOCK[pipe->pipe_idx], val, + REG_SET_2(BLND_V_UPDATE_LOCK[pipe->stream_res.tg->inst], val, BLND_BLND_V_UPDATE_LOCK, blnd, BLND_V_UPDATE_LOCK_MODE, update_lock_mode); if (hws->wa.blnd_crtc_trigger) { if (!lock) { - uint32_t value = REG_READ(CRTC_H_BLANK_START_END[pipe->pipe_idx]); - REG_WRITE(CRTC_H_BLANK_START_END[pipe->pipe_idx], value); + uint32_t value = REG_READ(CRTC_H_BLANK_START_END[pipe->stream_res.tg->inst]); + REG_WRITE(CRTC_H_BLANK_START_END[pipe->stream_res.tg->inst], value); } } }
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h index b003984..057b8af 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
@@ -140,42 +140,8 @@ BL_REG_LIST() #define HWSEQ_DCN_REG_LIST()\ - SRII(DCHUBP_CNTL, HUBP, 0), \ - SRII(DCHUBP_CNTL, HUBP, 1), \ - SRII(DCHUBP_CNTL, HUBP, 2), \ - SRII(DCHUBP_CNTL, HUBP, 3), \ - SRII(HUBP_CLK_CNTL, HUBP, 0), \ - SRII(HUBP_CLK_CNTL, HUBP, 1), \ - SRII(HUBP_CLK_CNTL, HUBP, 2), \ - SRII(HUBP_CLK_CNTL, HUBP, 3), \ - SRII(DPP_CONTROL, DPP_TOP, 0), \ - SRII(DPP_CONTROL, DPP_TOP, 1), \ - SRII(DPP_CONTROL, DPP_TOP, 2), \ - SRII(DPP_CONTROL, DPP_TOP, 3), \ - SRII(OPP_PIPE_CONTROL, OPP_PIPE, 0), \ - SRII(OPP_PIPE_CONTROL, OPP_PIPE, 1), \ - SRII(OPP_PIPE_CONTROL, OPP_PIPE, 2), \ - SRII(OPP_PIPE_CONTROL, OPP_PIPE, 3), \ SR(REFCLK_CNTL), \ - SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A),\ - SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A),\ - SR(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A),\ - SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B),\ - SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B),\ - SR(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B),\ - SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C),\ - SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C),\ - SR(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C),\ - SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D),\ - SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D),\ - SR(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D),\ - SR(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL),\ - SR(DCHUBBUB_ARB_DRAM_STATE_CNTL),\ - SR(DCHUBBUB_ARB_SAT_LEVEL),\ - SR(DCHUBBUB_ARB_DF_REQ_OUTSTAND),\ SR(DCHUBBUB_GLOBAL_TIMER_CNTL), \ - SR(DCHUBBUB_TEST_DEBUG_INDEX), \ - SR(DCHUBBUB_TEST_DEBUG_DATA), \ SR(DIO_MEM_PWR_CTRL), \ SR(DCCG_GATE_DISABLE_CNTL), \ SR(DCCG_GATE_DISABLE_CNTL2), \ @@ -195,22 +161,10 @@ MMHUB_SR(MC_VM_SYSTEM_APERTURE_LOW_ADDR),\ MMHUB_SR(MC_VM_SYSTEM_APERTURE_HIGH_ADDR) -#define HWSEQ_SR_WATERMARK_REG_LIST()\ - SR(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A),\ - SR(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A),\ - SR(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B),\ - SR(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B),\ - SR(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C),\ - SR(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C),\ - SR(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D),\ - SR(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D) - #define HWSEQ_DCN1_REG_LIST()\ HWSEQ_DCN_REG_LIST(), \ - HWSEQ_SR_WATERMARK_REG_LIST(), \ HWSEQ_PIXEL_RATE_REG_LIST(OTG), \ HWSEQ_PHYPLL_REG_LIST(OTG), \ - SR(DCHUBBUB_SDPIF_FB_TOP),\ SR(DCHUBBUB_SDPIF_FB_BASE),\ SR(DCHUBBUB_SDPIF_FB_OFFSET),\ SR(DCHUBBUB_SDPIF_AGP_BASE),\ @@ -261,39 +215,9 @@ struct dce_hwseq_registers { uint32_t DCHUB_AGP_BOT; uint32_t DCHUB_AGP_TOP; - uint32_t DCHUBP_CNTL[4]; - uint32_t HUBP_CLK_CNTL[4]; - uint32_t DPP_CONTROL[4]; - uint32_t OPP_PIPE_CONTROL[4]; uint32_t REFCLK_CNTL; - uint32_t DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A; - uint32_t DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A; - uint32_t DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A; - uint32_t DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A; - uint32_t DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A; - uint32_t DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B; - uint32_t DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B; - uint32_t DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B; - uint32_t DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B; - uint32_t DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B; - uint32_t DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C; - uint32_t DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C; - uint32_t DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C; - uint32_t DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C; - uint32_t DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C; - uint32_t DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D; - uint32_t DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D; - uint32_t DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D; - uint32_t DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D; - uint32_t DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D; - uint32_t DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL; - uint32_t DCHUBBUB_ARB_SAT_LEVEL; - uint32_t DCHUBBUB_ARB_DF_REQ_OUTSTAND; + uint32_t DCHUBBUB_GLOBAL_TIMER_CNTL; - uint32_t DCHUBBUB_ARB_DRAM_STATE_CNTL; - uint32_t DCHUBBUB_TEST_DEBUG_INDEX; - uint32_t DCHUBBUB_TEST_DEBUG_DATA; - uint32_t DCHUBBUB_SDPIF_FB_TOP; uint32_t DCHUBBUB_SDPIF_FB_BASE; uint32_t DCHUBBUB_SDPIF_FB_OFFSET; uint32_t DCHUBBUB_SDPIF_AGP_BASE; @@ -405,6 +329,8 @@ struct dce_hwseq_registers { HWSEQ_DCE10_MASK_SH_LIST(mask_sh),\ SF(DCFEV_CLOCK_CONTROL, DCFEV_CLOCK_ENABLE, mask_sh),\ HWS_SF(, LVTMA_PWRSEQ_CNTL, LVTMA_BLON, mask_sh),\ + HWS_SF(, LVTMA_PWRSEQ_CNTL, LVTMA_DIGON, mask_sh),\ + HWS_SF(, LVTMA_PWRSEQ_CNTL, LVTMA_DIGON_OVRD, mask_sh),\ HWS_SF(, LVTMA_PWRSEQ_STATE, LVTMA_PWRSEQ_TARGET_STATE_R, mask_sh),\ HWSEQ_PIXEL_RATE_MASK_SH_LIST(mask_sh, CRTC0_) @@ -435,31 +361,17 @@ struct dce_hwseq_registers { #define HWSEQ_DCN_MASK_SH_LIST(mask_sh)\ HWSEQ_PIXEL_RATE_MASK_SH_LIST(mask_sh, OTG0_),\ HWS_SF1(OTG0_, PHYPLL_PIXEL_RATE_CNTL, PHYPLL_PIXEL_RATE_SOURCE, mask_sh), \ - HWS_SF(HUBP0_, DCHUBP_CNTL, HUBP_VTG_SEL, mask_sh), \ - HWS_SF(HUBP0_, HUBP_CLK_CNTL, HUBP_CLOCK_ENABLE, mask_sh), \ - HWS_SF(DPP_TOP0_, DPP_CONTROL, DPP_CLOCK_ENABLE, mask_sh), \ - HWS_SF(OPP_PIPE0_, OPP_PIPE_CONTROL, OPP_PIPE_CLOCK_EN, mask_sh),\ HWS_SF(, DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_ENABLE, mask_sh), \ - HWS_SF(, DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, mask_sh), \ - HWS_SF(, DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, DCHUBBUB_ARB_WATERMARK_CHANGE_DONE_INTERRUPT_DISABLE, mask_sh), \ - HWS_SF(, DCHUBBUB_ARB_DRAM_STATE_CNTL, DCHUBBUB_ARB_ALLOW_SELF_REFRESH_FORCE_VALUE, mask_sh), \ - HWS_SF(, DCHUBBUB_ARB_DRAM_STATE_CNTL, DCHUBBUB_ARB_ALLOW_SELF_REFRESH_FORCE_ENABLE, mask_sh), \ - HWS_SF(, DCHUBBUB_ARB_DRAM_STATE_CNTL, DCHUBBUB_ARB_ALLOW_PSTATE_CHANGE_FORCE_VALUE, mask_sh), \ - HWS_SF(, DCHUBBUB_ARB_DRAM_STATE_CNTL, DCHUBBUB_ARB_ALLOW_PSTATE_CHANGE_FORCE_ENABLE, mask_sh), \ - HWS_SF(, DCHUBBUB_ARB_SAT_LEVEL, DCHUBBUB_ARB_SAT_LEVEL, mask_sh), \ - HWS_SF(, DCHUBBUB_ARB_DF_REQ_OUTSTAND, DCHUBBUB_ARB_MIN_REQ_OUTSTAND, mask_sh), \ HWS_SF(, DCFCLK_CNTL, DCFCLK_GATE_DIS, mask_sh) #define HWSEQ_DCN1_MASK_SH_LIST(mask_sh)\ HWSEQ_DCN_MASK_SH_LIST(mask_sh), \ HWS_SF1(OTG0_, PHYPLL_PIXEL_RATE_CNTL, PIXEL_RATE_PLL_SOURCE, mask_sh), \ - HWS_SF(, DCHUBBUB_SDPIF_FB_TOP, SDPIF_FB_TOP, mask_sh), \ HWS_SF(, DCHUBBUB_SDPIF_FB_BASE, SDPIF_FB_BASE, mask_sh), \ HWS_SF(, DCHUBBUB_SDPIF_FB_OFFSET, SDPIF_FB_OFFSET, mask_sh), \ HWS_SF(, DCHUBBUB_SDPIF_AGP_BASE, SDPIF_AGP_BASE, mask_sh), \ HWS_SF(, DCHUBBUB_SDPIF_AGP_BOT, SDPIF_AGP_BOT, mask_sh), \ HWS_SF(, DCHUBBUB_SDPIF_AGP_TOP, SDPIF_AGP_TOP, mask_sh), \ - HWS_SF(DPP_TOP0_, DPP_CONTROL, DPPCLK_RATE_CONTROL, mask_sh), \ /* todo: get these from GVM instead of reading registers ourselves */\ HWS_SF(, VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32, PAGE_DIRECTORY_ENTRY_HI32, mask_sh),\ HWS_SF(, VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32, PAGE_DIRECTORY_ENTRY_LO32, mask_sh),\ @@ -502,6 +414,8 @@ struct dce_hwseq_registers { HWS_SF(, VGA_TEST_CONTROL, VGA_TEST_ENABLE, mask_sh),\ HWS_SF(, VGA_TEST_CONTROL, VGA_TEST_RENDER_START, mask_sh),\ HWS_SF(, LVTMA_PWRSEQ_CNTL, LVTMA_BLON, mask_sh), \ + HWS_SF(, LVTMA_PWRSEQ_CNTL, LVTMA_DIGON, mask_sh), \ + HWS_SF(, LVTMA_PWRSEQ_CNTL, LVTMA_DIGON_OVRD, mask_sh), \ HWS_SF(, LVTMA_PWRSEQ_STATE, LVTMA_PWRSEQ_TARGET_STATE_R, mask_sh) #define HWSEQ_REG_FIELD_LIST(type) \ @@ -534,14 +448,14 @@ struct dce_hwseq_registers { type ENABLE_L1_TLB;\ type SYSTEM_ACCESS_MODE;\ type LVTMA_BLON;\ - type LVTMA_PWRSEQ_TARGET_STATE_R; + type LVTMA_PWRSEQ_TARGET_STATE_R;\ + type LVTMA_DIGON;\ + type LVTMA_DIGON_OVRD; #define HWSEQ_DCN_REG_FIELD_LIST(type) \ type HUBP_VTG_SEL; \ type HUBP_CLOCK_ENABLE; \ type DPP_CLOCK_ENABLE; \ - type DPPCLK_RATE_CONTROL; \ - type SDPIF_FB_TOP;\ type SDPIF_FB_BASE;\ type SDPIF_FB_OFFSET;\ type SDPIF_AGP_BASE;\ @@ -554,14 +468,6 @@ struct dce_hwseq_registers { type AGP_BOT;\ type AGP_TOP;\ type DCHUBBUB_GLOBAL_TIMER_ENABLE; \ - type DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST;\ - type DCHUBBUB_ARB_WATERMARK_CHANGE_DONE_INTERRUPT_DISABLE;\ - type DCHUBBUB_ARB_ALLOW_SELF_REFRESH_FORCE_VALUE;\ - type DCHUBBUB_ARB_ALLOW_SELF_REFRESH_FORCE_ENABLE;\ - type DCHUBBUB_ARB_ALLOW_PSTATE_CHANGE_FORCE_VALUE;\ - type DCHUBBUB_ARB_ALLOW_PSTATE_CHANGE_FORCE_ENABLE;\ - type DCHUBBUB_ARB_SAT_LEVEL;\ - type DCHUBBUB_ARB_MIN_REQ_OUTSTAND;\ type OPP_PIPE_CLOCK_EN;\ type IP_REQUEST_EN; \ type DOMAIN0_POWER_FORCEON; \
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c index a4b09dd..daba1ba 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c
@@ -56,6 +56,8 @@ #define CTX \ enc110->base.ctx +#define DC_LOGGER \ + enc110->base.ctx->logger #define REG(reg)\ (enc110->link_regs->reg) @@ -769,8 +771,7 @@ void dce110_link_encoder_construct( bp_cap_info.DP_HBR3_EN; enc110->base.features.flags.bits.HDMI_6GB_EN = bp_cap_info.HDMI_6GB_EN; } else { - dm_logger_write(enc110->base.ctx->logger, LOG_WARNING, - "%s: Failed to get encoder_cap_info from VBIOS with error code %d!\n", + DC_LOG_WARNING("%s: Failed to get encoder_cap_info from VBIOS with error code %d!\n", __func__, result); } @@ -823,7 +824,6 @@ void dce110_link_encoder_hw_init( struct link_encoder *enc) { struct dce110_link_encoder *enc110 = TO_DCE110_LINK_ENC(enc); - struct dc_context *ctx = enc110->base.ctx; struct bp_transmitter_control cntl = { 0 }; enum bp_result result; @@ -835,11 +835,13 @@ void dce110_link_encoder_hw_init( cntl.coherent = false; cntl.hpd_sel = enc110->base.hpd_source; + if (enc110->base.connector.id == CONNECTOR_ID_EDP) + cntl.signal = SIGNAL_TYPE_EDP; + result = link_transmitter_control(enc110, &cntl); if (result != BP_RESULT_OK) { - dm_logger_write(ctx->logger, LOG_ERROR, - "%s: Failed to execute VBIOS command table!\n", + DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n", __func__); BREAK_TO_DEBUGGER(); return; @@ -915,7 +917,6 @@ void dce110_link_encoder_enable_tmds_output( uint32_t pixel_clock) { struct dce110_link_encoder *enc110 = TO_DCE110_LINK_ENC(enc); - struct dc_context *ctx = enc110->base.ctx; struct bp_transmitter_control cntl = { 0 }; enum bp_result result; @@ -939,8 +940,7 @@ void dce110_link_encoder_enable_tmds_output( result = link_transmitter_control(enc110, &cntl); if (result != BP_RESULT_OK) { - dm_logger_write(ctx->logger, LOG_ERROR, - "%s: Failed to execute VBIOS command table!\n", + DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n", __func__); BREAK_TO_DEBUGGER(); } @@ -953,7 +953,6 @@ void dce110_link_encoder_enable_dp_output( enum clock_source_id clock_source) { struct dce110_link_encoder *enc110 = TO_DCE110_LINK_ENC(enc); - struct dc_context *ctx = enc110->base.ctx; struct bp_transmitter_control cntl = { 0 }; enum bp_result result; @@ -980,8 +979,7 @@ void dce110_link_encoder_enable_dp_output( result = link_transmitter_control(enc110, &cntl); if (result != BP_RESULT_OK) { - dm_logger_write(ctx->logger, LOG_ERROR, - "%s: Failed to execute VBIOS command table!\n", + DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n", __func__); BREAK_TO_DEBUGGER(); } @@ -994,7 +992,6 @@ void dce110_link_encoder_enable_dp_mst_output( enum clock_source_id clock_source) { struct dce110_link_encoder *enc110 = TO_DCE110_LINK_ENC(enc); - struct dc_context *ctx = enc110->base.ctx; struct bp_transmitter_control cntl = { 0 }; enum bp_result result; @@ -1021,8 +1018,7 @@ void dce110_link_encoder_enable_dp_mst_output( result = link_transmitter_control(enc110, &cntl); if (result != BP_RESULT_OK) { - dm_logger_write(ctx->logger, LOG_ERROR, - "%s: Failed to execute VBIOS command table!\n", + DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n", __func__); BREAK_TO_DEBUGGER(); } @@ -1036,7 +1032,6 @@ void dce110_link_encoder_disable_output( enum signal_type signal) { struct dce110_link_encoder *enc110 = TO_DCE110_LINK_ENC(enc); - struct dc_context *ctx = enc110->base.ctx; struct bp_transmitter_control cntl = { 0 }; enum bp_result result; @@ -1064,8 +1059,7 @@ void dce110_link_encoder_disable_output( result = link_transmitter_control(enc110, &cntl); if (result != BP_RESULT_OK) { - dm_logger_write(ctx->logger, LOG_ERROR, - "%s: Failed to execute VBIOS command table!\n", + DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n", __func__); BREAK_TO_DEBUGGER(); return;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c index 83bae20..162f6a6 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -26,7 +26,8 @@ #include "dc_bios_types.h" #include "dce_stream_encoder.h" #include "reg_helper.h" - +#define DC_LOGGER \ + enc110->base.ctx->logger enum DP_PIXEL_ENCODING { DP_PIXEL_ENCODING_RGB444 = 0x00000000, DP_PIXEL_ENCODING_YCBCR422 = 0x00000001, @@ -197,7 +198,6 @@ static void dce110_update_hdmi_info_packet( uint32_t packet_index, const struct encoder_info_packet *info_packet) { - struct dc_context *ctx = enc110->base.ctx; uint32_t cont, send, line; if (info_packet->valid) { @@ -277,8 +277,7 @@ static void dce110_update_hdmi_info_packet( #endif default: /* invalid HW packet index */ - dm_logger_write( - ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "Invalid HW packet index: %s()\n", __func__); return; @@ -736,6 +735,8 @@ static void dce110_stream_encoder_update_hdmi_info_packets( if (info_frame->avi.valid) { const uint32_t *content = (const uint32_t *) &info_frame->avi.sb[0]; + /*we need turn on clock before programming AFMT block*/ + REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1); REG_WRITE(AFMT_AVI_INFO0, content[0]); @@ -920,6 +921,7 @@ static void dce110_stream_encoder_dp_blank( { struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc); uint32_t retries = 0; + uint32_t reg1 = 0; uint32_t max_retries = DP_BLANK_MAX_RETRY * 10; /* Note: For CZ, we are changing driver default to disable @@ -928,7 +930,10 @@ static void dce110_stream_encoder_dp_blank( * handful of panels that cannot handle disable stream at * HBLANK and will result in a white line flash across the * screen on stream disable. */ - + REG_GET(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, ®1); + if ((reg1 & 0x1) == 0) + /*stream not enabled*/ + return; /* Specify the video stream disable point * (2 = start of the next vertical blank) */ REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_DIS_DEFER, 2); @@ -1382,7 +1387,7 @@ static void dce110_se_setup_hdmi_audio( crtc_info->requested_pixel_clock, crtc_info->calculated_pixel_clock, &audio_clock_info); - dm_logger_write(enc->ctx->logger, LOG_HW_AUDIO, + DC_LOG_HW_AUDIO( "\n%s:Input::requested_pixel_clock = %d" \ "calculated_pixel_clock = %d \n", __func__, \ crtc_info->requested_pixel_clock, \
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c index ac28113..832c5daa 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c
@@ -38,6 +38,8 @@ #define CTX \ xfm_dce->base.ctx +#define DC_LOGGER \ + xfm_dce->base.ctx->logger #define IDENTITY_RATIO(ratio) (dal_fixed31_32_u2d19(ratio) == (1 << 19)) #define GAMUT_MATRIX_SIZE 12 @@ -618,80 +620,48 @@ static void program_bit_depth_reduction( enum dc_color_depth depth, const struct bit_depth_reduction_params *bit_depth_params) { - enum dcp_bit_depth_reduction_mode depth_reduction_mode; - enum dcp_spatial_dither_mode spatial_dither_mode; - bool frame_random_enable; - bool rgb_random_enable; - bool highpass_random_enable; + enum dcp_out_trunc_round_depth trunc_round_depth; + enum dcp_out_trunc_round_mode trunc_mode; + bool spatial_dither_enable; ASSERT(depth < COLOR_DEPTH_121212); /* Invalid clamp bit depth */ - if (bit_depth_params->flags.SPATIAL_DITHER_ENABLED) { - depth_reduction_mode = DCP_BIT_DEPTH_REDUCTION_MODE_DITHER; - frame_random_enable = true; - rgb_random_enable = true; - highpass_random_enable = true; + spatial_dither_enable = bit_depth_params->flags.SPATIAL_DITHER_ENABLED; + /* Default to 12 bit truncation without rounding */ + trunc_round_depth = DCP_OUT_TRUNC_ROUND_DEPTH_12BIT; + trunc_mode = DCP_OUT_TRUNC_ROUND_MODE_TRUNCATE; - } else { - depth_reduction_mode = DCP_BIT_DEPTH_REDUCTION_MODE_DISABLED; - frame_random_enable = false; - rgb_random_enable = false; - highpass_random_enable = false; + if (bit_depth_params->flags.TRUNCATE_ENABLED) { + /* Don't enable dithering if truncation is enabled */ + spatial_dither_enable = false; + trunc_mode = bit_depth_params->flags.TRUNCATE_MODE ? + DCP_OUT_TRUNC_ROUND_MODE_ROUND : + DCP_OUT_TRUNC_ROUND_MODE_TRUNCATE; + + if (bit_depth_params->flags.TRUNCATE_DEPTH == 0 || + bit_depth_params->flags.TRUNCATE_DEPTH == 1) + trunc_round_depth = DCP_OUT_TRUNC_ROUND_DEPTH_8BIT; + else if (bit_depth_params->flags.TRUNCATE_DEPTH == 2) + trunc_round_depth = DCP_OUT_TRUNC_ROUND_DEPTH_10BIT; + else { + /* + * Invalid truncate/round depth. Setting here to 12bit + * to prevent use-before-initialize errors. + */ + trunc_round_depth = DCP_OUT_TRUNC_ROUND_DEPTH_12BIT; + BREAK_TO_DEBUGGER(); + } } - spatial_dither_mode = DCP_SPATIAL_DITHER_MODE_A_AA_A; - set_clamp(xfm_dce, depth); - - switch (depth_reduction_mode) { - case DCP_BIT_DEPTH_REDUCTION_MODE_DITHER: - /* Spatial Dither: Set round/truncate to bypass (12bit), - * enable Dither (30bpp) */ - set_round(xfm_dce, - DCP_OUT_TRUNC_ROUND_MODE_TRUNCATE, - DCP_OUT_TRUNC_ROUND_DEPTH_12BIT); - - set_dither(xfm_dce, true, spatial_dither_mode, - DCP_SPATIAL_DITHER_DEPTH_30BPP, frame_random_enable, - rgb_random_enable, highpass_random_enable); - break; - case DCP_BIT_DEPTH_REDUCTION_MODE_ROUND: - /* Round: Enable round (10bit), disable Dither */ - set_round(xfm_dce, - DCP_OUT_TRUNC_ROUND_MODE_ROUND, - DCP_OUT_TRUNC_ROUND_DEPTH_10BIT); - - set_dither(xfm_dce, false, spatial_dither_mode, - DCP_SPATIAL_DITHER_DEPTH_30BPP, frame_random_enable, - rgb_random_enable, highpass_random_enable); - break; - case DCP_BIT_DEPTH_REDUCTION_MODE_TRUNCATE: /* Truncate */ - /* Truncate: Enable truncate (10bit), disable Dither */ - set_round(xfm_dce, - DCP_OUT_TRUNC_ROUND_MODE_TRUNCATE, - DCP_OUT_TRUNC_ROUND_DEPTH_10BIT); - - set_dither(xfm_dce, false, spatial_dither_mode, - DCP_SPATIAL_DITHER_DEPTH_30BPP, frame_random_enable, - rgb_random_enable, highpass_random_enable); - break; - - case DCP_BIT_DEPTH_REDUCTION_MODE_DISABLED: /* Disabled */ - /* Truncate: Set round/truncate to bypass (12bit), - * disable Dither */ - set_round(xfm_dce, - DCP_OUT_TRUNC_ROUND_MODE_TRUNCATE, - DCP_OUT_TRUNC_ROUND_DEPTH_12BIT); - - set_dither(xfm_dce, false, spatial_dither_mode, - DCP_SPATIAL_DITHER_DEPTH_30BPP, frame_random_enable, - rgb_random_enable, highpass_random_enable); - break; - default: - /* Invalid DCP Depth reduction mode */ - BREAK_TO_DEBUGGER(); - break; - } + set_round(xfm_dce, trunc_mode, trunc_round_depth); + set_dither(xfm_dce, + spatial_dither_enable, + DCP_SPATIAL_DITHER_MODE_A_AA_A, + DCP_SPATIAL_DITHER_DEPTH_30BPP, + bit_depth_params->flags.FRAME_RANDOM, + bit_depth_params->flags.RGB_RANDOM, + bit_depth_params->flags.HIGHPASS_RANDOM); } static int dce_transform_get_max_num_of_supported_lines( @@ -725,8 +695,7 @@ static int dce_transform_get_max_num_of_supported_lines( break; default: - dm_logger_write(xfm_dce->base.ctx->logger, LOG_WARNING, - "%s: Invalid LB pixel depth", + DC_LOG_WARNING("%s: Invalid LB pixel depth", __func__); BREAK_TO_DEBUGGER(); break; @@ -823,8 +792,7 @@ static void dce_transform_set_pixel_storage_depth( if (!(xfm_dce->lb_pixel_depth_supported & depth)) { /*we should use unsupported capabilities * unless it is required by w/a*/ - dm_logger_write(xfm->ctx->logger, LOG_WARNING, - "%s: Capability not supported", + DC_LOG_WARNING("%s: Capability not supported", __func__); } } @@ -879,6 +847,7 @@ static void dce_transform_set_gamut_remap( const struct xfm_grph_csc_adjustment *adjust) { struct dce_transform *xfm_dce = TO_DCE_TRANSFORM(xfm); + int i = 0; if (adjust->gamut_adjust_type != GRAPHICS_GAMUT_ADJUST_TYPE_SW) /* Bypass if type is bypass or hw */ @@ -887,20 +856,8 @@ static void dce_transform_set_gamut_remap( struct fixed31_32 arr_matrix[GAMUT_MATRIX_SIZE]; uint16_t arr_reg_val[GAMUT_MATRIX_SIZE]; - arr_matrix[0] = adjust->temperature_matrix[0]; - arr_matrix[1] = adjust->temperature_matrix[1]; - arr_matrix[2] = adjust->temperature_matrix[2]; - arr_matrix[3] = dal_fixed31_32_zero; - - arr_matrix[4] = adjust->temperature_matrix[3]; - arr_matrix[5] = adjust->temperature_matrix[4]; - arr_matrix[6] = adjust->temperature_matrix[5]; - arr_matrix[7] = dal_fixed31_32_zero; - - arr_matrix[8] = adjust->temperature_matrix[6]; - arr_matrix[9] = adjust->temperature_matrix[7]; - arr_matrix[10] = adjust->temperature_matrix[8]; - arr_matrix[11] = dal_fixed31_32_zero; + for (i = 0; i < GAMUT_MATRIX_SIZE; i++) + arr_matrix[i] = adjust->temperature_matrix[i]; convert_float_matrix( arr_reg_val, arr_matrix, GAMUT_MATRIX_SIZE); @@ -1215,8 +1172,7 @@ static void program_pwl(struct dce_transform *xfm_dce, } if (counter == max_tries) { - dm_logger_write(xfm_dce->base.ctx->logger, LOG_WARNING, - "%s: regamma lut was not powered on " + DC_LOG_WARNING("%s: regamma lut was not powered on " "in a timely manner," " programming still proceeds\n", __func__);
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.h b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.h index bfc94b4..948281d 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.h
@@ -248,6 +248,7 @@ XFM_SF(DCP0_REGAMMA_CNTLA_REGION_0_1, REGAMMA_CNTLA_EXP_REGION1_LUT_OFFSET, mask_sh),\ XFM_SF(DCP0_REGAMMA_CNTLA_REGION_0_1, REGAMMA_CNTLA_EXP_REGION1_NUM_SEGMENTS, mask_sh),\ XFM_SF(DCP0_REGAMMA_CONTROL, GRPH_REGAMMA_MODE, mask_sh),\ + XFM_SF(DCP0_REGAMMA_LUT_WRITE_EN_MASK, REGAMMA_LUT_WRITE_EN_MASK, mask_sh),\ XFM_SF(SCL0_SCL_MODE, SCL_MODE, mask_sh), \ XFM_SF(SCL0_SCL_TAP_CONTROL, SCL_H_NUM_OF_TAPS, mask_sh), \ XFM_SF(SCL0_SCL_TAP_CONTROL, SCL_V_NUM_OF_TAPS, mask_sh), \
diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c index 469af05..41f83ec 100644 --- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c
@@ -69,7 +69,7 @@ static const struct dce100_hw_seq_reg_offsets reg_offsets[] = { ******************************************************************************/ /***************************PIPE_CONTROL***********************************/ -static bool dce100_enable_display_power_gating( +bool dce100_enable_display_power_gating( struct dc *dc, uint8_t controller_id, struct dc_bios *dcb,
diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.h index cb5384e..c6ec0ed 100644 --- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.h
@@ -38,5 +38,9 @@ void dce100_set_bandwidth( struct dc_state *context, bool decrease_allowed); +bool dce100_enable_display_power_gating(struct dc *dc, uint8_t controller_id, + struct dc_bios *dcb, + enum pipe_gating_control power_gating); + #endif /* __DC_HWSS_DCE100_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c index 442dd2d..3092f76 100644 --- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
@@ -51,6 +51,9 @@ #include "dce/dce_10_0_d.h" #include "dce/dce_10_0_sh_mask.h" +#include "dce/dce_dmcu.h" +#include "dce/dce_abm.h" + #ifndef mmMC_HUB_RDREQ_DMIF_LIMIT #include "gmc/gmc_8_2_d.h" #include "gmc/gmc_8_2_sh_mask.h" @@ -320,7 +323,29 @@ static const struct dce110_clk_src_mask cs_mask = { CS_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK) }; +static const struct dce_dmcu_registers dmcu_regs = { + DMCU_DCE110_COMMON_REG_LIST() +}; +static const struct dce_dmcu_shift dmcu_shift = { + DMCU_MASK_SH_LIST_DCE110(__SHIFT) +}; + +static const struct dce_dmcu_mask dmcu_mask = { + DMCU_MASK_SH_LIST_DCE110(_MASK) +}; + +static const struct dce_abm_registers abm_regs = { + ABM_DCE110_COMMON_REG_LIST() +}; + +static const struct dce_abm_shift abm_shift = { + ABM_MASK_SH_LIST_DCE110(__SHIFT) +}; + +static const struct dce_abm_mask abm_mask = { + ABM_MASK_SH_LIST_DCE110(_MASK) +}; #define DCFE_MEM_PWR_CTRL_REG_BASE 0x1b03 @@ -622,6 +647,12 @@ static void destruct(struct dce110_resource_pool *pool) if (pool->base.display_clock != NULL) dce_disp_clk_destroy(&pool->base.display_clock); + if (pool->base.abm != NULL) + dce_abm_destroy(&pool->base.abm); + + if (pool->base.dmcu != NULL) + dce_dmcu_destroy(&pool->base.dmcu); + if (pool->base.irqs != NULL) dal_irq_service_destroy(&pool->base.irqs); } @@ -829,6 +860,25 @@ static bool construct( goto res_create_fail; } + pool->base.dmcu = dce_dmcu_create(ctx, + &dmcu_regs, + &dmcu_shift, + &dmcu_mask); + if (pool->base.dmcu == NULL) { + dm_error("DC: failed to create dmcu!\n"); + BREAK_TO_DEBUGGER(); + goto res_create_fail; + } + + pool->base.abm = dce_abm_create(ctx, + &abm_regs, + &abm_shift, + &abm_mask); + if (pool->base.abm == NULL) { + dm_error("DC: failed to create abm!\n"); + BREAK_TO_DEBUGGER(); + goto res_create_fail; + } /* get static clock information for PPLIB or firmware, save * max_clock_state @@ -849,6 +899,7 @@ static bool construct( *************************************************/ pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE; pool->base.pipe_count = res_cap.num_timing_generator; + pool->base.timing_generator_count = pool->base.res_cap->num_timing_generator; dc->caps.max_downscale_ratio = 200; dc->caps.i2c_speed_in_khz = 40; dc->caps.max_cursor_size = 128;
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c index 78bae5e..1fd0e96 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
@@ -34,6 +34,8 @@ #include "dce110_compressor.h" +#define DC_LOGGER \ + cp110->base.ctx->logger #define DCP_REG(reg)\ (reg + cp110->offsets.dcp_offset) #define DMIF_REG(reg)\ @@ -157,14 +159,10 @@ static void wait_for_fbc_state_changed( } if (counter == 1000) { - dm_logger_write( - cp110->base.ctx->logger, LOG_WARNING, - "%s: wait counter exceeded, changes to HW not applied", + DC_LOG_WARNING("%s: wait counter exceeded, changes to HW not applied", __func__); } else { - dm_logger_write( - cp110->base.ctx->logger, LOG_SYNC, - "FBC status changed to %d", enabled); + DC_LOG_SYNC("FBC status changed to %d", enabled); } @@ -351,9 +349,7 @@ void dce110_compressor_program_compressed_surface_address_and_pitch( if (compressor->min_compress_ratio == FBC_COMPRESS_RATIO_1TO1) fbc_pitch = fbc_pitch / 8; else - dm_logger_write( - compressor->ctx->logger, LOG_WARNING, - "%s: Unexpected DCE11 compression ratio", + DC_LOG_WARNING("%s: Unexpected DCE11 compression ratio", __func__); /* Clear content first. */
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c index 4142239..4c658f7 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -57,6 +57,8 @@ #include "dce/dce_11_0_sh_mask.h" #include "custom_float.h" +#include "atomfirmware.h" + /* * All values are in milliseconds; * For eDP, after power-up/power/down, @@ -68,6 +70,8 @@ #define CTX \ hws->ctx +#define DC_LOGGER \ + ctx->logger #define REG(reg)\ hws->regs->reg @@ -407,6 +411,10 @@ static bool convert_to_custom_float(struct pwl_result_data *rgb_resulted, return true; } +#define MAX_LOW_POINT 25 +#define NUMBER_REGIONS 16 +#define NUMBER_SW_SEGMENTS 16 + static bool dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf, struct pwl_params *regamma_params) @@ -421,8 +429,8 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf, struct fixed31_32 y1_min; struct fixed31_32 y3_max; - int32_t segment_start, segment_end; - uint32_t i, j, k, seg_distr[16], increment, start_index, hw_points; + int32_t region_start, region_end; + uint32_t i, j, k, seg_distr[NUMBER_REGIONS], increment, start_index, hw_points; if (output_tf == NULL || regamma_params == NULL || output_tf->type == TF_TYPE_BYPASS) return false; @@ -437,34 +445,23 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf, /* 16 segments * segments are from 2^-11 to 2^5 */ - segment_start = -11; - segment_end = 5; + region_start = -11; + region_end = region_start + NUMBER_REGIONS; - seg_distr[0] = 2; - seg_distr[1] = 2; - seg_distr[2] = 2; - seg_distr[3] = 2; - seg_distr[4] = 2; - seg_distr[5] = 2; - seg_distr[6] = 3; - seg_distr[7] = 4; - seg_distr[8] = 4; - seg_distr[9] = 4; - seg_distr[10] = 4; - seg_distr[11] = 5; - seg_distr[12] = 5; - seg_distr[13] = 5; - seg_distr[14] = 5; - seg_distr[15] = 5; + for (i = 0; i < NUMBER_REGIONS; i++) + seg_distr[i] = 4; } else { /* 10 segments - * segment is from 2^-10 to 2^0 + * segment is from 2^-10 to 2^1 + * We include an extra segment for range [2^0, 2^1). This is to + * ensure that colors with normalized values of 1 don't miss the + * LUT. */ - segment_start = -10; - segment_end = 0; + region_start = -10; + region_end = 1; - seg_distr[0] = 3; + seg_distr[0] = 4; seg_distr[1] = 4; seg_distr[2] = 4; seg_distr[3] = 4; @@ -472,9 +469,9 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf, seg_distr[5] = 4; seg_distr[6] = 4; seg_distr[7] = 4; - seg_distr[8] = 5; - seg_distr[9] = 5; - seg_distr[10] = -1; + seg_distr[8] = 4; + seg_distr[9] = 4; + seg_distr[10] = 0; seg_distr[11] = -1; seg_distr[12] = -1; seg_distr[13] = -1; @@ -488,10 +485,12 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf, } j = 0; - for (k = 0; k < (segment_end - segment_start); k++) { - increment = 32 / (1 << seg_distr[k]); - start_index = (segment_start + k + 25) * 32; - for (i = start_index; i < start_index + 32; i += increment) { + for (k = 0; k < (region_end - region_start); k++) { + increment = NUMBER_SW_SEGMENTS / (1 << seg_distr[k]); + start_index = (region_start + k + MAX_LOW_POINT) * + NUMBER_SW_SEGMENTS; + for (i = start_index; i < start_index + NUMBER_SW_SEGMENTS; + i += increment) { if (j == hw_points - 1) break; rgb_resulted[j].red = output_tf->tf_pts.red[i]; @@ -502,15 +501,15 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf, } /* last point */ - start_index = (segment_end + 25) * 32; + start_index = (region_end + MAX_LOW_POINT) * NUMBER_SW_SEGMENTS; rgb_resulted[hw_points - 1].red = output_tf->tf_pts.red[start_index]; rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index]; rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index]; arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2), - dal_fixed31_32_from_int(segment_start)); + dal_fixed31_32_from_int(region_start)); arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2), - dal_fixed31_32_from_int(segment_end)); + dal_fixed31_32_from_int(region_end)); y_r = rgb_resulted[0].red; y_g = rgb_resulted[0].green; @@ -625,7 +624,7 @@ static enum dc_status bios_parser_crtc_source_select( const struct dc_sink *sink = pipe_ctx->stream->sink; crtc_source_select.engine_id = pipe_ctx->stream_res.stream_enc->id; - crtc_source_select.controller_id = pipe_ctx->pipe_idx + 1; + crtc_source_select.controller_id = pipe_ctx->stream_res.tg->inst + 1; /*TODO: Need to un-hardcode color depth, dp_audio and account for * the case where signal and sink signal is different (translator * encoder)*/ @@ -741,10 +740,14 @@ static bool is_panel_backlight_on(struct dce_hwseq *hws) static bool is_panel_powered_on(struct dce_hwseq *hws) { - uint32_t value; + uint32_t pwr_seq_state, dig_on, dig_on_ovrd; - REG_GET(LVTMA_PWRSEQ_STATE, LVTMA_PWRSEQ_TARGET_STATE_R, &value); - return value == 1; + + REG_GET(LVTMA_PWRSEQ_STATE, LVTMA_PWRSEQ_TARGET_STATE_R, &pwr_seq_state); + + REG_GET_2(LVTMA_PWRSEQ_CNTL, LVTMA_DIGON, &dig_on, LVTMA_DIGON_OVRD, &dig_on_ovrd); + + return (pwr_seq_state == 1) || (dig_on == 1 && dig_on_ovrd == 1); } static enum bp_result link_transmitter_control( @@ -825,7 +828,7 @@ void hwss_edp_wait_for_hpd_ready( dal_gpio_destroy_irq(&hpd); if (false == edp_hpd_high) { - dm_logger_write(ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "%s: wait timed out!\n", __func__); } } @@ -849,7 +852,7 @@ void hwss_edp_power_control( if (power_up != is_panel_powered_on(hwseq)) { /* Send VBIOS command to prompt eDP panel power */ - dm_logger_write(ctx->logger, LOG_HW_RESUME_S3, + DC_LOG_HW_RESUME_S3( "%s: Panel Power action: %s\n", __func__, (power_up ? "On":"Off")); @@ -865,11 +868,11 @@ void hwss_edp_power_control( bp_result = link_transmitter_control(ctx->dc_bios, &cntl); if (bp_result != BP_RESULT_OK) - dm_logger_write(ctx->logger, LOG_ERROR, + DC_LOG_ERROR( "%s: Panel Power bp_result: %d\n", __func__, bp_result); } else { - dm_logger_write(ctx->logger, LOG_HW_RESUME_S3, + DC_LOG_HW_RESUME_S3( "%s: Skipping Panel Power action: %s\n", __func__, (power_up ? "On":"Off")); } @@ -895,7 +898,7 @@ void hwss_edp_backlight_control( } if (enable && is_panel_backlight_on(hws)) { - dm_logger_write(ctx->logger, LOG_HW_RESUME_S3, + DC_LOG_HW_RESUME_S3( "%s: panel already powered up. Do nothing.\n", __func__); return; @@ -903,7 +906,7 @@ void hwss_edp_backlight_control( /* Send VBIOS command to control eDP panel backlight */ - dm_logger_write(ctx->logger, LOG_HW_RESUME_S3, + DC_LOG_HW_RESUME_S3( "%s: backlight action: %s\n", __func__, (enable ? "On":"Off")); @@ -917,6 +920,7 @@ void hwss_edp_backlight_control( /*todo: unhardcode*/ cntl.lanes_number = LANE_COUNT_FOUR; cntl.hpd_sel = link->link_enc->hpd_source; + cntl.signal = SIGNAL_TYPE_EDP; /* For eDP, the following delays might need to be considered * after link training completed: @@ -929,7 +933,13 @@ void hwss_edp_backlight_control( * Enable it in the future if necessary. */ /* dc_service_sleep_in_milliseconds(50); */ + /*edp 1.2*/ + if (cntl.action == TRANSMITTER_CONTROL_BACKLIGHT_ON) + edp_receiver_ready_T7(link); link_transmitter_control(ctx->dc_bios, &cntl); + /*edp 1.2*/ + if (cntl.action == TRANSMITTER_CONTROL_BACKLIGHT_OFF) + edp_receiver_ready_T9(link); } void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option) @@ -949,7 +959,11 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option) pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control( pipe_ctx->stream_res.stream_enc, true); if (pipe_ctx->stream_res.audio) { - pipe_ctx->stream_res.audio->funcs->az_disable(pipe_ctx->stream_res.audio); + if (option != KEEP_ACQUIRED_RESOURCE || + !dc->debug.az_endpoint_mute_only) { + /*only disalbe az_endpoint if power down or free*/ + pipe_ctx->stream_res.audio->funcs->az_disable(pipe_ctx->stream_res.audio); + } if (dc_is_dp_signal(pipe_ctx->stream->signal)) pipe_ctx->stream_res.stream_enc->funcs->dp_audio_disable( @@ -972,9 +986,6 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option) */ } - /* blank at encoder level */ - if (dc_is_dp_signal(pipe_ctx->stream->signal)) - pipe_ctx->stream_res.stream_enc->funcs->dp_blank(pipe_ctx->stream_res.stream_enc); link->link_enc->funcs->connect_dig_be_to_fe( link->link_enc, @@ -987,12 +998,34 @@ void dce110_unblank_stream(struct pipe_ctx *pipe_ctx, struct dc_link_settings *link_settings) { struct encoder_unblank_param params = { { 0 } }; + struct dc_stream_state *stream = pipe_ctx->stream; + struct dc_link *link = stream->sink->link; /* only 3 items below are used by unblank */ params.pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_khz; params.link_settings.link_rate = link_settings->link_rate; - pipe_ctx->stream_res.stream_enc->funcs->dp_unblank(pipe_ctx->stream_res.stream_enc, ¶ms); + + if (dc_is_dp_signal(pipe_ctx->stream->signal)) + pipe_ctx->stream_res.stream_enc->funcs->dp_unblank(pipe_ctx->stream_res.stream_enc, ¶ms); + + if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP) { + link->dc->hwss.edp_backlight_control(link, true); + stream->bl_pwm_level = 0; + } +} +void dce110_blank_stream(struct pipe_ctx *pipe_ctx) +{ + struct dc_stream_state *stream = pipe_ctx->stream; + struct dc_link *link = stream->sink->link; + + if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP) { + link->dc->hwss.edp_backlight_control(link, false); + dc_link_set_abm_disable(link); + } + + if (dc_is_dp_signal(pipe_ctx->stream->signal)) + pipe_ctx->stream_res.stream_enc->funcs->dp_blank(pipe_ctx->stream_res.stream_enc); } @@ -1094,7 +1127,7 @@ static void build_audio_output( audio_output->pll_info.dto_source = translate_to_dto_source( - pipe_ctx->pipe_idx + 1); + pipe_ctx->stream_res.tg->inst + 1); /* TODO hard code to enable for now. Need get from stream */ audio_output->pll_info.ss_enabled = true; @@ -1106,7 +1139,7 @@ static void build_audio_output( static void get_surface_visual_confirm_color(const struct pipe_ctx *pipe_ctx, struct tg_color *color) { - uint32_t color_value = MAX_TG_COLOR_VALUE * (4 - pipe_ctx->pipe_idx) / 4; + uint32_t color_value = MAX_TG_COLOR_VALUE * (4 - pipe_ctx->stream_res.tg->inst) / 4; switch (pipe_ctx->plane_res.scl_data.format) { case PIXEL_FORMAT_ARGB8888: @@ -1303,10 +1336,8 @@ static enum dc_status apply_single_controller_ctx_to_hw( resource_build_info_frame(pipe_ctx); dce110_update_info_frame(pipe_ctx); - if (!pipe_ctx_old->stream) { - if (!pipe_ctx->stream->dpms_off) - core_link_enable_stream(context, pipe_ctx); - } + if (!pipe_ctx_old->stream) + core_link_enable_stream(context, pipe_ctx); pipe_ctx->plane_res.scl_data.lb_params.alpha_en = pipe_ctx->bottom_pipe != 0; @@ -1410,6 +1441,31 @@ static void disable_vga_and_power_gate_all_controllers( } } +static struct dc_link *get_link_for_edp_not_in_use( + struct dc *dc, + struct dc_state *context) +{ + int i; + struct dc_link *link = NULL; + + /* check if eDP panel is suppose to be set mode, if yes, no need to disable */ + for (i = 0; i < context->stream_count; i++) { + if (context->streams[i]->signal == SIGNAL_TYPE_EDP) + return NULL; + } + + /* check if there is an eDP panel not in use */ + for (i = 0; i < dc->link_count; i++) { + if (dc->links[i]->local_sink && + dc->links[i]->local_sink->sink_signal == SIGNAL_TYPE_EDP) { + link = dc->links[i]; + break; + } + } + + return link; +} + /** * When ASIC goes from VBIOS/VGA mode to driver/accelerated mode we need: * 1. Power down all DC HW blocks @@ -1417,11 +1473,37 @@ static void disable_vga_and_power_gate_all_controllers( * 3. Enable power gating for controller * 4. Set acc_mode_change bit (VBIOS will clear this bit when going to FSDOS) */ -void dce110_enable_accelerated_mode(struct dc *dc) +void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context) { - power_down_all_hw_blocks(dc); + struct dc_bios *dcb = dc->ctx->dc_bios; - disable_vga_and_power_gate_all_controllers(dc); + /* vbios already light up eDP, so we can leverage vbios and skip eDP + * programming + */ + bool can_eDP_fast_boot_optimize = + (dcb->funcs->get_vga_enabled_displays(dc->ctx->dc_bios) == ATOM_DISPLAY_LCD1_ACTIVE); + + /* if OS doesn't light up eDP and eDP link is available, we want to disable */ + struct dc_link *edp_link_to_turnoff = NULL; + + if (can_eDP_fast_boot_optimize) { + edp_link_to_turnoff = get_link_for_edp_not_in_use(dc, context); + + if (!edp_link_to_turnoff) + dc->apply_edp_fast_boot_optimization = true; + } + + if (!dc->apply_edp_fast_boot_optimization) { + if (edp_link_to_turnoff) { + /*turn off backlight before DP_blank and encoder powered down*/ + dc->hwss.edp_backlight_control(edp_link_to_turnoff, false); + } + /*resume from S3, no vbios posting, no need to power down again*/ + power_down_all_hw_blocks(dc); + disable_vga_and_power_gate_all_controllers(dc); + if (edp_link_to_turnoff) + dc->hwss.edp_power_control(edp_link_to_turnoff, false); + } bios_set_scratch_acc_mode_change(dc->ctx->dc_bios); } @@ -1442,7 +1524,7 @@ static uint32_t compute_pstate_blackout_duration( return total_dest_line_time_ns; } -void dce110_set_displaymarks( +static void dce110_set_displaymarks( const struct dc *dc, struct dc_state *context) { @@ -1556,6 +1638,8 @@ static void set_static_screen_control(struct pipe_ctx **pipe_ctx, value |= 0x80; if (events->cursor_update) value |= 0x2; + if (events->force_trigger) + value |= 0x1; #if defined(CONFIG_DRM_AMD_DC_FBC) value |= 0x84; @@ -1766,36 +1850,6 @@ static void enable_fbc(struct dc *dc, } #endif -static enum dc_status apply_ctx_to_hw_fpga( - struct dc *dc, - struct dc_state *context) -{ - enum dc_status status = DC_ERROR_UNEXPECTED; - int i; - - for (i = 0; i < MAX_PIPES; i++) { - struct pipe_ctx *pipe_ctx_old = - &dc->current_state->res_ctx.pipe_ctx[i]; - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; - - if (pipe_ctx->stream == NULL) - continue; - - if (pipe_ctx->stream == pipe_ctx_old->stream) - continue; - - status = apply_single_controller_ctx_to_hw( - pipe_ctx, - context, - dc); - - if (status != DC_OK) - return status; - } - - return DC_OK; -} - static void dce110_reset_hw_ctx_wrap( struct dc *dc, struct dc_state *context) @@ -1865,11 +1919,6 @@ enum dc_status dce110_apply_ctx_to_hw( if (context->stream_count <= 0) return DC_OK; - if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { - apply_ctx_to_hw_fpga(dc, context); - return DC_OK; - } - /* Apply new context */ dcb->funcs->set_scratch_critical_state(dcb, true); @@ -2077,9 +2126,6 @@ enum dc_status dce110_apply_ctx_to_hw( return status; } - /* pplib is notified if disp_num changed */ - dc->hwss.set_bandwidth(dc, context, true); - /* to save power */ apply_min_clocks(dc, context, &clocks_state, false); @@ -2161,13 +2207,14 @@ static void program_surface_visibility(const struct dc *dc, } else if (!pipe_ctx->plane_state->visible) blank_target = true; - dce_set_blender_mode(dc->hwseq, pipe_ctx->pipe_idx, blender_mode); + dce_set_blender_mode(dc->hwseq, pipe_ctx->stream_res.tg->inst, blender_mode); pipe_ctx->stream_res.tg->funcs->set_blank(pipe_ctx->stream_res.tg, blank_target); } static void program_gamut_remap(struct pipe_ctx *pipe_ctx) { + int i = 0; struct xfm_grph_csc_adjustment adjust; memset(&adjust, 0, sizeof(adjust)); adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_BYPASS; @@ -2175,33 +2222,10 @@ static void program_gamut_remap(struct pipe_ctx *pipe_ctx) if (pipe_ctx->stream->gamut_remap_matrix.enable_remap == true) { adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_SW; - adjust.temperature_matrix[0] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[0]; - adjust.temperature_matrix[1] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[1]; - adjust.temperature_matrix[2] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[2]; - adjust.temperature_matrix[3] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[4]; - adjust.temperature_matrix[4] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[5]; - adjust.temperature_matrix[5] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[6]; - adjust.temperature_matrix[6] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[8]; - adjust.temperature_matrix[7] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[9]; - adjust.temperature_matrix[8] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[10]; + + for (i = 0; i < CSC_TEMPERATURE_MATRIX_SIZE; i++) + adjust.temperature_matrix[i] = + pipe_ctx->stream->gamut_remap_matrix.matrix[i]; } pipe_ctx->plane_res.xfm->funcs->transform_set_gamut_remap(pipe_ctx->plane_res.xfm, &adjust); @@ -2225,7 +2249,7 @@ static void set_plane_config( memset(&tbl_entry, 0, sizeof(tbl_entry)); adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_BYPASS; - dce_enable_fe_clock(dc->hwseq, pipe_ctx->pipe_idx, true); + dce_enable_fe_clock(dc->hwseq, mi->inst, true); set_default_colors(pipe_ctx); if (pipe_ctx->stream->csc_color_matrix.enable_adjustment == true) { @@ -2242,33 +2266,10 @@ static void set_plane_config( if (pipe_ctx->stream->gamut_remap_matrix.enable_remap == true) { adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_SW; - adjust.temperature_matrix[0] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[0]; - adjust.temperature_matrix[1] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[1]; - adjust.temperature_matrix[2] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[2]; - adjust.temperature_matrix[3] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[4]; - adjust.temperature_matrix[4] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[5]; - adjust.temperature_matrix[5] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[6]; - adjust.temperature_matrix[6] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[8]; - adjust.temperature_matrix[7] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[9]; - adjust.temperature_matrix[8] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[10]; + + for (i = 0; i < CSC_TEMPERATURE_MATRIX_SIZE; i++) + adjust.temperature_matrix[i] = + pipe_ctx->stream->gamut_remap_matrix.matrix[i]; } pipe_ctx->plane_res.xfm->funcs->transform_set_gamut_remap(pipe_ctx->plane_res.xfm, &adjust); @@ -2313,7 +2314,7 @@ static void update_plane_addr(const struct dc *dc, plane_state->status.requested_address = plane_state->address; } -void dce110_update_pending_status(struct pipe_ctx *pipe_ctx) +static void dce110_update_pending_status(struct pipe_ctx *pipe_ctx) { struct dc_plane_state *plane_state = pipe_ctx->plane_state; @@ -2554,7 +2555,7 @@ void dce110_fill_display_configs( num_cfgs++; cfg->signal = pipe_ctx->stream->signal; - cfg->pipe_idx = pipe_ctx->pipe_idx; + cfg->pipe_idx = pipe_ctx->stream_res.tg->inst; cfg->src_height = stream->src.height; cfg->src_width = stream->src.width; cfg->ddi_channel_mapping = @@ -2717,6 +2718,7 @@ static void dce110_program_front_end_for_pipe( struct xfm_grph_csc_adjustment adjust; struct out_csc_color_matrix tbl_entry; unsigned int i; + struct dc_context *ctx = dc->ctx; #if defined(CONFIG_DRM_AMD_DC_FBC) unsigned int underlay_idx = dc->res_pool->underlay_pipe_index; #endif @@ -2728,7 +2730,7 @@ static void dce110_program_front_end_for_pipe( memset(&adjust, 0, sizeof(adjust)); adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_BYPASS; - dce_enable_fe_clock(dc->hwseq, pipe_ctx->pipe_idx, true); + dce_enable_fe_clock(dc->hwseq, mi->inst, true); set_default_colors(pipe_ctx); if (pipe_ctx->stream->csc_color_matrix.enable_adjustment @@ -2746,33 +2748,10 @@ static void dce110_program_front_end_for_pipe( if (pipe_ctx->stream->gamut_remap_matrix.enable_remap == true) { adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_SW; - adjust.temperature_matrix[0] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[0]; - adjust.temperature_matrix[1] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[1]; - adjust.temperature_matrix[2] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[2]; - adjust.temperature_matrix[3] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[4]; - adjust.temperature_matrix[4] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[5]; - adjust.temperature_matrix[5] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[6]; - adjust.temperature_matrix[6] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[8]; - adjust.temperature_matrix[7] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[9]; - adjust.temperature_matrix[8] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[10]; + + for (i = 0; i < CSC_TEMPERATURE_MATRIX_SIZE; i++) + adjust.temperature_matrix[i] = + pipe_ctx->stream->gamut_remap_matrix.matrix[i]; } pipe_ctx->plane_res.xfm->funcs->transform_set_gamut_remap(pipe_ctx->plane_res.xfm, &adjust); @@ -2819,7 +2798,7 @@ static void dce110_program_front_end_for_pipe( if (pipe_ctx->plane_state->update_flags.bits.full_update) dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream); - dm_logger_write(dc->ctx->logger, LOG_SURFACE, + DC_LOG_SURFACE( "Pipe:%d 0x%x: addr hi:0x%x, " "addr low:0x%x, " "src: %d, %d, %d," @@ -2842,7 +2821,7 @@ static void dce110_program_front_end_for_pipe( pipe_ctx->plane_state->clip_rect.width, pipe_ctx->plane_state->clip_rect.height); - dm_logger_write(dc->ctx->logger, LOG_SURFACE, + DC_LOG_SURFACE( "Pipe %d: width, height, x, y\n" "viewport:%d, %d, %d, %d\n" "recout: %d, %d, %d, %d\n", @@ -2914,7 +2893,8 @@ static void dce110_apply_ctx_for_surface( static void dce110_power_down_fe(struct dc *dc, struct pipe_ctx *pipe_ctx) { - int fe_idx = pipe_ctx->pipe_idx; + int fe_idx = pipe_ctx->plane_res.mi ? + pipe_ctx->plane_res.mi->inst : pipe_ctx->pipe_idx; /* Do not power down fe when stream is active on dce*/ if (dc->current_state->res_ctx.pipe_ctx[fe_idx].stream) @@ -2987,15 +2967,18 @@ void dce110_set_cursor_attribute(struct pipe_ctx *pipe_ctx) { struct dc_cursor_attributes *attributes = &pipe_ctx->stream->cursor_attributes; - if (pipe_ctx->plane_res.ipp->funcs->ipp_cursor_set_attributes) + if (pipe_ctx->plane_res.ipp && + pipe_ctx->plane_res.ipp->funcs->ipp_cursor_set_attributes) pipe_ctx->plane_res.ipp->funcs->ipp_cursor_set_attributes( pipe_ctx->plane_res.ipp, attributes); - if (pipe_ctx->plane_res.mi->funcs->set_cursor_attributes) + if (pipe_ctx->plane_res.mi && + pipe_ctx->plane_res.mi->funcs->set_cursor_attributes) pipe_ctx->plane_res.mi->funcs->set_cursor_attributes( pipe_ctx->plane_res.mi, attributes); - if (pipe_ctx->plane_res.xfm->funcs->set_cursor_attributes) + if (pipe_ctx->plane_res.xfm && + pipe_ctx->plane_res.xfm->funcs->set_cursor_attributes) pipe_ctx->plane_res.xfm->funcs->set_cursor_attributes( pipe_ctx->plane_res.xfm, attributes); } @@ -3023,6 +3006,7 @@ static const struct hw_sequencer_funcs dce110_funcs = { .enable_stream = dce110_enable_stream, .disable_stream = dce110_disable_stream, .unblank_stream = dce110_unblank_stream, + .blank_stream = dce110_blank_stream, .enable_display_pipe_clock_gating = enable_display_pipe_clock_gating, .enable_display_power_gating = dce110_enable_display_power_gating, .disable_plane = dce110_power_down_fe,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h index fc63764..5d7e9f5 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h
@@ -39,11 +39,7 @@ enum dc_status dce110_apply_ctx_to_hw( struct dc *dc, struct dc_state *context); -void dce110_set_display_clock(struct dc_state *context); -void dce110_set_displaymarks( - const struct dc *dc, - struct dc_state *context); void dce110_enable_stream(struct pipe_ctx *pipe_ctx); @@ -52,15 +48,14 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option); void dce110_unblank_stream(struct pipe_ctx *pipe_ctx, struct dc_link_settings *link_settings); +void dce110_blank_stream(struct pipe_ctx *pipe_ctx); void dce110_update_info_frame(struct pipe_ctx *pipe_ctx); void dce110_set_avmute(struct pipe_ctx *pipe_ctx, bool enable); -void dce110_enable_accelerated_mode(struct dc *dc); +void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context); void dce110_power_down(struct dc *dc); -void dce110_update_pending_status(struct pipe_ctx *pipe_ctx); - void dce110_fill_display_configs( const struct dc_state *context, struct dm_pp_display_configuration *pp_display_cfg);
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c index 59c8ccf..94ac43e 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
@@ -52,6 +52,8 @@ #include "dce/dce_abm.h" #include "dce/dce_dmcu.h" +#define DC_LOGGER \ + dc->ctx->logger #if defined(CONFIG_DRM_AMD_DC_FBC) #include "dce110/dce110_compressor.h" #endif @@ -700,7 +702,7 @@ static void get_pixel_clock_parameters( pixel_clk_params->requested_pix_clk = stream->timing.pix_clk_khz; pixel_clk_params->encoder_object_id = stream->sink->link->link_enc->id; pixel_clk_params->signal_type = pipe_ctx->stream->signal; - pixel_clk_params->controller_id = pipe_ctx->pipe_idx + 1; + pixel_clk_params->controller_id = pipe_ctx->stream_res.tg->inst + 1; /* TODO: un-hardcode*/ pixel_clk_params->requested_sym_clk = LINK_RATE_LOW * LINK_RATE_REF_FREQ_IN_KHZ; @@ -771,8 +773,7 @@ static bool dce110_validate_bandwidth( { bool result = false; - dm_logger_write( - dc->ctx->logger, LOG_BANDWIDTH_CALCS, + DC_LOG_BANDWIDTH_CALCS( "%s: start", __func__); @@ -786,8 +787,7 @@ static bool dce110_validate_bandwidth( result = true; if (!result) - dm_logger_write(dc->ctx->logger, LOG_BANDWIDTH_VALIDATION, - "%s: %dx%d@%d Bandwidth validation failed!\n", + DC_LOG_BANDWIDTH_VALIDATION("%s: %dx%d@%d Bandwidth validation failed!\n", __func__, context->streams[0]->timing.h_addressable, context->streams[0]->timing.v_addressable, @@ -990,7 +990,7 @@ static struct pipe_ctx *dce110_acquire_underlay( dc->hwss.enable_display_power_gating( dc, - pipe_ctx->pipe_idx, + pipe_ctx->stream_res.tg->inst, dcb, PIPE_GATING_CONTROL_DISABLE); /* @@ -1170,7 +1170,7 @@ static bool construct( pool->base.pipe_count = pool->base.res_cap->num_timing_generator; pool->base.underlay_pipe_index = pool->base.pipe_count; - + pool->base.timing_generator_count = pool->base.res_cap->num_timing_generator; dc->caps.max_downscale_ratio = 150; dc->caps.i2c_speed_in_khz = 100; dc->caps.max_cursor_size = 128;
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c index 25ca721..be71539 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
@@ -2077,6 +2077,125 @@ bool dce110_arm_vert_intr(struct timing_generator *tg, uint8_t width) return true; } +static bool dce110_is_tg_enabled(struct timing_generator *tg) +{ + uint32_t addr = 0; + uint32_t value = 0; + uint32_t field = 0; + struct dce110_timing_generator *tg110 = DCE110TG_FROM_TG(tg); + + addr = CRTC_REG(mmCRTC_CONTROL); + value = dm_read_reg(tg->ctx, addr); + field = get_reg_field_value(value, CRTC_CONTROL, + CRTC_CURRENT_MASTER_EN_STATE); + return field == 1; +} + +bool dce110_configure_crc(struct timing_generator *tg, + const struct crc_params *params) +{ + uint32_t cntl_addr = 0; + uint32_t addr = 0; + uint32_t value; + struct dce110_timing_generator *tg110 = DCE110TG_FROM_TG(tg); + + /* Cannot configure crc on a CRTC that is disabled */ + if (!dce110_is_tg_enabled(tg)) + return false; + + cntl_addr = CRTC_REG(mmCRTC_CRC_CNTL); + + /* First, disable CRC before we configure it. */ + dm_write_reg(tg->ctx, cntl_addr, 0); + + if (!params->enable) + return true; + + /* Program frame boundaries */ + /* Window A x axis start and end. */ + value = 0; + addr = CRTC_REG(mmCRTC_CRC0_WINDOWA_X_CONTROL); + set_reg_field_value(value, params->windowa_x_start, + CRTC_CRC0_WINDOWA_X_CONTROL, + CRTC_CRC0_WINDOWA_X_START); + set_reg_field_value(value, params->windowa_x_end, + CRTC_CRC0_WINDOWA_X_CONTROL, + CRTC_CRC0_WINDOWA_X_END); + dm_write_reg(tg->ctx, addr, value); + + /* Window A y axis start and end. */ + value = 0; + addr = CRTC_REG(mmCRTC_CRC0_WINDOWA_Y_CONTROL); + set_reg_field_value(value, params->windowa_y_start, + CRTC_CRC0_WINDOWA_Y_CONTROL, + CRTC_CRC0_WINDOWA_Y_START); + set_reg_field_value(value, params->windowa_y_end, + CRTC_CRC0_WINDOWA_Y_CONTROL, + CRTC_CRC0_WINDOWA_Y_END); + dm_write_reg(tg->ctx, addr, value); + + /* Window B x axis start and end. */ + value = 0; + addr = CRTC_REG(mmCRTC_CRC0_WINDOWB_X_CONTROL); + set_reg_field_value(value, params->windowb_x_start, + CRTC_CRC0_WINDOWB_X_CONTROL, + CRTC_CRC0_WINDOWB_X_START); + set_reg_field_value(value, params->windowb_x_end, + CRTC_CRC0_WINDOWB_X_CONTROL, + CRTC_CRC0_WINDOWB_X_END); + dm_write_reg(tg->ctx, addr, value); + + /* Window B y axis start and end. */ + value = 0; + addr = CRTC_REG(mmCRTC_CRC0_WINDOWB_Y_CONTROL); + set_reg_field_value(value, params->windowb_y_start, + CRTC_CRC0_WINDOWB_Y_CONTROL, + CRTC_CRC0_WINDOWB_Y_START); + set_reg_field_value(value, params->windowb_y_end, + CRTC_CRC0_WINDOWB_Y_CONTROL, + CRTC_CRC0_WINDOWB_Y_END); + dm_write_reg(tg->ctx, addr, value); + + /* Set crc mode and selection, and enable. Only using CRC0*/ + value = 0; + set_reg_field_value(value, params->continuous_mode ? 1 : 0, + CRTC_CRC_CNTL, CRTC_CRC_CONT_EN); + set_reg_field_value(value, params->selection, + CRTC_CRC_CNTL, CRTC_CRC0_SELECT); + set_reg_field_value(value, 1, CRTC_CRC_CNTL, CRTC_CRC_EN); + dm_write_reg(tg->ctx, cntl_addr, value); + + return true; +} + +bool dce110_get_crc(struct timing_generator *tg, + uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb) +{ + uint32_t addr = 0; + uint32_t value = 0; + uint32_t field = 0; + struct dce110_timing_generator *tg110 = DCE110TG_FROM_TG(tg); + + addr = CRTC_REG(mmCRTC_CRC_CNTL); + value = dm_read_reg(tg->ctx, addr); + field = get_reg_field_value(value, CRTC_CRC_CNTL, CRTC_CRC_EN); + + /* Early return if CRC is not enabled for this CRTC */ + if (!field) + return false; + + addr = CRTC_REG(mmCRTC_CRC0_DATA_RG); + value = dm_read_reg(tg->ctx, addr); + *r_cr = get_reg_field_value(value, CRTC_CRC0_DATA_RG, CRC0_R_CR); + *g_y = get_reg_field_value(value, CRTC_CRC0_DATA_RG, CRC0_G_Y); + + addr = CRTC_REG(mmCRTC_CRC0_DATA_B); + value = dm_read_reg(tg->ctx, addr); + *b_cb = get_reg_field_value(value, CRTC_CRC0_DATA_B, CRC0_B_CB); + + return true; +} + static const struct timing_generator_funcs dce110_tg_funcs = { .validate_timing = dce110_tg_validate_timing, .program_timing = dce110_tg_program_timing, @@ -2112,6 +2231,9 @@ static const struct timing_generator_funcs dce110_tg_funcs = { dce110_timing_generator_set_static_screen_control, .set_test_pattern = dce110_timing_generator_set_test_pattern, .arm_vert_intr = dce110_arm_vert_intr, + .is_tg_enabled = dce110_is_tg_enabled, + .configure_crc = dce110_configure_crc, + .get_crc = dce110_get_crc, }; void dce110_timing_generator_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h index 232747c..734d496 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h
@@ -276,4 +276,10 @@ void dce110_tg_set_colors(struct timing_generator *tg, bool dce110_arm_vert_intr( struct timing_generator *tg, uint8_t width); +bool dce110_configure_crc(struct timing_generator *tg, + const struct crc_params *params); + +bool dce110_get_crc(struct timing_generator *tg, + uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb); + #endif /* __DC_TIMING_GENERATOR_DCE110_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c index 31cbd1f..0658de7 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c
@@ -38,6 +38,8 @@ #include "timing_generator.h" +#define DC_LOGGER \ + tg->ctx->logger /** ******************************************************************************** * * DCE11 Timing Generator Implementation @@ -607,8 +609,7 @@ static uint32_t dce110_timing_generator_v_get_vblank_counter(struct timing_gener static bool dce110_timing_generator_v_did_triggered_reset_occur( struct timing_generator *tg) { - dm_logger_write(tg->ctx->logger, LOG_ERROR, - "Timing Sync not supported on underlay pipe\n"); + DC_LOG_ERROR("Timing Sync not supported on underlay pipe\n"); return false; } @@ -616,8 +617,7 @@ static void dce110_timing_generator_v_setup_global_swap_lock( struct timing_generator *tg, const struct dcp_gsl_params *gsl_params) { - dm_logger_write(tg->ctx->logger, LOG_ERROR, - "Timing Sync not supported on underlay pipe\n"); + DC_LOG_ERROR("Timing Sync not supported on underlay pipe\n"); return; } @@ -625,24 +625,21 @@ static void dce110_timing_generator_v_enable_reset_trigger( struct timing_generator *tg, int source_tg_inst) { - dm_logger_write(tg->ctx->logger, LOG_ERROR, - "Timing Sync not supported on underlay pipe\n"); + DC_LOG_ERROR("Timing Sync not supported on underlay pipe\n"); return; } static void dce110_timing_generator_v_disable_reset_trigger( struct timing_generator *tg) { - dm_logger_write(tg->ctx->logger, LOG_ERROR, - "Timing Sync not supported on underlay pipe\n"); + DC_LOG_ERROR("Timing Sync not supported on underlay pipe\n"); return; } static void dce110_timing_generator_v_tear_down_global_swap_lock( struct timing_generator *tg) { - dm_logger_write(tg->ctx->logger, LOG_ERROR, - "Timing Sync not supported on underlay pipe\n"); + DC_LOG_ERROR("Timing Sync not supported on underlay pipe\n"); return; }
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c index 47390dc..8ba3c12 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c
@@ -30,6 +30,8 @@ #include "dce/dce_11_0_sh_mask.h" #define SCLV_PHASES 64 +#define DC_LOGGER \ + xfm->ctx->logger struct sclv_ratios_inits { uint32_t h_int_scale_ratio_luma; @@ -670,8 +672,7 @@ static void dce110_xfmv_set_pixel_storage_depth( if (!(xfm_dce->lb_pixel_depth_supported & depth)) { /*we should use unsupported capabilities * unless it is required by w/a*/ - dm_logger_write(xfm->ctx->logger, LOG_WARNING, - "%s: Capability not supported", + DC_LOG_WARNING("%s: Capability not supported", __func__); } }
diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_compressor.c b/drivers/gpu/drm/amd/display/dc/dce112/dce112_compressor.c index 6964992..faae12c 100644 --- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_compressor.c +++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_compressor.c
@@ -33,7 +33,8 @@ #include "include/logger_interface.h" #include "dce112_compressor.h" - +#define DC_LOGGER \ + cp110->base.ctx->logger #define DCP_REG(reg)\ (reg + cp110->offsets.dcp_offset) #define DMIF_REG(reg)\ @@ -129,8 +130,7 @@ static uint32_t lpt_memory_control_config(struct dce112_compressor *cp110, LOW_POWER_TILING_NUM_PIPES); break; default: - dm_logger_write( - cp110->base.ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "%s: Invalid LPT NUM_PIPES!!!", __func__); break; @@ -175,8 +175,7 @@ static uint32_t lpt_memory_control_config(struct dce112_compressor *cp110, LOW_POWER_TILING_NUM_BANKS); break; default: - dm_logger_write( - cp110->base.ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "%s: Invalid LPT NUM_BANKS!!!", __func__); break; @@ -209,8 +208,7 @@ static uint32_t lpt_memory_control_config(struct dce112_compressor *cp110, LOW_POWER_TILING_PIPE_INTERLEAVE_SIZE); break; default: - dm_logger_write( - cp110->base.ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "%s: Invalid LPT INTERLEAVE_SIZE!!!", __func__); break; @@ -253,15 +251,13 @@ static uint32_t lpt_memory_control_config(struct dce112_compressor *cp110, LOW_POWER_TILING_ROW_SIZE); break; default: - dm_logger_write( - cp110->base.ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "%s: Invalid LPT ROW_SIZE!!!", __func__); break; } } else { - dm_logger_write( - cp110->base.ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "%s: LPT MC Configuration is not provided", __func__); } @@ -311,8 +307,7 @@ static void wait_for_fbc_state_changed( } if (counter == 10) { - dm_logger_write( - cp110->base.ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "%s: wait counter exceeded, changes to HW not applied", __func__); } @@ -525,8 +520,7 @@ void dce112_compressor_program_compressed_surface_address_and_pitch( if (compressor->min_compress_ratio == FBC_COMPRESS_RATIO_1TO1) fbc_pitch = fbc_pitch / 8; else - dm_logger_write( - compressor->ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "%s: Unexpected DCE11 compression ratio", __func__); @@ -690,8 +684,7 @@ void dce112_compressor_program_lpt_control( LOW_POWER_TILING_MODE); break; default: - dm_logger_write( - compressor->ctx->logger, LOG_WARNING, + DC_LOG_WARNING( "%s: Invalid selected DRAM channels for LPT!!!", __func__); break;
diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c index 98d9cd0..cd1e3f7 100644 --- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
@@ -56,6 +56,8 @@ #include "dce/dce_11_2_sh_mask.h" #include "dce100/dce100_resource.h" +#define DC_LOGGER \ + dc->ctx->logger #ifndef mmDP_DPHY_INTERNAL_CTRL #define mmDP_DPHY_INTERNAL_CTRL 0x4aa7 @@ -722,8 +724,7 @@ bool dce112_validate_bandwidth( { bool result = false; - dm_logger_write( - dc->ctx->logger, LOG_BANDWIDTH_CALCS, + DC_LOG_BANDWIDTH_CALCS( "%s: start", __func__); @@ -737,7 +738,7 @@ bool dce112_validate_bandwidth( result = true; if (!result) - dm_logger_write(dc->ctx->logger, LOG_BANDWIDTH_VALIDATION, + DC_LOG_BANDWIDTH_VALIDATION( "%s: Bandwidth validation failed!", __func__); @@ -1100,6 +1101,7 @@ static bool construct( *************************************************/ pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE; pool->base.pipe_count = pool->base.res_cap->num_timing_generator; + pool->base.timing_generator_count = pool->base.res_cap->num_timing_generator; dc->caps.max_downscale_ratio = 200; dc->caps.i2c_speed_in_khz = 100; dc->caps.max_cursor_size = 128;
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c index 75d02974..e96ff86 100644 --- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c
@@ -33,7 +33,8 @@ #include "dce/dce_12_0_offset.h" #include "dce/dce_12_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #include "reg_helper.h" #define CTX \
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c index 5aab01db..4659a4b 100644 --- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
@@ -56,7 +56,8 @@ #include "dce/dce_12_0_offset.h" #include "dce/dce_12_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #include "nbio/nbio_6_1_offset.h" #include "reg_helper.h" @@ -830,6 +831,7 @@ static bool construct( /* TODO: Fill more data from GreenlandAsicCapability.cpp */ pool->base.pipe_count = res_cap.num_timing_generator; + pool->base.timing_generator_count = pool->base.res_cap->num_timing_generator; pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE; dc->caps.max_downscale_ratio = 200;
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c index 0aa60e5..7bee781 100644 --- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c +++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
@@ -27,7 +27,8 @@ #include "dce/dce_12_0_offset.h" #include "dce/dce_12_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #include "dc_types.h" #include "dc_bios_types.h"
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/Makefile b/drivers/gpu/drm/amd/display/dc/dce80/Makefile index bc388aa..666fcb2 100644 --- a/drivers/gpu/drm/amd/display/dc/dce80/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dce80/Makefile
@@ -23,7 +23,7 @@ # Makefile for the 'controller' sub-component of DAL. # It provides the control and status of HW CRTC block. -DCE80 = dce80_timing_generator.o dce80_compressor.o dce80_hw_sequencer.o \ +DCE80 = dce80_timing_generator.o dce80_hw_sequencer.o \ dce80_resource.o AMD_DAL_DCE80 = $(addprefix $(AMDDALPATH)/dc/dce80/,$(DCE80))
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_compressor.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_compressor.c deleted file mode 100644 index 951f2ca..0000000 --- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_compressor.c +++ /dev/null
@@ -1,834 +0,0 @@ -/* - * Copyright 2012-15 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ - -#include "dm_services.h" - -#include "dce/dce_8_0_d.h" -#include "dce/dce_8_0_sh_mask.h" -#include "gmc/gmc_7_1_sh_mask.h" -#include "gmc/gmc_7_1_d.h" - -#include "include/logger_interface.h" -#include "dce80_compressor.h" - -#define DCP_REG(reg)\ - (reg + cp80->offsets.dcp_offset) -#define DMIF_REG(reg)\ - (reg + cp80->offsets.dmif_offset) - -static const struct dce80_compressor_reg_offsets reg_offsets[] = { -{ - .dcp_offset = (mmDCP0_GRPH_CONTROL - mmDCP0_GRPH_CONTROL), - .dmif_offset = (mmDMIF_PG0_DPG_PIPE_DPM_CONTROL - - mmDMIF_PG0_DPG_PIPE_DPM_CONTROL), -}, -{ - .dcp_offset = (mmDCP1_GRPH_CONTROL - mmDCP0_GRPH_CONTROL), - .dmif_offset = (mmDMIF_PG1_DPG_PIPE_DPM_CONTROL - - mmDMIF_PG0_DPG_PIPE_DPM_CONTROL), -}, -{ - .dcp_offset = (mmDCP2_GRPH_CONTROL - mmDCP0_GRPH_CONTROL), - .dmif_offset = (mmDMIF_PG2_DPG_PIPE_DPM_CONTROL - - mmDMIF_PG0_DPG_PIPE_DPM_CONTROL), -}, -{ - .dcp_offset = (mmDCP3_GRPH_CONTROL - mmDCP0_GRPH_CONTROL), - .dmif_offset = (mmDMIF_PG3_DPG_PIPE_DPM_CONTROL - - mmDMIF_PG0_DPG_PIPE_DPM_CONTROL), -}, -{ - .dcp_offset = (mmDCP4_GRPH_CONTROL - mmDCP0_GRPH_CONTROL), - .dmif_offset = (mmDMIF_PG4_DPG_PIPE_DPM_CONTROL - - mmDMIF_PG0_DPG_PIPE_DPM_CONTROL), -}, -{ - .dcp_offset = (mmDCP5_GRPH_CONTROL - mmDCP0_GRPH_CONTROL), - .dmif_offset = (mmDMIF_PG5_DPG_PIPE_DPM_CONTROL - - mmDMIF_PG0_DPG_PIPE_DPM_CONTROL), -} -}; - -static const uint32_t dce8_one_lpt_channel_max_resolution = 2048 * 1200; - -enum fbc_idle_force { - /* Bit 0 - Display registers updated */ - FBC_IDLE_FORCE_DISPLAY_REGISTER_UPDATE = 0x00000001, - - /* Bit 2 - FBC_GRPH_COMP_EN register updated */ - FBC_IDLE_FORCE_GRPH_COMP_EN = 0x00000002, - /* Bit 3 - FBC_SRC_SEL register updated */ - FBC_IDLE_FORCE_SRC_SEL_CHANGE = 0x00000004, - /* Bit 4 - FBC_MIN_COMPRESSION register updated */ - FBC_IDLE_FORCE_MIN_COMPRESSION_CHANGE = 0x00000008, - /* Bit 5 - FBC_ALPHA_COMP_EN register updated */ - FBC_IDLE_FORCE_ALPHA_COMP_EN = 0x00000010, - /* Bit 6 - FBC_ZERO_ALPHA_CHUNK_SKIP_EN register updated */ - FBC_IDLE_FORCE_ZERO_ALPHA_CHUNK_SKIP_EN = 0x00000020, - /* Bit 7 - FBC_FORCE_COPY_TO_COMP_BUF register updated */ - FBC_IDLE_FORCE_FORCE_COPY_TO_COMP_BUF = 0x00000040, - - /* Bit 24 - Memory write to region 0 defined by MC registers. */ - FBC_IDLE_FORCE_MEMORY_WRITE_TO_REGION0 = 0x01000000, - /* Bit 25 - Memory write to region 1 defined by MC registers */ - FBC_IDLE_FORCE_MEMORY_WRITE_TO_REGION1 = 0x02000000, - /* Bit 26 - Memory write to region 2 defined by MC registers */ - FBC_IDLE_FORCE_MEMORY_WRITE_TO_REGION2 = 0x04000000, - /* Bit 27 - Memory write to region 3 defined by MC registers. */ - FBC_IDLE_FORCE_MEMORY_WRITE_TO_REGION3 = 0x08000000, - - /* Bit 28 - Memory write from any client other than MCIF */ - FBC_IDLE_FORCE_MEMORY_WRITE_OTHER_THAN_MCIF = 0x10000000, - /* Bit 29 - CG statics screen signal is inactive */ - FBC_IDLE_FORCE_CG_STATIC_SCREEN_IS_INACTIVE = 0x20000000, -}; - -static uint32_t lpt_size_alignment(struct dce80_compressor *cp80) -{ - /*LPT_ALIGNMENT (in bytes) = ROW_SIZE * #BANKS * # DRAM CHANNELS. */ - return cp80->base.raw_size * cp80->base.banks_num * - cp80->base.dram_channels_num; -} - -static uint32_t lpt_memory_control_config(struct dce80_compressor *cp80, - uint32_t lpt_control) -{ - /*LPT MC Config */ - if (cp80->base.options.bits.LPT_MC_CONFIG == 1) { - /* POSSIBLE VALUES for LPT NUM_PIPES (DRAM CHANNELS): - * 00 - 1 CHANNEL - * 01 - 2 CHANNELS - * 02 - 4 OR 6 CHANNELS - * (Only for discrete GPU, N/A for CZ) - * 03 - 8 OR 12 CHANNELS - * (Only for discrete GPU, N/A for CZ) */ - switch (cp80->base.dram_channels_num) { - case 2: - set_reg_field_value( - lpt_control, - 1, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_NUM_PIPES); - break; - case 1: - set_reg_field_value( - lpt_control, - 0, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_NUM_PIPES); - break; - default: - dm_logger_write( - cp80->base.ctx->logger, LOG_WARNING, - "%s: Invalid LPT NUM_PIPES!!!", - __func__); - break; - } - - /* The mapping for LPT NUM_BANKS is in - * GRPH_CONTROL.GRPH_NUM_BANKS register field - * Specifies the number of memory banks for tiling - * purposes. Only applies to 2D and 3D tiling modes. - * POSSIBLE VALUES: - * 00 - DCP_GRPH_NUM_BANKS_2BANK: ADDR_SURF_2_BANK - * 01 - DCP_GRPH_NUM_BANKS_4BANK: ADDR_SURF_4_BANK - * 02 - DCP_GRPH_NUM_BANKS_8BANK: ADDR_SURF_8_BANK - * 03 - DCP_GRPH_NUM_BANKS_16BANK: ADDR_SURF_16_BANK */ - switch (cp80->base.banks_num) { - case 16: - set_reg_field_value( - lpt_control, - 3, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_NUM_BANKS); - break; - case 8: - set_reg_field_value( - lpt_control, - 2, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_NUM_BANKS); - break; - case 4: - set_reg_field_value( - lpt_control, - 1, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_NUM_BANKS); - break; - case 2: - set_reg_field_value( - lpt_control, - 0, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_NUM_BANKS); - break; - default: - dm_logger_write( - cp80->base.ctx->logger, LOG_WARNING, - "%s: Invalid LPT NUM_BANKS!!!", - __func__); - break; - } - - /* The mapping is in DMIF_ADDR_CALC. - * ADDR_CONFIG_PIPE_INTERLEAVE_SIZE register field for - * Carrizo specifies the memory interleave per pipe. - * It effectively specifies the location of pipe bits in - * the memory address. - * POSSIBLE VALUES: - * 00 - ADDR_CONFIG_PIPE_INTERLEAVE_256B: 256 byte - * interleave - * 01 - ADDR_CONFIG_PIPE_INTERLEAVE_512B: 512 byte - * interleave - */ - switch (cp80->base.channel_interleave_size) { - case 256: /*256B */ - set_reg_field_value( - lpt_control, - 0, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_PIPE_INTERLEAVE_SIZE); - break; - case 512: /*512B */ - set_reg_field_value( - lpt_control, - 1, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_PIPE_INTERLEAVE_SIZE); - break; - default: - dm_logger_write( - cp80->base.ctx->logger, LOG_WARNING, - "%s: Invalid LPT INTERLEAVE_SIZE!!!", - __func__); - break; - } - - /* The mapping for LOW_POWER_TILING_ROW_SIZE is in - * DMIF_ADDR_CALC.ADDR_CONFIG_ROW_SIZE register field - * for Carrizo. Specifies the size of dram row in bytes. - * This should match up with NOOFCOLS field in - * MC_ARB_RAMCFG (ROW_SIZE = 4 * 2 ^^ columns). - * This register DMIF_ADDR_CALC is not used by the - * hardware as it is only used for addrlib assertions. - * POSSIBLE VALUES: - * 00 - ADDR_CONFIG_1KB_ROW: Treat 1KB as DRAM row - * boundary - * 01 - ADDR_CONFIG_2KB_ROW: Treat 2KB as DRAM row - * boundary - * 02 - ADDR_CONFIG_4KB_ROW: Treat 4KB as DRAM row - * boundary */ - switch (cp80->base.raw_size) { - case 4096: /*4 KB */ - set_reg_field_value( - lpt_control, - 2, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_ROW_SIZE); - break; - case 2048: - set_reg_field_value( - lpt_control, - 1, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_ROW_SIZE); - break; - case 1024: - set_reg_field_value( - lpt_control, - 0, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_ROW_SIZE); - break; - default: - dm_logger_write( - cp80->base.ctx->logger, LOG_WARNING, - "%s: Invalid LPT ROW_SIZE!!!", - __func__); - break; - } - } else { - dm_logger_write( - cp80->base.ctx->logger, LOG_WARNING, - "%s: LPT MC Configuration is not provided", - __func__); - } - - return lpt_control; -} - -static bool is_source_bigger_than_epanel_size( - struct dce80_compressor *cp80, - uint32_t source_view_width, - uint32_t source_view_height) -{ - if (cp80->base.embedded_panel_h_size != 0 && - cp80->base.embedded_panel_v_size != 0 && - ((source_view_width * source_view_height) > - (cp80->base.embedded_panel_h_size * - cp80->base.embedded_panel_v_size))) - return true; - - return false; -} - -static uint32_t align_to_chunks_number_per_line( - struct dce80_compressor *cp80, - uint32_t pixels) -{ - return 256 * ((pixels + 255) / 256); -} - -static void wait_for_fbc_state_changed( - struct dce80_compressor *cp80, - bool enabled) -{ - uint8_t counter = 0; - uint32_t addr = mmFBC_STATUS; - uint32_t value; - - while (counter < 10) { - value = dm_read_reg(cp80->base.ctx, addr); - if (get_reg_field_value( - value, - FBC_STATUS, - FBC_ENABLE_STATUS) == enabled) - break; - udelay(10); - counter++; - } - - if (counter == 10) { - dm_logger_write( - cp80->base.ctx->logger, LOG_WARNING, - "%s: wait counter exceeded, changes to HW not applied", - __func__); - } -} - -void dce80_compressor_power_up_fbc(struct compressor *compressor) -{ - uint32_t value; - uint32_t addr; - - addr = mmFBC_CNTL; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value(value, 0, FBC_CNTL, FBC_GRPH_COMP_EN); - set_reg_field_value(value, 1, FBC_CNTL, FBC_EN); - set_reg_field_value(value, 2, FBC_CNTL, FBC_COHERENCY_MODE); - dm_write_reg(compressor->ctx, addr, value); - - addr = mmFBC_COMP_MODE; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value(value, 1, FBC_COMP_MODE, FBC_RLE_EN); - set_reg_field_value(value, 1, FBC_COMP_MODE, FBC_DPCM4_RGB_EN); - set_reg_field_value(value, 1, FBC_COMP_MODE, FBC_IND_EN); - dm_write_reg(compressor->ctx, addr, value); - - addr = mmFBC_COMP_CNTL; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value(value, 1, FBC_COMP_CNTL, FBC_DEPTH_RGB08_EN); - dm_write_reg(compressor->ctx, addr, value); - /*FBC_MIN_COMPRESSION 0 ==> 2:1 */ - /* 1 ==> 4:1 */ - /* 2 ==> 8:1 */ - /* 0xF ==> 1:1 */ - set_reg_field_value(value, 0xF, FBC_COMP_CNTL, FBC_MIN_COMPRESSION); - dm_write_reg(compressor->ctx, addr, value); - compressor->min_compress_ratio = FBC_COMPRESS_RATIO_1TO1; - - value = 0; - dm_write_reg(compressor->ctx, mmFBC_IND_LUT0, value); - - value = 0xFFFFFF; - dm_write_reg(compressor->ctx, mmFBC_IND_LUT1, value); -} - -void dce80_compressor_enable_fbc( - struct compressor *compressor, - uint32_t paths_num, - struct compr_addr_and_pitch_params *params) -{ - struct dce80_compressor *cp80 = TO_DCE80_COMPRESSOR(compressor); - - if (compressor->options.bits.FBC_SUPPORT && - (compressor->options.bits.DUMMY_BACKEND == 0) && - (!dce80_compressor_is_fbc_enabled_in_hw(compressor, NULL)) && - (!is_source_bigger_than_epanel_size( - cp80, - params->source_view_width, - params->source_view_height))) { - - uint32_t addr; - uint32_t value; - - /* Before enabling FBC first need to enable LPT if applicable - * LPT state should always be changed (enable/disable) while FBC - * is disabled */ - if (compressor->options.bits.LPT_SUPPORT && (paths_num < 2) && - (params->source_view_width * - params->source_view_height <= - dce8_one_lpt_channel_max_resolution)) { - dce80_compressor_enable_lpt(compressor); - } - - addr = mmFBC_CNTL; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value(value, 1, FBC_CNTL, FBC_GRPH_COMP_EN); - set_reg_field_value( - value, - params->inst, - FBC_CNTL, FBC_SRC_SEL); - dm_write_reg(compressor->ctx, addr, value); - - /* Keep track of enum controller_id FBC is attached to */ - compressor->is_enabled = true; - compressor->attached_inst = params->inst; - cp80->offsets = reg_offsets[params->inst]; - - /*Toggle it as there is bug in HW */ - set_reg_field_value(value, 0, FBC_CNTL, FBC_GRPH_COMP_EN); - dm_write_reg(compressor->ctx, addr, value); - set_reg_field_value(value, 1, FBC_CNTL, FBC_GRPH_COMP_EN); - dm_write_reg(compressor->ctx, addr, value); - - wait_for_fbc_state_changed(cp80, true); - } -} - -void dce80_compressor_disable_fbc(struct compressor *compressor) -{ - struct dce80_compressor *cp80 = TO_DCE80_COMPRESSOR(compressor); - - if (compressor->options.bits.FBC_SUPPORT && - dce80_compressor_is_fbc_enabled_in_hw(compressor, NULL)) { - uint32_t reg_data; - /* Turn off compression */ - reg_data = dm_read_reg(compressor->ctx, mmFBC_CNTL); - set_reg_field_value(reg_data, 0, FBC_CNTL, FBC_GRPH_COMP_EN); - dm_write_reg(compressor->ctx, mmFBC_CNTL, reg_data); - - /* Reset enum controller_id to undefined */ - compressor->attached_inst = 0; - compressor->is_enabled = false; - - /* Whenever disabling FBC make sure LPT is disabled if LPT - * supported */ - if (compressor->options.bits.LPT_SUPPORT) - dce80_compressor_disable_lpt(compressor); - - wait_for_fbc_state_changed(cp80, false); - } -} - -bool dce80_compressor_is_fbc_enabled_in_hw( - struct compressor *compressor, - uint32_t *inst) -{ - /* Check the hardware register */ - uint32_t value; - - value = dm_read_reg(compressor->ctx, mmFBC_STATUS); - if (get_reg_field_value(value, FBC_STATUS, FBC_ENABLE_STATUS)) { - if (inst != NULL) - *inst = compressor->attached_inst; - return true; - } - - value = dm_read_reg(compressor->ctx, mmFBC_CNTL); - if (get_reg_field_value(value, FBC_CNTL, FBC_GRPH_COMP_EN)) { - if (inst != NULL) - *inst = compressor->attached_inst; - return true; - } - - return false; -} - -bool dce80_compressor_is_lpt_enabled_in_hw(struct compressor *compressor) -{ - /* Check the hardware register */ - uint32_t value = dm_read_reg(compressor->ctx, - mmLOW_POWER_TILING_CONTROL); - - return get_reg_field_value( - value, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_ENABLE); -} - -void dce80_compressor_program_compressed_surface_address_and_pitch( - struct compressor *compressor, - struct compr_addr_and_pitch_params *params) -{ - struct dce80_compressor *cp80 = TO_DCE80_COMPRESSOR(compressor); - uint32_t value = 0; - uint32_t fbc_pitch = 0; - uint32_t compressed_surf_address_low_part = - compressor->compr_surface_address.addr.low_part; - - /* Clear content first. */ - dm_write_reg( - compressor->ctx, - DCP_REG(mmGRPH_COMPRESS_SURFACE_ADDRESS_HIGH), - 0); - dm_write_reg(compressor->ctx, - DCP_REG(mmGRPH_COMPRESS_SURFACE_ADDRESS), 0); - - if (compressor->options.bits.LPT_SUPPORT) { - uint32_t lpt_alignment = lpt_size_alignment(cp80); - - if (lpt_alignment != 0) { - compressed_surf_address_low_part = - ((compressed_surf_address_low_part - + (lpt_alignment - 1)) / lpt_alignment) - * lpt_alignment; - } - } - - /* Write address, HIGH has to be first. */ - dm_write_reg(compressor->ctx, - DCP_REG(mmGRPH_COMPRESS_SURFACE_ADDRESS_HIGH), - compressor->compr_surface_address.addr.high_part); - dm_write_reg(compressor->ctx, - DCP_REG(mmGRPH_COMPRESS_SURFACE_ADDRESS), - compressed_surf_address_low_part); - - fbc_pitch = align_to_chunks_number_per_line( - cp80, - params->source_view_width); - - if (compressor->min_compress_ratio == FBC_COMPRESS_RATIO_1TO1) - fbc_pitch = fbc_pitch / 8; - else - dm_logger_write( - compressor->ctx->logger, LOG_WARNING, - "%s: Unexpected DCE8 compression ratio", - __func__); - - /* Clear content first. */ - dm_write_reg(compressor->ctx, DCP_REG(mmGRPH_COMPRESS_PITCH), 0); - - /* Write FBC Pitch. */ - set_reg_field_value( - value, - fbc_pitch, - GRPH_COMPRESS_PITCH, - GRPH_COMPRESS_PITCH); - dm_write_reg(compressor->ctx, DCP_REG(mmGRPH_COMPRESS_PITCH), value); - -} - -void dce80_compressor_disable_lpt(struct compressor *compressor) -{ - struct dce80_compressor *cp80 = TO_DCE80_COMPRESSOR(compressor); - uint32_t value; - uint32_t addr; - uint32_t inx; - - /* Disable all pipes LPT Stutter */ - for (inx = 0; inx < 3; inx++) { - value = - dm_read_reg( - compressor->ctx, - DMIF_REG(mmDPG_PIPE_STUTTER_CONTROL_NONLPTCH)); - set_reg_field_value( - value, - 0, - DPG_PIPE_STUTTER_CONTROL_NONLPTCH, - STUTTER_ENABLE_NONLPTCH); - dm_write_reg( - compressor->ctx, - DMIF_REG(mmDPG_PIPE_STUTTER_CONTROL_NONLPTCH), - value); - } - - /* Disable LPT */ - addr = mmLOW_POWER_TILING_CONTROL; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value( - value, - 0, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_ENABLE); - dm_write_reg(compressor->ctx, addr, value); - - /* Clear selection of Channel(s) containing Compressed Surface */ - addr = mmGMCON_LPT_TARGET; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value( - value, - 0xFFFFFFFF, - GMCON_LPT_TARGET, - STCTRL_LPT_TARGET); - dm_write_reg(compressor->ctx, mmGMCON_LPT_TARGET, value); -} - -void dce80_compressor_enable_lpt(struct compressor *compressor) -{ - struct dce80_compressor *cp80 = TO_DCE80_COMPRESSOR(compressor); - uint32_t value; - uint32_t addr; - uint32_t value_control; - uint32_t channels; - - /* Enable LPT Stutter from Display pipe */ - value = dm_read_reg(compressor->ctx, - DMIF_REG(mmDPG_PIPE_STUTTER_CONTROL_NONLPTCH)); - set_reg_field_value( - value, - 1, - DPG_PIPE_STUTTER_CONTROL_NONLPTCH, - STUTTER_ENABLE_NONLPTCH); - dm_write_reg(compressor->ctx, - DMIF_REG(mmDPG_PIPE_STUTTER_CONTROL_NONLPTCH), value); - - /* Selection of Channel(s) containing Compressed Surface: 0xfffffff - * will disable LPT. - * STCTRL_LPT_TARGETn corresponds to channel n. */ - addr = mmLOW_POWER_TILING_CONTROL; - value_control = dm_read_reg(compressor->ctx, addr); - channels = get_reg_field_value(value_control, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_MODE); - - addr = mmGMCON_LPT_TARGET; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value( - value, - channels + 1, /* not mentioned in programming guide, - but follow DCE8.1 */ - GMCON_LPT_TARGET, - STCTRL_LPT_TARGET); - dm_write_reg(compressor->ctx, addr, value); - - /* Enable LPT */ - addr = mmLOW_POWER_TILING_CONTROL; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value( - value, - 1, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_ENABLE); - dm_write_reg(compressor->ctx, addr, value); -} - -void dce80_compressor_program_lpt_control( - struct compressor *compressor, - struct compr_addr_and_pitch_params *params) -{ - struct dce80_compressor *cp80 = TO_DCE80_COMPRESSOR(compressor); - uint32_t rows_per_channel; - uint32_t lpt_alignment; - uint32_t source_view_width; - uint32_t source_view_height; - uint32_t lpt_control = 0; - - if (!compressor->options.bits.LPT_SUPPORT) - return; - - lpt_control = dm_read_reg(compressor->ctx, - mmLOW_POWER_TILING_CONTROL); - - /* POSSIBLE VALUES for Low Power Tiling Mode: - * 00 - Use channel 0 - * 01 - Use Channel 0 and 1 - * 02 - Use Channel 0,1,2,3 - * 03 - reserved */ - switch (compressor->lpt_channels_num) { - /* case 2: - * Use Channel 0 & 1 / Not used for DCE 11 */ - case 1: - /*Use Channel 0 for LPT for DCE 11 */ - set_reg_field_value( - lpt_control, - 0, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_MODE); - break; - default: - dm_logger_write( - compressor->ctx->logger, LOG_WARNING, - "%s: Invalid selected DRAM channels for LPT!!!", - __func__); - break; - } - - lpt_control = lpt_memory_control_config(cp80, lpt_control); - - /* Program LOW_POWER_TILING_ROWS_PER_CHAN field which depends on - * FBC compressed surface pitch. - * LOW_POWER_TILING_ROWS_PER_CHAN = Roundup ((Surface Height * - * Surface Pitch) / (Row Size * Number of Channels * - * Number of Banks)). */ - rows_per_channel = 0; - lpt_alignment = lpt_size_alignment(cp80); - source_view_width = - align_to_chunks_number_per_line( - cp80, - params->source_view_width); - source_view_height = (params->source_view_height + 1) & (~0x1); - - if (lpt_alignment != 0) { - rows_per_channel = source_view_width * source_view_height * 4; - rows_per_channel = - (rows_per_channel % lpt_alignment) ? - (rows_per_channel / lpt_alignment + 1) : - rows_per_channel / lpt_alignment; - } - - set_reg_field_value( - lpt_control, - rows_per_channel, - LOW_POWER_TILING_CONTROL, - LOW_POWER_TILING_ROWS_PER_CHAN); - - dm_write_reg(compressor->ctx, - mmLOW_POWER_TILING_CONTROL, lpt_control); -} - -/* - * DCE 11 Frame Buffer Compression Implementation - */ - -void dce80_compressor_set_fbc_invalidation_triggers( - struct compressor *compressor, - uint32_t fbc_trigger) -{ - /* Disable region hit event, FBC_MEMORY_REGION_MASK = 0 (bits 16-19) - * for DCE 11 regions cannot be used - does not work with S/G - */ - uint32_t addr = mmFBC_CLIENT_REGION_MASK; - uint32_t value = dm_read_reg(compressor->ctx, addr); - - set_reg_field_value( - value, - 0, - FBC_CLIENT_REGION_MASK, - FBC_MEMORY_REGION_MASK); - dm_write_reg(compressor->ctx, addr, value); - - /* Setup events when to clear all CSM entries (effectively marking - * current compressed data invalid) - * For DCE 11 CSM metadata 11111 means - "Not Compressed" - * Used as the initial value of the metadata sent to the compressor - * after invalidation, to indicate that the compressor should attempt - * to compress all chunks on the current pass. Also used when the chunk - * is not successfully written to memory. - * When this CSM value is detected, FBC reads from the uncompressed - * buffer. Set events according to passed in value, these events are - * valid for DCE8: - * - bit 0 - display register updated - * - bit 28 - memory write from any client except from MCIF - * - bit 29 - CG static screen signal is inactive - * In addition, DCE8.1 also needs to set new DCE8.1 specific events - * that are used to trigger invalidation on certain register changes, - * for example enabling of Alpha Compression may trigger invalidation of - * FBC once bit is set. These events are as follows: - * - Bit 2 - FBC_GRPH_COMP_EN register updated - * - Bit 3 - FBC_SRC_SEL register updated - * - Bit 4 - FBC_MIN_COMPRESSION register updated - * - Bit 5 - FBC_ALPHA_COMP_EN register updated - * - Bit 6 - FBC_ZERO_ALPHA_CHUNK_SKIP_EN register updated - * - Bit 7 - FBC_FORCE_COPY_TO_COMP_BUF register updated - */ - addr = mmFBC_IDLE_FORCE_CLEAR_MASK; - value = dm_read_reg(compressor->ctx, addr); - set_reg_field_value( - value, - fbc_trigger | - FBC_IDLE_FORCE_GRPH_COMP_EN | - FBC_IDLE_FORCE_SRC_SEL_CHANGE | - FBC_IDLE_FORCE_MIN_COMPRESSION_CHANGE | - FBC_IDLE_FORCE_ALPHA_COMP_EN | - FBC_IDLE_FORCE_ZERO_ALPHA_CHUNK_SKIP_EN | - FBC_IDLE_FORCE_FORCE_COPY_TO_COMP_BUF, - FBC_IDLE_FORCE_CLEAR_MASK, - FBC_IDLE_FORCE_CLEAR_MASK); - dm_write_reg(compressor->ctx, addr, value); -} - -void dce80_compressor_construct(struct dce80_compressor *compressor, - struct dc_context *ctx) -{ - struct dc_bios *bp = ctx->dc_bios; - struct embedded_panel_info panel_info; - - compressor->base.options.raw = 0; - compressor->base.options.bits.FBC_SUPPORT = true; - compressor->base.options.bits.LPT_SUPPORT = true; - /* For DCE 11 always use one DRAM channel for LPT */ - compressor->base.lpt_channels_num = 1; - compressor->base.options.bits.DUMMY_BACKEND = false; - - /* Check if this system has more than 1 DRAM channel; if only 1 then LPT - * should not be supported */ - if (compressor->base.memory_bus_width == 64) - compressor->base.options.bits.LPT_SUPPORT = false; - - compressor->base.options.bits.CLK_GATING_DISABLED = false; - - compressor->base.ctx = ctx; - compressor->base.embedded_panel_h_size = 0; - compressor->base.embedded_panel_v_size = 0; - compressor->base.memory_bus_width = ctx->asic_id.vram_width; - compressor->base.allocated_size = 0; - compressor->base.preferred_requested_size = 0; - compressor->base.min_compress_ratio = FBC_COMPRESS_RATIO_INVALID; - compressor->base.banks_num = 0; - compressor->base.raw_size = 0; - compressor->base.channel_interleave_size = 0; - compressor->base.dram_channels_num = 0; - compressor->base.lpt_channels_num = 0; - compressor->base.attached_inst = 0; - compressor->base.is_enabled = false; - - if (BP_RESULT_OK == - bp->funcs->get_embedded_panel_info(bp, &panel_info)) { - compressor->base.embedded_panel_h_size = - panel_info.lcd_timing.horizontal_addressable; - compressor->base.embedded_panel_v_size = - panel_info.lcd_timing.vertical_addressable; - } -} - -struct compressor *dce80_compressor_create(struct dc_context *ctx) -{ - struct dce80_compressor *cp80 = - kzalloc(sizeof(struct dce80_compressor), GFP_KERNEL); - - if (!cp80) - return NULL; - - dce80_compressor_construct(cp80, ctx); - return &cp80->base; -} - -void dce80_compressor_destroy(struct compressor **compressor) -{ - kfree(TO_DCE80_COMPRESSOR(*compressor)); - *compressor = NULL; -}
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_compressor.h b/drivers/gpu/drm/amd/display/dc/dce80/dce80_compressor.h deleted file mode 100644 index cca58b0..0000000 --- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_compressor.h +++ /dev/null
@@ -1,78 +0,0 @@ -/* Copyright 2012-15 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ - -#ifndef __DC_COMPRESSOR_DCE80_H__ -#define __DC_COMPRESSOR_DCE80_H__ - -#include "../inc/compressor.h" - -#define TO_DCE80_COMPRESSOR(compressor)\ - container_of(compressor, struct dce80_compressor, base) - -struct dce80_compressor_reg_offsets { - uint32_t dcp_offset; - uint32_t dmif_offset; -}; - -struct dce80_compressor { - struct compressor base; - struct dce80_compressor_reg_offsets offsets; -}; - -struct compressor *dce80_compressor_create(struct dc_context *ctx); - -void dce80_compressor_construct(struct dce80_compressor *cp80, - struct dc_context *ctx); - -void dce80_compressor_destroy(struct compressor **cp); - -/* FBC RELATED */ -void dce80_compressor_power_up_fbc(struct compressor *cp); - -void dce80_compressor_enable_fbc(struct compressor *cp, uint32_t paths_num, - struct compr_addr_and_pitch_params *params); - -void dce80_compressor_disable_fbc(struct compressor *cp); - -void dce80_compressor_set_fbc_invalidation_triggers(struct compressor *cp, - uint32_t fbc_trigger); - -void dce80_compressor_program_compressed_surface_address_and_pitch( - struct compressor *cp, - struct compr_addr_and_pitch_params *params); - -bool dce80_compressor_is_fbc_enabled_in_hw(struct compressor *cp, - uint32_t *fbc_mapped_crtc_id); - -/* LPT RELATED */ -void dce80_compressor_enable_lpt(struct compressor *cp); - -void dce80_compressor_disable_lpt(struct compressor *cp); - -void dce80_compressor_program_lpt_control(struct compressor *cp, - struct compr_addr_and_pitch_params *params); - -bool dce80_compressor_is_lpt_enabled_in_hw(struct compressor *cp); - -#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_hw_sequencer.c index ccfcf1c..6c6a1a16 100644 --- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_hw_sequencer.c
@@ -70,47 +70,11 @@ static const struct dce80_hw_seq_reg_offsets reg_offsets[] = { /***************************PIPE_CONTROL***********************************/ -static bool dce80_enable_display_power_gating( - struct dc *dc, - uint8_t controller_id, - struct dc_bios *dcb, - enum pipe_gating_control power_gating) -{ - enum bp_result bp_result = BP_RESULT_OK; - enum bp_pipe_control_action cntl; - struct dc_context *ctx = dc->ctx; - - if (power_gating == PIPE_GATING_CONTROL_INIT) - cntl = ASIC_PIPE_INIT; - else if (power_gating == PIPE_GATING_CONTROL_ENABLE) - cntl = ASIC_PIPE_ENABLE; - else - cntl = ASIC_PIPE_DISABLE; - - if (!(power_gating == PIPE_GATING_CONTROL_INIT && controller_id != 0)){ - - bp_result = dcb->funcs->enable_disp_power_gating( - dcb, controller_id + 1, cntl); - - /* Revert MASTER_UPDATE_MODE to 0 because bios sets it 2 - * by default when command table is called - */ - dm_write_reg(ctx, - HW_REG_CRTC(mmMASTER_UPDATE_MODE, controller_id), - 0); - } - - if (bp_result == BP_RESULT_OK) - return true; - else - return false; -} - void dce80_hw_sequencer_construct(struct dc *dc) { dce110_hw_sequencer_construct(dc); - dc->hwss.enable_display_power_gating = dce80_enable_display_power_gating; + dc->hwss.enable_display_power_gating = dce100_enable_display_power_gating; dc->hwss.pipe_control_lock = dce_pipe_control_lock; dc->hwss.set_bandwidth = dce100_set_bandwidth; }
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c index 25d7eb1..5d854a3 100644 --- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
@@ -53,6 +53,8 @@ #include "reg_helper.h" +#include "dce/dce_dmcu.h" +#include "dce/dce_abm.h" /* TODO remove this include */ #ifndef mmMC_HUB_RDREQ_DMIF_LIMIT @@ -364,6 +366,29 @@ static const struct resource_caps res_cap_83 = { .num_pll = 2, }; +static const struct dce_dmcu_registers dmcu_regs = { + DMCU_DCE80_REG_LIST() +}; + +static const struct dce_dmcu_shift dmcu_shift = { + DMCU_MASK_SH_LIST_DCE80(__SHIFT) +}; + +static const struct dce_dmcu_mask dmcu_mask = { + DMCU_MASK_SH_LIST_DCE80(_MASK) +}; +static const struct dce_abm_registers abm_regs = { + ABM_DCE110_COMMON_REG_LIST() +}; + +static const struct dce_abm_shift abm_shift = { + ABM_MASK_SH_LIST_DCE110(__SHIFT) +}; + +static const struct dce_abm_mask abm_mask = { + ABM_MASK_SH_LIST_DCE110(_MASK) +}; + #define CTX ctx #define REG(reg) mm ## reg @@ -643,6 +668,12 @@ static void destruct(struct dce110_resource_pool *pool) } } + if (pool->base.abm != NULL) + dce_abm_destroy(&pool->base.abm); + + if (pool->base.dmcu != NULL) + dce_dmcu_destroy(&pool->base.dmcu); + if (pool->base.dp_clock_source != NULL) dce80_clock_source_destroy(&pool->base.dp_clock_source); @@ -790,6 +821,7 @@ static bool dce80_construct( *************************************************/ pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE; pool->base.pipe_count = res_cap.num_timing_generator; + pool->base.timing_generator_count = res_cap.num_timing_generator; dc->caps.max_downscale_ratio = 200; dc->caps.i2c_speed_in_khz = 40; dc->caps.max_cursor_size = 128; @@ -849,7 +881,25 @@ static bool dce80_construct( goto res_create_fail; } + pool->base.dmcu = dce_dmcu_create(ctx, + &dmcu_regs, + &dmcu_shift, + &dmcu_mask); + if (pool->base.dmcu == NULL) { + dm_error("DC: failed to create dmcu!\n"); + BREAK_TO_DEBUGGER(); + goto res_create_fail; + } + pool->base.abm = dce_abm_create(ctx, + &abm_regs, + &abm_shift, + &abm_mask); + if (pool->base.abm == NULL) { + dm_error("DC: failed to create abm!\n"); + BREAK_TO_DEBUGGER(); + goto res_create_fail; + } if (dm_pp_get_static_clocks(ctx, &static_clk_info)) pool->base.display_clock->max_clks_state = static_clk_info.max_clocks_state; @@ -955,6 +1005,7 @@ static bool dce81_construct( *************************************************/ pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE; pool->base.pipe_count = res_cap_81.num_timing_generator; + pool->base.timing_generator_count = res_cap_81.num_timing_generator; dc->caps.max_downscale_ratio = 200; dc->caps.i2c_speed_in_khz = 40; dc->caps.max_cursor_size = 128; @@ -1014,6 +1065,25 @@ static bool dce81_construct( goto res_create_fail; } + pool->base.dmcu = dce_dmcu_create(ctx, + &dmcu_regs, + &dmcu_shift, + &dmcu_mask); + if (pool->base.dmcu == NULL) { + dm_error("DC: failed to create dmcu!\n"); + BREAK_TO_DEBUGGER(); + goto res_create_fail; + } + + pool->base.abm = dce_abm_create(ctx, + &abm_regs, + &abm_shift, + &abm_mask); + if (pool->base.abm == NULL) { + dm_error("DC: failed to create abm!\n"); + BREAK_TO_DEBUGGER(); + goto res_create_fail; + } if (dm_pp_get_static_clocks(ctx, &static_clk_info)) pool->base.display_clock->max_clks_state = @@ -1120,6 +1190,7 @@ static bool dce83_construct( *************************************************/ pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE; pool->base.pipe_count = res_cap_83.num_timing_generator; + pool->base.timing_generator_count = res_cap_83.num_timing_generator; dc->caps.max_downscale_ratio = 200; dc->caps.i2c_speed_in_khz = 40; dc->caps.max_cursor_size = 128; @@ -1175,6 +1246,25 @@ static bool dce83_construct( goto res_create_fail; } + pool->base.dmcu = dce_dmcu_create(ctx, + &dmcu_regs, + &dmcu_shift, + &dmcu_mask); + if (pool->base.dmcu == NULL) { + dm_error("DC: failed to create dmcu!\n"); + BREAK_TO_DEBUGGER(); + goto res_create_fail; + } + + pool->base.abm = dce_abm_create(ctx, + &abm_regs, + &abm_shift, + &abm_mask); + if (pool->base.abm == NULL) { + dm_error("DC: failed to create abm!\n"); + BREAK_TO_DEBUGGER(); + goto res_create_fail; + } if (dm_pp_get_static_clocks(ctx, &static_clk_info)) pool->base.display_clock->max_clks_state =
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c index 2658948..3ba4712 100644 --- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c +++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c
@@ -84,7 +84,7 @@ static const struct dce110_timing_generator_offsets reg_offsets[] = { #define DCP_REG(reg) (reg + tg110->offsets.dcp) #define DMIF_REG(reg) (reg + tg110->offsets.dmif) -void program_pix_dur(struct timing_generator *tg, uint32_t pix_clk_khz) +static void program_pix_dur(struct timing_generator *tg, uint32_t pix_clk_khz) { uint64_t pix_dur; uint32_t addr = mmDMIF_PG0_DPG_PIPE_ARBITRATION_CONTROL1 @@ -115,68 +115,7 @@ static void program_timing(struct timing_generator *tg, dce110_tg_program_timing(tg, timing, use_vbios); } -static const struct timing_generator_funcs dce80_tg_funcs = { - .validate_timing = dce110_tg_validate_timing, - .program_timing = program_timing, - .enable_crtc = dce110_timing_generator_enable_crtc, - .disable_crtc = dce110_timing_generator_disable_crtc, - .is_counter_moving = dce110_timing_generator_is_counter_moving, - .get_position = dce110_timing_generator_get_position, - .get_frame_count = dce110_timing_generator_get_vblank_counter, - .get_scanoutpos = dce110_timing_generator_get_crtc_scanoutpos, - .set_early_control = dce110_timing_generator_set_early_control, - .wait_for_state = dce110_tg_wait_for_state, - .set_blank = dce110_tg_set_blank, - .is_blanked = dce110_tg_is_blanked, - .set_colors = dce110_tg_set_colors, - .set_overscan_blank_color = - dce110_timing_generator_set_overscan_color_black, - .set_blank_color = dce110_timing_generator_program_blank_color, - .disable_vga = dce110_timing_generator_disable_vga, - .did_triggered_reset_occur = - dce110_timing_generator_did_triggered_reset_occur, - .setup_global_swap_lock = - dce110_timing_generator_setup_global_swap_lock, - .enable_reset_trigger = dce110_timing_generator_enable_reset_trigger, - .disable_reset_trigger = dce110_timing_generator_disable_reset_trigger, - .tear_down_global_swap_lock = - dce110_timing_generator_tear_down_global_swap_lock, - .set_drr = dce110_timing_generator_set_drr, - .set_static_screen_control = - dce110_timing_generator_set_static_screen_control, - .set_test_pattern = dce110_timing_generator_set_test_pattern, - .arm_vert_intr = dce110_arm_vert_intr, - - /* DCE8.0 overrides */ - .enable_advanced_request = - dce80_timing_generator_enable_advanced_request, -}; - -void dce80_timing_generator_construct( - struct dce110_timing_generator *tg110, - struct dc_context *ctx, - uint32_t instance, - const struct dce110_timing_generator_offsets *offsets) -{ - tg110->controller_id = CONTROLLER_ID_D0 + instance; - tg110->base.inst = instance; - tg110->offsets = *offsets; - tg110->derived_offsets = reg_offsets[instance]; - - tg110->base.funcs = &dce80_tg_funcs; - - tg110->base.ctx = ctx; - tg110->base.bp = ctx->dc_bios; - - tg110->max_h_total = CRTC_H_TOTAL__CRTC_H_TOTAL_MASK + 1; - tg110->max_v_total = CRTC_V_TOTAL__CRTC_V_TOTAL_MASK + 1; - - tg110->min_h_blank = 56; - tg110->min_h_front_porch = 4; - tg110->min_h_back_porch = 4; -} - -void dce80_timing_generator_enable_advanced_request( +static void dce80_timing_generator_enable_advanced_request( struct timing_generator *tg, bool enable, const struct dc_crtc_timing *timing) @@ -237,3 +176,67 @@ void dce80_timing_generator_enable_advanced_request( dm_write_reg(tg->ctx, addr, value); } + +static const struct timing_generator_funcs dce80_tg_funcs = { + .validate_timing = dce110_tg_validate_timing, + .program_timing = program_timing, + .enable_crtc = dce110_timing_generator_enable_crtc, + .disable_crtc = dce110_timing_generator_disable_crtc, + .is_counter_moving = dce110_timing_generator_is_counter_moving, + .get_position = dce110_timing_generator_get_position, + .get_frame_count = dce110_timing_generator_get_vblank_counter, + .get_scanoutpos = dce110_timing_generator_get_crtc_scanoutpos, + .set_early_control = dce110_timing_generator_set_early_control, + .wait_for_state = dce110_tg_wait_for_state, + .set_blank = dce110_tg_set_blank, + .is_blanked = dce110_tg_is_blanked, + .set_colors = dce110_tg_set_colors, + .set_overscan_blank_color = + dce110_timing_generator_set_overscan_color_black, + .set_blank_color = dce110_timing_generator_program_blank_color, + .disable_vga = dce110_timing_generator_disable_vga, + .did_triggered_reset_occur = + dce110_timing_generator_did_triggered_reset_occur, + .setup_global_swap_lock = + dce110_timing_generator_setup_global_swap_lock, + .enable_reset_trigger = dce110_timing_generator_enable_reset_trigger, + .disable_reset_trigger = dce110_timing_generator_disable_reset_trigger, + .tear_down_global_swap_lock = + dce110_timing_generator_tear_down_global_swap_lock, + .set_drr = dce110_timing_generator_set_drr, + .set_static_screen_control = + dce110_timing_generator_set_static_screen_control, + .set_test_pattern = dce110_timing_generator_set_test_pattern, + .arm_vert_intr = dce110_arm_vert_intr, + + /* DCE8.0 overrides */ + .enable_advanced_request = + dce80_timing_generator_enable_advanced_request, + .configure_crc = dce110_configure_crc, + .get_crc = dce110_get_crc, +}; + +void dce80_timing_generator_construct( + struct dce110_timing_generator *tg110, + struct dc_context *ctx, + uint32_t instance, + const struct dce110_timing_generator_offsets *offsets) +{ + tg110->controller_id = CONTROLLER_ID_D0 + instance; + tg110->base.inst = instance; + tg110->offsets = *offsets; + tg110->derived_offsets = reg_offsets[instance]; + + tg110->base.funcs = &dce80_tg_funcs; + + tg110->base.ctx = ctx; + tg110->base.bp = ctx->dc_bios; + + tg110->max_h_total = CRTC_H_TOTAL__CRTC_H_TOTAL_MASK + 1; + tg110->max_v_total = CRTC_V_TOTAL__CRTC_V_TOTAL_MASK + 1; + + tg110->min_h_blank = 56; + tg110->min_h_front_porch = 4; + tg110->min_h_back_porch = 4; +} +
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.h b/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.h index 9cebb24c..8ff1b06 100644 --- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.h
@@ -36,10 +36,4 @@ void dce80_timing_generator_construct( uint32_t instance, const struct dce110_timing_generator_offsets *offsets); -/******** HW programming ************/ -void dce80_timing_generator_enable_advanced_request( - struct timing_generator *tg, - bool enable, - const struct dc_crtc_timing *timing); - #endif /* __DC_TIMING_GENERATOR_DCE80_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c index 53ba360..881a1bf 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
@@ -232,10 +232,11 @@ bool cm_helper_convert_to_custom_float( return true; } - +/* driver uses 32 regions or less, but DCN HW has 34, extra 2 are set to 0 */ #define MAX_REGIONS_NUMBER 34 #define MAX_LOW_POINT 25 -#define NUMBER_SEGMENTS 32 +#define NUMBER_REGIONS 32 +#define NUMBER_SW_SEGMENTS 16 bool cm_helper_translate_curve_to_hw_format( const struct dc_transfer_func *output_tf, @@ -251,7 +252,7 @@ bool cm_helper_translate_curve_to_hw_format( struct fixed31_32 y1_min; struct fixed31_32 y3_max; - int32_t segment_start, segment_end; + int32_t region_start, region_end; int32_t i; uint32_t j, k, seg_distr[MAX_REGIONS_NUMBER], increment, start_index, hw_points; @@ -271,11 +272,11 @@ bool cm_helper_translate_curve_to_hw_format( /* 32 segments * segments are from 2^-25 to 2^7 */ - for (i = 0; i < 32 ; i++) + for (i = 0; i < NUMBER_REGIONS ; i++) seg_distr[i] = 3; - segment_start = -25; - segment_end = 7; + region_start = -MAX_LOW_POINT; + region_end = NUMBER_REGIONS - MAX_LOW_POINT; } else { /* 10 segments * segment is from 2^-10 to 2^0 @@ -289,14 +290,14 @@ bool cm_helper_translate_curve_to_hw_format( seg_distr[5] = 4; seg_distr[6] = 4; seg_distr[7] = 4; - seg_distr[8] = 5; - seg_distr[9] = 5; + seg_distr[8] = 4; + seg_distr[9] = 4; - segment_start = -10; - segment_end = 0; + region_start = -10; + region_end = 0; } - for (i = segment_end - segment_start; i < MAX_REGIONS_NUMBER ; i++) + for (i = region_end - region_start; i < MAX_REGIONS_NUMBER ; i++) seg_distr[i] = -1; for (k = 0; k < MAX_REGIONS_NUMBER; k++) { @@ -305,10 +306,12 @@ bool cm_helper_translate_curve_to_hw_format( } j = 0; - for (k = 0; k < (segment_end - segment_start); k++) { - increment = NUMBER_SEGMENTS / (1 << seg_distr[k]); - start_index = (segment_start + k + MAX_LOW_POINT) * NUMBER_SEGMENTS; - for (i = start_index; i < start_index + NUMBER_SEGMENTS; i += increment) { + for (k = 0; k < (region_end - region_start); k++) { + increment = NUMBER_SW_SEGMENTS / (1 << seg_distr[k]); + start_index = (region_start + k + MAX_LOW_POINT) * + NUMBER_SW_SEGMENTS; + for (i = start_index; i < start_index + NUMBER_SW_SEGMENTS; + i += increment) { if (j == hw_points - 1) break; rgb_resulted[j].red = output_tf->tf_pts.red[i]; @@ -319,15 +322,15 @@ bool cm_helper_translate_curve_to_hw_format( } /* last point */ - start_index = (segment_end + MAX_LOW_POINT) * NUMBER_SEGMENTS; + start_index = (region_end + MAX_LOW_POINT) * NUMBER_SW_SEGMENTS; rgb_resulted[hw_points - 1].red = output_tf->tf_pts.red[start_index]; rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index]; rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index]; arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2), - dal_fixed31_32_from_int(segment_start)); + dal_fixed31_32_from_int(region_start)); arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2), - dal_fixed31_32_from_int(segment_end)); + dal_fixed31_32_from_int(region_end)); y_r = rgb_resulted[0].red; y_g = rgb_resulted[0].green; @@ -413,3 +416,156 @@ bool cm_helper_translate_curve_to_hw_format( return true; } + +#define NUM_DEGAMMA_REGIONS 12 + + +bool cm_helper_translate_curve_to_degamma_hw_format( + const struct dc_transfer_func *output_tf, + struct pwl_params *lut_params) +{ + struct curve_points *arr_points; + struct pwl_result_data *rgb_resulted; + struct pwl_result_data *rgb; + struct pwl_result_data *rgb_plus_1; + struct fixed31_32 y_r; + struct fixed31_32 y_g; + struct fixed31_32 y_b; + struct fixed31_32 y1_min; + struct fixed31_32 y3_max; + + int32_t region_start, region_end; + int32_t i; + uint32_t j, k, seg_distr[MAX_REGIONS_NUMBER], increment, start_index, hw_points; + + if (output_tf == NULL || lut_params == NULL || output_tf->type == TF_TYPE_BYPASS) + return false; + + PERF_TRACE(); + + arr_points = lut_params->arr_points; + rgb_resulted = lut_params->rgb_resulted; + hw_points = 0; + + memset(lut_params, 0, sizeof(struct pwl_params)); + memset(seg_distr, 0, sizeof(seg_distr)); + + region_start = -NUM_DEGAMMA_REGIONS; + region_end = 0; + + + for (i = region_end - region_start; i < MAX_REGIONS_NUMBER ; i++) + seg_distr[i] = -1; + /* 12 segments + * segments are from 2^-12 to 0 + */ + for (i = 0; i < NUM_DEGAMMA_REGIONS ; i++) + seg_distr[i] = 4; + + for (k = 0; k < MAX_REGIONS_NUMBER; k++) { + if (seg_distr[k] != -1) + hw_points += (1 << seg_distr[k]); + } + + j = 0; + for (k = 0; k < (region_end - region_start); k++) { + increment = NUMBER_SW_SEGMENTS / (1 << seg_distr[k]); + start_index = (region_start + k + MAX_LOW_POINT) * + NUMBER_SW_SEGMENTS; + for (i = start_index; i < start_index + NUMBER_SW_SEGMENTS; + i += increment) { + if (j == hw_points - 1) + break; + rgb_resulted[j].red = output_tf->tf_pts.red[i]; + rgb_resulted[j].green = output_tf->tf_pts.green[i]; + rgb_resulted[j].blue = output_tf->tf_pts.blue[i]; + j++; + } + } + + /* last point */ + start_index = (region_end + MAX_LOW_POINT) * NUMBER_SW_SEGMENTS; + rgb_resulted[hw_points - 1].red = output_tf->tf_pts.red[start_index]; + rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index]; + rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index]; + + arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2), + dal_fixed31_32_from_int(region_start)); + arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2), + dal_fixed31_32_from_int(region_end)); + + y_r = rgb_resulted[0].red; + y_g = rgb_resulted[0].green; + y_b = rgb_resulted[0].blue; + + y1_min = dal_fixed31_32_min(y_r, dal_fixed31_32_min(y_g, y_b)); + + arr_points[0].y = y1_min; + arr_points[0].slope = dal_fixed31_32_div(arr_points[0].y, arr_points[0].x); + y_r = rgb_resulted[hw_points - 1].red; + y_g = rgb_resulted[hw_points - 1].green; + y_b = rgb_resulted[hw_points - 1].blue; + + /* see comment above, m_arrPoints[1].y should be the Y value for the + * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1) + */ + y3_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b)); + + arr_points[1].y = y3_max; + + arr_points[1].slope = dal_fixed31_32_zero; + + if (output_tf->tf == TRANSFER_FUNCTION_PQ) { + /* for PQ, we want to have a straight line from last HW X point, + * and the slope to be such that we hit 1.0 at 10000 nits. + */ + const struct fixed31_32 end_value = + dal_fixed31_32_from_int(125); + + arr_points[1].slope = dal_fixed31_32_div( + dal_fixed31_32_sub(dal_fixed31_32_one, arr_points[1].y), + dal_fixed31_32_sub(end_value, arr_points[1].x)); + } + + lut_params->hw_points_num = hw_points; + + i = 1; + for (k = 0; k < MAX_REGIONS_NUMBER && i < MAX_REGIONS_NUMBER; k++) { + if (seg_distr[k] != -1) { + lut_params->arr_curve_points[k].segments_num = + seg_distr[k]; + lut_params->arr_curve_points[i].offset = + lut_params->arr_curve_points[k].offset + (1 << seg_distr[k]); + } + i++; + } + + if (seg_distr[k] != -1) + lut_params->arr_curve_points[k].segments_num = seg_distr[k]; + + rgb = rgb_resulted; + rgb_plus_1 = rgb_resulted + 1; + + i = 1; + while (i != hw_points + 1) { + if (dal_fixed31_32_lt(rgb_plus_1->red, rgb->red)) + rgb_plus_1->red = rgb->red; + if (dal_fixed31_32_lt(rgb_plus_1->green, rgb->green)) + rgb_plus_1->green = rgb->green; + if (dal_fixed31_32_lt(rgb_plus_1->blue, rgb->blue)) + rgb_plus_1->blue = rgb->blue; + + rgb->delta_red = dal_fixed31_32_sub(rgb_plus_1->red, rgb->red); + rgb->delta_green = dal_fixed31_32_sub(rgb_plus_1->green, rgb->green); + rgb->delta_blue = dal_fixed31_32_sub(rgb_plus_1->blue, rgb->blue); + + ++rgb_plus_1; + ++rgb; + ++i; + } + cm_helper_convert_to_custom_float(rgb_resulted, + lut_params->arr_points, + hw_points, false); + + return true; +}
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.h index 64e476b8..7a531b0 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.h
@@ -106,4 +106,9 @@ bool cm_helper_translate_curve_to_hw_format( const struct dc_transfer_func *output_tf, struct pwl_params *lut_params, bool fixpoint); +bool cm_helper_translate_curve_to_degamma_hw_format( + const struct dc_transfer_func *output_tf, + struct pwl_params *lut_params); + + #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c index f2a08b1..e305c28 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -196,7 +196,7 @@ static void dpp1_cm_set_regamma_pwl( case OPP_REGAMMA_SRGB: re_mode = 1; break; - case OPP_REGAMMA_3_6: + case OPP_REGAMMA_XVYCC: re_mode = 2; break; case OPP_REGAMMA_USER: @@ -424,6 +424,24 @@ void dpp1_set_cursor_position( } +void dpp1_dppclk_control( + struct dpp *dpp_base, + bool dppclk_div, + bool enable) +{ + struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base); + + if (enable) { + if (dpp->tf_mask->DPPCLK_RATE_CONTROL) + REG_UPDATE_2(DPP_CONTROL, + DPPCLK_RATE_CONTROL, dppclk_div, + DPP_CLOCK_ENABLE, 1); + else + REG_UPDATE(DPP_CONTROL, DPP_CLOCK_ENABLE, 1); + } else + REG_UPDATE(DPP_CONTROL, DPP_CLOCK_ENABLE, 0); +} + static const struct dpp_funcs dcn10_dpp_funcs = { .dpp_reset = dpp_reset, .dpp_set_scaler = dpp1_dscl_set_scaler_manual_scale, @@ -445,6 +463,8 @@ static const struct dpp_funcs dcn10_dpp_funcs = { .dpp_full_bypass = dpp1_full_bypass, .set_cursor_attributes = dpp1_set_cursor_attributes, .set_cursor_position = dpp1_set_cursor_position, + .dpp_dppclk_control = dpp1_dppclk_control, + .dpp_set_hdr_multiplier = dpp1_set_hdr_multiplier, }; static struct dpp_caps dcn10_dpp_cap = {
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h index f56ee4d..17b062a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
@@ -112,7 +112,9 @@ SRI(CNVC_SURFACE_PIXEL_FORMAT, CNVC_CFG, id), \ SRI(CURSOR0_CONTROL, CNVC_CUR, id), \ SRI(CURSOR0_COLOR0, CNVC_CUR, id), \ - SRI(CURSOR0_COLOR1, CNVC_CUR, id) + SRI(CURSOR0_COLOR1, CNVC_CUR, id), \ + SRI(DPP_CONTROL, DPP_TOP, id), \ + SRI(CM_HDR_MULT_COEF, CM, id) @@ -306,7 +308,9 @@ TF_SF(CNVC_CUR0_CURSOR0_CONTROL, CUR0_EXPANSION_MODE, mask_sh), \ TF_SF(CNVC_CUR0_CURSOR0_CONTROL, CUR0_ENABLE, mask_sh), \ TF_SF(CNVC_CUR0_CURSOR0_COLOR0, CUR0_COLOR0, mask_sh), \ - TF_SF(CNVC_CUR0_CURSOR0_COLOR1, CUR0_COLOR1, mask_sh) + TF_SF(CNVC_CUR0_CURSOR0_COLOR1, CUR0_COLOR1, mask_sh), \ + TF_SF(DPP_TOP0_DPP_CONTROL, DPP_CLOCK_ENABLE, mask_sh), \ + TF_SF(CM0_CM_HDR_MULT_COEF, CM_HDR_MULT_COEF, mask_sh) #define TF_REG_LIST_SH_MASK_DCN10(mask_sh)\ TF_REG_LIST_SH_MASK_DCN(mask_sh),\ @@ -410,7 +414,8 @@ TF_SF(CURSOR0_CURSOR_CONTROL, CURSOR_MODE, mask_sh), \ TF_SF(CURSOR0_CURSOR_CONTROL, CURSOR_PITCH, mask_sh), \ TF_SF(CURSOR0_CURSOR_CONTROL, CURSOR_LINES_PER_CHUNK, mask_sh), \ - TF_SF(CURSOR0_CURSOR_CONTROL, CURSOR_ENABLE, mask_sh) + TF_SF(CURSOR0_CURSOR_CONTROL, CURSOR_ENABLE, mask_sh), \ + TF_SF(DPP_TOP0_DPP_CONTROL, DPPCLK_RATE_CONTROL, mask_sh) #define TF_REG_FIELD_LIST(type) \ type EXT_OVERSCAN_LEFT; \ @@ -1007,7 +1012,10 @@ type CM_BYPASS; \ type FORMAT_CONTROL__ALPHA_EN; \ type CUR0_COLOR0; \ - type CUR0_COLOR1; + type CUR0_COLOR1; \ + type DPPCLK_RATE_CONTROL; \ + type DPP_CLOCK_ENABLE; \ + type CM_HDR_MULT_COEF; struct dcn_dpp_shift { TF_REG_FIELD_LIST(uint8_t) @@ -1252,7 +1260,9 @@ struct dcn_dpp_mask { uint32_t CURSOR_CONTROL; \ uint32_t CURSOR0_CONTROL; \ uint32_t CURSOR0_COLOR0; \ - uint32_t CURSOR0_COLOR1; + uint32_t CURSOR0_COLOR1; \ + uint32_t DPP_CONTROL; \ + uint32_t CM_HDR_MULT_COEF; struct dcn_dpp_registers { DPP_COMMON_REG_VARIABLE_LIST @@ -1287,6 +1297,12 @@ void dpp1_set_cursor_attributes( struct dpp *dpp_base, enum dc_cursor_color_format color_format); +void dpp1_set_cursor_position( + struct dpp *dpp_base, + const struct dc_cursor_position *pos, + const struct dc_cursor_mi_param *param, + uint32_t width); + bool dpp1_dscl_is_lb_conf_valid( int ceil_vratio, int num_partitions, @@ -1397,6 +1413,15 @@ void dpp1_cnv_setup ( void dpp1_full_bypass(struct dpp *dpp_base); +void dpp1_dppclk_control( + struct dpp *dpp_base, + bool dppclk_div, + bool enable); + +void dpp1_set_hdr_multiplier( + struct dpp *dpp_base, + uint32_t multiplier); + void dpp1_construct(struct dcn10_dpp *dpp1, struct dc_context *ctx, uint32_t inst,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c index a5b0990..fb32975e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
@@ -193,6 +193,7 @@ void dpp1_cm_set_gamut_remap( const struct dpp_grph_csc_adjustment *adjust) { struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base); + int i = 0; if (adjust->gamut_adjust_type != GRAPHICS_GAMUT_ADJUST_TYPE_SW) /* Bypass if type is bypass or hw */ @@ -201,20 +202,8 @@ void dpp1_cm_set_gamut_remap( struct fixed31_32 arr_matrix[12]; uint16_t arr_reg_val[12]; - arr_matrix[0] = adjust->temperature_matrix[0]; - arr_matrix[1] = adjust->temperature_matrix[1]; - arr_matrix[2] = adjust->temperature_matrix[2]; - arr_matrix[3] = dal_fixed31_32_zero; - - arr_matrix[4] = adjust->temperature_matrix[3]; - arr_matrix[5] = adjust->temperature_matrix[4]; - arr_matrix[6] = adjust->temperature_matrix[5]; - arr_matrix[7] = dal_fixed31_32_zero; - - arr_matrix[8] = adjust->temperature_matrix[6]; - arr_matrix[9] = adjust->temperature_matrix[7]; - arr_matrix[10] = adjust->temperature_matrix[8]; - arr_matrix[11] = dal_fixed31_32_zero; + for (i = 0; i < 12; i++) + arr_matrix[i] = adjust->temperature_matrix[i]; convert_float_matrix( arr_reg_val, arr_matrix, 12); @@ -309,6 +298,32 @@ static void dpp1_cm_get_reg_field( reg->masks.exp_resion_start_segment = dpp->tf_mask->CM_RGAM_RAMB_EXP_REGION_START_SEGMENT_B; } +static void dpp1_cm_get_degamma_reg_field( + struct dcn10_dpp *dpp, + struct xfer_func_reg *reg) +{ + reg->shifts.exp_region0_lut_offset = dpp->tf_shift->CM_DGAM_RAMA_EXP_REGION0_LUT_OFFSET; + reg->masks.exp_region0_lut_offset = dpp->tf_mask->CM_DGAM_RAMA_EXP_REGION0_LUT_OFFSET; + reg->shifts.exp_region0_num_segments = dpp->tf_shift->CM_DGAM_RAMA_EXP_REGION0_NUM_SEGMENTS; + reg->masks.exp_region0_num_segments = dpp->tf_mask->CM_DGAM_RAMA_EXP_REGION0_NUM_SEGMENTS; + reg->shifts.exp_region1_lut_offset = dpp->tf_shift->CM_DGAM_RAMA_EXP_REGION1_LUT_OFFSET; + reg->masks.exp_region1_lut_offset = dpp->tf_mask->CM_DGAM_RAMA_EXP_REGION1_LUT_OFFSET; + reg->shifts.exp_region1_num_segments = dpp->tf_shift->CM_DGAM_RAMA_EXP_REGION1_NUM_SEGMENTS; + reg->masks.exp_region1_num_segments = dpp->tf_mask->CM_DGAM_RAMA_EXP_REGION1_NUM_SEGMENTS; + + reg->shifts.field_region_end = dpp->tf_shift->CM_DGAM_RAMB_EXP_REGION_END_B; + reg->masks.field_region_end = dpp->tf_mask->CM_DGAM_RAMB_EXP_REGION_END_B; + reg->shifts.field_region_end_slope = dpp->tf_shift->CM_DGAM_RAMB_EXP_REGION_END_SLOPE_B; + reg->masks.field_region_end_slope = dpp->tf_mask->CM_DGAM_RAMB_EXP_REGION_END_SLOPE_B; + reg->shifts.field_region_end_base = dpp->tf_shift->CM_DGAM_RAMB_EXP_REGION_END_BASE_B; + reg->masks.field_region_end_base = dpp->tf_mask->CM_DGAM_RAMB_EXP_REGION_END_BASE_B; + reg->shifts.field_region_linear_slope = dpp->tf_shift->CM_DGAM_RAMB_EXP_REGION_LINEAR_SLOPE_B; + reg->masks.field_region_linear_slope = dpp->tf_mask->CM_DGAM_RAMB_EXP_REGION_LINEAR_SLOPE_B; + reg->shifts.exp_region_start = dpp->tf_shift->CM_DGAM_RAMB_EXP_REGION_START_B; + reg->masks.exp_region_start = dpp->tf_mask->CM_DGAM_RAMB_EXP_REGION_START_B; + reg->shifts.exp_resion_start_segment = dpp->tf_shift->CM_DGAM_RAMB_EXP_REGION_START_SEGMENT_B; + reg->masks.exp_resion_start_segment = dpp->tf_mask->CM_DGAM_RAMB_EXP_REGION_START_SEGMENT_B; +} void dpp1_cm_set_output_csc_adjustment( struct dpp *dpp_base, const uint16_t *regval) @@ -513,7 +528,7 @@ void dpp1_program_degamma_lutb_settings( struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base); struct xfer_func_reg gam_regs; - dpp1_cm_get_reg_field(dpp, &gam_regs); + dpp1_cm_get_degamma_reg_field(dpp, &gam_regs); gam_regs.start_cntl_b = REG(CM_DGAM_RAMB_START_CNTL_B); gam_regs.start_cntl_g = REG(CM_DGAM_RAMB_START_CNTL_G); @@ -542,7 +557,7 @@ void dpp1_program_degamma_luta_settings( struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base); struct xfer_func_reg gam_regs; - dpp1_cm_get_reg_field(dpp, &gam_regs); + dpp1_cm_get_degamma_reg_field(dpp, &gam_regs); gam_regs.start_cntl_b = REG(CM_DGAM_RAMA_START_CNTL_B); gam_regs.start_cntl_g = REG(CM_DGAM_RAMA_START_CNTL_G); @@ -789,3 +804,12 @@ void dpp1_program_input_lut( REG_UPDATE(CM_IGAM_CONTROL, CM_IGAM_LUT_MODE, rama_occupied ? 3 : 2); REG_GET(CM_IGAM_CONTROL, CM_IGAM_LUT_MODE, &ram_num); } + +void dpp1_set_hdr_multiplier( + struct dpp *dpp_base, + uint32_t multiplier) +{ + struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base); + + REG_UPDATE(CM_HDR_MULT_COEF, CM_HDR_MULT_COEF, multiplier); +}
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c index eb83171..738f67f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
@@ -30,6 +30,8 @@ #define CTX \ hubbub->ctx +#define DC_LOGGER \ + hubbub->ctx->logger #define REG(reg)\ hubbub->regs->reg @@ -100,7 +102,6 @@ bool hubbub1_verify_allow_pstate_change_high( static unsigned int max_sampled_pstate_wait_us; /* data collection */ static bool forced_pstate_allow; /* help with revert wa */ - unsigned int debug_index = 0x7; unsigned int debug_data; unsigned int i; @@ -115,7 +116,9 @@ bool hubbub1_verify_allow_pstate_change_high( forced_pstate_allow = false; } - /* description "3-0: Pipe0 cursor0 QOS + /* RV1: + * dchubbubdebugind, at: 0x7 + * description "3-0: Pipe0 cursor0 QOS * 7-4: Pipe1 cursor0 QOS * 11-8: Pipe2 cursor0 QOS * 15-12: Pipe3 cursor0 QOS @@ -137,7 +140,8 @@ bool hubbub1_verify_allow_pstate_change_high( * 31: SOC pstate change request */ - REG_WRITE(DCHUBBUB_TEST_DEBUG_INDEX, debug_index); + + REG_WRITE(DCHUBBUB_TEST_DEBUG_INDEX, hubbub->debug_test_index_pstate); for (i = 0; i < pstate_wait_timeout_us; i++) { debug_data = REG_READ(DCHUBBUB_TEST_DEBUG_DATA); @@ -145,8 +149,7 @@ bool hubbub1_verify_allow_pstate_change_high( if (debug_data & (1 << 30)) { if (i > pstate_wait_expected_timeout_us) - dm_logger_write(hubbub->ctx->logger, LOG_WARNING, - "pstate took longer than expected ~%dus\n", + DC_LOG_WARNING("pstate took longer than expected ~%dus\n", i); return true; @@ -165,8 +168,7 @@ bool hubbub1_verify_allow_pstate_change_high( DCHUBBUB_ARB_ALLOW_PSTATE_CHANGE_FORCE_ENABLE, 1); forced_pstate_allow = true; - dm_logger_write(hubbub->ctx->logger, LOG_WARNING, - "pstate TEST_DEBUG_DATA: 0x%X\n", + DC_LOG_WARNING("pstate TEST_DEBUG_DATA: 0x%X\n", debug_data); return false; @@ -209,16 +211,14 @@ void hubbub1_program_watermarks( refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "URGENCY_WATERMARK_A calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_A calculated =%d\n" "HW register value = 0x%x\n", watermarks->a.urgent_ns, prog_wm_value); prog_wm_value = convert_and_clamp(watermarks->a.pte_meta_urgent_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "PTE_META_URGENCY_WATERMARK_A calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_A calculated =%d\n" "HW register value = 0x%x\n", watermarks->a.pte_meta_urgent_ns, prog_wm_value); @@ -227,8 +227,7 @@ void hubbub1_program_watermarks( watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "SR_ENTER_EXIT_WATERMARK_A calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_A calculated =%d\n" "HW register value = 0x%x\n", watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); @@ -237,8 +236,7 @@ void hubbub1_program_watermarks( watermarks->a.cstate_pstate.cstate_exit_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "SR_EXIT_WATERMARK_A calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_A calculated =%d\n" "HW register value = 0x%x\n", watermarks->a.cstate_pstate.cstate_exit_ns, prog_wm_value); } @@ -247,8 +245,7 @@ void hubbub1_program_watermarks( watermarks->a.cstate_pstate.pstate_change_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "DRAM_CLK_CHANGE_WATERMARK_A calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_A calculated =%d\n" "HW register value = 0x%x\n\n", watermarks->a.cstate_pstate.pstate_change_ns, prog_wm_value); @@ -257,8 +254,7 @@ void hubbub1_program_watermarks( prog_wm_value = convert_and_clamp( watermarks->b.urgent_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "URGENCY_WATERMARK_B calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_B calculated =%d\n" "HW register value = 0x%x\n", watermarks->b.urgent_ns, prog_wm_value); @@ -267,8 +263,7 @@ void hubbub1_program_watermarks( watermarks->b.pte_meta_urgent_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "PTE_META_URGENCY_WATERMARK_B calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_B calculated =%d\n" "HW register value = 0x%x\n", watermarks->b.pte_meta_urgent_ns, prog_wm_value); @@ -278,8 +273,7 @@ void hubbub1_program_watermarks( watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "SR_ENTER_WATERMARK_B calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_B calculated =%d\n" "HW register value = 0x%x\n", watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); @@ -288,8 +282,7 @@ void hubbub1_program_watermarks( watermarks->b.cstate_pstate.cstate_exit_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "SR_EXIT_WATERMARK_B calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_B calculated =%d\n" "HW register value = 0x%x\n", watermarks->b.cstate_pstate.cstate_exit_ns, prog_wm_value); } @@ -298,8 +291,7 @@ void hubbub1_program_watermarks( watermarks->b.cstate_pstate.pstate_change_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "DRAM_CLK_CHANGE_WATERMARK_B calculated =%d\n\n" + DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_B calculated =%d\n\n" "HW register value = 0x%x\n", watermarks->b.cstate_pstate.pstate_change_ns, prog_wm_value); @@ -307,8 +299,7 @@ void hubbub1_program_watermarks( prog_wm_value = convert_and_clamp( watermarks->c.urgent_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "URGENCY_WATERMARK_C calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_C calculated =%d\n" "HW register value = 0x%x\n", watermarks->c.urgent_ns, prog_wm_value); @@ -317,8 +308,7 @@ void hubbub1_program_watermarks( watermarks->c.pte_meta_urgent_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "PTE_META_URGENCY_WATERMARK_C calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_C calculated =%d\n" "HW register value = 0x%x\n", watermarks->c.pte_meta_urgent_ns, prog_wm_value); @@ -328,8 +318,7 @@ void hubbub1_program_watermarks( watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "SR_ENTER_WATERMARK_C calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_C calculated =%d\n" "HW register value = 0x%x\n", watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); @@ -338,8 +327,7 @@ void hubbub1_program_watermarks( watermarks->c.cstate_pstate.cstate_exit_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "SR_EXIT_WATERMARK_C calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_C calculated =%d\n" "HW register value = 0x%x\n", watermarks->c.cstate_pstate.cstate_exit_ns, prog_wm_value); } @@ -348,8 +336,7 @@ void hubbub1_program_watermarks( watermarks->c.cstate_pstate.pstate_change_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "DRAM_CLK_CHANGE_WATERMARK_C calculated =%d\n\n" + DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_C calculated =%d\n\n" "HW register value = 0x%x\n", watermarks->c.cstate_pstate.pstate_change_ns, prog_wm_value); @@ -357,8 +344,7 @@ void hubbub1_program_watermarks( prog_wm_value = convert_and_clamp( watermarks->d.urgent_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "URGENCY_WATERMARK_D calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_D calculated =%d\n" "HW register value = 0x%x\n", watermarks->d.urgent_ns, prog_wm_value); @@ -366,8 +352,7 @@ void hubbub1_program_watermarks( watermarks->d.pte_meta_urgent_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "PTE_META_URGENCY_WATERMARK_D calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_D calculated =%d\n" "HW register value = 0x%x\n", watermarks->d.pte_meta_urgent_ns, prog_wm_value); @@ -377,8 +362,7 @@ void hubbub1_program_watermarks( watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "SR_ENTER_WATERMARK_D calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_D calculated =%d\n" "HW register value = 0x%x\n", watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); @@ -387,8 +371,7 @@ void hubbub1_program_watermarks( watermarks->d.cstate_pstate.cstate_exit_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "SR_EXIT_WATERMARK_D calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_D calculated =%d\n" "HW register value = 0x%x\n", watermarks->d.cstate_pstate.cstate_exit_ns, prog_wm_value); } @@ -398,8 +381,7 @@ void hubbub1_program_watermarks( watermarks->d.cstate_pstate.pstate_change_ns, refclk_mhz, 0x1fffff); REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, prog_wm_value); - dm_logger_write(hubbub->ctx->logger, LOG_BANDWIDTH_CALCS, - "DRAM_CLK_CHANGE_WATERMARK_D calculated =%d\n" + DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_D calculated =%d\n" "HW register value = 0x%x\n\n", watermarks->d.cstate_pstate.pstate_change_ns, prog_wm_value); @@ -512,5 +494,6 @@ void hubbub1_construct(struct hubbub *hubbub, hubbub->shifts = hubbub_shift; hubbub->masks = hubbub_mask; + hubbub->debug_test_index_pstate = 0x7; }
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h index d5c9784..a16e908 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
@@ -185,6 +185,7 @@ struct hubbub { const struct dcn_hubbub_registers *regs; const struct dcn_hubbub_shift *shifts; const struct dcn_hubbub_mask *masks; + unsigned int debug_test_index_pstate; }; void hubbub1_update_dchub(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c index 585b333..39b72f6 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -73,6 +73,9 @@ static void hubp1_disconnect(struct hubp *hubp) REG_UPDATE(DCHUBP_CNTL, HUBP_TTU_DISABLE, 1); + + REG_UPDATE(CURSOR_CONTROL, + CURSOR_ENABLE, 0); } static void hubp1_set_hubp_blank_en(struct hubp *hubp, bool blank) @@ -296,8 +299,9 @@ bool hubp1_program_surface_flip_and_addr( if (address->grph.addr.quad_part == 0) break; - REG_UPDATE(DCSURF_SURFACE_CONTROL, - PRIMARY_SURFACE_TMZ, address->tmz_surface); + REG_UPDATE_2(DCSURF_SURFACE_CONTROL, + PRIMARY_SURFACE_TMZ, address->tmz_surface, + PRIMARY_META_SURFACE_TMZ, address->tmz_surface); if (address->grph.meta_addr.quad_part != 0) { REG_SET(DCSURF_PRIMARY_META_SURFACE_ADDRESS_HIGH, 0, @@ -322,8 +326,11 @@ bool hubp1_program_surface_flip_and_addr( || address->video_progressive.chroma_addr.quad_part == 0) break; - REG_UPDATE(DCSURF_SURFACE_CONTROL, - PRIMARY_SURFACE_TMZ, address->tmz_surface); + REG_UPDATE_4(DCSURF_SURFACE_CONTROL, + PRIMARY_SURFACE_TMZ, address->tmz_surface, + PRIMARY_SURFACE_TMZ_C, address->tmz_surface, + PRIMARY_META_SURFACE_TMZ, address->tmz_surface, + PRIMARY_META_SURFACE_TMZ_C, address->tmz_surface); if (address->video_progressive.luma_meta_addr.quad_part != 0) { REG_SET(DCSURF_PRIMARY_META_SURFACE_ADDRESS_HIGH_C, 0, @@ -365,8 +372,11 @@ bool hubp1_program_surface_flip_and_addr( if (address->grph_stereo.right_addr.quad_part == 0) break; - REG_UPDATE(DCSURF_SURFACE_CONTROL, - PRIMARY_SURFACE_TMZ, address->tmz_surface); + REG_UPDATE_4(DCSURF_SURFACE_CONTROL, + PRIMARY_SURFACE_TMZ, address->tmz_surface, + PRIMARY_SURFACE_TMZ_C, address->tmz_surface, + PRIMARY_META_SURFACE_TMZ, address->tmz_surface, + PRIMARY_META_SURFACE_TMZ_C, address->tmz_surface); if (address->grph_stereo.right_meta_addr.quad_part != 0) { @@ -909,6 +919,21 @@ void hubp1_cursor_set_position( /* TODO Handle surface pixel formats other than 4:4:4 */ } +void hubp1_clk_cntl(struct hubp *hubp, bool enable) +{ + struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp); + uint32_t clk_enable = enable ? 1 : 0; + + REG_UPDATE(HUBP_CLK_CNTL, HUBP_CLOCK_ENABLE, clk_enable); +} + +void hubp1_vtg_sel(struct hubp *hubp, uint32_t otg_inst) +{ + struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp); + + REG_UPDATE(DCHUBP_CNTL, HUBP_VTG_SEL, otg_inst); +} + static struct hubp_funcs dcn10_hubp_funcs = { .hubp_program_surface_flip_and_addr = hubp1_program_surface_flip_and_addr, @@ -925,6 +950,8 @@ static struct hubp_funcs dcn10_hubp_funcs = { .set_cursor_attributes = hubp1_cursor_set_attributes, .set_cursor_position = hubp1_cursor_set_position, .hubp_disconnect = hubp1_disconnect, + .hubp_clk_cntl = hubp1_clk_cntl, + .hubp_vtg_sel = hubp1_vtg_sel, }; /*****************************************/
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h index 33e91d9..4a3703e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -96,7 +96,8 @@ SRI(DCN_SURF0_TTU_CNTL1, HUBPREQ, id),\ SRI(DCN_SURF1_TTU_CNTL0, HUBPREQ, id),\ SRI(DCN_SURF1_TTU_CNTL1, HUBPREQ, id),\ - SRI(DCN_VM_MX_L1_TLB_CNTL, HUBPREQ, id) + SRI(DCN_VM_MX_L1_TLB_CNTL, HUBPREQ, id),\ + SRI(HUBP_CLK_CNTL, HUBP, id) #define HUBP_REG_LIST_DCN10(id)\ HUBP_REG_LIST_DCN(id),\ @@ -230,7 +231,8 @@ uint32_t CURSOR_CONTROL; \ uint32_t CURSOR_POSITION; \ uint32_t CURSOR_HOT_SPOT; \ - uint32_t CURSOR_DST_OFFSET + uint32_t CURSOR_DST_OFFSET; \ + uint32_t HUBP_CLK_CNTL #define HUBP_SF(reg_name, field_name, post_fix)\ .field_name = reg_name ## __ ## field_name ## post_fix @@ -240,6 +242,7 @@ HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_TTU_DISABLE, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_UNDERFLOW_STATUS, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_NO_OUTSTANDING_REQ, mask_sh),\ + HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_VTG_SEL, mask_sh),\ HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, NUM_PIPES, mask_sh),\ HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, NUM_BANKS, mask_sh),\ HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, PIPE_INTERLEAVE, mask_sh),\ @@ -293,6 +296,9 @@ HUBP_SF(HUBPREQ0_DCSURF_SURFACE_EARLIEST_INUSE_C, SURFACE_EARLIEST_INUSE_ADDRESS_C, mask_sh),\ HUBP_SF(HUBPREQ0_DCSURF_SURFACE_EARLIEST_INUSE_HIGH_C, SURFACE_EARLIEST_INUSE_ADDRESS_HIGH_C, mask_sh),\ HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_SURFACE_TMZ, mask_sh),\ + HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_SURFACE_TMZ_C, mask_sh),\ + HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_META_SURFACE_TMZ, mask_sh),\ + HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_META_SURFACE_TMZ_C, mask_sh),\ HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_SURFACE_DCC_EN, mask_sh),\ HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_SURFACE_DCC_IND_64B_BLK, mask_sh),\ HUBP_SF(HUBPRET0_HUBPRET_CONTROL, DET_BUF_PLANE1_BASE_ADDRESS, mask_sh),\ @@ -352,7 +358,8 @@ HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL0, QoS_RAMP_DISABLE, mask_sh),\ HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL1, REFCYC_PER_REQ_DELIVERY_PRE, mask_sh),\ HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, ENABLE_L1_TLB, mask_sh),\ - HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE, mask_sh) + HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE, mask_sh),\ + HUBP_SF(HUBP0_HUBP_CLK_CNTL, HUBP_CLOCK_ENABLE, mask_sh) #define HUBP_MASK_SH_LIST_DCN10(mask_sh)\ HUBP_MASK_SH_LIST_DCN(mask_sh),\ @@ -398,6 +405,7 @@ type HUBP_BLANK_EN;\ type HUBP_TTU_DISABLE;\ type HUBP_NO_OUTSTANDING_REQ;\ + type HUBP_VTG_SEL;\ type HUBP_UNDERFLOW_STATUS;\ type NUM_PIPES;\ type NUM_BANKS;\ @@ -452,6 +460,13 @@ type SURFACE_EARLIEST_INUSE_ADDRESS_C;\ type SURFACE_EARLIEST_INUSE_ADDRESS_HIGH_C;\ type PRIMARY_SURFACE_TMZ;\ + type PRIMARY_SURFACE_TMZ_C;\ + type SECONDARY_SURFACE_TMZ;\ + type SECONDARY_SURFACE_TMZ_C;\ + type PRIMARY_META_SURFACE_TMZ;\ + type PRIMARY_META_SURFACE_TMZ_C;\ + type SECONDARY_META_SURFACE_TMZ;\ + type SECONDARY_META_SURFACE_TMZ_C;\ type PRIMARY_SURFACE_DCC_EN;\ type PRIMARY_SURFACE_DCC_IND_64B_BLK;\ type DET_BUF_PLANE1_BASE_ADDRESS;\ @@ -524,6 +539,7 @@ type VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR_LSB;\ type ENABLE_L1_TLB;\ type SYSTEM_ACCESS_MODE;\ + type HUBP_CLOCK_ENABLE;\ type MC_VM_SYSTEM_APERTURE_DEFAULT_SYSTEM;\ type MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_MSB;\ type MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB;\ @@ -653,6 +669,9 @@ void min_set_viewport(struct hubp *hubp, const struct rect *viewport, const struct rect *viewport_c); +void hubp1_clk_cntl(struct hubp *hubp, bool enable); +void hubp1_vtg_sel(struct hubp *hubp, uint32_t otg_inst); + void dcn10_hubp_construct( struct dcn10_hubp *hubp1, struct dc_context *ctx,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index bbe8d91..8b0f6b8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -45,6 +45,8 @@ #include "dcn10_hubbub.h" #include "dcn10_cm_common.h" +#define DC_LOGGER \ + ctx->logger #define CTX \ hws->ctx #define REG(reg)\ @@ -133,7 +135,7 @@ void dcn10_log_hw_state(struct dc *dc) DTN_INFO("[%d]:\t %xh \t %xh \t %d \t %d \t " "%xh \t %xh \t %xh \t " "%d \t %d \t %d \t %xh \t", - i, + hubp->inst, s.pixel_format, s.inuse_addr_hi, s.viewport_width, @@ -155,7 +157,7 @@ void dcn10_log_hw_state(struct dc *dc) DTN_INFO("OTG:\t v_bs \t v_be \t v_ss \t v_se \t vpol \t vmax \t vmin \t " "h_bs \t h_be \t h_ss \t h_se \t hpol \t htot \t vtot \t underflow\n"); - for (i = 0; i < pool->res_cap->num_timing_generator; i++) { + for (i = 0; i < pool->timing_generator_count; i++) { struct timing_generator *tg = pool->timing_generators[i]; struct dcn_otg_state s = {0}; @@ -168,7 +170,7 @@ void dcn10_log_hw_state(struct dc *dc) DTN_INFO("[%d]:\t %d \t %d \t %d \t %d \t " "%d \t %d \t %d \t %d \t %d \t %d \t " "%d \t %d \t %d \t %d \t %d \t ", - i, + tg->inst, s.v_blank_start, s.v_blank_end, s.v_sync_a_start, @@ -193,26 +195,6 @@ void dcn10_log_hw_state(struct dc *dc) DTN_INFO_END(); } -static void enable_dppclk( - struct dce_hwseq *hws, - uint8_t plane_id, - uint32_t requested_pix_clk, - bool dppclk_div) -{ - dm_logger_write(hws->ctx->logger, LOG_SURFACE, - "dppclk_rate_control for pipe %d programed to %d\n", - plane_id, - dppclk_div); - - if (hws->shifts->DPPCLK_RATE_CONTROL) - REG_UPDATE_2(DPP_CONTROL[plane_id], - DPPCLK_RATE_CONTROL, dppclk_div, - DPP_CLOCK_ENABLE, 1); - else - REG_UPDATE(DPP_CONTROL[plane_id], - DPP_CLOCK_ENABLE, 1); -} - static void enable_power_gating_plane( struct dce_hwseq *hws, bool enable) @@ -372,6 +354,7 @@ static void power_on_plane( struct dce_hwseq *hws, int plane_id) { + struct dc_context *ctx = hws->ctx; if (REG(DC_IP_REQUEST_CNTL)) { REG_SET(DC_IP_REQUEST_CNTL, 0, IP_REQUEST_EN, 1); @@ -379,7 +362,7 @@ static void power_on_plane( hubp_pg_control(hws, plane_id, true); REG_SET(DC_IP_REQUEST_CNTL, 0, IP_REQUEST_EN, 0); - dm_logger_write(hws->ctx->logger, LOG_DEBUG, + DC_LOG_DEBUG( "Un-gated front end for pipe %d\n", plane_id); } } @@ -570,7 +553,7 @@ static void reset_back_end_for_pipe( struct dc_state *context) { int i; - + struct dc_context *ctx = dc->ctx; if (pipe_ctx->stream_res.stream_enc == NULL) { pipe_ctx->stream = NULL; return; @@ -580,6 +563,22 @@ static void reset_back_end_for_pipe( /* DPMS may already disable */ if (!pipe_ctx->stream->dpms_off) core_link_disable_stream(pipe_ctx, FREE_ACQUIRED_RESOURCE); + else if (pipe_ctx->stream_res.audio) { + /* + * if stream is already disabled outside of commit streams path, + * audio disable was skipped. Need to do it here + */ + pipe_ctx->stream_res.audio->funcs->az_disable(pipe_ctx->stream_res.audio); + + if (dc->caps.dynamic_audio == true) { + /*we have to dynamic arbitrate the audio endpoints*/ + pipe_ctx->stream_res.audio = NULL; + /*we free the resource, need reset is_audio_acquired*/ + update_audio_usage(&dc->current_state->res_ctx, dc->res_pool, pipe_ctx->stream_res.audio, false); + } + + } + } /* by upper caller loop, parent pipe: pipe0, will be reset last. @@ -600,8 +599,7 @@ static void reset_back_end_for_pipe( return; pipe_ctx->stream = NULL; - dm_logger_write(dc->ctx->logger, LOG_DEBUG, - "Reset back end for pipe %d, tg:%d\n", + DC_LOG_DEBUG("Reset back end for pipe %d, tg:%d\n", pipe_ctx->pipe_idx, pipe_ctx->stream_res.tg->inst); } @@ -621,29 +619,22 @@ static void dcn10_verify_allow_pstate_change_high(struct dc *dc) /* trigger HW to start disconnect plane from stream on the next vsync */ static void plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx) { - int fe_idx = pipe_ctx->pipe_idx; - struct hubp *hubp = dc->res_pool->hubps[fe_idx]; + struct hubp *hubp = pipe_ctx->plane_res.hubp; + int dpp_id = pipe_ctx->plane_res.dpp->inst; struct mpc *mpc = dc->res_pool->mpc; - int opp_id; struct mpc_tree *mpc_tree_params; struct mpcc *mpcc_to_remove = NULL; + struct output_pixel_processor *opp = pipe_ctx->stream_res.opp; - /* look at tree rather than mi here to know if we already reset */ - for (opp_id = 0; opp_id < dc->res_pool->pipe_count; opp_id++) { - struct output_pixel_processor *opp = dc->res_pool->opps[opp_id]; - - mpc_tree_params = &(opp->mpc_tree_params); - mpcc_to_remove = mpc->funcs->get_mpcc_for_dpp(mpc_tree_params, fe_idx); - if (mpcc_to_remove != NULL) - break; - } + mpc_tree_params = &(opp->mpc_tree_params); + mpcc_to_remove = mpc->funcs->get_mpcc_for_dpp(mpc_tree_params, dpp_id); /*Already reset*/ - if (opp_id == dc->res_pool->pipe_count) + if (mpcc_to_remove == NULL) return; mpc->funcs->remove_mpcc(mpc, mpc_tree_params, mpcc_to_remove); - dc->res_pool->opps[opp_id]->mpcc_disconnect_pending[fe_idx] = true; + opp->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true; dc->optimized_required = true; @@ -654,21 +645,22 @@ static void plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx) dcn10_verify_allow_pstate_change_high(dc); } -static void plane_atomic_power_down(struct dc *dc, int fe_idx) +static void plane_atomic_power_down(struct dc *dc, struct pipe_ctx *pipe_ctx) { struct dce_hwseq *hws = dc->hwseq; - struct dpp *dpp = dc->res_pool->dpps[fe_idx]; + struct dpp *dpp = pipe_ctx->plane_res.dpp; + struct dc_context *ctx = dc->ctx; if (REG(DC_IP_REQUEST_CNTL)) { REG_SET(DC_IP_REQUEST_CNTL, 0, IP_REQUEST_EN, 1); - dpp_pg_control(hws, fe_idx, false); - hubp_pg_control(hws, fe_idx, false); + dpp_pg_control(hws, dpp->inst, false); + hubp_pg_control(hws, pipe_ctx->plane_res.hubp->inst, false); dpp->funcs->dpp_reset(dpp); REG_SET(DC_IP_REQUEST_CNTL, 0, IP_REQUEST_EN, 0); - dm_logger_write(dc->ctx->logger, LOG_DEBUG, - "Power gated front end %d\n", fe_idx); + DC_LOG_DEBUG( + "Power gated front end %d\n", pipe_ctx->pipe_idx); } } @@ -677,26 +669,25 @@ static void plane_atomic_power_down(struct dc *dc, int fe_idx) */ static void plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx) { - int fe_idx = pipe_ctx->pipe_idx; - struct dce_hwseq *hws = dc->hwseq; - struct hubp *hubp = dc->res_pool->hubps[fe_idx]; + struct hubp *hubp = pipe_ctx->plane_res.hubp; + struct dpp *dpp = pipe_ctx->plane_res.dpp; int opp_id = hubp->opp_id; dc->hwss.wait_for_mpcc_disconnect(dc, dc->res_pool, pipe_ctx); - REG_UPDATE(HUBP_CLK_CNTL[fe_idx], - HUBP_CLOCK_ENABLE, 0); - REG_UPDATE(DPP_CONTROL[fe_idx], - DPP_CLOCK_ENABLE, 0); + hubp->funcs->hubp_clk_cntl(hubp, false); - if (opp_id != 0xf && dc->res_pool->opps[opp_id]->mpc_tree_params.opp_list == NULL) - REG_UPDATE(OPP_PIPE_CONTROL[opp_id], - OPP_PIPE_CLOCK_EN, 0); + dpp->funcs->dpp_dppclk_control(dpp, false, false); + + if (opp_id != 0xf && pipe_ctx->stream_res.opp->mpc_tree_params.opp_list == NULL) + pipe_ctx->stream_res.opp->funcs->opp_pipe_clock_control( + pipe_ctx->stream_res.opp, + false); hubp->power_gated = true; dc->optimized_required = false; /* We're powering off, no need to optimize */ - plane_atomic_power_down(dc, fe_idx); + plane_atomic_power_down(dc, pipe_ctx); pipe_ctx->stream = NULL; memset(&pipe_ctx->stream_res, 0, sizeof(pipe_ctx->stream_res)); @@ -708,15 +699,16 @@ static void plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx) static void dcn10_disable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx) { - if (dc->res_pool->hubps[pipe_ctx->pipe_idx]->power_gated) + struct dc_context *ctx = dc->ctx; + + if (!pipe_ctx->plane_res.hubp || pipe_ctx->plane_res.hubp->power_gated) return; plane_atomic_disable(dc, pipe_ctx); apply_DEGVIDCN10_253_wa(dc); - dm_logger_write(dc->ctx->logger, LOG_DC, - "Power down front end %d\n", + DC_LOG_DC("Power down front end %d\n", pipe_ctx->pipe_idx); } @@ -744,26 +736,25 @@ static void dcn10_init_hw(struct dc *dc) } enable_power_gating_plane(dc->hwseq, true); - return; - } - /* end of FPGA. Below if real ASIC */ + } else { - if (!dcb->funcs->is_accelerated_mode(dcb)) { - bios_golden_init(dc); - disable_vga(dc->hwseq); - } + if (!dcb->funcs->is_accelerated_mode(dcb)) { + bios_golden_init(dc); + disable_vga(dc->hwseq); + } - for (i = 0; i < dc->link_count; i++) { - /* Power up AND update implementation according to the - * required signal (which may be different from the - * default signal on connector). - */ - struct dc_link *link = dc->links[i]; + for (i = 0; i < dc->link_count; i++) { + /* Power up AND update implementation according to the + * required signal (which may be different from the + * default signal on connector). + */ + struct dc_link *link = dc->links[i]; - if (link->link_enc->connector.id == CONNECTOR_ID_EDP) - dc->hwss.edp_power_control(link, true); + if (link->link_enc->connector.id == CONNECTOR_ID_EDP) + dc->hwss.edp_power_control(link, true); - link->link_enc->funcs->hw_init(link->link_enc); + link->link_enc->funcs->hw_init(link->link_enc); + } } for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -792,18 +783,21 @@ static void dcn10_init_hw(struct dc *dc) struct timing_generator *tg = dc->res_pool->timing_generators[i]; struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; struct hubp *hubp = dc->res_pool->hubps[i]; + struct dpp *dpp = dc->res_pool->dpps[i]; pipe_ctx->stream_res.tg = tg; pipe_ctx->pipe_idx = i; pipe_ctx->plane_res.hubp = hubp; - hubp->mpcc_id = i; + pipe_ctx->plane_res.dpp = dpp; + pipe_ctx->plane_res.mpcc_inst = dpp->inst; + hubp->mpcc_id = dpp->inst; hubp->opp_id = 0xf; hubp->power_gated = false; dc->res_pool->opps[i]->mpc_tree_params.opp_id = dc->res_pool->opps[i]->inst; dc->res_pool->opps[i]->mpc_tree_params.opp_list = NULL; - dc->res_pool->opps[i]->mpcc_disconnect_pending[i] = true; + dc->res_pool->opps[i]->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true; pipe_ctx->stream_res.opp = dc->res_pool->opps[i]; plane_atomic_disconnect(dc, pipe_ctx); @@ -828,6 +822,10 @@ static void dcn10_init_hw(struct dc *dc) tg->funcs->tg_init(tg); } + /* end of FPGA. Below if real ASIC */ + if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) + return; + for (i = 0; i < dc->res_pool->audio_count; i++) { struct audio *audio = dc->res_pool->audios[i]; @@ -946,7 +944,10 @@ static bool dcn10_set_input_transfer_func(struct pipe_ctx *pipe_ctx, if (plane_state->in_transfer_func) tf = plane_state->in_transfer_func; - if (plane_state->gamma_correction && dce_use_lut(plane_state->format)) + if (plane_state->gamma_correction && + plane_state->gamma_correction->is_identity) + dpp_base->funcs->dpp_set_degamma(dpp_base, IPP_DEGAMMA_MODE_BYPASS); + else if (plane_state->gamma_correction && dce_use_lut(plane_state->format)) dpp_base->funcs->dpp_program_input_lut(dpp_base, plane_state->gamma_correction); if (tf == NULL) @@ -1017,8 +1018,6 @@ static void dcn10_pipe_control_lock( struct pipe_ctx *pipe, bool lock) { - struct hubp *hubp = NULL; - hubp = dc->res_pool->hubps[pipe->pipe_idx]; /* use TG master update lock to lock everything on the TG * therefore only top pipe need to lock */ @@ -1121,7 +1120,7 @@ static void dcn10_enable_per_frame_crtc_position_reset( DC_SYNC_INFO("Waiting for trigger\n"); - for (i = 1; i < group_size; i++) + for (i = 0; i < group_size; i++) wait_for_reset_trigger_to_occur(dc_ctx, grouped_pipes[i]->stream_res.tg); DC_SYNC_INFO("Multi-display sync is complete\n"); @@ -1131,7 +1130,7 @@ static void dcn10_enable_per_frame_crtc_position_reset( struct dc *core_dc, struct pipe_ctx *pipe_ctx) { - dm_logger_write(core_dc->ctx->logger, LOG_BANDWIDTH_CALCS, + DC_LOG_BANDWIDTH_CALCS(core_dc->ctx->logger, "\n============== DML TTU Output parameters [%d] ==============\n" "qos_level_low_wm: %d, \n" "qos_level_high_wm: %d, \n" @@ -1161,7 +1160,7 @@ static void dcn10_enable_per_frame_crtc_position_reset( pipe_ctx->ttu_regs.refcyc_per_req_delivery_pre_c ); - dm_logger_write(core_dc->ctx->logger, LOG_BANDWIDTH_CALCS, + DC_LOG_BANDWIDTH_CALCS(core_dc->ctx->logger, "\n============== DML DLG Output parameters [%d] ==============\n" "refcyc_h_blank_end: %d, \n" "dlg_vblank_end: %d, \n" @@ -1196,7 +1195,7 @@ static void dcn10_enable_per_frame_crtc_position_reset( pipe_ctx->dlg_regs.refcyc_per_pte_group_nom_l ); - dm_logger_write(core_dc->ctx->logger, LOG_BANDWIDTH_CALCS, + DC_LOG_BANDWIDTH_CALCS(core_dc->ctx->logger, "\ndst_y_per_meta_row_nom_l: %d, \n" "refcyc_per_meta_chunk_nom_l: %d, \n" "refcyc_per_line_delivery_pre_l: %d, \n" @@ -1226,7 +1225,7 @@ static void dcn10_enable_per_frame_crtc_position_reset( pipe_ctx->dlg_regs.refcyc_per_line_delivery_c ); - dm_logger_write(core_dc->ctx->logger, LOG_BANDWIDTH_CALCS, + DC_LOG_BANDWIDTH_CALCS(core_dc->ctx->logger, "\n============== DML RQ Output parameters [%d] ==============\n" "chunk_size: %d \n" "min_chunk_size: %d \n" @@ -1347,19 +1346,19 @@ static void dcn10_enable_plane( undo_DEGVIDCN10_253_wa(dc); power_on_plane(dc->hwseq, - pipe_ctx->pipe_idx); + pipe_ctx->plane_res.hubp->inst); /* enable DCFCLK current DCHUB */ - REG_UPDATE(HUBP_CLK_CNTL[pipe_ctx->pipe_idx], - HUBP_CLOCK_ENABLE, 1); + pipe_ctx->plane_res.hubp->funcs->hubp_clk_cntl(pipe_ctx->plane_res.hubp, true); /* make sure OPP_PIPE_CLOCK_EN = 1 */ - REG_UPDATE(OPP_PIPE_CONTROL[pipe_ctx->stream_res.tg->inst], - OPP_PIPE_CLOCK_EN, 1); + pipe_ctx->stream_res.opp->funcs->opp_pipe_clock_control( + pipe_ctx->stream_res.opp, + true); /* TODO: enable/disable in dm as per update type. if (plane_state) { - dm_logger_write(dc->ctx->logger, LOG_DC, + DC_LOG_DC(dc->ctx->logger, "Pipe:%d 0x%x: addr hi:0x%x, " "addr low:0x%x, " "src: %d, %d, %d," @@ -1377,7 +1376,7 @@ static void dcn10_enable_plane( plane_state->dst_rect.width, plane_state->dst_rect.height); - dm_logger_write(dc->ctx->logger, LOG_DC, + DC_LOG_DC(dc->ctx->logger, "Pipe %d: width, height, x, y format:%d\n" "viewport:%d, %d, %d, %d\n" "recout: %d, %d, %d, %d\n", @@ -1404,6 +1403,7 @@ static void dcn10_enable_plane( static void program_gamut_remap(struct pipe_ctx *pipe_ctx) { + int i = 0; struct dpp_grph_csc_adjustment adjust; memset(&adjust, 0, sizeof(adjust)); adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_BYPASS; @@ -1411,33 +1411,9 @@ static void program_gamut_remap(struct pipe_ctx *pipe_ctx) if (pipe_ctx->stream->gamut_remap_matrix.enable_remap == true) { adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_SW; - adjust.temperature_matrix[0] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[0]; - adjust.temperature_matrix[1] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[1]; - adjust.temperature_matrix[2] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[2]; - adjust.temperature_matrix[3] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[4]; - adjust.temperature_matrix[4] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[5]; - adjust.temperature_matrix[5] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[6]; - adjust.temperature_matrix[6] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[8]; - adjust.temperature_matrix[7] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[9]; - adjust.temperature_matrix[8] = - pipe_ctx->stream-> - gamut_remap_matrix.matrix[10]; + for (i = 0; i < CSC_TEMPERATURE_MATRIX_SIZE; i++) + adjust.temperature_matrix[i] = + pipe_ctx->stream->gamut_remap_matrix.matrix[i]; } pipe_ctx->plane_res.dpp->funcs->dpp_set_gamut_remap(pipe_ctx->plane_res.dpp, &adjust); @@ -1498,7 +1474,7 @@ static bool is_pipe_tree_visible(struct pipe_ctx *pipe_ctx) return false; } -static bool is_rgb_cspace(enum dc_color_space output_color_space) +bool is_rgb_cspace(enum dc_color_space output_color_space) { switch (output_color_space) { case COLOR_SPACE_SRGB: @@ -1636,6 +1612,7 @@ static void update_dpp(struct dpp *dpp, struct dc_plane_state *plane_state) dpp->funcs->dpp_program_bias_and_scale(dpp, &bns_params); } + static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx) { struct hubp *hubp = pipe_ctx->plane_res.hubp; @@ -1726,22 +1703,28 @@ static void update_dchubp_dpp( struct pipe_ctx *pipe_ctx, struct dc_state *context) { - struct dce_hwseq *hws = dc->hwseq; struct hubp *hubp = pipe_ctx->plane_res.hubp; struct dpp *dpp = pipe_ctx->plane_res.dpp; struct dc_plane_state *plane_state = pipe_ctx->plane_state; union plane_size size = plane_state->plane_size; /* depends on DML calculation, DPP clock value may change dynamically */ + /* If request max dpp clk is lower than current dispclk, no need to + * divided by 2 + */ if (plane_state->update_flags.bits.full_update) { - enable_dppclk( - dc->hwseq, - pipe_ctx->pipe_idx, - pipe_ctx->stream_res.pix_clk_params.requested_pix_clk, - context->bw.dcn.calc_clk.dppclk_div); - dc->current_state->bw.dcn.cur_clk.dppclk_div = - context->bw.dcn.calc_clk.dppclk_div; - context->bw.dcn.cur_clk.dppclk_div = context->bw.dcn.calc_clk.dppclk_div; + bool should_divided_by_2 = context->bw.dcn.calc_clk.dppclk_khz <= + context->bw.dcn.cur_clk.dispclk_khz / 2; + + dpp->funcs->dpp_dppclk_control( + dpp, + should_divided_by_2, + true); + + dc->current_state->bw.dcn.cur_clk.dppclk_khz = + should_divided_by_2 ? + context->bw.dcn.cur_clk.dispclk_khz / 2 : + context->bw.dcn.cur_clk.dispclk_khz; } /* TODO: Need input parameter to tell current DCHUB pipe tie to which OTG @@ -1749,7 +1732,7 @@ static void update_dchubp_dpp( * VTG is 1:1 mapping with OTG. Each pipe HUBP will select which VTG */ if (plane_state->update_flags.bits.full_update) { - REG_UPDATE(DCHUBP_CNTL[pipe_ctx->pipe_idx], HUBP_VTG_SEL, pipe_ctx->stream_res.tg->inst); + hubp->funcs->hubp_vtg_sel(hubp, pipe_ctx->stream_res.tg->inst); hubp->funcs->hubp_setup( hubp, @@ -1802,6 +1785,7 @@ static void update_dchubp_dpp( } if (plane_state->update_flags.bits.full_update || + plane_state->update_flags.bits.pixel_format_change || plane_state->update_flags.bits.horizontal_mirror_change || plane_state->update_flags.bits.rotation_change || plane_state->update_flags.bits.swizzle_change || @@ -1826,14 +1810,62 @@ static void update_dchubp_dpp( hubp->funcs->set_blank(hubp, false); } +static void dcn10_otg_blank( + struct dc *dc, + struct stream_resource stream_res, + struct dc_stream_state *stream, + bool blank) +{ + enum dc_color_space color_space; + struct tg_color black_color = {0}; + + /* program otg blank color */ + color_space = stream->output_color_space; + color_space_to_black_color(dc, color_space, &black_color); + + if (stream_res.tg->funcs->set_blank_color) + stream_res.tg->funcs->set_blank_color( + stream_res.tg, + &black_color); + + if (!blank) { + if (stream_res.tg->funcs->set_blank) + stream_res.tg->funcs->set_blank(stream_res.tg, blank); + if (stream_res.abm) + stream_res.abm->funcs->set_abm_level(stream_res.abm, stream->abm_level); + } else if (blank) { + if (stream_res.abm) + stream_res.abm->funcs->set_abm_immediate_disable(stream_res.abm); + if (stream_res.tg->funcs->set_blank) + stream_res.tg->funcs->set_blank(stream_res.tg, blank); + } +} + +static void set_hdr_multiplier(struct pipe_ctx *pipe_ctx) +{ + struct fixed31_32 multiplier = dal_fixed31_32_from_fraction( + pipe_ctx->plane_state->sdr_white_level, 80); + uint32_t hw_mult = 0x1f000; // 1.0 default multiplier + struct custom_float_format fmt; + + fmt.exponenta_bits = 6; + fmt.mantissa_bits = 12; + fmt.sign = true; + + if (pipe_ctx->plane_state->sdr_white_level > 80) + convert_to_custom_float_format(multiplier, &fmt, &hw_mult); + + pipe_ctx->plane_res.dpp->funcs->dpp_set_hdr_multiplier( + pipe_ctx->plane_res.dpp, hw_mult); +} static void program_all_pipe_in_tree( struct dc *dc, struct pipe_ctx *pipe_ctx, struct dc_state *context) { - if (pipe_ctx->top_pipe == NULL) { + bool blank = !is_pipe_tree_visible(pipe_ctx); pipe_ctx->stream_res.tg->dlg_otg_param.vready_offset = pipe_ctx->pipe_dlg_param.vready_offset; pipe_ctx->stream_res.tg->dlg_otg_param.vstartup_start = pipe_ctx->pipe_dlg_param.vstartup_start; @@ -1844,10 +1876,8 @@ static void program_all_pipe_in_tree( pipe_ctx->stream_res.tg->funcs->program_global_sync( pipe_ctx->stream_res.tg); - if (pipe_ctx->stream_res.tg->funcs->set_blank) - pipe_ctx->stream_res.tg->funcs->set_blank( - pipe_ctx->stream_res.tg, - !is_pipe_tree_visible(pipe_ctx)); + dcn10_otg_blank(dc, pipe_ctx->stream_res, + pipe_ctx->stream, blank); } if (pipe_ctx->plane_state != NULL) { @@ -1856,6 +1886,8 @@ static void program_all_pipe_in_tree( update_dchubp_dpp(dc, pipe_ctx, context); + set_hdr_multiplier(pipe_ctx); + if (pipe_ctx->plane_state->update_flags.bits.full_update || pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change || pipe_ctx->plane_state->update_flags.bits.gamma_change) @@ -1871,8 +1903,9 @@ static void program_all_pipe_in_tree( dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream); } - if (pipe_ctx->bottom_pipe != NULL && pipe_ctx->bottom_pipe != pipe_ctx) + if (pipe_ctx->bottom_pipe != NULL && pipe_ctx->bottom_pipe != pipe_ctx) { program_all_pipe_in_tree(dc, pipe_ctx->bottom_pipe, context); + } } static void dcn10_pplib_apply_display_requirements( @@ -1881,16 +1914,10 @@ static void dcn10_pplib_apply_display_requirements( { struct dm_pp_display_configuration *pp_display_cfg = &context->pp_display_cfg; - pp_display_cfg->all_displays_in_sync = false;/*todo*/ - pp_display_cfg->nb_pstate_switch_disable = false; pp_display_cfg->min_engine_clock_khz = context->bw.dcn.cur_clk.dcfclk_khz; pp_display_cfg->min_memory_clock_khz = context->bw.dcn.cur_clk.fclk_khz; pp_display_cfg->min_engine_clock_deep_sleep_khz = context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz; pp_display_cfg->min_dcfc_deep_sleep_clock_khz = context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz; - pp_display_cfg->avail_mclk_switch_time_us = - context->bw.dcn.cur_clk.dram_ccm_us > 0 ? context->bw.dcn.cur_clk.dram_ccm_us : 0; - pp_display_cfg->avail_mclk_switch_time_in_disp_active_us = - context->bw.dcn.cur_clk.min_active_dram_ccm_us > 0 ? context->bw.dcn.cur_clk.min_active_dram_ccm_us : 0; pp_display_cfg->min_dcfclock_khz = context->bw.dcn.cur_clk.dcfclk_khz; pp_display_cfg->disp_clk_khz = context->bw.dcn.cur_clk.dispclk_khz; dce110_fill_display_configs(context, pp_display_cfg); @@ -1953,28 +1980,23 @@ static void dcn10_apply_ctx_for_surface( { int i; struct timing_generator *tg; - struct output_pixel_processor *opp; bool removed_pipe[4] = { false }; unsigned int ref_clk_mhz = dc->res_pool->ref_clock_inKhz/1000; bool program_water_mark = false; - + struct dc_context *ctx = dc->ctx; struct pipe_ctx *top_pipe_to_program = find_top_pipe_for_stream(dc, context, stream); if (!top_pipe_to_program) return; - opp = top_pipe_to_program->stream_res.opp; - tg = top_pipe_to_program->stream_res.tg; - tg->funcs->lock(tg); + dcn10_pipe_control_lock(dc, top_pipe_to_program, true); if (num_planes == 0) { - /* OTG blank before remove all front end */ - if (tg->funcs->set_blank) - tg->funcs->set_blank(tg, true); + dcn10_otg_blank(dc, top_pipe_to_program->stream_res, top_pipe_to_program->stream, true); } /* Disconnect unused mpcc */ @@ -1992,7 +2014,7 @@ static void dcn10_apply_ctx_for_surface( if (old_pipe_ctx->stream_res.tg == tg && old_pipe_ctx->plane_res.hubp && old_pipe_ctx->plane_res.hubp->opp_id != 0xf) { - dcn10_disable_plane(dc, pipe_ctx); + dcn10_disable_plane(dc, old_pipe_ctx); /* * power down fe will unlock when calling reset, need * to lock it back here. Messy, need rework. @@ -2008,7 +2030,7 @@ static void dcn10_apply_ctx_for_surface( plane_atomic_disconnect(dc, old_pipe_ctx); removed_pipe[i] = true; - dm_logger_write(dc->ctx->logger, LOG_DC, + DC_LOG_DC( "Reset mpcc for pipe %d\n", old_pipe_ctx->pipe_idx); } @@ -2017,7 +2039,7 @@ static void dcn10_apply_ctx_for_surface( if (num_planes > 0) program_all_pipe_in_tree(dc, top_pipe_to_program, context); - tg->funcs->unlock(tg); + dcn10_pipe_control_lock(dc, top_pipe_to_program, false); if (num_planes == 0) false_optc_underflow_wa(dc, stream, tg); @@ -2051,7 +2073,7 @@ static void dcn10_apply_ctx_for_surface( dcn10_verify_allow_pstate_change_high(dc); } } -/* dm_logger_write(dc->ctx->logger, LOG_BANDWIDTH_CALCS, +/* DC_LOG_BANDWIDTH_CALCS(dc->ctx->logger, "\n============== Watermark parameters ==============\n" "a.urgent_ns: %d \n" "a.cstate_enter_plus_exit: %d \n" @@ -2074,7 +2096,7 @@ static void dcn10_apply_ctx_for_surface( context->bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns, context->bw.dcn.watermarks.b.pte_meta_urgent_ns ); - dm_logger_write(dc->ctx->logger, LOG_BANDWIDTH_CALCS, + DC_LOG_BANDWIDTH_CALCS(dc->ctx->logger, "\nc.urgent_ns: %d \n" "c.cstate_enter_plus_exit: %d \n" "c.cstate_exit: %d \n" @@ -2100,6 +2122,101 @@ static void dcn10_apply_ctx_for_surface( */ } +static inline bool should_set_clock(bool decrease_allowed, int calc_clk, int cur_clk) +{ + return ((decrease_allowed && calc_clk < cur_clk) || calc_clk > cur_clk); +} + +static int determine_dppclk_threshold(struct dc *dc, struct dc_state *context) +{ + bool request_dpp_div = context->bw.dcn.calc_clk.dispclk_khz > + context->bw.dcn.calc_clk.dppclk_khz; + bool dispclk_increase = context->bw.dcn.calc_clk.dispclk_khz > + context->bw.dcn.cur_clk.dispclk_khz; + int disp_clk_threshold = context->bw.dcn.calc_clk.max_supported_dppclk_khz; + bool cur_dpp_div = context->bw.dcn.cur_clk.dispclk_khz > + context->bw.dcn.cur_clk.dppclk_khz; + + /* increase clock, looking for div is 0 for current, request div is 1*/ + if (dispclk_increase) { + /* already divided by 2, no need to reach target clk with 2 steps*/ + if (cur_dpp_div) + return context->bw.dcn.calc_clk.dispclk_khz; + + /* request disp clk is lower than maximum supported dpp clk, + * no need to reach target clk with two steps. + */ + if (context->bw.dcn.calc_clk.dispclk_khz <= disp_clk_threshold) + return context->bw.dcn.calc_clk.dispclk_khz; + + /* target dpp clk not request divided by 2, still within threshold */ + if (!request_dpp_div) + return context->bw.dcn.calc_clk.dispclk_khz; + + } else { + /* decrease clock, looking for current dppclk divided by 2, + * request dppclk not divided by 2. + */ + + /* current dpp clk not divided by 2, no need to ramp*/ + if (!cur_dpp_div) + return context->bw.dcn.calc_clk.dispclk_khz; + + /* current disp clk is lower than current maximum dpp clk, + * no need to ramp + */ + if (context->bw.dcn.cur_clk.dispclk_khz <= disp_clk_threshold) + return context->bw.dcn.calc_clk.dispclk_khz; + + /* request dpp clk need to be divided by 2 */ + if (request_dpp_div) + return context->bw.dcn.calc_clk.dispclk_khz; + } + + return disp_clk_threshold; +} + +static void ramp_up_dispclk_with_dpp(struct dc *dc, struct dc_state *context) +{ + int i; + bool request_dpp_div = context->bw.dcn.calc_clk.dispclk_khz > + context->bw.dcn.calc_clk.dppclk_khz; + + int dispclk_to_dpp_threshold = determine_dppclk_threshold(dc, context); + + /* set disp clk to dpp clk threshold */ + dc->res_pool->display_clock->funcs->set_clock( + dc->res_pool->display_clock, + dispclk_to_dpp_threshold); + + /* update request dpp clk division option */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i]; + + if (!pipe_ctx->plane_state) + continue; + + pipe_ctx->plane_res.dpp->funcs->dpp_dppclk_control( + pipe_ctx->plane_res.dpp, + request_dpp_div, + true); + } + + /* If target clk not same as dppclk threshold, set to target clock */ + if (dispclk_to_dpp_threshold != context->bw.dcn.calc_clk.dispclk_khz) { + dc->res_pool->display_clock->funcs->set_clock( + dc->res_pool->display_clock, + context->bw.dcn.calc_clk.dispclk_khz); + } + + context->bw.dcn.cur_clk.dispclk_khz = + context->bw.dcn.calc_clk.dispclk_khz; + context->bw.dcn.cur_clk.dppclk_khz = + context->bw.dcn.calc_clk.dppclk_khz; + context->bw.dcn.cur_clk.max_supported_dppclk_khz = + context->bw.dcn.calc_clk.max_supported_dppclk_khz; +} + static void dcn10_set_bandwidth( struct dc *dc, struct dc_state *context, @@ -2117,31 +2234,33 @@ static void dcn10_set_bandwidth( if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) return; - if (decrease_allowed || context->bw.dcn.calc_clk.dispclk_khz - > dc->current_state->bw.dcn.cur_clk.dispclk_khz) { - dc->res_pool->display_clock->funcs->set_clock( - dc->res_pool->display_clock, - context->bw.dcn.calc_clk.dispclk_khz); - dc->current_state->bw.dcn.cur_clk.dispclk_khz = - context->bw.dcn.calc_clk.dispclk_khz; - } - if (decrease_allowed || context->bw.dcn.calc_clk.dcfclk_khz - > dc->current_state->bw.dcn.cur_clk.dcfclk_khz) { + if (should_set_clock( + decrease_allowed, + context->bw.dcn.calc_clk.dcfclk_khz, + dc->current_state->bw.dcn.cur_clk.dcfclk_khz)) { + context->bw.dcn.cur_clk.dcfclk_khz = + context->bw.dcn.calc_clk.dcfclk_khz; smu_req.hard_min_dcefclk_khz = context->bw.dcn.calc_clk.dcfclk_khz; } - if (decrease_allowed || context->bw.dcn.calc_clk.fclk_khz - > dc->current_state->bw.dcn.cur_clk.fclk_khz) { - smu_req.hard_min_fclk_khz = context->bw.dcn.calc_clk.fclk_khz; - } - if (decrease_allowed || context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz - > dc->current_state->bw.dcn.cur_clk.dcfclk_deep_sleep_khz) { - dc->current_state->bw.dcn.calc_clk.dcfclk_deep_sleep_khz = - context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz; + + if (should_set_clock( + decrease_allowed, + context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz, + dc->current_state->bw.dcn.cur_clk.dcfclk_deep_sleep_khz)) { context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz = context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz; } + if (should_set_clock( + decrease_allowed, + context->bw.dcn.calc_clk.fclk_khz, + dc->current_state->bw.dcn.cur_clk.fclk_khz)) { + context->bw.dcn.cur_clk.fclk_khz = + context->bw.dcn.calc_clk.fclk_khz; + smu_req.hard_min_fclk_khz = context->bw.dcn.calc_clk.fclk_khz; + } + smu_req.display_count = context->stream_count; if (pp_smu->set_display_requirement) @@ -2149,21 +2268,17 @@ static void dcn10_set_bandwidth( *smu_req_cur = smu_req; - /* Decrease in freq is increase in period so opposite comparison for dram_ccm */ - if (decrease_allowed || context->bw.dcn.calc_clk.dram_ccm_us - < dc->current_state->bw.dcn.cur_clk.dram_ccm_us) { - dc->current_state->bw.dcn.calc_clk.dram_ccm_us = - context->bw.dcn.calc_clk.dram_ccm_us; - context->bw.dcn.cur_clk.dram_ccm_us = - context->bw.dcn.calc_clk.dram_ccm_us; + /* make sure dcf clk is before dpp clk to + * make sure we have enough voltage to run dpp clk + */ + if (should_set_clock( + decrease_allowed, + context->bw.dcn.calc_clk.dispclk_khz, + dc->current_state->bw.dcn.cur_clk.dispclk_khz)) { + + ramp_up_dispclk_with_dpp(dc, context); } - if (decrease_allowed || context->bw.dcn.calc_clk.min_active_dram_ccm_us - < dc->current_state->bw.dcn.cur_clk.min_active_dram_ccm_us) { - dc->current_state->bw.dcn.calc_clk.min_active_dram_ccm_us = - context->bw.dcn.calc_clk.min_active_dram_ccm_us; - context->bw.dcn.cur_clk.min_active_dram_ccm_us = - context->bw.dcn.calc_clk.min_active_dram_ccm_us; - } + dcn10_pplib_apply_display_requirements(dc, context); if (dc->debug.sanity_checks) { @@ -2212,6 +2327,8 @@ static void set_static_screen_control(struct pipe_ctx **pipe_ctx, value |= 0x80; if (events->cursor_update) value |= 0x2; + if (events->force_trigger) + value |= 0x1; for (i = 0; i < num_pipes; i++) pipe_ctx[i]->stream_res.tg->funcs-> @@ -2284,12 +2401,24 @@ static void dcn10_setup_stereo(struct pipe_ctx *pipe_ctx, struct dc *dc) return; } +static struct hubp *get_hubp_by_inst(struct resource_pool *res_pool, int mpcc_inst) +{ + int i; + + for (i = 0; i < res_pool->pipe_count; i++) { + if (res_pool->hubps[i]->inst == mpcc_inst) + return res_pool->hubps[i]; + } + ASSERT(false); + return NULL; +} + static void dcn10_wait_for_mpcc_disconnect( struct dc *dc, struct resource_pool *res_pool, struct pipe_ctx *pipe_ctx) { - int i; + int mpcc_inst; if (dc->debug.sanity_checks) { dcn10_verify_allow_pstate_change_high(dc); @@ -2298,12 +2427,14 @@ static void dcn10_wait_for_mpcc_disconnect( if (!pipe_ctx->stream_res.opp) return; - for (i = 0; i < MAX_PIPES; i++) { - if (pipe_ctx->stream_res.opp->mpcc_disconnect_pending[i]) { - res_pool->mpc->funcs->wait_for_idle(res_pool->mpc, i); - pipe_ctx->stream_res.opp->mpcc_disconnect_pending[i] = false; - res_pool->hubps[i]->funcs->set_blank(res_pool->hubps[i], true); - /*dm_logger_write(dc->ctx->logger, LOG_ERROR, + for (mpcc_inst = 0; mpcc_inst < MAX_PIPES; mpcc_inst++) { + if (pipe_ctx->stream_res.opp->mpcc_disconnect_pending[mpcc_inst]) { + struct hubp *hubp = get_hubp_by_inst(res_pool, mpcc_inst); + + res_pool->mpc->funcs->wait_for_idle(res_pool->mpc, mpcc_inst); + pipe_ctx->stream_res.opp->mpcc_disconnect_pending[mpcc_inst] = false; + hubp->funcs->set_blank(hubp, true); + /*DC_LOG_ERROR(dc->ctx->logger, "[debug_mpo: wait_for_mpcc finished waiting on mpcc %d]\n", i);*/ } @@ -2404,6 +2535,7 @@ static const struct hw_sequencer_funcs dcn10_funcs = { .enable_stream = dce110_enable_stream, .disable_stream = dce110_disable_stream, .unblank_stream = dce110_unblank_stream, + .blank_stream = dce110_blank_stream, .enable_display_power_gating = dcn10_dummy_display_power_gating, .disable_plane = dcn10_disable_plane, .pipe_control_lock = dcn10_pipe_control_lock,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h index b9d3260..6c526b5 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
@@ -35,5 +35,6 @@ extern void fill_display_configs( const struct dc_state *context, struct dm_pp_display_configuration *pp_display_cfg); +bool is_rgb_cspace(enum dc_color_space output_color_space); #endif /* __DC_HWSS_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_ipp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_ipp.h index d7b5bd2..819b749 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_ipp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_ipp.h
@@ -33,7 +33,6 @@ #define IPP_REG_LIST_DCN(id) \ SRI(FORMAT_CONTROL, CNVC_CFG, id), \ - SRI(DPP_CONTROL, DPP_TOP, id), \ SRI(CNVC_SURFACE_PIXEL_FORMAT, CNVC_CFG, id), \ SRI(CURSOR0_CONTROL, CNVC_CUR, id), \ SRI(CURSOR0_COLOR0, CNVC_CUR, id), \ @@ -130,7 +129,6 @@ struct dcn10_ipp_mask { }; struct dcn10_ipp_registers { - uint32_t DPP_CONTROL; uint32_t CURSOR_SETTINS; uint32_t CURSOR_SETTINGS; uint32_t CNVC_SURFACE_PIXEL_FORMAT;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c index f6ba0ee..77a1a9d 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c
@@ -367,6 +367,14 @@ void opp1_program_oppbuf( } +void opp1_pipe_clock_control(struct output_pixel_processor *opp, bool enable) +{ + struct dcn10_opp *oppn10 = TO_DCN10_OPP(opp); + uint32_t regval = enable ? 1 : 0; + + REG_UPDATE(OPP_PIPE_CONTROL, OPP_PIPE_CLOCK_EN, regval); +} + /*****************************************/ /* Constructor, Destructor */ /*****************************************/ @@ -382,6 +390,7 @@ static struct opp_funcs dcn10_opp_funcs = { .opp_program_fmt = opp1_program_fmt, .opp_program_bit_depth_reduction = opp1_program_bit_depth_reduction, .opp_program_stereo = opp1_program_stereo, + .opp_pipe_clock_control = opp1_pipe_clock_control, .opp_destroy = opp1_destroy };
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h index bc5058a..0f10ade 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h
@@ -44,7 +44,8 @@ SRI(FMT_MAP420_MEMORY_CONTROL, FMT, id), \ SRI(OPPBUF_CONTROL, OPPBUF, id),\ SRI(OPPBUF_3D_PARAMETERS_0, OPPBUF, id), \ - SRI(OPPBUF_3D_PARAMETERS_1, OPPBUF, id) + SRI(OPPBUF_3D_PARAMETERS_1, OPPBUF, id), \ + SRI(OPP_PIPE_CONTROL, OPP_PIPE, id) #define OPP_REG_LIST_DCN10(id) \ OPP_REG_LIST_DCN(id) @@ -61,7 +62,8 @@ uint32_t OPPBUF_CONTROL; \ uint32_t OPPBUF_CONTROL1; \ uint32_t OPPBUF_3D_PARAMETERS_0; \ - uint32_t OPPBUF_3D_PARAMETERS_1 + uint32_t OPPBUF_3D_PARAMETERS_1; \ + uint32_t OPP_PIPE_CONTROL #define OPP_MASK_SH_LIST_DCN(mask_sh) \ OPP_SF(FMT0_FMT_BIT_DEPTH_CONTROL, FMT_TRUNCATE_EN, mask_sh), \ @@ -89,7 +91,8 @@ OPP_SF(OPPBUF0_OPPBUF_CONTROL, OPPBUF_ACTIVE_WIDTH, mask_sh),\ OPP_SF(OPPBUF0_OPPBUF_CONTROL, OPPBUF_PIXEL_REPETITION, mask_sh),\ OPP_SF(OPPBUF0_OPPBUF_3D_PARAMETERS_0, OPPBUF_3D_VACT_SPACE1_SIZE, mask_sh), \ - OPP_SF(OPPBUF0_OPPBUF_3D_PARAMETERS_0, OPPBUF_3D_VACT_SPACE2_SIZE, mask_sh) + OPP_SF(OPPBUF0_OPPBUF_3D_PARAMETERS_0, OPPBUF_3D_VACT_SPACE2_SIZE, mask_sh), \ + OPP_SF(OPP_PIPE0_OPP_PIPE_CONTROL, OPP_PIPE_CLOCK_EN, mask_sh) #define OPP_MASK_SH_LIST_DCN10(mask_sh) \ OPP_MASK_SH_LIST_DCN(mask_sh), \ @@ -125,7 +128,8 @@ type OPPBUF_OVERLAP_PIXEL_NUM;\ type OPPBUF_NUM_SEGMENT_PADDED_PIXELS; \ type OPPBUF_3D_VACT_SPACE1_SIZE; \ - type OPPBUF_3D_VACT_SPACE2_SIZE + type OPPBUF_3D_VACT_SPACE2_SIZE; \ + type OPP_PIPE_CLOCK_EN struct dcn10_opp_registers { OPP_COMMON_REG_VARIABLE_LIST; @@ -176,6 +180,8 @@ void opp1_program_stereo( bool enable, const struct dc_crtc_timing *timing); +void opp1_pipe_clock_control(struct output_pixel_processor *opp, bool enable); + void opp1_destroy(struct output_pixel_processor **opp); #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h index a3c7c20..d25e7bf 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
@@ -83,6 +83,8 @@ struct dcn_optc_registers { + uint32_t OTG_GLOBAL_CONTROL1; + uint32_t OTG_GLOBAL_CONTROL2; uint32_t OTG_VERT_SYNC_CONTROL; uint32_t OTG_MASTER_UPDATE_MODE; uint32_t OTG_GSL_CONTROL; @@ -126,12 +128,12 @@ struct dcn_optc_registers { uint32_t OTG_VERTICAL_INTERRUPT2_POSITION; uint32_t OPTC_INPUT_CLOCK_CONTROL; uint32_t OPTC_DATA_SOURCE_SELECT; + uint32_t OPTC_MEMORY_CONFIG; uint32_t OPTC_INPUT_GLOBAL_CONTROL; uint32_t CONTROL; uint32_t OTG_GSL_WINDOW_X; uint32_t OTG_GSL_WINDOW_Y; uint32_t OTG_VUPDATE_KEEPOUT; - uint32_t OTG_DSC_START_POSITION; }; #define TG_COMMON_MASK_SH_LIST_DCN(mask_sh)\ @@ -241,7 +243,7 @@ struct dcn_optc_registers { SF(OTG0_OTG_TEST_PATTERN_COLOR, OTG_TEST_PATTERN_DATA, mask_sh),\ SF(ODM0_OPTC_DATA_SOURCE_SELECT, OPTC_SRC_SEL, mask_sh) -#define TG_REG_FIELD_LIST(type) \ +#define TG_REG_FIELD_LIST_DCN1_0(type) \ type VSTARTUP_START;\ type VUPDATE_OFFSET;\ type VUPDATE_WIDTH;\ @@ -326,10 +328,9 @@ struct dcn_optc_registers { type OPTC_INPUT_CLK_EN;\ type OPTC_INPUT_CLK_ON;\ type OPTC_INPUT_CLK_GATE_DIS;\ - type OPTC_SRC_SEL;\ - type OPTC_SEG0_SRC_SEL;\ type OPTC_UNDERFLOW_OCCURRED_STATUS;\ type OPTC_UNDERFLOW_CLEAR;\ + type OPTC_SRC_SEL;\ type VTG0_ENABLE;\ type VTG0_FP2;\ type VTG0_VCOUNT_INIT;\ @@ -352,10 +353,11 @@ struct dcn_optc_registers { type OTG_MASTER_UPDATE_LOCK_GSL_EN;\ type MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_START_OFFSET;\ type MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_END_OFFSET;\ - type OTG_DSC_START_POSITION_X;\ - type OTG_DSC_START_POSITION_LINE_NUM;\ type OTG_MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_EN; +#define TG_REG_FIELD_LIST(type) \ + TG_REG_FIELD_LIST_DCN1_0(type) + struct dcn_optc_shift { TG_REG_FIELD_LIST(uint8_t)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c index 9b81b2e..b0ef20a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -50,7 +50,8 @@ #include "dcn10_hubp.h" #include "dcn10_hubbub.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #include "dcn/dcn_1_0_offset.h" #include "dcn/dcn_1_0_sh_mask.h" @@ -365,6 +366,7 @@ static const struct dcn_optc_mask tg_mask = { static const struct bios_registers bios_regs = { + NBIO_SR(BIOS_SCRATCH_3), NBIO_SR(BIOS_SCRATCH_6) }; @@ -438,7 +440,11 @@ static const struct dc_debug debug_defaults_drv = { .timing_trace = false, .clock_trace = true, - .min_disp_clk_khz = 300000, + /* raven smu dones't allow 0 disp clk, + * smu min disp clk limit is 50Mhz + * keep min disp clk 100Mhz avoid smu hang + */ + .min_disp_clk_khz = 100000, .disable_pplib_clock_request = true, .disable_pplib_wm_range = false, @@ -450,6 +456,7 @@ static const struct dc_debug debug_defaults_drv = { .disable_stereo_support = true, .vsr_support = true, .performance_trace = false, + .az_endpoint_mute_only = true, }; static const struct dc_debug debug_defaults_diags = { @@ -818,7 +825,7 @@ static void get_pixel_clock_parameters( pixel_clk_params->requested_pix_clk = stream->timing.pix_clk_khz; pixel_clk_params->encoder_object_id = stream->sink->link->link_enc->id; pixel_clk_params->signal_type = pipe_ctx->stream->signal; - pixel_clk_params->controller_id = pipe_ctx->pipe_idx + 1; + pixel_clk_params->controller_id = pipe_ctx->stream_res.tg->inst + 1; /* TODO: un-hardcode*/ pixel_clk_params->requested_sym_clk = LINK_RATE_LOW * LINK_RATE_REF_FREQ_IN_KHZ; @@ -960,11 +967,13 @@ static struct pipe_ctx *dcn10_acquire_idle_pipe_for_layer( idle_pipe->stream = head_pipe->stream; idle_pipe->stream_res.tg = head_pipe->stream_res.tg; + idle_pipe->stream_res.abm = head_pipe->stream_res.abm; idle_pipe->stream_res.opp = head_pipe->stream_res.opp; idle_pipe->plane_res.hubp = pool->hubps[idle_pipe->pipe_idx]; idle_pipe->plane_res.ipp = pool->ipps[idle_pipe->pipe_idx]; idle_pipe->plane_res.dpp = pool->dpps[idle_pipe->pipe_idx]; + idle_pipe->plane_res.mpcc_inst = pool->dpps[idle_pipe->pipe_idx]->inst; return idle_pipe; } @@ -1319,13 +1328,11 @@ static bool construct( } } - if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { - pool->base.display_clock = dce120_disp_clk_create(ctx); - if (pool->base.display_clock == NULL) { - dm_error("DC: failed to create display clock!\n"); - BREAK_TO_DEBUGGER(); - goto fail; - } + pool->base.display_clock = dce120_disp_clk_create(ctx); + if (pool->base.display_clock == NULL) { + dm_error("DC: failed to create display clock!\n"); + BREAK_TO_DEBUGGER(); + goto fail; } pool->base.dmcu = dcn10_dmcu_create(ctx, @@ -1448,6 +1455,7 @@ static bool construct( /* valid pipe num */ pool->base.pipe_count = j; + pool->base.timing_generator_count = j; /* within dml lib, it is hard code to 4. If ASIC pipe is fused, * the value may be changed
diff --git a/drivers/gpu/drm/amd/display/dc/dm_helpers.h b/drivers/gpu/drm/amd/display/dc/dm_helpers.h index ab88f07..034369f 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_helpers.h +++ b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
@@ -50,6 +50,13 @@ bool dm_helpers_dp_mst_write_payload_allocation_table( bool enable); /* + * Clear payload allocation table before enable MST DP link. + */ +void dm_helpers_dp_mst_clear_payload_allocation_table( + struct dc_context *ctx, + const struct dc_link *link); + +/* * Polls for ACT (allocation change trigger) handled and */ bool dm_helpers_dp_mst_poll_for_allocation_change_trigger( @@ -101,5 +108,8 @@ enum dc_edid_status dm_helpers_read_local_edid( struct dc_link *link, struct dc_sink *sink); +void dm_set_dcn_clocks( + struct dc_context *ctx, + struct dc_clocks *clks); #endif /* __DM_HELPERS__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h index bbfa832..eac4bfe 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h +++ b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h
@@ -91,7 +91,8 @@ struct pp_smu_funcs_rv { /* which SMU message? are reader and writer WM separate SMU msg? */ void (*set_wm_ranges)(struct pp_smu *pp, struct pp_smu_wm_range_sets *ranges); - + /* PME w/a */ + void (*set_pme_wa_enable)(struct pp_smu *pp); }; #if 0
diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h index 225b7bf..22e7ee7 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_services.h +++ b/drivers/gpu/drm/amd/display/dc/dm_services.h
@@ -192,37 +192,6 @@ unsigned int generic_reg_wait(const struct dc_context *ctx, * Power Play (PP) interfaces **************************************/ -/* DAL calls this function to notify PP about clocks it needs for the Mode Set. - * This is done *before* it changes DCE clock. - * - * If required clock is higher than current, then PP will increase the voltage. - * - * If required clock is lower than current, then PP will defer reduction of - * voltage until the call to dc_service_pp_post_dce_clock_change(). - * - * \input - Contains clocks needed for Mode Set. - * - * \output - Contains clocks adjusted by PP which DAL should use for Mode Set. - * Valid only if function returns zero. - * - * \returns true - call is successful - * false - call failed - */ -bool dm_pp_pre_dce_clock_change( - struct dc_context *ctx, - struct dm_pp_gpu_clock_range *requested_state, - struct dm_pp_gpu_clock_range *actual_state); - -/* The returned clocks range are 'static' system clocks which will be used for - * mode validation purposes. - * - * \returns true - call is successful - * false - call failed - */ -bool dc_service_get_system_clocks_range( - const struct dc_context *ctx, - struct dm_pp_gpu_clock_range *sys_clks); - /* Gets valid clocks levels from pplib * * input: clk_type - display clk / sclk / mem clk
diff --git a/drivers/gpu/drm/amd/display/dc/dm_services_types.h b/drivers/gpu/drm/amd/display/dc/dm_services_types.h index 25ece53..ab8c77d 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_services_types.h +++ b/drivers/gpu/drm/amd/display/dc/dm_services_types.h
@@ -29,7 +29,7 @@ #include "os_types.h" #include "dc_types.h" -#include "dm_pp_smu.h" +struct pp_smu_funcs_rv; struct dm_pp_clock_range { int min_khz;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index 782886c..f83a608 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile
@@ -32,18 +32,15 @@ dml_ccflags := -mhard-float -msse $(cc_stack_align) -CFLAGS_display_mode_vba.o := $(dml_ccflags) CFLAGS_display_mode_lib.o := $(dml_ccflags) CFLAGS_display_pipe_clocks.o := $(dml_ccflags) -CFLAGS_display_rq_dlg_calc.o := $(dml_ccflags) CFLAGS_dml1_display_rq_dlg_calc.o := $(dml_ccflags) CFLAGS_display_rq_dlg_helpers.o := $(dml_ccflags) CFLAGS_soc_bounding_box.o := $(dml_ccflags) CFLAGS_dml_common_defs.o := $(dml_ccflags) -DML = display_mode_lib.o display_rq_dlg_calc.o \ - display_rq_dlg_helpers.o dml1_display_rq_dlg_calc.o \ - soc_bounding_box.o dml_common_defs.o display_mode_vba.o +DML = display_mode_lib.o display_rq_dlg_helpers.o dml1_display_rq_dlg_calc.o \ + soc_bounding_box.o dml_common_defs.o AMD_DAL_DML = $(addprefix $(AMDDALPATH)/dc/dml/,$(DML))
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h index 26f4f2a..3c2abcb 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h
@@ -28,8 +28,6 @@ #include "dml_common_defs.h" #include "soc_bounding_box.h" -#include "display_mode_vba.h" -#include "display_rq_dlg_calc.h" #include "dml1_display_rq_dlg_calc.h" enum dml_project { @@ -41,7 +39,6 @@ struct display_mode_lib { struct _vcs_dpi_ip_params_st ip; struct _vcs_dpi_soc_bounding_box_st soc; enum dml_project project; - struct vba_vars_st vba; struct dal_logger *logger; };
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h index aeebd8b..09affa1 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
@@ -140,7 +140,6 @@ struct _vcs_dpi_ip_params_st { unsigned int max_hscl_taps; unsigned int max_vscl_taps; unsigned int xfc_supported; - unsigned int ptoi_supported; unsigned int xfc_fill_constant_bytes; double dispclk_ramp_margin_percent; double xfc_fill_bw_overhead_percent;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c deleted file mode 100644 index 260e113f..0000000 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c +++ /dev/null
@@ -1,6085 +0,0 @@ -/* - * Copyright 2017 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ - -#include "display_mode_lib.h" -#include "display_mode_vba.h" - -#include "dml_inline_defs.h" - -/* - * NOTE: - * This file is gcc-parseable HW gospel, coming straight from HW engineers. - * - * It doesn't adhere to Linux kernel style and sometimes will do things in odd - * ways. Unless there is something clearly wrong with it the code should - * remain as-is as it provides us with a guarantee from HW that it is correct. - */ - -#define BPP_INVALID 0 -#define BPP_BLENDED_PIPE 0xffffffff -static const unsigned int NumberOfStates = DC__VOLTAGE_STATES; - -static void fetch_socbb_params(struct display_mode_lib *mode_lib); -static void fetch_ip_params(struct display_mode_lib *mode_lib); -static void fetch_pipe_params(struct display_mode_lib *mode_lib); -static void recalculate_params( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes); -static void recalculate(struct display_mode_lib *mode_lib); -static double adjust_ReturnBW( - struct display_mode_lib *mode_lib, - double ReturnBW, - bool DCCEnabledAnyPlane, - double ReturnBandwidthToDCN); -static void ModeSupportAndSystemConfiguration(struct display_mode_lib *mode_lib); -static void DisplayPipeConfiguration(struct display_mode_lib *mode_lib); -static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( - struct display_mode_lib *mode_lib); -static unsigned int dscceComputeDelay( - unsigned int bpc, - double bpp, - unsigned int sliceWidth, - unsigned int numSlices, - enum output_format_class pixelFormat); -static unsigned int dscComputeDelay(enum output_format_class pixelFormat); -// Super monster function with some 45 argument -static bool CalculatePrefetchSchedule( - struct display_mode_lib *mode_lib, - double DPPCLK, - double DISPCLK, - double PixelClock, - double DCFClkDeepSleep, - unsigned int DSCDelay, - unsigned int DPPPerPlane, - bool ScalerEnabled, - unsigned int NumberOfCursors, - double DPPCLKDelaySubtotal, - double DPPCLKDelaySCL, - double DPPCLKDelaySCLLBOnly, - double DPPCLKDelayCNVCFormater, - double DPPCLKDelayCNVCCursor, - double DISPCLKDelaySubtotal, - unsigned int ScalerRecoutWidth, - enum output_format_class OutputFormat, - unsigned int VBlank, - unsigned int HTotal, - unsigned int MaxInterDCNTileRepeaters, - unsigned int VStartup, - unsigned int PageTableLevels, - bool VirtualMemoryEnable, - bool DynamicMetadataEnable, - unsigned int DynamicMetadataLinesBeforeActiveRequired, - unsigned int DynamicMetadataTransmittedBytes, - bool DCCEnable, - double UrgentLatency, - double UrgentExtraLatency, - double TCalc, - unsigned int PDEAndMetaPTEBytesFrame, - unsigned int MetaRowByte, - unsigned int PixelPTEBytesPerRow, - double PrefetchSourceLinesY, - unsigned int SwathWidthY, - double BytePerPixelDETY, - double VInitPreFillY, - unsigned int MaxNumSwathY, - double PrefetchSourceLinesC, - double BytePerPixelDETC, - double VInitPreFillC, - unsigned int MaxNumSwathC, - unsigned int SwathHeightY, - unsigned int SwathHeightC, - double TWait, - bool XFCEnabled, - double XFCRemoteSurfaceFlipDelay, - bool InterlaceEnable, - bool ProgressiveToInterlaceUnitInOPP, - double *DSTXAfterScaler, - double *DSTYAfterScaler, - double *DestinationLinesForPrefetch, - double *PrefetchBandwidth, - double *DestinationLinesToRequestVMInVBlank, - double *DestinationLinesToRequestRowInVBlank, - double *VRatioPrefetchY, - double *VRatioPrefetchC, - double *RequiredPrefetchPixDataBW, - unsigned int *VStartupRequiredWhenNotEnoughTimeForDynamicMetadata, - double *Tno_bw, - unsigned int *VUpdateOffsetPix, - unsigned int *VUpdateWidthPix, - unsigned int *VReadyOffsetPix); -static double RoundToDFSGranularityUp(double Clock, double VCOSpeed); -static double RoundToDFSGranularityDown(double Clock, double VCOSpeed); -static double CalculatePrefetchSourceLines( - struct display_mode_lib *mode_lib, - double VRatio, - double vtaps, - bool Interlace, - bool ProgressiveToInterlaceUnitInOPP, - unsigned int SwathHeight, - unsigned int ViewportYStart, - double *VInitPreFill, - unsigned int *MaxNumSwath); -static unsigned int CalculateVMAndRowBytes( - struct display_mode_lib *mode_lib, - bool DCCEnable, - unsigned int BlockHeight256Bytes, - unsigned int BlockWidth256Bytes, - enum source_format_class SourcePixelFormat, - unsigned int SurfaceTiling, - unsigned int BytePerPixel, - enum scan_direction_class ScanDirection, - unsigned int ViewportWidth, - unsigned int ViewportHeight, - unsigned int SwathWidthY, - bool VirtualMemoryEnable, - unsigned int VMMPageSize, - unsigned int PTEBufferSizeInRequests, - unsigned int PDEProcessingBufIn64KBReqs, - unsigned int Pitch, - unsigned int DCCMetaPitch, - unsigned int *MacroTileWidth, - unsigned int *MetaRowByte, - unsigned int *PixelPTEBytesPerRow, - bool *PTEBufferSizeNotExceeded, - unsigned int *dpte_row_height, - unsigned int *meta_row_height); -static double CalculateTWait( - unsigned int PrefetchMode, - double DRAMClockChangeLatency, - double UrgentLatency, - double SREnterPlusExitTime); -static double CalculateRemoteSurfaceFlipDelay( - struct display_mode_lib *mode_lib, - double VRatio, - double SwathWidth, - double Bpp, - double LineTime, - double XFCTSlvVupdateOffset, - double XFCTSlvVupdateWidth, - double XFCTSlvVreadyOffset, - double XFCXBUFLatencyTolerance, - double XFCFillBWOverhead, - double XFCSlvChunkSize, - double XFCBusTransportTime, - double TCalc, - double TWait, - double *SrcActiveDrainRate, - double *TInitXFill, - double *TslvChk); -static double CalculateWriteBackDISPCLK( - enum source_format_class WritebackPixelFormat, - double PixelClock, - double WritebackHRatio, - double WritebackVRatio, - unsigned int WritebackLumaHTaps, - unsigned int WritebackLumaVTaps, - unsigned int WritebackChromaHTaps, - unsigned int WritebackChromaVTaps, - double WritebackDestinationWidth, - unsigned int HTotal, - unsigned int WritebackChromaLineBufferWidth); -static void CalculateActiveRowBandwidth( - bool VirtualMemoryEnable, - enum source_format_class SourcePixelFormat, - double VRatio, - bool DCCEnable, - double LineTime, - unsigned int MetaRowByteLuma, - unsigned int MetaRowByteChroma, - unsigned int meta_row_height_luma, - unsigned int meta_row_height_chroma, - unsigned int PixelPTEBytesPerRowLuma, - unsigned int PixelPTEBytesPerRowChroma, - unsigned int dpte_row_height_luma, - unsigned int dpte_row_height_chroma, - double *meta_row_bw, - double *dpte_row_bw, - double *qual_row_bw); -static void CalculateFlipSchedule( - struct display_mode_lib *mode_lib, - double UrgentExtraLatency, - double UrgentLatency, - unsigned int MaxPageTableLevels, - bool VirtualMemoryEnable, - double BandwidthAvailableForImmediateFlip, - unsigned int TotImmediateFlipBytes, - enum source_format_class SourcePixelFormat, - unsigned int ImmediateFlipBytes, - double LineTime, - double Tno_bw, - double VRatio, - double PDEAndMetaPTEBytesFrame, - unsigned int MetaRowByte, - unsigned int PixelPTEBytesPerRow, - bool DCCEnable, - unsigned int dpte_row_height, - unsigned int meta_row_height, - double qual_row_bw, - double *DestinationLinesToRequestVMInImmediateFlip, - double *DestinationLinesToRequestRowInImmediateFlip, - double *final_flip_bw, - bool *ImmediateFlipSupportedForPipe); -static double CalculateWriteBackDelay( - enum source_format_class WritebackPixelFormat, - double WritebackHRatio, - double WritebackVRatio, - unsigned int WritebackLumaHTaps, - unsigned int WritebackLumaVTaps, - unsigned int WritebackChromaHTaps, - unsigned int WritebackChromaVTaps, - unsigned int WritebackDestinationWidth); -static void PixelClockAdjustmentForProgressiveToInterlaceUnit(struct display_mode_lib *mode_lib); -static unsigned int CursorBppEnumToBits(enum cursor_bpp ebpp); -static void ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib); - -void set_prefetch_mode( - struct display_mode_lib *mode_lib, - bool cstate_en, - bool pstate_en, - bool ignore_viewport_pos, - bool immediate_flip_support) -{ - unsigned int prefetch_mode; - - if (cstate_en && pstate_en) - prefetch_mode = 0; - else if (cstate_en) - prefetch_mode = 1; - else - prefetch_mode = 2; - if (prefetch_mode != mode_lib->vba.PrefetchMode - || ignore_viewport_pos != mode_lib->vba.IgnoreViewportPositioning - || immediate_flip_support != mode_lib->vba.ImmediateFlipSupport) { - DTRACE( - " Prefetch mode has changed from %i to %i. Recalculating.", - prefetch_mode, - mode_lib->vba.PrefetchMode); - mode_lib->vba.PrefetchMode = prefetch_mode; - mode_lib->vba.IgnoreViewportPositioning = ignore_viewport_pos; - mode_lib->vba.ImmediateFlipSupport = immediate_flip_support; - recalculate(mode_lib); - } -} - -unsigned int dml_get_voltage_level( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes) -{ - bool need_recalculate = memcmp(&mode_lib->soc, &mode_lib->vba.soc, sizeof(mode_lib->vba.soc)) != 0 - || memcmp(&mode_lib->ip, &mode_lib->vba.ip, sizeof(mode_lib->vba.ip)) != 0 - || num_pipes != mode_lib->vba.cache_num_pipes - || memcmp(pipes, mode_lib->vba.cache_pipes, - sizeof(display_e2e_pipe_params_st) * num_pipes) != 0; - - mode_lib->vba.soc = mode_lib->soc; - mode_lib->vba.ip = mode_lib->ip; - memcpy(mode_lib->vba.cache_pipes, pipes, sizeof(*pipes) * num_pipes); - mode_lib->vba.cache_num_pipes = num_pipes; - - if (need_recalculate && pipes[0].clks_cfg.dppclk_mhz != 0) - recalculate(mode_lib); - else { - fetch_socbb_params(mode_lib); - fetch_ip_params(mode_lib); - fetch_pipe_params(mode_lib); - } - ModeSupportAndSystemConfigurationFull(mode_lib); - - return mode_lib->vba.VoltageLevel; -} - -#define dml_get_attr_func(attr, var) double get_##attr(struct display_mode_lib *mode_lib, const display_e2e_pipe_params_st *pipes, unsigned int num_pipes) \ -{ \ - recalculate_params(mode_lib, pipes, num_pipes); \ - return var; \ -} - -dml_get_attr_func(clk_dcf_deepsleep, mode_lib->vba.DCFClkDeepSleep); -dml_get_attr_func(wm_urgent, mode_lib->vba.UrgentWatermark); -dml_get_attr_func(wm_memory_trip, mode_lib->vba.MemoryTripWatermark); -dml_get_attr_func(wm_writeback_urgent, mode_lib->vba.WritebackUrgentWatermark); -dml_get_attr_func(wm_stutter_exit, mode_lib->vba.StutterExitWatermark); -dml_get_attr_func(wm_stutter_enter_exit, mode_lib->vba.StutterEnterPlusExitWatermark); -dml_get_attr_func(wm_dram_clock_change, mode_lib->vba.DRAMClockChangeWatermark); -dml_get_attr_func(wm_writeback_dram_clock_change, mode_lib->vba.WritebackDRAMClockChangeWatermark); -dml_get_attr_func(wm_xfc_underflow, mode_lib->vba.UrgentWatermark); // xfc_underflow maps to urgent -dml_get_attr_func(stutter_efficiency, mode_lib->vba.StutterEfficiency); -dml_get_attr_func(stutter_efficiency_no_vblank, mode_lib->vba.StutterEfficiencyNotIncludingVBlank); -dml_get_attr_func(urgent_latency, mode_lib->vba.MinUrgentLatencySupportUs); -dml_get_attr_func(urgent_extra_latency, mode_lib->vba.UrgentExtraLatency); -dml_get_attr_func(nonurgent_latency, mode_lib->vba.NonUrgentLatencyTolerance); -dml_get_attr_func( - dram_clock_change_latency, - mode_lib->vba.MinActiveDRAMClockChangeLatencySupported); -dml_get_attr_func(dispclk_calculated, mode_lib->vba.DISPCLK_calculated); -dml_get_attr_func(total_data_read_bw, mode_lib->vba.TotalDataReadBandwidth); -dml_get_attr_func(return_bw, mode_lib->vba.ReturnBW); -dml_get_attr_func(tcalc, mode_lib->vba.TCalc); - -#define dml_get_pipe_attr_func(attr, var) double get_##attr(struct display_mode_lib *mode_lib, const display_e2e_pipe_params_st *pipes, unsigned int num_pipes, unsigned int which_pipe) \ -{\ - unsigned int which_plane; \ - recalculate_params(mode_lib, pipes, num_pipes); \ - which_plane = mode_lib->vba.pipe_plane[which_pipe]; \ - return var[which_plane]; \ -} - -dml_get_pipe_attr_func(dsc_delay, mode_lib->vba.DSCDelay); -dml_get_pipe_attr_func(dppclk_calculated, mode_lib->vba.DPPCLK_calculated); -dml_get_pipe_attr_func(dscclk_calculated, mode_lib->vba.DSCCLK_calculated); -dml_get_pipe_attr_func(min_ttu_vblank, mode_lib->vba.MinTTUVBlank); -dml_get_pipe_attr_func(vratio_prefetch_l, mode_lib->vba.VRatioPrefetchY); -dml_get_pipe_attr_func(vratio_prefetch_c, mode_lib->vba.VRatioPrefetchC); -dml_get_pipe_attr_func(dst_x_after_scaler, mode_lib->vba.DSTXAfterScaler); -dml_get_pipe_attr_func(dst_y_after_scaler, mode_lib->vba.DSTYAfterScaler); -dml_get_pipe_attr_func(dst_y_per_vm_vblank, mode_lib->vba.DestinationLinesToRequestVMInVBlank); -dml_get_pipe_attr_func(dst_y_per_row_vblank, mode_lib->vba.DestinationLinesToRequestRowInVBlank); -dml_get_pipe_attr_func(dst_y_prefetch, mode_lib->vba.DestinationLinesForPrefetch); -dml_get_pipe_attr_func(dst_y_per_vm_flip, mode_lib->vba.DestinationLinesToRequestVMInImmediateFlip); -dml_get_pipe_attr_func( - dst_y_per_row_flip, - mode_lib->vba.DestinationLinesToRequestRowInImmediateFlip); - -dml_get_pipe_attr_func(xfc_transfer_delay, mode_lib->vba.XFCTransferDelay); -dml_get_pipe_attr_func(xfc_precharge_delay, mode_lib->vba.XFCPrechargeDelay); -dml_get_pipe_attr_func(xfc_remote_surface_flip_latency, mode_lib->vba.XFCRemoteSurfaceFlipLatency); -dml_get_pipe_attr_func(xfc_prefetch_margin, mode_lib->vba.XFCPrefetchMargin); - -unsigned int get_vstartup_calculated( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes, - unsigned int which_pipe) -{ - unsigned int which_plane; - - recalculate_params(mode_lib, pipes, num_pipes); - which_plane = mode_lib->vba.pipe_plane[which_pipe]; - return mode_lib->vba.VStartup[which_plane]; -} - -double get_total_immediate_flip_bytes( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes) -{ - recalculate_params(mode_lib, pipes, num_pipes); - return mode_lib->vba.TotImmediateFlipBytes; -} - -double get_total_immediate_flip_bw( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes) -{ - recalculate_params(mode_lib, pipes, num_pipes); - return mode_lib->vba.ImmediateFlipBW; -} - -double get_total_prefetch_bw( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes) -{ - unsigned int k; - double total_prefetch_bw = 0.0; - - recalculate_params(mode_lib, pipes, num_pipes); - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) - total_prefetch_bw += mode_lib->vba.PrefetchBandwidth[k]; - return total_prefetch_bw; -} - -static void fetch_socbb_params(struct display_mode_lib *mode_lib) -{ - soc_bounding_box_st *soc = &mode_lib->vba.soc; - unsigned int i; - - // SOC Bounding Box Parameters - mode_lib->vba.ReturnBusWidth = soc->return_bus_width_bytes; - mode_lib->vba.NumberOfChannels = soc->num_chans; - mode_lib->vba.PercentOfIdealDRAMAndFabricBWReceivedAfterUrgLatency = - soc->ideal_dram_bw_after_urgent_percent; // there's always that one bastard variable that's so long it throws everything out of alignment! - mode_lib->vba.UrgentLatency = soc->urgent_latency_us; - mode_lib->vba.RoundTripPingLatencyCycles = soc->round_trip_ping_latency_dcfclk_cycles; - mode_lib->vba.UrgentOutOfOrderReturnPerChannel = - soc->urgent_out_of_order_return_per_channel_bytes; - mode_lib->vba.WritebackLatency = soc->writeback_latency_us; - mode_lib->vba.SRExitTime = soc->sr_exit_time_us; - mode_lib->vba.SREnterPlusExitTime = soc->sr_enter_plus_exit_time_us; - mode_lib->vba.DRAMClockChangeLatency = soc->dram_clock_change_latency_us; - mode_lib->vba.Downspreading = soc->downspread_percent; - mode_lib->vba.DRAMChannelWidth = soc->dram_channel_width_bytes; // new! - mode_lib->vba.FabricDatapathToDCNDataReturn = soc->fabric_datapath_to_dcn_data_return_bytes; // new! - mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading = soc->dcn_downspread_percent; // new - mode_lib->vba.DISPCLKDPPCLKVCOSpeed = soc->dispclk_dppclk_vco_speed_mhz; // new - mode_lib->vba.VMMPageSize = soc->vmm_page_size_bytes; - // Set the voltage scaling clocks as the defaults. Most of these will - // be set to different values by the test - for (i = 0; i < DC__VOLTAGE_STATES; i++) - if (soc->clock_limits[i].state == mode_lib->vba.VoltageLevel) - break; - - mode_lib->vba.DCFCLK = soc->clock_limits[i].dcfclk_mhz; - mode_lib->vba.SOCCLK = soc->clock_limits[i].socclk_mhz; - mode_lib->vba.DRAMSpeed = soc->clock_limits[i].dram_speed_mhz; - mode_lib->vba.FabricClock = soc->clock_limits[i].fabricclk_mhz; - - mode_lib->vba.XFCBusTransportTime = soc->xfc_bus_transport_time_us; - mode_lib->vba.XFCXBUFLatencyTolerance = soc->xfc_xbuf_latency_tolerance_us; - - mode_lib->vba.SupportGFX7CompatibleTilingIn32bppAnd64bpp = false; - mode_lib->vba.MaxHSCLRatio = 4; - mode_lib->vba.MaxVSCLRatio = 4; - mode_lib->vba.MaxNumWriteback = 0; /*TODO*/ - mode_lib->vba.WritebackLumaAndChromaScalingSupported = true; - mode_lib->vba.Cursor64BppSupport = true; - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.DCFCLKPerState[i] = soc->clock_limits[i].dcfclk_mhz; - mode_lib->vba.FabricClockPerState[i] = soc->clock_limits[i].fabricclk_mhz; - mode_lib->vba.SOCCLKPerState[i] = soc->clock_limits[i].socclk_mhz; - mode_lib->vba.PHYCLKPerState[i] = soc->clock_limits[i].phyclk_mhz; - mode_lib->vba.MaxDppclk[i] = soc->clock_limits[i].dppclk_mhz; - mode_lib->vba.MaxDSCCLK[i] = soc->clock_limits[i].dscclk_mhz; - mode_lib->vba.DRAMSpeedPerState[i] = soc->clock_limits[i].dram_speed_mhz; - mode_lib->vba.MaxDispclk[i] = soc->clock_limits[i].dispclk_mhz; - } -} - -static void fetch_ip_params(struct display_mode_lib *mode_lib) -{ - ip_params_st *ip = &mode_lib->vba.ip; - - // IP Parameters - mode_lib->vba.MaxNumDPP = ip->max_num_dpp; - mode_lib->vba.MaxNumOTG = ip->max_num_otg; - mode_lib->vba.CursorChunkSize = ip->cursor_chunk_size; - mode_lib->vba.CursorBufferSize = ip->cursor_buffer_size; - - mode_lib->vba.MaxDCHUBToPSCLThroughput = ip->max_dchub_pscl_bw_pix_per_clk; - mode_lib->vba.MaxPSCLToLBThroughput = ip->max_pscl_lb_bw_pix_per_clk; - mode_lib->vba.ROBBufferSizeInKByte = ip->rob_buffer_size_kbytes; - mode_lib->vba.DETBufferSizeInKByte = ip->det_buffer_size_kbytes; - mode_lib->vba.PixelChunkSizeInKByte = ip->pixel_chunk_size_kbytes; - mode_lib->vba.MetaChunkSize = ip->meta_chunk_size_kbytes; - mode_lib->vba.PTEChunkSize = ip->pte_chunk_size_kbytes; - mode_lib->vba.WritebackChunkSize = ip->writeback_chunk_size_kbytes; - mode_lib->vba.LineBufferSize = ip->line_buffer_size_bits; - mode_lib->vba.MaxLineBufferLines = ip->max_line_buffer_lines; - mode_lib->vba.PTEBufferSizeInRequests = ip->dpte_buffer_size_in_pte_reqs; - mode_lib->vba.DPPOutputBufferPixels = ip->dpp_output_buffer_pixels; - mode_lib->vba.OPPOutputBufferLines = ip->opp_output_buffer_lines; - mode_lib->vba.WritebackInterfaceLumaBufferSize = ip->writeback_luma_buffer_size_kbytes; - mode_lib->vba.WritebackInterfaceChromaBufferSize = ip->writeback_chroma_buffer_size_kbytes; - mode_lib->vba.WritebackChromaLineBufferWidth = - ip->writeback_chroma_line_buffer_width_pixels; - mode_lib->vba.MaxPageTableLevels = ip->max_page_table_levels; - mode_lib->vba.MaxInterDCNTileRepeaters = ip->max_inter_dcn_tile_repeaters; - mode_lib->vba.NumberOfDSC = ip->num_dsc; - mode_lib->vba.ODMCapability = ip->odm_capable; - mode_lib->vba.DISPCLKRampingMargin = ip->dispclk_ramp_margin_percent; - - mode_lib->vba.XFCSupported = ip->xfc_supported; - mode_lib->vba.XFCFillBWOverhead = ip->xfc_fill_bw_overhead_percent; - mode_lib->vba.XFCFillConstant = ip->xfc_fill_constant_bytes; - mode_lib->vba.DPPCLKDelaySubtotal = ip->dppclk_delay_subtotal; - mode_lib->vba.DPPCLKDelaySCL = ip->dppclk_delay_scl; - mode_lib->vba.DPPCLKDelaySCLLBOnly = ip->dppclk_delay_scl_lb_only; - mode_lib->vba.DPPCLKDelayCNVCFormater = ip->dppclk_delay_cnvc_formatter; - mode_lib->vba.DPPCLKDelayCNVCCursor = ip->dppclk_delay_cnvc_cursor; - mode_lib->vba.DISPCLKDelaySubtotal = ip->dispclk_delay_subtotal; - - mode_lib->vba.ProgressiveToInterlaceUnitInOPP = ip->ptoi_supported; - - mode_lib->vba.PDEProcessingBufIn64KBReqs = ip->pde_proc_buffer_size_64k_reqs; -} - -static void fetch_pipe_params(struct display_mode_lib *mode_lib) -{ - display_e2e_pipe_params_st *pipes = mode_lib->vba.cache_pipes; - ip_params_st *ip = &mode_lib->vba.ip; - - unsigned int OTGInstPlane[DC__NUM_DPP__MAX]; - unsigned int j, k; - bool PlaneVisited[DC__NUM_DPP__MAX]; - bool visited[DC__NUM_DPP__MAX]; - - // Convert Pipes to Planes - for (k = 0; k < mode_lib->vba.cache_num_pipes; ++k) - visited[k] = false; - - mode_lib->vba.NumberOfActivePlanes = 0; - for (j = 0; j < mode_lib->vba.cache_num_pipes; ++j) { - display_pipe_source_params_st *src = &pipes[j].pipe.src; - display_pipe_dest_params_st *dst = &pipes[j].pipe.dest; - scaler_ratio_depth_st *scl = &pipes[j].pipe.scale_ratio_depth; - scaler_taps_st *taps = &pipes[j].pipe.scale_taps; - display_output_params_st *dout = &pipes[j].dout; - display_clocks_and_cfg_st *clks = &pipes[j].clks_cfg; - - if (visited[j]) - continue; - visited[j] = true; - - mode_lib->vba.pipe_plane[j] = mode_lib->vba.NumberOfActivePlanes; - - mode_lib->vba.DPPPerPlane[mode_lib->vba.NumberOfActivePlanes] = 1; - mode_lib->vba.SourceScan[mode_lib->vba.NumberOfActivePlanes] = - (enum scan_direction_class) (src->source_scan); - mode_lib->vba.ViewportWidth[mode_lib->vba.NumberOfActivePlanes] = - src->viewport_width; - mode_lib->vba.ViewportHeight[mode_lib->vba.NumberOfActivePlanes] = - src->viewport_height; - mode_lib->vba.ViewportYStartY[mode_lib->vba.NumberOfActivePlanes] = - src->viewport_y_y; - mode_lib->vba.ViewportYStartC[mode_lib->vba.NumberOfActivePlanes] = - src->viewport_y_c; - mode_lib->vba.PitchY[mode_lib->vba.NumberOfActivePlanes] = src->data_pitch; - mode_lib->vba.PitchC[mode_lib->vba.NumberOfActivePlanes] = src->data_pitch_c; - mode_lib->vba.DCCMetaPitchY[mode_lib->vba.NumberOfActivePlanes] = src->meta_pitch; - mode_lib->vba.HRatio[mode_lib->vba.NumberOfActivePlanes] = scl->hscl_ratio; - mode_lib->vba.VRatio[mode_lib->vba.NumberOfActivePlanes] = scl->vscl_ratio; - mode_lib->vba.ScalerEnabled[mode_lib->vba.NumberOfActivePlanes] = scl->scl_enable; - mode_lib->vba.Interlace[mode_lib->vba.NumberOfActivePlanes] = dst->interlaced; - if (mode_lib->vba.Interlace[mode_lib->vba.NumberOfActivePlanes]) - mode_lib->vba.VRatio[mode_lib->vba.NumberOfActivePlanes] *= 2.0; - mode_lib->vba.htaps[mode_lib->vba.NumberOfActivePlanes] = taps->htaps; - mode_lib->vba.vtaps[mode_lib->vba.NumberOfActivePlanes] = taps->vtaps; - mode_lib->vba.HTAPsChroma[mode_lib->vba.NumberOfActivePlanes] = taps->htaps_c; - mode_lib->vba.VTAPsChroma[mode_lib->vba.NumberOfActivePlanes] = taps->vtaps_c; - mode_lib->vba.HTotal[mode_lib->vba.NumberOfActivePlanes] = dst->htotal; - mode_lib->vba.VTotal[mode_lib->vba.NumberOfActivePlanes] = dst->vtotal; - mode_lib->vba.DCCEnable[mode_lib->vba.NumberOfActivePlanes] = - src->dcc_use_global ? - ip->dcc_supported : src->dcc && ip->dcc_supported; - mode_lib->vba.DCCRate[mode_lib->vba.NumberOfActivePlanes] = src->dcc_rate; - mode_lib->vba.SourcePixelFormat[mode_lib->vba.NumberOfActivePlanes] = - (enum source_format_class) (src->source_format); - mode_lib->vba.HActive[mode_lib->vba.NumberOfActivePlanes] = dst->hactive; - mode_lib->vba.VActive[mode_lib->vba.NumberOfActivePlanes] = dst->vactive; - mode_lib->vba.SurfaceTiling[mode_lib->vba.NumberOfActivePlanes] = - (enum dm_swizzle_mode) (src->sw_mode); - mode_lib->vba.ScalerRecoutWidth[mode_lib->vba.NumberOfActivePlanes] = - dst->recout_width; // TODO: or should this be full_recout_width???...maybe only when in hsplit mode? - mode_lib->vba.ODMCombineEnabled[mode_lib->vba.NumberOfActivePlanes] = - dst->odm_combine; - mode_lib->vba.OutputFormat[mode_lib->vba.NumberOfActivePlanes] = - (enum output_format_class) (dout->output_format); - mode_lib->vba.Output[mode_lib->vba.NumberOfActivePlanes] = - (enum output_encoder_class) (dout->output_type); - mode_lib->vba.OutputBpp[mode_lib->vba.NumberOfActivePlanes] = dout->output_bpp; - mode_lib->vba.OutputLinkDPLanes[mode_lib->vba.NumberOfActivePlanes] = - dout->dp_lanes; - mode_lib->vba.DSCEnabled[mode_lib->vba.NumberOfActivePlanes] = dout->dsc_enable; - mode_lib->vba.NumberOfDSCSlices[mode_lib->vba.NumberOfActivePlanes] = - dout->dsc_slices; - mode_lib->vba.DSCInputBitPerComponent[mode_lib->vba.NumberOfActivePlanes] = - dout->opp_input_bpc == 0 ? 12 : dout->opp_input_bpc; - mode_lib->vba.WritebackEnable[mode_lib->vba.NumberOfActivePlanes] = dout->wb_enable; - mode_lib->vba.WritebackSourceHeight[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_src_height; - mode_lib->vba.WritebackDestinationWidth[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_dst_width; - mode_lib->vba.WritebackDestinationHeight[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_dst_height; - mode_lib->vba.WritebackPixelFormat[mode_lib->vba.NumberOfActivePlanes] = - (enum source_format_class) (dout->wb.wb_pixel_format); - mode_lib->vba.WritebackLumaHTaps[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_htaps_luma; - mode_lib->vba.WritebackLumaVTaps[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_vtaps_luma; - mode_lib->vba.WritebackChromaHTaps[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_htaps_chroma; - mode_lib->vba.WritebackChromaVTaps[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_vtaps_chroma; - mode_lib->vba.WritebackHRatio[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_hratio; - mode_lib->vba.WritebackVRatio[mode_lib->vba.NumberOfActivePlanes] = - dout->wb.wb_vratio; - - mode_lib->vba.DynamicMetadataEnable[mode_lib->vba.NumberOfActivePlanes] = - src->dynamic_metadata_enable; - mode_lib->vba.DynamicMetadataLinesBeforeActiveRequired[mode_lib->vba.NumberOfActivePlanes] = - src->dynamic_metadata_lines_before_active; - mode_lib->vba.DynamicMetadataTransmittedBytes[mode_lib->vba.NumberOfActivePlanes] = - src->dynamic_metadata_xmit_bytes; - - mode_lib->vba.XFCEnabled[mode_lib->vba.NumberOfActivePlanes] = src->xfc_enable - && ip->xfc_supported; - mode_lib->vba.XFCSlvChunkSize = src->xfc_params.xfc_slv_chunk_size_bytes; - mode_lib->vba.XFCTSlvVupdateOffset = src->xfc_params.xfc_tslv_vupdate_offset_us; - mode_lib->vba.XFCTSlvVupdateWidth = src->xfc_params.xfc_tslv_vupdate_width_us; - mode_lib->vba.XFCTSlvVreadyOffset = src->xfc_params.xfc_tslv_vready_offset_us; - mode_lib->vba.PixelClock[mode_lib->vba.NumberOfActivePlanes] = dst->pixel_rate_mhz; - mode_lib->vba.DPPCLK[mode_lib->vba.NumberOfActivePlanes] = clks->dppclk_mhz; - if (ip->is_line_buffer_bpp_fixed) - mode_lib->vba.LBBitPerPixel[mode_lib->vba.NumberOfActivePlanes] = - ip->line_buffer_fixed_bpp; - else { - unsigned int lb_depth; - - switch (scl->lb_depth) { - case dm_lb_6: - lb_depth = 18; - break; - case dm_lb_8: - lb_depth = 24; - break; - case dm_lb_10: - lb_depth = 30; - break; - case dm_lb_12: - lb_depth = 36; - break; - case dm_lb_16: - lb_depth = 48; - break; - default: - lb_depth = 36; - } - mode_lib->vba.LBBitPerPixel[mode_lib->vba.NumberOfActivePlanes] = lb_depth; - } - mode_lib->vba.NumberOfCursors[mode_lib->vba.NumberOfActivePlanes] = 0; - // The DML spreadsheet assumes that the two cursors utilize the same amount of bandwidth. We'll - // calculate things a little more accurately - for (k = 0; k < DC__NUM_CURSOR__MAX; ++k) { - switch (k) { - case 0: - mode_lib->vba.CursorBPP[mode_lib->vba.NumberOfActivePlanes][0] = - CursorBppEnumToBits( - (enum cursor_bpp) (src->cur0_bpp)); - mode_lib->vba.CursorWidth[mode_lib->vba.NumberOfActivePlanes][0] = - src->cur0_src_width; - if (src->cur0_src_width > 0) - mode_lib->vba.NumberOfCursors[mode_lib->vba.NumberOfActivePlanes]++; - break; - case 1: - mode_lib->vba.CursorBPP[mode_lib->vba.NumberOfActivePlanes][1] = - CursorBppEnumToBits( - (enum cursor_bpp) (src->cur1_bpp)); - mode_lib->vba.CursorWidth[mode_lib->vba.NumberOfActivePlanes][1] = - src->cur1_src_width; - if (src->cur1_src_width > 0) - mode_lib->vba.NumberOfCursors[mode_lib->vba.NumberOfActivePlanes]++; - break; - default: - dml_print( - "ERROR: Number of cursors specified exceeds supported maximum\n") - ; - } - } - - OTGInstPlane[mode_lib->vba.NumberOfActivePlanes] = dst->otg_inst; - - if (dst->odm_combine && !src->is_hsplit) - dml_print( - "ERROR: ODM Combine is specified but is_hsplit has not be specified for pipe %i\n", - j); - - if (src->is_hsplit) { - for (k = j + 1; k < mode_lib->vba.cache_num_pipes; ++k) { - display_pipe_source_params_st *src_k = &pipes[k].pipe.src; - display_output_params_st *dout_k = &pipes[k].dout; - - if (src_k->is_hsplit && !visited[k] - && src->hsplit_grp == src_k->hsplit_grp) { - mode_lib->vba.pipe_plane[k] = - mode_lib->vba.NumberOfActivePlanes; - mode_lib->vba.DPPPerPlane[mode_lib->vba.NumberOfActivePlanes]++; - if (mode_lib->vba.SourceScan[mode_lib->vba.NumberOfActivePlanes] - == dm_horz) - mode_lib->vba.ViewportWidth[mode_lib->vba.NumberOfActivePlanes] += - src_k->viewport_width; - else - mode_lib->vba.ViewportHeight[mode_lib->vba.NumberOfActivePlanes] += - src_k->viewport_height; - - mode_lib->vba.NumberOfDSCSlices[mode_lib->vba.NumberOfActivePlanes] += - dout_k->dsc_slices; - visited[k] = true; - } - } - } - - mode_lib->vba.NumberOfActivePlanes++; - } - - // handle overlays through dml_ml->vba.BlendingAndTiming - // dml_ml->vba.BlendingAndTiming tells you which instance to look at to get timing, the so called 'master' - - for (j = 0; j < mode_lib->vba.NumberOfActivePlanes; ++j) - PlaneVisited[j] = false; - - for (j = 0; j < mode_lib->vba.NumberOfActivePlanes; ++j) { - for (k = j + 1; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (!PlaneVisited[k] && OTGInstPlane[j] == OTGInstPlane[k]) { - // doesn't matter, so choose the smaller one - mode_lib->vba.BlendingAndTiming[j] = j; - PlaneVisited[j] = true; - mode_lib->vba.BlendingAndTiming[k] = j; - PlaneVisited[k] = true; - } - } - - if (!PlaneVisited[j]) { - mode_lib->vba.BlendingAndTiming[j] = j; - PlaneVisited[j] = true; - } - } - - // TODO: dml_ml->vba.ODMCombineEnabled => 2 * dml_ml->vba.DPPPerPlane...actually maybe not since all pipes are specified - // Do we want the dscclk to automatically be halved? Guess not since the value is specified - - mode_lib->vba.SynchronizedVBlank = pipes[0].pipe.dest.synchronized_vblank_all_planes; - for (k = 1; k < mode_lib->vba.cache_num_pipes; ++k) - ASSERT(mode_lib->vba.SynchronizedVBlank == pipes[k].pipe.dest.synchronized_vblank_all_planes); - - mode_lib->vba.VirtualMemoryEnable = false; - mode_lib->vba.OverridePageTableLevels = 0; - - for (k = 0; k < mode_lib->vba.cache_num_pipes; ++k) { - mode_lib->vba.VirtualMemoryEnable = mode_lib->vba.VirtualMemoryEnable - || !!pipes[k].pipe.src.vm; - mode_lib->vba.OverridePageTableLevels = - (pipes[k].pipe.src.vm_levels_force_en - && mode_lib->vba.OverridePageTableLevels - < pipes[k].pipe.src.vm_levels_force) ? - pipes[k].pipe.src.vm_levels_force : - mode_lib->vba.OverridePageTableLevels; - } - - if (mode_lib->vba.OverridePageTableLevels) - mode_lib->vba.MaxPageTableLevels = mode_lib->vba.OverridePageTableLevels; - - mode_lib->vba.VirtualMemoryEnable = mode_lib->vba.VirtualMemoryEnable && !!ip->pte_enable; - - mode_lib->vba.FabricAndDRAMBandwidth = dml_min( - mode_lib->vba.DRAMSpeed * mode_lib->vba.NumberOfChannels - * mode_lib->vba.DRAMChannelWidth, - mode_lib->vba.FabricClock * mode_lib->vba.FabricDatapathToDCNDataReturn) - / 1000.0; - - // TODO: Must be consistent across all pipes - // DCCProgrammingAssumesScanDirectionUnknown = src.dcc_scan_dir_unknown; -} - -static void recalculate(struct display_mode_lib *mode_lib) -{ - ModeSupportAndSystemConfiguration(mode_lib); - PixelClockAdjustmentForProgressiveToInterlaceUnit(mode_lib); - DisplayPipeConfiguration(mode_lib); - DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation(mode_lib); -} - -// in wm mode we pull the parameters needed from the display_e2e_pipe_params_st structs -// rather than working them out as in recalculate_ms -static void recalculate_params( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes) -{ - // This is only safe to use memcmp because there are non-POD types in struct display_mode_lib - if (memcmp(&mode_lib->soc, &mode_lib->vba.soc, sizeof(mode_lib->vba.soc)) != 0 - || memcmp(&mode_lib->ip, &mode_lib->vba.ip, sizeof(mode_lib->vba.ip)) != 0 - || num_pipes != mode_lib->vba.cache_num_pipes - || memcmp( - pipes, - mode_lib->vba.cache_pipes, - sizeof(display_e2e_pipe_params_st) * num_pipes) != 0) { - mode_lib->vba.soc = mode_lib->soc; - mode_lib->vba.ip = mode_lib->ip; - memcpy(mode_lib->vba.cache_pipes, pipes, sizeof(*pipes) * num_pipes); - mode_lib->vba.cache_num_pipes = num_pipes; - recalculate(mode_lib); - } -} - -static void ModeSupportAndSystemConfiguration(struct display_mode_lib *mode_lib) -{ - soc_bounding_box_st *soc = &mode_lib->vba.soc; - unsigned int i, k; - unsigned int total_pipes = 0; - - mode_lib->vba.VoltageLevel = mode_lib->vba.cache_pipes[0].clks_cfg.voltage; - for (i = 1; i < mode_lib->vba.cache_num_pipes; ++i) - ASSERT(mode_lib->vba.VoltageLevel == -1 || mode_lib->vba.VoltageLevel == mode_lib->vba.cache_pipes[i].clks_cfg.voltage); - - mode_lib->vba.DCFCLK = mode_lib->vba.cache_pipes[0].clks_cfg.dcfclk_mhz; - mode_lib->vba.SOCCLK = mode_lib->vba.cache_pipes[0].clks_cfg.socclk_mhz; - - if (mode_lib->vba.cache_pipes[0].clks_cfg.dispclk_mhz > 0.0) - mode_lib->vba.DISPCLK = mode_lib->vba.cache_pipes[0].clks_cfg.dispclk_mhz; - else - mode_lib->vba.DISPCLK = soc->clock_limits[mode_lib->vba.VoltageLevel].dispclk_mhz; - - fetch_socbb_params(mode_lib); - fetch_ip_params(mode_lib); - fetch_pipe_params(mode_lib); - - // Total Available Pipes Support Check - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) - total_pipes += mode_lib->vba.DPPPerPlane[k]; - ASSERT(total_pipes <= DC__NUM_DPP__MAX); -} - -static double adjust_ReturnBW( - struct display_mode_lib *mode_lib, - double ReturnBW, - bool DCCEnabledAnyPlane, - double ReturnBandwidthToDCN) -{ - double CriticalCompression; - - if (DCCEnabledAnyPlane - && ReturnBandwidthToDCN - > mode_lib->vba.DCFCLK * mode_lib->vba.ReturnBusWidth / 4.0) - ReturnBW = - dml_min( - ReturnBW, - ReturnBandwidthToDCN * 4 - * (1.0 - - mode_lib->vba.UrgentLatency - / ((mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024 - / ReturnBandwidthToDCN - - mode_lib->vba.DCFCLK - * mode_lib->vba.ReturnBusWidth - / 4) - + mode_lib->vba.UrgentLatency)); - - CriticalCompression = 2.0 * mode_lib->vba.ReturnBusWidth * mode_lib->vba.DCFCLK - * mode_lib->vba.UrgentLatency - / (ReturnBandwidthToDCN * mode_lib->vba.UrgentLatency - + (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024); - - if (DCCEnabledAnyPlane && CriticalCompression > 1.0 && CriticalCompression < 4.0) - ReturnBW = - dml_min( - ReturnBW, - 4.0 * ReturnBandwidthToDCN - * (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024 - * mode_lib->vba.ReturnBusWidth - * mode_lib->vba.DCFCLK - * mode_lib->vba.UrgentLatency - / dml_pow( - (ReturnBandwidthToDCN - * mode_lib->vba.UrgentLatency - + (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024), - 2)); - - return ReturnBW; -} - -static unsigned int dscceComputeDelay( - unsigned int bpc, - double bpp, - unsigned int sliceWidth, - unsigned int numSlices, - enum output_format_class pixelFormat) -{ - // valid bpc = source bits per component in the set of {8, 10, 12} - // valid bpp = increments of 1/16 of a bit - // min = 6/7/8 in N420/N422/444, respectively - // max = such that compression is 1:1 - //valid sliceWidth = number of pixels per slice line, must be less than or equal to 5184/numSlices (or 4096/numSlices in 420 mode) - //valid numSlices = number of slices in the horiziontal direction per DSC engine in the set of {1, 2, 3, 4} - //valid pixelFormat = pixel/color format in the set of {:N444_RGB, :S422, :N422, :N420} - - // fixed value - unsigned int rcModelSize = 8192; - - // N422/N420 operate at 2 pixels per clock - unsigned int pixelsPerClock, lstall, D, initalXmitDelay, w, s, ix, wx, p, l0, a, ax, l, - Delay, pixels; - - if (pixelFormat == dm_n422 || pixelFormat == dm_420) - pixelsPerClock = 2; - // #all other modes operate at 1 pixel per clock - else - pixelsPerClock = 1; - - //initial transmit delay as per PPS - initalXmitDelay = dml_round(rcModelSize / 2.0 / bpp / pixelsPerClock); - - //compute ssm delay - if (bpc == 8) - D = 81; - else if (bpc == 10) - D = 89; - else - D = 113; - - //divide by pixel per cycle to compute slice width as seen by DSC - w = sliceWidth / pixelsPerClock; - - //422 mode has an additional cycle of delay - if (pixelFormat == dm_s422) - s = 1; - else - s = 0; - - //main calculation for the dscce - ix = initalXmitDelay + 45; - wx = (w + 2) / 3; - p = 3 * wx - w; - l0 = ix / w; - a = ix + p * l0; - ax = (a + 2) / 3 + D + 6 + 1; - l = (ax + wx - 1) / wx; - if ((ix % w) == 0 && p != 0) - lstall = 1; - else - lstall = 0; - Delay = l * wx * (numSlices - 1) + ax + s + lstall + 22; - - //dsc processes 3 pixel containers per cycle and a container can contain 1 or 2 pixels - pixels = Delay * 3 * pixelsPerClock; - return pixels; -} - -static unsigned int dscComputeDelay(enum output_format_class pixelFormat) -{ - unsigned int Delay = 0; - - if (pixelFormat == dm_420) { - // sfr - Delay = Delay + 2; - // dsccif - Delay = Delay + 0; - // dscc - input deserializer - Delay = Delay + 3; - // dscc gets pixels every other cycle - Delay = Delay + 2; - // dscc - input cdc fifo - Delay = Delay + 12; - // dscc gets pixels every other cycle - Delay = Delay + 13; - // dscc - cdc uncertainty - Delay = Delay + 2; - // dscc - output cdc fifo - Delay = Delay + 7; - // dscc gets pixels every other cycle - Delay = Delay + 3; - // dscc - cdc uncertainty - Delay = Delay + 2; - // dscc - output serializer - Delay = Delay + 1; - // sft - Delay = Delay + 1; - } else if (pixelFormat == dm_n422) { - // sfr - Delay = Delay + 2; - // dsccif - Delay = Delay + 1; - // dscc - input deserializer - Delay = Delay + 5; - // dscc - input cdc fifo - Delay = Delay + 25; - // dscc - cdc uncertainty - Delay = Delay + 2; - // dscc - output cdc fifo - Delay = Delay + 10; - // dscc - cdc uncertainty - Delay = Delay + 2; - // dscc - output serializer - Delay = Delay + 1; - // sft - Delay = Delay + 1; - } else { - // sfr - Delay = Delay + 2; - // dsccif - Delay = Delay + 0; - // dscc - input deserializer - Delay = Delay + 3; - // dscc - input cdc fifo - Delay = Delay + 12; - // dscc - cdc uncertainty - Delay = Delay + 2; - // dscc - output cdc fifo - Delay = Delay + 7; - // dscc - output serializer - Delay = Delay + 1; - // dscc - cdc uncertainty - Delay = Delay + 2; - // sft - Delay = Delay + 1; - } - - return Delay; -} - -static bool CalculatePrefetchSchedule( - struct display_mode_lib *mode_lib, - double DPPCLK, - double DISPCLK, - double PixelClock, - double DCFClkDeepSleep, - unsigned int DSCDelay, - unsigned int DPPPerPlane, - bool ScalerEnabled, - unsigned int NumberOfCursors, - double DPPCLKDelaySubtotal, - double DPPCLKDelaySCL, - double DPPCLKDelaySCLLBOnly, - double DPPCLKDelayCNVCFormater, - double DPPCLKDelayCNVCCursor, - double DISPCLKDelaySubtotal, - unsigned int ScalerRecoutWidth, - enum output_format_class OutputFormat, - unsigned int VBlank, - unsigned int HTotal, - unsigned int MaxInterDCNTileRepeaters, - unsigned int VStartup, - unsigned int PageTableLevels, - bool VirtualMemoryEnable, - bool DynamicMetadataEnable, - unsigned int DynamicMetadataLinesBeforeActiveRequired, - unsigned int DynamicMetadataTransmittedBytes, - bool DCCEnable, - double UrgentLatency, - double UrgentExtraLatency, - double TCalc, - unsigned int PDEAndMetaPTEBytesFrame, - unsigned int MetaRowByte, - unsigned int PixelPTEBytesPerRow, - double PrefetchSourceLinesY, - unsigned int SwathWidthY, - double BytePerPixelDETY, - double VInitPreFillY, - unsigned int MaxNumSwathY, - double PrefetchSourceLinesC, - double BytePerPixelDETC, - double VInitPreFillC, - unsigned int MaxNumSwathC, - unsigned int SwathHeightY, - unsigned int SwathHeightC, - double TWait, - bool XFCEnabled, - double XFCRemoteSurfaceFlipDelay, - bool InterlaceEnable, - bool ProgressiveToInterlaceUnitInOPP, - double *DSTXAfterScaler, - double *DSTYAfterScaler, - double *DestinationLinesForPrefetch, - double *PrefetchBandwidth, - double *DestinationLinesToRequestVMInVBlank, - double *DestinationLinesToRequestRowInVBlank, - double *VRatioPrefetchY, - double *VRatioPrefetchC, - double *RequiredPrefetchPixDataBW, - unsigned int *VStartupRequiredWhenNotEnoughTimeForDynamicMetadata, - double *Tno_bw, - unsigned int *VUpdateOffsetPix, - unsigned int *VUpdateWidthPix, - unsigned int *VReadyOffsetPix) -{ - bool MyError = false; - unsigned int DPPCycles, DISPCLKCycles; - double DSTTotalPixelsAfterScaler, TotalRepeaterDelayTime; - double Tdm, LineTime, Tsetup; - double dst_y_prefetch_equ; - double Tsw_oto; - double prefetch_bw_oto; - double Tvm_oto; - double Tr0_oto; - double Tpre_oto; - double dst_y_prefetch_oto; - double TimeForFetchingMetaPTE = 0; - double TimeForFetchingRowInVBlank = 0; - double LinesToRequestPrefetchPixelData = 0; - - if (ScalerEnabled) - DPPCycles = DPPCLKDelaySubtotal + DPPCLKDelaySCL; - else - DPPCycles = DPPCLKDelaySubtotal + DPPCLKDelaySCLLBOnly; - - DPPCycles = DPPCycles + DPPCLKDelayCNVCFormater + NumberOfCursors * DPPCLKDelayCNVCCursor; - - DISPCLKCycles = DISPCLKDelaySubtotal; - - if (DPPCLK == 0.0 || DISPCLK == 0.0) - return true; - - *DSTXAfterScaler = DPPCycles * PixelClock / DPPCLK + DISPCLKCycles * PixelClock / DISPCLK - + DSCDelay; - - if (DPPPerPlane > 1) - *DSTXAfterScaler = *DSTXAfterScaler + ScalerRecoutWidth; - - if (OutputFormat == dm_420 || (InterlaceEnable && ProgressiveToInterlaceUnitInOPP)) - *DSTYAfterScaler = 1; - else - *DSTYAfterScaler = 0; - - DSTTotalPixelsAfterScaler = ((double) (*DSTYAfterScaler * HTotal)) + *DSTXAfterScaler; - *DSTYAfterScaler = dml_floor(DSTTotalPixelsAfterScaler / HTotal, 1); - *DSTXAfterScaler = DSTTotalPixelsAfterScaler - ((double) (*DSTYAfterScaler * HTotal)); - - *VUpdateOffsetPix = dml_ceil(HTotal / 4.0, 1); - TotalRepeaterDelayTime = MaxInterDCNTileRepeaters * (2.0 / DPPCLK + 3.0 / DISPCLK); - *VUpdateWidthPix = (14.0 / DCFClkDeepSleep + 12.0 / DPPCLK + TotalRepeaterDelayTime) - * PixelClock; - - *VReadyOffsetPix = dml_max( - 150.0 / DPPCLK, - TotalRepeaterDelayTime + 20.0 / DCFClkDeepSleep + 10.0 / DPPCLK) - * PixelClock; - - Tsetup = (double) (*VUpdateOffsetPix + *VUpdateWidthPix + *VReadyOffsetPix) / PixelClock; - - LineTime = (double) HTotal / PixelClock; - - if (DynamicMetadataEnable) { - double Tdmbf, Tdmec, Tdmsks; - - Tdm = dml_max(0.0, UrgentExtraLatency - TCalc); - Tdmbf = DynamicMetadataTransmittedBytes / 4.0 / DISPCLK; - Tdmec = LineTime; - if (DynamicMetadataLinesBeforeActiveRequired == 0) - Tdmsks = VBlank * LineTime / 2.0; - else - Tdmsks = DynamicMetadataLinesBeforeActiveRequired * LineTime; - if (InterlaceEnable && !ProgressiveToInterlaceUnitInOPP) - Tdmsks = Tdmsks / 2; - if (VStartup * LineTime - < Tsetup + TWait + UrgentExtraLatency + Tdmbf + Tdmec + Tdmsks) { - MyError = true; - *VStartupRequiredWhenNotEnoughTimeForDynamicMetadata = (Tsetup + TWait - + UrgentExtraLatency + Tdmbf + Tdmec + Tdmsks) / LineTime; - } else - *VStartupRequiredWhenNotEnoughTimeForDynamicMetadata = 0.0; - } else - Tdm = 0; - - if (VirtualMemoryEnable) { - if (PageTableLevels == 4) - *Tno_bw = UrgentExtraLatency + UrgentLatency; - else if (PageTableLevels == 3) - *Tno_bw = UrgentExtraLatency; - else - *Tno_bw = 0; - } else if (DCCEnable) - *Tno_bw = LineTime; - else - *Tno_bw = LineTime / 4; - - dst_y_prefetch_equ = VStartup - dml_max(TCalc + TWait, XFCRemoteSurfaceFlipDelay) / LineTime - - (Tsetup + Tdm) / LineTime - - (*DSTYAfterScaler + *DSTXAfterScaler / HTotal); - - Tsw_oto = dml_max(PrefetchSourceLinesY, PrefetchSourceLinesC) * LineTime; - - prefetch_bw_oto = (MetaRowByte + PixelPTEBytesPerRow - + PrefetchSourceLinesY * SwathWidthY * dml_ceil(BytePerPixelDETY, 1) - + PrefetchSourceLinesC * SwathWidthY / 2 * dml_ceil(BytePerPixelDETC, 2)) - / Tsw_oto; - - if (VirtualMemoryEnable == true) { - Tvm_oto = - dml_max( - *Tno_bw + PDEAndMetaPTEBytesFrame / prefetch_bw_oto, - dml_max( - UrgentExtraLatency - + UrgentLatency - * (PageTableLevels - - 1), - LineTime / 4.0)); - } else - Tvm_oto = LineTime / 4.0; - - if ((VirtualMemoryEnable == true || DCCEnable == true)) { - Tr0_oto = dml_max( - (MetaRowByte + PixelPTEBytesPerRow) / prefetch_bw_oto, - dml_max(UrgentLatency, dml_max(LineTime - Tvm_oto, LineTime / 4))); - } else - Tr0_oto = LineTime - Tvm_oto; - - Tpre_oto = Tvm_oto + Tr0_oto + Tsw_oto; - - dst_y_prefetch_oto = Tpre_oto / LineTime; - - if (dst_y_prefetch_oto < dst_y_prefetch_equ) - *DestinationLinesForPrefetch = dst_y_prefetch_oto; - else - *DestinationLinesForPrefetch = dst_y_prefetch_equ; - - *DestinationLinesForPrefetch = dml_floor(4.0 * (*DestinationLinesForPrefetch + 0.125), 1) - / 4; - - dml_print("DML: VStartup: %d\n", VStartup); - dml_print("DML: TCalc: %f\n", TCalc); - dml_print("DML: TWait: %f\n", TWait); - dml_print("DML: XFCRemoteSurfaceFlipDelay: %f\n", XFCRemoteSurfaceFlipDelay); - dml_print("DML: LineTime: %f\n", LineTime); - dml_print("DML: Tsetup: %f\n", Tsetup); - dml_print("DML: Tdm: %f\n", Tdm); - dml_print("DML: DSTYAfterScaler: %f\n", *DSTYAfterScaler); - dml_print("DML: DSTXAfterScaler: %f\n", *DSTXAfterScaler); - dml_print("DML: HTotal: %d\n", HTotal); - - *PrefetchBandwidth = 0; - *DestinationLinesToRequestVMInVBlank = 0; - *DestinationLinesToRequestRowInVBlank = 0; - *VRatioPrefetchY = 0; - *VRatioPrefetchC = 0; - *RequiredPrefetchPixDataBW = 0; - if (*DestinationLinesForPrefetch > 1) { - *PrefetchBandwidth = (PDEAndMetaPTEBytesFrame + 2 * MetaRowByte - + 2 * PixelPTEBytesPerRow - + PrefetchSourceLinesY * SwathWidthY * dml_ceil(BytePerPixelDETY, 1) - + PrefetchSourceLinesC * SwathWidthY / 2 - * dml_ceil(BytePerPixelDETC, 2)) - / (*DestinationLinesForPrefetch * LineTime - *Tno_bw); - if (VirtualMemoryEnable) { - TimeForFetchingMetaPTE = - dml_max( - *Tno_bw - + (double) PDEAndMetaPTEBytesFrame - / *PrefetchBandwidth, - dml_max( - UrgentExtraLatency - + UrgentLatency - * (PageTableLevels - - 1), - LineTime / 4)); - } else { - if (NumberOfCursors > 0 || XFCEnabled) - TimeForFetchingMetaPTE = LineTime / 4; - else - TimeForFetchingMetaPTE = 0.0; - } - - if ((VirtualMemoryEnable == true || DCCEnable == true)) { - TimeForFetchingRowInVBlank = - dml_max( - (MetaRowByte + PixelPTEBytesPerRow) - / *PrefetchBandwidth, - dml_max( - UrgentLatency, - dml_max( - LineTime - - TimeForFetchingMetaPTE, - LineTime - / 4.0))); - } else { - if (NumberOfCursors > 0 || XFCEnabled) - TimeForFetchingRowInVBlank = LineTime - TimeForFetchingMetaPTE; - else - TimeForFetchingRowInVBlank = 0.0; - } - - *DestinationLinesToRequestVMInVBlank = dml_floor( - 4.0 * (TimeForFetchingMetaPTE / LineTime + 0.125), - 1) / 4.0; - - *DestinationLinesToRequestRowInVBlank = dml_floor( - 4.0 * (TimeForFetchingRowInVBlank / LineTime + 0.125), - 1) / 4.0; - - LinesToRequestPrefetchPixelData = - *DestinationLinesForPrefetch - - ((NumberOfCursors > 0 || VirtualMemoryEnable - || DCCEnable) ? - (*DestinationLinesToRequestVMInVBlank - + *DestinationLinesToRequestRowInVBlank) : - 0.0); - - if (LinesToRequestPrefetchPixelData > 0) { - - *VRatioPrefetchY = (double) PrefetchSourceLinesY - / LinesToRequestPrefetchPixelData; - *VRatioPrefetchY = dml_max(*VRatioPrefetchY, 1.0); - if ((SwathHeightY > 4) && (VInitPreFillY > 3)) { - if (LinesToRequestPrefetchPixelData > (VInitPreFillY - 3.0) / 2.0) { - *VRatioPrefetchY = - dml_max( - (double) PrefetchSourceLinesY - / LinesToRequestPrefetchPixelData, - (double) MaxNumSwathY - * SwathHeightY - / (LinesToRequestPrefetchPixelData - - (VInitPreFillY - - 3.0) - / 2.0)); - *VRatioPrefetchY = dml_max(*VRatioPrefetchY, 1.0); - } else { - MyError = true; - *VRatioPrefetchY = 0; - } - } - - *VRatioPrefetchC = (double) PrefetchSourceLinesC - / LinesToRequestPrefetchPixelData; - *VRatioPrefetchC = dml_max(*VRatioPrefetchC, 1.0); - - if ((SwathHeightC > 4)) { - if (LinesToRequestPrefetchPixelData > (VInitPreFillC - 3.0) / 2.0) { - *VRatioPrefetchC = - dml_max( - *VRatioPrefetchC, - (double) MaxNumSwathC - * SwathHeightC - / (LinesToRequestPrefetchPixelData - - (VInitPreFillC - - 3.0) - / 2.0)); - *VRatioPrefetchC = dml_max(*VRatioPrefetchC, 1.0); - } else { - MyError = true; - *VRatioPrefetchC = 0; - } - } - - *RequiredPrefetchPixDataBW = - DPPPerPlane - * ((double) PrefetchSourceLinesY - / LinesToRequestPrefetchPixelData - * dml_ceil( - BytePerPixelDETY, - 1) - + (double) PrefetchSourceLinesC - / LinesToRequestPrefetchPixelData - * dml_ceil( - BytePerPixelDETC, - 2) - / 2) - * SwathWidthY / LineTime; - } else { - MyError = true; - *VRatioPrefetchY = 0; - *VRatioPrefetchC = 0; - *RequiredPrefetchPixDataBW = 0; - } - - } else { - MyError = true; - } - - if (MyError) { - *PrefetchBandwidth = 0; - TimeForFetchingMetaPTE = 0; - TimeForFetchingRowInVBlank = 0; - *DestinationLinesToRequestVMInVBlank = 0; - *DestinationLinesToRequestRowInVBlank = 0; - *DestinationLinesForPrefetch = 0; - LinesToRequestPrefetchPixelData = 0; - *VRatioPrefetchY = 0; - *VRatioPrefetchC = 0; - *RequiredPrefetchPixDataBW = 0; - } - - return MyError; -} - -static double RoundToDFSGranularityUp(double Clock, double VCOSpeed) -{ - return VCOSpeed * 4 / dml_floor(VCOSpeed * 4 / Clock, 1); -} - -static double RoundToDFSGranularityDown(double Clock, double VCOSpeed) -{ - return VCOSpeed * 4 / dml_ceil(VCOSpeed * 4 / Clock, 1); -} - -static double CalculatePrefetchSourceLines( - struct display_mode_lib *mode_lib, - double VRatio, - double vtaps, - bool Interlace, - bool ProgressiveToInterlaceUnitInOPP, - unsigned int SwathHeight, - unsigned int ViewportYStart, - double *VInitPreFill, - unsigned int *MaxNumSwath) -{ - unsigned int MaxPartialSwath; - - if (ProgressiveToInterlaceUnitInOPP) - *VInitPreFill = dml_floor((VRatio + vtaps + 1) / 2.0, 1); - else - *VInitPreFill = dml_floor((VRatio + vtaps + 1 + Interlace * 0.5 * VRatio) / 2.0, 1); - - if (!mode_lib->vba.IgnoreViewportPositioning) { - - *MaxNumSwath = dml_ceil((*VInitPreFill - 1.0) / SwathHeight, 1) + 1.0; - - if (*VInitPreFill > 1.0) - MaxPartialSwath = (unsigned int) (*VInitPreFill - 2) % SwathHeight; - else - MaxPartialSwath = (unsigned int) (*VInitPreFill + SwathHeight - 2) - % SwathHeight; - MaxPartialSwath = dml_max(1U, MaxPartialSwath); - - } else { - - if (ViewportYStart != 0) - dml_print( - "WARNING DML: using viewport y position of 0 even though actual viewport y position is non-zero in prefetch source lines calculation\n"); - - *MaxNumSwath = dml_ceil(*VInitPreFill / SwathHeight, 1); - - if (*VInitPreFill > 1.0) - MaxPartialSwath = (unsigned int) (*VInitPreFill - 1) % SwathHeight; - else - MaxPartialSwath = (unsigned int) (*VInitPreFill + SwathHeight - 1) - % SwathHeight; - } - - return *MaxNumSwath * SwathHeight + MaxPartialSwath; -} - -static unsigned int CalculateVMAndRowBytes( - struct display_mode_lib *mode_lib, - bool DCCEnable, - unsigned int BlockHeight256Bytes, - unsigned int BlockWidth256Bytes, - enum source_format_class SourcePixelFormat, - unsigned int SurfaceTiling, - unsigned int BytePerPixel, - enum scan_direction_class ScanDirection, - unsigned int ViewportWidth, - unsigned int ViewportHeight, - unsigned int SwathWidth, - bool VirtualMemoryEnable, - unsigned int VMMPageSize, - unsigned int PTEBufferSizeInRequests, - unsigned int PDEProcessingBufIn64KBReqs, - unsigned int Pitch, - unsigned int DCCMetaPitch, - unsigned int *MacroTileWidth, - unsigned int *MetaRowByte, - unsigned int *PixelPTEBytesPerRow, - bool *PTEBufferSizeNotExceeded, - unsigned int *dpte_row_height, - unsigned int *meta_row_height) -{ - unsigned int MetaRequestHeight; - unsigned int MetaRequestWidth; - unsigned int MetaSurfWidth; - unsigned int MetaSurfHeight; - unsigned int MPDEBytesFrame; - unsigned int MetaPTEBytesFrame; - unsigned int DCCMetaSurfaceBytes; - - unsigned int MacroTileSizeBytes; - unsigned int MacroTileHeight; - unsigned int DPDE0BytesFrame; - unsigned int ExtraDPDEBytesFrame; - unsigned int PDEAndMetaPTEBytesFrame; - - if (DCCEnable == true) { - MetaRequestHeight = 8 * BlockHeight256Bytes; - MetaRequestWidth = 8 * BlockWidth256Bytes; - if (ScanDirection == dm_horz) { - *meta_row_height = MetaRequestHeight; - MetaSurfWidth = dml_ceil((double) SwathWidth - 1, MetaRequestWidth) - + MetaRequestWidth; - *MetaRowByte = MetaSurfWidth * MetaRequestHeight * BytePerPixel / 256.0; - } else { - *meta_row_height = MetaRequestWidth; - MetaSurfHeight = dml_ceil((double) SwathWidth - 1, MetaRequestHeight) - + MetaRequestHeight; - *MetaRowByte = MetaSurfHeight * MetaRequestWidth * BytePerPixel / 256.0; - } - if (ScanDirection == dm_horz) { - DCCMetaSurfaceBytes = DCCMetaPitch - * (dml_ceil(ViewportHeight - 1, 64 * BlockHeight256Bytes) - + 64 * BlockHeight256Bytes) * BytePerPixel - / 256; - } else { - DCCMetaSurfaceBytes = DCCMetaPitch - * (dml_ceil( - (double) ViewportHeight - 1, - 64 * BlockHeight256Bytes) - + 64 * BlockHeight256Bytes) * BytePerPixel - / 256; - } - if (VirtualMemoryEnable == true) { - MetaPTEBytesFrame = (dml_ceil( - (double) (DCCMetaSurfaceBytes - VMMPageSize) - / (8 * VMMPageSize), - 1) + 1) * 64; - MPDEBytesFrame = 128 * (mode_lib->vba.MaxPageTableLevels - 1); - } else { - MetaPTEBytesFrame = 0; - MPDEBytesFrame = 0; - } - } else { - MetaPTEBytesFrame = 0; - MPDEBytesFrame = 0; - *MetaRowByte = 0; - } - - if (SurfaceTiling == dm_sw_linear) { - MacroTileSizeBytes = 256; - MacroTileHeight = 1; - } else if (SurfaceTiling == dm_sw_4kb_s || SurfaceTiling == dm_sw_4kb_s_x - || SurfaceTiling == dm_sw_4kb_d || SurfaceTiling == dm_sw_4kb_d_x) { - MacroTileSizeBytes = 4096; - MacroTileHeight = 4 * BlockHeight256Bytes; - } else if (SurfaceTiling == dm_sw_64kb_s || SurfaceTiling == dm_sw_64kb_s_t - || SurfaceTiling == dm_sw_64kb_s_x || SurfaceTiling == dm_sw_64kb_d - || SurfaceTiling == dm_sw_64kb_d_t || SurfaceTiling == dm_sw_64kb_d_x - || SurfaceTiling == dm_sw_64kb_r_x) { - MacroTileSizeBytes = 65536; - MacroTileHeight = 16 * BlockHeight256Bytes; - } else { - MacroTileSizeBytes = 262144; - MacroTileHeight = 32 * BlockHeight256Bytes; - } - *MacroTileWidth = MacroTileSizeBytes / BytePerPixel / MacroTileHeight; - - if (VirtualMemoryEnable == true && mode_lib->vba.MaxPageTableLevels > 1) { - if (ScanDirection == dm_horz) { - DPDE0BytesFrame = - 64 - * (dml_ceil( - ((Pitch - * (dml_ceil( - ViewportHeight - - 1, - MacroTileHeight) - + MacroTileHeight) - * BytePerPixel) - - MacroTileSizeBytes) - / (8 - * 2097152), - 1) + 1); - } else { - DPDE0BytesFrame = - 64 - * (dml_ceil( - ((Pitch - * (dml_ceil( - (double) SwathWidth - - 1, - MacroTileHeight) - + MacroTileHeight) - * BytePerPixel) - - MacroTileSizeBytes) - / (8 - * 2097152), - 1) + 1); - } - ExtraDPDEBytesFrame = 128 * (mode_lib->vba.MaxPageTableLevels - 2); - } else { - DPDE0BytesFrame = 0; - ExtraDPDEBytesFrame = 0; - } - - PDEAndMetaPTEBytesFrame = MetaPTEBytesFrame + MPDEBytesFrame + DPDE0BytesFrame - + ExtraDPDEBytesFrame; - - if (VirtualMemoryEnable == true) { - unsigned int PTERequestSize; - unsigned int PixelPTEReqHeight; - unsigned int PixelPTEReqWidth; - double FractionOfPTEReturnDrop; - unsigned int EffectivePDEProcessingBufIn64KBReqs; - - if (SurfaceTiling == dm_sw_linear) { - PixelPTEReqHeight = 1; - PixelPTEReqWidth = 8.0 * VMMPageSize / BytePerPixel; - PTERequestSize = 64; - FractionOfPTEReturnDrop = 0; - } else if (MacroTileSizeBytes == 4096) { - PixelPTEReqHeight = MacroTileHeight; - PixelPTEReqWidth = 8 * *MacroTileWidth; - PTERequestSize = 64; - if (ScanDirection == dm_horz) - FractionOfPTEReturnDrop = 0; - else - FractionOfPTEReturnDrop = 7 / 8; - } else if (VMMPageSize == 4096 && MacroTileSizeBytes > 4096) { - PixelPTEReqHeight = 16 * BlockHeight256Bytes; - PixelPTEReqWidth = 16 * BlockWidth256Bytes; - PTERequestSize = 128; - FractionOfPTEReturnDrop = 0; - } else { - PixelPTEReqHeight = MacroTileHeight; - PixelPTEReqWidth = 8 * *MacroTileWidth; - PTERequestSize = 64; - FractionOfPTEReturnDrop = 0; - } - - if (SourcePixelFormat == dm_420_8 || SourcePixelFormat == dm_420_10) - EffectivePDEProcessingBufIn64KBReqs = PDEProcessingBufIn64KBReqs / 2; - else - EffectivePDEProcessingBufIn64KBReqs = PDEProcessingBufIn64KBReqs; - - if (SurfaceTiling == dm_sw_linear) { - *dpte_row_height = - dml_min( - 128, - 1 - << (unsigned int) dml_floor( - dml_log2( - dml_min( - (double) PTEBufferSizeInRequests - * PixelPTEReqWidth, - EffectivePDEProcessingBufIn64KBReqs - * 65536.0 - / BytePerPixel) - / Pitch), - 1)); - *PixelPTEBytesPerRow = PTERequestSize - * (dml_ceil( - (double) (Pitch * *dpte_row_height - 1) - / PixelPTEReqWidth, - 1) + 1); - } else if (ScanDirection == dm_horz) { - *dpte_row_height = PixelPTEReqHeight; - *PixelPTEBytesPerRow = PTERequestSize - * (dml_ceil(((double) SwathWidth - 1) / PixelPTEReqWidth, 1) - + 1); - } else { - *dpte_row_height = dml_min(PixelPTEReqWidth, *MacroTileWidth); - *PixelPTEBytesPerRow = PTERequestSize - * (dml_ceil( - ((double) SwathWidth - 1) - / PixelPTEReqHeight, - 1) + 1); - } - if (*PixelPTEBytesPerRow * (1 - FractionOfPTEReturnDrop) - <= 64 * PTEBufferSizeInRequests) { - *PTEBufferSizeNotExceeded = true; - } else { - *PTEBufferSizeNotExceeded = false; - } - } else { - *PixelPTEBytesPerRow = 0; - *PTEBufferSizeNotExceeded = true; - } - - return PDEAndMetaPTEBytesFrame; -} - -static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( - struct display_mode_lib *mode_lib) -{ - unsigned int j, k; - - mode_lib->vba.WritebackDISPCLK = 0.0; - mode_lib->vba.DISPCLKWithRamping = 0; - mode_lib->vba.DISPCLKWithoutRamping = 0; - mode_lib->vba.GlobalDPPCLK = 0.0; - - // dml_ml->vba.DISPCLK and dml_ml->vba.DPPCLK Calculation - // - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.WritebackEnable[k]) { - mode_lib->vba.WritebackDISPCLK = - dml_max( - mode_lib->vba.WritebackDISPCLK, - CalculateWriteBackDISPCLK( - mode_lib->vba.WritebackPixelFormat[k], - mode_lib->vba.PixelClock[k], - mode_lib->vba.WritebackHRatio[k], - mode_lib->vba.WritebackVRatio[k], - mode_lib->vba.WritebackLumaHTaps[k], - mode_lib->vba.WritebackLumaVTaps[k], - mode_lib->vba.WritebackChromaHTaps[k], - mode_lib->vba.WritebackChromaVTaps[k], - mode_lib->vba.WritebackDestinationWidth[k], - mode_lib->vba.HTotal[k], - mode_lib->vba.WritebackChromaLineBufferWidth)); - } - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.HRatio[k] > 1) { - mode_lib->vba.PSCL_THROUGHPUT_LUMA[k] = dml_min( - mode_lib->vba.MaxDCHUBToPSCLThroughput, - mode_lib->vba.MaxPSCLToLBThroughput - * mode_lib->vba.HRatio[k] - / dml_ceil( - mode_lib->vba.htaps[k] - / 6.0, - 1)); - } else { - mode_lib->vba.PSCL_THROUGHPUT_LUMA[k] = dml_min( - mode_lib->vba.MaxDCHUBToPSCLThroughput, - mode_lib->vba.MaxPSCLToLBThroughput); - } - - mode_lib->vba.DPPCLKUsingSingleDPPLuma = - mode_lib->vba.PixelClock[k] - * dml_max( - mode_lib->vba.vtaps[k] / 6.0 - * dml_min( - 1.0, - mode_lib->vba.HRatio[k]), - dml_max( - mode_lib->vba.HRatio[k] - * mode_lib->vba.VRatio[k] - / mode_lib->vba.PSCL_THROUGHPUT_LUMA[k], - 1.0)); - - if ((mode_lib->vba.htaps[k] > 6 || mode_lib->vba.vtaps[k] > 6) - && mode_lib->vba.DPPCLKUsingSingleDPPLuma - < 2 * mode_lib->vba.PixelClock[k]) { - mode_lib->vba.DPPCLKUsingSingleDPPLuma = 2 * mode_lib->vba.PixelClock[k]; - } - - if ((mode_lib->vba.SourcePixelFormat[k] != dm_420_8 - && mode_lib->vba.SourcePixelFormat[k] != dm_420_10)) { - mode_lib->vba.PSCL_THROUGHPUT_CHROMA[k] = 0.0; - mode_lib->vba.DPPCLKUsingSingleDPP[k] = - mode_lib->vba.DPPCLKUsingSingleDPPLuma; - } else { - if (mode_lib->vba.HRatio[k] > 1) { - mode_lib->vba.PSCL_THROUGHPUT_CHROMA[k] = - dml_min( - mode_lib->vba.MaxDCHUBToPSCLThroughput, - mode_lib->vba.MaxPSCLToLBThroughput - * mode_lib->vba.HRatio[k] - / 2 - / dml_ceil( - mode_lib->vba.HTAPsChroma[k] - / 6.0, - 1.0)); - } else { - mode_lib->vba.PSCL_THROUGHPUT_CHROMA[k] = dml_min( - mode_lib->vba.MaxDCHUBToPSCLThroughput, - mode_lib->vba.MaxPSCLToLBThroughput); - } - mode_lib->vba.DPPCLKUsingSingleDPPChroma = - mode_lib->vba.PixelClock[k] - * dml_max( - mode_lib->vba.VTAPsChroma[k] - / 6.0 - * dml_min( - 1.0, - mode_lib->vba.HRatio[k] - / 2), - dml_max( - mode_lib->vba.HRatio[k] - * mode_lib->vba.VRatio[k] - / 4 - / mode_lib->vba.PSCL_THROUGHPUT_CHROMA[k], - 1.0)); - - if ((mode_lib->vba.HTAPsChroma[k] > 6 || mode_lib->vba.VTAPsChroma[k] > 6) - && mode_lib->vba.DPPCLKUsingSingleDPPChroma - < 2 * mode_lib->vba.PixelClock[k]) { - mode_lib->vba.DPPCLKUsingSingleDPPChroma = 2 - * mode_lib->vba.PixelClock[k]; - } - - mode_lib->vba.DPPCLKUsingSingleDPP[k] = dml_max( - mode_lib->vba.DPPCLKUsingSingleDPPLuma, - mode_lib->vba.DPPCLKUsingSingleDPPChroma); - } - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.BlendingAndTiming[k] != k) - continue; - if (mode_lib->vba.ODMCombineEnabled[k]) { - mode_lib->vba.DISPCLKWithRamping = - dml_max( - mode_lib->vba.DISPCLKWithRamping, - mode_lib->vba.PixelClock[k] / 2 - * (1 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100) - * (1 - + mode_lib->vba.DISPCLKRampingMargin - / 100)); - mode_lib->vba.DISPCLKWithoutRamping = - dml_max( - mode_lib->vba.DISPCLKWithoutRamping, - mode_lib->vba.PixelClock[k] / 2 - * (1 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100)); - } else if (!mode_lib->vba.ODMCombineEnabled[k]) { - mode_lib->vba.DISPCLKWithRamping = - dml_max( - mode_lib->vba.DISPCLKWithRamping, - mode_lib->vba.PixelClock[k] - * (1 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100) - * (1 - + mode_lib->vba.DISPCLKRampingMargin - / 100)); - mode_lib->vba.DISPCLKWithoutRamping = - dml_max( - mode_lib->vba.DISPCLKWithoutRamping, - mode_lib->vba.PixelClock[k] - * (1 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100)); - } - } - - mode_lib->vba.DISPCLKWithRamping = dml_max( - mode_lib->vba.DISPCLKWithRamping, - mode_lib->vba.WritebackDISPCLK); - mode_lib->vba.DISPCLKWithoutRamping = dml_max( - mode_lib->vba.DISPCLKWithoutRamping, - mode_lib->vba.WritebackDISPCLK); - - ASSERT(mode_lib->vba.DISPCLKDPPCLKVCOSpeed != 0); - mode_lib->vba.DISPCLKWithRampingRoundedToDFSGranularity = RoundToDFSGranularityUp( - mode_lib->vba.DISPCLKWithRamping, - mode_lib->vba.DISPCLKDPPCLKVCOSpeed); - mode_lib->vba.DISPCLKWithoutRampingRoundedToDFSGranularity = RoundToDFSGranularityUp( - mode_lib->vba.DISPCLKWithoutRamping, - mode_lib->vba.DISPCLKDPPCLKVCOSpeed); - mode_lib->vba.MaxDispclkRoundedToDFSGranularity = RoundToDFSGranularityDown( - mode_lib->vba.soc.clock_limits[NumberOfStates - 1].dispclk_mhz, - mode_lib->vba.DISPCLKDPPCLKVCOSpeed); - if (mode_lib->vba.DISPCLKWithoutRampingRoundedToDFSGranularity - > mode_lib->vba.MaxDispclkRoundedToDFSGranularity) { - mode_lib->vba.DISPCLK_calculated = - mode_lib->vba.DISPCLKWithoutRampingRoundedToDFSGranularity; - } else if (mode_lib->vba.DISPCLKWithRampingRoundedToDFSGranularity - > mode_lib->vba.MaxDispclkRoundedToDFSGranularity) { - mode_lib->vba.DISPCLK_calculated = mode_lib->vba.MaxDispclkRoundedToDFSGranularity; - } else { - mode_lib->vba.DISPCLK_calculated = - mode_lib->vba.DISPCLKWithRampingRoundedToDFSGranularity; - } - DTRACE(" dispclk_mhz (calculated) = %f", mode_lib->vba.DISPCLK_calculated); - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.DPPCLK_calculated[k] = mode_lib->vba.DPPCLKUsingSingleDPP[k] - / mode_lib->vba.DPPPerPlane[k] - * (1 + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading / 100); - mode_lib->vba.GlobalDPPCLK = dml_max( - mode_lib->vba.GlobalDPPCLK, - mode_lib->vba.DPPCLK_calculated[k]); - } - mode_lib->vba.GlobalDPPCLK = RoundToDFSGranularityUp( - mode_lib->vba.GlobalDPPCLK, - mode_lib->vba.DISPCLKDPPCLKVCOSpeed); - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.DPPCLK_calculated[k] = mode_lib->vba.GlobalDPPCLK / 255 - * dml_ceil( - mode_lib->vba.DPPCLK_calculated[k] * 255 - / mode_lib->vba.GlobalDPPCLK, - 1); - DTRACE(" dppclk_mhz[%i] (calculated) = %f", k, mode_lib->vba.DPPCLK_calculated[k]); - } - - // Urgent Watermark - mode_lib->vba.DCCEnabledAnyPlane = false; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) - if (mode_lib->vba.DCCEnable[k]) - mode_lib->vba.DCCEnabledAnyPlane = true; - - mode_lib->vba.ReturnBandwidthToDCN = dml_min( - mode_lib->vba.ReturnBusWidth * mode_lib->vba.DCFCLK, - mode_lib->vba.FabricAndDRAMBandwidth * 1000) - * mode_lib->vba.PercentOfIdealDRAMAndFabricBWReceivedAfterUrgLatency / 100; - - mode_lib->vba.ReturnBW = mode_lib->vba.ReturnBandwidthToDCN; - mode_lib->vba.ReturnBW = adjust_ReturnBW( - mode_lib, - mode_lib->vba.ReturnBW, - mode_lib->vba.DCCEnabledAnyPlane, - mode_lib->vba.ReturnBandwidthToDCN); - - // Let's do this calculation again?? - mode_lib->vba.ReturnBandwidthToDCN = dml_min( - mode_lib->vba.ReturnBusWidth * mode_lib->vba.DCFCLK, - mode_lib->vba.FabricAndDRAMBandwidth * 1000); - mode_lib->vba.ReturnBW = adjust_ReturnBW( - mode_lib, - mode_lib->vba.ReturnBW, - mode_lib->vba.DCCEnabledAnyPlane, - mode_lib->vba.ReturnBandwidthToDCN); - - DTRACE(" dcfclk_mhz = %f", mode_lib->vba.DCFCLK); - DTRACE(" return_bw_to_dcn = %f", mode_lib->vba.ReturnBandwidthToDCN); - DTRACE(" return_bus_bw = %f", mode_lib->vba.ReturnBW); - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - bool MainPlaneDoesODMCombine = false; - - if (mode_lib->vba.SourceScan[k] == dm_horz) - mode_lib->vba.SwathWidthSingleDPPY[k] = mode_lib->vba.ViewportWidth[k]; - else - mode_lib->vba.SwathWidthSingleDPPY[k] = mode_lib->vba.ViewportHeight[k]; - - if (mode_lib->vba.ODMCombineEnabled[k] == true) - MainPlaneDoesODMCombine = true; - for (j = 0; j < mode_lib->vba.NumberOfActivePlanes; ++j) - if (mode_lib->vba.BlendingAndTiming[k] == j - && mode_lib->vba.ODMCombineEnabled[j] == true) - MainPlaneDoesODMCombine = true; - - if (MainPlaneDoesODMCombine == true) - mode_lib->vba.SwathWidthY[k] = dml_min( - (double) mode_lib->vba.SwathWidthSingleDPPY[k], - dml_round( - mode_lib->vba.HActive[k] / 2.0 - * mode_lib->vba.HRatio[k])); - else - mode_lib->vba.SwathWidthY[k] = mode_lib->vba.SwathWidthSingleDPPY[k] - / mode_lib->vba.DPPPerPlane[k]; - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.SourcePixelFormat[k] == dm_444_64) { - mode_lib->vba.BytePerPixelDETY[k] = 8; - mode_lib->vba.BytePerPixelDETC[k] = 0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_32) { - mode_lib->vba.BytePerPixelDETY[k] = 4; - mode_lib->vba.BytePerPixelDETC[k] = 0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_16) { - mode_lib->vba.BytePerPixelDETY[k] = 2; - mode_lib->vba.BytePerPixelDETC[k] = 0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_8) { - mode_lib->vba.BytePerPixelDETY[k] = 1; - mode_lib->vba.BytePerPixelDETC[k] = 0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_420_8) { - mode_lib->vba.BytePerPixelDETY[k] = 1; - mode_lib->vba.BytePerPixelDETC[k] = 2; - } else { // dm_420_10 - mode_lib->vba.BytePerPixelDETY[k] = 4.0 / 3.0; - mode_lib->vba.BytePerPixelDETC[k] = 8.0 / 3.0; - } - } - - mode_lib->vba.TotalDataReadBandwidth = 0.0; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.ReadBandwidthPlaneLuma[k] = mode_lib->vba.SwathWidthSingleDPPY[k] - * dml_ceil(mode_lib->vba.BytePerPixelDETY[k], 1) - / (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]) - * mode_lib->vba.VRatio[k]; - mode_lib->vba.ReadBandwidthPlaneChroma[k] = mode_lib->vba.SwathWidthSingleDPPY[k] - / 2 * dml_ceil(mode_lib->vba.BytePerPixelDETC[k], 2) - / (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]) - * mode_lib->vba.VRatio[k] / 2; - DTRACE( - " read_bw[%i] = %fBps", - k, - mode_lib->vba.ReadBandwidthPlaneLuma[k] - + mode_lib->vba.ReadBandwidthPlaneChroma[k]); - mode_lib->vba.TotalDataReadBandwidth += mode_lib->vba.ReadBandwidthPlaneLuma[k] - + mode_lib->vba.ReadBandwidthPlaneChroma[k]; - } - - mode_lib->vba.TotalDCCActiveDPP = 0; - mode_lib->vba.TotalActiveDPP = 0; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.TotalActiveDPP = mode_lib->vba.TotalActiveDPP - + mode_lib->vba.DPPPerPlane[k]; - if (mode_lib->vba.DCCEnable[k]) - mode_lib->vba.TotalDCCActiveDPP = mode_lib->vba.TotalDCCActiveDPP - + mode_lib->vba.DPPPerPlane[k]; - } - - mode_lib->vba.UrgentRoundTripAndOutOfOrderLatency = - (mode_lib->vba.RoundTripPingLatencyCycles + 32) / mode_lib->vba.DCFCLK - + mode_lib->vba.UrgentOutOfOrderReturnPerChannel - * mode_lib->vba.NumberOfChannels - / mode_lib->vba.ReturnBW; - - mode_lib->vba.LastPixelOfLineExtraWatermark = 0; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - double DataFabricLineDeliveryTimeLuma, DataFabricLineDeliveryTimeChroma; - - if (mode_lib->vba.VRatio[k] <= 1.0) - mode_lib->vba.DisplayPipeLineDeliveryTimeLuma[k] = - (double) mode_lib->vba.SwathWidthY[k] - * mode_lib->vba.DPPPerPlane[k] - / mode_lib->vba.HRatio[k] - / mode_lib->vba.PixelClock[k]; - else - mode_lib->vba.DisplayPipeLineDeliveryTimeLuma[k] = - (double) mode_lib->vba.SwathWidthY[k] - / mode_lib->vba.PSCL_THROUGHPUT_LUMA[k] - / mode_lib->vba.DPPCLK[k]; - - DataFabricLineDeliveryTimeLuma = mode_lib->vba.SwathWidthSingleDPPY[k] - * mode_lib->vba.SwathHeightY[k] - * dml_ceil(mode_lib->vba.BytePerPixelDETY[k], 1) - / (mode_lib->vba.ReturnBW * mode_lib->vba.ReadBandwidthPlaneLuma[k] - / mode_lib->vba.TotalDataReadBandwidth); - mode_lib->vba.LastPixelOfLineExtraWatermark = dml_max( - mode_lib->vba.LastPixelOfLineExtraWatermark, - DataFabricLineDeliveryTimeLuma - - mode_lib->vba.DisplayPipeLineDeliveryTimeLuma[k]); - - if (mode_lib->vba.BytePerPixelDETC[k] == 0) - mode_lib->vba.DisplayPipeLineDeliveryTimeChroma[k] = 0.0; - else if (mode_lib->vba.VRatio[k] / 2.0 <= 1.0) - mode_lib->vba.DisplayPipeLineDeliveryTimeChroma[k] = - mode_lib->vba.SwathWidthY[k] / 2.0 - * mode_lib->vba.DPPPerPlane[k] - / (mode_lib->vba.HRatio[k] / 2.0) - / mode_lib->vba.PixelClock[k]; - else - mode_lib->vba.DisplayPipeLineDeliveryTimeChroma[k] = - mode_lib->vba.SwathWidthY[k] / 2.0 - / mode_lib->vba.PSCL_THROUGHPUT_CHROMA[k] - / mode_lib->vba.DPPCLK[k]; - - DataFabricLineDeliveryTimeChroma = mode_lib->vba.SwathWidthSingleDPPY[k] / 2.0 - * mode_lib->vba.SwathHeightC[k] - * dml_ceil(mode_lib->vba.BytePerPixelDETC[k], 2) - / (mode_lib->vba.ReturnBW - * mode_lib->vba.ReadBandwidthPlaneChroma[k] - / mode_lib->vba.TotalDataReadBandwidth); - mode_lib->vba.LastPixelOfLineExtraWatermark = - dml_max( - mode_lib->vba.LastPixelOfLineExtraWatermark, - DataFabricLineDeliveryTimeChroma - - mode_lib->vba.DisplayPipeLineDeliveryTimeChroma[k]); - } - - mode_lib->vba.UrgentExtraLatency = mode_lib->vba.UrgentRoundTripAndOutOfOrderLatency - + (mode_lib->vba.TotalActiveDPP * mode_lib->vba.PixelChunkSizeInKByte - + mode_lib->vba.TotalDCCActiveDPP - * mode_lib->vba.MetaChunkSize) * 1024.0 - / mode_lib->vba.ReturnBW; - - if (mode_lib->vba.VirtualMemoryEnable) - mode_lib->vba.UrgentExtraLatency += mode_lib->vba.TotalActiveDPP - * mode_lib->vba.PTEChunkSize * 1024.0 / mode_lib->vba.ReturnBW; - - mode_lib->vba.UrgentWatermark = mode_lib->vba.UrgentLatency - + mode_lib->vba.LastPixelOfLineExtraWatermark - + mode_lib->vba.UrgentExtraLatency; - - DTRACE(" urgent_extra_latency = %fus", mode_lib->vba.UrgentExtraLatency); - DTRACE(" wm_urgent = %fus", mode_lib->vba.UrgentWatermark); - - mode_lib->vba.MemoryTripWatermark = mode_lib->vba.UrgentLatency; - - mode_lib->vba.TotalActiveWriteback = 0; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.WritebackEnable[k]) - mode_lib->vba.TotalActiveWriteback = mode_lib->vba.TotalActiveWriteback + 1; - } - - if (mode_lib->vba.TotalActiveWriteback <= 1) - mode_lib->vba.WritebackUrgentWatermark = mode_lib->vba.WritebackLatency; - else - mode_lib->vba.WritebackUrgentWatermark = mode_lib->vba.WritebackLatency - + mode_lib->vba.WritebackChunkSize * 1024.0 / 32 - / mode_lib->vba.SOCCLK; - - DTRACE(" wm_wb_urgent = %fus", mode_lib->vba.WritebackUrgentWatermark); - - // NB P-State/DRAM Clock Change Watermark - mode_lib->vba.DRAMClockChangeWatermark = mode_lib->vba.DRAMClockChangeLatency - + mode_lib->vba.UrgentWatermark; - - DTRACE(" wm_pstate_change = %fus", mode_lib->vba.DRAMClockChangeWatermark); - - DTRACE(" calculating wb pstate watermark"); - DTRACE(" total wb outputs %d", mode_lib->vba.TotalActiveWriteback); - DTRACE(" socclk frequency %f Mhz", mode_lib->vba.SOCCLK); - - if (mode_lib->vba.TotalActiveWriteback <= 1) - mode_lib->vba.WritebackDRAMClockChangeWatermark = - mode_lib->vba.DRAMClockChangeLatency - + mode_lib->vba.WritebackLatency; - else - mode_lib->vba.WritebackDRAMClockChangeWatermark = - mode_lib->vba.DRAMClockChangeLatency - + mode_lib->vba.WritebackLatency - + mode_lib->vba.WritebackChunkSize * 1024.0 / 32 - / mode_lib->vba.SOCCLK; - - DTRACE(" wm_wb_pstate %fus", mode_lib->vba.WritebackDRAMClockChangeWatermark); - - // Stutter Efficiency - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.LinesInDETY[k] = mode_lib->vba.DETBufferSizeY[k] - / mode_lib->vba.BytePerPixelDETY[k] / mode_lib->vba.SwathWidthY[k]; - mode_lib->vba.LinesInDETYRoundedDownToSwath[k] = dml_floor( - mode_lib->vba.LinesInDETY[k], - mode_lib->vba.SwathHeightY[k]); - mode_lib->vba.FullDETBufferingTimeY[k] = - mode_lib->vba.LinesInDETYRoundedDownToSwath[k] - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) - / mode_lib->vba.VRatio[k]; - if (mode_lib->vba.BytePerPixelDETC[k] > 0) { - mode_lib->vba.LinesInDETC[k] = mode_lib->vba.DETBufferSizeC[k] - / mode_lib->vba.BytePerPixelDETC[k] - / (mode_lib->vba.SwathWidthY[k] / 2); - mode_lib->vba.LinesInDETCRoundedDownToSwath[k] = dml_floor( - mode_lib->vba.LinesInDETC[k], - mode_lib->vba.SwathHeightC[k]); - mode_lib->vba.FullDETBufferingTimeC[k] = - mode_lib->vba.LinesInDETCRoundedDownToSwath[k] - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) - / (mode_lib->vba.VRatio[k] / 2); - } else { - mode_lib->vba.LinesInDETC[k] = 0; - mode_lib->vba.LinesInDETCRoundedDownToSwath[k] = 0; - mode_lib->vba.FullDETBufferingTimeC[k] = 999999; - } - } - - mode_lib->vba.MinFullDETBufferingTime = 999999.0; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.FullDETBufferingTimeY[k] - < mode_lib->vba.MinFullDETBufferingTime) { - mode_lib->vba.MinFullDETBufferingTime = - mode_lib->vba.FullDETBufferingTimeY[k]; - mode_lib->vba.FrameTimeForMinFullDETBufferingTime = - (double) mode_lib->vba.VTotal[k] * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]; - } - if (mode_lib->vba.FullDETBufferingTimeC[k] - < mode_lib->vba.MinFullDETBufferingTime) { - mode_lib->vba.MinFullDETBufferingTime = - mode_lib->vba.FullDETBufferingTimeC[k]; - mode_lib->vba.FrameTimeForMinFullDETBufferingTime = - (double) mode_lib->vba.VTotal[k] * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]; - } - } - - mode_lib->vba.AverageReadBandwidthGBytePerSecond = 0.0; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.DCCEnable[k]) { - mode_lib->vba.AverageReadBandwidthGBytePerSecond = - mode_lib->vba.AverageReadBandwidthGBytePerSecond - + mode_lib->vba.ReadBandwidthPlaneLuma[k] - / mode_lib->vba.DCCRate[k] - / 1000 - + mode_lib->vba.ReadBandwidthPlaneChroma[k] - / mode_lib->vba.DCCRate[k] - / 1000; - } else { - mode_lib->vba.AverageReadBandwidthGBytePerSecond = - mode_lib->vba.AverageReadBandwidthGBytePerSecond - + mode_lib->vba.ReadBandwidthPlaneLuma[k] - / 1000 - + mode_lib->vba.ReadBandwidthPlaneChroma[k] - / 1000; - } - if (mode_lib->vba.DCCEnable[k]) { - mode_lib->vba.AverageReadBandwidthGBytePerSecond = - mode_lib->vba.AverageReadBandwidthGBytePerSecond - + mode_lib->vba.ReadBandwidthPlaneLuma[k] - / 1000 / 256 - + mode_lib->vba.ReadBandwidthPlaneChroma[k] - / 1000 / 256; - } - if (mode_lib->vba.VirtualMemoryEnable) { - mode_lib->vba.AverageReadBandwidthGBytePerSecond = - mode_lib->vba.AverageReadBandwidthGBytePerSecond - + mode_lib->vba.ReadBandwidthPlaneLuma[k] - / 1000 / 512 - + mode_lib->vba.ReadBandwidthPlaneChroma[k] - / 1000 / 512; - } - } - - mode_lib->vba.PartOfBurstThatFitsInROB = - dml_min( - mode_lib->vba.MinFullDETBufferingTime - * mode_lib->vba.TotalDataReadBandwidth, - mode_lib->vba.ROBBufferSizeInKByte * 1024 - * mode_lib->vba.TotalDataReadBandwidth - / (mode_lib->vba.AverageReadBandwidthGBytePerSecond - * 1000)); - mode_lib->vba.StutterBurstTime = mode_lib->vba.PartOfBurstThatFitsInROB - * (mode_lib->vba.AverageReadBandwidthGBytePerSecond * 1000) - / mode_lib->vba.TotalDataReadBandwidth / mode_lib->vba.ReturnBW - + (mode_lib->vba.MinFullDETBufferingTime - * mode_lib->vba.TotalDataReadBandwidth - - mode_lib->vba.PartOfBurstThatFitsInROB) - / (mode_lib->vba.DCFCLK * 64); - if (mode_lib->vba.TotalActiveWriteback == 0) { - mode_lib->vba.StutterEfficiencyNotIncludingVBlank = (1 - - (mode_lib->vba.SRExitTime + mode_lib->vba.StutterBurstTime) - / mode_lib->vba.MinFullDETBufferingTime) * 100; - } else { - mode_lib->vba.StutterEfficiencyNotIncludingVBlank = 0; - } - - mode_lib->vba.SmallestVBlank = 999999; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.SynchronizedVBlank || mode_lib->vba.NumberOfActivePlanes == 1) { - mode_lib->vba.VBlankTime = (double) (mode_lib->vba.VTotal[k] - - mode_lib->vba.VActive[k]) * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]; - } else { - mode_lib->vba.VBlankTime = 0; - } - mode_lib->vba.SmallestVBlank = dml_min( - mode_lib->vba.SmallestVBlank, - mode_lib->vba.VBlankTime); - } - - mode_lib->vba.StutterEfficiency = (mode_lib->vba.StutterEfficiencyNotIncludingVBlank / 100 - * (mode_lib->vba.FrameTimeForMinFullDETBufferingTime - - mode_lib->vba.SmallestVBlank) - + mode_lib->vba.SmallestVBlank) - / mode_lib->vba.FrameTimeForMinFullDETBufferingTime * 100; - - // dml_ml->vba.DCFCLK Deep Sleep - mode_lib->vba.DCFClkDeepSleep = 8.0; - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; k++) { - if (mode_lib->vba.BytePerPixelDETC[k] > 0) { - mode_lib->vba.DCFCLKDeepSleepPerPlane = - dml_max( - 1.1 * mode_lib->vba.SwathWidthY[k] - * dml_ceil( - mode_lib->vba.BytePerPixelDETY[k], - 1) / 32 - / mode_lib->vba.DisplayPipeLineDeliveryTimeLuma[k], - 1.1 * mode_lib->vba.SwathWidthY[k] / 2.0 - * dml_ceil( - mode_lib->vba.BytePerPixelDETC[k], - 2) / 32 - / mode_lib->vba.DisplayPipeLineDeliveryTimeChroma[k]); - } else - mode_lib->vba.DCFCLKDeepSleepPerPlane = 1.1 * mode_lib->vba.SwathWidthY[k] - * dml_ceil(mode_lib->vba.BytePerPixelDETY[k], 1) / 64.0 - / mode_lib->vba.DisplayPipeLineDeliveryTimeLuma[k]; - mode_lib->vba.DCFCLKDeepSleepPerPlane = dml_max( - mode_lib->vba.DCFCLKDeepSleepPerPlane, - mode_lib->vba.PixelClock[k] / 16.0); - mode_lib->vba.DCFClkDeepSleep = dml_max( - mode_lib->vba.DCFClkDeepSleep, - mode_lib->vba.DCFCLKDeepSleepPerPlane); - - DTRACE( - " dcfclk_deepsleep_per_plane[%i] = %fMHz", - k, - mode_lib->vba.DCFCLKDeepSleepPerPlane); - } - - DTRACE(" dcfclk_deepsleep_mhz = %fMHz", mode_lib->vba.DCFClkDeepSleep); - - // Stutter Watermark - mode_lib->vba.StutterExitWatermark = mode_lib->vba.SRExitTime - + mode_lib->vba.LastPixelOfLineExtraWatermark - + mode_lib->vba.UrgentExtraLatency + 10 / mode_lib->vba.DCFClkDeepSleep; - mode_lib->vba.StutterEnterPlusExitWatermark = mode_lib->vba.SREnterPlusExitTime - + mode_lib->vba.LastPixelOfLineExtraWatermark - + mode_lib->vba.UrgentExtraLatency; - - DTRACE(" wm_cstate_exit = %fus", mode_lib->vba.StutterExitWatermark); - DTRACE(" wm_cstate_enter_exit = %fus", mode_lib->vba.StutterEnterPlusExitWatermark); - - // Urgent Latency Supported - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.EffectiveDETPlusLBLinesLuma = - dml_floor( - mode_lib->vba.LinesInDETY[k] - + dml_min( - mode_lib->vba.LinesInDETY[k] - * mode_lib->vba.DPPCLK[k] - * mode_lib->vba.BytePerPixelDETY[k] - * mode_lib->vba.PSCL_THROUGHPUT_LUMA[k] - / (mode_lib->vba.ReturnBW - / mode_lib->vba.DPPPerPlane[k]), - (double) mode_lib->vba.EffectiveLBLatencyHidingSourceLinesLuma), - mode_lib->vba.SwathHeightY[k]); - - mode_lib->vba.UrgentLatencySupportUsLuma = mode_lib->vba.EffectiveDETPlusLBLinesLuma - * (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]) - / mode_lib->vba.VRatio[k] - - mode_lib->vba.EffectiveDETPlusLBLinesLuma - * mode_lib->vba.SwathWidthY[k] - * mode_lib->vba.BytePerPixelDETY[k] - / (mode_lib->vba.ReturnBW - / mode_lib->vba.DPPPerPlane[k]); - - if (mode_lib->vba.BytePerPixelDETC[k] > 0) { - mode_lib->vba.EffectiveDETPlusLBLinesChroma = - dml_floor( - mode_lib->vba.LinesInDETC[k] - + dml_min( - mode_lib->vba.LinesInDETC[k] - * mode_lib->vba.DPPCLK[k] - * mode_lib->vba.BytePerPixelDETC[k] - * mode_lib->vba.PSCL_THROUGHPUT_CHROMA[k] - / (mode_lib->vba.ReturnBW - / mode_lib->vba.DPPPerPlane[k]), - (double) mode_lib->vba.EffectiveLBLatencyHidingSourceLinesChroma), - mode_lib->vba.SwathHeightC[k]); - mode_lib->vba.UrgentLatencySupportUsChroma = - mode_lib->vba.EffectiveDETPlusLBLinesChroma - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) - / (mode_lib->vba.VRatio[k] / 2) - - mode_lib->vba.EffectiveDETPlusLBLinesChroma - * (mode_lib->vba.SwathWidthY[k] - / 2) - * mode_lib->vba.BytePerPixelDETC[k] - / (mode_lib->vba.ReturnBW - / mode_lib->vba.DPPPerPlane[k]); - mode_lib->vba.UrgentLatencySupportUs[k] = dml_min( - mode_lib->vba.UrgentLatencySupportUsLuma, - mode_lib->vba.UrgentLatencySupportUsChroma); - } else { - mode_lib->vba.UrgentLatencySupportUs[k] = - mode_lib->vba.UrgentLatencySupportUsLuma; - } - } - - mode_lib->vba.MinUrgentLatencySupportUs = 999999; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.MinUrgentLatencySupportUs = dml_min( - mode_lib->vba.MinUrgentLatencySupportUs, - mode_lib->vba.UrgentLatencySupportUs[k]); - } - - // Non-Urgent Latency Tolerance - mode_lib->vba.NonUrgentLatencyTolerance = mode_lib->vba.MinUrgentLatencySupportUs - - mode_lib->vba.UrgentWatermark; - - // DSCCLK - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if ((mode_lib->vba.BlendingAndTiming[k] != k) || !mode_lib->vba.DSCEnabled[k]) { - mode_lib->vba.DSCCLK_calculated[k] = 0.0; - } else { - if (mode_lib->vba.OutputFormat[k] == dm_420 - || mode_lib->vba.OutputFormat[k] == dm_n422) - mode_lib->vba.DSCFormatFactor = 2; - else - mode_lib->vba.DSCFormatFactor = 1; - if (mode_lib->vba.ODMCombineEnabled[k]) - mode_lib->vba.DSCCLK_calculated[k] = - mode_lib->vba.PixelClockBackEnd[k] / 6 - / mode_lib->vba.DSCFormatFactor - / (1 - - mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100); - else - mode_lib->vba.DSCCLK_calculated[k] = - mode_lib->vba.PixelClockBackEnd[k] / 3 - / mode_lib->vba.DSCFormatFactor - / (1 - - mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100); - } - } - - // DSC Delay - // TODO - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - double bpp = mode_lib->vba.OutputBpp[k]; - unsigned int slices = mode_lib->vba.NumberOfDSCSlices[k]; - - if (mode_lib->vba.DSCEnabled[k] && bpp != 0) { - if (!mode_lib->vba.ODMCombineEnabled[k]) { - mode_lib->vba.DSCDelay[k] = - dscceComputeDelay( - mode_lib->vba.DSCInputBitPerComponent[k], - bpp, - dml_ceil( - (double) mode_lib->vba.HActive[k] - / mode_lib->vba.NumberOfDSCSlices[k], - 1), - slices, - mode_lib->vba.OutputFormat[k]) - + dscComputeDelay( - mode_lib->vba.OutputFormat[k]); - } else { - mode_lib->vba.DSCDelay[k] = - 2 - * (dscceComputeDelay( - mode_lib->vba.DSCInputBitPerComponent[k], - bpp, - dml_ceil( - (double) mode_lib->vba.HActive[k] - / mode_lib->vba.NumberOfDSCSlices[k], - 1), - slices / 2.0, - mode_lib->vba.OutputFormat[k]) - + dscComputeDelay( - mode_lib->vba.OutputFormat[k])); - } - mode_lib->vba.DSCDelay[k] = mode_lib->vba.DSCDelay[k] - * mode_lib->vba.PixelClock[k] - / mode_lib->vba.PixelClockBackEnd[k]; - } else { - mode_lib->vba.DSCDelay[k] = 0; - } - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) - for (j = 0; j < mode_lib->vba.NumberOfActivePlanes; ++j) // NumberOfPlanes - if (j != k && mode_lib->vba.BlendingAndTiming[k] == j - && mode_lib->vba.DSCEnabled[j]) - mode_lib->vba.DSCDelay[k] = mode_lib->vba.DSCDelay[j]; - - // Prefetch - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - unsigned int PDEAndMetaPTEBytesFrameY; - unsigned int PixelPTEBytesPerRowY; - unsigned int MetaRowByteY; - unsigned int MetaRowByteC; - unsigned int PDEAndMetaPTEBytesFrameC; - unsigned int PixelPTEBytesPerRowC; - - Calculate256BBlockSizes( - mode_lib->vba.SourcePixelFormat[k], - mode_lib->vba.SurfaceTiling[k], - dml_ceil(mode_lib->vba.BytePerPixelDETY[k], 1), - dml_ceil(mode_lib->vba.BytePerPixelDETC[k], 2), - &mode_lib->vba.BlockHeight256BytesY[k], - &mode_lib->vba.BlockHeight256BytesC[k], - &mode_lib->vba.BlockWidth256BytesY[k], - &mode_lib->vba.BlockWidth256BytesC[k]); - PDEAndMetaPTEBytesFrameY = CalculateVMAndRowBytes( - mode_lib, - mode_lib->vba.DCCEnable[k], - mode_lib->vba.BlockHeight256BytesY[k], - mode_lib->vba.BlockWidth256BytesY[k], - mode_lib->vba.SourcePixelFormat[k], - mode_lib->vba.SurfaceTiling[k], - dml_ceil(mode_lib->vba.BytePerPixelDETY[k], 1), - mode_lib->vba.SourceScan[k], - mode_lib->vba.ViewportWidth[k], - mode_lib->vba.ViewportHeight[k], - mode_lib->vba.SwathWidthY[k], - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.VMMPageSize, - mode_lib->vba.PTEBufferSizeInRequests, - mode_lib->vba.PDEProcessingBufIn64KBReqs, - mode_lib->vba.PitchY[k], - mode_lib->vba.DCCMetaPitchY[k], - &mode_lib->vba.MacroTileWidthY[k], - &MetaRowByteY, - &PixelPTEBytesPerRowY, - &mode_lib->vba.PTEBufferSizeNotExceeded[mode_lib->vba.VoltageLevel], - &mode_lib->vba.dpte_row_height[k], - &mode_lib->vba.meta_row_height[k]); - mode_lib->vba.PrefetchSourceLinesY[k] = CalculatePrefetchSourceLines( - mode_lib, - mode_lib->vba.VRatio[k], - mode_lib->vba.vtaps[k], - mode_lib->vba.Interlace[k], - mode_lib->vba.ProgressiveToInterlaceUnitInOPP, - mode_lib->vba.SwathHeightY[k], - mode_lib->vba.ViewportYStartY[k], - &mode_lib->vba.VInitPreFillY[k], - &mode_lib->vba.MaxNumSwathY[k]); - - if ((mode_lib->vba.SourcePixelFormat[k] != dm_444_64 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_32 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_8)) { - PDEAndMetaPTEBytesFrameC = - CalculateVMAndRowBytes( - mode_lib, - mode_lib->vba.DCCEnable[k], - mode_lib->vba.BlockHeight256BytesC[k], - mode_lib->vba.BlockWidth256BytesC[k], - mode_lib->vba.SourcePixelFormat[k], - mode_lib->vba.SurfaceTiling[k], - dml_ceil( - mode_lib->vba.BytePerPixelDETC[k], - 2), - mode_lib->vba.SourceScan[k], - mode_lib->vba.ViewportWidth[k] / 2, - mode_lib->vba.ViewportHeight[k] / 2, - mode_lib->vba.SwathWidthY[k] / 2, - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.VMMPageSize, - mode_lib->vba.PTEBufferSizeInRequests, - mode_lib->vba.PDEProcessingBufIn64KBReqs, - mode_lib->vba.PitchC[k], - 0, - &mode_lib->vba.MacroTileWidthC[k], - &MetaRowByteC, - &PixelPTEBytesPerRowC, - &mode_lib->vba.PTEBufferSizeNotExceeded[mode_lib->vba.VoltageLevel], - &mode_lib->vba.dpte_row_height_chroma[k], - &mode_lib->vba.meta_row_height_chroma[k]); - mode_lib->vba.PrefetchSourceLinesC[k] = CalculatePrefetchSourceLines( - mode_lib, - mode_lib->vba.VRatio[k] / 2, - mode_lib->vba.VTAPsChroma[k], - mode_lib->vba.Interlace[k], - mode_lib->vba.ProgressiveToInterlaceUnitInOPP, - mode_lib->vba.SwathHeightC[k], - mode_lib->vba.ViewportYStartC[k], - &mode_lib->vba.VInitPreFillC[k], - &mode_lib->vba.MaxNumSwathC[k]); - } else { - PixelPTEBytesPerRowC = 0; - PDEAndMetaPTEBytesFrameC = 0; - MetaRowByteC = 0; - mode_lib->vba.MaxNumSwathC[k] = 0; - mode_lib->vba.PrefetchSourceLinesC[k] = 0; - } - - mode_lib->vba.PixelPTEBytesPerRow[k] = PixelPTEBytesPerRowY + PixelPTEBytesPerRowC; - mode_lib->vba.PDEAndMetaPTEBytesFrame[k] = PDEAndMetaPTEBytesFrameY - + PDEAndMetaPTEBytesFrameC; - mode_lib->vba.MetaRowByte[k] = MetaRowByteY + MetaRowByteC; - - CalculateActiveRowBandwidth( - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.SourcePixelFormat[k], - mode_lib->vba.VRatio[k], - mode_lib->vba.DCCEnable[k], - mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k], - MetaRowByteY, - MetaRowByteC, - mode_lib->vba.meta_row_height[k], - mode_lib->vba.meta_row_height_chroma[k], - PixelPTEBytesPerRowY, - PixelPTEBytesPerRowC, - mode_lib->vba.dpte_row_height[k], - mode_lib->vba.dpte_row_height_chroma[k], - &mode_lib->vba.meta_row_bw[k], - &mode_lib->vba.dpte_row_bw[k], - &mode_lib->vba.qual_row_bw[k]); - } - - mode_lib->vba.TCalc = 24.0 / mode_lib->vba.DCFClkDeepSleep; - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.BlendingAndTiming[k] == k) { - if (mode_lib->vba.WritebackEnable[k] == true) { - mode_lib->vba.WritebackDelay[mode_lib->vba.VoltageLevel][k] = - mode_lib->vba.WritebackLatency - + CalculateWriteBackDelay( - mode_lib->vba.WritebackPixelFormat[k], - mode_lib->vba.WritebackHRatio[k], - mode_lib->vba.WritebackVRatio[k], - mode_lib->vba.WritebackLumaHTaps[k], - mode_lib->vba.WritebackLumaVTaps[k], - mode_lib->vba.WritebackChromaHTaps[k], - mode_lib->vba.WritebackChromaVTaps[k], - mode_lib->vba.WritebackDestinationWidth[k]) - / mode_lib->vba.DISPCLK; - } else - mode_lib->vba.WritebackDelay[mode_lib->vba.VoltageLevel][k] = 0; - for (j = 0; j < mode_lib->vba.NumberOfActivePlanes; ++j) { - if (mode_lib->vba.BlendingAndTiming[j] == k - && mode_lib->vba.WritebackEnable[j] == true) { - mode_lib->vba.WritebackDelay[mode_lib->vba.VoltageLevel][k] = - dml_max( - mode_lib->vba.WritebackDelay[mode_lib->vba.VoltageLevel][k], - mode_lib->vba.WritebackLatency - + CalculateWriteBackDelay( - mode_lib->vba.WritebackPixelFormat[j], - mode_lib->vba.WritebackHRatio[j], - mode_lib->vba.WritebackVRatio[j], - mode_lib->vba.WritebackLumaHTaps[j], - mode_lib->vba.WritebackLumaVTaps[j], - mode_lib->vba.WritebackChromaHTaps[j], - mode_lib->vba.WritebackChromaVTaps[j], - mode_lib->vba.WritebackDestinationWidth[j]) - / mode_lib->vba.DISPCLK); - } - } - } - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) - for (j = 0; j < mode_lib->vba.NumberOfActivePlanes; ++j) - if (mode_lib->vba.BlendingAndTiming[k] == j) - mode_lib->vba.WritebackDelay[mode_lib->vba.VoltageLevel][k] = - mode_lib->vba.WritebackDelay[mode_lib->vba.VoltageLevel][j]; - - mode_lib->vba.VStartupLines = 13; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.MaxVStartupLines[k] = - mode_lib->vba.VTotal[k] - mode_lib->vba.VActive[k] - - dml_max( - 1.0, - dml_ceil( - mode_lib->vba.WritebackDelay[mode_lib->vba.VoltageLevel][k] - / (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]), - 1)); - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) - mode_lib->vba.MaximumMaxVStartupLines = dml_max( - mode_lib->vba.MaximumMaxVStartupLines, - mode_lib->vba.MaxVStartupLines[k]); - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.cursor_bw[k] = 0.0; - for (j = 0; j < mode_lib->vba.NumberOfCursors[k]; ++j) - mode_lib->vba.cursor_bw[k] += mode_lib->vba.CursorWidth[k][j] - * mode_lib->vba.CursorBPP[k][j] / 8.0 - / (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]) - * mode_lib->vba.VRatio[k]; - } - - do { - double MaxTotalRDBandwidth = 0; - bool DestinationLineTimesForPrefetchLessThan2 = false; - bool VRatioPrefetchMoreThan4 = false; - bool prefetch_vm_bw_valid = true; - bool prefetch_row_bw_valid = true; - double TWait = CalculateTWait( - mode_lib->vba.PrefetchMode, - mode_lib->vba.DRAMClockChangeLatency, - mode_lib->vba.UrgentLatency, - mode_lib->vba.SREnterPlusExitTime); - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.XFCEnabled[k] == true) { - mode_lib->vba.XFCRemoteSurfaceFlipDelay = - CalculateRemoteSurfaceFlipDelay( - mode_lib, - mode_lib->vba.VRatio[k], - mode_lib->vba.SwathWidthY[k], - dml_ceil( - mode_lib->vba.BytePerPixelDETY[k], - 1), - mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k], - mode_lib->vba.XFCTSlvVupdateOffset, - mode_lib->vba.XFCTSlvVupdateWidth, - mode_lib->vba.XFCTSlvVreadyOffset, - mode_lib->vba.XFCXBUFLatencyTolerance, - mode_lib->vba.XFCFillBWOverhead, - mode_lib->vba.XFCSlvChunkSize, - mode_lib->vba.XFCBusTransportTime, - mode_lib->vba.TCalc, - TWait, - &mode_lib->vba.SrcActiveDrainRate, - &mode_lib->vba.TInitXFill, - &mode_lib->vba.TslvChk); - } else { - mode_lib->vba.XFCRemoteSurfaceFlipDelay = 0; - } - mode_lib->vba.ErrorResult[k] = - CalculatePrefetchSchedule( - mode_lib, - mode_lib->vba.DPPCLK[k], - mode_lib->vba.DISPCLK, - mode_lib->vba.PixelClock[k], - mode_lib->vba.DCFClkDeepSleep, - mode_lib->vba.DSCDelay[k], - mode_lib->vba.DPPPerPlane[k], - mode_lib->vba.ScalerEnabled[k], - mode_lib->vba.NumberOfCursors[k], - mode_lib->vba.DPPCLKDelaySubtotal, - mode_lib->vba.DPPCLKDelaySCL, - mode_lib->vba.DPPCLKDelaySCLLBOnly, - mode_lib->vba.DPPCLKDelayCNVCFormater, - mode_lib->vba.DPPCLKDelayCNVCCursor, - mode_lib->vba.DISPCLKDelaySubtotal, - (unsigned int) (mode_lib->vba.SwathWidthY[k] - / mode_lib->vba.HRatio[k]), - mode_lib->vba.OutputFormat[k], - mode_lib->vba.VTotal[k] - - mode_lib->vba.VActive[k], - mode_lib->vba.HTotal[k], - mode_lib->vba.MaxInterDCNTileRepeaters, - dml_min( - mode_lib->vba.VStartupLines, - mode_lib->vba.MaxVStartupLines[k]), - mode_lib->vba.MaxPageTableLevels, - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.DynamicMetadataEnable[k], - mode_lib->vba.DynamicMetadataLinesBeforeActiveRequired[k], - mode_lib->vba.DynamicMetadataTransmittedBytes[k], - mode_lib->vba.DCCEnable[k], - mode_lib->vba.UrgentLatency, - mode_lib->vba.UrgentExtraLatency, - mode_lib->vba.TCalc, - mode_lib->vba.PDEAndMetaPTEBytesFrame[k], - mode_lib->vba.MetaRowByte[k], - mode_lib->vba.PixelPTEBytesPerRow[k], - mode_lib->vba.PrefetchSourceLinesY[k], - mode_lib->vba.SwathWidthY[k], - mode_lib->vba.BytePerPixelDETY[k], - mode_lib->vba.VInitPreFillY[k], - mode_lib->vba.MaxNumSwathY[k], - mode_lib->vba.PrefetchSourceLinesC[k], - mode_lib->vba.BytePerPixelDETC[k], - mode_lib->vba.VInitPreFillC[k], - mode_lib->vba.MaxNumSwathC[k], - mode_lib->vba.SwathHeightY[k], - mode_lib->vba.SwathHeightC[k], - TWait, - mode_lib->vba.XFCEnabled[k], - mode_lib->vba.XFCRemoteSurfaceFlipDelay, - mode_lib->vba.Interlace[k], - mode_lib->vba.ProgressiveToInterlaceUnitInOPP, - &mode_lib->vba.DSTXAfterScaler[k], - &mode_lib->vba.DSTYAfterScaler[k], - &mode_lib->vba.DestinationLinesForPrefetch[k], - &mode_lib->vba.PrefetchBandwidth[k], - &mode_lib->vba.DestinationLinesToRequestVMInVBlank[k], - &mode_lib->vba.DestinationLinesToRequestRowInVBlank[k], - &mode_lib->vba.VRatioPrefetchY[k], - &mode_lib->vba.VRatioPrefetchC[k], - &mode_lib->vba.RequiredPrefetchPixDataBW[k], - &mode_lib->vba.VStartupRequiredWhenNotEnoughTimeForDynamicMetadata, - &mode_lib->vba.Tno_bw[k], - &mode_lib->vba.VUpdateOffsetPix[k], - &mode_lib->vba.VUpdateWidthPix[k], - &mode_lib->vba.VReadyOffsetPix[k]); - if (mode_lib->vba.BlendingAndTiming[k] == k) { - mode_lib->vba.VStartup[k] = dml_min( - mode_lib->vba.VStartupLines, - mode_lib->vba.MaxVStartupLines[k]); - if (mode_lib->vba.VStartupRequiredWhenNotEnoughTimeForDynamicMetadata - != 0) { - mode_lib->vba.VStartup[k] = - mode_lib->vba.VStartupRequiredWhenNotEnoughTimeForDynamicMetadata; - } - } else { - mode_lib->vba.VStartup[k] = - dml_min( - mode_lib->vba.VStartupLines, - mode_lib->vba.MaxVStartupLines[mode_lib->vba.BlendingAndTiming[k]]); - } - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - - if (mode_lib->vba.PDEAndMetaPTEBytesFrame[k] == 0) - mode_lib->vba.prefetch_vm_bw[k] = 0; - else if (mode_lib->vba.DestinationLinesToRequestVMInVBlank[k] > 0) { - mode_lib->vba.prefetch_vm_bw[k] = - (double) mode_lib->vba.PDEAndMetaPTEBytesFrame[k] - / (mode_lib->vba.DestinationLinesToRequestVMInVBlank[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]); - } else { - mode_lib->vba.prefetch_vm_bw[k] = 0; - prefetch_vm_bw_valid = false; - } - if (mode_lib->vba.MetaRowByte[k] + mode_lib->vba.PixelPTEBytesPerRow[k] - == 0) - mode_lib->vba.prefetch_row_bw[k] = 0; - else if (mode_lib->vba.DestinationLinesToRequestRowInVBlank[k] > 0) { - mode_lib->vba.prefetch_row_bw[k] = - (double) (mode_lib->vba.MetaRowByte[k] - + mode_lib->vba.PixelPTEBytesPerRow[k]) - / (mode_lib->vba.DestinationLinesToRequestRowInVBlank[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]); - } else { - mode_lib->vba.prefetch_row_bw[k] = 0; - prefetch_row_bw_valid = false; - } - - MaxTotalRDBandwidth = - MaxTotalRDBandwidth + mode_lib->vba.cursor_bw[k] - + dml_max( - mode_lib->vba.prefetch_vm_bw[k], - dml_max( - mode_lib->vba.prefetch_row_bw[k], - dml_max( - mode_lib->vba.ReadBandwidthPlaneLuma[k] - + mode_lib->vba.ReadBandwidthPlaneChroma[k], - mode_lib->vba.RequiredPrefetchPixDataBW[k]) - + mode_lib->vba.meta_row_bw[k] - + mode_lib->vba.dpte_row_bw[k])); - - if (mode_lib->vba.DestinationLinesForPrefetch[k] < 2) - DestinationLineTimesForPrefetchLessThan2 = true; - if (mode_lib->vba.VRatioPrefetchY[k] > 4 - || mode_lib->vba.VRatioPrefetchC[k] > 4) - VRatioPrefetchMoreThan4 = true; - } - - if (MaxTotalRDBandwidth <= mode_lib->vba.ReturnBW && prefetch_vm_bw_valid - && prefetch_row_bw_valid && !VRatioPrefetchMoreThan4 - && !DestinationLineTimesForPrefetchLessThan2) - mode_lib->vba.PrefetchModeSupported = true; - else { - mode_lib->vba.PrefetchModeSupported = false; - dml_print( - "DML: CalculatePrefetchSchedule ***failed***. Bandwidth violation. Results are NOT valid\n"); - } - - if (mode_lib->vba.PrefetchModeSupported == true) { - double final_flip_bw[DC__NUM_DPP__MAX]; - unsigned int ImmediateFlipBytes[DC__NUM_DPP__MAX]; - double total_dcn_read_bw_with_flip = 0; - - mode_lib->vba.BandwidthAvailableForImmediateFlip = mode_lib->vba.ReturnBW; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.BandwidthAvailableForImmediateFlip = - mode_lib->vba.BandwidthAvailableForImmediateFlip - - mode_lib->vba.cursor_bw[k] - - dml_max( - mode_lib->vba.ReadBandwidthPlaneLuma[k] - + mode_lib->vba.ReadBandwidthPlaneChroma[k] - + mode_lib->vba.qual_row_bw[k], - mode_lib->vba.PrefetchBandwidth[k]); - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - ImmediateFlipBytes[k] = 0; - if ((mode_lib->vba.SourcePixelFormat[k] != dm_420_8 - && mode_lib->vba.SourcePixelFormat[k] != dm_420_10)) { - ImmediateFlipBytes[k] = - mode_lib->vba.PDEAndMetaPTEBytesFrame[k] - + mode_lib->vba.MetaRowByte[k] - + mode_lib->vba.PixelPTEBytesPerRow[k]; - } - } - mode_lib->vba.TotImmediateFlipBytes = 0; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if ((mode_lib->vba.SourcePixelFormat[k] != dm_420_8 - && mode_lib->vba.SourcePixelFormat[k] != dm_420_10)) { - mode_lib->vba.TotImmediateFlipBytes = - mode_lib->vba.TotImmediateFlipBytes - + ImmediateFlipBytes[k]; - } - } - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - CalculateFlipSchedule( - mode_lib, - mode_lib->vba.UrgentExtraLatency, - mode_lib->vba.UrgentLatency, - mode_lib->vba.MaxPageTableLevels, - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.BandwidthAvailableForImmediateFlip, - mode_lib->vba.TotImmediateFlipBytes, - mode_lib->vba.SourcePixelFormat[k], - ImmediateFlipBytes[k], - mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k], - mode_lib->vba.VRatio[k], - mode_lib->vba.Tno_bw[k], - mode_lib->vba.PDEAndMetaPTEBytesFrame[k], - mode_lib->vba.MetaRowByte[k], - mode_lib->vba.PixelPTEBytesPerRow[k], - mode_lib->vba.DCCEnable[k], - mode_lib->vba.dpte_row_height[k], - mode_lib->vba.meta_row_height[k], - mode_lib->vba.qual_row_bw[k], - &mode_lib->vba.DestinationLinesToRequestVMInImmediateFlip[k], - &mode_lib->vba.DestinationLinesToRequestRowInImmediateFlip[k], - &final_flip_bw[k], - &mode_lib->vba.ImmediateFlipSupportedForPipe[k]); - } - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - total_dcn_read_bw_with_flip = - total_dcn_read_bw_with_flip - + mode_lib->vba.cursor_bw[k] - + dml_max( - mode_lib->vba.prefetch_vm_bw[k], - dml_max( - mode_lib->vba.prefetch_row_bw[k], - final_flip_bw[k] - + dml_max( - mode_lib->vba.ReadBandwidthPlaneLuma[k] - + mode_lib->vba.ReadBandwidthPlaneChroma[k], - mode_lib->vba.RequiredPrefetchPixDataBW[k]))); - } - mode_lib->vba.ImmediateFlipSupported = true; - if (total_dcn_read_bw_with_flip > mode_lib->vba.ReturnBW) { - mode_lib->vba.ImmediateFlipSupported = false; - } - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.ImmediateFlipSupportedForPipe[k] == false) { - mode_lib->vba.ImmediateFlipSupported = false; - } - } - } else { - mode_lib->vba.ImmediateFlipSupported = false; - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.ErrorResult[k]) { - mode_lib->vba.PrefetchModeSupported = false; - dml_print( - "DML: CalculatePrefetchSchedule ***failed***. Prefetch schedule violation. Results are NOT valid\n"); - } - } - - mode_lib->vba.VStartupLines = mode_lib->vba.VStartupLines + 1; - } while (!((mode_lib->vba.PrefetchModeSupported - && (!mode_lib->vba.ImmediateFlipSupport - || mode_lib->vba.ImmediateFlipSupported)) - || mode_lib->vba.MaximumMaxVStartupLines < mode_lib->vba.VStartupLines)); - - //Display Pipeline Delivery Time in Prefetch - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.VRatioPrefetchY[k] <= 1) { - mode_lib->vba.DisplayPipeLineDeliveryTimeLumaPrefetch[k] = - mode_lib->vba.SwathWidthY[k] * mode_lib->vba.DPPPerPlane[k] - / mode_lib->vba.HRatio[k] - / mode_lib->vba.PixelClock[k]; - } else { - mode_lib->vba.DisplayPipeLineDeliveryTimeLumaPrefetch[k] = - mode_lib->vba.SwathWidthY[k] - / mode_lib->vba.PSCL_THROUGHPUT_LUMA[k] - / mode_lib->vba.DPPCLK[k]; - } - if (mode_lib->vba.BytePerPixelDETC[k] == 0) { - mode_lib->vba.DisplayPipeLineDeliveryTimeChromaPrefetch[k] = 0; - } else { - if (mode_lib->vba.VRatioPrefetchC[k] <= 1) { - mode_lib->vba.DisplayPipeLineDeliveryTimeChromaPrefetch[k] = - mode_lib->vba.SwathWidthY[k] - * mode_lib->vba.DPPPerPlane[k] - / mode_lib->vba.HRatio[k] - / mode_lib->vba.PixelClock[k]; - } else { - mode_lib->vba.DisplayPipeLineDeliveryTimeChromaPrefetch[k] = - mode_lib->vba.SwathWidthY[k] - / mode_lib->vba.PSCL_THROUGHPUT_LUMA[k] - / mode_lib->vba.DPPCLK[k]; - } - } - } - - // Min TTUVBlank - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.PrefetchMode == 0) { - mode_lib->vba.AllowDRAMClockChangeDuringVBlank[k] = true; - mode_lib->vba.AllowDRAMSelfRefreshDuringVBlank[k] = true; - mode_lib->vba.MinTTUVBlank[k] = dml_max( - mode_lib->vba.DRAMClockChangeWatermark, - dml_max( - mode_lib->vba.StutterEnterPlusExitWatermark, - mode_lib->vba.UrgentWatermark)); - } else if (mode_lib->vba.PrefetchMode == 1) { - mode_lib->vba.AllowDRAMClockChangeDuringVBlank[k] = false; - mode_lib->vba.AllowDRAMSelfRefreshDuringVBlank[k] = true; - mode_lib->vba.MinTTUVBlank[k] = dml_max( - mode_lib->vba.StutterEnterPlusExitWatermark, - mode_lib->vba.UrgentWatermark); - } else { - mode_lib->vba.AllowDRAMClockChangeDuringVBlank[k] = false; - mode_lib->vba.AllowDRAMSelfRefreshDuringVBlank[k] = false; - mode_lib->vba.MinTTUVBlank[k] = mode_lib->vba.UrgentWatermark; - } - if (!mode_lib->vba.DynamicMetadataEnable[k]) - mode_lib->vba.MinTTUVBlank[k] = mode_lib->vba.TCalc - + mode_lib->vba.MinTTUVBlank[k]; - } - - // DCC Configuration - mode_lib->vba.ActiveDPPs = 0; - // NB P-State/DRAM Clock Change Support - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.ActiveDPPs = mode_lib->vba.ActiveDPPs + mode_lib->vba.DPPPerPlane[k]; - } - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - double EffectiveLBLatencyHidingY; - double EffectiveLBLatencyHidingC; - double DPPOutputBufferLinesY; - double DPPOutputBufferLinesC; - double DPPOPPBufferingY; - double MaxDETBufferingTimeY; - double ActiveDRAMClockChangeLatencyMarginY; - - mode_lib->vba.LBLatencyHidingSourceLinesY = - dml_min( - mode_lib->vba.MaxLineBufferLines, - (unsigned int) dml_floor( - (double) mode_lib->vba.LineBufferSize - / mode_lib->vba.LBBitPerPixel[k] - / (mode_lib->vba.SwathWidthY[k] - / dml_max( - mode_lib->vba.HRatio[k], - 1.0)), - 1)) - (mode_lib->vba.vtaps[k] - 1); - - mode_lib->vba.LBLatencyHidingSourceLinesC = - dml_min( - mode_lib->vba.MaxLineBufferLines, - (unsigned int) dml_floor( - (double) mode_lib->vba.LineBufferSize - / mode_lib->vba.LBBitPerPixel[k] - / (mode_lib->vba.SwathWidthY[k] - / 2.0 - / dml_max( - mode_lib->vba.HRatio[k] - / 2, - 1.0)), - 1)) - - (mode_lib->vba.VTAPsChroma[k] - 1); - - EffectiveLBLatencyHidingY = mode_lib->vba.LBLatencyHidingSourceLinesY - / mode_lib->vba.VRatio[k] - * (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]); - - EffectiveLBLatencyHidingC = mode_lib->vba.LBLatencyHidingSourceLinesC - / (mode_lib->vba.VRatio[k] / 2) - * (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]); - - if (mode_lib->vba.SwathWidthY[k] > 2 * mode_lib->vba.DPPOutputBufferPixels) { - DPPOutputBufferLinesY = mode_lib->vba.DPPOutputBufferPixels - / mode_lib->vba.SwathWidthY[k]; - } else if (mode_lib->vba.SwathWidthY[k] > mode_lib->vba.DPPOutputBufferPixels) { - DPPOutputBufferLinesY = 0.5; - } else { - DPPOutputBufferLinesY = 1; - } - - if (mode_lib->vba.SwathWidthY[k] / 2 > 2 * mode_lib->vba.DPPOutputBufferPixels) { - DPPOutputBufferLinesC = mode_lib->vba.DPPOutputBufferPixels - / (mode_lib->vba.SwathWidthY[k] / 2); - } else if (mode_lib->vba.SwathWidthY[k] / 2 > mode_lib->vba.DPPOutputBufferPixels) { - DPPOutputBufferLinesC = 0.5; - } else { - DPPOutputBufferLinesC = 1; - } - - DPPOPPBufferingY = (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]) - * (DPPOutputBufferLinesY + mode_lib->vba.OPPOutputBufferLines); - MaxDETBufferingTimeY = mode_lib->vba.FullDETBufferingTimeY[k] - + (mode_lib->vba.LinesInDETY[k] - - mode_lib->vba.LinesInDETYRoundedDownToSwath[k]) - / mode_lib->vba.SwathHeightY[k] - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]); - - ActiveDRAMClockChangeLatencyMarginY = DPPOPPBufferingY + EffectiveLBLatencyHidingY - + MaxDETBufferingTimeY - mode_lib->vba.DRAMClockChangeWatermark; - - if (mode_lib->vba.ActiveDPPs > 1) { - ActiveDRAMClockChangeLatencyMarginY = - ActiveDRAMClockChangeLatencyMarginY - - (1 - 1 / (mode_lib->vba.ActiveDPPs - 1)) - * mode_lib->vba.SwathHeightY[k] - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]); - } - - if (mode_lib->vba.BytePerPixelDETC[k] > 0) { - double DPPOPPBufferingC = (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) - * (DPPOutputBufferLinesC - + mode_lib->vba.OPPOutputBufferLines); - double MaxDETBufferingTimeC = - mode_lib->vba.FullDETBufferingTimeC[k] - + (mode_lib->vba.LinesInDETC[k] - - mode_lib->vba.LinesInDETCRoundedDownToSwath[k]) - / mode_lib->vba.SwathHeightC[k] - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]); - double ActiveDRAMClockChangeLatencyMarginC = DPPOPPBufferingC - + EffectiveLBLatencyHidingC + MaxDETBufferingTimeC - - mode_lib->vba.DRAMClockChangeWatermark; - - if (mode_lib->vba.ActiveDPPs > 1) { - ActiveDRAMClockChangeLatencyMarginC = - ActiveDRAMClockChangeLatencyMarginC - - (1 - - 1 - / (mode_lib->vba.ActiveDPPs - - 1)) - * mode_lib->vba.SwathHeightC[k] - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]); - } - mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k] = dml_min( - ActiveDRAMClockChangeLatencyMarginY, - ActiveDRAMClockChangeLatencyMarginC); - } else { - mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k] = - ActiveDRAMClockChangeLatencyMarginY; - } - - if (mode_lib->vba.WritebackEnable[k]) { - double WritebackDRAMClockChangeLatencyMargin; - - if (mode_lib->vba.WritebackPixelFormat[k] == dm_444_32) { - WritebackDRAMClockChangeLatencyMargin = - (double) (mode_lib->vba.WritebackInterfaceLumaBufferSize - + mode_lib->vba.WritebackInterfaceChromaBufferSize) - / (mode_lib->vba.WritebackDestinationWidth[k] - * mode_lib->vba.WritebackDestinationHeight[k] - / (mode_lib->vba.WritebackSourceHeight[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) - * 4) - - mode_lib->vba.WritebackDRAMClockChangeWatermark; - } else if (mode_lib->vba.WritebackPixelFormat[k] == dm_420_10) { - WritebackDRAMClockChangeLatencyMargin = - dml_min( - (double) mode_lib->vba.WritebackInterfaceLumaBufferSize - * 8.0 / 10, - 2.0 - * mode_lib->vba.WritebackInterfaceChromaBufferSize - * 8 / 10) - / (mode_lib->vba.WritebackDestinationWidth[k] - * mode_lib->vba.WritebackDestinationHeight[k] - / (mode_lib->vba.WritebackSourceHeight[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k])) - - mode_lib->vba.WritebackDRAMClockChangeWatermark; - } else { - WritebackDRAMClockChangeLatencyMargin = - dml_min( - (double) mode_lib->vba.WritebackInterfaceLumaBufferSize, - 2.0 - * mode_lib->vba.WritebackInterfaceChromaBufferSize) - / (mode_lib->vba.WritebackDestinationWidth[k] - * mode_lib->vba.WritebackDestinationHeight[k] - / (mode_lib->vba.WritebackSourceHeight[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k])) - - mode_lib->vba.WritebackDRAMClockChangeWatermark; - } - mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k] = dml_min( - mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k], - WritebackDRAMClockChangeLatencyMargin); - } - } - - mode_lib->vba.MinActiveDRAMClockChangeMargin = 999999; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k] - < mode_lib->vba.MinActiveDRAMClockChangeMargin) { - mode_lib->vba.MinActiveDRAMClockChangeMargin = - mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k]; - } - } - - mode_lib->vba.MinActiveDRAMClockChangeLatencySupported = - mode_lib->vba.MinActiveDRAMClockChangeMargin - + mode_lib->vba.DRAMClockChangeLatency; - - if (mode_lib->vba.MinActiveDRAMClockChangeMargin > 0) { - mode_lib->vba.DRAMClockChangeSupport = dm_dram_clock_change_vactive; - } else { - if (mode_lib->vba.SynchronizedVBlank || mode_lib->vba.NumberOfActivePlanes == 1) { - mode_lib->vba.DRAMClockChangeSupport = dm_dram_clock_change_vblank; - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (!mode_lib->vba.AllowDRAMClockChangeDuringVBlank[k]) { - mode_lib->vba.DRAMClockChangeSupport = - dm_dram_clock_change_unsupported; - } - } - } else { - mode_lib->vba.DRAMClockChangeSupport = dm_dram_clock_change_unsupported; - } - } - - //XFC Parameters: - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - if (mode_lib->vba.XFCEnabled[k] == true) { - double TWait; - - mode_lib->vba.XFCSlaveVUpdateOffset[k] = mode_lib->vba.XFCTSlvVupdateOffset; - mode_lib->vba.XFCSlaveVupdateWidth[k] = mode_lib->vba.XFCTSlvVupdateWidth; - mode_lib->vba.XFCSlaveVReadyOffset[k] = mode_lib->vba.XFCTSlvVreadyOffset; - TWait = CalculateTWait( - mode_lib->vba.PrefetchMode, - mode_lib->vba.DRAMClockChangeLatency, - mode_lib->vba.UrgentLatency, - mode_lib->vba.SREnterPlusExitTime); - mode_lib->vba.XFCRemoteSurfaceFlipDelay = CalculateRemoteSurfaceFlipDelay( - mode_lib, - mode_lib->vba.VRatio[k], - mode_lib->vba.SwathWidthY[k], - dml_ceil(mode_lib->vba.BytePerPixelDETY[k], 1), - mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k], - mode_lib->vba.XFCTSlvVupdateOffset, - mode_lib->vba.XFCTSlvVupdateWidth, - mode_lib->vba.XFCTSlvVreadyOffset, - mode_lib->vba.XFCXBUFLatencyTolerance, - mode_lib->vba.XFCFillBWOverhead, - mode_lib->vba.XFCSlvChunkSize, - mode_lib->vba.XFCBusTransportTime, - mode_lib->vba.TCalc, - TWait, - &mode_lib->vba.SrcActiveDrainRate, - &mode_lib->vba.TInitXFill, - &mode_lib->vba.TslvChk); - mode_lib->vba.XFCRemoteSurfaceFlipLatency[k] = - dml_floor( - mode_lib->vba.XFCRemoteSurfaceFlipDelay - / (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]), - 1); - mode_lib->vba.XFCTransferDelay[k] = - dml_ceil( - mode_lib->vba.XFCBusTransportTime - / (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]), - 1); - mode_lib->vba.XFCPrechargeDelay[k] = - dml_ceil( - (mode_lib->vba.XFCBusTransportTime - + mode_lib->vba.TInitXFill - + mode_lib->vba.TslvChk) - / (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]), - 1); - mode_lib->vba.InitFillLevel = mode_lib->vba.XFCXBUFLatencyTolerance - * mode_lib->vba.SrcActiveDrainRate; - mode_lib->vba.FinalFillMargin = - (mode_lib->vba.DestinationLinesToRequestVMInVBlank[k] - + mode_lib->vba.DestinationLinesToRequestRowInVBlank[k]) - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k] - * mode_lib->vba.SrcActiveDrainRate - + mode_lib->vba.XFCFillConstant; - mode_lib->vba.FinalFillLevel = mode_lib->vba.XFCRemoteSurfaceFlipDelay - * mode_lib->vba.SrcActiveDrainRate - + mode_lib->vba.FinalFillMargin; - mode_lib->vba.RemainingFillLevel = dml_max( - 0.0, - mode_lib->vba.FinalFillLevel - mode_lib->vba.InitFillLevel); - mode_lib->vba.TFinalxFill = mode_lib->vba.RemainingFillLevel - / (mode_lib->vba.SrcActiveDrainRate - * mode_lib->vba.XFCFillBWOverhead / 100); - mode_lib->vba.XFCPrefetchMargin[k] = - mode_lib->vba.XFCRemoteSurfaceFlipDelay - + mode_lib->vba.TFinalxFill - + (mode_lib->vba.DestinationLinesToRequestVMInVBlank[k] - + mode_lib->vba.DestinationLinesToRequestRowInVBlank[k]) - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]; - } else { - mode_lib->vba.XFCSlaveVUpdateOffset[k] = 0; - mode_lib->vba.XFCSlaveVupdateWidth[k] = 0; - mode_lib->vba.XFCSlaveVReadyOffset[k] = 0; - mode_lib->vba.XFCRemoteSurfaceFlipLatency[k] = 0; - mode_lib->vba.XFCPrechargeDelay[k] = 0; - mode_lib->vba.XFCTransferDelay[k] = 0; - mode_lib->vba.XFCPrefetchMargin[k] = 0; - } - } -} - -static void DisplayPipeConfiguration(struct display_mode_lib *mode_lib) -{ - double BytePerPixDETY; - double BytePerPixDETC; - double Read256BytesBlockHeightY; - double Read256BytesBlockHeightC; - double Read256BytesBlockWidthY; - double Read256BytesBlockWidthC; - double MaximumSwathHeightY; - double MaximumSwathHeightC; - double MinimumSwathHeightY; - double MinimumSwathHeightC; - double SwathWidth; - double SwathWidthGranularityY; - double SwathWidthGranularityC; - double RoundedUpMaxSwathSizeBytesY; - double RoundedUpMaxSwathSizeBytesC; - unsigned int j, k; - - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - bool MainPlaneDoesODMCombine = false; - - if (mode_lib->vba.SourcePixelFormat[k] == dm_444_64) { - BytePerPixDETY = 8; - BytePerPixDETC = 0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_32) { - BytePerPixDETY = 4; - BytePerPixDETC = 0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_16) { - BytePerPixDETY = 2; - BytePerPixDETC = 0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_8) { - BytePerPixDETY = 1; - BytePerPixDETC = 0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_420_8) { - BytePerPixDETY = 1; - BytePerPixDETC = 2; - } else { - BytePerPixDETY = 4.0 / 3.0; - BytePerPixDETC = 8.0 / 3.0; - } - - if ((mode_lib->vba.SourcePixelFormat[k] == dm_444_64 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_32 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_16 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_8)) { - if (mode_lib->vba.SurfaceTiling[k] == dm_sw_linear) { - Read256BytesBlockHeightY = 1; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_64) { - Read256BytesBlockHeightY = 4; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_32 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_16) { - Read256BytesBlockHeightY = 8; - } else { - Read256BytesBlockHeightY = 16; - } - Read256BytesBlockWidthY = 256 / dml_ceil(BytePerPixDETY, 1) - / Read256BytesBlockHeightY; - Read256BytesBlockHeightC = 0; - Read256BytesBlockWidthC = 0; - } else { - if (mode_lib->vba.SurfaceTiling[k] == dm_sw_linear) { - Read256BytesBlockHeightY = 1; - Read256BytesBlockHeightC = 1; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_420_8) { - Read256BytesBlockHeightY = 16; - Read256BytesBlockHeightC = 8; - } else { - Read256BytesBlockHeightY = 8; - Read256BytesBlockHeightC = 8; - } - Read256BytesBlockWidthY = 256 / dml_ceil(BytePerPixDETY, 1) - / Read256BytesBlockHeightY; - Read256BytesBlockWidthC = 256 / dml_ceil(BytePerPixDETC, 2) - / Read256BytesBlockHeightC; - } - - if (mode_lib->vba.SourceScan[k] == dm_horz) { - MaximumSwathHeightY = Read256BytesBlockHeightY; - MaximumSwathHeightC = Read256BytesBlockHeightC; - } else { - MaximumSwathHeightY = Read256BytesBlockWidthY; - MaximumSwathHeightC = Read256BytesBlockWidthC; - } - - if ((mode_lib->vba.SourcePixelFormat[k] == dm_444_64 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_32 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_16 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_8)) { - if (mode_lib->vba.SurfaceTiling[k] == dm_sw_linear - || (mode_lib->vba.SourcePixelFormat[k] == dm_444_64 - && (mode_lib->vba.SurfaceTiling[k] - == dm_sw_4kb_s - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_4kb_s_x - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_64kb_s - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_64kb_s_t - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_64kb_s_x - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_var_s - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_var_s_x) - && mode_lib->vba.SourceScan[k] == dm_horz)) { - MinimumSwathHeightY = MaximumSwathHeightY; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_8 - && mode_lib->vba.SourceScan[k] != dm_horz) { - MinimumSwathHeightY = MaximumSwathHeightY; - } else { - MinimumSwathHeightY = MaximumSwathHeightY / 2.0; - } - MinimumSwathHeightC = MaximumSwathHeightC; - } else { - if (mode_lib->vba.SurfaceTiling[k] == dm_sw_linear) { - MinimumSwathHeightY = MaximumSwathHeightY; - MinimumSwathHeightC = MaximumSwathHeightC; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_420_8 - && mode_lib->vba.SourceScan[k] == dm_horz) { - MinimumSwathHeightY = MaximumSwathHeightY / 2.0; - MinimumSwathHeightC = MaximumSwathHeightC; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_420_10 - && mode_lib->vba.SourceScan[k] == dm_horz) { - MinimumSwathHeightC = MaximumSwathHeightC / 2.0; - MinimumSwathHeightY = MaximumSwathHeightY; - } else { - MinimumSwathHeightY = MaximumSwathHeightY; - MinimumSwathHeightC = MaximumSwathHeightC; - } - } - - if (mode_lib->vba.SourceScan[k] == dm_horz) { - SwathWidth = mode_lib->vba.ViewportWidth[k]; - } else { - SwathWidth = mode_lib->vba.ViewportHeight[k]; - } - - if (mode_lib->vba.ODMCombineEnabled[k] == true) { - MainPlaneDoesODMCombine = true; - } - for (j = 0; j < mode_lib->vba.NumberOfActivePlanes; ++j) { - if (mode_lib->vba.BlendingAndTiming[k] == j - && mode_lib->vba.ODMCombineEnabled[j] == true) { - MainPlaneDoesODMCombine = true; - } - } - - if (MainPlaneDoesODMCombine == true) { - SwathWidth = dml_min( - SwathWidth, - mode_lib->vba.HActive[k] / 2.0 * mode_lib->vba.HRatio[k]); - } else { - SwathWidth = SwathWidth / mode_lib->vba.DPPPerPlane[k]; - } - - SwathWidthGranularityY = 256 / dml_ceil(BytePerPixDETY, 1) / MaximumSwathHeightY; - RoundedUpMaxSwathSizeBytesY = (dml_ceil( - (double) (SwathWidth - 1), - SwathWidthGranularityY) + SwathWidthGranularityY) * BytePerPixDETY - * MaximumSwathHeightY; - if (mode_lib->vba.SourcePixelFormat[k] == dm_420_10) { - RoundedUpMaxSwathSizeBytesY = dml_ceil(RoundedUpMaxSwathSizeBytesY, 256) - + 256; - } - if (MaximumSwathHeightC > 0) { - SwathWidthGranularityC = 256.0 / dml_ceil(BytePerPixDETC, 2) - / MaximumSwathHeightC; - RoundedUpMaxSwathSizeBytesC = (dml_ceil( - (double) (SwathWidth / 2.0 - 1), - SwathWidthGranularityC) + SwathWidthGranularityC) - * BytePerPixDETC * MaximumSwathHeightC; - if (mode_lib->vba.SourcePixelFormat[k] == dm_420_10) { - RoundedUpMaxSwathSizeBytesC = dml_ceil( - RoundedUpMaxSwathSizeBytesC, - 256) + 256; - } - } else - RoundedUpMaxSwathSizeBytesC = 0.0; - - if (RoundedUpMaxSwathSizeBytesY + RoundedUpMaxSwathSizeBytesC - <= mode_lib->vba.DETBufferSizeInKByte * 1024.0 / 2.0) { - mode_lib->vba.SwathHeightY[k] = MaximumSwathHeightY; - mode_lib->vba.SwathHeightC[k] = MaximumSwathHeightC; - } else { - mode_lib->vba.SwathHeightY[k] = MinimumSwathHeightY; - mode_lib->vba.SwathHeightC[k] = MinimumSwathHeightC; - } - - if (mode_lib->vba.SwathHeightC[k] == 0) { - mode_lib->vba.DETBufferSizeY[k] = mode_lib->vba.DETBufferSizeInKByte * 1024; - mode_lib->vba.DETBufferSizeC[k] = 0; - } else if (mode_lib->vba.SwathHeightY[k] <= mode_lib->vba.SwathHeightC[k]) { - mode_lib->vba.DETBufferSizeY[k] = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 / 2; - mode_lib->vba.DETBufferSizeC[k] = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 / 2; - } else { - mode_lib->vba.DETBufferSizeY[k] = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 * 2 / 3; - mode_lib->vba.DETBufferSizeC[k] = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 / 3; - } - } -} - -bool Calculate256BBlockSizes( - enum source_format_class SourcePixelFormat, - enum dm_swizzle_mode SurfaceTiling, - unsigned int BytePerPixelY, - unsigned int BytePerPixelC, - unsigned int *BlockHeight256BytesY, - unsigned int *BlockHeight256BytesC, - unsigned int *BlockWidth256BytesY, - unsigned int *BlockWidth256BytesC) -{ - if ((SourcePixelFormat == dm_444_64 || SourcePixelFormat == dm_444_32 - || SourcePixelFormat == dm_444_16 - || SourcePixelFormat == dm_444_8)) { - if (SurfaceTiling == dm_sw_linear) { - *BlockHeight256BytesY = 1; - } else if (SourcePixelFormat == dm_444_64) { - *BlockHeight256BytesY = 4; - } else if (SourcePixelFormat == dm_444_8) { - *BlockHeight256BytesY = 16; - } else { - *BlockHeight256BytesY = 8; - } - *BlockWidth256BytesY = 256 / BytePerPixelY / *BlockHeight256BytesY; - *BlockHeight256BytesC = 0; - *BlockWidth256BytesC = 0; - } else { - if (SurfaceTiling == dm_sw_linear) { - *BlockHeight256BytesY = 1; - *BlockHeight256BytesC = 1; - } else if (SourcePixelFormat == dm_420_8) { - *BlockHeight256BytesY = 16; - *BlockHeight256BytesC = 8; - } else { - *BlockHeight256BytesY = 8; - *BlockHeight256BytesC = 8; - } - *BlockWidth256BytesY = 256 / BytePerPixelY / *BlockHeight256BytesY; - *BlockWidth256BytesC = 256 / BytePerPixelC / *BlockHeight256BytesC; - } - return true; -} - -static double CalculateTWait( - unsigned int PrefetchMode, - double DRAMClockChangeLatency, - double UrgentLatency, - double SREnterPlusExitTime) -{ - if (PrefetchMode == 0) { - return dml_max( - DRAMClockChangeLatency + UrgentLatency, - dml_max(SREnterPlusExitTime, UrgentLatency)); - } else if (PrefetchMode == 1) { - return dml_max(SREnterPlusExitTime, UrgentLatency); - } else { - return UrgentLatency; - } -} - -static double CalculateRemoteSurfaceFlipDelay( - struct display_mode_lib *mode_lib, - double VRatio, - double SwathWidth, - double Bpp, - double LineTime, - double XFCTSlvVupdateOffset, - double XFCTSlvVupdateWidth, - double XFCTSlvVreadyOffset, - double XFCXBUFLatencyTolerance, - double XFCFillBWOverhead, - double XFCSlvChunkSize, - double XFCBusTransportTime, - double TCalc, - double TWait, - double *SrcActiveDrainRate, - double *TInitXFill, - double *TslvChk) -{ - double TSlvSetup, AvgfillRate, result; - - *SrcActiveDrainRate = VRatio * SwathWidth * Bpp / LineTime; - TSlvSetup = XFCTSlvVupdateOffset + XFCTSlvVupdateWidth + XFCTSlvVreadyOffset; - *TInitXFill = XFCXBUFLatencyTolerance / (1 + XFCFillBWOverhead / 100); - AvgfillRate = *SrcActiveDrainRate * (1 + XFCFillBWOverhead / 100); - *TslvChk = XFCSlvChunkSize / AvgfillRate; - dml_print( - "DML::CalculateRemoteSurfaceFlipDelay: SrcActiveDrainRate: %f\n", - *SrcActiveDrainRate); - dml_print("DML::CalculateRemoteSurfaceFlipDelay: TSlvSetup: %f\n", TSlvSetup); - dml_print("DML::CalculateRemoteSurfaceFlipDelay: TInitXFill: %f\n", *TInitXFill); - dml_print("DML::CalculateRemoteSurfaceFlipDelay: AvgfillRate: %f\n", AvgfillRate); - dml_print("DML::CalculateRemoteSurfaceFlipDelay: TslvChk: %f\n", *TslvChk); - result = 2 * XFCBusTransportTime + TSlvSetup + TCalc + TWait + *TslvChk + *TInitXFill; // TODO: This doesn't seem to match programming guide - dml_print("DML::CalculateRemoteSurfaceFlipDelay: RemoteSurfaceFlipDelay: %f\n", result); - return result; -} - -static double CalculateWriteBackDISPCLK( - enum source_format_class WritebackPixelFormat, - double PixelClock, - double WritebackHRatio, - double WritebackVRatio, - unsigned int WritebackLumaHTaps, - unsigned int WritebackLumaVTaps, - unsigned int WritebackChromaHTaps, - unsigned int WritebackChromaVTaps, - double WritebackDestinationWidth, - unsigned int HTotal, - unsigned int WritebackChromaLineBufferWidth) -{ - double CalculateWriteBackDISPCLK = - 1.01 * PixelClock - * dml_max( - dml_ceil(WritebackLumaHTaps / 4.0, 1) - / WritebackHRatio, - dml_max( - (WritebackLumaVTaps - * dml_ceil( - 1.0 - / WritebackVRatio, - 1) - * dml_ceil( - WritebackDestinationWidth - / 4.0, - 1) - + dml_ceil( - WritebackDestinationWidth - / 4.0, - 1)) - / (double) HTotal - + dml_ceil( - 1.0 - / WritebackVRatio, - 1) - * (dml_ceil( - WritebackLumaVTaps - / 4.0, - 1) - + 4.0) - / (double) HTotal, - dml_ceil( - 1.0 - / WritebackVRatio, - 1) - * WritebackDestinationWidth - / (double) HTotal)); - if (WritebackPixelFormat != dm_444_32) { - CalculateWriteBackDISPCLK = - dml_max( - CalculateWriteBackDISPCLK, - 1.01 * PixelClock - * dml_max( - dml_ceil( - WritebackChromaHTaps - / 2.0, - 1) - / (2 - * WritebackHRatio), - dml_max( - (WritebackChromaVTaps - * dml_ceil( - 1 - / (2 - * WritebackVRatio), - 1) - * dml_ceil( - WritebackDestinationWidth - / 2.0 - / 2.0, - 1) - + dml_ceil( - WritebackDestinationWidth - / 2.0 - / WritebackChromaLineBufferWidth, - 1)) - / HTotal - + dml_ceil( - 1 - / (2 - * WritebackVRatio), - 1) - * (dml_ceil( - WritebackChromaVTaps - / 4.0, - 1) - + 4) - / HTotal, - dml_ceil( - 1.0 - / (2 - * WritebackVRatio), - 1) - * WritebackDestinationWidth - / 2.0 - / HTotal))); - } - return CalculateWriteBackDISPCLK; -} - -static double CalculateWriteBackDelay( - enum source_format_class WritebackPixelFormat, - double WritebackHRatio, - double WritebackVRatio, - unsigned int WritebackLumaHTaps, - unsigned int WritebackLumaVTaps, - unsigned int WritebackChromaHTaps, - unsigned int WritebackChromaVTaps, - unsigned int WritebackDestinationWidth) -{ - double CalculateWriteBackDelay = - dml_max( - dml_ceil(WritebackLumaHTaps / 4.0, 1) / WritebackHRatio, - WritebackLumaVTaps * dml_ceil(1.0 / WritebackVRatio, 1) - * dml_ceil( - WritebackDestinationWidth - / 4.0, - 1) - + dml_ceil(1.0 / WritebackVRatio, 1) - * (dml_ceil( - WritebackLumaVTaps - / 4.0, - 1) + 4)); - - if (WritebackPixelFormat != dm_444_32) { - CalculateWriteBackDelay = - dml_max( - CalculateWriteBackDelay, - dml_max( - dml_ceil( - WritebackChromaHTaps - / 2.0, - 1) - / (2 - * WritebackHRatio), - WritebackChromaVTaps - * dml_ceil( - 1 - / (2 - * WritebackVRatio), - 1) - * dml_ceil( - WritebackDestinationWidth - / 2.0 - / 2.0, - 1) - + dml_ceil( - 1 - / (2 - * WritebackVRatio), - 1) - * (dml_ceil( - WritebackChromaVTaps - / 4.0, - 1) - + 4))); - } - return CalculateWriteBackDelay; -} - -static void CalculateActiveRowBandwidth( - bool VirtualMemoryEnable, - enum source_format_class SourcePixelFormat, - double VRatio, - bool DCCEnable, - double LineTime, - unsigned int MetaRowByteLuma, - unsigned int MetaRowByteChroma, - unsigned int meta_row_height_luma, - unsigned int meta_row_height_chroma, - unsigned int PixelPTEBytesPerRowLuma, - unsigned int PixelPTEBytesPerRowChroma, - unsigned int dpte_row_height_luma, - unsigned int dpte_row_height_chroma, - double *meta_row_bw, - double *dpte_row_bw, - double *qual_row_bw) -{ - if (DCCEnable != true) { - *meta_row_bw = 0; - } else if (SourcePixelFormat == dm_420_8 || SourcePixelFormat == dm_420_10) { - *meta_row_bw = VRatio * MetaRowByteLuma / (meta_row_height_luma * LineTime) - + VRatio / 2 * MetaRowByteChroma - / (meta_row_height_chroma * LineTime); - } else { - *meta_row_bw = VRatio * MetaRowByteLuma / (meta_row_height_luma * LineTime); - } - - if (VirtualMemoryEnable != true) { - *dpte_row_bw = 0; - } else if (SourcePixelFormat == dm_420_8 || SourcePixelFormat == dm_420_10) { - *dpte_row_bw = VRatio * PixelPTEBytesPerRowLuma / (dpte_row_height_luma * LineTime) - + VRatio / 2 * PixelPTEBytesPerRowChroma - / (dpte_row_height_chroma * LineTime); - } else { - *dpte_row_bw = VRatio * PixelPTEBytesPerRowLuma / (dpte_row_height_luma * LineTime); - } - - if ((SourcePixelFormat == dm_420_8 || SourcePixelFormat == dm_420_10)) { - *qual_row_bw = *meta_row_bw + *dpte_row_bw; - } else { - *qual_row_bw = 0; - } -} - -static void CalculateFlipSchedule( - struct display_mode_lib *mode_lib, - double UrgentExtraLatency, - double UrgentLatency, - unsigned int MaxPageTableLevels, - bool VirtualMemoryEnable, - double BandwidthAvailableForImmediateFlip, - unsigned int TotImmediateFlipBytes, - enum source_format_class SourcePixelFormat, - unsigned int ImmediateFlipBytes, - double LineTime, - double Tno_bw, - double VRatio, - double PDEAndMetaPTEBytesFrame, - unsigned int MetaRowByte, - unsigned int PixelPTEBytesPerRow, - bool DCCEnable, - unsigned int dpte_row_height, - unsigned int meta_row_height, - double qual_row_bw, - double *DestinationLinesToRequestVMInImmediateFlip, - double *DestinationLinesToRequestRowInImmediateFlip, - double *final_flip_bw, - bool *ImmediateFlipSupportedForPipe) -{ - double min_row_time = 0.0; - - if (SourcePixelFormat == dm_420_8 || SourcePixelFormat == dm_420_10) { - *DestinationLinesToRequestVMInImmediateFlip = 0.0; - *DestinationLinesToRequestRowInImmediateFlip = 0.0; - *final_flip_bw = qual_row_bw; - *ImmediateFlipSupportedForPipe = true; - } else { - double TimeForFetchingMetaPTEImmediateFlip; - double TimeForFetchingRowInVBlankImmediateFlip; - - if (VirtualMemoryEnable == true) { - mode_lib->vba.ImmediateFlipBW = BandwidthAvailableForImmediateFlip - * ImmediateFlipBytes / TotImmediateFlipBytes; - TimeForFetchingMetaPTEImmediateFlip = - dml_max( - Tno_bw - + PDEAndMetaPTEBytesFrame - / mode_lib->vba.ImmediateFlipBW, - dml_max( - UrgentExtraLatency - + UrgentLatency - * (MaxPageTableLevels - - 1), - LineTime / 4.0)); - } else { - TimeForFetchingMetaPTEImmediateFlip = 0; - } - - *DestinationLinesToRequestVMInImmediateFlip = dml_floor( - 4.0 * (TimeForFetchingMetaPTEImmediateFlip / LineTime + 0.125), - 1) / 4.0; - - if ((VirtualMemoryEnable == true || DCCEnable == true)) { - mode_lib->vba.ImmediateFlipBW = BandwidthAvailableForImmediateFlip - * ImmediateFlipBytes / TotImmediateFlipBytes; - TimeForFetchingRowInVBlankImmediateFlip = dml_max( - (MetaRowByte + PixelPTEBytesPerRow) - / mode_lib->vba.ImmediateFlipBW, - dml_max(UrgentLatency, LineTime / 4.0)); - } else { - TimeForFetchingRowInVBlankImmediateFlip = 0; - } - - *DestinationLinesToRequestRowInImmediateFlip = dml_floor( - 4.0 * (TimeForFetchingRowInVBlankImmediateFlip / LineTime + 0.125), - 1) / 4.0; - - if (VirtualMemoryEnable == true) { - *final_flip_bw = - dml_max( - PDEAndMetaPTEBytesFrame - / (*DestinationLinesToRequestVMInImmediateFlip - * LineTime), - (MetaRowByte + PixelPTEBytesPerRow) - / (TimeForFetchingRowInVBlankImmediateFlip - * LineTime)); - } else if (MetaRowByte + PixelPTEBytesPerRow > 0) { - *final_flip_bw = (MetaRowByte + PixelPTEBytesPerRow) - / (TimeForFetchingRowInVBlankImmediateFlip * LineTime); - } else { - *final_flip_bw = 0; - } - - if (VirtualMemoryEnable && !DCCEnable) - min_row_time = dpte_row_height * LineTime / VRatio; - else if (!VirtualMemoryEnable && DCCEnable) - min_row_time = meta_row_height * LineTime / VRatio; - else - min_row_time = dml_min(dpte_row_height, meta_row_height) * LineTime - / VRatio; - - if (*DestinationLinesToRequestVMInImmediateFlip >= 8 - || *DestinationLinesToRequestRowInImmediateFlip >= 16 - || TimeForFetchingMetaPTEImmediateFlip - + 2 * TimeForFetchingRowInVBlankImmediateFlip - > min_row_time) - *ImmediateFlipSupportedForPipe = false; - else - *ImmediateFlipSupportedForPipe = true; - } -} - -static void PixelClockAdjustmentForProgressiveToInterlaceUnit(struct display_mode_lib *mode_lib) -{ - unsigned int k; - - //Progressive To dml_ml->vba.Interlace Unit Effect - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { - mode_lib->vba.PixelClockBackEnd[k] = mode_lib->vba.PixelClock[k]; - if (mode_lib->vba.Interlace[k] == 1 - && mode_lib->vba.ProgressiveToInterlaceUnitInOPP == true) { - mode_lib->vba.PixelClock[k] = 2 * mode_lib->vba.PixelClock[k]; - } - } -} - -static unsigned int CursorBppEnumToBits(enum cursor_bpp ebpp) -{ - switch (ebpp) { - case dm_cur_2bit: - return 2; - case dm_cur_32bit: - return 32; - case dm_cur_64bit: - return 64; - default: - return 0; - } -} - -static unsigned int TruncToValidBPP( - double DecimalBPP, - bool DSCEnabled, - enum output_encoder_class Output, - enum output_format_class Format, - unsigned int DSCInputBitPerComponent) -{ - if (Output == dm_hdmi) { - if (Format == dm_420) { - if (DecimalBPP >= 18) - return 18; - else if (DecimalBPP >= 15) - return 15; - else if (DecimalBPP >= 12) - return 12; - else - return BPP_INVALID; - } else if (Format == dm_444) { - if (DecimalBPP >= 36) - return 36; - else if (DecimalBPP >= 30) - return 30; - else if (DecimalBPP >= 24) - return 24; - else - return BPP_INVALID; - } else { - if (DecimalBPP / 1.5 >= 24) - return 24; - else if (DecimalBPP / 1.5 >= 20) - return 20; - else if (DecimalBPP / 1.5 >= 16) - return 16; - else - return BPP_INVALID; - } - } else { - if (DSCEnabled) { - if (Format == dm_420) { - if (DecimalBPP < 6) - return BPP_INVALID; - else if (DecimalBPP >= 1.5 * DSCInputBitPerComponent - 1 / 16) - return 1.5 * DSCInputBitPerComponent - 1 / 16; - else - return dml_floor(16 * DecimalBPP, 1) / 16; - } else if (Format == dm_n422) { - if (DecimalBPP < 7) - return BPP_INVALID; - else if (DecimalBPP >= 2 * DSCInputBitPerComponent - 1 / 16) - return 2 * DSCInputBitPerComponent - 1 / 16; - else - return dml_floor(16 * DecimalBPP, 1) / 16; - } else { - if (DecimalBPP < 8) - return BPP_INVALID; - else if (DecimalBPP >= 3 * DSCInputBitPerComponent - 1 / 16) - return 3 * DSCInputBitPerComponent - 1 / 16; - else - return dml_floor(16 * DecimalBPP, 1) / 16; - } - } else if (Format == dm_420) { - if (DecimalBPP >= 18) - return 18; - else if (DecimalBPP >= 15) - return 15; - else if (DecimalBPP >= 12) - return 12; - else - return BPP_INVALID; - } else if (Format == dm_s422 || Format == dm_n422) { - if (DecimalBPP >= 24) - return 24; - else if (DecimalBPP >= 20) - return 20; - else if (DecimalBPP >= 16) - return 16; - else - return BPP_INVALID; - } else { - if (DecimalBPP >= 36) - return 36; - else if (DecimalBPP >= 30) - return 30; - else if (DecimalBPP >= 24) - return 24; - else - return BPP_INVALID; - } - } -} - -static void ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) -{ - int i; - unsigned int j, k; - /*MODE SUPPORT, VOLTAGE STATE AND SOC CONFIGURATION*/ - - /*Scale Ratio, taps Support Check*/ - - mode_lib->vba.ScaleRatioAndTapsSupport = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.ScalerEnabled[k] == false - && ((mode_lib->vba.SourcePixelFormat[k] != dm_444_64 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_32 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_mono_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_mono_8) - || mode_lib->vba.HRatio[k] != 1.0 - || mode_lib->vba.htaps[k] != 1.0 - || mode_lib->vba.VRatio[k] != 1.0 - || mode_lib->vba.vtaps[k] != 1.0)) { - mode_lib->vba.ScaleRatioAndTapsSupport = false; - } else if (mode_lib->vba.vtaps[k] < 1.0 || mode_lib->vba.vtaps[k] > 8.0 - || mode_lib->vba.htaps[k] < 1.0 || mode_lib->vba.htaps[k] > 8.0 - || (mode_lib->vba.htaps[k] > 1.0 - && (mode_lib->vba.htaps[k] % 2) == 1) - || mode_lib->vba.HRatio[k] > mode_lib->vba.MaxHSCLRatio - || mode_lib->vba.VRatio[k] > mode_lib->vba.MaxVSCLRatio - || mode_lib->vba.HRatio[k] > mode_lib->vba.htaps[k] - || mode_lib->vba.VRatio[k] > mode_lib->vba.vtaps[k] - || (mode_lib->vba.SourcePixelFormat[k] != dm_444_64 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_32 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_mono_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_mono_8 - && (mode_lib->vba.HRatio[k] / 2.0 - > mode_lib->vba.HTAPsChroma[k] - || mode_lib->vba.VRatio[k] / 2.0 - > mode_lib->vba.VTAPsChroma[k]))) { - mode_lib->vba.ScaleRatioAndTapsSupport = false; - } - } - /*Source Format, Pixel Format and Scan Support Check*/ - - mode_lib->vba.SourceFormatPixelAndScanSupport = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if ((mode_lib->vba.SurfaceTiling[k] == dm_sw_linear - && mode_lib->vba.SourceScan[k] != dm_horz) - || ((mode_lib->vba.SurfaceTiling[k] == dm_sw_4kb_d - || mode_lib->vba.SurfaceTiling[k] == dm_sw_4kb_d_x - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_d - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_d_t - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_d_x - || mode_lib->vba.SurfaceTiling[k] == dm_sw_var_d - || mode_lib->vba.SurfaceTiling[k] == dm_sw_var_d_x) - && mode_lib->vba.SourcePixelFormat[k] != dm_444_64) - || (mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_r_x - && (mode_lib->vba.SourcePixelFormat[k] == dm_mono_8 - || mode_lib->vba.SourcePixelFormat[k] - == dm_420_8 - || mode_lib->vba.SourcePixelFormat[k] - == dm_420_10)) - || (((mode_lib->vba.SurfaceTiling[k] - == dm_sw_gfx7_2d_thin_gl - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_gfx7_2d_thin_lvp) - && !((mode_lib->vba.SourcePixelFormat[k] - == dm_444_64 - || mode_lib->vba.SourcePixelFormat[k] - == dm_444_32) - && mode_lib->vba.SourceScan[k] - == dm_horz - && mode_lib->vba.SupportGFX7CompatibleTilingIn32bppAnd64bpp - == true - && mode_lib->vba.DCCEnable[k] - == false)) - || (mode_lib->vba.DCCEnable[k] == true - && (mode_lib->vba.SurfaceTiling[k] - == dm_sw_linear - || mode_lib->vba.SourcePixelFormat[k] - == dm_420_8 - || mode_lib->vba.SourcePixelFormat[k] - == dm_420_10)))) { - mode_lib->vba.SourceFormatPixelAndScanSupport = false; - } - } - /*Bandwidth Support Check*/ - - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.SourceScan[k] == dm_horz) { - mode_lib->vba.SwathWidthYSingleDPP[k] = mode_lib->vba.ViewportWidth[k]; - } else { - mode_lib->vba.SwathWidthYSingleDPP[k] = mode_lib->vba.ViewportHeight[k]; - } - if (mode_lib->vba.SourcePixelFormat[k] == dm_444_64) { - mode_lib->vba.BytePerPixelInDETY[k] = 8.0; - mode_lib->vba.BytePerPixelInDETC[k] = 0.0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_32) { - mode_lib->vba.BytePerPixelInDETY[k] = 4.0; - mode_lib->vba.BytePerPixelInDETC[k] = 0.0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_444_16 - || mode_lib->vba.SourcePixelFormat[k] == dm_mono_16) { - mode_lib->vba.BytePerPixelInDETY[k] = 2.0; - mode_lib->vba.BytePerPixelInDETC[k] = 0.0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_mono_8) { - mode_lib->vba.BytePerPixelInDETY[k] = 1.0; - mode_lib->vba.BytePerPixelInDETC[k] = 0.0; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_420_8) { - mode_lib->vba.BytePerPixelInDETY[k] = 1.0; - mode_lib->vba.BytePerPixelInDETC[k] = 2.0; - } else { - mode_lib->vba.BytePerPixelInDETY[k] = 4.0 / 3; - mode_lib->vba.BytePerPixelInDETC[k] = 8.0 / 3; - } - } - mode_lib->vba.TotalReadBandwidthConsumedGBytePerSecond = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.ReadBandwidth[k] = mode_lib->vba.SwathWidthYSingleDPP[k] - * (dml_ceil(mode_lib->vba.BytePerPixelInDETY[k], 1.0) - * mode_lib->vba.VRatio[k] - + dml_ceil(mode_lib->vba.BytePerPixelInDETC[k], 2.0) - / 2.0 * mode_lib->vba.VRatio[k] / 2) - / (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]); - if (mode_lib->vba.DCCEnable[k] == true) { - mode_lib->vba.ReadBandwidth[k] = mode_lib->vba.ReadBandwidth[k] - * (1 + 1 / 256); - } - if (mode_lib->vba.VirtualMemoryEnable == true - && mode_lib->vba.SourceScan[k] != dm_horz - && (mode_lib->vba.SurfaceTiling[k] == dm_sw_4kb_s - || mode_lib->vba.SurfaceTiling[k] == dm_sw_4kb_s_x - || mode_lib->vba.SurfaceTiling[k] == dm_sw_4kb_d - || mode_lib->vba.SurfaceTiling[k] == dm_sw_4kb_d_x)) { - mode_lib->vba.ReadBandwidth[k] = mode_lib->vba.ReadBandwidth[k] - * (1 + 1 / 64); - } else if (mode_lib->vba.VirtualMemoryEnable == true - && mode_lib->vba.SourceScan[k] == dm_horz - && (mode_lib->vba.SourcePixelFormat[k] == dm_444_64 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_32) - && (mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_s - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_s_t - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_s_x - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_d - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_d_t - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_d_x - || mode_lib->vba.SurfaceTiling[k] == dm_sw_64kb_r_x)) { - mode_lib->vba.ReadBandwidth[k] = mode_lib->vba.ReadBandwidth[k] - * (1 + 1 / 256); - } else if (mode_lib->vba.VirtualMemoryEnable == true) { - mode_lib->vba.ReadBandwidth[k] = mode_lib->vba.ReadBandwidth[k] - * (1 + 1 / 512); - } - mode_lib->vba.TotalReadBandwidthConsumedGBytePerSecond = - mode_lib->vba.TotalReadBandwidthConsumedGBytePerSecond - + mode_lib->vba.ReadBandwidth[k] / 1000.0; - } - mode_lib->vba.TotalWriteBandwidthConsumedGBytePerSecond = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.WritebackEnable[k] == true - && mode_lib->vba.WritebackPixelFormat[k] == dm_444_32) { - mode_lib->vba.WriteBandwidth[k] = mode_lib->vba.WritebackDestinationWidth[k] - * mode_lib->vba.WritebackDestinationHeight[k] - / (mode_lib->vba.WritebackSourceHeight[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) * 4.0; - } else if (mode_lib->vba.WritebackEnable[k] == true - && mode_lib->vba.WritebackPixelFormat[k] == dm_420_10) { - mode_lib->vba.WriteBandwidth[k] = mode_lib->vba.WritebackDestinationWidth[k] - * mode_lib->vba.WritebackDestinationHeight[k] - / (mode_lib->vba.WritebackSourceHeight[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) * 3.0; - } else if (mode_lib->vba.WritebackEnable[k] == true) { - mode_lib->vba.WriteBandwidth[k] = mode_lib->vba.WritebackDestinationWidth[k] - * mode_lib->vba.WritebackDestinationHeight[k] - / (mode_lib->vba.WritebackSourceHeight[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) * 1.5; - } else { - mode_lib->vba.WriteBandwidth[k] = 0.0; - } - mode_lib->vba.TotalWriteBandwidthConsumedGBytePerSecond = - mode_lib->vba.TotalWriteBandwidthConsumedGBytePerSecond - + mode_lib->vba.WriteBandwidth[k] / 1000.0; - } - mode_lib->vba.TotalBandwidthConsumedGBytePerSecond = - mode_lib->vba.TotalReadBandwidthConsumedGBytePerSecond - + mode_lib->vba.TotalWriteBandwidthConsumedGBytePerSecond; - mode_lib->vba.DCCEnabledInAnyPlane = false; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.DCCEnable[k] == true) { - mode_lib->vba.DCCEnabledInAnyPlane = true; - } - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.FabricAndDRAMBandwidthPerState[i] = dml_min( - mode_lib->vba.DRAMSpeedPerState[i] * mode_lib->vba.NumberOfChannels - * mode_lib->vba.DRAMChannelWidth, - mode_lib->vba.FabricClockPerState[i] - * mode_lib->vba.FabricDatapathToDCNDataReturn) - / 1000; - mode_lib->vba.ReturnBWToDCNPerState = dml_min( - mode_lib->vba.ReturnBusWidth * mode_lib->vba.DCFCLKPerState[i], - mode_lib->vba.FabricAndDRAMBandwidthPerState[i] * 1000.0) - * mode_lib->vba.PercentOfIdealDRAMAndFabricBWReceivedAfterUrgLatency - / 100; - mode_lib->vba.ReturnBWPerState[i] = mode_lib->vba.ReturnBWToDCNPerState; - if (mode_lib->vba.DCCEnabledInAnyPlane == true - && mode_lib->vba.ReturnBWToDCNPerState - > mode_lib->vba.DCFCLKPerState[i] - * mode_lib->vba.ReturnBusWidth - / 4.0) { - mode_lib->vba.ReturnBWPerState[i] = - dml_min( - mode_lib->vba.ReturnBWPerState[i], - mode_lib->vba.ReturnBWToDCNPerState * 4.0 - * (1.0 - - mode_lib->vba.UrgentLatency - / ((mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0 - / (mode_lib->vba.ReturnBWToDCNPerState - - mode_lib->vba.DCFCLKPerState[i] - * mode_lib->vba.ReturnBusWidth - / 4.0) - + mode_lib->vba.UrgentLatency))); - } - mode_lib->vba.CriticalPoint = - 2.0 * mode_lib->vba.ReturnBusWidth * mode_lib->vba.DCFCLKPerState[i] - * mode_lib->vba.UrgentLatency - / (mode_lib->vba.ReturnBWToDCNPerState - * mode_lib->vba.UrgentLatency - + (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0); - if (mode_lib->vba.DCCEnabledInAnyPlane == true && mode_lib->vba.CriticalPoint > 1.0 - && mode_lib->vba.CriticalPoint < 4.0) { - mode_lib->vba.ReturnBWPerState[i] = - dml_min( - mode_lib->vba.ReturnBWPerState[i], - dml_pow( - 4.0 - * mode_lib->vba.ReturnBWToDCNPerState - * (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0 - * mode_lib->vba.ReturnBusWidth - * mode_lib->vba.DCFCLKPerState[i] - * mode_lib->vba.UrgentLatency - / (mode_lib->vba.ReturnBWToDCNPerState - * mode_lib->vba.UrgentLatency - + (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0), - 2)); - } - mode_lib->vba.ReturnBWToDCNPerState = dml_min( - mode_lib->vba.ReturnBusWidth * mode_lib->vba.DCFCLKPerState[i], - mode_lib->vba.FabricAndDRAMBandwidthPerState[i] * 1000.0); - if (mode_lib->vba.DCCEnabledInAnyPlane == true - && mode_lib->vba.ReturnBWToDCNPerState - > mode_lib->vba.DCFCLKPerState[i] - * mode_lib->vba.ReturnBusWidth - / 4.0) { - mode_lib->vba.ReturnBWPerState[i] = - dml_min( - mode_lib->vba.ReturnBWPerState[i], - mode_lib->vba.ReturnBWToDCNPerState * 4.0 - * (1.0 - - mode_lib->vba.UrgentLatency - / ((mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0 - / (mode_lib->vba.ReturnBWToDCNPerState - - mode_lib->vba.DCFCLKPerState[i] - * mode_lib->vba.ReturnBusWidth - / 4.0) - + mode_lib->vba.UrgentLatency))); - } - mode_lib->vba.CriticalPoint = - 2.0 * mode_lib->vba.ReturnBusWidth * mode_lib->vba.DCFCLKPerState[i] - * mode_lib->vba.UrgentLatency - / (mode_lib->vba.ReturnBWToDCNPerState - * mode_lib->vba.UrgentLatency - + (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0); - if (mode_lib->vba.DCCEnabledInAnyPlane == true && mode_lib->vba.CriticalPoint > 1.0 - && mode_lib->vba.CriticalPoint < 4.0) { - mode_lib->vba.ReturnBWPerState[i] = - dml_min( - mode_lib->vba.ReturnBWPerState[i], - dml_pow( - 4.0 - * mode_lib->vba.ReturnBWToDCNPerState - * (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0 - * mode_lib->vba.ReturnBusWidth - * mode_lib->vba.DCFCLKPerState[i] - * mode_lib->vba.UrgentLatency - / (mode_lib->vba.ReturnBWToDCNPerState - * mode_lib->vba.UrgentLatency - + (mode_lib->vba.ROBBufferSizeInKByte - - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0), - 2)); - } - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - if ((mode_lib->vba.TotalReadBandwidthConsumedGBytePerSecond * 1000.0 - <= mode_lib->vba.ReturnBWPerState[i]) - && (mode_lib->vba.TotalBandwidthConsumedGBytePerSecond * 1000.0 - <= mode_lib->vba.FabricAndDRAMBandwidthPerState[i] - * 1000.0 - * mode_lib->vba.PercentOfIdealDRAMAndFabricBWReceivedAfterUrgLatency - / 100.0)) { - mode_lib->vba.BandwidthSupport[i] = true; - } else { - mode_lib->vba.BandwidthSupport[i] = false; - } - } - /*Writeback Latency support check*/ - - mode_lib->vba.WritebackLatencySupport = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.WritebackEnable[k] == true) { - if (mode_lib->vba.WritebackPixelFormat[k] == dm_444_32) { - if (mode_lib->vba.WriteBandwidth[k] - > (mode_lib->vba.WritebackInterfaceLumaBufferSize - + mode_lib->vba.WritebackInterfaceChromaBufferSize) - / mode_lib->vba.WritebackLatency) { - mode_lib->vba.WritebackLatencySupport = false; - } - } else { - if (mode_lib->vba.WriteBandwidth[k] - > 1.5 - * dml_min( - mode_lib->vba.WritebackInterfaceLumaBufferSize, - 2.0 - * mode_lib->vba.WritebackInterfaceChromaBufferSize) - / mode_lib->vba.WritebackLatency) { - mode_lib->vba.WritebackLatencySupport = false; - } - } - } - } - /*Re-ordering Buffer Support Check*/ - - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.UrgentRoundTripAndOutOfOrderLatencyPerState[i] = - (mode_lib->vba.RoundTripPingLatencyCycles + 32.0) - / mode_lib->vba.DCFCLKPerState[i] - + mode_lib->vba.UrgentOutOfOrderReturnPerChannel - * mode_lib->vba.NumberOfChannels - / mode_lib->vba.ReturnBWPerState[i]; - if ((mode_lib->vba.ROBBufferSizeInKByte - mode_lib->vba.PixelChunkSizeInKByte) - * 1024.0 / mode_lib->vba.ReturnBWPerState[i] - > mode_lib->vba.UrgentRoundTripAndOutOfOrderLatencyPerState[i]) { - mode_lib->vba.ROBSupport[i] = true; - } else { - mode_lib->vba.ROBSupport[i] = false; - } - } - /*Writeback Mode Support Check*/ - - mode_lib->vba.TotalNumberOfActiveWriteback = 0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.WritebackEnable[k] == true) { - mode_lib->vba.TotalNumberOfActiveWriteback = - mode_lib->vba.TotalNumberOfActiveWriteback + 1; - } - } - mode_lib->vba.WritebackModeSupport = true; - if (mode_lib->vba.TotalNumberOfActiveWriteback > mode_lib->vba.MaxNumWriteback) { - mode_lib->vba.WritebackModeSupport = false; - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.WritebackEnable[k] == true - && mode_lib->vba.Writeback10bpc420Supported != true - && mode_lib->vba.WritebackPixelFormat[k] == dm_420_10) { - mode_lib->vba.WritebackModeSupport = false; - } - } - /*Writeback Scale Ratio and Taps Support Check*/ - - mode_lib->vba.WritebackScaleRatioAndTapsSupport = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.WritebackEnable[k] == true) { - if (mode_lib->vba.WritebackLumaAndChromaScalingSupported == false - && (mode_lib->vba.WritebackHRatio[k] != 1.0 - || mode_lib->vba.WritebackVRatio[k] != 1.0)) { - mode_lib->vba.WritebackScaleRatioAndTapsSupport = false; - } - if (mode_lib->vba.WritebackHRatio[k] > mode_lib->vba.WritebackMaxHSCLRatio - || mode_lib->vba.WritebackVRatio[k] - > mode_lib->vba.WritebackMaxVSCLRatio - || mode_lib->vba.WritebackHRatio[k] - < mode_lib->vba.WritebackMinHSCLRatio - || mode_lib->vba.WritebackVRatio[k] - < mode_lib->vba.WritebackMinVSCLRatio - || mode_lib->vba.WritebackLumaHTaps[k] - > mode_lib->vba.WritebackMaxHSCLTaps - || mode_lib->vba.WritebackLumaVTaps[k] - > mode_lib->vba.WritebackMaxVSCLTaps - || mode_lib->vba.WritebackHRatio[k] - > mode_lib->vba.WritebackLumaHTaps[k] - || mode_lib->vba.WritebackVRatio[k] - > mode_lib->vba.WritebackLumaVTaps[k] - || (mode_lib->vba.WritebackLumaHTaps[k] > 2.0 - && ((mode_lib->vba.WritebackLumaHTaps[k] % 2) - == 1)) - || (mode_lib->vba.WritebackPixelFormat[k] != dm_444_32 - && (mode_lib->vba.WritebackChromaHTaps[k] - > mode_lib->vba.WritebackMaxHSCLTaps - || mode_lib->vba.WritebackChromaVTaps[k] - > mode_lib->vba.WritebackMaxVSCLTaps - || 2.0 - * mode_lib->vba.WritebackHRatio[k] - > mode_lib->vba.WritebackChromaHTaps[k] - || 2.0 - * mode_lib->vba.WritebackVRatio[k] - > mode_lib->vba.WritebackChromaVTaps[k] - || (mode_lib->vba.WritebackChromaHTaps[k] > 2.0 - && ((mode_lib->vba.WritebackChromaHTaps[k] % 2) == 1))))) { - mode_lib->vba.WritebackScaleRatioAndTapsSupport = false; - } - if (mode_lib->vba.WritebackVRatio[k] < 1.0) { - mode_lib->vba.WritebackLumaVExtra = - dml_max(1.0 - 2.0 / dml_ceil(1.0 / mode_lib->vba.WritebackVRatio[k], 1.0), 0.0); - } else { - mode_lib->vba.WritebackLumaVExtra = -1; - } - if ((mode_lib->vba.WritebackPixelFormat[k] == dm_444_32 - && mode_lib->vba.WritebackLumaVTaps[k] - > (mode_lib->vba.WritebackLineBufferLumaBufferSize - + mode_lib->vba.WritebackLineBufferChromaBufferSize) - / 3.0 - / mode_lib->vba.WritebackDestinationWidth[k] - - mode_lib->vba.WritebackLumaVExtra) - || (mode_lib->vba.WritebackPixelFormat[k] == dm_420_8 - && mode_lib->vba.WritebackLumaVTaps[k] - > mode_lib->vba.WritebackLineBufferLumaBufferSize - / mode_lib->vba.WritebackDestinationWidth[k] - - mode_lib->vba.WritebackLumaVExtra) - || (mode_lib->vba.WritebackPixelFormat[k] == dm_420_10 - && mode_lib->vba.WritebackLumaVTaps[k] - > mode_lib->vba.WritebackLineBufferLumaBufferSize - * 8.0 / 10.0 - / mode_lib->vba.WritebackDestinationWidth[k] - - mode_lib->vba.WritebackLumaVExtra)) { - mode_lib->vba.WritebackScaleRatioAndTapsSupport = false; - } - if (2.0 * mode_lib->vba.WritebackVRatio[k] < 1) { - mode_lib->vba.WritebackChromaVExtra = 0.0; - } else { - mode_lib->vba.WritebackChromaVExtra = -1; - } - if ((mode_lib->vba.WritebackPixelFormat[k] == dm_420_8 - && mode_lib->vba.WritebackChromaVTaps[k] - > mode_lib->vba.WritebackLineBufferChromaBufferSize - / mode_lib->vba.WritebackDestinationWidth[k] - - mode_lib->vba.WritebackChromaVExtra) - || (mode_lib->vba.WritebackPixelFormat[k] == dm_420_10 - && mode_lib->vba.WritebackChromaVTaps[k] - > mode_lib->vba.WritebackLineBufferChromaBufferSize - * 8.0 / 10.0 - / mode_lib->vba.WritebackDestinationWidth[k] - - mode_lib->vba.WritebackChromaVExtra)) { - mode_lib->vba.WritebackScaleRatioAndTapsSupport = false; - } - } - } - /*Maximum DISPCLK/DPPCLK Support check*/ - - mode_lib->vba.WritebackRequiredDISPCLK = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.WritebackEnable[k] == true) { - mode_lib->vba.WritebackRequiredDISPCLK = - dml_max( - mode_lib->vba.WritebackRequiredDISPCLK, - CalculateWriteBackDISPCLK( - mode_lib->vba.WritebackPixelFormat[k], - mode_lib->vba.PixelClock[k], - mode_lib->vba.WritebackHRatio[k], - mode_lib->vba.WritebackVRatio[k], - mode_lib->vba.WritebackLumaHTaps[k], - mode_lib->vba.WritebackLumaVTaps[k], - mode_lib->vba.WritebackChromaHTaps[k], - mode_lib->vba.WritebackChromaVTaps[k], - mode_lib->vba.WritebackDestinationWidth[k], - mode_lib->vba.HTotal[k], - mode_lib->vba.WritebackChromaLineBufferWidth)); - } - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.HRatio[k] > 1.0) { - mode_lib->vba.PSCL_FACTOR[k] = dml_min( - mode_lib->vba.MaxDCHUBToPSCLThroughput, - mode_lib->vba.MaxPSCLToLBThroughput - * mode_lib->vba.HRatio[k] - / dml_ceil( - mode_lib->vba.htaps[k] - / 6.0, - 1.0)); - } else { - mode_lib->vba.PSCL_FACTOR[k] = dml_min( - mode_lib->vba.MaxDCHUBToPSCLThroughput, - mode_lib->vba.MaxPSCLToLBThroughput); - } - if (mode_lib->vba.BytePerPixelInDETC[k] == 0.0) { - mode_lib->vba.PSCL_FACTOR_CHROMA[k] = 0.0; - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] = - mode_lib->vba.PixelClock[k] - * dml_max3( - mode_lib->vba.vtaps[k] / 6.0 - * dml_min( - 1.0, - mode_lib->vba.HRatio[k]), - mode_lib->vba.HRatio[k] - * mode_lib->vba.VRatio[k] - / mode_lib->vba.PSCL_FACTOR[k], - 1.0); - if ((mode_lib->vba.htaps[k] > 6.0 || mode_lib->vba.vtaps[k] > 6.0) - && mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - < 2.0 * mode_lib->vba.PixelClock[k]) { - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] = 2.0 - * mode_lib->vba.PixelClock[k]; - } - } else { - if (mode_lib->vba.HRatio[k] / 2.0 > 1.0) { - mode_lib->vba.PSCL_FACTOR_CHROMA[k] = - dml_min( - mode_lib->vba.MaxDCHUBToPSCLThroughput, - mode_lib->vba.MaxPSCLToLBThroughput - * mode_lib->vba.HRatio[k] - / 2.0 - / dml_ceil( - mode_lib->vba.HTAPsChroma[k] - / 6.0, - 1.0)); - } else { - mode_lib->vba.PSCL_FACTOR_CHROMA[k] = dml_min( - mode_lib->vba.MaxDCHUBToPSCLThroughput, - mode_lib->vba.MaxPSCLToLBThroughput); - } - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] = - mode_lib->vba.PixelClock[k] - * dml_max5( - mode_lib->vba.vtaps[k] / 6.0 - * dml_min( - 1.0, - mode_lib->vba.HRatio[k]), - mode_lib->vba.HRatio[k] - * mode_lib->vba.VRatio[k] - / mode_lib->vba.PSCL_FACTOR[k], - mode_lib->vba.VTAPsChroma[k] - / 6.0 - * dml_min( - 1.0, - mode_lib->vba.HRatio[k] - / 2.0), - mode_lib->vba.HRatio[k] - * mode_lib->vba.VRatio[k] - / 4.0 - / mode_lib->vba.PSCL_FACTOR_CHROMA[k], - 1.0); - if ((mode_lib->vba.htaps[k] > 6.0 || mode_lib->vba.vtaps[k] > 6.0 - || mode_lib->vba.HTAPsChroma[k] > 6.0 - || mode_lib->vba.VTAPsChroma[k] > 6.0) - && mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - < 2.0 * mode_lib->vba.PixelClock[k]) { - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] = 2.0 - * mode_lib->vba.PixelClock[k]; - } - } - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - Calculate256BBlockSizes( - mode_lib->vba.SourcePixelFormat[k], - mode_lib->vba.SurfaceTiling[k], - dml_ceil(mode_lib->vba.BytePerPixelInDETY[k], 1.0), - dml_ceil(mode_lib->vba.BytePerPixelInDETC[k], 2.0), - &mode_lib->vba.Read256BlockHeightY[k], - &mode_lib->vba.Read256BlockHeightC[k], - &mode_lib->vba.Read256BlockWidthY[k], - &mode_lib->vba.Read256BlockWidthC[k]); - if (mode_lib->vba.SourceScan[k] == dm_horz) { - mode_lib->vba.MaxSwathHeightY[k] = mode_lib->vba.Read256BlockHeightY[k]; - mode_lib->vba.MaxSwathHeightC[k] = mode_lib->vba.Read256BlockHeightC[k]; - } else { - mode_lib->vba.MaxSwathHeightY[k] = mode_lib->vba.Read256BlockWidthY[k]; - mode_lib->vba.MaxSwathHeightC[k] = mode_lib->vba.Read256BlockWidthC[k]; - } - if ((mode_lib->vba.SourcePixelFormat[k] == dm_444_64 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_32 - || mode_lib->vba.SourcePixelFormat[k] == dm_444_16 - || mode_lib->vba.SourcePixelFormat[k] == dm_mono_16 - || mode_lib->vba.SourcePixelFormat[k] == dm_mono_8)) { - if (mode_lib->vba.SurfaceTiling[k] == dm_sw_linear - || (mode_lib->vba.SourcePixelFormat[k] == dm_444_64 - && (mode_lib->vba.SurfaceTiling[k] - == dm_sw_4kb_s - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_4kb_s_x - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_64kb_s - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_64kb_s_t - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_64kb_s_x - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_var_s - || mode_lib->vba.SurfaceTiling[k] - == dm_sw_var_s_x) - && mode_lib->vba.SourceScan[k] == dm_horz)) { - mode_lib->vba.MinSwathHeightY[k] = mode_lib->vba.MaxSwathHeightY[k]; - } else { - mode_lib->vba.MinSwathHeightY[k] = mode_lib->vba.MaxSwathHeightY[k] - / 2.0; - } - mode_lib->vba.MinSwathHeightC[k] = mode_lib->vba.MaxSwathHeightC[k]; - } else { - if (mode_lib->vba.SurfaceTiling[k] == dm_sw_linear) { - mode_lib->vba.MinSwathHeightY[k] = mode_lib->vba.MaxSwathHeightY[k]; - mode_lib->vba.MinSwathHeightC[k] = mode_lib->vba.MaxSwathHeightC[k]; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_420_8 - && mode_lib->vba.SourceScan[k] == dm_horz) { - mode_lib->vba.MinSwathHeightY[k] = mode_lib->vba.MaxSwathHeightY[k] - / 2.0; - mode_lib->vba.MinSwathHeightC[k] = mode_lib->vba.MaxSwathHeightC[k]; - } else if (mode_lib->vba.SourcePixelFormat[k] == dm_420_10 - && mode_lib->vba.SourceScan[k] == dm_horz) { - mode_lib->vba.MinSwathHeightC[k] = mode_lib->vba.MaxSwathHeightC[k] - / 2.0; - mode_lib->vba.MinSwathHeightY[k] = mode_lib->vba.MaxSwathHeightY[k]; - } else { - mode_lib->vba.MinSwathHeightY[k] = mode_lib->vba.MaxSwathHeightY[k]; - mode_lib->vba.MinSwathHeightC[k] = mode_lib->vba.MaxSwathHeightC[k]; - } - } - if (mode_lib->vba.SurfaceTiling[k] == dm_sw_linear) { - mode_lib->vba.MaximumSwathWidthSupport = 8192.0; - } else { - mode_lib->vba.MaximumSwathWidthSupport = 5120.0; - } - mode_lib->vba.MaximumSwathWidthInDETBuffer = - dml_min( - mode_lib->vba.MaximumSwathWidthSupport, - mode_lib->vba.DETBufferSizeInKByte * 1024.0 / 2.0 - / (mode_lib->vba.BytePerPixelInDETY[k] - * mode_lib->vba.MinSwathHeightY[k] - + mode_lib->vba.BytePerPixelInDETC[k] - / 2.0 - * mode_lib->vba.MinSwathHeightC[k])); - if (mode_lib->vba.BytePerPixelInDETC[k] == 0.0) { - mode_lib->vba.MaximumSwathWidthInLineBuffer = - mode_lib->vba.LineBufferSize - * dml_max(mode_lib->vba.HRatio[k], 1.0) - / mode_lib->vba.LBBitPerPixel[k] - / (mode_lib->vba.vtaps[k] - + dml_max( - dml_ceil( - mode_lib->vba.VRatio[k], - 1.0) - - 2, - 0.0)); - } else { - mode_lib->vba.MaximumSwathWidthInLineBuffer = - dml_min( - mode_lib->vba.LineBufferSize - * dml_max( - mode_lib->vba.HRatio[k], - 1.0) - / mode_lib->vba.LBBitPerPixel[k] - / (mode_lib->vba.vtaps[k] - + dml_max( - dml_ceil( - mode_lib->vba.VRatio[k], - 1.0) - - 2, - 0.0)), - 2.0 * mode_lib->vba.LineBufferSize - * dml_max( - mode_lib->vba.HRatio[k] - / 2.0, - 1.0) - / mode_lib->vba.LBBitPerPixel[k] - / (mode_lib->vba.VTAPsChroma[k] - + dml_max( - dml_ceil( - mode_lib->vba.VRatio[k] - / 2.0, - 1.0) - - 2, - 0.0))); - } - mode_lib->vba.MaximumSwathWidth[k] = dml_min( - mode_lib->vba.MaximumSwathWidthInDETBuffer, - mode_lib->vba.MaximumSwathWidthInLineBuffer); - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity = RoundToDFSGranularityDown( - mode_lib->vba.MaxDispclk[i], - mode_lib->vba.DISPCLKDPPCLKVCOSpeed); - mode_lib->vba.MaxDppclkRoundedDownToDFSGranularity = RoundToDFSGranularityDown( - mode_lib->vba.MaxDppclk[i], - mode_lib->vba.DISPCLKDPPCLKVCOSpeed); - mode_lib->vba.RequiredDISPCLK[i] = 0.0; - mode_lib->vba.DISPCLK_DPPCLK_Support[i] = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine = - mode_lib->vba.PixelClock[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - * (1.0 - + mode_lib->vba.DISPCLKRampingMargin - / 100.0); - if (mode_lib->vba.ODMCapability == true - && mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine - > mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity) { - mode_lib->vba.ODMCombineEnablePerState[i][k] = true; - mode_lib->vba.PlaneRequiredDISPCLK = - mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine - / 2.0; - } else { - mode_lib->vba.ODMCombineEnablePerState[i][k] = false; - mode_lib->vba.PlaneRequiredDISPCLK = - mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine; - } - if (mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - <= mode_lib->vba.MaxDppclkRoundedDownToDFSGranularity - && mode_lib->vba.SwathWidthYSingleDPP[k] - <= mode_lib->vba.MaximumSwathWidth[k] - && mode_lib->vba.ODMCombineEnablePerState[i][k] == false) { - mode_lib->vba.NoOfDPP[i][k] = 1; - mode_lib->vba.RequiredDPPCLK[i][k] = - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0); - } else { - mode_lib->vba.NoOfDPP[i][k] = 2; - mode_lib->vba.RequiredDPPCLK[i][k] = - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - / 2.0; - } - mode_lib->vba.RequiredDISPCLK[i] = dml_max( - mode_lib->vba.RequiredDISPCLK[i], - mode_lib->vba.PlaneRequiredDISPCLK); - if ((mode_lib->vba.MinDPPCLKUsingSingleDPP[k] / mode_lib->vba.NoOfDPP[i][k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - > mode_lib->vba.MaxDppclkRoundedDownToDFSGranularity) - || (mode_lib->vba.PlaneRequiredDISPCLK - > mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity)) { - mode_lib->vba.DISPCLK_DPPCLK_Support[i] = false; - } - } - mode_lib->vba.TotalNumberOfActiveDPP[i] = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.TotalNumberOfActiveDPP[i] = - mode_lib->vba.TotalNumberOfActiveDPP[i] - + mode_lib->vba.NoOfDPP[i][k]; - } - if ((mode_lib->vba.MaxDispclk[i] == mode_lib->vba.MaxDispclk[DC__VOLTAGE_STATES] - && mode_lib->vba.MaxDppclk[i] - == mode_lib->vba.MaxDppclk[DC__VOLTAGE_STATES]) - && (mode_lib->vba.TotalNumberOfActiveDPP[i] - > mode_lib->vba.MaxNumDPP - || mode_lib->vba.DISPCLK_DPPCLK_Support[i] == false)) { - mode_lib->vba.RequiredDISPCLK[i] = 0.0; - mode_lib->vba.DISPCLK_DPPCLK_Support[i] = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine = - mode_lib->vba.PixelClock[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0); - if (mode_lib->vba.ODMCapability == true - && mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine - > mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity) { - mode_lib->vba.ODMCombineEnablePerState[i][k] = true; - mode_lib->vba.PlaneRequiredDISPCLK = - mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine - / 2.0; - } else { - mode_lib->vba.ODMCombineEnablePerState[i][k] = false; - mode_lib->vba.PlaneRequiredDISPCLK = - mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine; - } - if (mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - <= mode_lib->vba.MaxDppclkRoundedDownToDFSGranularity - && mode_lib->vba.SwathWidthYSingleDPP[k] - <= mode_lib->vba.MaximumSwathWidth[k] - && mode_lib->vba.ODMCombineEnablePerState[i][k] - == false) { - mode_lib->vba.NoOfDPP[i][k] = 1; - mode_lib->vba.RequiredDPPCLK[i][k] = - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0); - } else { - mode_lib->vba.NoOfDPP[i][k] = 2; - mode_lib->vba.RequiredDPPCLK[i][k] = - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - / 2.0; - } - mode_lib->vba.RequiredDISPCLK[i] = dml_max( - mode_lib->vba.RequiredDISPCLK[i], - mode_lib->vba.PlaneRequiredDISPCLK); - if ((mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - / mode_lib->vba.NoOfDPP[i][k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - > mode_lib->vba.MaxDppclkRoundedDownToDFSGranularity) - || (mode_lib->vba.PlaneRequiredDISPCLK - > mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity)) { - mode_lib->vba.DISPCLK_DPPCLK_Support[i] = false; - } - } - mode_lib->vba.TotalNumberOfActiveDPP[i] = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.TotalNumberOfActiveDPP[i] = - mode_lib->vba.TotalNumberOfActiveDPP[i] - + mode_lib->vba.NoOfDPP[i][k]; - } - } - if (mode_lib->vba.TotalNumberOfActiveDPP[i] > mode_lib->vba.MaxNumDPP) { - mode_lib->vba.RequiredDISPCLK[i] = 0.0; - mode_lib->vba.DISPCLK_DPPCLK_Support[i] = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.ODMCombineEnablePerState[i][k] = false; - if (mode_lib->vba.SwathWidthYSingleDPP[k] - <= mode_lib->vba.MaximumSwathWidth[k]) { - mode_lib->vba.NoOfDPP[i][k] = 1; - mode_lib->vba.RequiredDPPCLK[i][k] = - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0); - } else { - mode_lib->vba.NoOfDPP[i][k] = 2; - mode_lib->vba.RequiredDPPCLK[i][k] = - mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - / 2.0; - } - if (!(mode_lib->vba.MaxDispclk[i] - == mode_lib->vba.MaxDispclk[DC__VOLTAGE_STATES] - && mode_lib->vba.MaxDppclk[i] - == mode_lib->vba.MaxDppclk[DC__VOLTAGE_STATES])) { - mode_lib->vba.PlaneRequiredDISPCLK = - mode_lib->vba.PixelClock[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - * (1.0 - + mode_lib->vba.DISPCLKRampingMargin - / 100.0); - } else { - mode_lib->vba.PlaneRequiredDISPCLK = - mode_lib->vba.PixelClock[k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0); - } - mode_lib->vba.RequiredDISPCLK[i] = dml_max( - mode_lib->vba.RequiredDISPCLK[i], - mode_lib->vba.PlaneRequiredDISPCLK); - if ((mode_lib->vba.MinDPPCLKUsingSingleDPP[k] - / mode_lib->vba.NoOfDPP[i][k] - * (1.0 - + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - > mode_lib->vba.MaxDppclkRoundedDownToDFSGranularity) - || (mode_lib->vba.PlaneRequiredDISPCLK - > mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity)) { - mode_lib->vba.DISPCLK_DPPCLK_Support[i] = false; - } - } - mode_lib->vba.TotalNumberOfActiveDPP[i] = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.TotalNumberOfActiveDPP[i] = - mode_lib->vba.TotalNumberOfActiveDPP[i] - + mode_lib->vba.NoOfDPP[i][k]; - } - } - mode_lib->vba.RequiredDISPCLK[i] = dml_max( - mode_lib->vba.RequiredDISPCLK[i], - mode_lib->vba.WritebackRequiredDISPCLK); - if (mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity - < mode_lib->vba.WritebackRequiredDISPCLK) { - mode_lib->vba.DISPCLK_DPPCLK_Support[i] = false; - } - } - /*Viewport Size Check*/ - - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.ViewportSizeSupport[i] = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.ODMCombineEnablePerState[i][k] == true) { - if (dml_min(mode_lib->vba.SwathWidthYSingleDPP[k], dml_round(mode_lib->vba.HActive[k] / 2.0 * mode_lib->vba.HRatio[k])) - > mode_lib->vba.MaximumSwathWidth[k]) { - mode_lib->vba.ViewportSizeSupport[i] = false; - } - } else { - if (mode_lib->vba.SwathWidthYSingleDPP[k] / 2.0 - > mode_lib->vba.MaximumSwathWidth[k]) { - mode_lib->vba.ViewportSizeSupport[i] = false; - } - } - } - } - /*Total Available Pipes Support Check*/ - - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - if (mode_lib->vba.TotalNumberOfActiveDPP[i] <= mode_lib->vba.MaxNumDPP) { - mode_lib->vba.TotalAvailablePipesSupport[i] = true; - } else { - mode_lib->vba.TotalAvailablePipesSupport[i] = false; - } - } - /*Total Available OTG Support Check*/ - - mode_lib->vba.TotalNumberOfActiveOTG = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.BlendingAndTiming[k] == k) { - mode_lib->vba.TotalNumberOfActiveOTG = mode_lib->vba.TotalNumberOfActiveOTG - + 1.0; - } - } - if (mode_lib->vba.TotalNumberOfActiveOTG <= mode_lib->vba.MaxNumOTG) { - mode_lib->vba.NumberOfOTGSupport = true; - } else { - mode_lib->vba.NumberOfOTGSupport = false; - } - /*Display IO and DSC Support Check*/ - - mode_lib->vba.NonsupportedDSCInputBPC = false; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (!(mode_lib->vba.DSCInputBitPerComponent[k] == 12.0 - || mode_lib->vba.DSCInputBitPerComponent[k] == 10.0 - || mode_lib->vba.DSCInputBitPerComponent[k] == 8.0)) { - mode_lib->vba.NonsupportedDSCInputBPC = true; - } - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.RequiresDSC[i][k] = 0; - mode_lib->vba.RequiresFEC[i][k] = 0; - if (mode_lib->vba.BlendingAndTiming[k] == k) { - if (mode_lib->vba.Output[k] == dm_hdmi) { - mode_lib->vba.RequiresDSC[i][k] = 0; - mode_lib->vba.RequiresFEC[i][k] = 0; - mode_lib->vba.OutputBppPerState[i][k] = - TruncToValidBPP(dml_min(600.0, mode_lib->vba.PHYCLKPerState[i]) - / mode_lib->vba.PixelClockBackEnd[k] * 24, - false, - mode_lib->vba.Output[k], - mode_lib->vba.OutputFormat[k], - mode_lib->vba.DSCInputBitPerComponent[k]); - } else if (mode_lib->vba.Output[k] == dm_dp - || mode_lib->vba.Output[k] == dm_edp) { - if (mode_lib->vba.Output[k] == dm_edp) { - mode_lib->vba.EffectiveFECOverhead = 0.0; - } else { - mode_lib->vba.EffectiveFECOverhead = - mode_lib->vba.FECOverhead; - } - if (mode_lib->vba.PHYCLKPerState[i] >= 270.0) { - mode_lib->vba.Outbpp = - TruncToValidBPP((1.0 - mode_lib->vba.Downspreading / 100.0) * 270.0 - * mode_lib->vba.OutputLinkDPLanes[k] / mode_lib->vba.PixelClockBackEnd[k] * 8.0, - false, - mode_lib->vba.Output[k], - mode_lib->vba.OutputFormat[k], - mode_lib->vba.DSCInputBitPerComponent[k]); - mode_lib->vba.OutbppDSC = - TruncToValidBPP((1.0 - mode_lib->vba.Downspreading / 100.0) - * (1.0 - mode_lib->vba.EffectiveFECOverhead / 100.0) * 270.0 - * mode_lib->vba.OutputLinkDPLanes[k] / mode_lib->vba.PixelClockBackEnd[k] * 8.0, - true, - mode_lib->vba.Output[k], - mode_lib->vba.OutputFormat[k], - mode_lib->vba.DSCInputBitPerComponent[k]); - if (mode_lib->vba.DSCEnabled[k] == true) { - mode_lib->vba.RequiresDSC[i][k] = true; - if (mode_lib->vba.Output[k] == dm_dp) { - mode_lib->vba.RequiresFEC[i][k] = - true; - } else { - mode_lib->vba.RequiresFEC[i][k] = - false; - } - mode_lib->vba.Outbpp = - mode_lib->vba.OutbppDSC; - } else { - mode_lib->vba.RequiresDSC[i][k] = false; - mode_lib->vba.RequiresFEC[i][k] = false; - } - mode_lib->vba.OutputBppPerState[i][k] = - mode_lib->vba.Outbpp; - } - if (mode_lib->vba.Outbpp == BPP_INVALID) { - mode_lib->vba.Outbpp = - TruncToValidBPP((1.0 - mode_lib->vba.Downspreading / 100.0) * 540.0 - * mode_lib->vba.OutputLinkDPLanes[k] / mode_lib->vba.PixelClockBackEnd[k] * 8.0, - false, - mode_lib->vba.Output[k], - mode_lib->vba.OutputFormat[k], - mode_lib->vba.DSCInputBitPerComponent[k]); - mode_lib->vba.OutbppDSC = - TruncToValidBPP((1.0 - mode_lib->vba.Downspreading / 100.0) - * (1.0 - mode_lib->vba.EffectiveFECOverhead / 100.0) * 540.0 - * mode_lib->vba.OutputLinkDPLanes[k] / mode_lib->vba.PixelClockBackEnd[k] * 8.0, - true, - mode_lib->vba.Output[k], - mode_lib->vba.OutputFormat[k], - mode_lib->vba.DSCInputBitPerComponent[k]); - if (mode_lib->vba.DSCEnabled[k] == true) { - mode_lib->vba.RequiresDSC[i][k] = true; - if (mode_lib->vba.Output[k] == dm_dp) { - mode_lib->vba.RequiresFEC[i][k] = - true; - } else { - mode_lib->vba.RequiresFEC[i][k] = - false; - } - mode_lib->vba.Outbpp = - mode_lib->vba.OutbppDSC; - } else { - mode_lib->vba.RequiresDSC[i][k] = false; - mode_lib->vba.RequiresFEC[i][k] = false; - } - mode_lib->vba.OutputBppPerState[i][k] = - mode_lib->vba.Outbpp; - } - if (mode_lib->vba.Outbpp == BPP_INVALID - && mode_lib->vba.PHYCLKPerState[i] - >= 810.0) { - mode_lib->vba.Outbpp = - TruncToValidBPP((1.0 - mode_lib->vba.Downspreading / 100.0) * 810.0 - * mode_lib->vba.OutputLinkDPLanes[k] / mode_lib->vba.PixelClockBackEnd[k] * 8.0, - false, - mode_lib->vba.Output[k], - mode_lib->vba.OutputFormat[k], - mode_lib->vba.DSCInputBitPerComponent[k]); - mode_lib->vba.OutbppDSC = - TruncToValidBPP((1.0 - mode_lib->vba.Downspreading / 100.0) - * (1.0 - mode_lib->vba.EffectiveFECOverhead / 100.0) * 810.0 - * mode_lib->vba.OutputLinkDPLanes[k] / mode_lib->vba.PixelClockBackEnd[k] * 8.0, - true, - mode_lib->vba.Output[k], - mode_lib->vba.OutputFormat[k], - mode_lib->vba.DSCInputBitPerComponent[k]); - if (mode_lib->vba.DSCEnabled[k] == true - || mode_lib->vba.Outbpp == BPP_INVALID) { - mode_lib->vba.RequiresDSC[i][k] = true; - if (mode_lib->vba.Output[k] == dm_dp) { - mode_lib->vba.RequiresFEC[i][k] = - true; - } else { - mode_lib->vba.RequiresFEC[i][k] = - false; - } - mode_lib->vba.Outbpp = - mode_lib->vba.OutbppDSC; - } else { - mode_lib->vba.RequiresDSC[i][k] = false; - mode_lib->vba.RequiresFEC[i][k] = false; - } - mode_lib->vba.OutputBppPerState[i][k] = - mode_lib->vba.Outbpp; - } - } - } else { - mode_lib->vba.OutputBppPerState[i][k] = BPP_BLENDED_PIPE; - } - } - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.DIOSupport[i] = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.OutputBppPerState[i][k] == BPP_INVALID - || (mode_lib->vba.OutputFormat[k] == dm_420 - && mode_lib->vba.ProgressiveToInterlaceUnitInOPP - == true)) { - mode_lib->vba.DIOSupport[i] = false; - } - } - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.DSCCLKRequiredMoreThanSupported[i] = false; - if (mode_lib->vba.BlendingAndTiming[k] == k) { - if ((mode_lib->vba.Output[k] == dm_dp - || mode_lib->vba.Output[k] == dm_edp)) { - if (mode_lib->vba.OutputFormat[k] == dm_420 - || mode_lib->vba.OutputFormat[k] - == dm_n422) { - mode_lib->vba.DSCFormatFactor = 2; - } else { - mode_lib->vba.DSCFormatFactor = 1; - } - if (mode_lib->vba.RequiresDSC[i][k] == true) { - if (mode_lib->vba.ODMCombineEnablePerState[i][k] - == true) { - if (mode_lib->vba.PixelClockBackEnd[k] / 6.0 - / mode_lib->vba.DSCFormatFactor - > (1.0 - - mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - * mode_lib->vba.MaxDSCCLK[i]) { - mode_lib->vba.DSCCLKRequiredMoreThanSupported[i] = - true; - } - } else { - if (mode_lib->vba.PixelClockBackEnd[k] / 3.0 - / mode_lib->vba.DSCFormatFactor - > (1.0 - - mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading - / 100.0) - * mode_lib->vba.MaxDSCCLK[i]) { - mode_lib->vba.DSCCLKRequiredMoreThanSupported[i] = - true; - } - } - } - } - } - } - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.NotEnoughDSCUnits[i] = false; - mode_lib->vba.TotalDSCUnitsRequired = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.RequiresDSC[i][k] == true) { - if (mode_lib->vba.ODMCombineEnablePerState[i][k] == true) { - mode_lib->vba.TotalDSCUnitsRequired = - mode_lib->vba.TotalDSCUnitsRequired + 2.0; - } else { - mode_lib->vba.TotalDSCUnitsRequired = - mode_lib->vba.TotalDSCUnitsRequired + 1.0; - } - } - } - if (mode_lib->vba.TotalDSCUnitsRequired > mode_lib->vba.NumberOfDSC) { - mode_lib->vba.NotEnoughDSCUnits[i] = true; - } - } - /*DSC Delay per state*/ - - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.BlendingAndTiming[k] != k) { - mode_lib->vba.slices = 0; - } else if (mode_lib->vba.RequiresDSC[i][k] == 0 - || mode_lib->vba.RequiresDSC[i][k] == false) { - mode_lib->vba.slices = 0; - } else if (mode_lib->vba.PixelClockBackEnd[k] > 3200.0) { - mode_lib->vba.slices = dml_ceil( - mode_lib->vba.PixelClockBackEnd[k] / 400.0, - 4.0); - } else if (mode_lib->vba.PixelClockBackEnd[k] > 1360.0) { - mode_lib->vba.slices = 8.0; - } else if (mode_lib->vba.PixelClockBackEnd[k] > 680.0) { - mode_lib->vba.slices = 4.0; - } else if (mode_lib->vba.PixelClockBackEnd[k] > 340.0) { - mode_lib->vba.slices = 2.0; - } else { - mode_lib->vba.slices = 1.0; - } - if (mode_lib->vba.OutputBppPerState[i][k] == BPP_BLENDED_PIPE - || mode_lib->vba.OutputBppPerState[i][k] == BPP_INVALID) { - mode_lib->vba.bpp = 0.0; - } else { - mode_lib->vba.bpp = mode_lib->vba.OutputBppPerState[i][k]; - } - if (mode_lib->vba.RequiresDSC[i][k] == true && mode_lib->vba.bpp != 0.0) { - if (mode_lib->vba.ODMCombineEnablePerState[i][k] == false) { - mode_lib->vba.DSCDelayPerState[i][k] = - dscceComputeDelay( - mode_lib->vba.DSCInputBitPerComponent[k], - mode_lib->vba.bpp, - dml_ceil( - mode_lib->vba.HActive[k] - / mode_lib->vba.slices, - 1.0), - mode_lib->vba.slices, - mode_lib->vba.OutputFormat[k]) - + dscComputeDelay( - mode_lib->vba.OutputFormat[k]); - } else { - mode_lib->vba.DSCDelayPerState[i][k] = - 2.0 - * (dscceComputeDelay( - mode_lib->vba.DSCInputBitPerComponent[k], - mode_lib->vba.bpp, - dml_ceil( - mode_lib->vba.HActive[k] - / mode_lib->vba.slices, - 1.0), - mode_lib->vba.slices - / 2, - mode_lib->vba.OutputFormat[k]) - + dscComputeDelay( - mode_lib->vba.OutputFormat[k])); - } - mode_lib->vba.DSCDelayPerState[i][k] = - mode_lib->vba.DSCDelayPerState[i][k] - * mode_lib->vba.PixelClock[k] - / mode_lib->vba.PixelClockBackEnd[k]; - } else { - mode_lib->vba.DSCDelayPerState[i][k] = 0.0; - } - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - for (j = 0; j <= mode_lib->vba.NumberOfActivePlanes - 1; j++) { - if (mode_lib->vba.BlendingAndTiming[k] == j - && mode_lib->vba.RequiresDSC[i][j] == true) { - mode_lib->vba.DSCDelayPerState[i][k] = - mode_lib->vba.DSCDelayPerState[i][j]; - } - } - } - } - /*Urgent Latency Support Check*/ - - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - if (mode_lib->vba.ODMCombineEnablePerState[i][k] == true) { - mode_lib->vba.SwathWidthYPerState[i][k] = - dml_min( - mode_lib->vba.SwathWidthYSingleDPP[k], - dml_round( - mode_lib->vba.HActive[k] - / 2.0 - * mode_lib->vba.HRatio[k])); - } else { - mode_lib->vba.SwathWidthYPerState[i][k] = - mode_lib->vba.SwathWidthYSingleDPP[k] - / mode_lib->vba.NoOfDPP[i][k]; - } - mode_lib->vba.SwathWidthGranularityY = 256.0 - / dml_ceil(mode_lib->vba.BytePerPixelInDETY[k], 1.0) - / mode_lib->vba.MaxSwathHeightY[k]; - mode_lib->vba.RoundedUpMaxSwathSizeBytesY = (dml_ceil( - mode_lib->vba.SwathWidthYPerState[i][k] - 1.0, - mode_lib->vba.SwathWidthGranularityY) - + mode_lib->vba.SwathWidthGranularityY) - * mode_lib->vba.BytePerPixelInDETY[k] - * mode_lib->vba.MaxSwathHeightY[k]; - if (mode_lib->vba.SourcePixelFormat[k] == dm_420_10) { - mode_lib->vba.RoundedUpMaxSwathSizeBytesY = dml_ceil( - mode_lib->vba.RoundedUpMaxSwathSizeBytesY, - 256.0) + 256; - } - if (mode_lib->vba.MaxSwathHeightC[k] > 0.0) { - mode_lib->vba.SwathWidthGranularityC = 256.0 - / dml_ceil(mode_lib->vba.BytePerPixelInDETC[k], 2.0) - / mode_lib->vba.MaxSwathHeightC[k]; - mode_lib->vba.RoundedUpMaxSwathSizeBytesC = (dml_ceil( - mode_lib->vba.SwathWidthYPerState[i][k] / 2.0 - 1.0, - mode_lib->vba.SwathWidthGranularityC) - + mode_lib->vba.SwathWidthGranularityC) - * mode_lib->vba.BytePerPixelInDETC[k] - * mode_lib->vba.MaxSwathHeightC[k]; - if (mode_lib->vba.SourcePixelFormat[k] == dm_420_10) { - mode_lib->vba.RoundedUpMaxSwathSizeBytesC = dml_ceil( - mode_lib->vba.RoundedUpMaxSwathSizeBytesC, - 256.0) + 256; - } - } else { - mode_lib->vba.RoundedUpMaxSwathSizeBytesC = 0.0; - } - if (mode_lib->vba.RoundedUpMaxSwathSizeBytesY - + mode_lib->vba.RoundedUpMaxSwathSizeBytesC - <= mode_lib->vba.DETBufferSizeInKByte * 1024.0 / 2.0) { - mode_lib->vba.SwathHeightYPerState[i][k] = - mode_lib->vba.MaxSwathHeightY[k]; - mode_lib->vba.SwathHeightCPerState[i][k] = - mode_lib->vba.MaxSwathHeightC[k]; - } else { - mode_lib->vba.SwathHeightYPerState[i][k] = - mode_lib->vba.MinSwathHeightY[k]; - mode_lib->vba.SwathHeightCPerState[i][k] = - mode_lib->vba.MinSwathHeightC[k]; - } - if (mode_lib->vba.BytePerPixelInDETC[k] == 0.0) { - mode_lib->vba.LinesInDETLuma = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 / mode_lib->vba.BytePerPixelInDETY[k] - / mode_lib->vba.SwathWidthYPerState[i][k]; - mode_lib->vba.LinesInDETChroma = 0.0; - } else if (mode_lib->vba.SwathHeightYPerState[i][k] - <= mode_lib->vba.SwathHeightCPerState[i][k]) { - mode_lib->vba.LinesInDETLuma = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 / 2.0 / mode_lib->vba.BytePerPixelInDETY[k] - / mode_lib->vba.SwathWidthYPerState[i][k]; - mode_lib->vba.LinesInDETChroma = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 / 2.0 / mode_lib->vba.BytePerPixelInDETC[k] - / (mode_lib->vba.SwathWidthYPerState[i][k] / 2.0); - } else { - mode_lib->vba.LinesInDETLuma = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 * 2.0 / 3.0 - / mode_lib->vba.BytePerPixelInDETY[k] - / mode_lib->vba.SwathWidthYPerState[i][k]; - mode_lib->vba.LinesInDETChroma = mode_lib->vba.DETBufferSizeInKByte - * 1024.0 / 3.0 / mode_lib->vba.BytePerPixelInDETY[k] - / (mode_lib->vba.SwathWidthYPerState[i][k] / 2.0); - } - mode_lib->vba.EffectiveLBLatencyHidingSourceLinesLuma = - dml_min( - mode_lib->vba.MaxLineBufferLines, - dml_floor( - mode_lib->vba.LineBufferSize - / mode_lib->vba.LBBitPerPixel[k] - / (mode_lib->vba.SwathWidthYPerState[i][k] - / dml_max( - mode_lib->vba.HRatio[k], - 1.0)), - 1.0)) - - (mode_lib->vba.vtaps[k] - 1.0); - mode_lib->vba.EffectiveLBLatencyHidingSourceLinesChroma = - dml_min( - mode_lib->vba.MaxLineBufferLines, - dml_floor( - mode_lib->vba.LineBufferSize - / mode_lib->vba.LBBitPerPixel[k] - / (mode_lib->vba.SwathWidthYPerState[i][k] - / 2.0 - / dml_max( - mode_lib->vba.HRatio[k] - / 2.0, - 1.0)), - 1.0)) - - (mode_lib->vba.VTAPsChroma[k] - 1.0); - mode_lib->vba.EffectiveDETLBLinesLuma = - dml_floor( - mode_lib->vba.LinesInDETLuma - + dml_min( - mode_lib->vba.LinesInDETLuma - * mode_lib->vba.RequiredDISPCLK[i] - * mode_lib->vba.BytePerPixelInDETY[k] - * mode_lib->vba.PSCL_FACTOR[k] - / mode_lib->vba.ReturnBWPerState[i], - mode_lib->vba.EffectiveLBLatencyHidingSourceLinesLuma), - mode_lib->vba.SwathHeightYPerState[i][k]); - mode_lib->vba.EffectiveDETLBLinesChroma = - dml_floor( - mode_lib->vba.LinesInDETChroma - + dml_min( - mode_lib->vba.LinesInDETChroma - * mode_lib->vba.RequiredDISPCLK[i] - * mode_lib->vba.BytePerPixelInDETC[k] - * mode_lib->vba.PSCL_FACTOR_CHROMA[k] - / mode_lib->vba.ReturnBWPerState[i], - mode_lib->vba.EffectiveLBLatencyHidingSourceLinesChroma), - mode_lib->vba.SwathHeightCPerState[i][k]); - if (mode_lib->vba.BytePerPixelInDETC[k] == 0.0) { - mode_lib->vba.UrgentLatencySupportUsPerState[i][k] = - mode_lib->vba.EffectiveDETLBLinesLuma - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) - / mode_lib->vba.VRatio[k] - - mode_lib->vba.EffectiveDETLBLinesLuma - * mode_lib->vba.SwathWidthYPerState[i][k] - * dml_ceil( - mode_lib->vba.BytePerPixelInDETY[k], - 1.0) - / (mode_lib->vba.ReturnBWPerState[i] - / mode_lib->vba.NoOfDPP[i][k]); - } else { - mode_lib->vba.UrgentLatencySupportUsPerState[i][k] = - dml_min( - mode_lib->vba.EffectiveDETLBLinesLuma - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) - / mode_lib->vba.VRatio[k] - - mode_lib->vba.EffectiveDETLBLinesLuma - * mode_lib->vba.SwathWidthYPerState[i][k] - * dml_ceil( - mode_lib->vba.BytePerPixelInDETY[k], - 1.0) - / (mode_lib->vba.ReturnBWPerState[i] - / mode_lib->vba.NoOfDPP[i][k]), - mode_lib->vba.EffectiveDETLBLinesChroma - * (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]) - / (mode_lib->vba.VRatio[k] - / 2.0) - - mode_lib->vba.EffectiveDETLBLinesChroma - * mode_lib->vba.SwathWidthYPerState[i][k] - / 2.0 - * dml_ceil( - mode_lib->vba.BytePerPixelInDETC[k], - 2.0) - / (mode_lib->vba.ReturnBWPerState[i] - / mode_lib->vba.NoOfDPP[i][k])); - } - } - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.UrgentLatencySupport[i] = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.UrgentLatencySupportUsPerState[i][k] - < mode_lib->vba.UrgentLatency / 1.0) { - mode_lib->vba.UrgentLatencySupport[i] = false; - } - } - } - /*Prefetch Check*/ - - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.TotalNumberOfDCCActiveDPP[i] = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.DCCEnable[k] == true) { - mode_lib->vba.TotalNumberOfDCCActiveDPP[i] = - mode_lib->vba.TotalNumberOfDCCActiveDPP[i] - + mode_lib->vba.NoOfDPP[i][k]; - } - } - } - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.ProjectedDCFCLKDeepSleep = 8.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.ProjectedDCFCLKDeepSleep = dml_max( - mode_lib->vba.ProjectedDCFCLKDeepSleep, - mode_lib->vba.PixelClock[k] / 16.0); - if (mode_lib->vba.BytePerPixelInDETC[k] == 0.0) { - if (mode_lib->vba.VRatio[k] <= 1.0) { - mode_lib->vba.ProjectedDCFCLKDeepSleep = - dml_max( - mode_lib->vba.ProjectedDCFCLKDeepSleep, - 1.1 - * dml_ceil( - mode_lib->vba.BytePerPixelInDETY[k], - 1.0) - / 64.0 - * mode_lib->vba.HRatio[k] - * mode_lib->vba.PixelClock[k] - / mode_lib->vba.NoOfDPP[i][k]); - } else { - mode_lib->vba.ProjectedDCFCLKDeepSleep = - dml_max( - mode_lib->vba.ProjectedDCFCLKDeepSleep, - 1.1 - * dml_ceil( - mode_lib->vba.BytePerPixelInDETY[k], - 1.0) - / 64.0 - * mode_lib->vba.PSCL_FACTOR[k] - * mode_lib->vba.RequiredDPPCLK[i][k]); - } - } else { - if (mode_lib->vba.VRatio[k] <= 1.0) { - mode_lib->vba.ProjectedDCFCLKDeepSleep = - dml_max( - mode_lib->vba.ProjectedDCFCLKDeepSleep, - 1.1 - * dml_ceil( - mode_lib->vba.BytePerPixelInDETY[k], - 1.0) - / 32.0 - * mode_lib->vba.HRatio[k] - * mode_lib->vba.PixelClock[k] - / mode_lib->vba.NoOfDPP[i][k]); - } else { - mode_lib->vba.ProjectedDCFCLKDeepSleep = - dml_max( - mode_lib->vba.ProjectedDCFCLKDeepSleep, - 1.1 - * dml_ceil( - mode_lib->vba.BytePerPixelInDETY[k], - 1.0) - / 32.0 - * mode_lib->vba.PSCL_FACTOR[k] - * mode_lib->vba.RequiredDPPCLK[i][k]); - } - if (mode_lib->vba.VRatio[k] / 2.0 <= 1.0) { - mode_lib->vba.ProjectedDCFCLKDeepSleep = - dml_max( - mode_lib->vba.ProjectedDCFCLKDeepSleep, - 1.1 - * dml_ceil( - mode_lib->vba.BytePerPixelInDETC[k], - 2.0) - / 32.0 - * mode_lib->vba.HRatio[k] - / 2.0 - * mode_lib->vba.PixelClock[k] - / mode_lib->vba.NoOfDPP[i][k]); - } else { - mode_lib->vba.ProjectedDCFCLKDeepSleep = - dml_max( - mode_lib->vba.ProjectedDCFCLKDeepSleep, - 1.1 - * dml_ceil( - mode_lib->vba.BytePerPixelInDETC[k], - 2.0) - / 32.0 - * mode_lib->vba.PSCL_FACTOR_CHROMA[k] - * mode_lib->vba.RequiredDPPCLK[i][k]); - } - } - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.PDEAndMetaPTEBytesPerFrameY = CalculateVMAndRowBytes( - mode_lib, - mode_lib->vba.DCCEnable[k], - mode_lib->vba.Read256BlockHeightY[k], - mode_lib->vba.Read256BlockWidthY[k], - mode_lib->vba.SourcePixelFormat[k], - mode_lib->vba.SurfaceTiling[k], - dml_ceil(mode_lib->vba.BytePerPixelInDETY[k], 1.0), - mode_lib->vba.SourceScan[k], - mode_lib->vba.ViewportWidth[k], - mode_lib->vba.ViewportHeight[k], - mode_lib->vba.SwathWidthYPerState[i][k], - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.VMMPageSize, - mode_lib->vba.PTEBufferSizeInRequests, - mode_lib->vba.PDEProcessingBufIn64KBReqs, - mode_lib->vba.PitchY[k], - mode_lib->vba.DCCMetaPitchY[k], - &mode_lib->vba.MacroTileWidthY[k], - &mode_lib->vba.MetaRowBytesY, - &mode_lib->vba.DPTEBytesPerRowY, - &mode_lib->vba.PTEBufferSizeNotExceededY[i][k], - &mode_lib->vba.dpte_row_height[k], - &mode_lib->vba.meta_row_height[k]); - mode_lib->vba.PrefetchLinesY[k] = CalculatePrefetchSourceLines( - mode_lib, - mode_lib->vba.VRatio[k], - mode_lib->vba.vtaps[k], - mode_lib->vba.Interlace[k], - mode_lib->vba.ProgressiveToInterlaceUnitInOPP, - mode_lib->vba.SwathHeightYPerState[i][k], - mode_lib->vba.ViewportYStartY[k], - &mode_lib->vba.PrefillY[k], - &mode_lib->vba.MaxNumSwY[k]); - if ((mode_lib->vba.SourcePixelFormat[k] != dm_444_64 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_32 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_mono_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_mono_8)) { - mode_lib->vba.PDEAndMetaPTEBytesPerFrameC = CalculateVMAndRowBytes( - mode_lib, - mode_lib->vba.DCCEnable[k], - mode_lib->vba.Read256BlockHeightY[k], - mode_lib->vba.Read256BlockWidthY[k], - mode_lib->vba.SourcePixelFormat[k], - mode_lib->vba.SurfaceTiling[k], - dml_ceil(mode_lib->vba.BytePerPixelInDETC[k], 2.0), - mode_lib->vba.SourceScan[k], - mode_lib->vba.ViewportWidth[k] / 2.0, - mode_lib->vba.ViewportHeight[k] / 2.0, - mode_lib->vba.SwathWidthYPerState[i][k] / 2.0, - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.VMMPageSize, - mode_lib->vba.PTEBufferSizeInRequests, - mode_lib->vba.PDEProcessingBufIn64KBReqs, - mode_lib->vba.PitchC[k], - 0.0, - &mode_lib->vba.MacroTileWidthC[k], - &mode_lib->vba.MetaRowBytesC, - &mode_lib->vba.DPTEBytesPerRowC, - &mode_lib->vba.PTEBufferSizeNotExceededC[i][k], - &mode_lib->vba.dpte_row_height_chroma[k], - &mode_lib->vba.meta_row_height_chroma[k]); - mode_lib->vba.PrefetchLinesC[k] = CalculatePrefetchSourceLines( - mode_lib, - mode_lib->vba.VRatio[k] / 2.0, - mode_lib->vba.VTAPsChroma[k], - mode_lib->vba.Interlace[k], - mode_lib->vba.ProgressiveToInterlaceUnitInOPP, - mode_lib->vba.SwathHeightCPerState[i][k], - mode_lib->vba.ViewportYStartC[k], - &mode_lib->vba.PrefillC[k], - &mode_lib->vba.MaxNumSwC[k]); - } else { - mode_lib->vba.PDEAndMetaPTEBytesPerFrameC = 0.0; - mode_lib->vba.MetaRowBytesC = 0.0; - mode_lib->vba.DPTEBytesPerRowC = 0.0; - mode_lib->vba.PrefetchLinesC[k] = 0.0; - mode_lib->vba.PTEBufferSizeNotExceededC[i][k] = true; - } - mode_lib->vba.PDEAndMetaPTEBytesPerFrame[k] = - mode_lib->vba.PDEAndMetaPTEBytesPerFrameY - + mode_lib->vba.PDEAndMetaPTEBytesPerFrameC; - mode_lib->vba.MetaRowBytes[k] = mode_lib->vba.MetaRowBytesY - + mode_lib->vba.MetaRowBytesC; - mode_lib->vba.DPTEBytesPerRow[k] = mode_lib->vba.DPTEBytesPerRowY - + mode_lib->vba.DPTEBytesPerRowC; - } - mode_lib->vba.ExtraLatency = - mode_lib->vba.UrgentRoundTripAndOutOfOrderLatencyPerState[i] - + (mode_lib->vba.TotalNumberOfActiveDPP[i] - * mode_lib->vba.PixelChunkSizeInKByte - + mode_lib->vba.TotalNumberOfDCCActiveDPP[i] - * mode_lib->vba.MetaChunkSize) - * 1024.0 - / mode_lib->vba.ReturnBWPerState[i]; - if (mode_lib->vba.VirtualMemoryEnable == true) { - mode_lib->vba.ExtraLatency = mode_lib->vba.ExtraLatency - + mode_lib->vba.TotalNumberOfActiveDPP[i] - * mode_lib->vba.PTEChunkSize * 1024.0 - / mode_lib->vba.ReturnBWPerState[i]; - } - mode_lib->vba.TimeCalc = 24.0 / mode_lib->vba.ProjectedDCFCLKDeepSleep; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.BlendingAndTiming[k] == k) { - if (mode_lib->vba.WritebackEnable[k] == true) { - mode_lib->vba.WritebackDelay[i][k] = - mode_lib->vba.WritebackLatency - + CalculateWriteBackDelay( - mode_lib->vba.WritebackPixelFormat[k], - mode_lib->vba.WritebackHRatio[k], - mode_lib->vba.WritebackVRatio[k], - mode_lib->vba.WritebackLumaHTaps[k], - mode_lib->vba.WritebackLumaVTaps[k], - mode_lib->vba.WritebackChromaHTaps[k], - mode_lib->vba.WritebackChromaVTaps[k], - mode_lib->vba.WritebackDestinationWidth[k]) - / mode_lib->vba.RequiredDISPCLK[i]; - } else { - mode_lib->vba.WritebackDelay[i][k] = 0.0; - } - for (j = 0; j <= mode_lib->vba.NumberOfActivePlanes - 1; j++) { - if (mode_lib->vba.BlendingAndTiming[j] == k - && mode_lib->vba.WritebackEnable[j] - == true) { - mode_lib->vba.WritebackDelay[i][k] = - dml_max( - mode_lib->vba.WritebackDelay[i][k], - mode_lib->vba.WritebackLatency - + CalculateWriteBackDelay( - mode_lib->vba.WritebackPixelFormat[j], - mode_lib->vba.WritebackHRatio[j], - mode_lib->vba.WritebackVRatio[j], - mode_lib->vba.WritebackLumaHTaps[j], - mode_lib->vba.WritebackLumaVTaps[j], - mode_lib->vba.WritebackChromaHTaps[j], - mode_lib->vba.WritebackChromaVTaps[j], - mode_lib->vba.WritebackDestinationWidth[j]) - / mode_lib->vba.RequiredDISPCLK[i]); - } - } - } - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - for (j = 0; j <= mode_lib->vba.NumberOfActivePlanes - 1; j++) { - if (mode_lib->vba.BlendingAndTiming[k] == j) { - mode_lib->vba.WritebackDelay[i][k] = - mode_lib->vba.WritebackDelay[i][j]; - } - } - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.MaximumVStartup[k] = - mode_lib->vba.VTotal[k] - mode_lib->vba.VActive[k] - - dml_max( - 1.0, - dml_ceil( - mode_lib->vba.WritebackDelay[i][k] - / (mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]), - 1.0)); - } - mode_lib->vba.TWait = CalculateTWait( - mode_lib->vba.PrefetchMode, - mode_lib->vba.DRAMClockChangeLatency, - mode_lib->vba.UrgentLatency, - mode_lib->vba.SREnterPlusExitTime); - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.XFCEnabled[k] == true) { - mode_lib->vba.XFCRemoteSurfaceFlipDelay = - CalculateRemoteSurfaceFlipDelay( - mode_lib, - mode_lib->vba.VRatio[k], - mode_lib->vba.SwathWidthYPerState[i][k], - dml_ceil( - mode_lib->vba.BytePerPixelInDETY[k], - 1.0), - mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k], - mode_lib->vba.XFCTSlvVupdateOffset, - mode_lib->vba.XFCTSlvVupdateWidth, - mode_lib->vba.XFCTSlvVreadyOffset, - mode_lib->vba.XFCXBUFLatencyTolerance, - mode_lib->vba.XFCFillBWOverhead, - mode_lib->vba.XFCSlvChunkSize, - mode_lib->vba.XFCBusTransportTime, - mode_lib->vba.TimeCalc, - mode_lib->vba.TWait, - &mode_lib->vba.SrcActiveDrainRate, - &mode_lib->vba.TInitXFill, - &mode_lib->vba.TslvChk); - } else { - mode_lib->vba.XFCRemoteSurfaceFlipDelay = 0.0; - } - mode_lib->vba.IsErrorResult[i][k] = - CalculatePrefetchSchedule( - mode_lib, - mode_lib->vba.RequiredDPPCLK[i][k], - mode_lib->vba.RequiredDISPCLK[i], - mode_lib->vba.PixelClock[k], - mode_lib->vba.ProjectedDCFCLKDeepSleep, - mode_lib->vba.DSCDelayPerState[i][k], - mode_lib->vba.NoOfDPP[i][k], - mode_lib->vba.ScalerEnabled[k], - mode_lib->vba.NumberOfCursors[k], - mode_lib->vba.DPPCLKDelaySubtotal, - mode_lib->vba.DPPCLKDelaySCL, - mode_lib->vba.DPPCLKDelaySCLLBOnly, - mode_lib->vba.DPPCLKDelayCNVCFormater, - mode_lib->vba.DPPCLKDelayCNVCCursor, - mode_lib->vba.DISPCLKDelaySubtotal, - mode_lib->vba.SwathWidthYPerState[i][k] - / mode_lib->vba.HRatio[k], - mode_lib->vba.OutputFormat[k], - mode_lib->vba.VTotal[k] - - mode_lib->vba.VActive[k], - mode_lib->vba.HTotal[k], - mode_lib->vba.MaxInterDCNTileRepeaters, - mode_lib->vba.MaximumVStartup[k], - mode_lib->vba.MaxPageTableLevels, - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.DynamicMetadataEnable[k], - mode_lib->vba.DynamicMetadataLinesBeforeActiveRequired[k], - mode_lib->vba.DynamicMetadataTransmittedBytes[k], - mode_lib->vba.DCCEnable[k], - mode_lib->vba.UrgentLatency, - mode_lib->vba.ExtraLatency, - mode_lib->vba.TimeCalc, - mode_lib->vba.PDEAndMetaPTEBytesPerFrame[k], - mode_lib->vba.MetaRowBytes[k], - mode_lib->vba.DPTEBytesPerRow[k], - mode_lib->vba.PrefetchLinesY[k], - mode_lib->vba.SwathWidthYPerState[i][k], - mode_lib->vba.BytePerPixelInDETY[k], - mode_lib->vba.PrefillY[k], - mode_lib->vba.MaxNumSwY[k], - mode_lib->vba.PrefetchLinesC[k], - mode_lib->vba.BytePerPixelInDETC[k], - mode_lib->vba.PrefillC[k], - mode_lib->vba.MaxNumSwC[k], - mode_lib->vba.SwathHeightYPerState[i][k], - mode_lib->vba.SwathHeightCPerState[i][k], - mode_lib->vba.TWait, - mode_lib->vba.XFCEnabled[k], - mode_lib->vba.XFCRemoteSurfaceFlipDelay, - mode_lib->vba.Interlace[k], - mode_lib->vba.ProgressiveToInterlaceUnitInOPP, - mode_lib->vba.DSTXAfterScaler, - mode_lib->vba.DSTYAfterScaler, - &mode_lib->vba.LineTimesForPrefetch[k], - &mode_lib->vba.PrefetchBW[k], - &mode_lib->vba.LinesForMetaPTE[k], - &mode_lib->vba.LinesForMetaAndDPTERow[k], - &mode_lib->vba.VRatioPreY[i][k], - &mode_lib->vba.VRatioPreC[i][k], - &mode_lib->vba.RequiredPrefetchPixelDataBW[i][k], - &mode_lib->vba.VStartupRequiredWhenNotEnoughTimeForDynamicMetadata, - &mode_lib->vba.Tno_bw[k], - &mode_lib->vba.VUpdateOffsetPix[k], - &mode_lib->vba.VUpdateWidthPix[k], - &mode_lib->vba.VReadyOffsetPix[k]); - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.cursor_bw[k] = mode_lib->vba.NumberOfCursors[k] - * mode_lib->vba.CursorWidth[k][0] - * mode_lib->vba.CursorBPP[k][0] / 8.0 - / (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]) - * mode_lib->vba.VRatio[k]; - } - mode_lib->vba.MaximumReadBandwidthWithPrefetch = 0.0; - mode_lib->vba.prefetch_vm_bw_valid = true; - mode_lib->vba.prefetch_row_bw_valid = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.PDEAndMetaPTEBytesPerFrame[k] == 0.0) { - mode_lib->vba.prefetch_vm_bw[k] = 0.0; - } else if (mode_lib->vba.LinesForMetaPTE[k] > 0.0) { - mode_lib->vba.prefetch_vm_bw[k] = - mode_lib->vba.PDEAndMetaPTEBytesPerFrame[k] - / (mode_lib->vba.LinesForMetaPTE[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]); - } else { - mode_lib->vba.prefetch_vm_bw[k] = 0.0; - mode_lib->vba.prefetch_vm_bw_valid = false; - } - if (mode_lib->vba.MetaRowBytes[k] + mode_lib->vba.DPTEBytesPerRow[k] - == 0.0) { - mode_lib->vba.prefetch_row_bw[k] = 0.0; - } else if (mode_lib->vba.LinesForMetaAndDPTERow[k] > 0.0) { - mode_lib->vba.prefetch_row_bw[k] = (mode_lib->vba.MetaRowBytes[k] - + mode_lib->vba.DPTEBytesPerRow[k]) - / (mode_lib->vba.LinesForMetaAndDPTERow[k] - * mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k]); - } else { - mode_lib->vba.prefetch_row_bw[k] = 0.0; - mode_lib->vba.prefetch_row_bw_valid = false; - } - mode_lib->vba.MaximumReadBandwidthWithPrefetch = - mode_lib->vba.MaximumReadBandwidthWithPrefetch - + mode_lib->vba.cursor_bw[k] - + dml_max4( - mode_lib->vba.prefetch_vm_bw[k], - mode_lib->vba.prefetch_row_bw[k], - mode_lib->vba.ReadBandwidth[k], - mode_lib->vba.RequiredPrefetchPixelDataBW[i][k]); - } - mode_lib->vba.PrefetchSupported[i] = true; - if (mode_lib->vba.MaximumReadBandwidthWithPrefetch - > mode_lib->vba.ReturnBWPerState[i] - || mode_lib->vba.prefetch_vm_bw_valid == false - || mode_lib->vba.prefetch_row_bw_valid == false) { - mode_lib->vba.PrefetchSupported[i] = false; - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.LineTimesForPrefetch[k] < 2.0 - || mode_lib->vba.LinesForMetaPTE[k] >= 8.0 - || mode_lib->vba.LinesForMetaAndDPTERow[k] >= 16.0 - || mode_lib->vba.IsErrorResult[i][k] == true) { - mode_lib->vba.PrefetchSupported[i] = false; - } - } - mode_lib->vba.VRatioInPrefetchSupported[i] = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.VRatioPreY[i][k] > 4.0 - || mode_lib->vba.VRatioPreC[i][k] > 4.0 - || mode_lib->vba.IsErrorResult[i][k] == true) { - mode_lib->vba.VRatioInPrefetchSupported[i] = false; - } - } - if (mode_lib->vba.PrefetchSupported[i] == true - && mode_lib->vba.VRatioInPrefetchSupported[i] == true) { - mode_lib->vba.BandwidthAvailableForImmediateFlip = - mode_lib->vba.ReturnBWPerState[i]; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.BandwidthAvailableForImmediateFlip = - mode_lib->vba.BandwidthAvailableForImmediateFlip - - mode_lib->vba.cursor_bw[k] - - dml_max( - mode_lib->vba.ReadBandwidth[k], - mode_lib->vba.PrefetchBW[k]); - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.ImmediateFlipBytes[k] = 0.0; - if ((mode_lib->vba.SourcePixelFormat[k] != dm_420_8 - && mode_lib->vba.SourcePixelFormat[k] != dm_420_10)) { - mode_lib->vba.ImmediateFlipBytes[k] = - mode_lib->vba.PDEAndMetaPTEBytesPerFrame[k] - + mode_lib->vba.MetaRowBytes[k] - + mode_lib->vba.DPTEBytesPerRow[k]; - } - } - mode_lib->vba.TotImmediateFlipBytes = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if ((mode_lib->vba.SourcePixelFormat[k] != dm_420_8 - && mode_lib->vba.SourcePixelFormat[k] != dm_420_10)) { - mode_lib->vba.TotImmediateFlipBytes = - mode_lib->vba.TotImmediateFlipBytes - + mode_lib->vba.ImmediateFlipBytes[k]; - } - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - CalculateFlipSchedule( - mode_lib, - mode_lib->vba.ExtraLatency, - mode_lib->vba.UrgentLatency, - mode_lib->vba.MaxPageTableLevels, - mode_lib->vba.VirtualMemoryEnable, - mode_lib->vba.BandwidthAvailableForImmediateFlip, - mode_lib->vba.TotImmediateFlipBytes, - mode_lib->vba.SourcePixelFormat[k], - mode_lib->vba.ImmediateFlipBytes[k], - mode_lib->vba.HTotal[k] - / mode_lib->vba.PixelClock[k], - mode_lib->vba.VRatio[k], - mode_lib->vba.Tno_bw[k], - mode_lib->vba.PDEAndMetaPTEBytesPerFrame[k], - mode_lib->vba.MetaRowBytes[k], - mode_lib->vba.DPTEBytesPerRow[k], - mode_lib->vba.DCCEnable[k], - mode_lib->vba.dpte_row_height[k], - mode_lib->vba.meta_row_height[k], - mode_lib->vba.qual_row_bw[k], - &mode_lib->vba.DestinationLinesToRequestVMInImmediateFlip[k], - &mode_lib->vba.DestinationLinesToRequestRowInImmediateFlip[k], - &mode_lib->vba.final_flip_bw[k], - &mode_lib->vba.ImmediateFlipSupportedForPipe[k]); - } - mode_lib->vba.total_dcn_read_bw_with_flip = 0.0; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.total_dcn_read_bw_with_flip = - mode_lib->vba.total_dcn_read_bw_with_flip - + mode_lib->vba.cursor_bw[k] - + dml_max3( - mode_lib->vba.prefetch_vm_bw[k], - mode_lib->vba.prefetch_row_bw[k], - mode_lib->vba.final_flip_bw[k] - + dml_max( - mode_lib->vba.ReadBandwidth[k], - mode_lib->vba.RequiredPrefetchPixelDataBW[i][k])); - } - mode_lib->vba.ImmediateFlipSupportedForState[i] = true; - if (mode_lib->vba.total_dcn_read_bw_with_flip - > mode_lib->vba.ReturnBWPerState[i]) { - mode_lib->vba.ImmediateFlipSupportedForState[i] = false; - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.ImmediateFlipSupportedForPipe[k] == false) { - mode_lib->vba.ImmediateFlipSupportedForState[i] = false; - } - } - } else { - mode_lib->vba.ImmediateFlipSupportedForState[i] = false; - } - } - /*PTE Buffer Size Check*/ - - for (i = 0; i <= DC__VOLTAGE_STATES; i++) { - mode_lib->vba.PTEBufferSizeNotExceeded[i] = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.PTEBufferSizeNotExceededY[i][k] == false - || mode_lib->vba.PTEBufferSizeNotExceededC[i][k] == false) { - mode_lib->vba.PTEBufferSizeNotExceeded[i] = false; - } - } - } - /*Cursor Support Check*/ - - mode_lib->vba.CursorSupport = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.CursorWidth[k][0] > 0.0) { - if (dml_floor( - dml_floor( - mode_lib->vba.CursorBufferSize - - mode_lib->vba.CursorChunkSize, - mode_lib->vba.CursorChunkSize) * 1024.0 - / (mode_lib->vba.CursorWidth[k][0] - * mode_lib->vba.CursorBPP[k][0] - / 8.0), - 1.0) - * (mode_lib->vba.HTotal[k] / mode_lib->vba.PixelClock[k]) - / mode_lib->vba.VRatio[k] < mode_lib->vba.UrgentLatency - || (mode_lib->vba.CursorBPP[k][0] == 64.0 - && mode_lib->vba.Cursor64BppSupport == false)) { - mode_lib->vba.CursorSupport = false; - } - } - } - /*Valid Pitch Check*/ - - mode_lib->vba.PitchSupport = true; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.AlignedYPitch[k] = dml_ceil( - dml_max(mode_lib->vba.PitchY[k], mode_lib->vba.ViewportWidth[k]), - mode_lib->vba.MacroTileWidthY[k]); - if (mode_lib->vba.AlignedYPitch[k] > mode_lib->vba.PitchY[k]) { - mode_lib->vba.PitchSupport = false; - } - if (mode_lib->vba.DCCEnable[k] == true) { - mode_lib->vba.AlignedDCCMetaPitch[k] = dml_ceil( - dml_max( - mode_lib->vba.DCCMetaPitchY[k], - mode_lib->vba.ViewportWidth[k]), - 64.0 * mode_lib->vba.Read256BlockWidthY[k]); - } else { - mode_lib->vba.AlignedDCCMetaPitch[k] = mode_lib->vba.DCCMetaPitchY[k]; - } - if (mode_lib->vba.AlignedDCCMetaPitch[k] > mode_lib->vba.DCCMetaPitchY[k]) { - mode_lib->vba.PitchSupport = false; - } - if (mode_lib->vba.SourcePixelFormat[k] != dm_444_64 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_32 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_mono_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_mono_8) { - mode_lib->vba.AlignedCPitch[k] = dml_ceil( - dml_max( - mode_lib->vba.PitchC[k], - mode_lib->vba.ViewportWidth[k] / 2.0), - mode_lib->vba.MacroTileWidthC[k]); - } else { - mode_lib->vba.AlignedCPitch[k] = mode_lib->vba.PitchC[k]; - } - if (mode_lib->vba.AlignedCPitch[k] > mode_lib->vba.PitchC[k]) { - mode_lib->vba.PitchSupport = false; - } - } - /*Mode Support, Voltage State and SOC Configuration*/ - - for (i = DC__VOLTAGE_STATES; i >= 0; i--) { - if (mode_lib->vba.ScaleRatioAndTapsSupport == true - && mode_lib->vba.SourceFormatPixelAndScanSupport == true - && mode_lib->vba.ViewportSizeSupport[i] == true - && mode_lib->vba.BandwidthSupport[i] == true - && mode_lib->vba.DIOSupport[i] == true - && mode_lib->vba.NotEnoughDSCUnits[i] == false - && mode_lib->vba.DSCCLKRequiredMoreThanSupported[i] == false - && mode_lib->vba.UrgentLatencySupport[i] == true - && mode_lib->vba.ROBSupport[i] == true - && mode_lib->vba.DISPCLK_DPPCLK_Support[i] == true - && mode_lib->vba.TotalAvailablePipesSupport[i] == true - && mode_lib->vba.NumberOfOTGSupport == true - && mode_lib->vba.WritebackModeSupport == true - && mode_lib->vba.WritebackLatencySupport == true - && mode_lib->vba.WritebackScaleRatioAndTapsSupport == true - && mode_lib->vba.CursorSupport == true - && mode_lib->vba.PitchSupport == true - && mode_lib->vba.PrefetchSupported[i] == true - && mode_lib->vba.VRatioInPrefetchSupported[i] == true - && mode_lib->vba.PTEBufferSizeNotExceeded[i] == true - && mode_lib->vba.NonsupportedDSCInputBPC == false) { - mode_lib->vba.ModeSupport[i] = true; - } else { - mode_lib->vba.ModeSupport[i] = false; - } - } - for (i = DC__VOLTAGE_STATES; i >= 0; i--) { - if (i == DC__VOLTAGE_STATES || mode_lib->vba.ModeSupport[i] == true) { - mode_lib->vba.VoltageLevel = i; - } - } - mode_lib->vba.DCFCLK = mode_lib->vba.DCFCLKPerState[mode_lib->vba.VoltageLevel]; - mode_lib->vba.DRAMSpeed = mode_lib->vba.DRAMSpeedPerState[mode_lib->vba.VoltageLevel]; - mode_lib->vba.FabricClock = mode_lib->vba.FabricClockPerState[mode_lib->vba.VoltageLevel]; - mode_lib->vba.SOCCLK = mode_lib->vba.SOCCLKPerState[mode_lib->vba.VoltageLevel]; - mode_lib->vba.FabricAndDRAMBandwidth = - mode_lib->vba.FabricAndDRAMBandwidthPerState[mode_lib->vba.VoltageLevel]; - mode_lib->vba.ImmediateFlipSupport = - mode_lib->vba.ImmediateFlipSupportedForState[mode_lib->vba.VoltageLevel]; - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - mode_lib->vba.DPPPerPlane[k] = mode_lib->vba.NoOfDPP[mode_lib->vba.VoltageLevel][k]; - } - for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - if (mode_lib->vba.BlendingAndTiming[k] == k) { - mode_lib->vba.ODMCombineEnabled[k] = - mode_lib->vba.ODMCombineEnablePerState[mode_lib->vba.VoltageLevel][k]; - } else { - mode_lib->vba.ODMCombineEnabled[k] = 0; - } - mode_lib->vba.DSCEnabled[k] = - mode_lib->vba.RequiresDSC[mode_lib->vba.VoltageLevel][k]; - mode_lib->vba.OutputBpp[k] = - mode_lib->vba.OutputBppPerState[mode_lib->vba.VoltageLevel][k]; - } -}
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h deleted file mode 100644 index 4112409..0000000 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h +++ /dev/null
@@ -1,598 +0,0 @@ -/* - * Copyright 2017 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ - -#ifndef __DML2_DISPLAY_MODE_VBA_H__ -#define __DML2_DISPLAY_MODE_VBA_H__ - -#include "dml_common_defs.h" - -struct display_mode_lib; - -void set_prefetch_mode(struct display_mode_lib *mode_lib, - bool cstate_en, - bool pstate_en, - bool ignore_viewport_pos, - bool immediate_flip_support); - -#define dml_get_attr_decl(attr) double get_##attr(struct display_mode_lib *mode_lib, const display_e2e_pipe_params_st *pipes, unsigned int num_pipes) - -dml_get_attr_decl(clk_dcf_deepsleep); -dml_get_attr_decl(wm_urgent); -dml_get_attr_decl(wm_memory_trip); -dml_get_attr_decl(wm_writeback_urgent); -dml_get_attr_decl(wm_stutter_exit); -dml_get_attr_decl(wm_stutter_enter_exit); -dml_get_attr_decl(wm_dram_clock_change); -dml_get_attr_decl(wm_writeback_dram_clock_change); -dml_get_attr_decl(wm_xfc_underflow); -dml_get_attr_decl(stutter_efficiency_no_vblank); -dml_get_attr_decl(stutter_efficiency); -dml_get_attr_decl(urgent_latency); -dml_get_attr_decl(urgent_extra_latency); -dml_get_attr_decl(nonurgent_latency); -dml_get_attr_decl(dram_clock_change_latency); -dml_get_attr_decl(dispclk_calculated); -dml_get_attr_decl(total_data_read_bw); -dml_get_attr_decl(return_bw); -dml_get_attr_decl(tcalc); - -#define dml_get_pipe_attr_decl(attr) double get_##attr(struct display_mode_lib *mode_lib, const display_e2e_pipe_params_st *pipes, unsigned int num_pipes, unsigned int which_pipe) - -dml_get_pipe_attr_decl(dsc_delay); -dml_get_pipe_attr_decl(dppclk_calculated); -dml_get_pipe_attr_decl(dscclk_calculated); -dml_get_pipe_attr_decl(min_ttu_vblank); -dml_get_pipe_attr_decl(vratio_prefetch_l); -dml_get_pipe_attr_decl(vratio_prefetch_c); -dml_get_pipe_attr_decl(dst_x_after_scaler); -dml_get_pipe_attr_decl(dst_y_after_scaler); -dml_get_pipe_attr_decl(dst_y_per_vm_vblank); -dml_get_pipe_attr_decl(dst_y_per_row_vblank); -dml_get_pipe_attr_decl(dst_y_prefetch); -dml_get_pipe_attr_decl(dst_y_per_vm_flip); -dml_get_pipe_attr_decl(dst_y_per_row_flip); -dml_get_pipe_attr_decl(xfc_transfer_delay); -dml_get_pipe_attr_decl(xfc_precharge_delay); -dml_get_pipe_attr_decl(xfc_remote_surface_flip_latency); -dml_get_pipe_attr_decl(xfc_prefetch_margin); - -unsigned int get_vstartup_calculated( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes, - unsigned int which_pipe); - -double get_total_immediate_flip_bytes( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes); -double get_total_immediate_flip_bw( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes); -double get_total_prefetch_bw( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes); - -unsigned int dml_get_voltage_level( - struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *pipes, - unsigned int num_pipes); - -bool Calculate256BBlockSizes( - enum source_format_class SourcePixelFormat, - enum dm_swizzle_mode SurfaceTiling, - unsigned int BytePerPixelY, - unsigned int BytePerPixelC, - unsigned int *BlockHeight256BytesY, - unsigned int *BlockHeight256BytesC, - unsigned int *BlockWidth256BytesY, - unsigned int *BlockWidth256BytesC); - - -struct vba_vars_st { - ip_params_st ip; - soc_bounding_box_st soc; - - unsigned int MaximumMaxVStartupLines; - double cursor_bw[DC__NUM_DPP__MAX]; - double meta_row_bw[DC__NUM_DPP__MAX]; - double dpte_row_bw[DC__NUM_DPP__MAX]; - double qual_row_bw[DC__NUM_DPP__MAX]; - double WritebackDISPCLK; - double PSCL_THROUGHPUT_LUMA[DC__NUM_DPP__MAX]; - double PSCL_THROUGHPUT_CHROMA[DC__NUM_DPP__MAX]; - double DPPCLKUsingSingleDPPLuma; - double DPPCLKUsingSingleDPPChroma; - double DPPCLKUsingSingleDPP[DC__NUM_DPP__MAX]; - double DISPCLKWithRamping; - double DISPCLKWithoutRamping; - double GlobalDPPCLK; - double DISPCLKWithRampingRoundedToDFSGranularity; - double DISPCLKWithoutRampingRoundedToDFSGranularity; - double MaxDispclkRoundedToDFSGranularity; - bool DCCEnabledAnyPlane; - double ReturnBandwidthToDCN; - unsigned int SwathWidthY[DC__NUM_DPP__MAX]; - unsigned int SwathWidthSingleDPPY[DC__NUM_DPP__MAX]; - double BytePerPixelDETY[DC__NUM_DPP__MAX]; - double BytePerPixelDETC[DC__NUM_DPP__MAX]; - double ReadBandwidthPlaneLuma[DC__NUM_DPP__MAX]; - double ReadBandwidthPlaneChroma[DC__NUM_DPP__MAX]; - unsigned int TotalActiveDPP; - unsigned int TotalDCCActiveDPP; - double UrgentRoundTripAndOutOfOrderLatency; - double DisplayPipeLineDeliveryTimeLuma[DC__NUM_DPP__MAX]; // WM - double DisplayPipeLineDeliveryTimeChroma[DC__NUM_DPP__MAX]; // WM - double LinesInDETY[DC__NUM_DPP__MAX]; // WM - double LinesInDETC[DC__NUM_DPP__MAX]; // WM - unsigned int LinesInDETYRoundedDownToSwath[DC__NUM_DPP__MAX]; // WM - unsigned int LinesInDETCRoundedDownToSwath[DC__NUM_DPP__MAX]; // WM - double FullDETBufferingTimeY[DC__NUM_DPP__MAX]; // WM - double FullDETBufferingTimeC[DC__NUM_DPP__MAX]; // WM - double MinFullDETBufferingTime; - double FrameTimeForMinFullDETBufferingTime; - double AverageReadBandwidthGBytePerSecond; - double PartOfBurstThatFitsInROB; - double StutterBurstTime; - //unsigned int NextPrefetchMode; - double VBlankTime; - double SmallestVBlank; - double DCFCLKDeepSleepPerPlane; - double EffectiveDETPlusLBLinesLuma; - double EffectiveDETPlusLBLinesChroma; - double UrgentLatencySupportUsLuma; - double UrgentLatencySupportUsChroma; - double UrgentLatencySupportUs[DC__NUM_DPP__MAX]; - unsigned int DSCFormatFactor; - unsigned int BlockHeight256BytesY[DC__NUM_DPP__MAX]; - unsigned int BlockHeight256BytesC[DC__NUM_DPP__MAX]; - unsigned int BlockWidth256BytesY[DC__NUM_DPP__MAX]; - unsigned int BlockWidth256BytesC[DC__NUM_DPP__MAX]; - double VInitPreFillY[DC__NUM_DPP__MAX]; - double VInitPreFillC[DC__NUM_DPP__MAX]; - unsigned int MaxNumSwathY[DC__NUM_DPP__MAX]; - unsigned int MaxNumSwathC[DC__NUM_DPP__MAX]; - double PrefetchSourceLinesY[DC__NUM_DPP__MAX]; - double PrefetchSourceLinesC[DC__NUM_DPP__MAX]; - double PixelPTEBytesPerRow[DC__NUM_DPP__MAX]; - double MetaRowByte[DC__NUM_DPP__MAX]; - unsigned int dpte_row_height[DC__NUM_DPP__MAX]; - unsigned int dpte_row_height_chroma[DC__NUM_DPP__MAX]; - unsigned int meta_row_height[DC__NUM_DPP__MAX]; - unsigned int meta_row_height_chroma[DC__NUM_DPP__MAX]; - - unsigned int MacroTileWidthY[DC__NUM_DPP__MAX]; - unsigned int MacroTileWidthC[DC__NUM_DPP__MAX]; - unsigned int MaxVStartupLines[DC__NUM_DPP__MAX]; - double WritebackDelay[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - bool PrefetchModeSupported; - bool AllowDRAMClockChangeDuringVBlank[DC__NUM_DPP__MAX]; - bool AllowDRAMSelfRefreshDuringVBlank[DC__NUM_DPP__MAX]; - double RequiredPrefetchPixDataBW[DC__NUM_DPP__MAX]; - double XFCRemoteSurfaceFlipDelay; - double TInitXFill; - double TslvChk; - double SrcActiveDrainRate; - double Tno_bw[DC__NUM_DPP__MAX]; - bool ImmediateFlipSupported; - - double prefetch_vm_bw[DC__NUM_DPP__MAX]; - double prefetch_row_bw[DC__NUM_DPP__MAX]; - bool ImmediateFlipSupportedForPipe[DC__NUM_DPP__MAX]; - unsigned int VStartupLines; - double DisplayPipeLineDeliveryTimeLumaPrefetch[DC__NUM_DPP__MAX]; - double DisplayPipeLineDeliveryTimeChromaPrefetch[DC__NUM_DPP__MAX]; - unsigned int ActiveDPPs; - unsigned int LBLatencyHidingSourceLinesY; - unsigned int LBLatencyHidingSourceLinesC; - double ActiveDRAMClockChangeLatencyMargin[DC__NUM_DPP__MAX]; - double MinActiveDRAMClockChangeMargin; - double XFCSlaveVUpdateOffset[DC__NUM_DPP__MAX]; - double XFCSlaveVupdateWidth[DC__NUM_DPP__MAX]; - double XFCSlaveVReadyOffset[DC__NUM_DPP__MAX]; - double InitFillLevel; - double FinalFillMargin; - double FinalFillLevel; - double RemainingFillLevel; - double TFinalxFill; - - - // - // SOC Bounding Box Parameters - // - double SRExitTime; - double SREnterPlusExitTime; - double UrgentLatency; - double WritebackLatency; - double PercentOfIdealDRAMAndFabricBWReceivedAfterUrgLatency; - double NumberOfChannels; - double DRAMChannelWidth; - double FabricDatapathToDCNDataReturn; - double ReturnBusWidth; - double Downspreading; - double DISPCLKDPPCLKDSCCLKDownSpreading; - double DISPCLKDPPCLKVCOSpeed; - double RoundTripPingLatencyCycles; - double UrgentOutOfOrderReturnPerChannel; - unsigned int VMMPageSize; - double DRAMClockChangeLatency; - double XFCBusTransportTime; - double XFCXBUFLatencyTolerance; - - // - // IP Parameters - // - unsigned int ROBBufferSizeInKByte; - double DETBufferSizeInKByte; - unsigned int DPPOutputBufferPixels; - unsigned int OPPOutputBufferLines; - unsigned int PixelChunkSizeInKByte; - double ReturnBW; - bool VirtualMemoryEnable; - unsigned int MaxPageTableLevels; - unsigned int OverridePageTableLevels; - unsigned int PTEChunkSize; - unsigned int MetaChunkSize; - unsigned int WritebackChunkSize; - bool ODMCapability; - unsigned int NumberOfDSC; - unsigned int LineBufferSize; - unsigned int MaxLineBufferLines; - unsigned int WritebackInterfaceLumaBufferSize; - unsigned int WritebackInterfaceChromaBufferSize; - unsigned int WritebackChromaLineBufferWidth; - double MaxDCHUBToPSCLThroughput; - double MaxPSCLToLBThroughput; - unsigned int PTEBufferSizeInRequests; - double DISPCLKRampingMargin; - unsigned int MaxInterDCNTileRepeaters; - bool XFCSupported; - double XFCSlvChunkSize; - double XFCFillBWOverhead; - double XFCFillConstant; - double XFCTSlvVupdateOffset; - double XFCTSlvVupdateWidth; - double XFCTSlvVreadyOffset; - double DPPCLKDelaySubtotal; - double DPPCLKDelaySCL; - double DPPCLKDelaySCLLBOnly; - double DPPCLKDelayCNVCFormater; - double DPPCLKDelayCNVCCursor; - double DISPCLKDelaySubtotal; - bool ProgressiveToInterlaceUnitInOPP; - unsigned int PDEProcessingBufIn64KBReqs; - - // Pipe/Plane Parameters - int VoltageLevel; - double FabricAndDRAMBandwidth; - double FabricClock; - double DRAMSpeed; - double DISPCLK; - double SOCCLK; - double DCFCLK; - - unsigned int NumberOfActivePlanes; - unsigned int ViewportWidth[DC__NUM_DPP__MAX]; - unsigned int ViewportHeight[DC__NUM_DPP__MAX]; - unsigned int ViewportYStartY[DC__NUM_DPP__MAX]; - unsigned int ViewportYStartC[DC__NUM_DPP__MAX]; - unsigned int PitchY[DC__NUM_DPP__MAX]; - unsigned int PitchC[DC__NUM_DPP__MAX]; - double HRatio[DC__NUM_DPP__MAX]; - double VRatio[DC__NUM_DPP__MAX]; - unsigned int htaps[DC__NUM_DPP__MAX]; - unsigned int vtaps[DC__NUM_DPP__MAX]; - unsigned int HTAPsChroma[DC__NUM_DPP__MAX]; - unsigned int VTAPsChroma[DC__NUM_DPP__MAX]; - unsigned int HTotal[DC__NUM_DPP__MAX]; - unsigned int VTotal[DC__NUM_DPP__MAX]; - unsigned int DPPPerPlane[DC__NUM_DPP__MAX]; - double PixelClock[DC__NUM_DPP__MAX]; - double PixelClockBackEnd[DC__NUM_DPP__MAX]; - double DPPCLK[DC__NUM_DPP__MAX]; - bool DCCEnable[DC__NUM_DPP__MAX]; - unsigned int DCCMetaPitchY[DC__NUM_DPP__MAX]; - enum scan_direction_class SourceScan[DC__NUM_DPP__MAX]; - enum source_format_class SourcePixelFormat[DC__NUM_DPP__MAX]; - bool WritebackEnable[DC__NUM_DPP__MAX]; - double WritebackDestinationWidth[DC__NUM_DPP__MAX]; - double WritebackDestinationHeight[DC__NUM_DPP__MAX]; - double WritebackSourceHeight[DC__NUM_DPP__MAX]; - enum source_format_class WritebackPixelFormat[DC__NUM_DPP__MAX]; - unsigned int WritebackLumaHTaps[DC__NUM_DPP__MAX]; - unsigned int WritebackLumaVTaps[DC__NUM_DPP__MAX]; - unsigned int WritebackChromaHTaps[DC__NUM_DPP__MAX]; - unsigned int WritebackChromaVTaps[DC__NUM_DPP__MAX]; - double WritebackHRatio[DC__NUM_DPP__MAX]; - double WritebackVRatio[DC__NUM_DPP__MAX]; - unsigned int HActive[DC__NUM_DPP__MAX]; - unsigned int VActive[DC__NUM_DPP__MAX]; - bool Interlace[DC__NUM_DPP__MAX]; - enum dm_swizzle_mode SurfaceTiling[DC__NUM_DPP__MAX]; - unsigned int ScalerRecoutWidth[DC__NUM_DPP__MAX]; - bool DynamicMetadataEnable[DC__NUM_DPP__MAX]; - unsigned int DynamicMetadataLinesBeforeActiveRequired[DC__NUM_DPP__MAX]; - unsigned int DynamicMetadataTransmittedBytes[DC__NUM_DPP__MAX]; - double DCCRate[DC__NUM_DPP__MAX]; - bool ODMCombineEnabled[DC__NUM_DPP__MAX]; - double OutputBpp[DC__NUM_DPP__MAX]; - unsigned int NumberOfDSCSlices[DC__NUM_DPP__MAX]; - bool DSCEnabled[DC__NUM_DPP__MAX]; - unsigned int DSCDelay[DC__NUM_DPP__MAX]; - unsigned int DSCInputBitPerComponent[DC__NUM_DPP__MAX]; - enum output_format_class OutputFormat[DC__NUM_DPP__MAX]; - enum output_encoder_class Output[DC__NUM_DPP__MAX]; - unsigned int BlendingAndTiming[DC__NUM_DPP__MAX]; - bool SynchronizedVBlank; - unsigned int NumberOfCursors[DC__NUM_DPP__MAX]; - unsigned int CursorWidth[DC__NUM_DPP__MAX][DC__NUM_CURSOR__MAX]; - unsigned int CursorBPP[DC__NUM_DPP__MAX][DC__NUM_CURSOR__MAX]; - bool XFCEnabled[DC__NUM_DPP__MAX]; - bool ScalerEnabled[DC__NUM_DPP__MAX]; - - // Intermediates/Informational - bool ImmediateFlipSupport; - unsigned int SwathHeightY[DC__NUM_DPP__MAX]; - unsigned int SwathHeightC[DC__NUM_DPP__MAX]; - unsigned int DETBufferSizeY[DC__NUM_DPP__MAX]; - unsigned int DETBufferSizeC[DC__NUM_DPP__MAX]; - unsigned int LBBitPerPixel[DC__NUM_DPP__MAX]; - double LastPixelOfLineExtraWatermark; - double TotalDataReadBandwidth; - unsigned int TotalActiveWriteback; - unsigned int EffectiveLBLatencyHidingSourceLinesLuma; - unsigned int EffectiveLBLatencyHidingSourceLinesChroma; - double BandwidthAvailableForImmediateFlip; - unsigned int PrefetchMode; - bool IgnoreViewportPositioning; - double PrefetchBandwidth[DC__NUM_DPP__MAX]; - bool ErrorResult[DC__NUM_DPP__MAX]; - double PDEAndMetaPTEBytesFrame[DC__NUM_DPP__MAX]; - - // - // Calculated dml_ml->vba.Outputs - // - double DCFClkDeepSleep; - double UrgentWatermark; - double UrgentExtraLatency; - double MemoryTripWatermark; - double WritebackUrgentWatermark; - double StutterExitWatermark; - double StutterEnterPlusExitWatermark; - double DRAMClockChangeWatermark; - double WritebackDRAMClockChangeWatermark; - double StutterEfficiency; - double StutterEfficiencyNotIncludingVBlank; - double MinUrgentLatencySupportUs; - double NonUrgentLatencyTolerance; - double MinActiveDRAMClockChangeLatencySupported; - enum clock_change_support DRAMClockChangeSupport; - - // These are the clocks calcuated by the library but they are not actually - // used explicitly. They are fetched by tests and then possibly used. The - // ultimate values to use are the ones specified by the parameters to DML - double DISPCLK_calculated; - double DSCCLK_calculated[DC__NUM_DPP__MAX]; - double DPPCLK_calculated[DC__NUM_DPP__MAX]; - - unsigned int VStartup[DC__NUM_DPP__MAX]; - unsigned int VUpdateOffsetPix[DC__NUM_DPP__MAX]; - unsigned int VUpdateWidthPix[DC__NUM_DPP__MAX]; - unsigned int VReadyOffsetPix[DC__NUM_DPP__MAX]; - unsigned int VStartupRequiredWhenNotEnoughTimeForDynamicMetadata; - - double ImmediateFlipBW; - unsigned int TotImmediateFlipBytes; - double TCalc; - double MinTTUVBlank[DC__NUM_DPP__MAX]; - double VRatioPrefetchY[DC__NUM_DPP__MAX]; - double VRatioPrefetchC[DC__NUM_DPP__MAX]; - double DSTXAfterScaler[DC__NUM_DPP__MAX]; - double DSTYAfterScaler[DC__NUM_DPP__MAX]; - - double DestinationLinesToRequestVMInVBlank[DC__NUM_DPP__MAX]; - double DestinationLinesToRequestRowInVBlank[DC__NUM_DPP__MAX]; - double DestinationLinesForPrefetch[DC__NUM_DPP__MAX]; - double DestinationLinesToRequestRowInImmediateFlip[DC__NUM_DPP__MAX]; - double DestinationLinesToRequestVMInImmediateFlip[DC__NUM_DPP__MAX]; - - double XFCTransferDelay[DC__NUM_DPP__MAX]; - double XFCPrechargeDelay[DC__NUM_DPP__MAX]; - double XFCRemoteSurfaceFlipLatency[DC__NUM_DPP__MAX]; - double XFCPrefetchMargin[DC__NUM_DPP__MAX]; - - display_e2e_pipe_params_st cache_pipes[DC__NUM_DPP__MAX]; - unsigned int cache_num_pipes; - unsigned int pipe_plane[DC__NUM_DPP__MAX]; - - /* vba mode support */ - /*inputs*/ - bool SupportGFX7CompatibleTilingIn32bppAnd64bpp; - double MaxHSCLRatio; - double MaxVSCLRatio; - unsigned int MaxNumWriteback; - bool WritebackLumaAndChromaScalingSupported; - bool Cursor64BppSupport; - double DCFCLKPerState[DC__VOLTAGE_STATES + 1]; - double FabricClockPerState[DC__VOLTAGE_STATES + 1]; - double SOCCLKPerState[DC__VOLTAGE_STATES + 1]; - double PHYCLKPerState[DC__VOLTAGE_STATES + 1]; - double MaxDppclk[DC__VOLTAGE_STATES + 1]; - double MaxDSCCLK[DC__VOLTAGE_STATES + 1]; - double DRAMSpeedPerState[DC__VOLTAGE_STATES + 1]; - double MaxDispclk[DC__VOLTAGE_STATES + 1]; - - /*outputs*/ - bool ScaleRatioAndTapsSupport; - bool SourceFormatPixelAndScanSupport; - unsigned int SwathWidthYSingleDPP[DC__NUM_DPP__MAX]; - double BytePerPixelInDETY[DC__NUM_DPP__MAX]; - double BytePerPixelInDETC[DC__NUM_DPP__MAX]; - double TotalReadBandwidthConsumedGBytePerSecond; - double ReadBandwidth[DC__NUM_DPP__MAX]; - double TotalWriteBandwidthConsumedGBytePerSecond; - double WriteBandwidth[DC__NUM_DPP__MAX]; - double TotalBandwidthConsumedGBytePerSecond; - bool DCCEnabledInAnyPlane; - bool WritebackLatencySupport; - bool WritebackModeSupport; - bool Writeback10bpc420Supported; - bool BandwidthSupport[DC__VOLTAGE_STATES + 1]; - unsigned int TotalNumberOfActiveWriteback; - double CriticalPoint; - double ReturnBWToDCNPerState; - double FabricAndDRAMBandwidthPerState[DC__VOLTAGE_STATES + 1]; - double ReturnBWPerState[DC__VOLTAGE_STATES + 1]; - double UrgentRoundTripAndOutOfOrderLatencyPerState[DC__VOLTAGE_STATES + 1]; - bool ODMCombineEnablePerState[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - bool PTEBufferSizeNotExceededY[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - bool PTEBufferSizeNotExceededC[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - bool PrefetchSupported[DC__VOLTAGE_STATES + 1]; - bool VRatioInPrefetchSupported[DC__VOLTAGE_STATES + 1]; - bool DISPCLK_DPPCLK_Support[DC__VOLTAGE_STATES + 1]; - bool TotalAvailablePipesSupport[DC__VOLTAGE_STATES + 1]; - bool UrgentLatencySupport[DC__VOLTAGE_STATES + 1]; - bool ModeSupport[DC__VOLTAGE_STATES + 1]; - bool DIOSupport[DC__VOLTAGE_STATES + 1]; - bool NotEnoughDSCUnits[DC__VOLTAGE_STATES + 1]; - bool DSCCLKRequiredMoreThanSupported[DC__VOLTAGE_STATES + 1]; - bool ROBSupport[DC__VOLTAGE_STATES + 1]; - bool PTEBufferSizeNotExceeded[DC__VOLTAGE_STATES + 1]; - bool RequiresDSC[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - bool IsErrorResult[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - bool ViewportSizeSupport[DC__VOLTAGE_STATES + 1]; - bool prefetch_vm_bw_valid; - bool prefetch_row_bw_valid; - bool NumberOfOTGSupport; - bool NonsupportedDSCInputBPC; - bool WritebackScaleRatioAndTapsSupport; - bool CursorSupport; - bool PitchSupport; - - double WritebackLineBufferLumaBufferSize; - double WritebackLineBufferChromaBufferSize; - double WritebackMinHSCLRatio; - double WritebackMinVSCLRatio; - double WritebackMaxHSCLRatio; - double WritebackMaxVSCLRatio; - double WritebackMaxHSCLTaps; - double WritebackMaxVSCLTaps; - unsigned int MaxNumDPP; - unsigned int MaxNumOTG; - double CursorBufferSize; - double CursorChunkSize; - unsigned int Mode; - unsigned int NoOfDPP[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double OutputLinkDPLanes[DC__NUM_DPP__MAX]; - double SwathWidthYPerState[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double SwathHeightYPerState[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double SwathHeightCPerState[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double UrgentLatencySupportUsPerState[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double VRatioPreY[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double VRatioPreC[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double RequiredPrefetchPixelDataBW[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double RequiredDPPCLK[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double RequiredDISPCLK[DC__VOLTAGE_STATES + 1]; - double TotalNumberOfActiveDPP[DC__VOLTAGE_STATES + 1]; - double TotalNumberOfDCCActiveDPP[DC__VOLTAGE_STATES + 1]; - double PrefetchBW[DC__NUM_DPP__MAX]; - double PDEAndMetaPTEBytesPerFrame[DC__NUM_DPP__MAX]; - double MetaRowBytes[DC__NUM_DPP__MAX]; - double DPTEBytesPerRow[DC__NUM_DPP__MAX]; - double PrefetchLinesY[DC__NUM_DPP__MAX]; - double PrefetchLinesC[DC__NUM_DPP__MAX]; - unsigned int MaxNumSwY[DC__NUM_DPP__MAX]; - unsigned int MaxNumSwC[DC__NUM_DPP__MAX]; - double PrefillY[DC__NUM_DPP__MAX]; - double PrefillC[DC__NUM_DPP__MAX]; - double LineTimesForPrefetch[DC__NUM_DPP__MAX]; - double LinesForMetaPTE[DC__NUM_DPP__MAX]; - double LinesForMetaAndDPTERow[DC__NUM_DPP__MAX]; - double MinDPPCLKUsingSingleDPP[DC__NUM_DPP__MAX]; - double RequiresFEC[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - unsigned int OutputBppPerState[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - double DSCDelayPerState[DC__VOLTAGE_STATES + 1][DC__NUM_DPP__MAX]; - unsigned int Read256BlockHeightY[DC__NUM_DPP__MAX]; - unsigned int Read256BlockWidthY[DC__NUM_DPP__MAX]; - unsigned int Read256BlockHeightC[DC__NUM_DPP__MAX]; - unsigned int Read256BlockWidthC[DC__NUM_DPP__MAX]; - unsigned int ImmediateFlipBytes[DC__NUM_DPP__MAX]; - double MaxSwathHeightY[DC__NUM_DPP__MAX]; - double MaxSwathHeightC[DC__NUM_DPP__MAX]; - double MinSwathHeightY[DC__NUM_DPP__MAX]; - double MinSwathHeightC[DC__NUM_DPP__MAX]; - double PSCL_FACTOR[DC__NUM_DPP__MAX]; - double PSCL_FACTOR_CHROMA[DC__NUM_DPP__MAX]; - double MaximumVStartup[DC__NUM_DPP__MAX]; - double AlignedDCCMetaPitch[DC__NUM_DPP__MAX]; - double AlignedYPitch[DC__NUM_DPP__MAX]; - double AlignedCPitch[DC__NUM_DPP__MAX]; - double MaximumSwathWidth[DC__NUM_DPP__MAX]; - double final_flip_bw[DC__NUM_DPP__MAX]; - double ImmediateFlipSupportedForState[DC__VOLTAGE_STATES + 1]; - - double WritebackLumaVExtra; - double WritebackChromaVExtra; - double WritebackRequiredDISPCLK; - double MaximumSwathWidthSupport; - double MaximumSwathWidthInDETBuffer; - double MaximumSwathWidthInLineBuffer; - double MaxDispclkRoundedDownToDFSGranularity; - double MaxDppclkRoundedDownToDFSGranularity; - double PlaneRequiredDISPCLKWithoutODMCombine; - double PlaneRequiredDISPCLK; - double TotalNumberOfActiveOTG; - double FECOverhead; - double EffectiveFECOverhead; - unsigned int Outbpp; - unsigned int OutbppDSC; - double TotalDSCUnitsRequired; - double bpp; - unsigned int slices; - double SwathWidthGranularityY; - double RoundedUpMaxSwathSizeBytesY; - double SwathWidthGranularityC; - double RoundedUpMaxSwathSizeBytesC; - double LinesInDETLuma; - double LinesInDETChroma; - double EffectiveDETLBLinesLuma; - double EffectiveDETLBLinesChroma; - double ProjectedDCFCLKDeepSleep; - double PDEAndMetaPTEBytesPerFrameY; - double PDEAndMetaPTEBytesPerFrameC; - unsigned int MetaRowBytesY; - unsigned int MetaRowBytesC; - unsigned int DPTEBytesPerRowC; - unsigned int DPTEBytesPerRowY; - double ExtraLatency; - double TimeCalc; - double TWait; - double MaximumReadBandwidthWithPrefetch; - double total_dcn_read_bw_with_flip; -}; - -#endif /* _DML2_DISPLAY_MODE_VBA_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_calc.c b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_calc.c deleted file mode 100644 index 325dd2b..0000000 --- a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_calc.c +++ /dev/null
@@ -1,1772 +0,0 @@ -/* - * Copyright 2017 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ - -#include "display_mode_lib.h" -#include "display_mode_vba.h" -#include "display_rq_dlg_calc.h" - -/* - * NOTE: - * This file is gcc-parseable HW gospel, coming straight from HW engineers. - * - * It doesn't adhere to Linux kernel style and sometimes will do things in odd - * ways. Unless there is something clearly wrong with it the code should - * remain as-is as it provides us with a guarantee from HW that it is correct. - */ - -static void calculate_ttu_cursor(struct display_mode_lib *mode_lib, - double *refcyc_per_req_delivery_pre_cur, - double *refcyc_per_req_delivery_cur, - double refclk_freq_in_mhz, - double ref_freq_to_pix_freq, - double hscale_pixel_rate_l, - double hscl_ratio, - double vratio_pre_l, - double vratio_l, - unsigned int cur_width, - enum cursor_bpp cur_bpp); - -#include "dml_inline_defs.h" - -static unsigned int get_bytes_per_element(enum source_format_class source_format, bool is_chroma) -{ - unsigned int ret_val = 0; - - if (source_format == dm_444_16) { - if (!is_chroma) - ret_val = 2; - } else if (source_format == dm_444_32) { - if (!is_chroma) - ret_val = 4; - } else if (source_format == dm_444_64) { - if (!is_chroma) - ret_val = 8; - } else if (source_format == dm_420_8) { - if (is_chroma) - ret_val = 2; - else - ret_val = 1; - } else if (source_format == dm_420_10) { - if (is_chroma) - ret_val = 4; - else - ret_val = 2; - } else if (source_format == dm_444_8) { - ret_val = 1; - } - return ret_val; -} - -static bool is_dual_plane(enum source_format_class source_format) -{ - bool ret_val = 0; - - if ((source_format == dm_420_8) || (source_format == dm_420_10)) - ret_val = 1; - - return ret_val; -} - -static double get_refcyc_per_delivery(struct display_mode_lib *mode_lib, - double refclk_freq_in_mhz, - double pclk_freq_in_mhz, - bool odm_combine, - unsigned int recout_width, - unsigned int hactive, - double vratio, - double hscale_pixel_rate, - unsigned int delivery_width, - unsigned int req_per_swath_ub) -{ - double refcyc_per_delivery = 0.0; - - if (vratio <= 1.0) { - if (odm_combine) - refcyc_per_delivery = (double) refclk_freq_in_mhz - * dml_min((double) recout_width, (double) hactive / 2.0) - / pclk_freq_in_mhz / (double) req_per_swath_ub; - else - refcyc_per_delivery = (double) refclk_freq_in_mhz * (double) recout_width - / pclk_freq_in_mhz / (double) req_per_swath_ub; - } else { - refcyc_per_delivery = (double) refclk_freq_in_mhz * (double) delivery_width - / (double) hscale_pixel_rate / (double) req_per_swath_ub; - } - - dml_print("DML_DLG: %s: refclk_freq_in_mhz = %3.2f\n", __func__, refclk_freq_in_mhz); - dml_print("DML_DLG: %s: pclk_freq_in_mhz = %3.2f\n", __func__, pclk_freq_in_mhz); - dml_print("DML_DLG: %s: recout_width = %d\n", __func__, recout_width); - dml_print("DML_DLG: %s: vratio = %3.2f\n", __func__, vratio); - dml_print("DML_DLG: %s: req_per_swath_ub = %d\n", __func__, req_per_swath_ub); - dml_print("DML_DLG: %s: refcyc_per_delivery= %3.2f\n", __func__, refcyc_per_delivery); - - return refcyc_per_delivery; - -} - -static unsigned int get_blk_size_bytes(const enum source_macro_tile_size tile_size) -{ - if (tile_size == dm_256k_tile) - return (256 * 1024); - else if (tile_size == dm_64k_tile) - return (64 * 1024); - else - return (4 * 1024); -} - -static void extract_rq_sizing_regs(struct display_mode_lib *mode_lib, - display_data_rq_regs_st *rq_regs, - const display_data_rq_sizing_params_st rq_sizing) -{ - dml_print("DML_DLG: %s: rq_sizing param\n", __func__); - print__data_rq_sizing_params_st(mode_lib, rq_sizing); - - rq_regs->chunk_size = dml_log2(rq_sizing.chunk_bytes) - 10; - - if (rq_sizing.min_chunk_bytes == 0) - rq_regs->min_chunk_size = 0; - else - rq_regs->min_chunk_size = dml_log2(rq_sizing.min_chunk_bytes) - 8 + 1; - - rq_regs->meta_chunk_size = dml_log2(rq_sizing.meta_chunk_bytes) - 10; - if (rq_sizing.min_meta_chunk_bytes == 0) - rq_regs->min_meta_chunk_size = 0; - else - rq_regs->min_meta_chunk_size = dml_log2(rq_sizing.min_meta_chunk_bytes) - 6 + 1; - - rq_regs->dpte_group_size = dml_log2(rq_sizing.dpte_group_bytes) - 6; - rq_regs->mpte_group_size = dml_log2(rq_sizing.mpte_group_bytes) - 6; -} - -static void extract_rq_regs(struct display_mode_lib *mode_lib, - display_rq_regs_st *rq_regs, - const display_rq_params_st rq_param) -{ - unsigned int detile_buf_size_in_bytes = mode_lib->ip.det_buffer_size_kbytes * 1024; - unsigned int detile_buf_plane1_addr = 0; - - extract_rq_sizing_regs(mode_lib, &(rq_regs->rq_regs_l), rq_param.sizing.rq_l); - - rq_regs->rq_regs_l.pte_row_height_linear = dml_floor(dml_log2(rq_param.dlg.rq_l.dpte_row_height), - 1) - 3; - - if (rq_param.yuv420) { - extract_rq_sizing_regs(mode_lib, &(rq_regs->rq_regs_c), rq_param.sizing.rq_c); - rq_regs->rq_regs_c.pte_row_height_linear = dml_floor(dml_log2(rq_param.dlg.rq_c.dpte_row_height), - 1) - 3; - } - - rq_regs->rq_regs_l.swath_height = dml_log2(rq_param.dlg.rq_l.swath_height); - rq_regs->rq_regs_c.swath_height = dml_log2(rq_param.dlg.rq_c.swath_height); - - // FIXME: take the max between luma, chroma chunk size? - // okay for now, as we are setting chunk_bytes to 8kb anyways - if (rq_param.sizing.rq_l.chunk_bytes >= 32 * 1024) { //32kb - rq_regs->drq_expansion_mode = 0; - } else { - rq_regs->drq_expansion_mode = 2; - } - rq_regs->prq_expansion_mode = 1; - rq_regs->mrq_expansion_mode = 1; - rq_regs->crq_expansion_mode = 1; - - if (rq_param.yuv420) { - if ((double) rq_param.misc.rq_l.stored_swath_bytes - / (double) rq_param.misc.rq_c.stored_swath_bytes <= 1.5) { - detile_buf_plane1_addr = (detile_buf_size_in_bytes / 2.0 / 64.0); // half to chroma - } else { - detile_buf_plane1_addr = dml_round_to_multiple((unsigned int) ((2.0 * detile_buf_size_in_bytes) / 3.0), - 256, - 0) / 64.0; // 2/3 to chroma - } - } - rq_regs->plane1_base_address = detile_buf_plane1_addr; -} - -static void handle_det_buf_split(struct display_mode_lib *mode_lib, - display_rq_params_st *rq_param, - const display_pipe_source_params_st pipe_src_param) -{ - unsigned int total_swath_bytes = 0; - unsigned int swath_bytes_l = 0; - unsigned int swath_bytes_c = 0; - unsigned int full_swath_bytes_packed_l = 0; - unsigned int full_swath_bytes_packed_c = 0; - bool req128_l = 0; - bool req128_c = 0; - bool surf_linear = (pipe_src_param.sw_mode == dm_sw_linear); - bool surf_vert = (pipe_src_param.source_scan == dm_vert); - unsigned int log2_swath_height_l = 0; - unsigned int log2_swath_height_c = 0; - unsigned int detile_buf_size_in_bytes = mode_lib->ip.det_buffer_size_kbytes * 1024; - - full_swath_bytes_packed_l = rq_param->misc.rq_l.full_swath_bytes; - full_swath_bytes_packed_c = rq_param->misc.rq_c.full_swath_bytes; - - if (rq_param->yuv420_10bpc) { - full_swath_bytes_packed_l = dml_round_to_multiple(rq_param->misc.rq_l.full_swath_bytes * 2 / 3, - 256, - 1) + 256; - full_swath_bytes_packed_c = dml_round_to_multiple(rq_param->misc.rq_c.full_swath_bytes * 2 / 3, - 256, - 1) + 256; - } - - if (rq_param->yuv420) { - total_swath_bytes = 2 * full_swath_bytes_packed_l + 2 * full_swath_bytes_packed_c; - - if (total_swath_bytes <= detile_buf_size_in_bytes) { //full 256b request - req128_l = 0; - req128_c = 0; - swath_bytes_l = full_swath_bytes_packed_l; - swath_bytes_c = full_swath_bytes_packed_c; - } else { //128b request (for luma only for yuv420 8bpc) - req128_l = 1; - req128_c = 0; - swath_bytes_l = full_swath_bytes_packed_l / 2; - swath_bytes_c = full_swath_bytes_packed_c; - } - // Note: assumption, the config that pass in will fit into - // the detiled buffer. - } else { - total_swath_bytes = 2 * full_swath_bytes_packed_l; - - if (total_swath_bytes <= detile_buf_size_in_bytes) - req128_l = 0; - else - req128_l = 1; - - swath_bytes_l = total_swath_bytes; - swath_bytes_c = 0; - } - rq_param->misc.rq_l.stored_swath_bytes = swath_bytes_l; - rq_param->misc.rq_c.stored_swath_bytes = swath_bytes_c; - - if (surf_linear) { - log2_swath_height_l = 0; - log2_swath_height_c = 0; - } else if (!surf_vert) { - log2_swath_height_l = dml_log2(rq_param->misc.rq_l.blk256_height) - req128_l; - log2_swath_height_c = dml_log2(rq_param->misc.rq_c.blk256_height) - req128_c; - } else { - log2_swath_height_l = dml_log2(rq_param->misc.rq_l.blk256_width) - req128_l; - log2_swath_height_c = dml_log2(rq_param->misc.rq_c.blk256_width) - req128_c; - } - rq_param->dlg.rq_l.swath_height = 1 << log2_swath_height_l; - rq_param->dlg.rq_c.swath_height = 1 << log2_swath_height_c; - - dml_print("DML_DLG: %s: req128_l = %0d\n", __func__, req128_l); - dml_print("DML_DLG: %s: req128_c = %0d\n", __func__, req128_c); - dml_print("DML_DLG: %s: full_swath_bytes_packed_l = %0d\n", - __func__, - full_swath_bytes_packed_l); - dml_print("DML_DLG: %s: full_swath_bytes_packed_c = %0d\n", - __func__, - full_swath_bytes_packed_c); -} - -static void get_meta_and_pte_attr(struct display_mode_lib *mode_lib, - display_data_rq_dlg_params_st *rq_dlg_param, - display_data_rq_misc_params_st *rq_misc_param, - display_data_rq_sizing_params_st *rq_sizing_param, - unsigned int vp_width, - unsigned int vp_height, - unsigned int data_pitch, - unsigned int meta_pitch, - unsigned int source_format, - unsigned int tiling, - unsigned int macro_tile_size, - unsigned int source_scan, - unsigned int is_chroma) -{ - bool surf_linear = (tiling == dm_sw_linear); - bool surf_vert = (source_scan == dm_vert); - - unsigned int bytes_per_element; - unsigned int bytes_per_element_y = get_bytes_per_element((enum source_format_class)(source_format), - false); - unsigned int bytes_per_element_c = get_bytes_per_element((enum source_format_class)(source_format), - true); - - unsigned int blk256_width = 0; - unsigned int blk256_height = 0; - - unsigned int blk256_width_y = 0; - unsigned int blk256_height_y = 0; - unsigned int blk256_width_c = 0; - unsigned int blk256_height_c = 0; - unsigned int log2_bytes_per_element; - unsigned int log2_blk256_width; - unsigned int log2_blk256_height; - unsigned int blk_bytes; - unsigned int log2_blk_bytes; - unsigned int log2_blk_height; - unsigned int log2_blk_width; - unsigned int log2_meta_req_bytes; - unsigned int log2_meta_req_height; - unsigned int log2_meta_req_width; - unsigned int meta_req_width; - unsigned int meta_req_height; - unsigned int log2_meta_row_height; - unsigned int meta_row_width_ub; - unsigned int log2_meta_chunk_bytes; - unsigned int log2_meta_chunk_height; - - //full sized meta chunk width in unit of data elements - unsigned int log2_meta_chunk_width; - unsigned int log2_min_meta_chunk_bytes; - unsigned int min_meta_chunk_width; - unsigned int meta_chunk_width; - unsigned int meta_chunk_per_row_int; - unsigned int meta_row_remainder; - unsigned int meta_chunk_threshold; - unsigned int meta_blk_bytes; - unsigned int meta_blk_height; - unsigned int meta_blk_width; - unsigned int meta_surface_bytes; - unsigned int vmpg_bytes; - unsigned int meta_pte_req_per_frame_ub; - unsigned int meta_pte_bytes_per_frame_ub; - const unsigned int log2_vmpg_bytes = dml_log2(mode_lib->soc.vmm_page_size_bytes); - const unsigned int dpte_buf_in_pte_reqs = mode_lib->ip.dpte_buffer_size_in_pte_reqs; - const unsigned int pde_proc_buffer_size_64k_reqs = - mode_lib->ip.pde_proc_buffer_size_64k_reqs; - - unsigned int log2_vmpg_height = 0; - unsigned int log2_vmpg_width = 0; - unsigned int log2_dpte_req_height_ptes = 0; - unsigned int log2_dpte_req_height = 0; - unsigned int log2_dpte_req_width = 0; - unsigned int log2_dpte_row_height_linear = 0; - unsigned int log2_dpte_row_height = 0; - unsigned int log2_dpte_group_width = 0; - unsigned int dpte_row_width_ub = 0; - unsigned int dpte_req_height = 0; - unsigned int dpte_req_width = 0; - unsigned int dpte_group_width = 0; - unsigned int log2_dpte_group_bytes = 0; - unsigned int log2_dpte_group_length = 0; - unsigned int pde_buf_entries; - bool yuv420 = (source_format == dm_420_8 || source_format == dm_420_10); - - Calculate256BBlockSizes((enum source_format_class)(source_format), - (enum dm_swizzle_mode)(tiling), - bytes_per_element_y, - bytes_per_element_c, - &blk256_height_y, - &blk256_height_c, - &blk256_width_y, - &blk256_width_c); - - if (!is_chroma) { - blk256_width = blk256_width_y; - blk256_height = blk256_height_y; - bytes_per_element = bytes_per_element_y; - } else { - blk256_width = blk256_width_c; - blk256_height = blk256_height_c; - bytes_per_element = bytes_per_element_c; - } - - log2_bytes_per_element = dml_log2(bytes_per_element); - - dml_print("DML_DLG: %s: surf_linear = %d\n", __func__, surf_linear); - dml_print("DML_DLG: %s: surf_vert = %d\n", __func__, surf_vert); - dml_print("DML_DLG: %s: blk256_width = %d\n", __func__, blk256_width); - dml_print("DML_DLG: %s: blk256_height = %d\n", __func__, blk256_height); - - log2_blk256_width = dml_log2((double) blk256_width); - log2_blk256_height = dml_log2((double) blk256_height); - blk_bytes = surf_linear ? - 256 : get_blk_size_bytes((enum source_macro_tile_size) macro_tile_size); - log2_blk_bytes = dml_log2((double) blk_bytes); - log2_blk_height = 0; - log2_blk_width = 0; - - // remember log rule - // "+" in log is multiply - // "-" in log is divide - // "/2" is like square root - // blk is vertical biased - if (tiling != dm_sw_linear) - log2_blk_height = log2_blk256_height - + dml_ceil((double) (log2_blk_bytes - 8) / 2.0, 1); - else - log2_blk_height = 0; // blk height of 1 - - log2_blk_width = log2_blk_bytes - log2_bytes_per_element - log2_blk_height; - - if (!surf_vert) { - rq_dlg_param->swath_width_ub = dml_round_to_multiple(vp_width - 1, blk256_width, 1) - + blk256_width; - rq_dlg_param->req_per_swath_ub = rq_dlg_param->swath_width_ub >> log2_blk256_width; - } else { - rq_dlg_param->swath_width_ub = dml_round_to_multiple(vp_height - 1, blk256_height, 1) - + blk256_height; - rq_dlg_param->req_per_swath_ub = rq_dlg_param->swath_width_ub >> log2_blk256_height; - } - - if (!surf_vert) - rq_misc_param->full_swath_bytes = rq_dlg_param->swath_width_ub * blk256_height - * bytes_per_element; - else - rq_misc_param->full_swath_bytes = rq_dlg_param->swath_width_ub * blk256_width - * bytes_per_element; - - rq_misc_param->blk256_height = blk256_height; - rq_misc_param->blk256_width = blk256_width; - - // ------- - // meta - // ------- - log2_meta_req_bytes = 6; // meta request is 64b and is 8x8byte meta element - - // each 64b meta request for dcn is 8x8 meta elements and - // a meta element covers one 256b block of the the data surface. - log2_meta_req_height = log2_blk256_height + 3; // meta req is 8x8 byte, each byte represent 1 blk256 - log2_meta_req_width = log2_meta_req_bytes + 8 - log2_bytes_per_element - - log2_meta_req_height; - meta_req_width = 1 << log2_meta_req_width; - meta_req_height = 1 << log2_meta_req_height; - log2_meta_row_height = 0; - meta_row_width_ub = 0; - - // the dimensions of a meta row are meta_row_width x meta_row_height in elements. - // calculate upper bound of the meta_row_width - if (!surf_vert) { - log2_meta_row_height = log2_meta_req_height; - meta_row_width_ub = dml_round_to_multiple(vp_width - 1, meta_req_width, 1) - + meta_req_width; - rq_dlg_param->meta_req_per_row_ub = meta_row_width_ub / meta_req_width; - } else { - log2_meta_row_height = log2_meta_req_width; - meta_row_width_ub = dml_round_to_multiple(vp_height - 1, meta_req_height, 1) - + meta_req_height; - rq_dlg_param->meta_req_per_row_ub = meta_row_width_ub / meta_req_height; - } - rq_dlg_param->meta_bytes_per_row_ub = rq_dlg_param->meta_req_per_row_ub * 64; - - rq_dlg_param->meta_row_height = 1 << log2_meta_row_height; - - log2_meta_chunk_bytes = dml_log2(rq_sizing_param->meta_chunk_bytes); - log2_meta_chunk_height = log2_meta_row_height; - - //full sized meta chunk width in unit of data elements - log2_meta_chunk_width = log2_meta_chunk_bytes + 8 - log2_bytes_per_element - - log2_meta_chunk_height; - log2_min_meta_chunk_bytes = dml_log2(rq_sizing_param->min_meta_chunk_bytes); - min_meta_chunk_width = 1 - << (log2_min_meta_chunk_bytes + 8 - log2_bytes_per_element - - log2_meta_chunk_height); - meta_chunk_width = 1 << log2_meta_chunk_width; - meta_chunk_per_row_int = (unsigned int) (meta_row_width_ub / meta_chunk_width); - meta_row_remainder = meta_row_width_ub % meta_chunk_width; - meta_chunk_threshold = 0; - meta_blk_bytes = 4096; - meta_blk_height = blk256_height * 64; - meta_blk_width = meta_blk_bytes * 256 / bytes_per_element / meta_blk_height; - meta_surface_bytes = meta_pitch - * (dml_round_to_multiple(vp_height - 1, meta_blk_height, 1) + meta_blk_height) - * bytes_per_element / 256; - vmpg_bytes = mode_lib->soc.vmm_page_size_bytes; - meta_pte_req_per_frame_ub = (dml_round_to_multiple(meta_surface_bytes - vmpg_bytes, - 8 * vmpg_bytes, - 1) + 8 * vmpg_bytes) / (8 * vmpg_bytes); - meta_pte_bytes_per_frame_ub = meta_pte_req_per_frame_ub * 64; //64B mpte request - rq_dlg_param->meta_pte_bytes_per_frame_ub = meta_pte_bytes_per_frame_ub; - - dml_print("DML_DLG: %s: meta_blk_height = %d\n", __func__, meta_blk_height); - dml_print("DML_DLG: %s: meta_blk_width = %d\n", __func__, meta_blk_width); - dml_print("DML_DLG: %s: meta_surface_bytes = %d\n", __func__, meta_surface_bytes); - dml_print("DML_DLG: %s: meta_pte_req_per_frame_ub = %d\n", - __func__, - meta_pte_req_per_frame_ub); - dml_print("DML_DLG: %s: meta_pte_bytes_per_frame_ub = %d\n", - __func__, - meta_pte_bytes_per_frame_ub); - - if (!surf_vert) - meta_chunk_threshold = 2 * min_meta_chunk_width - meta_req_width; - else - meta_chunk_threshold = 2 * min_meta_chunk_width - meta_req_height; - - if (meta_row_remainder <= meta_chunk_threshold) - rq_dlg_param->meta_chunks_per_row_ub = meta_chunk_per_row_int + 1; - else - rq_dlg_param->meta_chunks_per_row_ub = meta_chunk_per_row_int + 2; - - // ------ - // dpte - // ------ - if (surf_linear) { - log2_vmpg_height = 0; // one line high - } else { - log2_vmpg_height = (log2_vmpg_bytes - 8) / 2 + log2_blk256_height; - } - log2_vmpg_width = log2_vmpg_bytes - log2_bytes_per_element - log2_vmpg_height; - - // only 3 possible shapes for dpte request in dimensions of ptes: 8x1, 4x2, 2x4. - if (surf_linear) { //one 64B PTE request returns 8 PTEs - log2_dpte_req_height_ptes = 0; - log2_dpte_req_width = log2_vmpg_width + 3; - log2_dpte_req_height = 0; - } else if (log2_blk_bytes == 12) { //4KB tile means 4kB page size - //one 64B req gives 8x1 PTEs for 4KB tile - log2_dpte_req_height_ptes = 0; - log2_dpte_req_width = log2_blk_width + 3; - log2_dpte_req_height = log2_blk_height + 0; - } else if ((log2_blk_bytes >= 16) && (log2_vmpg_bytes == 12)) { // tile block >= 64KB - //two 64B reqs of 2x4 PTEs give 16 PTEs to cover 64KB - log2_dpte_req_height_ptes = 4; - log2_dpte_req_width = log2_blk256_width + 4; // log2_64KB_width - log2_dpte_req_height = log2_blk256_height + 4; // log2_64KB_height - } else { //64KB page size and must 64KB tile block - //one 64B req gives 8x1 PTEs for 64KB tile - log2_dpte_req_height_ptes = 0; - log2_dpte_req_width = log2_blk_width + 3; - log2_dpte_req_height = log2_blk_height + 0; - } - - // The dpte request dimensions in data elements is dpte_req_width x dpte_req_height - // log2_vmpg_width is how much 1 pte represent, now calculating how much a 64b pte req represent - // That depends on the pte shape (i.e. 8x1, 4x2, 2x4) - //log2_dpte_req_height = log2_vmpg_height + log2_dpte_req_height_ptes; - //log2_dpte_req_width = log2_vmpg_width + log2_dpte_req_width_ptes; - dpte_req_height = 1 << log2_dpte_req_height; - dpte_req_width = 1 << log2_dpte_req_width; - - // calculate pitch dpte row buffer can hold - // round the result down to a power of two. - pde_buf_entries = yuv420 ? (pde_proc_buffer_size_64k_reqs >> 1) : pde_proc_buffer_size_64k_reqs; - if (surf_linear) { - unsigned int dpte_row_height; - - log2_dpte_row_height_linear = dml_floor(dml_log2(dml_min(64 * 1024 * pde_buf_entries - / bytes_per_element, - dpte_buf_in_pte_reqs - * dpte_req_width) - / data_pitch), - 1); - - ASSERT(log2_dpte_row_height_linear >= 3); - - if (log2_dpte_row_height_linear > 7) - log2_dpte_row_height_linear = 7; - - log2_dpte_row_height = log2_dpte_row_height_linear; - // For linear, the dpte row is pitch dependent and the pte requests wrap at the pitch boundary. - // the dpte_row_width_ub is the upper bound of data_pitch*dpte_row_height in elements with this unique buffering. - dpte_row_height = 1 << log2_dpte_row_height; - dpte_row_width_ub = dml_round_to_multiple(data_pitch * dpte_row_height - 1, - dpte_req_width, - 1) + dpte_req_width; - rq_dlg_param->dpte_req_per_row_ub = dpte_row_width_ub / dpte_req_width; - } else { - // the upper bound of the dpte_row_width without dependency on viewport position follows. - // for tiled mode, row height is the same as req height and row store up to vp size upper bound - if (!surf_vert) { - log2_dpte_row_height = log2_dpte_req_height; - dpte_row_width_ub = dml_round_to_multiple(vp_width - 1, dpte_req_width, 1) - + dpte_req_width; - rq_dlg_param->dpte_req_per_row_ub = dpte_row_width_ub / dpte_req_width; - } else { - log2_dpte_row_height = - (log2_blk_width < log2_dpte_req_width) ? - log2_blk_width : log2_dpte_req_width; - dpte_row_width_ub = dml_round_to_multiple(vp_height - 1, dpte_req_height, 1) - + dpte_req_height; - rq_dlg_param->dpte_req_per_row_ub = dpte_row_width_ub / dpte_req_height; - } - } - if (log2_blk_bytes >= 16 && log2_vmpg_bytes == 12) // tile block >= 64KB - rq_dlg_param->dpte_bytes_per_row_ub = rq_dlg_param->dpte_req_per_row_ub * 128; //2*64B dpte request - else - rq_dlg_param->dpte_bytes_per_row_ub = rq_dlg_param->dpte_req_per_row_ub * 64; //64B dpte request - - rq_dlg_param->dpte_row_height = 1 << log2_dpte_row_height; - - // the dpte_group_bytes is reduced for the specific case of vertical - // access of a tile surface that has dpte request of 8x1 ptes. - if (!surf_linear & (log2_dpte_req_height_ptes == 0) & surf_vert) //reduced, in this case, will have page fault within a group - rq_sizing_param->dpte_group_bytes = 512; - else - //full size - rq_sizing_param->dpte_group_bytes = 2048; - - //since pte request size is 64byte, the number of data pte requests per full sized group is as follows. - log2_dpte_group_bytes = dml_log2(rq_sizing_param->dpte_group_bytes); - log2_dpte_group_length = log2_dpte_group_bytes - 6; //length in 64b requests - - // full sized data pte group width in elements - if (!surf_vert) - log2_dpte_group_width = log2_dpte_group_length + log2_dpte_req_width; - else - log2_dpte_group_width = log2_dpte_group_length + log2_dpte_req_height; - - //But if the tile block >=64KB and the page size is 4KB, then each dPTE request is 2*64B - if ((log2_blk_bytes >= 16) && (log2_vmpg_bytes == 12)) // tile block >= 64KB - log2_dpte_group_width = log2_dpte_group_width - 1; - - dpte_group_width = 1 << log2_dpte_group_width; - - // since dpte groups are only aligned to dpte_req_width and not dpte_group_width, - // the upper bound for the dpte groups per row is as follows. - rq_dlg_param->dpte_groups_per_row_ub = dml_ceil((double) dpte_row_width_ub / dpte_group_width, - 1); -} - -static void get_surf_rq_param(struct display_mode_lib *mode_lib, - display_data_rq_sizing_params_st *rq_sizing_param, - display_data_rq_dlg_params_st *rq_dlg_param, - display_data_rq_misc_params_st *rq_misc_param, - const display_pipe_source_params_st pipe_src_param, - bool is_chroma) -{ - bool mode_422 = 0; - unsigned int vp_width = 0; - unsigned int vp_height = 0; - unsigned int data_pitch = 0; - unsigned int meta_pitch = 0; - unsigned int ppe = mode_422 ? 2 : 1; - - // FIXME check if ppe apply for both luma and chroma in 422 case - if (is_chroma) { - vp_width = pipe_src_param.viewport_width_c / ppe; - vp_height = pipe_src_param.viewport_height_c; - data_pitch = pipe_src_param.data_pitch_c; - meta_pitch = pipe_src_param.meta_pitch_c; - } else { - vp_width = pipe_src_param.viewport_width / ppe; - vp_height = pipe_src_param.viewport_height; - data_pitch = pipe_src_param.data_pitch; - meta_pitch = pipe_src_param.meta_pitch; - } - - rq_sizing_param->chunk_bytes = 8192; - - if (rq_sizing_param->chunk_bytes == 64 * 1024) - rq_sizing_param->min_chunk_bytes = 0; - else - rq_sizing_param->min_chunk_bytes = 1024; - - rq_sizing_param->meta_chunk_bytes = 2048; - rq_sizing_param->min_meta_chunk_bytes = 256; - - rq_sizing_param->mpte_group_bytes = 2048; - - get_meta_and_pte_attr(mode_lib, - rq_dlg_param, - rq_misc_param, - rq_sizing_param, - vp_width, - vp_height, - data_pitch, - meta_pitch, - pipe_src_param.source_format, - pipe_src_param.sw_mode, - pipe_src_param.macro_tile_size, - pipe_src_param.source_scan, - is_chroma); -} - -void dml_rq_dlg_get_rq_params(struct display_mode_lib *mode_lib, - display_rq_params_st *rq_param, - const display_pipe_source_params_st pipe_src_param) -{ - // get param for luma surface - rq_param->yuv420 = pipe_src_param.source_format == dm_420_8 - || pipe_src_param.source_format == dm_420_10; - rq_param->yuv420_10bpc = pipe_src_param.source_format == dm_420_10; - - get_surf_rq_param(mode_lib, - &(rq_param->sizing.rq_l), - &(rq_param->dlg.rq_l), - &(rq_param->misc.rq_l), - pipe_src_param, - 0); - - if (is_dual_plane((enum source_format_class)(pipe_src_param.source_format))) { - // get param for chroma surface - get_surf_rq_param(mode_lib, - &(rq_param->sizing.rq_c), - &(rq_param->dlg.rq_c), - &(rq_param->misc.rq_c), - pipe_src_param, - 1); - } - - // calculate how to split the det buffer space between luma and chroma - handle_det_buf_split(mode_lib, rq_param, pipe_src_param); - print__rq_params_st(mode_lib, *rq_param); -} - -void dml_rq_dlg_get_rq_reg(struct display_mode_lib *mode_lib, - display_rq_regs_st *rq_regs, - const display_pipe_source_params_st pipe_src_param) -{ - display_rq_params_st rq_param = {0}; - - memset(rq_regs, 0, sizeof(*rq_regs)); - dml_rq_dlg_get_rq_params(mode_lib, &rq_param, pipe_src_param); - extract_rq_regs(mode_lib, rq_regs, rq_param); - - print__rq_regs_st(mode_lib, *rq_regs); -} - -// Note: currently taken in as is. -// Nice to decouple code from hw register implement and extract code that are repeated for luma and chroma. -void dml_rq_dlg_get_dlg_params(struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *e2e_pipe_param, - const unsigned int num_pipes, - const unsigned int pipe_idx, - display_dlg_regs_st *disp_dlg_regs, - display_ttu_regs_st *disp_ttu_regs, - const display_rq_dlg_params_st rq_dlg_param, - const display_dlg_sys_params_st dlg_sys_param, - const bool cstate_en, - const bool pstate_en, - const bool vm_en, - const bool ignore_viewport_pos, - const bool immediate_flip_support) -{ - const display_pipe_source_params_st *src = &e2e_pipe_param[pipe_idx].pipe.src; - const display_pipe_dest_params_st *dst = &e2e_pipe_param[pipe_idx].pipe.dest; - const display_output_params_st *dout = &e2e_pipe_param[pipe_idx].dout; - const display_clocks_and_cfg_st *clks = &e2e_pipe_param[pipe_idx].clks_cfg; - const scaler_ratio_depth_st *scl = &e2e_pipe_param[pipe_idx].pipe.scale_ratio_depth; - const scaler_taps_st *taps = &e2e_pipe_param[pipe_idx].pipe.scale_taps; - - // ------------------------- - // Section 1.15.2.1: OTG dependent Params - // ------------------------- - // Timing - unsigned int htotal = dst->htotal; -// unsigned int hblank_start = dst.hblank_start; // TODO: Remove - unsigned int hblank_end = dst->hblank_end; - unsigned int vblank_start = dst->vblank_start; - unsigned int vblank_end = dst->vblank_end; - unsigned int min_vblank = mode_lib->ip.min_vblank_lines; - - double dppclk_freq_in_mhz = clks->dppclk_mhz; - double dispclk_freq_in_mhz = clks->dispclk_mhz; - double refclk_freq_in_mhz = clks->refclk_mhz; - double pclk_freq_in_mhz = dst->pixel_rate_mhz; - bool interlaced = dst->interlaced; - - double ref_freq_to_pix_freq = refclk_freq_in_mhz / pclk_freq_in_mhz; - - double min_dcfclk_mhz; - double t_calc_us; - double min_ttu_vblank; - - double min_dst_y_ttu_vblank; - unsigned int dlg_vblank_start; - bool dual_plane; - bool mode_422; - unsigned int access_dir; - unsigned int vp_height_l; - unsigned int vp_width_l; - unsigned int vp_height_c; - unsigned int vp_width_c; - - // Scaling - unsigned int htaps_l; - unsigned int htaps_c; - double hratio_l; - double hratio_c; - double vratio_l; - double vratio_c; - bool scl_enable; - - double line_time_in_us; - // double vinit_l; - // double vinit_c; - // double vinit_bot_l; - // double vinit_bot_c; - - // unsigned int swath_height_l; - unsigned int swath_width_ub_l; - // unsigned int dpte_bytes_per_row_ub_l; - unsigned int dpte_groups_per_row_ub_l; - // unsigned int meta_pte_bytes_per_frame_ub_l; - // unsigned int meta_bytes_per_row_ub_l; - - // unsigned int swath_height_c; - unsigned int swath_width_ub_c; - // unsigned int dpte_bytes_per_row_ub_c; - unsigned int dpte_groups_per_row_ub_c; - - unsigned int meta_chunks_per_row_ub_l; - unsigned int meta_chunks_per_row_ub_c; - unsigned int vupdate_offset; - unsigned int vupdate_width; - unsigned int vready_offset; - - unsigned int dppclk_delay_subtotal; - unsigned int dispclk_delay_subtotal; - unsigned int pixel_rate_delay_subtotal; - - unsigned int vstartup_start; - unsigned int dst_x_after_scaler; - unsigned int dst_y_after_scaler; - double line_wait; - double dst_y_prefetch; - double dst_y_per_vm_vblank; - double dst_y_per_row_vblank; - double dst_y_per_vm_flip; - double dst_y_per_row_flip; - double min_dst_y_per_vm_vblank; - double min_dst_y_per_row_vblank; - double lsw; - double vratio_pre_l; - double vratio_pre_c; - unsigned int req_per_swath_ub_l; - unsigned int req_per_swath_ub_c; - unsigned int meta_row_height_l; - unsigned int meta_row_height_c; - unsigned int swath_width_pixels_ub_l; - unsigned int swath_width_pixels_ub_c; - unsigned int scaler_rec_in_width_l; - unsigned int scaler_rec_in_width_c; - unsigned int dpte_row_height_l; - unsigned int dpte_row_height_c; - double hscale_pixel_rate_l; - double hscale_pixel_rate_c; - double min_hratio_fact_l; - double min_hratio_fact_c; - double refcyc_per_line_delivery_pre_l; - double refcyc_per_line_delivery_pre_c; - double refcyc_per_line_delivery_l; - double refcyc_per_line_delivery_c; - - double refcyc_per_req_delivery_pre_l; - double refcyc_per_req_delivery_pre_c; - double refcyc_per_req_delivery_l; - double refcyc_per_req_delivery_c; - - unsigned int full_recout_width; - double xfc_transfer_delay; - double xfc_precharge_delay; - double xfc_remote_surface_flip_latency; - double xfc_dst_y_delta_drq_limit; - double xfc_prefetch_margin; - double refcyc_per_req_delivery_pre_cur0; - double refcyc_per_req_delivery_cur0; - double refcyc_per_req_delivery_pre_cur1; - double refcyc_per_req_delivery_cur1; - - memset(disp_dlg_regs, 0, sizeof(*disp_dlg_regs)); - memset(disp_ttu_regs, 0, sizeof(*disp_ttu_regs)); - - dml_print("DML_DLG: %s: cstate_en = %d\n", __func__, cstate_en); - dml_print("DML_DLG: %s: pstate_en = %d\n", __func__, pstate_en); - dml_print("DML_DLG: %s: vm_en = %d\n", __func__, vm_en); - dml_print("DML_DLG: %s: ignore_viewport_pos = %d\n", __func__, ignore_viewport_pos); - dml_print("DML_DLG: %s: immediate_flip_support = %d\n", __func__, immediate_flip_support); - - dml_print("DML_DLG: %s: dppclk_freq_in_mhz = %3.2f\n", __func__, dppclk_freq_in_mhz); - dml_print("DML_DLG: %s: dispclk_freq_in_mhz = %3.2f\n", __func__, dispclk_freq_in_mhz); - dml_print("DML_DLG: %s: refclk_freq_in_mhz = %3.2f\n", __func__, refclk_freq_in_mhz); - dml_print("DML_DLG: %s: pclk_freq_in_mhz = %3.2f\n", __func__, pclk_freq_in_mhz); - dml_print("DML_DLG: %s: interlaced = %d\n", __func__, interlaced); - ASSERT(ref_freq_to_pix_freq < 4.0); - - disp_dlg_regs->ref_freq_to_pix_freq = - (unsigned int) (ref_freq_to_pix_freq * dml_pow(2, 19)); - disp_dlg_regs->refcyc_per_htotal = (unsigned int) (ref_freq_to_pix_freq * (double) htotal - * dml_pow(2, 8)); - disp_dlg_regs->dlg_vblank_end = interlaced ? (vblank_end / 2) : vblank_end; // 15 bits - disp_dlg_regs->refcyc_h_blank_end = (unsigned int) ((double) hblank_end - * (double) ref_freq_to_pix_freq); - ASSERT(disp_dlg_regs->refcyc_h_blank_end < (unsigned int) dml_pow(2, 13)); - - min_dcfclk_mhz = dlg_sys_param.deepsleep_dcfclk_mhz; - set_prefetch_mode(mode_lib, cstate_en, pstate_en, ignore_viewport_pos, immediate_flip_support); - t_calc_us = get_tcalc(mode_lib, e2e_pipe_param, num_pipes); - min_ttu_vblank = get_min_ttu_vblank(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - - min_dst_y_ttu_vblank = min_ttu_vblank * pclk_freq_in_mhz / (double) htotal; - dlg_vblank_start = interlaced ? (vblank_start / 2) : vblank_start; - - disp_dlg_regs->min_dst_y_next_start = (unsigned int) (((double) dlg_vblank_start - + min_dst_y_ttu_vblank) * dml_pow(2, 2)); - ASSERT(disp_dlg_regs->min_dst_y_next_start < (unsigned int) dml_pow(2, 18)); - - dml_print("DML_DLG: %s: min_dcfclk_mhz = %3.2f\n", - __func__, - min_dcfclk_mhz); - dml_print("DML_DLG: %s: min_ttu_vblank = %3.2f\n", - __func__, - min_ttu_vblank); - dml_print("DML_DLG: %s: min_dst_y_ttu_vblank = %3.2f\n", - __func__, - min_dst_y_ttu_vblank); - dml_print("DML_DLG: %s: t_calc_us = %3.2f\n", - __func__, - t_calc_us); - dml_print("DML_DLG: %s: disp_dlg_regs->min_dst_y_next_start = 0x%0x\n", - __func__, - disp_dlg_regs->min_dst_y_next_start); - dml_print("DML_DLG: %s: ref_freq_to_pix_freq = %3.2f\n", - __func__, - ref_freq_to_pix_freq); - - // ------------------------- - // Section 1.15.2.2: Prefetch, Active and TTU - // ------------------------- - // Prefetch Calc - // Source -// dcc_en = src.dcc; - dual_plane = is_dual_plane((enum source_format_class)(src->source_format)); - mode_422 = 0; // FIXME - access_dir = (src->source_scan == dm_vert); // vp access direction: horizontal or vertical accessed -// bytes_per_element_l = get_bytes_per_element(source_format_class(src.source_format), 0); -// bytes_per_element_c = get_bytes_per_element(source_format_class(src.source_format), 1); - vp_height_l = src->viewport_height; - vp_width_l = src->viewport_width; - vp_height_c = src->viewport_height_c; - vp_width_c = src->viewport_width_c; - - // Scaling - htaps_l = taps->htaps; - htaps_c = taps->htaps_c; - hratio_l = scl->hscl_ratio; - hratio_c = scl->hscl_ratio_c; - vratio_l = scl->vscl_ratio; - vratio_c = scl->vscl_ratio_c; - scl_enable = scl->scl_enable; - - line_time_in_us = (htotal / pclk_freq_in_mhz); -// vinit_l = scl.vinit; -// vinit_c = scl.vinit_c; -// vinit_bot_l = scl.vinit_bot; -// vinit_bot_c = scl.vinit_bot_c; - -// unsigned int swath_height_l = rq_dlg_param.rq_l.swath_height; - swath_width_ub_l = rq_dlg_param.rq_l.swath_width_ub; -// unsigned int dpte_bytes_per_row_ub_l = rq_dlg_param.rq_l.dpte_bytes_per_row_ub; - dpte_groups_per_row_ub_l = rq_dlg_param.rq_l.dpte_groups_per_row_ub; -// unsigned int meta_pte_bytes_per_frame_ub_l = rq_dlg_param.rq_l.meta_pte_bytes_per_frame_ub; -// unsigned int meta_bytes_per_row_ub_l = rq_dlg_param.rq_l.meta_bytes_per_row_ub; - -// unsigned int swath_height_c = rq_dlg_param.rq_c.swath_height; - swath_width_ub_c = rq_dlg_param.rq_c.swath_width_ub; - // dpte_bytes_per_row_ub_c = rq_dlg_param.rq_c.dpte_bytes_per_row_ub; - dpte_groups_per_row_ub_c = rq_dlg_param.rq_c.dpte_groups_per_row_ub; - - meta_chunks_per_row_ub_l = rq_dlg_param.rq_l.meta_chunks_per_row_ub; - meta_chunks_per_row_ub_c = rq_dlg_param.rq_c.meta_chunks_per_row_ub; - vupdate_offset = dst->vupdate_offset; - vupdate_width = dst->vupdate_width; - vready_offset = dst->vready_offset; - - dppclk_delay_subtotal = mode_lib->ip.dppclk_delay_subtotal; - dispclk_delay_subtotal = mode_lib->ip.dispclk_delay_subtotal; - - if (scl_enable) - dppclk_delay_subtotal += mode_lib->ip.dppclk_delay_scl; - else - dppclk_delay_subtotal += mode_lib->ip.dppclk_delay_scl_lb_only; - - dppclk_delay_subtotal += mode_lib->ip.dppclk_delay_cnvc_formatter - + src->num_cursors * mode_lib->ip.dppclk_delay_cnvc_cursor; - - if (dout->dsc_enable) { - double dsc_delay = get_dsc_delay(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - - dispclk_delay_subtotal += dsc_delay; - } - - pixel_rate_delay_subtotal = dppclk_delay_subtotal * pclk_freq_in_mhz / dppclk_freq_in_mhz - + dispclk_delay_subtotal * pclk_freq_in_mhz / dispclk_freq_in_mhz; - - vstartup_start = dst->vstartup_start; - if (interlaced) { - if (vstartup_start / 2.0 - - (double) (vready_offset + vupdate_width + vupdate_offset) / htotal - <= vblank_end / 2.0) - disp_dlg_regs->vready_after_vcount0 = 1; - else - disp_dlg_regs->vready_after_vcount0 = 0; - } else { - if (vstartup_start - - (double) (vready_offset + vupdate_width + vupdate_offset) / htotal - <= vblank_end) - disp_dlg_regs->vready_after_vcount0 = 1; - else - disp_dlg_regs->vready_after_vcount0 = 0; - } - - // TODO: Where is this coming from? - if (interlaced) - vstartup_start = vstartup_start / 2; - - // TODO: What if this min_vblank doesn't match the value in the dml_config_settings.cpp? - if (vstartup_start >= min_vblank) { - dml_print("WARNING: DML_DLG: %s: vblank_start=%d vblank_end=%d\n", - __func__, - vblank_start, - vblank_end); - dml_print("WARNING: DML_DLG: %s: vstartup_start=%d should be less than min_vblank=%d\n", - __func__, - vstartup_start, - min_vblank); - min_vblank = vstartup_start + 1; - dml_print("WARNING: DML_DLG: %s: vstartup_start=%d should be less than min_vblank=%d\n", - __func__, - vstartup_start, - min_vblank); - } - - dst_x_after_scaler = get_dst_x_after_scaler(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - dst_y_after_scaler = get_dst_y_after_scaler(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - - dml_print("DML_DLG: %s: htotal = %d\n", __func__, htotal); - dml_print("DML_DLG: %s: pixel_rate_delay_subtotal = %d\n", - __func__, - pixel_rate_delay_subtotal); - dml_print("DML_DLG: %s: dst_x_after_scaler = %d\n", - __func__, - dst_x_after_scaler); - dml_print("DML_DLG: %s: dst_y_after_scaler = %d\n", - __func__, - dst_y_after_scaler); - - // Lwait - line_wait = mode_lib->soc.urgent_latency_us; - if (cstate_en) - line_wait = dml_max(mode_lib->soc.sr_enter_plus_exit_time_us, line_wait); - if (pstate_en) - line_wait = dml_max(mode_lib->soc.dram_clock_change_latency_us - + mode_lib->soc.urgent_latency_us, - line_wait); - line_wait = line_wait / line_time_in_us; - - dst_y_prefetch = get_dst_y_prefetch(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - dml_print("DML_DLG: %s: dst_y_prefetch (after rnd) = %3.2f\n", __func__, dst_y_prefetch); - - dst_y_per_vm_vblank = get_dst_y_per_vm_vblank(mode_lib, - e2e_pipe_param, - num_pipes, - pipe_idx); - dst_y_per_row_vblank = get_dst_y_per_row_vblank(mode_lib, - e2e_pipe_param, - num_pipes, - pipe_idx); - dst_y_per_vm_flip = get_dst_y_per_vm_flip(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - dst_y_per_row_flip = get_dst_y_per_row_flip(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - - min_dst_y_per_vm_vblank = 8.0; - min_dst_y_per_row_vblank = 16.0; - - // magic! - if (htotal <= 75) { - min_vblank = 300; - min_dst_y_per_vm_vblank = 100.0; - min_dst_y_per_row_vblank = 100.0; - } - - dml_print("DML_DLG: %s: dst_y_per_vm_vblank = %3.2f\n", __func__, dst_y_per_vm_vblank); - dml_print("DML_DLG: %s: dst_y_per_row_vblank = %3.2f\n", __func__, dst_y_per_row_vblank); - - ASSERT(dst_y_per_vm_vblank < min_dst_y_per_vm_vblank); - ASSERT(dst_y_per_row_vblank < min_dst_y_per_row_vblank); - - ASSERT(dst_y_prefetch > (dst_y_per_vm_vblank + dst_y_per_row_vblank)); - lsw = dst_y_prefetch - (dst_y_per_vm_vblank + dst_y_per_row_vblank); - - dml_print("DML_DLG: %s: lsw = %3.2f\n", __func__, lsw); - - vratio_pre_l = get_vratio_prefetch_l(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - vratio_pre_c = get_vratio_prefetch_c(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - - dml_print("DML_DLG: %s: vratio_pre_l=%3.2f\n", __func__, vratio_pre_l); - dml_print("DML_DLG: %s: vratio_pre_c=%3.2f\n", __func__, vratio_pre_c); - - // Active - req_per_swath_ub_l = rq_dlg_param.rq_l.req_per_swath_ub; - req_per_swath_ub_c = rq_dlg_param.rq_c.req_per_swath_ub; - meta_row_height_l = rq_dlg_param.rq_l.meta_row_height; - meta_row_height_c = rq_dlg_param.rq_c.meta_row_height; - swath_width_pixels_ub_l = 0; - swath_width_pixels_ub_c = 0; - scaler_rec_in_width_l = 0; - scaler_rec_in_width_c = 0; - dpte_row_height_l = rq_dlg_param.rq_l.dpte_row_height; - dpte_row_height_c = rq_dlg_param.rq_c.dpte_row_height; - - if (mode_422) { - swath_width_pixels_ub_l = swath_width_ub_l * 2; // *2 for 2 pixel per element - swath_width_pixels_ub_c = swath_width_ub_c * 2; - } else { - swath_width_pixels_ub_l = swath_width_ub_l * 1; - swath_width_pixels_ub_c = swath_width_ub_c * 1; - } - - hscale_pixel_rate_l = 0.; - hscale_pixel_rate_c = 0.; - min_hratio_fact_l = 1.0; - min_hratio_fact_c = 1.0; - - if (htaps_l <= 1) - min_hratio_fact_l = 2.0; - else if (htaps_l <= 6) { - if ((hratio_l * 2.0) > 4.0) - min_hratio_fact_l = 4.0; - else - min_hratio_fact_l = hratio_l * 2.0; - } else { - if (hratio_l > 4.0) - min_hratio_fact_l = 4.0; - else - min_hratio_fact_l = hratio_l; - } - - hscale_pixel_rate_l = min_hratio_fact_l * dppclk_freq_in_mhz; - - if (htaps_c <= 1) - min_hratio_fact_c = 2.0; - else if (htaps_c <= 6) { - if ((hratio_c * 2.0) > 4.0) - min_hratio_fact_c = 4.0; - else - min_hratio_fact_c = hratio_c * 2.0; - } else { - if (hratio_c > 4.0) - min_hratio_fact_c = 4.0; - else - min_hratio_fact_c = hratio_c; - } - - hscale_pixel_rate_c = min_hratio_fact_c * dppclk_freq_in_mhz; - - refcyc_per_line_delivery_pre_l = 0.; - refcyc_per_line_delivery_pre_c = 0.; - refcyc_per_line_delivery_l = 0.; - refcyc_per_line_delivery_c = 0.; - - refcyc_per_req_delivery_pre_l = 0.; - refcyc_per_req_delivery_pre_c = 0.; - refcyc_per_req_delivery_l = 0.; - refcyc_per_req_delivery_c = 0.; - - full_recout_width = 0; - // In ODM - if (src->is_hsplit) { - // This "hack" is only allowed (and valid) for MPC combine. In ODM - // combine, you MUST specify the full_recout_width...according to Oswin - if (dst->full_recout_width == 0 && !dst->odm_combine) { - dml_print("DML_DLG: %s: Warning: full_recout_width not set in hsplit mode\n", - __func__); - full_recout_width = dst->recout_width * 2; // assume half split for dcn1 - } else - full_recout_width = dst->full_recout_width; - } else - full_recout_width = dst->recout_width; - - // mpc_combine and odm_combine are mutually exclusive - refcyc_per_line_delivery_pre_l = get_refcyc_per_delivery(mode_lib, - refclk_freq_in_mhz, - pclk_freq_in_mhz, - dst->odm_combine, - full_recout_width, - dst->hactive, - vratio_pre_l, - hscale_pixel_rate_l, - swath_width_pixels_ub_l, - 1); // per line - - refcyc_per_line_delivery_l = get_refcyc_per_delivery(mode_lib, - refclk_freq_in_mhz, - pclk_freq_in_mhz, - dst->odm_combine, - full_recout_width, - dst->hactive, - vratio_l, - hscale_pixel_rate_l, - swath_width_pixels_ub_l, - 1); // per line - - dml_print("DML_DLG: %s: full_recout_width = %d\n", - __func__, - full_recout_width); - dml_print("DML_DLG: %s: hscale_pixel_rate_l = %3.2f\n", - __func__, - hscale_pixel_rate_l); - dml_print("DML_DLG: %s: refcyc_per_line_delivery_pre_l = %3.2f\n", - __func__, - refcyc_per_line_delivery_pre_l); - dml_print("DML_DLG: %s: refcyc_per_line_delivery_l = %3.2f\n", - __func__, - refcyc_per_line_delivery_l); - - if (dual_plane) { - refcyc_per_line_delivery_pre_c = get_refcyc_per_delivery(mode_lib, - refclk_freq_in_mhz, - pclk_freq_in_mhz, - dst->odm_combine, - full_recout_width, - dst->hactive, - vratio_pre_c, - hscale_pixel_rate_c, - swath_width_pixels_ub_c, - 1); // per line - - refcyc_per_line_delivery_c = get_refcyc_per_delivery(mode_lib, - refclk_freq_in_mhz, - pclk_freq_in_mhz, - dst->odm_combine, - full_recout_width, - dst->hactive, - vratio_c, - hscale_pixel_rate_c, - swath_width_pixels_ub_c, - 1); // per line - - dml_print("DML_DLG: %s: refcyc_per_line_delivery_pre_c = %3.2f\n", - __func__, - refcyc_per_line_delivery_pre_c); - dml_print("DML_DLG: %s: refcyc_per_line_delivery_c = %3.2f\n", - __func__, - refcyc_per_line_delivery_c); - } - - // TTU - Luma / Chroma - if (access_dir) { // vertical access - scaler_rec_in_width_l = vp_height_l; - scaler_rec_in_width_c = vp_height_c; - } else { - scaler_rec_in_width_l = vp_width_l; - scaler_rec_in_width_c = vp_width_c; - } - - refcyc_per_req_delivery_pre_l = get_refcyc_per_delivery(mode_lib, - refclk_freq_in_mhz, - pclk_freq_in_mhz, - dst->odm_combine, - full_recout_width, - dst->hactive, - vratio_pre_l, - hscale_pixel_rate_l, - scaler_rec_in_width_l, - req_per_swath_ub_l); // per req - refcyc_per_req_delivery_l = get_refcyc_per_delivery(mode_lib, - refclk_freq_in_mhz, - pclk_freq_in_mhz, - dst->odm_combine, - full_recout_width, - dst->hactive, - vratio_l, - hscale_pixel_rate_l, - scaler_rec_in_width_l, - req_per_swath_ub_l); // per req - - dml_print("DML_DLG: %s: refcyc_per_req_delivery_pre_l = %3.2f\n", - __func__, - refcyc_per_req_delivery_pre_l); - dml_print("DML_DLG: %s: refcyc_per_req_delivery_l = %3.2f\n", - __func__, - refcyc_per_req_delivery_l); - - ASSERT(refcyc_per_req_delivery_pre_l < dml_pow(2, 13)); - ASSERT(refcyc_per_req_delivery_l < dml_pow(2, 13)); - - if (dual_plane) { - refcyc_per_req_delivery_pre_c = get_refcyc_per_delivery(mode_lib, - refclk_freq_in_mhz, - pclk_freq_in_mhz, - dst->odm_combine, - full_recout_width, - dst->hactive, - vratio_pre_c, - hscale_pixel_rate_c, - scaler_rec_in_width_c, - req_per_swath_ub_c); // per req - refcyc_per_req_delivery_c = get_refcyc_per_delivery(mode_lib, - refclk_freq_in_mhz, - pclk_freq_in_mhz, - dst->odm_combine, - full_recout_width, - dst->hactive, - vratio_c, - hscale_pixel_rate_c, - scaler_rec_in_width_c, - req_per_swath_ub_c); // per req - - dml_print("DML_DLG: %s: refcyc_per_req_delivery_pre_c = %3.2f\n", - __func__, - refcyc_per_req_delivery_pre_c); - dml_print("DML_DLG: %s: refcyc_per_req_delivery_c = %3.2f\n", - __func__, - refcyc_per_req_delivery_c); - - ASSERT(refcyc_per_req_delivery_pre_c < dml_pow(2, 13)); - ASSERT(refcyc_per_req_delivery_c < dml_pow(2, 13)); - } - - // XFC - xfc_transfer_delay = get_xfc_transfer_delay(mode_lib, e2e_pipe_param, num_pipes, pipe_idx); - xfc_precharge_delay = get_xfc_precharge_delay(mode_lib, - e2e_pipe_param, - num_pipes, - pipe_idx); - xfc_remote_surface_flip_latency = get_xfc_remote_surface_flip_latency(mode_lib, - e2e_pipe_param, - num_pipes, - pipe_idx); - xfc_dst_y_delta_drq_limit = xfc_remote_surface_flip_latency; - xfc_prefetch_margin = get_xfc_prefetch_margin(mode_lib, - e2e_pipe_param, - num_pipes, - pipe_idx); - - // TTU - Cursor - refcyc_per_req_delivery_pre_cur0 = 0.0; - refcyc_per_req_delivery_cur0 = 0.0; - if (src->num_cursors > 0) { - calculate_ttu_cursor(mode_lib, - &refcyc_per_req_delivery_pre_cur0, - &refcyc_per_req_delivery_cur0, - refclk_freq_in_mhz, - ref_freq_to_pix_freq, - hscale_pixel_rate_l, - scl->hscl_ratio, - vratio_pre_l, - vratio_l, - src->cur0_src_width, - (enum cursor_bpp)(src->cur0_bpp)); - } - - refcyc_per_req_delivery_pre_cur1 = 0.0; - refcyc_per_req_delivery_cur1 = 0.0; - if (src->num_cursors > 1) { - calculate_ttu_cursor(mode_lib, - &refcyc_per_req_delivery_pre_cur1, - &refcyc_per_req_delivery_cur1, - refclk_freq_in_mhz, - ref_freq_to_pix_freq, - hscale_pixel_rate_l, - scl->hscl_ratio, - vratio_pre_l, - vratio_l, - src->cur1_src_width, - (enum cursor_bpp)(src->cur1_bpp)); - } - - // TTU - Misc - // all hard-coded - - // Assignment to register structures - disp_dlg_regs->dst_y_after_scaler = dst_y_after_scaler; // in terms of line - disp_dlg_regs->refcyc_x_after_scaler = dst_x_after_scaler * ref_freq_to_pix_freq; // in terms of refclk - ASSERT(disp_dlg_regs->refcyc_x_after_scaler < (unsigned int) dml_pow(2, 13)); - disp_dlg_regs->dst_y_prefetch = (unsigned int) (dst_y_prefetch * dml_pow(2, 2)); - disp_dlg_regs->dst_y_per_vm_vblank = (unsigned int) (dst_y_per_vm_vblank * dml_pow(2, 2)); - disp_dlg_regs->dst_y_per_row_vblank = (unsigned int) (dst_y_per_row_vblank * dml_pow(2, 2)); - disp_dlg_regs->dst_y_per_vm_flip = (unsigned int) (dst_y_per_vm_flip * dml_pow(2, 2)); - disp_dlg_regs->dst_y_per_row_flip = (unsigned int) (dst_y_per_row_flip * dml_pow(2, 2)); - - disp_dlg_regs->vratio_prefetch = (unsigned int) (vratio_pre_l * dml_pow(2, 19)); - disp_dlg_regs->vratio_prefetch_c = (unsigned int) (vratio_pre_c * dml_pow(2, 19)); - - disp_dlg_regs->refcyc_per_pte_group_vblank_l = - (unsigned int) (dst_y_per_row_vblank * (double) htotal - * ref_freq_to_pix_freq / (double) dpte_groups_per_row_ub_l); - ASSERT(disp_dlg_regs->refcyc_per_pte_group_vblank_l < (unsigned int) dml_pow(2, 13)); - - if (dual_plane) { - disp_dlg_regs->refcyc_per_pte_group_vblank_c = (unsigned int) (dst_y_per_row_vblank - * (double) htotal * ref_freq_to_pix_freq - / (double) dpte_groups_per_row_ub_c); - ASSERT(disp_dlg_regs->refcyc_per_pte_group_vblank_c - < (unsigned int) dml_pow(2, 13)); - } - - disp_dlg_regs->refcyc_per_meta_chunk_vblank_l = - (unsigned int) (dst_y_per_row_vblank * (double) htotal - * ref_freq_to_pix_freq / (double) meta_chunks_per_row_ub_l); - ASSERT(disp_dlg_regs->refcyc_per_meta_chunk_vblank_l < (unsigned int) dml_pow(2, 13)); - - disp_dlg_regs->refcyc_per_meta_chunk_vblank_c = - disp_dlg_regs->refcyc_per_meta_chunk_vblank_l; // dcc for 4:2:0 is not supported in dcn1.0. assigned to be the same as _l for now - - disp_dlg_regs->refcyc_per_pte_group_flip_l = (unsigned int) (dst_y_per_row_flip * htotal - * ref_freq_to_pix_freq) / dpte_groups_per_row_ub_l; - disp_dlg_regs->refcyc_per_meta_chunk_flip_l = (unsigned int) (dst_y_per_row_flip * htotal - * ref_freq_to_pix_freq) / meta_chunks_per_row_ub_l; - - if (dual_plane) { - disp_dlg_regs->refcyc_per_pte_group_flip_c = (unsigned int) (dst_y_per_row_flip - * htotal * ref_freq_to_pix_freq) / dpte_groups_per_row_ub_c; - disp_dlg_regs->refcyc_per_meta_chunk_flip_c = (unsigned int) (dst_y_per_row_flip - * htotal * ref_freq_to_pix_freq) / meta_chunks_per_row_ub_c; - } - - disp_dlg_regs->dst_y_per_pte_row_nom_l = (unsigned int) ((double) dpte_row_height_l - / (double) vratio_l * dml_pow(2, 2)); - ASSERT(disp_dlg_regs->dst_y_per_pte_row_nom_l < (unsigned int) dml_pow(2, 17)); - - if (dual_plane) { - disp_dlg_regs->dst_y_per_pte_row_nom_c = (unsigned int) ((double) dpte_row_height_c - / (double) vratio_c * dml_pow(2, 2)); - if (disp_dlg_regs->dst_y_per_pte_row_nom_c >= (unsigned int) dml_pow(2, 17)) { - dml_print("DML_DLG: %s: Warning dst_y_per_pte_row_nom_c %u larger than supported by register format U15.2 %u\n", - __func__, - disp_dlg_regs->dst_y_per_pte_row_nom_c, - (unsigned int) dml_pow(2, 17) - 1); - } - } - - disp_dlg_regs->dst_y_per_meta_row_nom_l = (unsigned int) ((double) meta_row_height_l - / (double) vratio_l * dml_pow(2, 2)); - ASSERT(disp_dlg_regs->dst_y_per_meta_row_nom_l < (unsigned int) dml_pow(2, 17)); - - disp_dlg_regs->dst_y_per_meta_row_nom_c = disp_dlg_regs->dst_y_per_meta_row_nom_l; // TODO: dcc for 4:2:0 is not supported in dcn1.0. assigned to be the same as _l for now - - disp_dlg_regs->refcyc_per_pte_group_nom_l = (unsigned int) ((double) dpte_row_height_l - / (double) vratio_l * (double) htotal * ref_freq_to_pix_freq - / (double) dpte_groups_per_row_ub_l); - if (disp_dlg_regs->refcyc_per_pte_group_nom_l >= (unsigned int) dml_pow(2, 23)) - disp_dlg_regs->refcyc_per_pte_group_nom_l = dml_pow(2, 23) - 1; - disp_dlg_regs->refcyc_per_meta_chunk_nom_l = (unsigned int) ((double) meta_row_height_l - / (double) vratio_l * (double) htotal * ref_freq_to_pix_freq - / (double) meta_chunks_per_row_ub_l); - if (disp_dlg_regs->refcyc_per_meta_chunk_nom_l >= (unsigned int) dml_pow(2, 23)) - disp_dlg_regs->refcyc_per_meta_chunk_nom_l = dml_pow(2, 23) - 1; - - if (dual_plane) { - disp_dlg_regs->refcyc_per_pte_group_nom_c = - (unsigned int) ((double) dpte_row_height_c / (double) vratio_c - * (double) htotal * ref_freq_to_pix_freq - / (double) dpte_groups_per_row_ub_c); - if (disp_dlg_regs->refcyc_per_pte_group_nom_c >= (unsigned int) dml_pow(2, 23)) - disp_dlg_regs->refcyc_per_pte_group_nom_c = dml_pow(2, 23) - 1; - - // TODO: Is this the right calculation? Does htotal need to be halved? - disp_dlg_regs->refcyc_per_meta_chunk_nom_c = - (unsigned int) ((double) meta_row_height_c / (double) vratio_c - * (double) htotal * ref_freq_to_pix_freq - / (double) meta_chunks_per_row_ub_c); - if (disp_dlg_regs->refcyc_per_meta_chunk_nom_c >= (unsigned int) dml_pow(2, 23)) - disp_dlg_regs->refcyc_per_meta_chunk_nom_c = dml_pow(2, 23) - 1; - } - - disp_dlg_regs->refcyc_per_line_delivery_pre_l = (unsigned int) dml_floor(refcyc_per_line_delivery_pre_l, - 1); - disp_dlg_regs->refcyc_per_line_delivery_l = (unsigned int) dml_floor(refcyc_per_line_delivery_l, - 1); - ASSERT(disp_dlg_regs->refcyc_per_line_delivery_pre_l < (unsigned int) dml_pow(2, 13)); - ASSERT(disp_dlg_regs->refcyc_per_line_delivery_l < (unsigned int) dml_pow(2, 13)); - - disp_dlg_regs->refcyc_per_line_delivery_pre_c = (unsigned int) dml_floor(refcyc_per_line_delivery_pre_c, - 1); - disp_dlg_regs->refcyc_per_line_delivery_c = (unsigned int) dml_floor(refcyc_per_line_delivery_c, - 1); - ASSERT(disp_dlg_regs->refcyc_per_line_delivery_pre_c < (unsigned int) dml_pow(2, 13)); - ASSERT(disp_dlg_regs->refcyc_per_line_delivery_c < (unsigned int) dml_pow(2, 13)); - - disp_dlg_regs->chunk_hdl_adjust_cur0 = 3; - disp_dlg_regs->dst_y_offset_cur0 = 0; - disp_dlg_regs->chunk_hdl_adjust_cur1 = 3; - disp_dlg_regs->dst_y_offset_cur1 = 0; - - disp_dlg_regs->xfc_reg_transfer_delay = xfc_transfer_delay; - disp_dlg_regs->xfc_reg_precharge_delay = xfc_precharge_delay; - disp_dlg_regs->xfc_reg_remote_surface_flip_latency = xfc_remote_surface_flip_latency; - disp_dlg_regs->xfc_reg_prefetch_margin = dml_ceil(xfc_prefetch_margin * refclk_freq_in_mhz, - 1); - - // slave has to have this value also set to off - if (src->xfc_enable && !src->xfc_slave) - disp_dlg_regs->dst_y_delta_drq_limit = dml_ceil(xfc_dst_y_delta_drq_limit, 1); - else - disp_dlg_regs->dst_y_delta_drq_limit = 0x7fff; // off - - disp_ttu_regs->refcyc_per_req_delivery_pre_l = (unsigned int) (refcyc_per_req_delivery_pre_l - * dml_pow(2, 10)); - disp_ttu_regs->refcyc_per_req_delivery_l = (unsigned int) (refcyc_per_req_delivery_l - * dml_pow(2, 10)); - disp_ttu_regs->refcyc_per_req_delivery_pre_c = (unsigned int) (refcyc_per_req_delivery_pre_c - * dml_pow(2, 10)); - disp_ttu_regs->refcyc_per_req_delivery_c = (unsigned int) (refcyc_per_req_delivery_c - * dml_pow(2, 10)); - disp_ttu_regs->refcyc_per_req_delivery_pre_cur0 = - (unsigned int) (refcyc_per_req_delivery_pre_cur0 * dml_pow(2, 10)); - disp_ttu_regs->refcyc_per_req_delivery_cur0 = (unsigned int) (refcyc_per_req_delivery_cur0 - * dml_pow(2, 10)); - disp_ttu_regs->refcyc_per_req_delivery_pre_cur1 = - (unsigned int) (refcyc_per_req_delivery_pre_cur1 * dml_pow(2, 10)); - disp_ttu_regs->refcyc_per_req_delivery_cur1 = (unsigned int) (refcyc_per_req_delivery_cur1 - * dml_pow(2, 10)); - disp_ttu_regs->qos_level_low_wm = 0; - ASSERT(disp_ttu_regs->qos_level_low_wm < dml_pow(2, 14)); - disp_ttu_regs->qos_level_high_wm = (unsigned int) (4.0 * (double) htotal - * ref_freq_to_pix_freq); - ASSERT(disp_ttu_regs->qos_level_high_wm < dml_pow(2, 14)); - - disp_ttu_regs->qos_level_flip = 14; - disp_ttu_regs->qos_level_fixed_l = 8; - disp_ttu_regs->qos_level_fixed_c = 8; - disp_ttu_regs->qos_level_fixed_cur0 = 8; - disp_ttu_regs->qos_ramp_disable_l = 0; - disp_ttu_regs->qos_ramp_disable_c = 0; - disp_ttu_regs->qos_ramp_disable_cur0 = 0; - - disp_ttu_regs->min_ttu_vblank = min_ttu_vblank * refclk_freq_in_mhz; - ASSERT(disp_ttu_regs->min_ttu_vblank < dml_pow(2, 24)); - - print__ttu_regs_st(mode_lib, *disp_ttu_regs); - print__dlg_regs_st(mode_lib, *disp_dlg_regs); -} - -void dml_rq_dlg_get_dlg_reg(struct display_mode_lib *mode_lib, - display_dlg_regs_st *dlg_regs, - display_ttu_regs_st *ttu_regs, - display_e2e_pipe_params_st *e2e_pipe_param, - const unsigned int num_pipes, - const unsigned int pipe_idx, - const bool cstate_en, - const bool pstate_en, - const bool vm_en, - const bool ignore_viewport_pos, - const bool immediate_flip_support) -{ - display_rq_params_st rq_param = {0}; - display_dlg_sys_params_st dlg_sys_param = {0}; - - // Get watermark and Tex. - dlg_sys_param.t_urg_wm_us = get_wm_urgent(mode_lib, e2e_pipe_param, num_pipes); - dlg_sys_param.deepsleep_dcfclk_mhz = get_clk_dcf_deepsleep(mode_lib, - e2e_pipe_param, - num_pipes); - dlg_sys_param.t_extra_us = get_urgent_extra_latency(mode_lib, e2e_pipe_param, num_pipes); - dlg_sys_param.mem_trip_us = get_wm_memory_trip(mode_lib, e2e_pipe_param, num_pipes); - dlg_sys_param.t_mclk_wm_us = get_wm_dram_clock_change(mode_lib, e2e_pipe_param, num_pipes); - dlg_sys_param.t_sr_wm_us = get_wm_stutter_enter_exit(mode_lib, e2e_pipe_param, num_pipes); - dlg_sys_param.total_flip_bw = get_total_immediate_flip_bw(mode_lib, - e2e_pipe_param, - num_pipes); - dlg_sys_param.total_flip_bytes = get_total_immediate_flip_bytes(mode_lib, - e2e_pipe_param, - num_pipes); - dlg_sys_param.t_srx_delay_us = mode_lib->ip.dcfclk_cstate_latency - / dlg_sys_param.deepsleep_dcfclk_mhz; // TODO: Deprecated - - print__dlg_sys_params_st(mode_lib, dlg_sys_param); - - // system parameter calculation done - - dml_print("DML_DLG: Calculation for pipe[%d] start\n\n", pipe_idx); - dml_rq_dlg_get_rq_params(mode_lib, &rq_param, e2e_pipe_param[pipe_idx].pipe.src); - dml_rq_dlg_get_dlg_params(mode_lib, - e2e_pipe_param, - num_pipes, - pipe_idx, - dlg_regs, - ttu_regs, - rq_param.dlg, - dlg_sys_param, - cstate_en, - pstate_en, - vm_en, - ignore_viewport_pos, - immediate_flip_support); - dml_print("DML_DLG: Calculation for pipe[%d] end\n", pipe_idx); -} - -void dml_rq_dlg_get_arb_params(struct display_mode_lib *mode_lib, display_arb_params_st *arb_param) -{ - memset(arb_param, 0, sizeof(*arb_param)); - arb_param->max_req_outstanding = 256; - arb_param->min_req_outstanding = 68; - arb_param->sat_level_us = 60; -} - -void calculate_ttu_cursor(struct display_mode_lib *mode_lib, - double *refcyc_per_req_delivery_pre_cur, - double *refcyc_per_req_delivery_cur, - double refclk_freq_in_mhz, - double ref_freq_to_pix_freq, - double hscale_pixel_rate_l, - double hscl_ratio, - double vratio_pre_l, - double vratio_l, - unsigned int cur_width, - enum cursor_bpp cur_bpp) -{ - unsigned int cur_src_width = cur_width; - unsigned int cur_req_size = 0; - unsigned int cur_req_width = 0; - double cur_width_ub = 0.0; - double cur_req_per_width = 0.0; - double hactive_cur = 0.0; - - ASSERT(cur_src_width <= 256); - - *refcyc_per_req_delivery_pre_cur = 0.0; - *refcyc_per_req_delivery_cur = 0.0; - if (cur_src_width > 0) { - unsigned int cur_bit_per_pixel = 0; - - if (cur_bpp == dm_cur_2bit) { - cur_req_size = 64; // byte - cur_bit_per_pixel = 2; - } else { // 32bit - cur_bit_per_pixel = 32; - if (cur_src_width >= 1 && cur_src_width <= 16) - cur_req_size = 64; - else if (cur_src_width >= 17 && cur_src_width <= 31) - cur_req_size = 128; - else - cur_req_size = 256; - } - - cur_req_width = (double) cur_req_size / ((double) cur_bit_per_pixel / 8.0); - cur_width_ub = dml_ceil((double) cur_src_width / (double) cur_req_width, 1) - * (double) cur_req_width; - cur_req_per_width = cur_width_ub / (double) cur_req_width; - hactive_cur = (double) cur_src_width / hscl_ratio; // FIXME: oswin to think about what to do for cursor - - if (vratio_pre_l <= 1.0) { - *refcyc_per_req_delivery_pre_cur = hactive_cur * ref_freq_to_pix_freq - / (double) cur_req_per_width; - } else { - *refcyc_per_req_delivery_pre_cur = (double) refclk_freq_in_mhz - * (double) cur_src_width / hscale_pixel_rate_l - / (double) cur_req_per_width; - } - - ASSERT(*refcyc_per_req_delivery_pre_cur < dml_pow(2, 13)); - - if (vratio_l <= 1.0) { - *refcyc_per_req_delivery_cur = hactive_cur * ref_freq_to_pix_freq - / (double) cur_req_per_width; - } else { - *refcyc_per_req_delivery_cur = (double) refclk_freq_in_mhz - * (double) cur_src_width / hscale_pixel_rate_l - / (double) cur_req_per_width; - } - - dml_print("DML_DLG: %s: cur_req_width = %d\n", - __func__, - cur_req_width); - dml_print("DML_DLG: %s: cur_width_ub = %3.2f\n", - __func__, - cur_width_ub); - dml_print("DML_DLG: %s: cur_req_per_width = %3.2f\n", - __func__, - cur_req_per_width); - dml_print("DML_DLG: %s: hactive_cur = %3.2f\n", - __func__, - hactive_cur); - dml_print("DML_DLG: %s: refcyc_per_req_delivery_pre_cur = %3.2f\n", - __func__, - *refcyc_per_req_delivery_pre_cur); - dml_print("DML_DLG: %s: refcyc_per_req_delivery_cur = %3.2f\n", - __func__, - *refcyc_per_req_delivery_cur); - - ASSERT(*refcyc_per_req_delivery_cur < dml_pow(2, 13)); - } -} - -unsigned int dml_rq_dlg_get_calculated_vstartup(struct display_mode_lib *mode_lib, - display_e2e_pipe_params_st *e2e_pipe_param, - const unsigned int num_pipes, - const unsigned int pipe_idx) -{ - unsigned int vstartup_pipe[DC__NUM_PIPES__MAX]; - bool visited[DC__NUM_PIPES__MAX]; - unsigned int pipe_inst = 0; - unsigned int i, j, k; - - for (k = 0; k < num_pipes; ++k) - visited[k] = false; - - for (i = 0; i < num_pipes; i++) { - if (e2e_pipe_param[i].pipe.src.is_hsplit && !visited[i]) { - unsigned int grp = e2e_pipe_param[i].pipe.src.hsplit_grp; - - for (j = i; j < num_pipes; j++) { - if (e2e_pipe_param[j].pipe.src.hsplit_grp == grp - && e2e_pipe_param[j].pipe.src.is_hsplit - && !visited[j]) { - vstartup_pipe[j] = get_vstartup_calculated(mode_lib, - e2e_pipe_param, - num_pipes, - pipe_inst); - visited[j] = true; - } - } - - pipe_inst++; - } - - if (!visited[i]) { - vstartup_pipe[i] = get_vstartup_calculated(mode_lib, - e2e_pipe_param, - num_pipes, - pipe_inst); - visited[i] = true; - pipe_inst++; - } - } - - return vstartup_pipe[pipe_idx]; - -} - -void dml_rq_dlg_get_row_heights(struct display_mode_lib *mode_lib, - unsigned int *o_dpte_row_height, - unsigned int *o_meta_row_height, - unsigned int vp_width, - unsigned int data_pitch, - int source_format, - int tiling, - int macro_tile_size, - int source_scan, - int is_chroma) -{ - display_data_rq_dlg_params_st rq_dlg_param; - display_data_rq_misc_params_st rq_misc_param; - display_data_rq_sizing_params_st rq_sizing_param; - - get_meta_and_pte_attr(mode_lib, - &rq_dlg_param, - &rq_misc_param, - &rq_sizing_param, - vp_width, - 0, // dummy - data_pitch, - 0, // dummy - source_format, - tiling, - macro_tile_size, - source_scan, - is_chroma); - - *o_dpte_row_height = rq_dlg_param.dpte_row_height; - *o_meta_row_height = rq_dlg_param.meta_row_height; -}
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_calc.h b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_calc.h deleted file mode 100644 index efdd4c73..0000000 --- a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_calc.h +++ /dev/null
@@ -1,148 +0,0 @@ -/* - * Copyright 2017 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ - -#ifndef __DML2_DISPLAY_RQ_DLG_CALC_H__ -#define __DML2_DISPLAY_RQ_DLG_CALC_H__ - -#include "dml_common_defs.h" -#include "display_rq_dlg_helpers.h" - -struct display_mode_lib; - -// Function: dml_rq_dlg_get_rq_params -// Calculate requestor related parameters that register definition agnostic -// (i.e. this layer does try to separate real values from register definition) -// Input: -// pipe_src_param - pipe source configuration (e.g. vp, pitch, etc.) -// Output: -// rq_param - values that can be used to setup RQ (e.g. swath_height, plane1_addr, etc.) -// -void dml_rq_dlg_get_rq_params( - struct display_mode_lib *mode_lib, - display_rq_params_st *rq_param, - const display_pipe_source_params_st pipe_src_param); - -// Function: dml_rq_dlg_get_rq_reg -// Main entry point for test to get the register values out of this DML class. -// This function calls <get_rq_param> and <extract_rq_regs> fucntions to calculate -// and then populate the rq_regs struct -// Input: -// pipe_src_param - pipe source configuration (e.g. vp, pitch, etc.) -// Output: -// rq_regs - struct that holds all the RQ registers field value. -// See also: <display_rq_regs_st> -void dml_rq_dlg_get_rq_reg( - struct display_mode_lib *mode_lib, - display_rq_regs_st *rq_regs, - const display_pipe_source_params_st pipe_src_param); - -// Function: dml_rq_dlg_get_dlg_params -// Calculate deadline related parameters -// -void dml_rq_dlg_get_dlg_params(struct display_mode_lib *mode_lib, - const display_e2e_pipe_params_st *e2e_pipe_param, - const unsigned int num_pipes, - const unsigned int pipe_idx, - display_dlg_regs_st *disp_dlg_regs, - display_ttu_regs_st *disp_ttu_regs, - const display_rq_dlg_params_st rq_dlg_param, - const display_dlg_sys_params_st dlg_sys_param, - const bool cstate_en, - const bool pstate_en, - const bool vm_en, - const bool ignore_viewport_pos, - const bool immediate_flip_support); - -// Function: dml_rq_dlg_get_dlg_param_prefetch -// For flip_bw programming guide change, now dml needs to calculate the flip_bytes and prefetch_bw -// for ALL pipes and use this info to calculate the prefetch programming. -// Output: prefetch_param.prefetch_bw and flip_bytes -void dml_rq_dlg_get_dlg_params_prefetch( - struct display_mode_lib *mode_lib, - display_dlg_prefetch_param_st *prefetch_param, - display_rq_dlg_params_st rq_dlg_param, - display_dlg_sys_params_st dlg_sys_param, - display_e2e_pipe_params_st e2e_pipe_param, - const bool cstate_en, - const bool pstate_en, - const bool vm_en); - -// Function: dml_rq_dlg_get_dlg_reg -// Calculate and return DLG and TTU register struct given the system setting -// Output: -// dlg_regs - output DLG register struct -// ttu_regs - output DLG TTU register struct -// Input: -// e2e_pipe_param - "compacted" array of e2e pipe param struct -// num_pipes - num of active "pipe" or "route" -// pipe_idx - index that identifies the e2e_pipe_param that corresponding to this dlg -// cstate - 0: when calculate min_ttu_vblank it is assumed cstate is not required. 1: Normal mode, cstate is considered. -// Added for legacy or unrealistic timing tests. -void dml_rq_dlg_get_dlg_reg( - struct display_mode_lib *mode_lib, - display_dlg_regs_st *dlg_regs, - display_ttu_regs_st *ttu_regs, - display_e2e_pipe_params_st *e2e_pipe_param, - const unsigned int num_pipes, - const unsigned int pipe_idx, - const bool cstate_en, - const bool pstate_en, - const bool vm_en, - const bool ignore_viewport_pos, - const bool immediate_flip_support); - -// Function: dml_rq_dlg_get_calculated_vstartup -// Calculate and return vstartup -// Output: -// unsigned int vstartup -// Input: -// e2e_pipe_param - "compacted" array of e2e pipe param struct -// num_pipes - num of active "pipe" or "route" -// pipe_idx - index that identifies the e2e_pipe_param that corresponding to this dlg -// NOTE: this MUST be called after setting the prefetch mode! -unsigned int dml_rq_dlg_get_calculated_vstartup( - struct display_mode_lib *mode_lib, - display_e2e_pipe_params_st *e2e_pipe_param, - const unsigned int num_pipes, - const unsigned int pipe_idx); - -// Function: dml_rq_dlg_get_row_heights -// Calculate dpte and meta row heights -void dml_rq_dlg_get_row_heights( - struct display_mode_lib *mode_lib, - unsigned int *o_dpte_row_height, - unsigned int *o_meta_row_height, - unsigned int vp_width, - unsigned int data_pitch, - int source_format, - int tiling, - int macro_tile_size, - int source_scan, - int is_chroma); - -// Function: dml_rq_dlg_get_arb_params -void dml_rq_dlg_get_arb_params(struct display_mode_lib *mode_lib, display_arb_params_st *arb_param); - -#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c index 189052e..48400d6 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c +++ b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c
@@ -24,6 +24,7 @@ */ #include "display_rq_dlg_helpers.h" +#include "dml_logger.h" void print__rq_params_st(struct display_mode_lib *mode_lib, display_rq_params_st rq_param) {
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.h b/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.h index 987d7671..3041649 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.h
@@ -27,10 +27,11 @@ #define __DISPLAY_RQ_DLG_CALC_H__ #include "dml_common_defs.h" -#include "display_rq_dlg_helpers.h" struct display_mode_lib; +#include "display_rq_dlg_helpers.h" + void dml1_extract_rq_regs( struct display_mode_lib *mode_lib, struct _vcs_dpi_display_rq_regs_st *rq_regs,
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml_common_defs.h b/drivers/gpu/drm/amd/display/dc/dml/dml_common_defs.h index b2847bc..f78cbae 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dml_common_defs.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dml_common_defs.h
@@ -31,8 +31,6 @@ #include "display_mode_structs.h" #include "display_mode_enums.h" -#define dml_print(str, ...) {dm_logger_write(mode_lib->logger, LOG_DML, str, ##__VA_ARGS__); } -#define DTRACE(str, ...) {dm_logger_write(mode_lib->logger, LOG_DML, str, ##__VA_ARGS__); } double dml_round(double a);
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h index e68086b..f9cf083 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
@@ -28,6 +28,7 @@ #include "dml_common_defs.h" #include "../calcs/dcn_calc_math.h" +#include "dml_logger.h" static inline double dml_min(double a, double b) {
diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_instance.h b/drivers/gpu/drm/amd/display/dc/dml/dml_logger.h similarity index 79% rename from drivers/gpu/drm/amd/powerplay/inc/pp_instance.h rename to drivers/gpu/drm/amd/display/dc/dml/dml_logger.h index 7d1eec5..465859b 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/pp_instance.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dml_logger.h
@@ -1,5 +1,5 @@ /* - * Copyright 2015 Advanced Micro Devices, Inc. + * Copyright 2018 Advanced Micro Devices, Inc. * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), @@ -19,20 +19,20 @@ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR * OTHER DEALINGS IN THE SOFTWARE. * + * Authors: AMD + * */ -#ifndef _PP_INSTANCE_H_ -#define _PP_INSTANCE_H_ -#include "hwmgr.h" -struct pp_instance { - uint32_t chip_family; - uint32_t chip_id; - bool pm_en; - uint32_t feature_mask; - void *device; - struct pp_hwmgr *hwmgr; - struct mutex pp_lock; -}; +#ifndef __DML_LOGGER_H_ +#define __DML_LOGGER_H_ + +#define DC_LOGGER \ + mode_lib->logger + +#define dml_print(str, ...) {DC_LOG_DML(str, ##__VA_ARGS__); } +#define DTRACE(str, ...) {DC_LOG_DML(str, ##__VA_ARGS__); } #endif + +
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/dce120/hw_factory_dce120.c b/drivers/gpu/drm/amd/display/dc/gpio/dce120/hw_factory_dce120.c index 0c2314e..ea3f888 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/dce120/hw_factory_dce120.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/dce120/hw_factory_dce120.c
@@ -36,7 +36,8 @@ #include "dce/dce_12_0_offset.h" #include "dce/dce_12_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #define block HPD #define reg_num 0
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/dce120/hw_translate_dce120.c b/drivers/gpu/drm/amd/display/dc/gpio/dce120/hw_translate_dce120.c index a225b02c..39ef5c7 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/dce120/hw_translate_dce120.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/dce120/hw_translate_dce120.c
@@ -35,7 +35,8 @@ #include "dce/dce_12_0_offset.h" #include "dce/dce_12_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" /* begin ********************* * macros to expend register list macro defined in HW object header file */
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/dcn10/hw_factory_dcn10.c b/drivers/gpu/drm/amd/display/dc/gpio/dcn10/hw_factory_dcn10.c index 5235f69..32aa47a 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/dcn10/hw_factory_dcn10.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/dcn10/hw_factory_dcn10.c
@@ -36,7 +36,8 @@ #include "dcn/dcn_1_0_offset.h" #include "dcn/dcn_1_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #define block HPD #define reg_num 0
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/dcn10/hw_translate_dcn10.c b/drivers/gpu/drm/amd/display/dc/gpio/dcn10/hw_translate_dcn10.c index 3478648..fecc8688 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/dcn10/hw_translate_dcn10.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/dcn10/hw_translate_dcn10.c
@@ -35,7 +35,8 @@ #include "dcn/dcn_1_0_offset.h" #include "dcn/dcn_1_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" /* begin ********************* * macros to expend register list macro defined in HW object header file */
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c index 4aad7bb..a6e4221 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c
@@ -55,6 +55,8 @@ enum { #define FROM_ENGINE(ptr) \ container_of((ptr), struct aux_engine, base) +#define DC_LOGGER \ + engine->base.ctx->logger enum i2caux_engine_type dal_aux_engine_get_engine_type( const struct engine *engine) @@ -281,6 +283,15 @@ static bool read_command( } while (ctx.operation_succeeded && !ctx.transaction_complete); request->payload.length = ctx.reply.length; + + if (request->payload.address_space == + I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD) { + DC_LOG_I2C_AUX("READ: addr:0x%x value:0x%x Result:%d", + request->payload.address, + request->payload.data[0], + ctx.operation_succeeded); + } + return ctx.operation_succeeded; } @@ -489,6 +500,14 @@ static bool write_command( msleep(engine->delay); } while (ctx.operation_succeeded && !ctx.transaction_complete); + if (request->payload.address_space == + I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD) { + DC_LOG_I2C_AUX("WRITE: addr:0x%x value:0x%x Result:%d", + request->payload.address, + request->payload.data[0], + ctx.operation_succeeded); + } + return ctx.operation_succeeded; }
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c index e61a65a..a2fc335 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c
@@ -453,10 +453,6 @@ static void construct( static void destruct( struct aux_engine_dce110 *engine) { - struct aux_engine_dce110 *aux110 = engine; -/*temp w/a, to do*/ - REG_UPDATE(AUX_ARB_CONTROL, AUX_DMCU_DONE_USING_AUX_REG, 1); - REG_UPDATE(AUX_ARB_CONTROL, AUX_SW_DONE_USING_AUX_REG, 1); dal_aux_engine_destruct(&engine->base); }
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c index 56e25b3..abd0095 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c
@@ -48,6 +48,8 @@ /* * This unit */ +#define DC_LOGGER \ + hw_engine->base.base.base.ctx->logger enum dc_i2c_status { DC_I2C_STATUS__DC_I2C_STATUS_IDLE, @@ -525,9 +527,7 @@ static void construct( REG_GET(MICROSECOND_TIME_BASE_DIV, XTAL_REF_DIV, &xtal_ref_div); if (xtal_ref_div == 0) { - dm_logger_write( - hw_engine->base.base.base.ctx->logger, LOG_WARNING, - "Invalid base timer divider\n", + DC_LOG_WARNING("Invalid base timer divider\n", __func__); xtal_ref_div = 2; }
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c index a401636..0e7b182 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c
@@ -38,7 +38,8 @@ #include "dce/dce_12_0_offset.h" #include "dce/dce_12_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" /* begin ********************* * macros to expend register list macro defined in HW object header file */
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c b/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c index bed7cc3..e44a890 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c
@@ -38,7 +38,8 @@ #include "dcn/dcn_1_0_offset.h" #include "dcn/dcn_1_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" /* begin ********************* * macros to expend register list macro defined in HW object header file */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h index d697105..8c51ad7 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h +++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
@@ -119,6 +119,11 @@ struct resource_funcs { struct dc *dc, struct dc_state *new_ctx, struct dc_stream_state *dc_stream); + + enum dc_status (*remove_stream_from_ctx)( + struct dc *dc, + struct dc_state *new_ctx, + struct dc_stream_state *stream); }; struct audio_support{ @@ -148,6 +153,7 @@ struct resource_pool { unsigned int underlay_pipe_index; unsigned int stream_enc_count; unsigned int ref_clock_inKhz; + unsigned int timing_generator_count; /* * reserved clock source for DP @@ -171,6 +177,15 @@ struct resource_pool { const struct resource_caps *res_cap; }; +struct dcn_fe_clocks { + int dppclk_khz; +}; + +struct dcn_fe_bandwidth { + struct dcn_fe_clocks calc; + struct dcn_fe_clocks cur; +}; + struct stream_resource { struct output_pixel_processor *opp; struct timing_generator *tg; @@ -179,6 +194,8 @@ struct stream_resource { struct pixel_clk_params pix_clk_params; struct encoder_info_frame encoder_info_frame; + + struct abm *abm; }; struct plane_resource { @@ -188,6 +205,9 @@ struct plane_resource { struct input_pixel_processor *ipp; struct transform *xfm; struct dpp *dpp; + uint8_t mpcc_inst; + + struct dcn_fe_bandwidth bw; }; struct pipe_ctx { @@ -238,20 +258,9 @@ struct dce_bw_output { int blackout_recovery_time_us; }; -struct dcn_bw_clocks { - int dispclk_khz; - int dppclk_khz; - bool dppclk_div; - int dcfclk_khz; - int dcfclk_deep_sleep_khz; - int fclk_khz; - int dram_ccm_us; - int min_active_dram_ccm_us; -}; - struct dcn_bw_output { - struct dcn_bw_clocks cur_clk; - struct dcn_bw_clocks calc_clk; + struct dc_clocks cur_clk; + struct dc_clocks calc_clk; struct dcn_watermark_set watermarks; };
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h index 616c73e..2f783c6 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h
@@ -53,7 +53,7 @@ bool perform_link_training_with_retries( bool is_mst_supported(struct dc_link *link); -void detect_dp_sink_caps(struct dc_link *link); +bool detect_dp_sink_caps(struct dc_link *link); void detect_edp_sink_caps(struct dc_link *link);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h index ae2399f..a9bfe9f 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
@@ -130,6 +130,9 @@ enum bw_defines { struct bw_calcs_dceip { enum bw_calcs_version version; + uint32_t percent_of_ideal_port_bw_received_after_urgent_latency; + uint32_t max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation; + uint32_t max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation; bool large_cursor; uint32_t cursor_max_outstanding_group_num; bool dmif_pipe_en_fbc_chunk_tracker; @@ -230,6 +233,7 @@ struct bw_calcs_vbios { struct bw_calcs_data { /* data for all displays */ + bool display_synchronization_enabled; uint32_t number_of_displays; enum bw_defines underlay_surface_type; enum bw_defines panning_and_bezel_adjustment; @@ -241,6 +245,7 @@ struct bw_calcs_data { bool d1_display_write_back_dwb_enable; enum bw_defines d1_underlay_mode; + bool increase_voltage_to_support_mclk_switch; bool cpup_state_change_enable; bool cpuc_state_change_enable; bool nbp_state_change_enable; @@ -449,6 +454,7 @@ struct bw_calcs_data { struct bw_fixed dram_speed_change_line_source_transfer_time[maximum_number_of_surfaces][3][8]; struct bw_fixed min_dram_speed_change_margin[3][8]; struct bw_fixed dispclk_required_for_dram_speed_change[3][8]; + struct bw_fixed dispclk_required_for_dram_speed_change_pipe[3][8]; struct bw_fixed blackout_duration_margin[3][8]; struct bw_fixed dispclk_required_for_blackout_duration[3][8]; struct bw_fixed dispclk_required_for_blackout_recovery[3][8];
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h index ce2063554..de60f94 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h
@@ -32,13 +32,6 @@ enum dmcu_state { DMCU_RUNNING = 1 }; -struct dmcu_version { - unsigned int day; - unsigned int month; - unsigned int year; - unsigned int interface_version; -}; - struct dmcu { struct dc_context *ctx; const struct dmcu_funcs *funcs;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h index 25edbde..9999560 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
@@ -35,6 +35,8 @@ struct dpp { int inst; struct dpp_caps *caps; struct pwl_params regamma_params; + struct pwl_params degamma_params; + }; struct dpp_grph_csc_adjustment { @@ -130,6 +132,14 @@ struct dpp_funcs { const struct dc_cursor_mi_param *param, uint32_t width ); + void (*dpp_set_hdr_multiplier)( + struct dpp *dpp_base, + uint32_t multiplier); + + void (*dpp_dppclk_control)( + struct dpp *dpp_base, + bool dppclk_div, + bool enable); };
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h index b7c7e70..9ced254 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
@@ -119,6 +119,9 @@ struct hubp_funcs { void (*hubp_disconnect)(struct hubp *hubp); + void (*hubp_clk_cntl)(struct hubp *hubp, bool enable); + void (*hubp_vtg_sel)(struct hubp *hubp, uint32_t otg_inst); + }; #endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h index e3f0b40..b221581 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
@@ -136,7 +136,7 @@ struct out_csc_color_matrix { enum opp_regamma { OPP_REGAMMA_BYPASS = 0, OPP_REGAMMA_SRGB, - OPP_REGAMMA_3_6, + OPP_REGAMMA_XVYCC, OPP_REGAMMA_USER };
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/opp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/opp.h index ab8fb77f..d974d9e 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/opp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/opp.h
@@ -297,6 +297,10 @@ struct opp_funcs { bool enable, const struct dc_crtc_timing *timing); + void (*opp_pipe_clock_control)( + struct output_pixel_processor *opp, + bool enable); + }; #endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h index 3ca34629..3217b5b 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
@@ -217,6 +217,7 @@ struct timing_generator_funcs { */ bool (*get_crc)(struct timing_generator *tg, uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb); + }; #endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h b/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h index 6f6c02b..c5b3623 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
@@ -30,7 +30,7 @@ #include "dc_hw_types.h" #include "fixed31_32.h" -#define CSC_TEMPERATURE_MATRIX_SIZE 9 +#define CSC_TEMPERATURE_MATRIX_SIZE 12 struct bit_depth_reduction_params;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h index 379c6ec..e764cba 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -114,7 +114,7 @@ struct hw_sequencer_funcs { void (*power_down)(struct dc *dc); - void (*enable_accelerated_mode)(struct dc *dc); + void (*enable_accelerated_mode)(struct dc *dc, struct dc_state *context); void (*enable_timing_synchronization)( struct dc *dc, @@ -149,6 +149,7 @@ struct hw_sequencer_funcs { void (*unblank_stream)(struct pipe_ctx *pipe_ctx, struct dc_link_settings *link_settings); + void (*blank_stream)(struct pipe_ctx *pipe_ctx); void (*pipe_control_lock)( struct dc *dc, struct pipe_ctx *pipe, @@ -200,7 +201,6 @@ struct hw_sequencer_funcs { void (*set_cursor_position)(struct pipe_ctx *pipe); void (*set_cursor_attribute)(struct pipe_ctx *pipe); - }; void color_space_to_black_color(
diff --git a/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h b/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h index f2b8c9a..30be7bb 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h +++ b/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h
@@ -51,6 +51,8 @@ void dp_enable_link_phy( const struct dc_link_settings *link_settings); void dp_receiver_power_ctrl(struct dc_link *link, bool on); +bool edp_receiver_ready_T9(struct dc_link *link); +bool edp_receiver_ready_T7(struct dc_link *link); void dp_disable_link_phy(struct dc_link *link, enum signal_type signal);
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c b/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c index d3e1923..afe0876 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c
@@ -36,27 +36,25 @@ #include "dc.h" #include "core_types.h" -static bool hpd_ack( - struct irq_service *irq_service, - const struct irq_source_info *info) +#define DC_LOGGER \ + irq_service->ctx->logger + +static bool hpd_ack(struct irq_service *irq_service, + const struct irq_source_info *info) { uint32_t addr = info->status_reg; uint32_t value = dm_read_reg(irq_service->ctx, addr); - uint32_t current_status = - get_reg_field_value( - value, - DC_HPD_INT_STATUS, - DC_HPD_SENSE_DELAYED); + uint32_t current_status = get_reg_field_value(value, + DC_HPD_INT_STATUS, + DC_HPD_SENSE_DELAYED); dal_irq_service_ack_generic(irq_service, info); value = dm_read_reg(irq_service->ctx, info->enable_reg); - set_reg_field_value( - value, - current_status ? 0 : 1, - DC_HPD_INT_CONTROL, - DC_HPD_INT_POLARITY); + set_reg_field_value(value, current_status ? 0 : 1, + DC_HPD_INT_CONTROL, + DC_HPD_INT_POLARITY); dm_write_reg(irq_service->ctx, info->enable_reg, value); @@ -176,41 +174,34 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { #define dc_underflow_int_entry(reg_num) \ [DC_IRQ_SOURCE_DC ## reg_num ## UNDERFLOW] = dummy_irq_entry() -bool dal_irq_service_dummy_set( - struct irq_service *irq_service, - const struct irq_source_info *info, - bool enable) +bool dal_irq_service_dummy_set(struct irq_service *irq_service, + const struct irq_source_info *info, + bool enable) { - dm_logger_write( - irq_service->ctx->logger, LOG_ERROR, - "%s: called for non-implemented irq source\n", - __func__); + DC_LOG_ERROR("%s: called for non-implemented irq source\n", + __func__); return false; } -bool dal_irq_service_dummy_ack( - struct irq_service *irq_service, - const struct irq_source_info *info) +bool dal_irq_service_dummy_ack(struct irq_service *irq_service, + const struct irq_source_info *info) { - dm_logger_write( - irq_service->ctx->logger, LOG_ERROR, - "%s: called for non-implemented irq source\n", - __func__); + DC_LOG_ERROR("%s: called for non-implemented irq source\n", + __func__); return false; } -bool dce110_vblank_set( - struct irq_service *irq_service, - const struct irq_source_info *info, - bool enable) +bool dce110_vblank_set(struct irq_service *irq_service, + const struct irq_source_info *info, + bool enable) { struct dc_context *dc_ctx = irq_service->ctx; struct dc *core_dc = irq_service->ctx->dc; - enum dc_irq_source dal_irq_src = dc_interrupt_to_irq_source( - irq_service->ctx->dc, - info->src_id, - info->ext_id); + enum dc_irq_source dal_irq_src = + dc_interrupt_to_irq_source(irq_service->ctx->dc, + info->src_id, + info->ext_id); uint8_t pipe_offset = dal_irq_src - IRQ_TYPE_VBLANK; struct timing_generator *tg = @@ -225,7 +216,6 @@ bool dce110_vblank_set( dal_irq_service_set_generic(irq_service, info, enable); return true; - } static const struct irq_source_info_funcs dummy_irq_info_funcs = { @@ -406,9 +396,8 @@ static const struct irq_service_funcs irq_service_funcs_dce110 = { .to_dal_irq_source = to_dal_irq_source_dce110 }; -static void construct( - struct irq_service *irq_service, - struct irq_service_init_data *init_data) +static void construct(struct irq_service *irq_service, + struct irq_service_init_data *init_data) { dal_irq_service_construct(irq_service, init_data); @@ -416,8 +405,8 @@ static void construct( irq_service->funcs = &irq_service_funcs_dce110; } -struct irq_service *dal_irq_service_dce110_create( - struct irq_service_init_data *init_data) +struct irq_service * +dal_irq_service_dce110_create(struct irq_service_init_data *init_data) { struct irq_service *irq_service = kzalloc(sizeof(*irq_service), GFP_KERNEL);
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c b/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c index 66d5258..1ea7256 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c
@@ -32,7 +32,8 @@ #include "dce/dce_12_0_offset.h" #include "dce/dce_12_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #include "ivsrcid/ivsrcid_vislands30.h"
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c b/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c index 7f7db66..e04ae49 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c
@@ -31,7 +31,8 @@ #include "dcn/dcn_1_0_offset.h" #include "dcn/dcn_1_0_sh_mask.h" -#include "soc15ip.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" #include "irq_service_dcn10.h"
diff --git a/drivers/gpu/drm/amd/display/dc/irq/irq_service.c b/drivers/gpu/drm/amd/display/dc/irq/irq_service.c index b106513..dcdfa0f 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/irq_service.c +++ b/drivers/gpu/drm/amd/display/dc/irq/irq_service.c
@@ -47,6 +47,8 @@ #define CTX \ irq_service->ctx +#define DC_LOGGER \ + irq_service->ctx->logger void dal_irq_service_construct( struct irq_service *irq_service, @@ -104,9 +106,7 @@ bool dal_irq_service_set( find_irq_source_info(irq_service, source); if (!info) { - dm_logger_write( - irq_service->ctx->logger, LOG_ERROR, - "%s: cannot find irq info table entry for %d\n", + DC_LOG_ERROR("%s: cannot find irq info table entry for %d\n", __func__, source); return false; @@ -142,9 +142,7 @@ bool dal_irq_service_ack( find_irq_source_info(irq_service, source); if (!info) { - dm_logger_write( - irq_service->ctx->logger, LOG_ERROR, - "%s: cannot find irq info table entry for %d\n", + DC_LOG_ERROR("%s: cannot find irq info table entry for %d\n", __func__, source); return false;
diff --git a/drivers/gpu/drm/amd/display/dc/os_types.h b/drivers/gpu/drm/amd/display/dc/os_types.h index 1fcbc99..a407892 100644 --- a/drivers/gpu/drm/amd/display/dc/os_types.h +++ b/drivers/gpu/drm/amd/display/dc/os_types.h
@@ -32,7 +32,7 @@ #include <linux/kref.h> -#include "cgs_linux.h" +#include "cgs_common.h" #if defined(__BIG_ENDIAN) && !defined(BIGENDIAN_CPU) #define BIGENDIAN_CPU
diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h index 7abe663..9b0a04f 100644 --- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h +++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -109,6 +109,19 @@ #define ASIC_REV_IS_STONEY(rev) \ ((rev >= STONEY_A0) && (rev < CZ_UNKNOWN)) +/* DCE12 */ + +#define AI_GREENLAND_P_A0 1 +#define AI_GREENLAND_P_A1 2 +#define AI_UNKNOWN 0xFF + +#define AI_VEGA12_P_A0 20 +#define ASICREV_IS_GREENLAND_M(eChipRev) (eChipRev < AI_VEGA12_P_A0) +#define ASICREV_IS_GREENLAND_P(eChipRev) (eChipRev < AI_VEGA12_P_A0) + +#define ASICREV_IS_VEGA12_P(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_UNKNOWN)) +#define ASICREV_IS_VEGA12_p(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_UNKNOWN)) + /* DCN1_0 */ #define INTERNAL_REV_RAVEN_A0 0x00 /* First spin of Raven */ #define RAVEN_A0 0x01
diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h index 4badaed..0de2586 100644 --- a/drivers/gpu/drm/amd/display/include/fixed31_32.h +++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -470,4 +470,7 @@ uint32_t dal_fixed31_32_clamp_u0d14( uint32_t dal_fixed31_32_clamp_u0d10( struct fixed31_32 arg); +int32_t dal_fixed31_32_s4d19( + struct fixed31_32 arg); + #endif
diff --git a/drivers/gpu/drm/amd/display/include/link_service_types.h b/drivers/gpu/drm/amd/display/include/link_service_types.h index adea1a5..80f0d93 100644 --- a/drivers/gpu/drm/amd/display/include/link_service_types.h +++ b/drivers/gpu/drm/amd/display/include/link_service_types.h
@@ -58,11 +58,14 @@ enum { enum link_training_result { LINK_TRAINING_SUCCESS, - LINK_TRAINING_CR_FAIL, + LINK_TRAINING_CR_FAIL_LANE0, + LINK_TRAINING_CR_FAIL_LANE1, + LINK_TRAINING_CR_FAIL_LANE23, /* CR DONE bit is cleared during EQ step */ LINK_TRAINING_EQ_FAIL_CR, /* other failure during EQ step */ LINK_TRAINING_EQ_FAIL_EQ, + LINK_TRAINING_LQA_FAIL, }; struct link_training_settings {
diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h index e2ff8cd..427796b 100644 --- a/drivers/gpu/drm/amd/display/include/logger_types.h +++ b/drivers/gpu/drm/amd/display/include/logger_types.h
@@ -29,6 +29,39 @@ #include "os_types.h" #define MAX_NAME_LEN 32 +#define DC_LOG_ERROR(a, ...) dm_logger_write(DC_LOGGER, LOG_ERROR, a, ## __VA_ARGS__) +#define DC_LOG_WARNING(a, ...) dm_logger_write(DC_LOGGER, LOG_WARNING, a, ## __VA_ARGS__) +#define DC_LOG_DEBUG(a, ...) dm_logger_write(DC_LOGGER, LOG_DEBUG, a, ## __VA_ARGS__) +#define DC_LOG_DC(a, ...) dm_logger_write(DC_LOGGER, LOG_DC, a, ## __VA_ARGS__) +#define DC_LOG_DTN(a, ...) dm_logger_write(DC_LOGGER, LOG_DTN, a, ## __VA_ARGS__) +#define DC_LOG_SURFACE(a, ...) dm_logger_write(DC_LOGGER, LOG_SURFACE, a, ## __VA_ARGS__) +#define DC_LOG_HW_HOTPLUG(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_HOTPLUG, a, ## __VA_ARGS__) +#define DC_LOG_HW_LINK_TRAINING(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_LINK_TRAINING, a, ## __VA_ARGS__) +#define DC_LOG_HW_SET_MODE(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_SET_MODE, a, ## __VA_ARGS__) +#define DC_LOG_HW_RESUME_S3(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_RESUME_S3, a, ## __VA_ARGS__) +#define DC_LOG_HW_AUDIO(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_AUDIO, a, ## __VA_ARGS__) +#define DC_LOG_HW_HPD_IRQ(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_HPD_IRQ, a, ## __VA_ARGS__) +#define DC_LOG_MST(a, ...) dm_logger_write(DC_LOGGER, LOG_MST, a, ## __VA_ARGS__) +#define DC_LOG_SCALER(a, ...) dm_logger_write(DC_LOGGER, LOG_SCALER, a, ## __VA_ARGS__) +#define DC_LOG_BIOS(a, ...) dm_logger_write(DC_LOGGER, LOG_BIOS, a, ## __VA_ARGS__) +#define DC_LOG_BANDWIDTH_CALCS(a, ...) dm_logger_write(DC_LOGGER, LOG_BANDWIDTH_CALCS, a, ## __VA_ARGS__) +#define DC_LOG_BANDWIDTH_VALIDATION(a, ...) dm_logger_write(DC_LOGGER, LOG_BANDWIDTH_VALIDATION, a, ## __VA_ARGS__) +#define DC_LOG_I2C_AUX(a, ...) dm_logger_write(DC_LOGGER, LOG_I2C_AUX, a, ## __VA_ARGS__) +#define DC_LOG_SYNC(a, ...) dm_logger_write(DC_LOGGER, LOG_SYNC, a, ## __VA_ARGS__) +#define DC_LOG_BACKLIGHT(a, ...) dm_logger_write(DC_LOGGER, LOG_BACKLIGHT, a, ## __VA_ARGS__) +#define DC_LOG_FEATURE_OVERRIDE(a, ...) dm_logger_write(DC_LOGGER, LOG_FEATURE_OVERRIDE, a, ## __VA_ARGS__) +#define DC_LOG_DETECTION_EDID_PARSER(a, ...) dm_logger_write(DC_LOGGER, LOG_DETECTION_EDID_PARSER, a, ## __VA_ARGS__) +#define DC_LOG_DETECTION_DP_CAPS(a, ...) dm_logger_write(DC_LOGGER, LOG_DETECTION_DP_CAPS, a, ## __VA_ARGS__) +#define DC_LOG_RESOURCE(a, ...) dm_logger_write(DC_LOGGER, LOG_RESOURCE, a, ## __VA_ARGS__) +#define DC_LOG_DML(a, ...) dm_logger_write(DC_LOGGER, LOG_DML, a, ## __VA_ARGS__) +#define DC_LOG_EVENT_MODE_SET(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_MODE_SET, a, ## __VA_ARGS__) +#define DC_LOG_EVENT_DETECTION(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_DETECTION, a, ## __VA_ARGS__) +#define DC_LOG_EVENT_LINK_TRAINING(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_LINK_TRAINING, a, ## __VA_ARGS__) +#define DC_LOG_EVENT_LINK_LOSS(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_LINK_LOSS, a, ## __VA_ARGS__) +#define DC_LOG_EVENT_UNDERFLOW(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_UNDERFLOW, a, ## __VA_ARGS__) +#define DC_LOG_IF_TRACE(a, ...) dm_logger_write(DC_LOGGER, LOG_IF_TRACE, a, ## __VA_ARGS__) +#define DC_LOG_PERF_TRACE(a, ...) dm_logger_write(DC_LOGGER, LOG_PERF_TRACE, a, ## __VA_ARGS__) + struct dal_logger; @@ -65,6 +98,7 @@ enum dc_log_type { LOG_EVENT_UNDERFLOW, LOG_IF_TRACE, LOG_PERF_TRACE, + LOG_PROFILING, LOG_SECTION_TOTAL_COUNT };
diff --git a/drivers/gpu/drm/amd/display/modules/color/Makefile b/drivers/gpu/drm/amd/display/modules/color/Makefile new file mode 100644 index 0000000..65c33a7 --- /dev/null +++ b/drivers/gpu/drm/amd/display/modules/color/Makefile
@@ -0,0 +1,31 @@ +# +# Copyright 2018 Advanced Micro Devices, Inc. +# +# Permission is hereby granted, free of charge, to any person obtaining a +# copy of this software and associated documentation files (the "Software"), +# to deal in the Software without restriction, including without limitation +# the rights to use, copy, modify, merge, publish, distribute, sublicense, +# and/or sell copies of the Software, and to permit persons to whom the +# Software is furnished to do so, subject to the following conditions: +# +# The above copyright notice and this permission notice shall be included in +# all copies or substantial portions of the Software. +# +# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +# THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR +# OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, +# ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +# OTHER DEALINGS IN THE SOFTWARE. +# +# +# Makefile for the color sub-module of DAL. +# + +MOD_COLOR = color_gamma.o + +AMD_DAL_MOD_COLOR = $(addprefix $(AMDDALPATH)/modules/color/,$(MOD_COLOR)) +#$(info ************ DAL COLOR MODULE MAKEFILE ************) + +AMD_DISPLAY_FILES += $(AMD_DAL_MOD_COLOR)
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c new file mode 100644 index 0000000..b3747a0 --- /dev/null +++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
@@ -0,0 +1,1400 @@ +/* + * Copyright 2016 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#include "dc.h" +#include "opp.h" +#include "color_gamma.h" + + +#define NUM_PTS_IN_REGION 16 +#define NUM_REGIONS 32 +#define MAX_HW_POINTS (NUM_PTS_IN_REGION*NUM_REGIONS) + +static struct hw_x_point coordinates_x[MAX_HW_POINTS + 2]; + +static struct fixed31_32 pq_table[MAX_HW_POINTS + 2]; +static struct fixed31_32 de_pq_table[MAX_HW_POINTS + 2]; + +static bool pq_initialized; /* = false; */ +static bool de_pq_initialized; /* = false; */ + +/* one-time setup of X points */ +void setup_x_points_distribution(void) +{ + struct fixed31_32 region_size = dal_fixed31_32_from_int(128); + int32_t segment; + uint32_t seg_offset; + uint32_t index; + struct fixed31_32 increment; + + coordinates_x[MAX_HW_POINTS].x = region_size; + coordinates_x[MAX_HW_POINTS + 1].x = region_size; + + for (segment = 6; segment > (6 - NUM_REGIONS); segment--) { + region_size = dal_fixed31_32_div_int(region_size, 2); + increment = dal_fixed31_32_div_int(region_size, + NUM_PTS_IN_REGION); + seg_offset = (segment + (NUM_REGIONS - 7)) * NUM_PTS_IN_REGION; + coordinates_x[seg_offset].x = region_size; + + for (index = seg_offset + 1; + index < seg_offset + NUM_PTS_IN_REGION; + index++) { + coordinates_x[index].x = dal_fixed31_32_add + (coordinates_x[index-1].x, increment); + } + } +} + +static void compute_pq(struct fixed31_32 in_x, struct fixed31_32 *out_y) +{ + /* consts for PQ gamma formula. */ + const struct fixed31_32 m1 = + dal_fixed31_32_from_fraction(159301758, 1000000000); + const struct fixed31_32 m2 = + dal_fixed31_32_from_fraction(7884375, 100000); + const struct fixed31_32 c1 = + dal_fixed31_32_from_fraction(8359375, 10000000); + const struct fixed31_32 c2 = + dal_fixed31_32_from_fraction(188515625, 10000000); + const struct fixed31_32 c3 = + dal_fixed31_32_from_fraction(186875, 10000); + + struct fixed31_32 l_pow_m1; + struct fixed31_32 base; + + if (dal_fixed31_32_lt(in_x, dal_fixed31_32_zero)) + in_x = dal_fixed31_32_zero; + + l_pow_m1 = dal_fixed31_32_pow(in_x, m1); + base = dal_fixed31_32_div( + dal_fixed31_32_add(c1, + (dal_fixed31_32_mul(c2, l_pow_m1))), + dal_fixed31_32_add(dal_fixed31_32_one, + (dal_fixed31_32_mul(c3, l_pow_m1)))); + *out_y = dal_fixed31_32_pow(base, m2); +} + +static void compute_de_pq(struct fixed31_32 in_x, struct fixed31_32 *out_y) +{ + /* consts for dePQ gamma formula. */ + const struct fixed31_32 m1 = + dal_fixed31_32_from_fraction(159301758, 1000000000); + const struct fixed31_32 m2 = + dal_fixed31_32_from_fraction(7884375, 100000); + const struct fixed31_32 c1 = + dal_fixed31_32_from_fraction(8359375, 10000000); + const struct fixed31_32 c2 = + dal_fixed31_32_from_fraction(188515625, 10000000); + const struct fixed31_32 c3 = + dal_fixed31_32_from_fraction(186875, 10000); + + struct fixed31_32 l_pow_m1; + struct fixed31_32 base, div; + + + if (dal_fixed31_32_lt(in_x, dal_fixed31_32_zero)) + in_x = dal_fixed31_32_zero; + + l_pow_m1 = dal_fixed31_32_pow(in_x, + dal_fixed31_32_div(dal_fixed31_32_one, m2)); + base = dal_fixed31_32_sub(l_pow_m1, c1); + + if (dal_fixed31_32_lt(base, dal_fixed31_32_zero)) + base = dal_fixed31_32_zero; + + div = dal_fixed31_32_sub(c2, dal_fixed31_32_mul(c3, l_pow_m1)); + + *out_y = dal_fixed31_32_pow(dal_fixed31_32_div(base, div), + dal_fixed31_32_div(dal_fixed31_32_one, m1)); + +} +/* one-time pre-compute PQ values - only for sdr_white_level 80 */ +void precompute_pq(void) +{ + int i; + struct fixed31_32 x; + const struct hw_x_point *coord_x = coordinates_x + 32; + struct fixed31_32 scaling_factor = + dal_fixed31_32_from_fraction(80, 10000); + + /* pow function has problems with arguments too small */ + for (i = 0; i < 32; i++) + pq_table[i] = dal_fixed31_32_zero; + + for (i = 32; i <= MAX_HW_POINTS; i++) { + x = dal_fixed31_32_mul(coord_x->x, scaling_factor); + compute_pq(x, &pq_table[i]); + ++coord_x; + } +} + +/* one-time pre-compute dePQ values - only for max pixel value 125 FP16 */ +void precompute_de_pq(void) +{ + int i; + struct fixed31_32 y; + uint32_t begin_index, end_index; + + struct fixed31_32 scaling_factor = dal_fixed31_32_from_int(125); + + /* X points is 2^-25 to 2^7 + * De-gamma X is 2^-12 to 2^0 – we are skipping first -12-(-25) = 13 regions + */ + begin_index = 13 * NUM_PTS_IN_REGION; + end_index = begin_index + 12 * NUM_PTS_IN_REGION; + + for (i = 0; i <= begin_index; i++) + de_pq_table[i] = dal_fixed31_32_zero; + + for (; i <= end_index; i++) { + compute_de_pq(coordinates_x[i].x, &y); + de_pq_table[i] = dal_fixed31_32_mul(y, scaling_factor); + } + + for (; i <= MAX_HW_POINTS; i++) + de_pq_table[i] = de_pq_table[i-1]; +} +struct dividers { + struct fixed31_32 divider1; + struct fixed31_32 divider2; + struct fixed31_32 divider3; +}; + +static void build_coefficients(struct gamma_coefficients *coefficients, bool is_2_4) +{ + static const int32_t numerator01[] = { 31308, 180000}; + static const int32_t numerator02[] = { 12920, 4500}; + static const int32_t numerator03[] = { 55, 99}; + static const int32_t numerator04[] = { 55, 99}; + static const int32_t numerator05[] = { 2400, 2200}; + + uint32_t i = 0; + uint32_t index = is_2_4 == true ? 0:1; + + do { + coefficients->a0[i] = dal_fixed31_32_from_fraction( + numerator01[index], 10000000); + coefficients->a1[i] = dal_fixed31_32_from_fraction( + numerator02[index], 1000); + coefficients->a2[i] = dal_fixed31_32_from_fraction( + numerator03[index], 1000); + coefficients->a3[i] = dal_fixed31_32_from_fraction( + numerator04[index], 1000); + coefficients->user_gamma[i] = dal_fixed31_32_from_fraction( + numerator05[index], 1000); + + ++i; + } while (i != ARRAY_SIZE(coefficients->a0)); +} + +static struct fixed31_32 translate_from_linear_space( + struct fixed31_32 arg, + struct fixed31_32 a0, + struct fixed31_32 a1, + struct fixed31_32 a2, + struct fixed31_32 a3, + struct fixed31_32 gamma) +{ + const struct fixed31_32 one = dal_fixed31_32_from_int(1); + + if (dal_fixed31_32_lt(one, arg)) + return one; + + if (dal_fixed31_32_le(arg, dal_fixed31_32_neg(a0))) + return dal_fixed31_32_sub( + a2, + dal_fixed31_32_mul( + dal_fixed31_32_add( + one, + a3), + dal_fixed31_32_pow( + dal_fixed31_32_neg(arg), + dal_fixed31_32_recip(gamma)))); + else if (dal_fixed31_32_le(a0, arg)) + return dal_fixed31_32_sub( + dal_fixed31_32_mul( + dal_fixed31_32_add( + one, + a3), + dal_fixed31_32_pow( + arg, + dal_fixed31_32_recip(gamma))), + a2); + else + return dal_fixed31_32_mul( + arg, + a1); +} + +static struct fixed31_32 translate_to_linear_space( + struct fixed31_32 arg, + struct fixed31_32 a0, + struct fixed31_32 a1, + struct fixed31_32 a2, + struct fixed31_32 a3, + struct fixed31_32 gamma) +{ + struct fixed31_32 linear; + + a0 = dal_fixed31_32_mul(a0, a1); + if (dal_fixed31_32_le(arg, dal_fixed31_32_neg(a0))) + + linear = dal_fixed31_32_neg( + dal_fixed31_32_pow( + dal_fixed31_32_div( + dal_fixed31_32_sub(a2, arg), + dal_fixed31_32_add( + dal_fixed31_32_one, a3)), gamma)); + + else if (dal_fixed31_32_le(dal_fixed31_32_neg(a0), arg) && + dal_fixed31_32_le(arg, a0)) + linear = dal_fixed31_32_div(arg, a1); + else + linear = dal_fixed31_32_pow( + dal_fixed31_32_div( + dal_fixed31_32_add(a2, arg), + dal_fixed31_32_add( + dal_fixed31_32_one, a3)), gamma); + + return linear; +} + +static inline struct fixed31_32 translate_from_linear_space_ex( + struct fixed31_32 arg, + struct gamma_coefficients *coeff, + uint32_t color_index) +{ + return translate_from_linear_space( + arg, + coeff->a0[color_index], + coeff->a1[color_index], + coeff->a2[color_index], + coeff->a3[color_index], + coeff->user_gamma[color_index]); +} + + +static inline struct fixed31_32 translate_to_linear_space_ex( + struct fixed31_32 arg, + struct gamma_coefficients *coeff, + uint32_t color_index) +{ + return translate_to_linear_space( + arg, + coeff->a0[color_index], + coeff->a1[color_index], + coeff->a2[color_index], + coeff->a3[color_index], + coeff->user_gamma[color_index]); +} + + +static bool find_software_points( + const struct dc_gamma *ramp, + const struct gamma_pixel *axis_x, + struct fixed31_32 hw_point, + enum channel_name channel, + uint32_t *index_to_start, + uint32_t *index_left, + uint32_t *index_right, + enum hw_point_position *pos) +{ + const uint32_t max_number = ramp->num_entries + 3; + + struct fixed31_32 left, right; + + uint32_t i = *index_to_start; + + while (i < max_number) { + if (channel == CHANNEL_NAME_RED) { + left = axis_x[i].r; + + if (i < max_number - 1) + right = axis_x[i + 1].r; + else + right = axis_x[max_number - 1].r; + } else if (channel == CHANNEL_NAME_GREEN) { + left = axis_x[i].g; + + if (i < max_number - 1) + right = axis_x[i + 1].g; + else + right = axis_x[max_number - 1].g; + } else { + left = axis_x[i].b; + + if (i < max_number - 1) + right = axis_x[i + 1].b; + else + right = axis_x[max_number - 1].b; + } + + if (dal_fixed31_32_le(left, hw_point) && + dal_fixed31_32_le(hw_point, right)) { + *index_to_start = i; + *index_left = i; + + if (i < max_number - 1) + *index_right = i + 1; + else + *index_right = max_number - 1; + + *pos = HW_POINT_POSITION_MIDDLE; + + return true; + } else if ((i == *index_to_start) && + dal_fixed31_32_le(hw_point, left)) { + *index_to_start = i; + *index_left = i; + *index_right = i; + + *pos = HW_POINT_POSITION_LEFT; + + return true; + } else if ((i == max_number - 1) && + dal_fixed31_32_le(right, hw_point)) { + *index_to_start = i; + *index_left = i; + *index_right = i; + + *pos = HW_POINT_POSITION_RIGHT; + + return true; + } + + ++i; + } + + return false; +} + +static bool build_custom_gamma_mapping_coefficients_worker( + const struct dc_gamma *ramp, + struct pixel_gamma_point *coeff, + const struct hw_x_point *coordinates_x, + const struct gamma_pixel *axis_x, + enum channel_name channel, + uint32_t number_of_points) +{ + uint32_t i = 0; + + while (i <= number_of_points) { + struct fixed31_32 coord_x; + + uint32_t index_to_start = 0; + uint32_t index_left = 0; + uint32_t index_right = 0; + + enum hw_point_position hw_pos; + + struct gamma_point *point; + + struct fixed31_32 left_pos; + struct fixed31_32 right_pos; + + if (channel == CHANNEL_NAME_RED) + coord_x = coordinates_x[i].regamma_y_red; + else if (channel == CHANNEL_NAME_GREEN) + coord_x = coordinates_x[i].regamma_y_green; + else + coord_x = coordinates_x[i].regamma_y_blue; + + if (!find_software_points( + ramp, axis_x, coord_x, channel, + &index_to_start, &index_left, &index_right, &hw_pos)) { + BREAK_TO_DEBUGGER(); + return false; + } + + if (index_left >= ramp->num_entries + 3) { + BREAK_TO_DEBUGGER(); + return false; + } + + if (index_right >= ramp->num_entries + 3) { + BREAK_TO_DEBUGGER(); + return false; + } + + if (channel == CHANNEL_NAME_RED) { + point = &coeff[i].r; + + left_pos = axis_x[index_left].r; + right_pos = axis_x[index_right].r; + } else if (channel == CHANNEL_NAME_GREEN) { + point = &coeff[i].g; + + left_pos = axis_x[index_left].g; + right_pos = axis_x[index_right].g; + } else { + point = &coeff[i].b; + + left_pos = axis_x[index_left].b; + right_pos = axis_x[index_right].b; + } + + if (hw_pos == HW_POINT_POSITION_MIDDLE) + point->coeff = dal_fixed31_32_div( + dal_fixed31_32_sub( + coord_x, + left_pos), + dal_fixed31_32_sub( + right_pos, + left_pos)); + else if (hw_pos == HW_POINT_POSITION_LEFT) + point->coeff = dal_fixed31_32_zero; + else if (hw_pos == HW_POINT_POSITION_RIGHT) + point->coeff = dal_fixed31_32_from_int(2); + else { + BREAK_TO_DEBUGGER(); + return false; + } + + point->left_index = index_left; + point->right_index = index_right; + point->pos = hw_pos; + + ++i; + } + + return true; +} + +static struct fixed31_32 calculate_mapped_value( + struct pwl_float_data *rgb, + const struct pixel_gamma_point *coeff, + enum channel_name channel, + uint32_t max_index) +{ + const struct gamma_point *point; + + struct fixed31_32 result; + + if (channel == CHANNEL_NAME_RED) + point = &coeff->r; + else if (channel == CHANNEL_NAME_GREEN) + point = &coeff->g; + else + point = &coeff->b; + + if ((point->left_index < 0) || (point->left_index > max_index)) { + BREAK_TO_DEBUGGER(); + return dal_fixed31_32_zero; + } + + if ((point->right_index < 0) || (point->right_index > max_index)) { + BREAK_TO_DEBUGGER(); + return dal_fixed31_32_zero; + } + + if (point->pos == HW_POINT_POSITION_MIDDLE) + if (channel == CHANNEL_NAME_RED) + result = dal_fixed31_32_add( + dal_fixed31_32_mul( + point->coeff, + dal_fixed31_32_sub( + rgb[point->right_index].r, + rgb[point->left_index].r)), + rgb[point->left_index].r); + else if (channel == CHANNEL_NAME_GREEN) + result = dal_fixed31_32_add( + dal_fixed31_32_mul( + point->coeff, + dal_fixed31_32_sub( + rgb[point->right_index].g, + rgb[point->left_index].g)), + rgb[point->left_index].g); + else + result = dal_fixed31_32_add( + dal_fixed31_32_mul( + point->coeff, + dal_fixed31_32_sub( + rgb[point->right_index].b, + rgb[point->left_index].b)), + rgb[point->left_index].b); + else if (point->pos == HW_POINT_POSITION_LEFT) { + BREAK_TO_DEBUGGER(); + result = dal_fixed31_32_zero; + } else { + BREAK_TO_DEBUGGER(); + result = dal_fixed31_32_one; + } + + return result; +} + +static void build_pq(struct pwl_float_data_ex *rgb_regamma, + uint32_t hw_points_num, + const struct hw_x_point *coordinate_x, + uint32_t sdr_white_level) +{ + uint32_t i, start_index; + + struct pwl_float_data_ex *rgb = rgb_regamma; + const struct hw_x_point *coord_x = coordinate_x; + struct fixed31_32 x; + struct fixed31_32 output; + struct fixed31_32 scaling_factor = + dal_fixed31_32_from_fraction(sdr_white_level, 10000); + + if (!pq_initialized && sdr_white_level == 80) { + precompute_pq(); + pq_initialized = true; + } + + /* TODO: start index is from segment 2^-24, skipping first segment + * due to x values too small for power calculations + */ + start_index = 32; + rgb += start_index; + coord_x += start_index; + + for (i = start_index; i <= hw_points_num; i++) { + /* Multiply 0.008 as regamma is 0-1 and FP16 input is 0-125. + * FP 1.0 = 80nits + */ + if (sdr_white_level == 80) { + output = pq_table[i]; + } else { + x = dal_fixed31_32_mul(coord_x->x, scaling_factor); + compute_pq(x, &output); + } + + /* should really not happen? */ + if (dal_fixed31_32_lt(output, dal_fixed31_32_zero)) + output = dal_fixed31_32_zero; + else if (dal_fixed31_32_lt(dal_fixed31_32_one, output)) + output = dal_fixed31_32_one; + + rgb->r = output; + rgb->g = output; + rgb->b = output; + + ++coord_x; + ++rgb; + } +} + +static void build_de_pq(struct pwl_float_data_ex *de_pq, + uint32_t hw_points_num, + const struct hw_x_point *coordinate_x) +{ + uint32_t i; + struct fixed31_32 output; + + struct fixed31_32 scaling_factor = dal_fixed31_32_from_int(125); + + if (!de_pq_initialized) { + precompute_de_pq(); + de_pq_initialized = true; + } + + + for (i = 0; i <= hw_points_num; i++) { + output = de_pq_table[i]; + /* should really not happen? */ + if (dal_fixed31_32_lt(output, dal_fixed31_32_zero)) + output = dal_fixed31_32_zero; + else if (dal_fixed31_32_lt(scaling_factor, output)) + output = scaling_factor; + de_pq[i].r = output; + de_pq[i].g = output; + de_pq[i].b = output; + } +} + +static void build_regamma(struct pwl_float_data_ex *rgb_regamma, + uint32_t hw_points_num, + const struct hw_x_point *coordinate_x, bool is_2_4) +{ + uint32_t i; + + struct gamma_coefficients coeff; + struct pwl_float_data_ex *rgb = rgb_regamma; + const struct hw_x_point *coord_x = coordinate_x; + + build_coefficients(&coeff, is_2_4); + + i = 0; + + while (i != hw_points_num + 1) { + /*TODO use y vs r,g,b*/ + rgb->r = translate_from_linear_space_ex( + coord_x->x, &coeff, 0); + rgb->g = rgb->r; + rgb->b = rgb->r; + ++coord_x; + ++rgb; + ++i; + } +} + +static void build_degamma(struct pwl_float_data_ex *curve, + uint32_t hw_points_num, + const struct hw_x_point *coordinate_x, bool is_2_4) +{ + uint32_t i; + struct gamma_coefficients coeff; + uint32_t begin_index, end_index; + + build_coefficients(&coeff, is_2_4); + i = 0; + + /* X points is 2^-25 to 2^7 + * De-gamma X is 2^-12 to 2^0 – we are skipping first -12-(-25) = 13 regions + */ + begin_index = 13 * NUM_PTS_IN_REGION; + end_index = begin_index + 12 * NUM_PTS_IN_REGION; + + while (i != begin_index) { + curve[i].r = dal_fixed31_32_zero; + curve[i].g = dal_fixed31_32_zero; + curve[i].b = dal_fixed31_32_zero; + i++; + } + + while (i != end_index) { + curve[i].r = translate_to_linear_space_ex( + coordinate_x[i].x, &coeff, 0); + curve[i].g = curve[i].r; + curve[i].b = curve[i].r; + i++; + } + while (i != hw_points_num + 1) { + curve[i].r = dal_fixed31_32_one; + curve[i].g = dal_fixed31_32_one; + curve[i].b = dal_fixed31_32_one; + i++; + } +} + +static bool scale_gamma(struct pwl_float_data *pwl_rgb, + const struct dc_gamma *ramp, + struct dividers dividers) +{ + const struct fixed31_32 max_driver = dal_fixed31_32_from_int(0xFFFF); + const struct fixed31_32 max_os = dal_fixed31_32_from_int(0xFF00); + struct fixed31_32 scaler = max_os; + uint32_t i; + struct pwl_float_data *rgb = pwl_rgb; + struct pwl_float_data *rgb_last = rgb + ramp->num_entries - 1; + + i = 0; + + do { + if (dal_fixed31_32_lt(max_os, ramp->entries.red[i]) || + dal_fixed31_32_lt(max_os, ramp->entries.green[i]) || + dal_fixed31_32_lt(max_os, ramp->entries.blue[i])) { + scaler = max_driver; + break; + } + ++i; + } while (i != ramp->num_entries); + + i = 0; + + do { + rgb->r = dal_fixed31_32_div( + ramp->entries.red[i], scaler); + rgb->g = dal_fixed31_32_div( + ramp->entries.green[i], scaler); + rgb->b = dal_fixed31_32_div( + ramp->entries.blue[i], scaler); + + ++rgb; + ++i; + } while (i != ramp->num_entries); + + rgb->r = dal_fixed31_32_mul(rgb_last->r, + dividers.divider1); + rgb->g = dal_fixed31_32_mul(rgb_last->g, + dividers.divider1); + rgb->b = dal_fixed31_32_mul(rgb_last->b, + dividers.divider1); + + ++rgb; + + rgb->r = dal_fixed31_32_mul(rgb_last->r, + dividers.divider2); + rgb->g = dal_fixed31_32_mul(rgb_last->g, + dividers.divider2); + rgb->b = dal_fixed31_32_mul(rgb_last->b, + dividers.divider2); + + ++rgb; + + rgb->r = dal_fixed31_32_mul(rgb_last->r, + dividers.divider3); + rgb->g = dal_fixed31_32_mul(rgb_last->g, + dividers.divider3); + rgb->b = dal_fixed31_32_mul(rgb_last->b, + dividers.divider3); + + return true; +} + +static bool scale_gamma_dx(struct pwl_float_data *pwl_rgb, + const struct dc_gamma *ramp, + struct dividers dividers) +{ + uint32_t i; + struct fixed31_32 min = dal_fixed31_32_zero; + struct fixed31_32 max = dal_fixed31_32_one; + + struct fixed31_32 delta = dal_fixed31_32_zero; + struct fixed31_32 offset = dal_fixed31_32_zero; + + for (i = 0 ; i < ramp->num_entries; i++) { + if (dal_fixed31_32_lt(ramp->entries.red[i], min)) + min = ramp->entries.red[i]; + + if (dal_fixed31_32_lt(ramp->entries.green[i], min)) + min = ramp->entries.green[i]; + + if (dal_fixed31_32_lt(ramp->entries.blue[i], min)) + min = ramp->entries.blue[i]; + + if (dal_fixed31_32_lt(max, ramp->entries.red[i])) + max = ramp->entries.red[i]; + + if (dal_fixed31_32_lt(max, ramp->entries.green[i])) + max = ramp->entries.green[i]; + + if (dal_fixed31_32_lt(max, ramp->entries.blue[i])) + max = ramp->entries.blue[i]; + } + + if (dal_fixed31_32_lt(min, dal_fixed31_32_zero)) + delta = dal_fixed31_32_neg(min); + + offset = dal_fixed31_32_add(min, max); + + for (i = 0 ; i < ramp->num_entries; i++) { + pwl_rgb[i].r = dal_fixed31_32_div( + dal_fixed31_32_add( + ramp->entries.red[i], delta), offset); + pwl_rgb[i].g = dal_fixed31_32_div( + dal_fixed31_32_add( + ramp->entries.green[i], delta), offset); + pwl_rgb[i].b = dal_fixed31_32_div( + dal_fixed31_32_add( + ramp->entries.blue[i], delta), offset); + + } + + pwl_rgb[i].r = dal_fixed31_32_sub(dal_fixed31_32_mul_int( + pwl_rgb[i-1].r, 2), pwl_rgb[i-2].r); + pwl_rgb[i].g = dal_fixed31_32_sub(dal_fixed31_32_mul_int( + pwl_rgb[i-1].g, 2), pwl_rgb[i-2].g); + pwl_rgb[i].b = dal_fixed31_32_sub(dal_fixed31_32_mul_int( + pwl_rgb[i-1].b, 2), pwl_rgb[i-2].b); + ++i; + pwl_rgb[i].r = dal_fixed31_32_sub(dal_fixed31_32_mul_int( + pwl_rgb[i-1].r, 2), pwl_rgb[i-2].r); + pwl_rgb[i].g = dal_fixed31_32_sub(dal_fixed31_32_mul_int( + pwl_rgb[i-1].g, 2), pwl_rgb[i-2].g); + pwl_rgb[i].b = dal_fixed31_32_sub(dal_fixed31_32_mul_int( + pwl_rgb[i-1].b, 2), pwl_rgb[i-2].b); + + return true; +} + +/* + * RS3+ color transform DDI - 1D LUT adjustment is composed with regamma here + * Input is evenly distributed in the output color space as specified in + * SetTimings + * + * Interpolation details: + * 1D LUT has 4096 values which give curve correction in 0-1 float range + * for evenly spaced points in 0-1 range. lut1D[index] gives correction + * for index/4095. + * First we find index for which: + * index/4095 < regamma_y < (index+1)/4095 => + * index < 4095*regamma_y < index + 1 + * norm_y = 4095*regamma_y, and index is just truncating to nearest integer + * lut1 = lut1D[index], lut2 = lut1D[index+1] + * + *adjustedY is then linearly interpolating regamma Y between lut1 and lut2 + */ +static void apply_lut_1d( + const struct dc_gamma *ramp, + uint32_t num_hw_points, + struct dc_transfer_func_distributed_points *tf_pts) +{ + int i = 0; + int color = 0; + struct fixed31_32 *regamma_y; + struct fixed31_32 norm_y; + struct fixed31_32 lut1; + struct fixed31_32 lut2; + const int max_lut_index = 4095; + const struct fixed31_32 max_lut_index_f = + dal_fixed31_32_from_int_nonconst(max_lut_index); + int32_t index = 0, index_next = 0; + struct fixed31_32 index_f; + struct fixed31_32 delta_lut; + struct fixed31_32 delta_index; + + if (ramp->type != GAMMA_CS_TFM_1D) + return; // this is not expected + + for (i = 0; i < num_hw_points; i++) { + for (color = 0; color < 3; color++) { + if (color == 0) + regamma_y = &tf_pts->red[i]; + else if (color == 1) + regamma_y = &tf_pts->green[i]; + else + regamma_y = &tf_pts->blue[i]; + + norm_y = dal_fixed31_32_mul(max_lut_index_f, + *regamma_y); + index = dal_fixed31_32_floor(norm_y); + index_f = dal_fixed31_32_from_int_nonconst(index); + + if (index < 0 || index > max_lut_index) + continue; + + index_next = (index == max_lut_index) ? index : index+1; + + if (color == 0) { + lut1 = ramp->entries.red[index]; + lut2 = ramp->entries.red[index_next]; + } else if (color == 1) { + lut1 = ramp->entries.green[index]; + lut2 = ramp->entries.green[index_next]; + } else { + lut1 = ramp->entries.blue[index]; + lut2 = ramp->entries.blue[index_next]; + } + + // we have everything now, so interpolate + delta_lut = dal_fixed31_32_sub(lut2, lut1); + delta_index = dal_fixed31_32_sub(norm_y, index_f); + + *regamma_y = dal_fixed31_32_add(lut1, + dal_fixed31_32_mul(delta_index, delta_lut)); + } + } +} + +static void build_evenly_distributed_points( + struct gamma_pixel *points, + uint32_t numberof_points, + struct dividers dividers) +{ + struct gamma_pixel *p = points; + struct gamma_pixel *p_last = p + numberof_points - 1; + + uint32_t i = 0; + + do { + struct fixed31_32 value = dal_fixed31_32_from_fraction(i, + numberof_points - 1); + + p->r = value; + p->g = value; + p->b = value; + + ++p; + ++i; + } while (i != numberof_points); + + p->r = dal_fixed31_32_div(p_last->r, dividers.divider1); + p->g = dal_fixed31_32_div(p_last->g, dividers.divider1); + p->b = dal_fixed31_32_div(p_last->b, dividers.divider1); + + ++p; + + p->r = dal_fixed31_32_div(p_last->r, dividers.divider2); + p->g = dal_fixed31_32_div(p_last->g, dividers.divider2); + p->b = dal_fixed31_32_div(p_last->b, dividers.divider2); + + ++p; + + p->r = dal_fixed31_32_div(p_last->r, dividers.divider3); + p->g = dal_fixed31_32_div(p_last->g, dividers.divider3); + p->b = dal_fixed31_32_div(p_last->b, dividers.divider3); +} + +static inline void copy_rgb_regamma_to_coordinates_x( + struct hw_x_point *coordinates_x, + uint32_t hw_points_num, + const struct pwl_float_data_ex *rgb_ex) +{ + struct hw_x_point *coords = coordinates_x; + uint32_t i = 0; + const struct pwl_float_data_ex *rgb_regamma = rgb_ex; + + while (i <= hw_points_num) { + coords->regamma_y_red = rgb_regamma->r; + coords->regamma_y_green = rgb_regamma->g; + coords->regamma_y_blue = rgb_regamma->b; + + ++coords; + ++rgb_regamma; + ++i; + } +} + +static bool calculate_interpolated_hardware_curve( + const struct dc_gamma *ramp, + struct pixel_gamma_point *coeff128, + struct pwl_float_data *rgb_user, + const struct hw_x_point *coordinates_x, + const struct gamma_pixel *axis_x, + uint32_t number_of_points, + struct dc_transfer_func_distributed_points *tf_pts) +{ + + const struct pixel_gamma_point *coeff = coeff128; + uint32_t max_entries = 3 - 1; + + uint32_t i = 0; + + for (i = 0; i < 3; i++) { + if (!build_custom_gamma_mapping_coefficients_worker( + ramp, coeff128, coordinates_x, axis_x, i, + number_of_points)) + return false; + } + + i = 0; + max_entries += ramp->num_entries; + + /* TODO: float point case */ + + while (i <= number_of_points) { + tf_pts->red[i] = calculate_mapped_value( + rgb_user, coeff, CHANNEL_NAME_RED, max_entries); + tf_pts->green[i] = calculate_mapped_value( + rgb_user, coeff, CHANNEL_NAME_GREEN, max_entries); + tf_pts->blue[i] = calculate_mapped_value( + rgb_user, coeff, CHANNEL_NAME_BLUE, max_entries); + + ++coeff; + ++i; + } + + return true; +} + +static void build_new_custom_resulted_curve( + uint32_t hw_points_num, + struct dc_transfer_func_distributed_points *tf_pts) +{ + uint32_t i; + + i = 0; + + while (i != hw_points_num + 1) { + tf_pts->red[i] = dal_fixed31_32_clamp( + tf_pts->red[i], dal_fixed31_32_zero, + dal_fixed31_32_one); + tf_pts->green[i] = dal_fixed31_32_clamp( + tf_pts->green[i], dal_fixed31_32_zero, + dal_fixed31_32_one); + tf_pts->blue[i] = dal_fixed31_32_clamp( + tf_pts->blue[i], dal_fixed31_32_zero, + dal_fixed31_32_one); + + ++i; + } +} + +static bool map_regamma_hw_to_x_user( + const struct dc_gamma *ramp, + struct pixel_gamma_point *coeff128, + struct pwl_float_data *rgb_user, + struct hw_x_point *coords_x, + const struct gamma_pixel *axis_x, + const struct pwl_float_data_ex *rgb_regamma, + uint32_t hw_points_num, + struct dc_transfer_func_distributed_points *tf_pts, + bool mapUserRamp) +{ + /* setup to spare calculated ideal regamma values */ + + int i = 0; + struct hw_x_point *coords = coords_x; + const struct pwl_float_data_ex *regamma = rgb_regamma; + + if (mapUserRamp) { + copy_rgb_regamma_to_coordinates_x(coords, + hw_points_num, + rgb_regamma); + + calculate_interpolated_hardware_curve( + ramp, coeff128, rgb_user, coords, axis_x, + hw_points_num, tf_pts); + } else { + /* just copy current rgb_regamma into tf_pts */ + while (i <= hw_points_num) { + tf_pts->red[i] = regamma->r; + tf_pts->green[i] = regamma->g; + tf_pts->blue[i] = regamma->b; + + ++regamma; + ++i; + } + } + + build_new_custom_resulted_curve(hw_points_num, tf_pts); + + return true; +} + +#define _EXTRA_POINTS 3 + +bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf, + const struct dc_gamma *ramp, bool mapUserRamp) +{ + struct dc_transfer_func_distributed_points *tf_pts = &output_tf->tf_pts; + struct dividers dividers; + + struct pwl_float_data *rgb_user = NULL; + struct pwl_float_data_ex *rgb_regamma = NULL; + struct gamma_pixel *axix_x = NULL; + struct pixel_gamma_point *coeff = NULL; + enum dc_transfer_func_predefined tf = TRANSFER_FUNCTION_SRGB; + bool ret = false; + + if (output_tf->type == TF_TYPE_BYPASS) + return false; + + /* we can use hardcoded curve for plain SRGB TF */ + if (output_tf->type == TF_TYPE_PREDEFINED && + output_tf->tf == TRANSFER_FUNCTION_SRGB && + (!mapUserRamp && ramp->type == GAMMA_RGB_256)) + return true; + + output_tf->type = TF_TYPE_DISTRIBUTED_POINTS; + + rgb_user = kvzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS), + GFP_KERNEL); + if (!rgb_user) + goto rgb_user_alloc_fail; + rgb_regamma = kvzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS + _EXTRA_POINTS), + GFP_KERNEL); + if (!rgb_regamma) + goto rgb_regamma_alloc_fail; + axix_x = kvzalloc(sizeof(*axix_x) * (ramp->num_entries + 3), + GFP_KERNEL); + if (!axix_x) + goto axix_x_alloc_fail; + coeff = kvzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL); + if (!coeff) + goto coeff_alloc_fail; + + dividers.divider1 = dal_fixed31_32_from_fraction(3, 2); + dividers.divider2 = dal_fixed31_32_from_int(2); + dividers.divider3 = dal_fixed31_32_from_fraction(5, 2); + + tf = output_tf->tf; + + build_evenly_distributed_points( + axix_x, + ramp->num_entries, + dividers); + + if (ramp->type == GAMMA_RGB_256 && mapUserRamp) + scale_gamma(rgb_user, ramp, dividers); + else if (ramp->type == GAMMA_RGB_FLOAT_1024) + scale_gamma_dx(rgb_user, ramp, dividers); + + if (tf == TRANSFER_FUNCTION_PQ) { + tf_pts->end_exponent = 7; + tf_pts->x_point_at_y1_red = 125; + tf_pts->x_point_at_y1_green = 125; + tf_pts->x_point_at_y1_blue = 125; + + build_pq(rgb_regamma, + MAX_HW_POINTS, + coordinates_x, + output_tf->sdr_ref_white_level); + } else { + tf_pts->end_exponent = 0; + tf_pts->x_point_at_y1_red = 1; + tf_pts->x_point_at_y1_green = 1; + tf_pts->x_point_at_y1_blue = 1; + + build_regamma(rgb_regamma, + MAX_HW_POINTS, + coordinates_x, tf == TRANSFER_FUNCTION_SRGB ? true:false); + } + + map_regamma_hw_to_x_user(ramp, coeff, rgb_user, + coordinates_x, axix_x, rgb_regamma, + MAX_HW_POINTS, tf_pts, + (mapUserRamp || ramp->type != GAMMA_RGB_256) && + ramp->type != GAMMA_CS_TFM_1D); + + if (ramp->type == GAMMA_CS_TFM_1D) + apply_lut_1d(ramp, MAX_HW_POINTS, tf_pts); + + ret = true; + + kvfree(coeff); +coeff_alloc_fail: + kvfree(axix_x); +axix_x_alloc_fail: + kvfree(rgb_regamma); +rgb_regamma_alloc_fail: + kvfree(rgb_user); +rgb_user_alloc_fail: + return ret; +} + +bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf, + const struct dc_gamma *ramp, bool mapUserRamp) +{ + struct dc_transfer_func_distributed_points *tf_pts = &input_tf->tf_pts; + struct dividers dividers; + + struct pwl_float_data *rgb_user = NULL; + struct pwl_float_data_ex *curve = NULL; + struct gamma_pixel *axix_x = NULL; + struct pixel_gamma_point *coeff = NULL; + enum dc_transfer_func_predefined tf = TRANSFER_FUNCTION_SRGB; + bool ret = false; + + if (input_tf->type == TF_TYPE_BYPASS) + return false; + + /* we can use hardcoded curve for plain SRGB TF */ + if (input_tf->type == TF_TYPE_PREDEFINED && + input_tf->tf == TRANSFER_FUNCTION_SRGB && + (!mapUserRamp && ramp->type == GAMMA_RGB_256)) + return true; + + input_tf->type = TF_TYPE_DISTRIBUTED_POINTS; + + rgb_user = kvzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS), + GFP_KERNEL); + if (!rgb_user) + goto rgb_user_alloc_fail; + curve = kvzalloc(sizeof(*curve) * (MAX_HW_POINTS + _EXTRA_POINTS), + GFP_KERNEL); + if (!curve) + goto curve_alloc_fail; + axix_x = kvzalloc(sizeof(*axix_x) * (ramp->num_entries + _EXTRA_POINTS), + GFP_KERNEL); + if (!axix_x) + goto axix_x_alloc_fail; + coeff = kvzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL); + if (!coeff) + goto coeff_alloc_fail; + + dividers.divider1 = dal_fixed31_32_from_fraction(3, 2); + dividers.divider2 = dal_fixed31_32_from_int(2); + dividers.divider3 = dal_fixed31_32_from_fraction(5, 2); + + tf = input_tf->tf; + + build_evenly_distributed_points( + axix_x, + ramp->num_entries, + dividers); + + if (ramp->type == GAMMA_RGB_256 && mapUserRamp) + scale_gamma(rgb_user, ramp, dividers); + else if (ramp->type == GAMMA_RGB_FLOAT_1024) + scale_gamma_dx(rgb_user, ramp, dividers); + + if (tf == TRANSFER_FUNCTION_PQ) + build_de_pq(curve, + MAX_HW_POINTS, + coordinates_x); + else + build_degamma(curve, + MAX_HW_POINTS, + coordinates_x, + tf == TRANSFER_FUNCTION_SRGB ? true:false); + + tf_pts->end_exponent = 0; + tf_pts->x_point_at_y1_red = 1; + tf_pts->x_point_at_y1_green = 1; + tf_pts->x_point_at_y1_blue = 1; + + map_regamma_hw_to_x_user(ramp, coeff, rgb_user, + coordinates_x, axix_x, curve, + MAX_HW_POINTS, tf_pts, + mapUserRamp); + + ret = true; + + kvfree(coeff); +coeff_alloc_fail: + kvfree(axix_x); +axix_x_alloc_fail: + kvfree(curve); +curve_alloc_fail: + kvfree(rgb_user); +rgb_user_alloc_fail: + + return ret; + +} + + +bool mod_color_calculate_curve(enum dc_transfer_func_predefined trans, + struct dc_transfer_func_distributed_points *points) +{ + uint32_t i; + bool ret = false; + struct pwl_float_data_ex *rgb_regamma = NULL; + + if (trans == TRANSFER_FUNCTION_UNITY || + trans == TRANSFER_FUNCTION_LINEAR) { + points->end_exponent = 0; + points->x_point_at_y1_red = 1; + points->x_point_at_y1_green = 1; + points->x_point_at_y1_blue = 1; + + for (i = 0; i <= MAX_HW_POINTS ; i++) { + points->red[i] = coordinates_x[i].x; + points->green[i] = coordinates_x[i].x; + points->blue[i] = coordinates_x[i].x; + } + ret = true; + } else if (trans == TRANSFER_FUNCTION_PQ) { + rgb_regamma = kvzalloc(sizeof(*rgb_regamma) * + (MAX_HW_POINTS + _EXTRA_POINTS), + GFP_KERNEL); + if (!rgb_regamma) + goto rgb_regamma_alloc_fail; + points->end_exponent = 7; + points->x_point_at_y1_red = 125; + points->x_point_at_y1_green = 125; + points->x_point_at_y1_blue = 125; + + + build_pq(rgb_regamma, + MAX_HW_POINTS, + coordinates_x, + 80); + for (i = 0; i <= MAX_HW_POINTS ; i++) { + points->red[i] = rgb_regamma[i].r; + points->green[i] = rgb_regamma[i].g; + points->blue[i] = rgb_regamma[i].b; + } + ret = true; + + kvfree(rgb_regamma); + } else if (trans == TRANSFER_FUNCTION_SRGB || + trans == TRANSFER_FUNCTION_BT709) { + rgb_regamma = kvzalloc(sizeof(*rgb_regamma) * + (MAX_HW_POINTS + _EXTRA_POINTS), + GFP_KERNEL); + if (!rgb_regamma) + goto rgb_regamma_alloc_fail; + points->end_exponent = 0; + points->x_point_at_y1_red = 1; + points->x_point_at_y1_green = 1; + points->x_point_at_y1_blue = 1; + + build_regamma(rgb_regamma, + MAX_HW_POINTS, + coordinates_x, trans == TRANSFER_FUNCTION_SRGB ? true:false); + for (i = 0; i <= MAX_HW_POINTS ; i++) { + points->red[i] = rgb_regamma[i].r; + points->green[i] = rgb_regamma[i].g; + points->blue[i] = rgb_regamma[i].b; + } + ret = true; + + kvfree(rgb_regamma); + } +rgb_regamma_alloc_fail: + return ret; +} + + +bool mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans, + struct dc_transfer_func_distributed_points *points) +{ + uint32_t i; + bool ret = false; + struct pwl_float_data_ex *rgb_degamma = NULL; + + if (trans == TRANSFER_FUNCTION_UNITY || + trans == TRANSFER_FUNCTION_LINEAR) { + + for (i = 0; i <= MAX_HW_POINTS ; i++) { + points->red[i] = coordinates_x[i].x; + points->green[i] = coordinates_x[i].x; + points->blue[i] = coordinates_x[i].x; + } + ret = true; + } else if (trans == TRANSFER_FUNCTION_PQ) { + rgb_degamma = kvzalloc(sizeof(*rgb_degamma) * + (MAX_HW_POINTS + _EXTRA_POINTS), + GFP_KERNEL); + if (!rgb_degamma) + goto rgb_degamma_alloc_fail; + + + build_de_pq(rgb_degamma, + MAX_HW_POINTS, + coordinates_x); + for (i = 0; i <= MAX_HW_POINTS ; i++) { + points->red[i] = rgb_degamma[i].r; + points->green[i] = rgb_degamma[i].g; + points->blue[i] = rgb_degamma[i].b; + } + ret = true; + + kvfree(rgb_degamma); + } else if (trans == TRANSFER_FUNCTION_SRGB || + trans == TRANSFER_FUNCTION_BT709) { + rgb_degamma = kvzalloc(sizeof(*rgb_degamma) * + (MAX_HW_POINTS + _EXTRA_POINTS), + GFP_KERNEL); + if (!rgb_degamma) + goto rgb_degamma_alloc_fail; + + build_degamma(rgb_degamma, + MAX_HW_POINTS, + coordinates_x, trans == TRANSFER_FUNCTION_SRGB ? true:false); + for (i = 0; i <= MAX_HW_POINTS ; i++) { + points->red[i] = rgb_degamma[i].r; + points->green[i] = rgb_degamma[i].g; + points->blue[i] = rgb_degamma[i].b; + } + ret = true; + + kvfree(rgb_degamma); + } + points->end_exponent = 0; + points->x_point_at_y1_red = 1; + points->x_point_at_y1_green = 1; + points->x_point_at_y1_blue = 1; + +rgb_degamma_alloc_fail: + return ret; +} + +
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.h b/drivers/gpu/drm/amd/display/modules/color/color_gamma.h new file mode 100644 index 0000000..b7f9bc2 --- /dev/null +++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.h
@@ -0,0 +1,53 @@ +/* + * Copyright 2016 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#ifndef COLOR_MOD_COLOR_GAMMA_H_ +#define COLOR_MOD_COLOR_GAMMA_H_ + +struct dc_transfer_func; +struct dc_gamma; +struct dc_transfer_func_distributed_points; +struct dc_rgb_fixed; +enum dc_transfer_func_predefined; + +void setup_x_points_distribution(void); +void precompute_pq(void); +void precompute_de_pq(void); + +bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf, + const struct dc_gamma *ramp, bool mapUserRamp); + +bool mod_color_calculate_degamma_params(struct dc_transfer_func *output_tf, + const struct dc_gamma *ramp, bool mapUserRamp); + +bool mod_color_calculate_curve(enum dc_transfer_func_predefined trans, + struct dc_transfer_func_distributed_points *points); + +bool mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans, + struct dc_transfer_func_distributed_points *points); + + + +#endif /* COLOR_MOD_COLOR_GAMMA_H_ */
diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index b4723af..27d4003 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
@@ -33,7 +33,7 @@ /* Refresh rate ramp at a fixed rate of 65 Hz/second */ #define STATIC_SCREEN_RAMP_DELTA_REFRESH_RATE_PER_FRAME ((1000 / 60) * 65) /* Number of elements in the render times cache array */ -#define RENDER_TIMES_MAX_COUNT 20 +#define RENDER_TIMES_MAX_COUNT 10 /* Threshold to exit BTR (to avoid frequent enter-exits at the lower limit) */ #define BTR_EXIT_MARGIN 2000 /* Number of consecutive frames to check before entering/exiting fixed refresh*/ @@ -46,13 +46,15 @@ #define FREESYNC_NO_STATIC_FOR_INTERNAL_REGKEY "DalFreeSyncNoStaticForInternal" +#define FREESYNC_DEFAULT_REGKEY "LCDFreeSyncDefault" + struct gradual_static_ramp { bool ramp_is_active; bool ramp_direction_is_up; unsigned int ramp_current_frame_duration_in_ns; }; -struct time_cache { +struct freesync_time { /* video (48Hz feature) related */ unsigned int update_duration_in_ns; @@ -64,6 +66,9 @@ struct time_cache { unsigned int render_times_index; unsigned int render_times[RENDER_TIMES_MAX_COUNT]; + + unsigned int min_window; + unsigned int max_window; }; struct below_the_range { @@ -98,11 +103,14 @@ struct freesync_state { bool static_screen; bool video; + unsigned int vmin; + unsigned int vmax; + + struct freesync_time time; + unsigned int nominal_refresh_rate_in_micro_hz; bool windowed_fullscreen; - struct time_cache time; - struct gradual_static_ramp static_ramp; struct below_the_range btr; struct fixed_refresh fixed_refresh; @@ -119,14 +127,16 @@ struct freesync_entity { struct freesync_registry_options { bool drr_external_supported; bool drr_internal_supported; + bool lcd_freesync_default_set; + int lcd_freesync_default_value; }; struct core_freesync { struct mod_freesync public; struct dc *dc; + struct freesync_registry_options opts; struct freesync_entity *map; int num_entities; - struct freesync_registry_options opts; }; #define MOD_FREESYNC_TO_CORE(mod_freesync)\ @@ -146,7 +156,7 @@ struct mod_freesync *mod_freesync_create(struct dc *dc) goto fail_alloc_context; core_freesync->map = kzalloc(sizeof(struct freesync_entity) * MOD_FREESYNC_MAX_CONCURRENT_STREAMS, - GFP_KERNEL); + GFP_KERNEL); if (core_freesync->map == NULL) goto fail_alloc_map; @@ -183,6 +193,16 @@ struct mod_freesync *mod_freesync_create(struct dc *dc) (data & 1) ? false : true; } + if (dm_read_persistent_data(dc->ctx, NULL, NULL, + FREESYNC_DEFAULT_REGKEY, + &data, sizeof(data), &flag)) { + core_freesync->opts.lcd_freesync_default_set = true; + core_freesync->opts.lcd_freesync_default_value = data; + } else { + core_freesync->opts.lcd_freesync_default_set = false; + core_freesync->opts.lcd_freesync_default_value = 0; + } + return &core_freesync->public; fail_construct: @@ -288,6 +308,18 @@ bool mod_freesync_add_stream(struct mod_freesync *mod_freesync, core_freesync->map[core_freesync->num_entities].user_enable. enable_for_video = (persistent_freesync_enable & 4) ? true : false; + /* If FreeSync display and LCDFreeSyncDefault is set, use as default values write back to userenable */ + } else if (caps->supported && (core_freesync->opts.lcd_freesync_default_set)) { + core_freesync->map[core_freesync->num_entities].user_enable.enable_for_gaming = + (core_freesync->opts.lcd_freesync_default_value & 1) ? true : false; + core_freesync->map[core_freesync->num_entities].user_enable.enable_for_static = + (core_freesync->opts.lcd_freesync_default_value & 2) ? true : false; + core_freesync->map[core_freesync->num_entities].user_enable.enable_for_video = + (core_freesync->opts.lcd_freesync_default_value & 4) ? true : false; + dm_write_persistent_data(dc->ctx, stream->sink, + FREESYNC_REGISTRY_NAME, + "userenable", &core_freesync->opts.lcd_freesync_default_value, + sizeof(int), &flag); } else { core_freesync->map[core_freesync->num_entities].user_enable. enable_for_gaming = false; @@ -330,6 +362,25 @@ bool mod_freesync_remove_stream(struct mod_freesync *mod_freesync, return true; } +static void adjust_vmin_vmax(struct core_freesync *core_freesync, + struct dc_stream_state **streams, + int num_streams, + int map_index, + unsigned int v_total_min, + unsigned int v_total_max) +{ + if (num_streams == 0 || streams == NULL || num_streams > 1) + return; + + core_freesync->map[map_index].state.vmin = v_total_min; + core_freesync->map[map_index].state.vmax = v_total_max; + + dc_stream_adjust_vmin_vmax(core_freesync->dc, streams, + num_streams, v_total_min, + v_total_max); +} + + static void update_stream_freesync_context(struct core_freesync *core_freesync, struct dc_stream_state *stream) { @@ -588,9 +639,10 @@ static bool set_freesync_on_streams(struct core_freesync *core_freesync, update_stream_freesync_context(core_freesync, streams[stream_idx]); - dc_stream_adjust_vmin_vmax(core_freesync->dc, streams, - num_streams, v_total_min, - v_total_max); + adjust_vmin_vmax(core_freesync, streams, + num_streams, map_index, + v_total_min, + v_total_max); return true; @@ -613,9 +665,10 @@ static bool set_freesync_on_streams(struct core_freesync *core_freesync, core_freesync, streams[stream_idx]); - dc_stream_adjust_vmin_vmax( - core_freesync->dc, streams, - num_streams, v_total_nominal, + adjust_vmin_vmax( + core_freesync, streams, + num_streams, map_index, + v_total_nominal, v_total_nominal); } return true; @@ -632,9 +685,10 @@ static bool set_freesync_on_streams(struct core_freesync *core_freesync, core_freesync, streams[stream_idx]); - dc_stream_adjust_vmin_vmax(core_freesync->dc, streams, - num_streams, v_total_nominal, - v_total_nominal); + adjust_vmin_vmax(core_freesync, streams, + num_streams, map_index, + v_total_nominal, + v_total_nominal); /* Reset the cached variables */ reset_freesync_state_variables(state); @@ -650,9 +704,10 @@ static bool set_freesync_on_streams(struct core_freesync *core_freesync, * not support freesync because a former stream has * be programmed */ - dc_stream_adjust_vmin_vmax(core_freesync->dc, streams, - num_streams, v_total_nominal, - v_total_nominal); + adjust_vmin_vmax(core_freesync, streams, + num_streams, map_index, + v_total_nominal, + v_total_nominal); /* Reset the cached variables */ reset_freesync_state_variables(state); } @@ -769,8 +824,9 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync, vmin = inserted_frame_v_total; /* Program V_TOTAL */ - dc_stream_adjust_vmin_vmax(core_freesync->dc, streams, - num_streams, vmin, vmax); + adjust_vmin_vmax(core_freesync, streams, + num_streams, index, + vmin, vmax); } if (state->btr.frame_counter > 0) @@ -804,9 +860,10 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync, update_stream_freesync_context(core_freesync, streams[0]); /* Program static screen ramp values */ - dc_stream_adjust_vmin_vmax(core_freesync->dc, streams, - num_streams, v_total, - v_total); + adjust_vmin_vmax(core_freesync, streams, + num_streams, index, + v_total, + v_total); triggers.overlay_update = true; triggers.surface_update = true; @@ -1063,9 +1120,9 @@ bool mod_freesync_override_min_max(struct mod_freesync *mod_freesync, max_refresh); /* Program vtotal min/max */ - dc_stream_adjust_vmin_vmax(core_freesync->dc, &streams, 1, - state->freesync_range.vmin, - state->freesync_range.vmax); + adjust_vmin_vmax(core_freesync, &streams, 1, index, + state->freesync_range.vmin, + state->freesync_range.vmax); } if (min_refresh != 0 && @@ -1399,11 +1456,9 @@ static void apply_fixed_refresh(struct core_freesync *core_freesync, } else { vmin = state->freesync_range.vmin; - vmax = vmin; - - dc_stream_adjust_vmin_vmax(core_freesync->dc, &stream, - 1, vmin, vmax); + adjust_vmin_vmax(core_freesync, &stream, map_index, + 1, vmin, vmax); } } @@ -1457,3 +1512,43 @@ void mod_freesync_pre_update_plane_addresses(struct mod_freesync *mod_freesync, } } + +void mod_freesync_get_settings(struct mod_freesync *mod_freesync, + struct dc_stream_state **streams, int num_streams, + unsigned int *v_total_min, unsigned int *v_total_max, + unsigned int *event_triggers, + unsigned int *window_min, unsigned int *window_max, + unsigned int *lfc_mid_point_in_us, + unsigned int *inserted_frames, + unsigned int *inserted_duration_in_us) +{ + unsigned int stream_index, map_index; + struct core_freesync *core_freesync = NULL; + + if (mod_freesync == NULL) + return; + + core_freesync = MOD_FREESYNC_TO_CORE(mod_freesync); + + for (stream_index = 0; stream_index < num_streams; stream_index++) { + + map_index = map_index_from_stream(core_freesync, + streams[stream_index]); + + if (core_freesync->map[map_index].caps->supported) { + struct freesync_state state = + core_freesync->map[map_index].state; + *v_total_min = state.vmin; + *v_total_max = state.vmax; + *event_triggers = 0; + *window_min = state.time.min_window; + *window_max = state.time.max_window; + *lfc_mid_point_in_us = state.btr.mid_point_in_us; + *inserted_frames = state.btr.frames_to_insert; + *inserted_duration_in_us = + state.btr.inserted_frame_duration_in_us; + } + + } +} +
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h index 84b5342..f083e16 100644 --- a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h +++ b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h
@@ -164,4 +164,13 @@ void mod_freesync_pre_update_plane_addresses(struct mod_freesync *mod_freesync, struct dc_stream_state **streams, int num_streams, unsigned int curr_time_stamp); +void mod_freesync_get_settings(struct mod_freesync *mod_freesync, + struct dc_stream_state **streams, int num_streams, + unsigned int *v_total_min, unsigned int *v_total_max, + unsigned int *event_triggers, + unsigned int *window_min, unsigned int *window_max, + unsigned int *lfc_mid_point_in_us, + unsigned int *inserted_frames, + unsigned int *inserted_duration_in_us); + #endif
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h b/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h new file mode 100644 index 0000000..3230e2a --- /dev/null +++ b/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h
@@ -0,0 +1,65 @@ +/* + * Copyright 2016 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#ifndef MODULES_INC_MOD_STATS_H_ +#define MODULES_INC_MOD_STATS_H_ + +#include "dm_services.h" + +struct mod_stats { + int dummy; +}; + +struct mod_stats_caps { + bool dummy; +}; + +struct mod_stats *mod_stats_create(struct dc *dc); + +void mod_stats_destroy(struct mod_stats *mod_stats); + +bool mod_stats_init(struct mod_stats *mod_stats); + +void mod_stats_dump(struct mod_stats *mod_stats); + +void mod_stats_reset_data(struct mod_stats *mod_stats); + +void mod_stats_update_flip(struct mod_stats *mod_stats, + unsigned long timestamp_in_ns); + +void mod_stats_update_vupdate(struct mod_stats *mod_stats, + unsigned long timestamp_in_ns); + +void mod_stats_update_freesync(struct mod_stats *mod_stats, + unsigned int v_total_min, + unsigned int v_total_max, + unsigned int event_triggers, + unsigned int window_min, + unsigned int window_max, + unsigned int lfc_mid_point_in_us, + unsigned int inserted_frames, + unsigned int inserted_frame_duration_in_us); + +#endif /* MODULES_INC_MOD_STATS_H_ */
diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c new file mode 100644 index 0000000..041f87b --- /dev/null +++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -0,0 +1,334 @@ +/* + * Copyright 2016 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#include "mod_stats.h" +#include "dm_services.h" +#include "dc.h" +#include "core_types.h" + +#define DAL_STATS_ENABLE_REGKEY "DalStatsEnable" +#define DAL_STATS_ENABLE_REGKEY_DEFAULT 0x00000001 +#define DAL_STATS_ENABLE_REGKEY_ENABLED 0x00000001 + +#define DAL_STATS_ENTRIES_REGKEY "DalStatsEntries" +#define DAL_STATS_ENTRIES_REGKEY_DEFAULT 0x00350000 +#define DAL_STATS_ENTRIES_REGKEY_MAX 0x01000000 + +#define MOD_STATS_NUM_VSYNCS 5 + +struct stats_time_cache { + unsigned long flip_timestamp_in_ns; + unsigned long vupdate_timestamp_in_ns; + + unsigned int render_time_in_us; + unsigned int avg_render_time_in_us_last_ten; + unsigned int v_sync_time_in_us[MOD_STATS_NUM_VSYNCS]; + unsigned int num_vsync_between_flips; + + unsigned int flip_to_vsync_time_in_us; + unsigned int vsync_to_flip_time_in_us; + + unsigned int min_window; + unsigned int max_window; + unsigned int v_total_min; + unsigned int v_total_max; + unsigned int event_triggers; + + unsigned int lfc_mid_point_in_us; + unsigned int num_frames_inserted; + unsigned int inserted_duration_in_us; + + unsigned int flags; +}; + +struct core_stats { + struct mod_stats public; + struct dc *dc; + + struct stats_time_cache *time; + unsigned int index; + + bool enabled; + unsigned int entries; +}; + +#define MOD_STATS_TO_CORE(mod_stats)\ + container_of(mod_stats, struct core_stats, public) + +bool mod_stats_init(struct mod_stats *mod_stats) +{ + bool result = false; + struct core_stats *core_stats = NULL; + struct dc *dc = NULL; + + if (mod_stats == NULL) + return false; + + core_stats = MOD_STATS_TO_CORE(mod_stats); + dc = core_stats->dc; + + return result; +} + +struct mod_stats *mod_stats_create(struct dc *dc) +{ + struct core_stats *core_stats = NULL; + struct persistent_data_flag flag; + unsigned int reg_data; + int i = 0; + + core_stats = kzalloc(sizeof(struct core_stats), GFP_KERNEL); + + if (core_stats == NULL) + goto fail_alloc_context; + + if (dc == NULL) + goto fail_construct; + + core_stats->dc = dc; + + core_stats->enabled = DAL_STATS_ENABLE_REGKEY_DEFAULT; + if (dm_read_persistent_data(dc->ctx, NULL, NULL, + DAL_STATS_ENABLE_REGKEY, + ®_data, sizeof(unsigned int), &flag)) + core_stats->enabled = reg_data; + + core_stats->entries = DAL_STATS_ENTRIES_REGKEY_DEFAULT; + if (dm_read_persistent_data(dc->ctx, NULL, NULL, + DAL_STATS_ENTRIES_REGKEY, + ®_data, sizeof(unsigned int), &flag)) { + if (reg_data > DAL_STATS_ENTRIES_REGKEY_MAX) + core_stats->entries = DAL_STATS_ENTRIES_REGKEY_MAX; + else + core_stats->entries = reg_data; + } + + core_stats->time = kzalloc(sizeof(struct stats_time_cache) * core_stats->entries, + GFP_KERNEL); + + if (core_stats->time == NULL) + goto fail_construct; + + /* Purposely leave index 0 unused so we don't need special logic to + * handle calculation cases that depend on previous flip data. + */ + core_stats->index = 1; + + return &core_stats->public; + +fail_construct: + kfree(core_stats); + +fail_alloc_context: + return NULL; +} + +void mod_stats_destroy(struct mod_stats *mod_stats) +{ + if (mod_stats != NULL) { + struct core_stats *core_stats = MOD_STATS_TO_CORE(mod_stats); + + if (core_stats->time != NULL) + kfree(core_stats->time); + + kfree(core_stats); + } +} + +void mod_stats_dump(struct mod_stats *mod_stats) +{ + struct dc *dc = NULL; + struct dal_logger *logger = NULL; + struct core_stats *core_stats = NULL; + struct stats_time_cache *time = NULL; + unsigned int index = 0; + + if (mod_stats == NULL) + return; + + core_stats = MOD_STATS_TO_CORE(mod_stats); + dc = core_stats->dc; + logger = dc->ctx->logger; + time = core_stats->time; + + //LogEntry* pLog = GetLog()->Open(LogMajor_ISR, LogMinor_ISR_FreeSyncSW); + + //if (!pLog->IsDummyEntry()) + { + dm_logger_write(logger, LOG_PROFILING, "==Display Caps==\n"); + dm_logger_write(logger, LOG_PROFILING, "\n"); + dm_logger_write(logger, LOG_PROFILING, "\n"); + + dm_logger_write(logger, LOG_PROFILING, "==Stats==\n"); + dm_logger_write(logger, LOG_PROFILING, + "render avgRender minWindow midPoint maxWindow vsyncToFlip flipToVsync #vsyncBetweenFlip #frame insertDuration vTotalMin vTotalMax eventTrigs vSyncTime1 vSyncTime2 vSyncTime3 vSyncTime4 vSyncTime5 flags\n"); + + for (int i = 0; i < core_stats->index && i < core_stats->entries; i++) { + dm_logger_write(logger, LOG_PROFILING, + "%u %u %u %u %u %u %u %u %u %u %u %u %u %u %u %u %u %u %u\n", + time[i].render_time_in_us, + time[i].avg_render_time_in_us_last_ten, + time[i].min_window, + time[i].lfc_mid_point_in_us, + time[i].max_window, + time[i].vsync_to_flip_time_in_us, + time[i].flip_to_vsync_time_in_us, + time[i].num_vsync_between_flips, + time[i].num_frames_inserted, + time[i].inserted_duration_in_us, + time[i].v_total_min, + time[i].v_total_max, + time[i].event_triggers, + time[i].v_sync_time_in_us[0], + time[i].v_sync_time_in_us[1], + time[i].v_sync_time_in_us[2], + time[i].v_sync_time_in_us[3], + time[i].v_sync_time_in_us[4], + time[i].flags); + } + } + //GetLog()->Close(pLog); + //GetLog()->UnSetLogMask(LogMajor_ISR, LogMinor_ISR_FreeSyncSW); +} + +void mod_stats_reset_data(struct mod_stats *mod_stats) +{ + struct core_stats *core_stats = NULL; + struct stats_time_cache *time = NULL; + unsigned int index = 0; + + if (mod_stats == NULL) + return; + + core_stats = MOD_STATS_TO_CORE(mod_stats); + + memset(core_stats->time, 0, + sizeof(struct stats_time_cache) * core_stats->entries); + + core_stats->index = 0; +} + +void mod_stats_update_flip(struct mod_stats *mod_stats, + unsigned long timestamp_in_ns) +{ + struct core_stats *core_stats = NULL; + struct stats_time_cache *time = NULL; + unsigned int index = 0; + + if (mod_stats == NULL) + return; + + core_stats = MOD_STATS_TO_CORE(mod_stats); + + if (core_stats->index >= core_stats->entries) + return; + + time = core_stats->time; + index = core_stats->index; + + time[index].flip_timestamp_in_ns = timestamp_in_ns; + time[index].render_time_in_us = + timestamp_in_ns - time[index - 1].flip_timestamp_in_ns; + + if (index >= 10) { + for (unsigned int i = 0; i < 10; i++) + time[index].avg_render_time_in_us_last_ten += + time[index - i].render_time_in_us; + time[index].avg_render_time_in_us_last_ten /= 10; + } + + if (time[index].num_vsync_between_flips > 0) + time[index].vsync_to_flip_time_in_us = + timestamp_in_ns - time[index].vupdate_timestamp_in_ns; + else + time[index].vsync_to_flip_time_in_us = + timestamp_in_ns - time[index - 1].vupdate_timestamp_in_ns; + + core_stats->index++; +} + +void mod_stats_update_vupdate(struct mod_stats *mod_stats, + unsigned long timestamp_in_ns) +{ + struct core_stats *core_stats = NULL; + struct stats_time_cache *time = NULL; + unsigned int index = 0; + + if (mod_stats == NULL) + return; + + core_stats = MOD_STATS_TO_CORE(mod_stats); + + if (core_stats->index >= core_stats->entries) + return; + + time = core_stats->time; + index = core_stats->index; + + time[index].vupdate_timestamp_in_ns = timestamp_in_ns; + if (time[index].num_vsync_between_flips < MOD_STATS_NUM_VSYNCS) + time[index].v_sync_time_in_us[time[index].num_vsync_between_flips] = + timestamp_in_ns - time[index - 1].vupdate_timestamp_in_ns; + time[index].flip_to_vsync_time_in_us = + timestamp_in_ns - time[index - 1].flip_timestamp_in_ns; + + time[index].num_vsync_between_flips++; +} + +void mod_stats_update_freesync(struct mod_stats *mod_stats, + unsigned int v_total_min, + unsigned int v_total_max, + unsigned int event_triggers, + unsigned int window_min, + unsigned int window_max, + unsigned int lfc_mid_point_in_us, + unsigned int inserted_frames, + unsigned int inserted_duration_in_us) +{ + struct core_stats *core_stats = NULL; + struct stats_time_cache *time = NULL; + unsigned int index = 0; + + if (mod_stats == NULL) + return; + + core_stats = MOD_STATS_TO_CORE(mod_stats); + + if (core_stats->index >= core_stats->entries) + return; + + time = core_stats->time; + index = core_stats->index; + + time[index].v_total_min = v_total_min; + time[index].v_total_max = v_total_max; + time[index].event_triggers = event_triggers; + time[index].min_window = window_min; + time[index].max_window = window_max; + time[index].lfc_mid_point_in_us = lfc_mid_point_in_us; + time[index].num_frames_inserted = inserted_frames; + time[index].inserted_duration_in_us = inserted_duration_in_us; +} +
diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h index b28d4b6..e2a2f11 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h
@@ -9364,17 +9364,31 @@ #define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_TMZ__SHIFT 0x0 #define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_DCC_EN__SHIFT 0x1 #define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_DCC_IND_64B_BLK__SHIFT 0x2 +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_TMZ_C__SHIFT 0x4 #define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_DCC_IND_64B_BLK_C__SHIFT 0x5 +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_TMZ__SHIFT 0x8 #define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_DCC_EN__SHIFT 0x9 #define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_DCC_IND_64B_BLK__SHIFT 0xa +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_TMZ_C__SHIFT 0xc #define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_DCC_IND_64B_BLK_C__SHIFT 0xd +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_META_SURFACE_TMZ__SHIFT 0x10 +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_META_SURFACE_TMZ_C__SHIFT 0x14 +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_META_SURFACE_TMZ__SHIFT 0x18 +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_META_SURFACE_TMZ_C__SHIFT 0x1c #define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_TMZ_MASK 0x00000001L #define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_DCC_EN_MASK 0x00000002L #define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_DCC_IND_64B_BLK_MASK 0x00000004L +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_TMZ_C_MASK 0x00000010L #define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_SURFACE_DCC_IND_64B_BLK_C_MASK 0x00000020L +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_TMZ_MASK 0x00000100L #define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_DCC_EN_MASK 0x00000200L #define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_DCC_IND_64B_BLK_MASK 0x00000400L +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_TMZ_C_MASK 0x00001000L #define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_SURFACE_DCC_IND_64B_BLK_C_MASK 0x00002000L +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_META_SURFACE_TMZ_MASK 0x00010000L +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__PRIMARY_META_SURFACE_TMZ_C_MASK 0x00100000L +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_META_SURFACE_TMZ_MASK 0x01000000L +#define HUBPREQ0_DCSURF_SURFACE_CONTROL__SECONDARY_META_SURFACE_TMZ_C_MASK 0x10000000L //HUBPREQ0_DCSURF_FLIP_CONTROL #define HUBPREQ0_DCSURF_FLIP_CONTROL__SURFACE_UPDATE_LOCK__SHIFT 0x0 #define HUBPREQ0_DCSURF_FLIP_CONTROL__SURFACE_FLIP_TYPE__SHIFT 0x1
diff --git a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_1_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_1_sh_mask.h new file mode 100644 index 0000000..13bfc2e --- /dev/null +++ b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_1_sh_mask.h
@@ -0,0 +1,31150 @@ +/* + * Copyright (C) 2017 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included + * in all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS + * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN + * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. + */ +#ifndef _gc_9_1_SH_MASK_HEADER +#define _gc_9_1_SH_MASK_HEADER + + +// addressBlock: gc_grbmdec +//GRBM_CNTL +#define GRBM_CNTL__READ_TIMEOUT__SHIFT 0x0 +#define GRBM_CNTL__REPORT_LAST_RDERR__SHIFT 0x1f +#define GRBM_CNTL__READ_TIMEOUT_MASK 0x000000FFL +#define GRBM_CNTL__REPORT_LAST_RDERR_MASK 0x80000000L +//GRBM_SKEW_CNTL +#define GRBM_SKEW_CNTL__SKEW_TOP_THRESHOLD__SHIFT 0x0 +#define GRBM_SKEW_CNTL__SKEW_COUNT__SHIFT 0x6 +#define GRBM_SKEW_CNTL__SKEW_TOP_THRESHOLD_MASK 0x0000003FL +#define GRBM_SKEW_CNTL__SKEW_COUNT_MASK 0x00000FC0L +//GRBM_STATUS2 +#define GRBM_STATUS2__ME0PIPE1_CMDFIFO_AVAIL__SHIFT 0x0 +#define GRBM_STATUS2__ME0PIPE1_CF_RQ_PENDING__SHIFT 0x4 +#define GRBM_STATUS2__ME0PIPE1_PF_RQ_PENDING__SHIFT 0x5 +#define GRBM_STATUS2__ME1PIPE0_RQ_PENDING__SHIFT 0x6 +#define GRBM_STATUS2__ME1PIPE1_RQ_PENDING__SHIFT 0x7 +#define GRBM_STATUS2__ME1PIPE2_RQ_PENDING__SHIFT 0x8 +#define GRBM_STATUS2__ME1PIPE3_RQ_PENDING__SHIFT 0x9 +#define GRBM_STATUS2__ME2PIPE0_RQ_PENDING__SHIFT 0xa +#define GRBM_STATUS2__ME2PIPE1_RQ_PENDING__SHIFT 0xb +#define GRBM_STATUS2__ME2PIPE2_RQ_PENDING__SHIFT 0xc +#define GRBM_STATUS2__ME2PIPE3_RQ_PENDING__SHIFT 0xd +#define GRBM_STATUS2__RLC_RQ_PENDING__SHIFT 0xe +#define GRBM_STATUS2__UTCL2_BUSY__SHIFT 0xf +#define GRBM_STATUS2__EA_BUSY__SHIFT 0x10 +#define GRBM_STATUS2__RMI_BUSY__SHIFT 0x11 +#define GRBM_STATUS2__UTCL2_RQ_PENDING__SHIFT 0x12 +#define GRBM_STATUS2__CPF_RQ_PENDING__SHIFT 0x13 +#define GRBM_STATUS2__EA_LINK_BUSY__SHIFT 0x14 +#define GRBM_STATUS2__RLC_BUSY__SHIFT 0x18 +#define GRBM_STATUS2__TC_BUSY__SHIFT 0x19 +#define GRBM_STATUS2__TCC_CC_RESIDENT__SHIFT 0x1a +#define GRBM_STATUS2__CPF_BUSY__SHIFT 0x1c +#define GRBM_STATUS2__CPC_BUSY__SHIFT 0x1d +#define GRBM_STATUS2__CPG_BUSY__SHIFT 0x1e +#define GRBM_STATUS2__CPAXI_BUSY__SHIFT 0x1f +#define GRBM_STATUS2__ME0PIPE1_CMDFIFO_AVAIL_MASK 0x0000000FL +#define GRBM_STATUS2__ME0PIPE1_CF_RQ_PENDING_MASK 0x00000010L +#define GRBM_STATUS2__ME0PIPE1_PF_RQ_PENDING_MASK 0x00000020L +#define GRBM_STATUS2__ME1PIPE0_RQ_PENDING_MASK 0x00000040L +#define GRBM_STATUS2__ME1PIPE1_RQ_PENDING_MASK 0x00000080L +#define GRBM_STATUS2__ME1PIPE2_RQ_PENDING_MASK 0x00000100L +#define GRBM_STATUS2__ME1PIPE3_RQ_PENDING_MASK 0x00000200L +#define GRBM_STATUS2__ME2PIPE0_RQ_PENDING_MASK 0x00000400L +#define GRBM_STATUS2__ME2PIPE1_RQ_PENDING_MASK 0x00000800L +#define GRBM_STATUS2__ME2PIPE2_RQ_PENDING_MASK 0x00001000L +#define GRBM_STATUS2__ME2PIPE3_RQ_PENDING_MASK 0x00002000L +#define GRBM_STATUS2__RLC_RQ_PENDING_MASK 0x00004000L +#define GRBM_STATUS2__UTCL2_BUSY_MASK 0x00008000L +#define GRBM_STATUS2__EA_BUSY_MASK 0x00010000L +#define GRBM_STATUS2__RMI_BUSY_MASK 0x00020000L +#define GRBM_STATUS2__UTCL2_RQ_PENDING_MASK 0x00040000L +#define GRBM_STATUS2__CPF_RQ_PENDING_MASK 0x00080000L +#define GRBM_STATUS2__EA_LINK_BUSY_MASK 0x00100000L +#define GRBM_STATUS2__RLC_BUSY_MASK 0x01000000L +#define GRBM_STATUS2__TC_BUSY_MASK 0x02000000L +#define GRBM_STATUS2__TCC_CC_RESIDENT_MASK 0x04000000L +#define GRBM_STATUS2__CPF_BUSY_MASK 0x10000000L +#define GRBM_STATUS2__CPC_BUSY_MASK 0x20000000L +#define GRBM_STATUS2__CPG_BUSY_MASK 0x40000000L +#define GRBM_STATUS2__CPAXI_BUSY_MASK 0x80000000L +//GRBM_PWR_CNTL +#define GRBM_PWR_CNTL__ALL_REQ_TYPE__SHIFT 0x0 +#define GRBM_PWR_CNTL__GFX_REQ_TYPE__SHIFT 0x2 +#define GRBM_PWR_CNTL__ALL_RSP_TYPE__SHIFT 0x4 +#define GRBM_PWR_CNTL__GFX_RSP_TYPE__SHIFT 0x6 +#define GRBM_PWR_CNTL__GFX_REQ_EN__SHIFT 0xe +#define GRBM_PWR_CNTL__ALL_REQ_EN__SHIFT 0xf +#define GRBM_PWR_CNTL__ALL_REQ_TYPE_MASK 0x00000003L +#define GRBM_PWR_CNTL__GFX_REQ_TYPE_MASK 0x0000000CL +#define GRBM_PWR_CNTL__ALL_RSP_TYPE_MASK 0x00000030L +#define GRBM_PWR_CNTL__GFX_RSP_TYPE_MASK 0x000000C0L +#define GRBM_PWR_CNTL__GFX_REQ_EN_MASK 0x00004000L +#define GRBM_PWR_CNTL__ALL_REQ_EN_MASK 0x00008000L +//GRBM_STATUS +#define GRBM_STATUS__ME0PIPE0_CMDFIFO_AVAIL__SHIFT 0x0 +#define GRBM_STATUS__RSMU_RQ_PENDING__SHIFT 0x5 +#define GRBM_STATUS__ME0PIPE0_CF_RQ_PENDING__SHIFT 0x7 +#define GRBM_STATUS__ME0PIPE0_PF_RQ_PENDING__SHIFT 0x8 +#define GRBM_STATUS__GDS_DMA_RQ_PENDING__SHIFT 0x9 +#define GRBM_STATUS__DB_CLEAN__SHIFT 0xc +#define GRBM_STATUS__CB_CLEAN__SHIFT 0xd +#define GRBM_STATUS__TA_BUSY__SHIFT 0xe +#define GRBM_STATUS__GDS_BUSY__SHIFT 0xf +#define GRBM_STATUS__WD_BUSY_NO_DMA__SHIFT 0x10 +#define GRBM_STATUS__VGT_BUSY__SHIFT 0x11 +#define GRBM_STATUS__IA_BUSY_NO_DMA__SHIFT 0x12 +#define GRBM_STATUS__IA_BUSY__SHIFT 0x13 +#define GRBM_STATUS__SX_BUSY__SHIFT 0x14 +#define GRBM_STATUS__WD_BUSY__SHIFT 0x15 +#define GRBM_STATUS__SPI_BUSY__SHIFT 0x16 +#define GRBM_STATUS__BCI_BUSY__SHIFT 0x17 +#define GRBM_STATUS__SC_BUSY__SHIFT 0x18 +#define GRBM_STATUS__PA_BUSY__SHIFT 0x19 +#define GRBM_STATUS__DB_BUSY__SHIFT 0x1a +#define GRBM_STATUS__CP_COHERENCY_BUSY__SHIFT 0x1c +#define GRBM_STATUS__CP_BUSY__SHIFT 0x1d +#define GRBM_STATUS__CB_BUSY__SHIFT 0x1e +#define GRBM_STATUS__GUI_ACTIVE__SHIFT 0x1f +#define GRBM_STATUS__ME0PIPE0_CMDFIFO_AVAIL_MASK 0x0000000FL +#define GRBM_STATUS__RSMU_RQ_PENDING_MASK 0x00000020L +#define GRBM_STATUS__ME0PIPE0_CF_RQ_PENDING_MASK 0x00000080L +#define GRBM_STATUS__ME0PIPE0_PF_RQ_PENDING_MASK 0x00000100L +#define GRBM_STATUS__GDS_DMA_RQ_PENDING_MASK 0x00000200L +#define GRBM_STATUS__DB_CLEAN_MASK 0x00001000L +#define GRBM_STATUS__CB_CLEAN_MASK 0x00002000L +#define GRBM_STATUS__TA_BUSY_MASK 0x00004000L +#define GRBM_STATUS__GDS_BUSY_MASK 0x00008000L +#define GRBM_STATUS__WD_BUSY_NO_DMA_MASK 0x00010000L +#define GRBM_STATUS__VGT_BUSY_MASK 0x00020000L +#define GRBM_STATUS__IA_BUSY_NO_DMA_MASK 0x00040000L +#define GRBM_STATUS__IA_BUSY_MASK 0x00080000L +#define GRBM_STATUS__SX_BUSY_MASK 0x00100000L +#define GRBM_STATUS__WD_BUSY_MASK 0x00200000L +#define GRBM_STATUS__SPI_BUSY_MASK 0x00400000L +#define GRBM_STATUS__BCI_BUSY_MASK 0x00800000L +#define GRBM_STATUS__SC_BUSY_MASK 0x01000000L +#define GRBM_STATUS__PA_BUSY_MASK 0x02000000L +#define GRBM_STATUS__DB_BUSY_MASK 0x04000000L +#define GRBM_STATUS__CP_COHERENCY_BUSY_MASK 0x10000000L +#define GRBM_STATUS__CP_BUSY_MASK 0x20000000L +#define GRBM_STATUS__CB_BUSY_MASK 0x40000000L +#define GRBM_STATUS__GUI_ACTIVE_MASK 0x80000000L +//GRBM_STATUS_SE0 +#define GRBM_STATUS_SE0__DB_CLEAN__SHIFT 0x1 +#define GRBM_STATUS_SE0__CB_CLEAN__SHIFT 0x2 +#define GRBM_STATUS_SE0__RMI_BUSY__SHIFT 0x15 +#define GRBM_STATUS_SE0__BCI_BUSY__SHIFT 0x16 +#define GRBM_STATUS_SE0__VGT_BUSY__SHIFT 0x17 +#define GRBM_STATUS_SE0__PA_BUSY__SHIFT 0x18 +#define GRBM_STATUS_SE0__TA_BUSY__SHIFT 0x19 +#define GRBM_STATUS_SE0__SX_BUSY__SHIFT 0x1a +#define GRBM_STATUS_SE0__SPI_BUSY__SHIFT 0x1b +#define GRBM_STATUS_SE0__SC_BUSY__SHIFT 0x1d +#define GRBM_STATUS_SE0__DB_BUSY__SHIFT 0x1e +#define GRBM_STATUS_SE0__CB_BUSY__SHIFT 0x1f +#define GRBM_STATUS_SE0__DB_CLEAN_MASK 0x00000002L +#define GRBM_STATUS_SE0__CB_CLEAN_MASK 0x00000004L +#define GRBM_STATUS_SE0__RMI_BUSY_MASK 0x00200000L +#define GRBM_STATUS_SE0__BCI_BUSY_MASK 0x00400000L +#define GRBM_STATUS_SE0__VGT_BUSY_MASK 0x00800000L +#define GRBM_STATUS_SE0__PA_BUSY_MASK 0x01000000L +#define GRBM_STATUS_SE0__TA_BUSY_MASK 0x02000000L +#define GRBM_STATUS_SE0__SX_BUSY_MASK 0x04000000L +#define GRBM_STATUS_SE0__SPI_BUSY_MASK 0x08000000L +#define GRBM_STATUS_SE0__SC_BUSY_MASK 0x20000000L +#define GRBM_STATUS_SE0__DB_BUSY_MASK 0x40000000L +#define GRBM_STATUS_SE0__CB_BUSY_MASK 0x80000000L +//GRBM_STATUS_SE1 +#define GRBM_STATUS_SE1__DB_CLEAN__SHIFT 0x1 +#define GRBM_STATUS_SE1__CB_CLEAN__SHIFT 0x2 +#define GRBM_STATUS_SE1__RMI_BUSY__SHIFT 0x15 +#define GRBM_STATUS_SE1__BCI_BUSY__SHIFT 0x16 +#define GRBM_STATUS_SE1__VGT_BUSY__SHIFT 0x17 +#define GRBM_STATUS_SE1__PA_BUSY__SHIFT 0x18 +#define GRBM_STATUS_SE1__TA_BUSY__SHIFT 0x19 +#define GRBM_STATUS_SE1__SX_BUSY__SHIFT 0x1a +#define GRBM_STATUS_SE1__SPI_BUSY__SHIFT 0x1b +#define GRBM_STATUS_SE1__SC_BUSY__SHIFT 0x1d +#define GRBM_STATUS_SE1__DB_BUSY__SHIFT 0x1e +#define GRBM_STATUS_SE1__CB_BUSY__SHIFT 0x1f +#define GRBM_STATUS_SE1__DB_CLEAN_MASK 0x00000002L +#define GRBM_STATUS_SE1__CB_CLEAN_MASK 0x00000004L +#define GRBM_STATUS_SE1__RMI_BUSY_MASK 0x00200000L +#define GRBM_STATUS_SE1__BCI_BUSY_MASK 0x00400000L +#define GRBM_STATUS_SE1__VGT_BUSY_MASK 0x00800000L +#define GRBM_STATUS_SE1__PA_BUSY_MASK 0x01000000L +#define GRBM_STATUS_SE1__TA_BUSY_MASK 0x02000000L +#define GRBM_STATUS_SE1__SX_BUSY_MASK 0x04000000L +#define GRBM_STATUS_SE1__SPI_BUSY_MASK 0x08000000L +#define GRBM_STATUS_SE1__SC_BUSY_MASK 0x20000000L +#define GRBM_STATUS_SE1__DB_BUSY_MASK 0x40000000L +#define GRBM_STATUS_SE1__CB_BUSY_MASK 0x80000000L +//GRBM_SOFT_RESET +#define GRBM_SOFT_RESET__SOFT_RESET_CP__SHIFT 0x0 +#define GRBM_SOFT_RESET__SOFT_RESET_RLC__SHIFT 0x2 +#define GRBM_SOFT_RESET__SOFT_RESET_GFX__SHIFT 0x10 +#define GRBM_SOFT_RESET__SOFT_RESET_CPF__SHIFT 0x11 +#define GRBM_SOFT_RESET__SOFT_RESET_CPC__SHIFT 0x12 +#define GRBM_SOFT_RESET__SOFT_RESET_CPG__SHIFT 0x13 +#define GRBM_SOFT_RESET__SOFT_RESET_CAC__SHIFT 0x14 +#define GRBM_SOFT_RESET__SOFT_RESET_CPAXI__SHIFT 0x15 +#define GRBM_SOFT_RESET__SOFT_RESET_EA__SHIFT 0x16 +#define GRBM_SOFT_RESET__SOFT_RESET_CP_MASK 0x00000001L +#define GRBM_SOFT_RESET__SOFT_RESET_RLC_MASK 0x00000004L +#define GRBM_SOFT_RESET__SOFT_RESET_GFX_MASK 0x00010000L +#define GRBM_SOFT_RESET__SOFT_RESET_CPF_MASK 0x00020000L +#define GRBM_SOFT_RESET__SOFT_RESET_CPC_MASK 0x00040000L +#define GRBM_SOFT_RESET__SOFT_RESET_CPG_MASK 0x00080000L +#define GRBM_SOFT_RESET__SOFT_RESET_CAC_MASK 0x00100000L +#define GRBM_SOFT_RESET__SOFT_RESET_CPAXI_MASK 0x00200000L +#define GRBM_SOFT_RESET__SOFT_RESET_EA_MASK 0x00400000L +//GRBM_CGTT_CLK_CNTL +#define GRBM_CGTT_CLK_CNTL__ON_DELAY__SHIFT 0x0 +#define GRBM_CGTT_CLK_CNTL__OFF_HYSTERESIS__SHIFT 0x4 +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE7__SHIFT 0x10 +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE6__SHIFT 0x11 +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE5__SHIFT 0x12 +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE4__SHIFT 0x13 +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE3__SHIFT 0x14 +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE2__SHIFT 0x15 +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE1__SHIFT 0x16 +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE0__SHIFT 0x17 +#define GRBM_CGTT_CLK_CNTL__SOFT_OVERRIDE_DYN__SHIFT 0x1e +#define GRBM_CGTT_CLK_CNTL__ON_DELAY_MASK 0x0000000FL +#define GRBM_CGTT_CLK_CNTL__OFF_HYSTERESIS_MASK 0x00000FF0L +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE7_MASK 0x00010000L +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE6_MASK 0x00020000L +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE5_MASK 0x00040000L +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE4_MASK 0x00080000L +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE3_MASK 0x00100000L +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE2_MASK 0x00200000L +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE1_MASK 0x00400000L +#define GRBM_CGTT_CLK_CNTL__SOFT_STALL_OVERRIDE0_MASK 0x00800000L +#define GRBM_CGTT_CLK_CNTL__SOFT_OVERRIDE_DYN_MASK 0x40000000L +//GRBM_GFX_CLKEN_CNTL +#define GRBM_GFX_CLKEN_CNTL__PREFIX_DELAY_CNT__SHIFT 0x0 +#define GRBM_GFX_CLKEN_CNTL__POST_DELAY_CNT__SHIFT 0x8 +#define GRBM_GFX_CLKEN_CNTL__PREFIX_DELAY_CNT_MASK 0x0000000FL +#define GRBM_GFX_CLKEN_CNTL__POST_DELAY_CNT_MASK 0x00001F00L +//GRBM_WAIT_IDLE_CLOCKS +#define GRBM_WAIT_IDLE_CLOCKS__WAIT_IDLE_CLOCKS__SHIFT 0x0 +#define GRBM_WAIT_IDLE_CLOCKS__WAIT_IDLE_CLOCKS_MASK 0x000000FFL +//GRBM_STATUS_SE2 +#define GRBM_STATUS_SE2__DB_CLEAN__SHIFT 0x1 +#define GRBM_STATUS_SE2__CB_CLEAN__SHIFT 0x2 +#define GRBM_STATUS_SE2__RMI_BUSY__SHIFT 0x15 +#define GRBM_STATUS_SE2__BCI_BUSY__SHIFT 0x16 +#define GRBM_STATUS_SE2__VGT_BUSY__SHIFT 0x17 +#define GRBM_STATUS_SE2__PA_BUSY__SHIFT 0x18 +#define GRBM_STATUS_SE2__TA_BUSY__SHIFT 0x19 +#define GRBM_STATUS_SE2__SX_BUSY__SHIFT 0x1a +#define GRBM_STATUS_SE2__SPI_BUSY__SHIFT 0x1b +#define GRBM_STATUS_SE2__SC_BUSY__SHIFT 0x1d +#define GRBM_STATUS_SE2__DB_BUSY__SHIFT 0x1e +#define GRBM_STATUS_SE2__CB_BUSY__SHIFT 0x1f +#define GRBM_STATUS_SE2__DB_CLEAN_MASK 0x00000002L +#define GRBM_STATUS_SE2__CB_CLEAN_MASK 0x00000004L +#define GRBM_STATUS_SE2__RMI_BUSY_MASK 0x00200000L +#define GRBM_STATUS_SE2__BCI_BUSY_MASK 0x00400000L +#define GRBM_STATUS_SE2__VGT_BUSY_MASK 0x00800000L +#define GRBM_STATUS_SE2__PA_BUSY_MASK 0x01000000L +#define GRBM_STATUS_SE2__TA_BUSY_MASK 0x02000000L +#define GRBM_STATUS_SE2__SX_BUSY_MASK 0x04000000L +#define GRBM_STATUS_SE2__SPI_BUSY_MASK 0x08000000L +#define GRBM_STATUS_SE2__SC_BUSY_MASK 0x20000000L +#define GRBM_STATUS_SE2__DB_BUSY_MASK 0x40000000L +#define GRBM_STATUS_SE2__CB_BUSY_MASK 0x80000000L +//GRBM_STATUS_SE3 +#define GRBM_STATUS_SE3__DB_CLEAN__SHIFT 0x1 +#define GRBM_STATUS_SE3__CB_CLEAN__SHIFT 0x2 +#define GRBM_STATUS_SE3__RMI_BUSY__SHIFT 0x15 +#define GRBM_STATUS_SE3__BCI_BUSY__SHIFT 0x16 +#define GRBM_STATUS_SE3__VGT_BUSY__SHIFT 0x17 +#define GRBM_STATUS_SE3__PA_BUSY__SHIFT 0x18 +#define GRBM_STATUS_SE3__TA_BUSY__SHIFT 0x19 +#define GRBM_STATUS_SE3__SX_BUSY__SHIFT 0x1a +#define GRBM_STATUS_SE3__SPI_BUSY__SHIFT 0x1b +#define GRBM_STATUS_SE3__SC_BUSY__SHIFT 0x1d +#define GRBM_STATUS_SE3__DB_BUSY__SHIFT 0x1e +#define GRBM_STATUS_SE3__CB_BUSY__SHIFT 0x1f +#define GRBM_STATUS_SE3__DB_CLEAN_MASK 0x00000002L +#define GRBM_STATUS_SE3__CB_CLEAN_MASK 0x00000004L +#define GRBM_STATUS_SE3__RMI_BUSY_MASK 0x00200000L +#define GRBM_STATUS_SE3__BCI_BUSY_MASK 0x00400000L +#define GRBM_STATUS_SE3__VGT_BUSY_MASK 0x00800000L +#define GRBM_STATUS_SE3__PA_BUSY_MASK 0x01000000L +#define GRBM_STATUS_SE3__TA_BUSY_MASK 0x02000000L +#define GRBM_STATUS_SE3__SX_BUSY_MASK 0x04000000L +#define GRBM_STATUS_SE3__SPI_BUSY_MASK 0x08000000L +#define GRBM_STATUS_SE3__SC_BUSY_MASK 0x20000000L +#define GRBM_STATUS_SE3__DB_BUSY_MASK 0x40000000L +#define GRBM_STATUS_SE3__CB_BUSY_MASK 0x80000000L +//GRBM_READ_ERROR +#define GRBM_READ_ERROR__READ_ADDRESS__SHIFT 0x2 +#define GRBM_READ_ERROR__READ_PIPEID__SHIFT 0x14 +#define GRBM_READ_ERROR__READ_MEID__SHIFT 0x16 +#define GRBM_READ_ERROR__READ_ERROR__SHIFT 0x1f +#define GRBM_READ_ERROR__READ_ADDRESS_MASK 0x0003FFFCL +#define GRBM_READ_ERROR__READ_PIPEID_MASK 0x00300000L +#define GRBM_READ_ERROR__READ_MEID_MASK 0x00C00000L +#define GRBM_READ_ERROR__READ_ERROR_MASK 0x80000000L +//GRBM_READ_ERROR2 +#define GRBM_READ_ERROR2__READ_REQUESTER_CPF__SHIFT 0x10 +#define GRBM_READ_ERROR2__READ_REQUESTER_RSMU__SHIFT 0x11 +#define GRBM_READ_ERROR2__READ_REQUESTER_RLC__SHIFT 0x12 +#define GRBM_READ_ERROR2__READ_REQUESTER_GDS_DMA__SHIFT 0x13 +#define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE0_CF__SHIFT 0x14 +#define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE0_PF__SHIFT 0x15 +#define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE1_CF__SHIFT 0x16 +#define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE1_PF__SHIFT 0x17 +#define GRBM_READ_ERROR2__READ_REQUESTER_ME1PIPE0__SHIFT 0x18 +#define GRBM_READ_ERROR2__READ_REQUESTER_ME1PIPE1__SHIFT 0x19 +#define GRBM_READ_ERROR2__READ_REQUESTER_ME1PIPE2__SHIFT 0x1a +#define GRBM_READ_ERROR2__READ_REQUESTER_ME1PIPE3__SHIFT 0x1b +#define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE0__SHIFT 0x1c +#define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE1__SHIFT 0x1d +#define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE2__SHIFT 0x1e +#define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE3__SHIFT 0x1f +#define GRBM_READ_ERROR2__READ_REQUESTER_CPF_MASK 0x00010000L +#define GRBM_READ_ERROR2__READ_REQUESTER_RSMU_MASK 0x00020000L +#define GRBM_READ_ERROR2__READ_REQUESTER_RLC_MASK 0x00040000L +#define GRBM_READ_ERROR2__READ_REQUESTER_GDS_DMA_MASK 0x00080000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE0_CF_MASK 0x00100000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE0_PF_MASK 0x00200000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE1_CF_MASK 0x00400000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE1_PF_MASK 0x00800000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME1PIPE0_MASK 0x01000000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME1PIPE1_MASK 0x02000000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME1PIPE2_MASK 0x04000000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME1PIPE3_MASK 0x08000000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE0_MASK 0x10000000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE1_MASK 0x20000000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE2_MASK 0x40000000L +#define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE3_MASK 0x80000000L +//GRBM_INT_CNTL +#define GRBM_INT_CNTL__RDERR_INT_ENABLE__SHIFT 0x0 +#define GRBM_INT_CNTL__GUI_IDLE_INT_ENABLE__SHIFT 0x13 +#define GRBM_INT_CNTL__RDERR_INT_ENABLE_MASK 0x00000001L +#define GRBM_INT_CNTL__GUI_IDLE_INT_ENABLE_MASK 0x00080000L +//GRBM_TRAP_OP +#define GRBM_TRAP_OP__RW__SHIFT 0x0 +#define GRBM_TRAP_OP__RW_MASK 0x00000001L +//GRBM_TRAP_ADDR +#define GRBM_TRAP_ADDR__DATA__SHIFT 0x0 +#define GRBM_TRAP_ADDR__DATA_MASK 0x0003FFFFL +//GRBM_TRAP_ADDR_MSK +#define GRBM_TRAP_ADDR_MSK__DATA__SHIFT 0x0 +#define GRBM_TRAP_ADDR_MSK__DATA_MASK 0x0003FFFFL +//GRBM_TRAP_WD +#define GRBM_TRAP_WD__DATA__SHIFT 0x0 +#define GRBM_TRAP_WD__DATA_MASK 0xFFFFFFFFL +//GRBM_TRAP_WD_MSK +#define GRBM_TRAP_WD_MSK__DATA__SHIFT 0x0 +#define GRBM_TRAP_WD_MSK__DATA_MASK 0xFFFFFFFFL +//GRBM_DSM_BYPASS +#define GRBM_DSM_BYPASS__BYPASS_BITS__SHIFT 0x0 +#define GRBM_DSM_BYPASS__BYPASS_EN__SHIFT 0x2 +#define GRBM_DSM_BYPASS__BYPASS_BITS_MASK 0x00000003L +#define GRBM_DSM_BYPASS__BYPASS_EN_MASK 0x00000004L +//GRBM_WRITE_ERROR +#define GRBM_WRITE_ERROR__WRITE_REQUESTER_RLC__SHIFT 0x0 +#define GRBM_WRITE_ERROR__WRITE_REQUESTER_RSMU__SHIFT 0x1 +#define GRBM_WRITE_ERROR__WRITE_SSRCID__SHIFT 0x2 +#define GRBM_WRITE_ERROR__WRITE_VFID__SHIFT 0x5 +#define GRBM_WRITE_ERROR__WRITE_VF__SHIFT 0xc +#define GRBM_WRITE_ERROR__WRITE_VMID__SHIFT 0xd +#define GRBM_WRITE_ERROR__WRITE_PIPEID__SHIFT 0x14 +#define GRBM_WRITE_ERROR__WRITE_MEID__SHIFT 0x16 +#define GRBM_WRITE_ERROR__WRITE_ERROR__SHIFT 0x1f +#define GRBM_WRITE_ERROR__WRITE_REQUESTER_RLC_MASK 0x00000001L +#define GRBM_WRITE_ERROR__WRITE_REQUESTER_RSMU_MASK 0x00000002L +#define GRBM_WRITE_ERROR__WRITE_SSRCID_MASK 0x0000001CL +#define GRBM_WRITE_ERROR__WRITE_VFID_MASK 0x000001E0L +#define GRBM_WRITE_ERROR__WRITE_VF_MASK 0x00001000L +#define GRBM_WRITE_ERROR__WRITE_VMID_MASK 0x0001E000L +#define GRBM_WRITE_ERROR__WRITE_PIPEID_MASK 0x00300000L +#define GRBM_WRITE_ERROR__WRITE_MEID_MASK 0x00C00000L +#define GRBM_WRITE_ERROR__WRITE_ERROR_MASK 0x80000000L +//GRBM_IOV_ERROR +#define GRBM_IOV_ERROR__IOV_ADDR__SHIFT 0x2 +#define GRBM_IOV_ERROR__IOV_VFID__SHIFT 0x14 +#define GRBM_IOV_ERROR__IOV_VF__SHIFT 0x1a +#define GRBM_IOV_ERROR__IOV_OP__SHIFT 0x1b +#define GRBM_IOV_ERROR__IOV_ERROR__SHIFT 0x1f +#define GRBM_IOV_ERROR__IOV_ADDR_MASK 0x000FFFFCL +#define GRBM_IOV_ERROR__IOV_VFID_MASK 0x03F00000L +#define GRBM_IOV_ERROR__IOV_VF_MASK 0x04000000L +#define GRBM_IOV_ERROR__IOV_OP_MASK 0x08000000L +#define GRBM_IOV_ERROR__IOV_ERROR_MASK 0x80000000L +//GRBM_CHIP_REVISION +#define GRBM_CHIP_REVISION__CHIP_REVISION__SHIFT 0x0 +#define GRBM_CHIP_REVISION__CHIP_REVISION_MASK 0x000000FFL +//GRBM_GFX_CNTL +#define GRBM_GFX_CNTL__PIPEID__SHIFT 0x0 +#define GRBM_GFX_CNTL__MEID__SHIFT 0x2 +#define GRBM_GFX_CNTL__VMID__SHIFT 0x4 +#define GRBM_GFX_CNTL__QUEUEID__SHIFT 0x8 +#define GRBM_GFX_CNTL__PIPEID_MASK 0x00000003L +#define GRBM_GFX_CNTL__MEID_MASK 0x0000000CL +#define GRBM_GFX_CNTL__VMID_MASK 0x000000F0L +#define GRBM_GFX_CNTL__QUEUEID_MASK 0x00000700L +//GRBM_RSMU_CFG +#define GRBM_RSMU_CFG__APERTURE_ID__SHIFT 0x0 +#define GRBM_RSMU_CFG__QOS__SHIFT 0xc +#define GRBM_RSMU_CFG__POSTED_WR__SHIFT 0x10 +#define GRBM_RSMU_CFG__APERTURE_ID_MASK 0x00000FFFL +#define GRBM_RSMU_CFG__QOS_MASK 0x0000F000L +#define GRBM_RSMU_CFG__POSTED_WR_MASK 0x00010000L +//GRBM_IH_CREDIT +#define GRBM_IH_CREDIT__CREDIT_VALUE__SHIFT 0x0 +#define GRBM_IH_CREDIT__IH_CLIENT_ID__SHIFT 0x10 +#define GRBM_IH_CREDIT__CREDIT_VALUE_MASK 0x00000003L +#define GRBM_IH_CREDIT__IH_CLIENT_ID_MASK 0x00FF0000L +//GRBM_PWR_CNTL2 +#define GRBM_PWR_CNTL2__PWR_REQUEST_HALT__SHIFT 0x10 +#define GRBM_PWR_CNTL2__PWR_GFX3D_REQUEST_HALT__SHIFT 0x14 +#define GRBM_PWR_CNTL2__PWR_REQUEST_HALT_MASK 0x00010000L +#define GRBM_PWR_CNTL2__PWR_GFX3D_REQUEST_HALT_MASK 0x00100000L +//GRBM_UTCL2_INVAL_RANGE_START +#define GRBM_UTCL2_INVAL_RANGE_START__DATA__SHIFT 0x0 +#define GRBM_UTCL2_INVAL_RANGE_START__DATA_MASK 0x0003FFFFL +//GRBM_UTCL2_INVAL_RANGE_END +#define GRBM_UTCL2_INVAL_RANGE_END__DATA__SHIFT 0x0 +#define GRBM_UTCL2_INVAL_RANGE_END__DATA_MASK 0x0003FFFFL +//GRBM_RSMU_READ_ERROR +#define GRBM_RSMU_READ_ERROR__RSMU_READ_ADDRESS__SHIFT 0x2 +#define GRBM_RSMU_READ_ERROR__RSMU_READ_VF__SHIFT 0x14 +#define GRBM_RSMU_READ_ERROR__RSMU_READ_VFID__SHIFT 0x15 +#define GRBM_RSMU_READ_ERROR__RSMU_READ_ERROR_TYPE__SHIFT 0x1b +#define GRBM_RSMU_READ_ERROR__RSMU_READ_ERROR__SHIFT 0x1f +#define GRBM_RSMU_READ_ERROR__RSMU_READ_ADDRESS_MASK 0x000FFFFCL +#define GRBM_RSMU_READ_ERROR__RSMU_READ_VF_MASK 0x00100000L +#define GRBM_RSMU_READ_ERROR__RSMU_READ_VFID_MASK 0x07E00000L +#define GRBM_RSMU_READ_ERROR__RSMU_READ_ERROR_TYPE_MASK 0x08000000L +#define GRBM_RSMU_READ_ERROR__RSMU_READ_ERROR_MASK 0x80000000L +//GRBM_CHICKEN_BITS +#define GRBM_CHICKEN_BITS__DISABLE_CP_VMID_RESET_REQ__SHIFT 0x0 +#define GRBM_CHICKEN_BITS__DISABLE_CP_VMID_RESET_REQ_MASK 0x00000001L +//GRBM_NOWHERE +#define GRBM_NOWHERE__DATA__SHIFT 0x0 +#define GRBM_NOWHERE__DATA_MASK 0xFFFFFFFFL +//GRBM_SCRATCH_REG0 +#define GRBM_SCRATCH_REG0__SCRATCH_REG0__SHIFT 0x0 +#define GRBM_SCRATCH_REG0__SCRATCH_REG0_MASK 0xFFFFFFFFL +//GRBM_SCRATCH_REG1 +#define GRBM_SCRATCH_REG1__SCRATCH_REG1__SHIFT 0x0 +#define GRBM_SCRATCH_REG1__SCRATCH_REG1_MASK 0xFFFFFFFFL +//GRBM_SCRATCH_REG2 +#define GRBM_SCRATCH_REG2__SCRATCH_REG2__SHIFT 0x0 +#define GRBM_SCRATCH_REG2__SCRATCH_REG2_MASK 0xFFFFFFFFL +//GRBM_SCRATCH_REG3 +#define GRBM_SCRATCH_REG3__SCRATCH_REG3__SHIFT 0x0 +#define GRBM_SCRATCH_REG3__SCRATCH_REG3_MASK 0xFFFFFFFFL +//GRBM_SCRATCH_REG4 +#define GRBM_SCRATCH_REG4__SCRATCH_REG4__SHIFT 0x0 +#define GRBM_SCRATCH_REG4__SCRATCH_REG4_MASK 0xFFFFFFFFL +//GRBM_SCRATCH_REG5 +#define GRBM_SCRATCH_REG5__SCRATCH_REG5__SHIFT 0x0 +#define GRBM_SCRATCH_REG5__SCRATCH_REG5_MASK 0xFFFFFFFFL +//GRBM_SCRATCH_REG6 +#define GRBM_SCRATCH_REG6__SCRATCH_REG6__SHIFT 0x0 +#define GRBM_SCRATCH_REG6__SCRATCH_REG6_MASK 0xFFFFFFFFL +//GRBM_SCRATCH_REG7 +#define GRBM_SCRATCH_REG7__SCRATCH_REG7__SHIFT 0x0 +#define GRBM_SCRATCH_REG7__SCRATCH_REG7_MASK 0xFFFFFFFFL + + +// addressBlock: gc_cpdec +//CP_CPC_STATUS +#define CP_CPC_STATUS__MEC1_BUSY__SHIFT 0x0 +#define CP_CPC_STATUS__MEC2_BUSY__SHIFT 0x1 +#define CP_CPC_STATUS__DC0_BUSY__SHIFT 0x2 +#define CP_CPC_STATUS__DC1_BUSY__SHIFT 0x3 +#define CP_CPC_STATUS__RCIU1_BUSY__SHIFT 0x4 +#define CP_CPC_STATUS__RCIU2_BUSY__SHIFT 0x5 +#define CP_CPC_STATUS__ROQ1_BUSY__SHIFT 0x6 +#define CP_CPC_STATUS__ROQ2_BUSY__SHIFT 0x7 +#define CP_CPC_STATUS__TCIU_BUSY__SHIFT 0xa +#define CP_CPC_STATUS__SCRATCH_RAM_BUSY__SHIFT 0xb +#define CP_CPC_STATUS__QU_BUSY__SHIFT 0xc +#define CP_CPC_STATUS__UTCL2IU_BUSY__SHIFT 0xd +#define CP_CPC_STATUS__SAVE_RESTORE_BUSY__SHIFT 0xe +#define CP_CPC_STATUS__CPG_CPC_BUSY__SHIFT 0x1d +#define CP_CPC_STATUS__CPF_CPC_BUSY__SHIFT 0x1e +#define CP_CPC_STATUS__CPC_BUSY__SHIFT 0x1f +#define CP_CPC_STATUS__MEC1_BUSY_MASK 0x00000001L +#define CP_CPC_STATUS__MEC2_BUSY_MASK 0x00000002L +#define CP_CPC_STATUS__DC0_BUSY_MASK 0x00000004L +#define CP_CPC_STATUS__DC1_BUSY_MASK 0x00000008L +#define CP_CPC_STATUS__RCIU1_BUSY_MASK 0x00000010L +#define CP_CPC_STATUS__RCIU2_BUSY_MASK 0x00000020L +#define CP_CPC_STATUS__ROQ1_BUSY_MASK 0x00000040L +#define CP_CPC_STATUS__ROQ2_BUSY_MASK 0x00000080L +#define CP_CPC_STATUS__TCIU_BUSY_MASK 0x00000400L +#define CP_CPC_STATUS__SCRATCH_RAM_BUSY_MASK 0x00000800L +#define CP_CPC_STATUS__QU_BUSY_MASK 0x00001000L +#define CP_CPC_STATUS__UTCL2IU_BUSY_MASK 0x00002000L +#define CP_CPC_STATUS__SAVE_RESTORE_BUSY_MASK 0x00004000L +#define CP_CPC_STATUS__CPG_CPC_BUSY_MASK 0x20000000L +#define CP_CPC_STATUS__CPF_CPC_BUSY_MASK 0x40000000L +#define CP_CPC_STATUS__CPC_BUSY_MASK 0x80000000L +//CP_CPC_BUSY_STAT +#define CP_CPC_BUSY_STAT__MEC1_LOAD_BUSY__SHIFT 0x0 +#define CP_CPC_BUSY_STAT__MEC1_SEMAPOHRE_BUSY__SHIFT 0x1 +#define CP_CPC_BUSY_STAT__MEC1_MUTEX_BUSY__SHIFT 0x2 +#define CP_CPC_BUSY_STAT__MEC1_MESSAGE_BUSY__SHIFT 0x3 +#define CP_CPC_BUSY_STAT__MEC1_EOP_QUEUE_BUSY__SHIFT 0x4 +#define CP_CPC_BUSY_STAT__MEC1_IQ_QUEUE_BUSY__SHIFT 0x5 +#define CP_CPC_BUSY_STAT__MEC1_IB_QUEUE_BUSY__SHIFT 0x6 +#define CP_CPC_BUSY_STAT__MEC1_TC_BUSY__SHIFT 0x7 +#define CP_CPC_BUSY_STAT__MEC1_DMA_BUSY__SHIFT 0x8 +#define CP_CPC_BUSY_STAT__MEC1_PARTIAL_FLUSH_BUSY__SHIFT 0x9 +#define CP_CPC_BUSY_STAT__MEC1_PIPE0_BUSY__SHIFT 0xa +#define CP_CPC_BUSY_STAT__MEC1_PIPE1_BUSY__SHIFT 0xb +#define CP_CPC_BUSY_STAT__MEC1_PIPE2_BUSY__SHIFT 0xc +#define CP_CPC_BUSY_STAT__MEC1_PIPE3_BUSY__SHIFT 0xd +#define CP_CPC_BUSY_STAT__MEC2_LOAD_BUSY__SHIFT 0x10 +#define CP_CPC_BUSY_STAT__MEC2_SEMAPOHRE_BUSY__SHIFT 0x11 +#define CP_CPC_BUSY_STAT__MEC2_MUTEX_BUSY__SHIFT 0x12 +#define CP_CPC_BUSY_STAT__MEC2_MESSAGE_BUSY__SHIFT 0x13 +#define CP_CPC_BUSY_STAT__MEC2_EOP_QUEUE_BUSY__SHIFT 0x14 +#define CP_CPC_BUSY_STAT__MEC2_IQ_QUEUE_BUSY__SHIFT 0x15 +#define CP_CPC_BUSY_STAT__MEC2_IB_QUEUE_BUSY__SHIFT 0x16 +#define CP_CPC_BUSY_STAT__MEC2_TC_BUSY__SHIFT 0x17 +#define CP_CPC_BUSY_STAT__MEC2_DMA_BUSY__SHIFT 0x18 +#define CP_CPC_BUSY_STAT__MEC2_PARTIAL_FLUSH_BUSY__SHIFT 0x19 +#define CP_CPC_BUSY_STAT__MEC2_PIPE0_BUSY__SHIFT 0x1a +#define CP_CPC_BUSY_STAT__MEC2_PIPE1_BUSY__SHIFT 0x1b +#define CP_CPC_BUSY_STAT__MEC2_PIPE2_BUSY__SHIFT 0x1c +#define CP_CPC_BUSY_STAT__MEC2_PIPE3_BUSY__SHIFT 0x1d +#define CP_CPC_BUSY_STAT__MEC1_LOAD_BUSY_MASK 0x00000001L +#define CP_CPC_BUSY_STAT__MEC1_SEMAPOHRE_BUSY_MASK 0x00000002L +#define CP_CPC_BUSY_STAT__MEC1_MUTEX_BUSY_MASK 0x00000004L +#define CP_CPC_BUSY_STAT__MEC1_MESSAGE_BUSY_MASK 0x00000008L +#define CP_CPC_BUSY_STAT__MEC1_EOP_QUEUE_BUSY_MASK 0x00000010L +#define CP_CPC_BUSY_STAT__MEC1_IQ_QUEUE_BUSY_MASK 0x00000020L +#define CP_CPC_BUSY_STAT__MEC1_IB_QUEUE_BUSY_MASK 0x00000040L +#define CP_CPC_BUSY_STAT__MEC1_TC_BUSY_MASK 0x00000080L +#define CP_CPC_BUSY_STAT__MEC1_DMA_BUSY_MASK 0x00000100L +#define CP_CPC_BUSY_STAT__MEC1_PARTIAL_FLUSH_BUSY_MASK 0x00000200L +#define CP_CPC_BUSY_STAT__MEC1_PIPE0_BUSY_MASK 0x00000400L +#define CP_CPC_BUSY_STAT__MEC1_PIPE1_BUSY_MASK 0x00000800L +#define CP_CPC_BUSY_STAT__MEC1_PIPE2_BUSY_MASK 0x00001000L +#define CP_CPC_BUSY_STAT__MEC1_PIPE3_BUSY_MASK 0x00002000L +#define CP_CPC_BUSY_STAT__MEC2_LOAD_BUSY_MASK 0x00010000L +#define CP_CPC_BUSY_STAT__MEC2_SEMAPOHRE_BUSY_MASK 0x00020000L +#define CP_CPC_BUSY_STAT__MEC2_MUTEX_BUSY_MASK 0x00040000L +#define CP_CPC_BUSY_STAT__MEC2_MESSAGE_BUSY_MASK 0x00080000L +#define CP_CPC_BUSY_STAT__MEC2_EOP_QUEUE_BUSY_MASK 0x00100000L +#define CP_CPC_BUSY_STAT__MEC2_IQ_QUEUE_BUSY_MASK 0x00200000L +#define CP_CPC_BUSY_STAT__MEC2_IB_QUEUE_BUSY_MASK 0x00400000L +#define CP_CPC_BUSY_STAT__MEC2_TC_BUSY_MASK 0x00800000L +#define CP_CPC_BUSY_STAT__MEC2_DMA_BUSY_MASK 0x01000000L +#define CP_CPC_BUSY_STAT__MEC2_PARTIAL_FLUSH_BUSY_MASK 0x02000000L +#define CP_CPC_BUSY_STAT__MEC2_PIPE0_BUSY_MASK 0x04000000L +#define CP_CPC_BUSY_STAT__MEC2_PIPE1_BUSY_MASK 0x08000000L +#define CP_CPC_BUSY_STAT__MEC2_PIPE2_BUSY_MASK 0x10000000L +#define CP_CPC_BUSY_STAT__MEC2_PIPE3_BUSY_MASK 0x20000000L +//CP_CPC_STALLED_STAT1 +#define CP_CPC_STALLED_STAT1__RCIU_TX_FREE_STALL__SHIFT 0x3 +#define CP_CPC_STALLED_STAT1__RCIU_PRIV_VIOLATION__SHIFT 0x4 +#define CP_CPC_STALLED_STAT1__TCIU_TX_FREE_STALL__SHIFT 0x6 +#define CP_CPC_STALLED_STAT1__MEC1_DECODING_PACKET__SHIFT 0x8 +#define CP_CPC_STALLED_STAT1__MEC1_WAIT_ON_RCIU__SHIFT 0x9 +#define CP_CPC_STALLED_STAT1__MEC1_WAIT_ON_RCIU_READ__SHIFT 0xa +#define CP_CPC_STALLED_STAT1__MEC1_WAIT_ON_ROQ_DATA__SHIFT 0xd +#define CP_CPC_STALLED_STAT1__MEC2_DECODING_PACKET__SHIFT 0x10 +#define CP_CPC_STALLED_STAT1__MEC2_WAIT_ON_RCIU__SHIFT 0x11 +#define CP_CPC_STALLED_STAT1__MEC2_WAIT_ON_RCIU_READ__SHIFT 0x12 +#define CP_CPC_STALLED_STAT1__MEC2_WAIT_ON_ROQ_DATA__SHIFT 0x15 +#define CP_CPC_STALLED_STAT1__UTCL2IU_WAITING_ON_FREE__SHIFT 0x16 +#define CP_CPC_STALLED_STAT1__UTCL2IU_WAITING_ON_TAGS__SHIFT 0x17 +#define CP_CPC_STALLED_STAT1__UTCL1_WAITING_ON_TRANS__SHIFT 0x18 +#define CP_CPC_STALLED_STAT1__RCIU_TX_FREE_STALL_MASK 0x00000008L +#define CP_CPC_STALLED_STAT1__RCIU_PRIV_VIOLATION_MASK 0x00000010L +#define CP_CPC_STALLED_STAT1__TCIU_TX_FREE_STALL_MASK 0x00000040L +#define CP_CPC_STALLED_STAT1__MEC1_DECODING_PACKET_MASK 0x00000100L +#define CP_CPC_STALLED_STAT1__MEC1_WAIT_ON_RCIU_MASK 0x00000200L +#define CP_CPC_STALLED_STAT1__MEC1_WAIT_ON_RCIU_READ_MASK 0x00000400L +#define CP_CPC_STALLED_STAT1__MEC1_WAIT_ON_ROQ_DATA_MASK 0x00002000L +#define CP_CPC_STALLED_STAT1__MEC2_DECODING_PACKET_MASK 0x00010000L +#define CP_CPC_STALLED_STAT1__MEC2_WAIT_ON_RCIU_MASK 0x00020000L +#define CP_CPC_STALLED_STAT1__MEC2_WAIT_ON_RCIU_READ_MASK 0x00040000L +#define CP_CPC_STALLED_STAT1__MEC2_WAIT_ON_ROQ_DATA_MASK 0x00200000L +#define CP_CPC_STALLED_STAT1__UTCL2IU_WAITING_ON_FREE_MASK 0x00400000L +#define CP_CPC_STALLED_STAT1__UTCL2IU_WAITING_ON_TAGS_MASK 0x00800000L +#define CP_CPC_STALLED_STAT1__UTCL1_WAITING_ON_TRANS_MASK 0x01000000L +//CP_CPF_STATUS +#define CP_CPF_STATUS__POST_WPTR_GFX_BUSY__SHIFT 0x0 +#define CP_CPF_STATUS__CSF_BUSY__SHIFT 0x1 +#define CP_CPF_STATUS__ROQ_ALIGN_BUSY__SHIFT 0x4 +#define CP_CPF_STATUS__ROQ_RING_BUSY__SHIFT 0x5 +#define CP_CPF_STATUS__ROQ_INDIRECT1_BUSY__SHIFT 0x6 +#define CP_CPF_STATUS__ROQ_INDIRECT2_BUSY__SHIFT 0x7 +#define CP_CPF_STATUS__ROQ_STATE_BUSY__SHIFT 0x8 +#define CP_CPF_STATUS__ROQ_CE_RING_BUSY__SHIFT 0x9 +#define CP_CPF_STATUS__ROQ_CE_INDIRECT1_BUSY__SHIFT 0xa +#define CP_CPF_STATUS__ROQ_CE_INDIRECT2_BUSY__SHIFT 0xb +#define CP_CPF_STATUS__SEMAPHORE_BUSY__SHIFT 0xc +#define CP_CPF_STATUS__INTERRUPT_BUSY__SHIFT 0xd +#define CP_CPF_STATUS__TCIU_BUSY__SHIFT 0xe +#define CP_CPF_STATUS__HQD_BUSY__SHIFT 0xf +#define CP_CPF_STATUS__PRT_BUSY__SHIFT 0x10 +#define CP_CPF_STATUS__UTCL2IU_BUSY__SHIFT 0x11 +#define CP_CPF_STATUS__CPF_GFX_BUSY__SHIFT 0x1a +#define CP_CPF_STATUS__CPF_CMP_BUSY__SHIFT 0x1b +#define CP_CPF_STATUS__GRBM_CPF_STAT_BUSY__SHIFT 0x1c +#define CP_CPF_STATUS__CPC_CPF_BUSY__SHIFT 0x1e +#define CP_CPF_STATUS__CPF_BUSY__SHIFT 0x1f +#define CP_CPF_STATUS__POST_WPTR_GFX_BUSY_MASK 0x00000001L +#define CP_CPF_STATUS__CSF_BUSY_MASK 0x00000002L +#define CP_CPF_STATUS__ROQ_ALIGN_BUSY_MASK 0x00000010L +#define CP_CPF_STATUS__ROQ_RING_BUSY_MASK 0x00000020L +#define CP_CPF_STATUS__ROQ_INDIRECT1_BUSY_MASK 0x00000040L +#define CP_CPF_STATUS__ROQ_INDIRECT2_BUSY_MASK 0x00000080L +#define CP_CPF_STATUS__ROQ_STATE_BUSY_MASK 0x00000100L +#define CP_CPF_STATUS__ROQ_CE_RING_BUSY_MASK 0x00000200L +#define CP_CPF_STATUS__ROQ_CE_INDIRECT1_BUSY_MASK 0x00000400L +#define CP_CPF_STATUS__ROQ_CE_INDIRECT2_BUSY_MASK 0x00000800L +#define CP_CPF_STATUS__SEMAPHORE_BUSY_MASK 0x00001000L +#define CP_CPF_STATUS__INTERRUPT_BUSY_MASK 0x00002000L +#define CP_CPF_STATUS__TCIU_BUSY_MASK 0x00004000L +#define CP_CPF_STATUS__HQD_BUSY_MASK 0x00008000L +#define CP_CPF_STATUS__PRT_BUSY_MASK 0x00010000L +#define CP_CPF_STATUS__UTCL2IU_BUSY_MASK 0x00020000L +#define CP_CPF_STATUS__CPF_GFX_BUSY_MASK 0x04000000L +#define CP_CPF_STATUS__CPF_CMP_BUSY_MASK 0x08000000L +#define CP_CPF_STATUS__GRBM_CPF_STAT_BUSY_MASK 0x30000000L +#define CP_CPF_STATUS__CPC_CPF_BUSY_MASK 0x40000000L +#define CP_CPF_STATUS__CPF_BUSY_MASK 0x80000000L +//CP_CPF_BUSY_STAT +#define CP_CPF_BUSY_STAT__REG_BUS_FIFO_BUSY__SHIFT 0x0 +#define CP_CPF_BUSY_STAT__CSF_RING_BUSY__SHIFT 0x1 +#define CP_CPF_BUSY_STAT__CSF_INDIRECT1_BUSY__SHIFT 0x2 +#define CP_CPF_BUSY_STAT__CSF_INDIRECT2_BUSY__SHIFT 0x3 +#define CP_CPF_BUSY_STAT__CSF_STATE_BUSY__SHIFT 0x4 +#define CP_CPF_BUSY_STAT__CSF_CE_INDR1_BUSY__SHIFT 0x5 +#define CP_CPF_BUSY_STAT__CSF_CE_INDR2_BUSY__SHIFT 0x6 +#define CP_CPF_BUSY_STAT__CSF_ARBITER_BUSY__SHIFT 0x7 +#define CP_CPF_BUSY_STAT__CSF_INPUT_BUSY__SHIFT 0x8 +#define CP_CPF_BUSY_STAT__OUTSTANDING_READ_TAGS__SHIFT 0x9 +#define CP_CPF_BUSY_STAT__HPD_PROCESSING_EOP_BUSY__SHIFT 0xb +#define CP_CPF_BUSY_STAT__HQD_DISPATCH_BUSY__SHIFT 0xc +#define CP_CPF_BUSY_STAT__HQD_IQ_TIMER_BUSY__SHIFT 0xd +#define CP_CPF_BUSY_STAT__HQD_DMA_OFFLOAD_BUSY__SHIFT 0xe +#define CP_CPF_BUSY_STAT__HQD_WAIT_SEMAPHORE_BUSY__SHIFT 0xf +#define CP_CPF_BUSY_STAT__HQD_SIGNAL_SEMAPHORE_BUSY__SHIFT 0x10 +#define CP_CPF_BUSY_STAT__HQD_MESSAGE_BUSY__SHIFT 0x11 +#define CP_CPF_BUSY_STAT__HQD_PQ_FETCHER_BUSY__SHIFT 0x12 +#define CP_CPF_BUSY_STAT__HQD_IB_FETCHER_BUSY__SHIFT 0x13 +#define CP_CPF_BUSY_STAT__HQD_IQ_FETCHER_BUSY__SHIFT 0x14 +#define CP_CPF_BUSY_STAT__HQD_EOP_FETCHER_BUSY__SHIFT 0x15 +#define CP_CPF_BUSY_STAT__HQD_CONSUMED_RPTR_BUSY__SHIFT 0x16 +#define CP_CPF_BUSY_STAT__HQD_FETCHER_ARB_BUSY__SHIFT 0x17 +#define CP_CPF_BUSY_STAT__HQD_ROQ_ALIGN_BUSY__SHIFT 0x18 +#define CP_CPF_BUSY_STAT__HQD_ROQ_EOP_BUSY__SHIFT 0x19 +#define CP_CPF_BUSY_STAT__HQD_ROQ_IQ_BUSY__SHIFT 0x1a +#define CP_CPF_BUSY_STAT__HQD_ROQ_PQ_BUSY__SHIFT 0x1b +#define CP_CPF_BUSY_STAT__HQD_ROQ_IB_BUSY__SHIFT 0x1c +#define CP_CPF_BUSY_STAT__HQD_WPTR_POLL_BUSY__SHIFT 0x1d +#define CP_CPF_BUSY_STAT__HQD_PQ_BUSY__SHIFT 0x1e +#define CP_CPF_BUSY_STAT__HQD_IB_BUSY__SHIFT 0x1f +#define CP_CPF_BUSY_STAT__REG_BUS_FIFO_BUSY_MASK 0x00000001L +#define CP_CPF_BUSY_STAT__CSF_RING_BUSY_MASK 0x00000002L +#define CP_CPF_BUSY_STAT__CSF_INDIRECT1_BUSY_MASK 0x00000004L +#define CP_CPF_BUSY_STAT__CSF_INDIRECT2_BUSY_MASK 0x00000008L +#define CP_CPF_BUSY_STAT__CSF_STATE_BUSY_MASK 0x00000010L +#define CP_CPF_BUSY_STAT__CSF_CE_INDR1_BUSY_MASK 0x00000020L +#define CP_CPF_BUSY_STAT__CSF_CE_INDR2_BUSY_MASK 0x00000040L +#define CP_CPF_BUSY_STAT__CSF_ARBITER_BUSY_MASK 0x00000080L +#define CP_CPF_BUSY_STAT__CSF_INPUT_BUSY_MASK 0x00000100L +#define CP_CPF_BUSY_STAT__OUTSTANDING_READ_TAGS_MASK 0x00000200L +#define CP_CPF_BUSY_STAT__HPD_PROCESSING_EOP_BUSY_MASK 0x00000800L +#define CP_CPF_BUSY_STAT__HQD_DISPATCH_BUSY_MASK 0x00001000L +#define CP_CPF_BUSY_STAT__HQD_IQ_TIMER_BUSY_MASK 0x00002000L +#define CP_CPF_BUSY_STAT__HQD_DMA_OFFLOAD_BUSY_MASK 0x00004000L +#define CP_CPF_BUSY_STAT__HQD_WAIT_SEMAPHORE_BUSY_MASK 0x00008000L +#define CP_CPF_BUSY_STAT__HQD_SIGNAL_SEMAPHORE_BUSY_MASK 0x00010000L +#define CP_CPF_BUSY_STAT__HQD_MESSAGE_BUSY_MASK 0x00020000L +#define CP_CPF_BUSY_STAT__HQD_PQ_FETCHER_BUSY_MASK 0x00040000L +#define CP_CPF_BUSY_STAT__HQD_IB_FETCHER_BUSY_MASK 0x00080000L +#define CP_CPF_BUSY_STAT__HQD_IQ_FETCHER_BUSY_MASK 0x00100000L +#define CP_CPF_BUSY_STAT__HQD_EOP_FETCHER_BUSY_MASK 0x00200000L +#define CP_CPF_BUSY_STAT__HQD_CONSUMED_RPTR_BUSY_MASK 0x00400000L +#define CP_CPF_BUSY_STAT__HQD_FETCHER_ARB_BUSY_MASK 0x00800000L +#define CP_CPF_BUSY_STAT__HQD_ROQ_ALIGN_BUSY_MASK 0x01000000L +#define CP_CPF_BUSY_STAT__HQD_ROQ_EOP_BUSY_MASK 0x02000000L +#define CP_CPF_BUSY_STAT__HQD_ROQ_IQ_BUSY_MASK 0x04000000L +#define CP_CPF_BUSY_STAT__HQD_ROQ_PQ_BUSY_MASK 0x08000000L +#define CP_CPF_BUSY_STAT__HQD_ROQ_IB_BUSY_MASK 0x10000000L +#define CP_CPF_BUSY_STAT__HQD_WPTR_POLL_BUSY_MASK 0x20000000L +#define CP_CPF_BUSY_STAT__HQD_PQ_BUSY_MASK 0x40000000L +#define CP_CPF_BUSY_STAT__HQD_IB_BUSY_MASK 0x80000000L +//CP_CPF_STALLED_STAT1 +#define CP_CPF_STALLED_STAT1__RING_FETCHING_DATA__SHIFT 0x0 +#define CP_CPF_STALLED_STAT1__INDR1_FETCHING_DATA__SHIFT 0x1 +#define CP_CPF_STALLED_STAT1__INDR2_FETCHING_DATA__SHIFT 0x2 +#define CP_CPF_STALLED_STAT1__STATE_FETCHING_DATA__SHIFT 0x3 +#define CP_CPF_STALLED_STAT1__TCIU_WAITING_ON_FREE__SHIFT 0x5 +#define CP_CPF_STALLED_STAT1__TCIU_WAITING_ON_TAGS__SHIFT 0x6 +#define CP_CPF_STALLED_STAT1__UTCL2IU_WAITING_ON_FREE__SHIFT 0x7 +#define CP_CPF_STALLED_STAT1__UTCL2IU_WAITING_ON_TAGS__SHIFT 0x8 +#define CP_CPF_STALLED_STAT1__GFX_UTCL1_WAITING_ON_TRANS__SHIFT 0x9 +#define CP_CPF_STALLED_STAT1__CMP_UTCL1_WAITING_ON_TRANS__SHIFT 0xa +#define CP_CPF_STALLED_STAT1__RCIU_WAITING_ON_FREE__SHIFT 0xb +#define CP_CPF_STALLED_STAT1__RING_FETCHING_DATA_MASK 0x00000001L +#define CP_CPF_STALLED_STAT1__INDR1_FETCHING_DATA_MASK 0x00000002L +#define CP_CPF_STALLED_STAT1__INDR2_FETCHING_DATA_MASK 0x00000004L +#define CP_CPF_STALLED_STAT1__STATE_FETCHING_DATA_MASK 0x00000008L +#define CP_CPF_STALLED_STAT1__TCIU_WAITING_ON_FREE_MASK 0x00000020L +#define CP_CPF_STALLED_STAT1__TCIU_WAITING_ON_TAGS_MASK 0x00000040L +#define CP_CPF_STALLED_STAT1__UTCL2IU_WAITING_ON_FREE_MASK 0x00000080L +#define CP_CPF_STALLED_STAT1__UTCL2IU_WAITING_ON_TAGS_MASK 0x00000100L +#define CP_CPF_STALLED_STAT1__GFX_UTCL1_WAITING_ON_TRANS_MASK 0x00000200L +#define CP_CPF_STALLED_STAT1__CMP_UTCL1_WAITING_ON_TRANS_MASK 0x00000400L +#define CP_CPF_STALLED_STAT1__RCIU_WAITING_ON_FREE_MASK 0x00000800L +//CP_CPC_GRBM_FREE_COUNT +#define CP_CPC_GRBM_FREE_COUNT__FREE_COUNT__SHIFT 0x0 +#define CP_CPC_GRBM_FREE_COUNT__FREE_COUNT_MASK 0x0000003FL +//CP_MEC_CNTL +#define CP_MEC_CNTL__MEC_INVALIDATE_ICACHE__SHIFT 0x4 +#define CP_MEC_CNTL__MEC_ME1_PIPE0_RESET__SHIFT 0x10 +#define CP_MEC_CNTL__MEC_ME1_PIPE1_RESET__SHIFT 0x11 +#define CP_MEC_CNTL__MEC_ME1_PIPE2_RESET__SHIFT 0x12 +#define CP_MEC_CNTL__MEC_ME1_PIPE3_RESET__SHIFT 0x13 +#define CP_MEC_CNTL__MEC_ME2_PIPE0_RESET__SHIFT 0x14 +#define CP_MEC_CNTL__MEC_ME2_PIPE1_RESET__SHIFT 0x15 +#define CP_MEC_CNTL__MEC_ME2_HALT__SHIFT 0x1c +#define CP_MEC_CNTL__MEC_ME2_STEP__SHIFT 0x1d +#define CP_MEC_CNTL__MEC_ME1_HALT__SHIFT 0x1e +#define CP_MEC_CNTL__MEC_ME1_STEP__SHIFT 0x1f +#define CP_MEC_CNTL__MEC_INVALIDATE_ICACHE_MASK 0x00000010L +#define CP_MEC_CNTL__MEC_ME1_PIPE0_RESET_MASK 0x00010000L +#define CP_MEC_CNTL__MEC_ME1_PIPE1_RESET_MASK 0x00020000L +#define CP_MEC_CNTL__MEC_ME1_PIPE2_RESET_MASK 0x00040000L +#define CP_MEC_CNTL__MEC_ME1_PIPE3_RESET_MASK 0x00080000L +#define CP_MEC_CNTL__MEC_ME2_PIPE0_RESET_MASK 0x00100000L +#define CP_MEC_CNTL__MEC_ME2_PIPE1_RESET_MASK 0x00200000L +#define CP_MEC_CNTL__MEC_ME2_HALT_MASK 0x10000000L +#define CP_MEC_CNTL__MEC_ME2_STEP_MASK 0x20000000L +#define CP_MEC_CNTL__MEC_ME1_HALT_MASK 0x40000000L +#define CP_MEC_CNTL__MEC_ME1_STEP_MASK 0x80000000L +//CP_MEC_ME1_HEADER_DUMP +#define CP_MEC_ME1_HEADER_DUMP__HEADER_DUMP__SHIFT 0x0 +#define CP_MEC_ME1_HEADER_DUMP__HEADER_DUMP_MASK 0xFFFFFFFFL +//CP_MEC_ME2_HEADER_DUMP +#define CP_MEC_ME2_HEADER_DUMP__HEADER_DUMP__SHIFT 0x0 +#define CP_MEC_ME2_HEADER_DUMP__HEADER_DUMP_MASK 0xFFFFFFFFL +//CP_CPC_SCRATCH_INDEX +#define CP_CPC_SCRATCH_INDEX__SCRATCH_INDEX__SHIFT 0x0 +#define CP_CPC_SCRATCH_INDEX__SCRATCH_INDEX_MASK 0x000001FFL +//CP_CPC_SCRATCH_DATA +#define CP_CPC_SCRATCH_DATA__SCRATCH_DATA__SHIFT 0x0 +#define CP_CPC_SCRATCH_DATA__SCRATCH_DATA_MASK 0xFFFFFFFFL +//CP_CPF_GRBM_FREE_COUNT +#define CP_CPF_GRBM_FREE_COUNT__FREE_COUNT__SHIFT 0x0 +#define CP_CPF_GRBM_FREE_COUNT__FREE_COUNT_MASK 0x00000007L +//CP_CPC_HALT_HYST_COUNT +#define CP_CPC_HALT_HYST_COUNT__COUNT__SHIFT 0x0 +#define CP_CPC_HALT_HYST_COUNT__COUNT_MASK 0x0000000FL +//CP_PRT_LOD_STATS_CNTL0 +#define CP_PRT_LOD_STATS_CNTL0__BU_SIZE__SHIFT 0x0 +#define CP_PRT_LOD_STATS_CNTL0__BU_SIZE_MASK 0xFFFFFFFFL +//CP_PRT_LOD_STATS_CNTL1 +#define CP_PRT_LOD_STATS_CNTL1__BASE_LO__SHIFT 0x0 +#define CP_PRT_LOD_STATS_CNTL1__BASE_LO_MASK 0xFFFFFFFFL +//CP_PRT_LOD_STATS_CNTL2 +#define CP_PRT_LOD_STATS_CNTL2__BASE_HI__SHIFT 0x0 +#define CP_PRT_LOD_STATS_CNTL2__BASE_HI_MASK 0x000003FFL +//CP_PRT_LOD_STATS_CNTL3 +#define CP_PRT_LOD_STATS_CNTL3__INTERVAL__SHIFT 0x2 +#define CP_PRT_LOD_STATS_CNTL3__RESET_CNT__SHIFT 0xa +#define CP_PRT_LOD_STATS_CNTL3__RESET_FORCE__SHIFT 0x12 +#define CP_PRT_LOD_STATS_CNTL3__REPORT_AND_RESET__SHIFT 0x13 +#define CP_PRT_LOD_STATS_CNTL3__MC_VMID__SHIFT 0x17 +#define CP_PRT_LOD_STATS_CNTL3__CACHE_POLICY__SHIFT 0x1c +#define CP_PRT_LOD_STATS_CNTL3__INTERVAL_MASK 0x000003FCL +#define CP_PRT_LOD_STATS_CNTL3__RESET_CNT_MASK 0x0003FC00L +#define CP_PRT_LOD_STATS_CNTL3__RESET_FORCE_MASK 0x00040000L +#define CP_PRT_LOD_STATS_CNTL3__REPORT_AND_RESET_MASK 0x00080000L +#define CP_PRT_LOD_STATS_CNTL3__MC_VMID_MASK 0x07800000L +#define CP_PRT_LOD_STATS_CNTL3__CACHE_POLICY_MASK 0x10000000L +//CP_CE_COMPARE_COUNT +#define CP_CE_COMPARE_COUNT__COMPARE_COUNT__SHIFT 0x0 +#define CP_CE_COMPARE_COUNT__COMPARE_COUNT_MASK 0xFFFFFFFFL +//CP_CE_DE_COUNT +#define CP_CE_DE_COUNT__DRAW_ENGINE_COUNT__SHIFT 0x0 +#define CP_CE_DE_COUNT__DRAW_ENGINE_COUNT_MASK 0xFFFFFFFFL +//CP_DE_CE_COUNT +#define CP_DE_CE_COUNT__CONST_ENGINE_COUNT__SHIFT 0x0 +#define CP_DE_CE_COUNT__CONST_ENGINE_COUNT_MASK 0xFFFFFFFFL +//CP_DE_LAST_INVAL_COUNT +#define CP_DE_LAST_INVAL_COUNT__LAST_INVAL_COUNT__SHIFT 0x0 +#define CP_DE_LAST_INVAL_COUNT__LAST_INVAL_COUNT_MASK 0xFFFFFFFFL +//CP_DE_DE_COUNT +#define CP_DE_DE_COUNT__DRAW_ENGINE_COUNT__SHIFT 0x0 +#define CP_DE_DE_COUNT__DRAW_ENGINE_COUNT_MASK 0xFFFFFFFFL +//CP_STALLED_STAT3 +#define CP_STALLED_STAT3__CE_TO_CSF_NOT_RDY_TO_RCV__SHIFT 0x0 +#define CP_STALLED_STAT3__CE_TO_RAM_INIT_FETCHER_NOT_RDY_TO_RCV__SHIFT 0x1 +#define CP_STALLED_STAT3__CE_WAITING_ON_DATA_FROM_RAM_INIT_FETCHER__SHIFT 0x2 +#define CP_STALLED_STAT3__CE_TO_RAM_INIT_NOT_RDY__SHIFT 0x3 +#define CP_STALLED_STAT3__CE_TO_RAM_DUMP_NOT_RDY__SHIFT 0x4 +#define CP_STALLED_STAT3__CE_TO_RAM_WRITE_NOT_RDY__SHIFT 0x5 +#define CP_STALLED_STAT3__CE_TO_INC_FIFO_NOT_RDY_TO_RCV__SHIFT 0x6 +#define CP_STALLED_STAT3__CE_TO_WR_FIFO_NOT_RDY_TO_RCV__SHIFT 0x7 +#define CP_STALLED_STAT3__CE_WAITING_ON_BUFFER_DATA__SHIFT 0xa +#define CP_STALLED_STAT3__CE_WAITING_ON_CE_BUFFER_FLAG__SHIFT 0xb +#define CP_STALLED_STAT3__CE_WAITING_ON_DE_COUNTER__SHIFT 0xc +#define CP_STALLED_STAT3__CE_WAITING_ON_DE_COUNTER_UNDERFLOW__SHIFT 0xd +#define CP_STALLED_STAT3__TCIU_WAITING_ON_FREE__SHIFT 0xe +#define CP_STALLED_STAT3__TCIU_WAITING_ON_TAGS__SHIFT 0xf +#define CP_STALLED_STAT3__CE_STALLED_ON_TC_WR_CONFIRM__SHIFT 0x10 +#define CP_STALLED_STAT3__CE_STALLED_ON_ATOMIC_RTN_DATA__SHIFT 0x11 +#define CP_STALLED_STAT3__UTCL2IU_WAITING_ON_FREE__SHIFT 0x12 +#define CP_STALLED_STAT3__UTCL2IU_WAITING_ON_TAGS__SHIFT 0x13 +#define CP_STALLED_STAT3__UTCL1_WAITING_ON_TRANS__SHIFT 0x14 +#define CP_STALLED_STAT3__CE_TO_CSF_NOT_RDY_TO_RCV_MASK 0x00000001L +#define CP_STALLED_STAT3__CE_TO_RAM_INIT_FETCHER_NOT_RDY_TO_RCV_MASK 0x00000002L +#define CP_STALLED_STAT3__CE_WAITING_ON_DATA_FROM_RAM_INIT_FETCHER_MASK 0x00000004L +#define CP_STALLED_STAT3__CE_TO_RAM_INIT_NOT_RDY_MASK 0x00000008L +#define CP_STALLED_STAT3__CE_TO_RAM_DUMP_NOT_RDY_MASK 0x00000010L +#define CP_STALLED_STAT3__CE_TO_RAM_WRITE_NOT_RDY_MASK 0x00000020L +#define CP_STALLED_STAT3__CE_TO_INC_FIFO_NOT_RDY_TO_RCV_MASK 0x00000040L +#define CP_STALLED_STAT3__CE_TO_WR_FIFO_NOT_RDY_TO_RCV_MASK 0x00000080L +#define CP_STALLED_STAT3__CE_WAITING_ON_BUFFER_DATA_MASK 0x00000400L +#define CP_STALLED_STAT3__CE_WAITING_ON_CE_BUFFER_FLAG_MASK 0x00000800L +#define CP_STALLED_STAT3__CE_WAITING_ON_DE_COUNTER_MASK 0x00001000L +#define CP_STALLED_STAT3__CE_WAITING_ON_DE_COUNTER_UNDERFLOW_MASK 0x00002000L +#define CP_STALLED_STAT3__TCIU_WAITING_ON_FREE_MASK 0x00004000L +#define CP_STALLED_STAT3__TCIU_WAITING_ON_TAGS_MASK 0x00008000L +#define CP_STALLED_STAT3__CE_STALLED_ON_TC_WR_CONFIRM_MASK 0x00010000L +#define CP_STALLED_STAT3__CE_STALLED_ON_ATOMIC_RTN_DATA_MASK 0x00020000L +#define CP_STALLED_STAT3__UTCL2IU_WAITING_ON_FREE_MASK 0x00040000L +#define CP_STALLED_STAT3__UTCL2IU_WAITING_ON_TAGS_MASK 0x00080000L +#define CP_STALLED_STAT3__UTCL1_WAITING_ON_TRANS_MASK 0x00100000L +//CP_STALLED_STAT1 +#define CP_STALLED_STAT1__RBIU_TO_DMA_NOT_RDY_TO_RCV__SHIFT 0x0 +#define CP_STALLED_STAT1__RBIU_TO_SEM_NOT_RDY_TO_RCV__SHIFT 0x2 +#define CP_STALLED_STAT1__RBIU_TO_MEMWR_NOT_RDY_TO_RCV__SHIFT 0x4 +#define CP_STALLED_STAT1__ME_HAS_ACTIVE_CE_BUFFER_FLAG__SHIFT 0xa +#define CP_STALLED_STAT1__ME_HAS_ACTIVE_DE_BUFFER_FLAG__SHIFT 0xb +#define CP_STALLED_STAT1__ME_STALLED_ON_TC_WR_CONFIRM__SHIFT 0xc +#define CP_STALLED_STAT1__ME_STALLED_ON_ATOMIC_RTN_DATA__SHIFT 0xd +#define CP_STALLED_STAT1__ME_WAITING_ON_TC_READ_DATA__SHIFT 0xe +#define CP_STALLED_STAT1__ME_WAITING_ON_REG_READ_DATA__SHIFT 0xf +#define CP_STALLED_STAT1__RCIU_WAITING_ON_GDS_FREE__SHIFT 0x17 +#define CP_STALLED_STAT1__RCIU_WAITING_ON_GRBM_FREE__SHIFT 0x18 +#define CP_STALLED_STAT1__RCIU_WAITING_ON_VGT_FREE__SHIFT 0x19 +#define CP_STALLED_STAT1__RCIU_STALLED_ON_ME_READ__SHIFT 0x1a +#define CP_STALLED_STAT1__RCIU_STALLED_ON_DMA_READ__SHIFT 0x1b +#define CP_STALLED_STAT1__RCIU_STALLED_ON_APPEND_READ__SHIFT 0x1c +#define CP_STALLED_STAT1__RCIU_HALTED_BY_REG_VIOLATION__SHIFT 0x1d +#define CP_STALLED_STAT1__RBIU_TO_DMA_NOT_RDY_TO_RCV_MASK 0x00000001L +#define CP_STALLED_STAT1__RBIU_TO_SEM_NOT_RDY_TO_RCV_MASK 0x00000004L +#define CP_STALLED_STAT1__RBIU_TO_MEMWR_NOT_RDY_TO_RCV_MASK 0x00000010L +#define CP_STALLED_STAT1__ME_HAS_ACTIVE_CE_BUFFER_FLAG_MASK 0x00000400L +#define CP_STALLED_STAT1__ME_HAS_ACTIVE_DE_BUFFER_FLAG_MASK 0x00000800L +#define CP_STALLED_STAT1__ME_STALLED_ON_TC_WR_CONFIRM_MASK 0x00001000L +#define CP_STALLED_STAT1__ME_STALLED_ON_ATOMIC_RTN_DATA_MASK 0x00002000L +#define CP_STALLED_STAT1__ME_WAITING_ON_TC_READ_DATA_MASK 0x00004000L +#define CP_STALLED_STAT1__ME_WAITING_ON_REG_READ_DATA_MASK 0x00008000L +#define CP_STALLED_STAT1__RCIU_WAITING_ON_GDS_FREE_MASK 0x00800000L +#define CP_STALLED_STAT1__RCIU_WAITING_ON_GRBM_FREE_MASK 0x01000000L +#define CP_STALLED_STAT1__RCIU_WAITING_ON_VGT_FREE_MASK 0x02000000L +#define CP_STALLED_STAT1__RCIU_STALLED_ON_ME_READ_MASK 0x04000000L +#define CP_STALLED_STAT1__RCIU_STALLED_ON_DMA_READ_MASK 0x08000000L +#define CP_STALLED_STAT1__RCIU_STALLED_ON_APPEND_READ_MASK 0x10000000L +#define CP_STALLED_STAT1__RCIU_HALTED_BY_REG_VIOLATION_MASK 0x20000000L +//CP_STALLED_STAT2 +#define CP_STALLED_STAT2__PFP_TO_CSF_NOT_RDY_TO_RCV__SHIFT 0x0 +#define CP_STALLED_STAT2__PFP_TO_MEQ_NOT_RDY_TO_RCV__SHIFT 0x1 +#define CP_STALLED_STAT2__PFP_TO_RCIU_NOT_RDY_TO_RCV__SHIFT 0x2 +#define CP_STALLED_STAT2__PFP_TO_VGT_WRITES_PENDING__SHIFT 0x4 +#define CP_STALLED_STAT2__PFP_RCIU_READ_PENDING__SHIFT 0x5 +#define CP_STALLED_STAT2__PFP_WAITING_ON_BUFFER_DATA__SHIFT 0x8 +#define CP_STALLED_STAT2__ME_WAIT_ON_CE_COUNTER__SHIFT 0x9 +#define CP_STALLED_STAT2__ME_WAIT_ON_AVAIL_BUFFER__SHIFT 0xa +#define CP_STALLED_STAT2__GFX_CNTX_NOT_AVAIL_TO_ME__SHIFT 0xb +#define CP_STALLED_STAT2__ME_RCIU_NOT_RDY_TO_RCV__SHIFT 0xc +#define CP_STALLED_STAT2__ME_TO_CONST_NOT_RDY_TO_RCV__SHIFT 0xd +#define CP_STALLED_STAT2__ME_WAITING_DATA_FROM_PFP__SHIFT 0xe +#define CP_STALLED_STAT2__ME_WAITING_ON_PARTIAL_FLUSH__SHIFT 0xf +#define CP_STALLED_STAT2__MEQ_TO_ME_NOT_RDY_TO_RCV__SHIFT 0x10 +#define CP_STALLED_STAT2__STQ_TO_ME_NOT_RDY_TO_RCV__SHIFT 0x11 +#define CP_STALLED_STAT2__ME_WAITING_DATA_FROM_STQ__SHIFT 0x12 +#define CP_STALLED_STAT2__PFP_STALLED_ON_TC_WR_CONFIRM__SHIFT 0x13 +#define CP_STALLED_STAT2__PFP_STALLED_ON_ATOMIC_RTN_DATA__SHIFT 0x14 +#define CP_STALLED_STAT2__EOPD_FIFO_NEEDS_SC_EOP_DONE__SHIFT 0x15 +#define CP_STALLED_STAT2__EOPD_FIFO_NEEDS_WR_CONFIRM__SHIFT 0x16 +#define CP_STALLED_STAT2__STRMO_WR_OF_PRIM_DATA_PENDING__SHIFT 0x17 +#define CP_STALLED_STAT2__PIPE_STATS_WR_DATA_PENDING__SHIFT 0x18 +#define CP_STALLED_STAT2__APPEND_RDY_WAIT_ON_CS_DONE__SHIFT 0x19 +#define CP_STALLED_STAT2__APPEND_RDY_WAIT_ON_PS_DONE__SHIFT 0x1a +#define CP_STALLED_STAT2__APPEND_WAIT_ON_WR_CONFIRM__SHIFT 0x1b +#define CP_STALLED_STAT2__APPEND_ACTIVE_PARTITION__SHIFT 0x1c +#define CP_STALLED_STAT2__APPEND_WAITING_TO_SEND_MEMWRITE__SHIFT 0x1d +#define CP_STALLED_STAT2__SURF_SYNC_NEEDS_IDLE_CNTXS__SHIFT 0x1e +#define CP_STALLED_STAT2__SURF_SYNC_NEEDS_ALL_CLEAN__SHIFT 0x1f +#define CP_STALLED_STAT2__PFP_TO_CSF_NOT_RDY_TO_RCV_MASK 0x00000001L +#define CP_STALLED_STAT2__PFP_TO_MEQ_NOT_RDY_TO_RCV_MASK 0x00000002L +#define CP_STALLED_STAT2__PFP_TO_RCIU_NOT_RDY_TO_RCV_MASK 0x00000004L +#define CP_STALLED_STAT2__PFP_TO_VGT_WRITES_PENDING_MASK 0x00000010L +#define CP_STALLED_STAT2__PFP_RCIU_READ_PENDING_MASK 0x00000020L +#define CP_STALLED_STAT2__PFP_WAITING_ON_BUFFER_DATA_MASK 0x00000100L +#define CP_STALLED_STAT2__ME_WAIT_ON_CE_COUNTER_MASK 0x00000200L +#define CP_STALLED_STAT2__ME_WAIT_ON_AVAIL_BUFFER_MASK 0x00000400L +#define CP_STALLED_STAT2__GFX_CNTX_NOT_AVAIL_TO_ME_MASK 0x00000800L +#define CP_STALLED_STAT2__ME_RCIU_NOT_RDY_TO_RCV_MASK 0x00001000L +#define CP_STALLED_STAT2__ME_TO_CONST_NOT_RDY_TO_RCV_MASK 0x00002000L +#define CP_STALLED_STAT2__ME_WAITING_DATA_FROM_PFP_MASK 0x00004000L +#define CP_STALLED_STAT2__ME_WAITING_ON_PARTIAL_FLUSH_MASK 0x00008000L +#define CP_STALLED_STAT2__MEQ_TO_ME_NOT_RDY_TO_RCV_MASK 0x00010000L +#define CP_STALLED_STAT2__STQ_TO_ME_NOT_RDY_TO_RCV_MASK 0x00020000L +#define CP_STALLED_STAT2__ME_WAITING_DATA_FROM_STQ_MASK 0x00040000L +#define CP_STALLED_STAT2__PFP_STALLED_ON_TC_WR_CONFIRM_MASK 0x00080000L +#define CP_STALLED_STAT2__PFP_STALLED_ON_ATOMIC_RTN_DATA_MASK 0x00100000L +#define CP_STALLED_STAT2__EOPD_FIFO_NEEDS_SC_EOP_DONE_MASK 0x00200000L +#define CP_STALLED_STAT2__EOPD_FIFO_NEEDS_WR_CONFIRM_MASK 0x00400000L +#define CP_STALLED_STAT2__STRMO_WR_OF_PRIM_DATA_PENDING_MASK 0x00800000L +#define CP_STALLED_STAT2__PIPE_STATS_WR_DATA_PENDING_MASK 0x01000000L +#define CP_STALLED_STAT2__APPEND_RDY_WAIT_ON_CS_DONE_MASK 0x02000000L +#define CP_STALLED_STAT2__APPEND_RDY_WAIT_ON_PS_DONE_MASK 0x04000000L +#define CP_STALLED_STAT2__APPEND_WAIT_ON_WR_CONFIRM_MASK 0x08000000L +#define CP_STALLED_STAT2__APPEND_ACTIVE_PARTITION_MASK 0x10000000L +#define CP_STALLED_STAT2__APPEND_WAITING_TO_SEND_MEMWRITE_MASK 0x20000000L +#define CP_STALLED_STAT2__SURF_SYNC_NEEDS_IDLE_CNTXS_MASK 0x40000000L +#define CP_STALLED_STAT2__SURF_SYNC_NEEDS_ALL_CLEAN_MASK 0x80000000L +//CP_BUSY_STAT +#define CP_BUSY_STAT__REG_BUS_FIFO_BUSY__SHIFT 0x0 +#define CP_BUSY_STAT__COHER_CNT_NEQ_ZERO__SHIFT 0x6 +#define CP_BUSY_STAT__PFP_PARSING_PACKETS__SHIFT 0x7 +#define CP_BUSY_STAT__ME_PARSING_PACKETS__SHIFT 0x8 +#define CP_BUSY_STAT__RCIU_PFP_BUSY__SHIFT 0x9 +#define CP_BUSY_STAT__RCIU_ME_BUSY__SHIFT 0xa +#define CP_BUSY_STAT__SEM_CMDFIFO_NOT_EMPTY__SHIFT 0xc +#define CP_BUSY_STAT__SEM_FAILED_AND_HOLDING__SHIFT 0xd +#define CP_BUSY_STAT__SEM_POLLING_FOR_PASS__SHIFT 0xe +#define CP_BUSY_STAT__GFX_CONTEXT_BUSY__SHIFT 0xf +#define CP_BUSY_STAT__ME_PARSER_BUSY__SHIFT 0x11 +#define CP_BUSY_STAT__EOP_DONE_BUSY__SHIFT 0x12 +#define CP_BUSY_STAT__STRM_OUT_BUSY__SHIFT 0x13 +#define CP_BUSY_STAT__PIPE_STATS_BUSY__SHIFT 0x14 +#define CP_BUSY_STAT__RCIU_CE_BUSY__SHIFT 0x15 +#define CP_BUSY_STAT__CE_PARSING_PACKETS__SHIFT 0x16 +#define CP_BUSY_STAT__REG_BUS_FIFO_BUSY_MASK 0x00000001L +#define CP_BUSY_STAT__COHER_CNT_NEQ_ZERO_MASK 0x00000040L +#define CP_BUSY_STAT__PFP_PARSING_PACKETS_MASK 0x00000080L +#define CP_BUSY_STAT__ME_PARSING_PACKETS_MASK 0x00000100L +#define CP_BUSY_STAT__RCIU_PFP_BUSY_MASK 0x00000200L +#define CP_BUSY_STAT__RCIU_ME_BUSY_MASK 0x00000400L +#define CP_BUSY_STAT__SEM_CMDFIFO_NOT_EMPTY_MASK 0x00001000L +#define CP_BUSY_STAT__SEM_FAILED_AND_HOLDING_MASK 0x00002000L +#define CP_BUSY_STAT__SEM_POLLING_FOR_PASS_MASK 0x00004000L +#define CP_BUSY_STAT__GFX_CONTEXT_BUSY_MASK 0x00008000L +#define CP_BUSY_STAT__ME_PARSER_BUSY_MASK 0x00020000L +#define CP_BUSY_STAT__EOP_DONE_BUSY_MASK 0x00040000L +#define CP_BUSY_STAT__STRM_OUT_BUSY_MASK 0x00080000L +#define CP_BUSY_STAT__PIPE_STATS_BUSY_MASK 0x00100000L +#define CP_BUSY_STAT__RCIU_CE_BUSY_MASK 0x00200000L +#define CP_BUSY_STAT__CE_PARSING_PACKETS_MASK 0x00400000L +//CP_STAT +#define CP_STAT__ROQ_RING_BUSY__SHIFT 0x9 +#define CP_STAT__ROQ_INDIRECT1_BUSY__SHIFT 0xa +#define CP_STAT__ROQ_INDIRECT2_BUSY__SHIFT 0xb +#define CP_STAT__ROQ_STATE_BUSY__SHIFT 0xc +#define CP_STAT__DC_BUSY__SHIFT 0xd +#define CP_STAT__UTCL2IU_BUSY__SHIFT 0xe +#define CP_STAT__PFP_BUSY__SHIFT 0xf +#define CP_STAT__MEQ_BUSY__SHIFT 0x10 +#define CP_STAT__ME_BUSY__SHIFT 0x11 +#define CP_STAT__QUERY_BUSY__SHIFT 0x12 +#define CP_STAT__SEMAPHORE_BUSY__SHIFT 0x13 +#define CP_STAT__INTERRUPT_BUSY__SHIFT 0x14 +#define CP_STAT__SURFACE_SYNC_BUSY__SHIFT 0x15 +#define CP_STAT__DMA_BUSY__SHIFT 0x16 +#define CP_STAT__RCIU_BUSY__SHIFT 0x17 +#define CP_STAT__SCRATCH_RAM_BUSY__SHIFT 0x18 +#define CP_STAT__CE_BUSY__SHIFT 0x1a +#define CP_STAT__TCIU_BUSY__SHIFT 0x1b +#define CP_STAT__ROQ_CE_RING_BUSY__SHIFT 0x1c +#define CP_STAT__ROQ_CE_INDIRECT1_BUSY__SHIFT 0x1d +#define CP_STAT__ROQ_CE_INDIRECT2_BUSY__SHIFT 0x1e +#define CP_STAT__CP_BUSY__SHIFT 0x1f +#define CP_STAT__ROQ_RING_BUSY_MASK 0x00000200L +#define CP_STAT__ROQ_INDIRECT1_BUSY_MASK 0x00000400L +#define CP_STAT__ROQ_INDIRECT2_BUSY_MASK 0x00000800L +#define CP_STAT__ROQ_STATE_BUSY_MASK 0x00001000L +#define CP_STAT__DC_BUSY_MASK 0x00002000L +#define CP_STAT__UTCL2IU_BUSY_MASK 0x00004000L +#define CP_STAT__PFP_BUSY_MASK 0x00008000L +#define CP_STAT__MEQ_BUSY_MASK 0x00010000L +#define CP_STAT__ME_BUSY_MASK 0x00020000L +#define CP_STAT__QUERY_BUSY_MASK 0x00040000L +#define CP_STAT__SEMAPHORE_BUSY_MASK 0x00080000L +#define CP_STAT__INTERRUPT_BUSY_MASK 0x00100000L +#define CP_STAT__SURFACE_SYNC_BUSY_MASK 0x00200000L +#define CP_STAT__DMA_BUSY_MASK 0x00400000L +#define CP_STAT__RCIU_BUSY_MASK 0x00800000L +#define CP_STAT__SCRATCH_RAM_BUSY_MASK 0x01000000L +#define CP_STAT__CE_BUSY_MASK 0x04000000L +#define CP_STAT__TCIU_BUSY_MASK 0x08000000L +#define CP_STAT__ROQ_CE_RING_BUSY_MASK 0x10000000L +#define CP_STAT__ROQ_CE_INDIRECT1_BUSY_MASK 0x20000000L +#define CP_STAT__ROQ_CE_INDIRECT2_BUSY_MASK 0x40000000L +#define CP_STAT__CP_BUSY_MASK 0x80000000L +//CP_ME_HEADER_DUMP +#define CP_ME_HEADER_DUMP__ME_HEADER_DUMP__SHIFT 0x0 +#define CP_ME_HEADER_DUMP__ME_HEADER_DUMP_MASK 0xFFFFFFFFL +//CP_PFP_HEADER_DUMP +#define CP_PFP_HEADER_DUMP__PFP_HEADER_DUMP__SHIFT 0x0 +#define CP_PFP_HEADER_DUMP__PFP_HEADER_DUMP_MASK 0xFFFFFFFFL +//CP_GRBM_FREE_COUNT +#define CP_GRBM_FREE_COUNT__FREE_COUNT__SHIFT 0x0 +#define CP_GRBM_FREE_COUNT__FREE_COUNT_GDS__SHIFT 0x8 +#define CP_GRBM_FREE_COUNT__FREE_COUNT_PFP__SHIFT 0x10 +#define CP_GRBM_FREE_COUNT__FREE_COUNT_MASK 0x0000003FL +#define CP_GRBM_FREE_COUNT__FREE_COUNT_GDS_MASK 0x00003F00L +#define CP_GRBM_FREE_COUNT__FREE_COUNT_PFP_MASK 0x003F0000L +//CP_CE_HEADER_DUMP +#define CP_CE_HEADER_DUMP__CE_HEADER_DUMP__SHIFT 0x0 +#define CP_CE_HEADER_DUMP__CE_HEADER_DUMP_MASK 0xFFFFFFFFL +//CP_PFP_INSTR_PNTR +#define CP_PFP_INSTR_PNTR__INSTR_PNTR__SHIFT 0x0 +#define CP_PFP_INSTR_PNTR__INSTR_PNTR_MASK 0x0000FFFFL +//CP_ME_INSTR_PNTR +#define CP_ME_INSTR_PNTR__INSTR_PNTR__SHIFT 0x0 +#define CP_ME_INSTR_PNTR__INSTR_PNTR_MASK 0x0000FFFFL +//CP_CE_INSTR_PNTR +#define CP_CE_INSTR_PNTR__INSTR_PNTR__SHIFT 0x0 +#define CP_CE_INSTR_PNTR__INSTR_PNTR_MASK 0x0000FFFFL +//CP_MEC1_INSTR_PNTR +#define CP_MEC1_INSTR_PNTR__INSTR_PNTR__SHIFT 0x0 +#define CP_MEC1_INSTR_PNTR__INSTR_PNTR_MASK 0x0000FFFFL +//CP_MEC2_INSTR_PNTR +#define CP_MEC2_INSTR_PNTR__INSTR_PNTR__SHIFT 0x0 +#define CP_MEC2_INSTR_PNTR__INSTR_PNTR_MASK 0x0000FFFFL +//CP_CSF_STAT +#define CP_CSF_STAT__BUFFER_REQUEST_COUNT__SHIFT 0x8 +#define CP_CSF_STAT__BUFFER_REQUEST_COUNT_MASK 0x0001FF00L +//CP_ME_CNTL +#define CP_ME_CNTL__CE_INVALIDATE_ICACHE__SHIFT 0x4 +#define CP_ME_CNTL__PFP_INVALIDATE_ICACHE__SHIFT 0x6 +#define CP_ME_CNTL__ME_INVALIDATE_ICACHE__SHIFT 0x8 +#define CP_ME_CNTL__CE_PIPE0_RESET__SHIFT 0x10 +#define CP_ME_CNTL__CE_PIPE1_RESET__SHIFT 0x11 +#define CP_ME_CNTL__PFP_PIPE0_RESET__SHIFT 0x12 +#define CP_ME_CNTL__PFP_PIPE1_RESET__SHIFT 0x13 +#define CP_ME_CNTL__ME_PIPE0_RESET__SHIFT 0x14 +#define CP_ME_CNTL__ME_PIPE1_RESET__SHIFT 0x15 +#define CP_ME_CNTL__CE_HALT__SHIFT 0x18 +#define CP_ME_CNTL__CE_STEP__SHIFT 0x19 +#define CP_ME_CNTL__PFP_HALT__SHIFT 0x1a +#define CP_ME_CNTL__PFP_STEP__SHIFT 0x1b +#define CP_ME_CNTL__ME_HALT__SHIFT 0x1c +#define CP_ME_CNTL__ME_STEP__SHIFT 0x1d +#define CP_ME_CNTL__CE_INVALIDATE_ICACHE_MASK 0x00000010L +#define CP_ME_CNTL__PFP_INVALIDATE_ICACHE_MASK 0x00000040L +#define CP_ME_CNTL__ME_INVALIDATE_ICACHE_MASK 0x00000100L +#define CP_ME_CNTL__CE_PIPE0_RESET_MASK 0x00010000L +#define CP_ME_CNTL__CE_PIPE1_RESET_MASK 0x00020000L +#define CP_ME_CNTL__PFP_PIPE0_RESET_MASK 0x00040000L +#define CP_ME_CNTL__PFP_PIPE1_RESET_MASK 0x00080000L +#define CP_ME_CNTL__ME_PIPE0_RESET_MASK 0x00100000L +#define CP_ME_CNTL__ME_PIPE1_RESET_MASK 0x00200000L +#define CP_ME_CNTL__CE_HALT_MASK 0x01000000L +#define CP_ME_CNTL__CE_STEP_MASK 0x02000000L +#define CP_ME_CNTL__PFP_HALT_MASK 0x04000000L +#define CP_ME_CNTL__PFP_STEP_MASK 0x08000000L +#define CP_ME_CNTL__ME_HALT_MASK 0x10000000L +#define CP_ME_CNTL__ME_STEP_MASK 0x20000000L +//CP_CNTX_STAT +#define CP_CNTX_STAT__ACTIVE_HP3D_CONTEXTS__SHIFT 0x0 +#define CP_CNTX_STAT__CURRENT_HP3D_CONTEXT__SHIFT 0x8 +#define CP_CNTX_STAT__ACTIVE_GFX_CONTEXTS__SHIFT 0x14 +#define CP_CNTX_STAT__CURRENT_GFX_CONTEXT__SHIFT 0x1c +#define CP_CNTX_STAT__ACTIVE_HP3D_CONTEXTS_MASK 0x000000FFL +#define CP_CNTX_STAT__CURRENT_HP3D_CONTEXT_MASK 0x00000700L +#define CP_CNTX_STAT__ACTIVE_GFX_CONTEXTS_MASK 0x0FF00000L +#define CP_CNTX_STAT__CURRENT_GFX_CONTEXT_MASK 0x70000000L +//CP_ME_PREEMPTION +#define CP_ME_PREEMPTION__OBSOLETE__SHIFT 0x0 +#define CP_ME_PREEMPTION__OBSOLETE_MASK 0x00000001L +//CP_ROQ_THRESHOLDS +#define CP_ROQ_THRESHOLDS__IB1_START__SHIFT 0x0 +#define CP_ROQ_THRESHOLDS__IB2_START__SHIFT 0x8 +#define CP_ROQ_THRESHOLDS__IB1_START_MASK 0x000000FFL +#define CP_ROQ_THRESHOLDS__IB2_START_MASK 0x0000FF00L +//CP_MEQ_STQ_THRESHOLD +#define CP_MEQ_STQ_THRESHOLD__STQ_START__SHIFT 0x0 +#define CP_MEQ_STQ_THRESHOLD__STQ_START_MASK 0x000000FFL +//CP_RB2_RPTR +#define CP_RB2_RPTR__RB_RPTR__SHIFT 0x0 +#define CP_RB2_RPTR__RB_RPTR_MASK 0x000FFFFFL +//CP_RB1_RPTR +#define CP_RB1_RPTR__RB_RPTR__SHIFT 0x0 +#define CP_RB1_RPTR__RB_RPTR_MASK 0x000FFFFFL +//CP_RB0_RPTR +#define CP_RB0_RPTR__RB_RPTR__SHIFT 0x0 +#define CP_RB0_RPTR__RB_RPTR_MASK 0x000FFFFFL +//CP_RB_RPTR +#define CP_RB_RPTR__RB_RPTR__SHIFT 0x0 +#define CP_RB_RPTR__RB_RPTR_MASK 0x000FFFFFL +//CP_RB_WPTR_DELAY +#define CP_RB_WPTR_DELAY__PRE_WRITE_TIMER__SHIFT 0x0 +#define CP_RB_WPTR_DELAY__PRE_WRITE_LIMIT__SHIFT 0x1c +#define CP_RB_WPTR_DELAY__PRE_WRITE_TIMER_MASK 0x0FFFFFFFL +#define CP_RB_WPTR_DELAY__PRE_WRITE_LIMIT_MASK 0xF0000000L +//CP_RB_WPTR_POLL_CNTL +#define CP_RB_WPTR_POLL_CNTL__POLL_FREQUENCY__SHIFT 0x0 +#define CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT__SHIFT 0x10 +#define CP_RB_WPTR_POLL_CNTL__POLL_FREQUENCY_MASK 0x0000FFFFL +#define CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT_MASK 0xFFFF0000L +//CP_ROQ1_THRESHOLDS +#define CP_ROQ1_THRESHOLDS__RB1_START__SHIFT 0x0 +#define CP_ROQ1_THRESHOLDS__RB2_START__SHIFT 0x8 +#define CP_ROQ1_THRESHOLDS__R0_IB1_START__SHIFT 0x10 +#define CP_ROQ1_THRESHOLDS__R1_IB1_START__SHIFT 0x18 +#define CP_ROQ1_THRESHOLDS__RB1_START_MASK 0x000000FFL +#define CP_ROQ1_THRESHOLDS__RB2_START_MASK 0x0000FF00L +#define CP_ROQ1_THRESHOLDS__R0_IB1_START_MASK 0x00FF0000L +#define CP_ROQ1_THRESHOLDS__R1_IB1_START_MASK 0xFF000000L +//CP_ROQ2_THRESHOLDS +#define CP_ROQ2_THRESHOLDS__R2_IB1_START__SHIFT 0x0 +#define CP_ROQ2_THRESHOLDS__R0_IB2_START__SHIFT 0x8 +#define CP_ROQ2_THRESHOLDS__R1_IB2_START__SHIFT 0x10 +#define CP_ROQ2_THRESHOLDS__R2_IB2_START__SHIFT 0x18 +#define CP_ROQ2_THRESHOLDS__R2_IB1_START_MASK 0x000000FFL +#define CP_ROQ2_THRESHOLDS__R0_IB2_START_MASK 0x0000FF00L +#define CP_ROQ2_THRESHOLDS__R1_IB2_START_MASK 0x00FF0000L +#define CP_ROQ2_THRESHOLDS__R2_IB2_START_MASK 0xFF000000L +//CP_STQ_THRESHOLDS +#define CP_STQ_THRESHOLDS__STQ0_START__SHIFT 0x0 +#define CP_STQ_THRESHOLDS__STQ1_START__SHIFT 0x8 +#define CP_STQ_THRESHOLDS__STQ2_START__SHIFT 0x10 +#define CP_STQ_THRESHOLDS__STQ0_START_MASK 0x000000FFL +#define CP_STQ_THRESHOLDS__STQ1_START_MASK 0x0000FF00L +#define CP_STQ_THRESHOLDS__STQ2_START_MASK 0x00FF0000L +//CP_QUEUE_THRESHOLDS +#define CP_QUEUE_THRESHOLDS__ROQ_IB1_START__SHIFT 0x0 +#define CP_QUEUE_THRESHOLDS__ROQ_IB2_START__SHIFT 0x8 +#define CP_QUEUE_THRESHOLDS__ROQ_IB1_START_MASK 0x0000003FL +#define CP_QUEUE_THRESHOLDS__ROQ_IB2_START_MASK 0x00003F00L +//CP_MEQ_THRESHOLDS +#define CP_MEQ_THRESHOLDS__MEQ1_START__SHIFT 0x0 +#define CP_MEQ_THRESHOLDS__MEQ2_START__SHIFT 0x8 +#define CP_MEQ_THRESHOLDS__MEQ1_START_MASK 0x000000FFL +#define CP_MEQ_THRESHOLDS__MEQ2_START_MASK 0x0000FF00L +//CP_ROQ_AVAIL +#define CP_ROQ_AVAIL__ROQ_CNT_RING__SHIFT 0x0 +#define CP_ROQ_AVAIL__ROQ_CNT_IB1__SHIFT 0x10 +#define CP_ROQ_AVAIL__ROQ_CNT_RING_MASK 0x000007FFL +#define CP_ROQ_AVAIL__ROQ_CNT_IB1_MASK 0x07FF0000L +//CP_STQ_AVAIL +#define CP_STQ_AVAIL__STQ_CNT__SHIFT 0x0 +#define CP_STQ_AVAIL__STQ_CNT_MASK 0x000001FFL +//CP_ROQ2_AVAIL +#define CP_ROQ2_AVAIL__ROQ_CNT_IB2__SHIFT 0x0 +#define CP_ROQ2_AVAIL__ROQ_CNT_IB2_MASK 0x000007FFL +//CP_MEQ_AVAIL +#define CP_MEQ_AVAIL__MEQ_CNT__SHIFT 0x0 +#define CP_MEQ_AVAIL__MEQ_CNT_MASK 0x000003FFL +//CP_CMD_INDEX +#define CP_CMD_INDEX__CMD_INDEX__SHIFT 0x0 +#define CP_CMD_INDEX__CMD_ME_SEL__SHIFT 0xc +#define CP_CMD_INDEX__CMD_QUEUE_SEL__SHIFT 0x10 +#define CP_CMD_INDEX__CMD_INDEX_MASK 0x000007FFL +#define CP_CMD_INDEX__CMD_ME_SEL_MASK 0x00003000L +#define CP_CMD_INDEX__CMD_QUEUE_SEL_MASK 0x00070000L +//CP_CMD_DATA +#define CP_CMD_DATA__CMD_DATA__SHIFT 0x0 +#define CP_CMD_DATA__CMD_DATA_MASK 0xFFFFFFFFL +//CP_ROQ_RB_STAT +#define CP_ROQ_RB_STAT__ROQ_RPTR_PRIMARY__SHIFT 0x0 +#define CP_ROQ_RB_STAT__ROQ_WPTR_PRIMARY__SHIFT 0x10 +#define CP_ROQ_RB_STAT__ROQ_RPTR_PRIMARY_MASK 0x000003FFL +#define CP_ROQ_RB_STAT__ROQ_WPTR_PRIMARY_MASK 0x03FF0000L +//CP_ROQ_IB1_STAT +#define CP_ROQ_IB1_STAT__ROQ_RPTR_INDIRECT1__SHIFT 0x0 +#define CP_ROQ_IB1_STAT__ROQ_WPTR_INDIRECT1__SHIFT 0x10 +#define CP_ROQ_IB1_STAT__ROQ_RPTR_INDIRECT1_MASK 0x000003FFL +#define CP_ROQ_IB1_STAT__ROQ_WPTR_INDIRECT1_MASK 0x03FF0000L +//CP_ROQ_IB2_STAT +#define CP_ROQ_IB2_STAT__ROQ_RPTR_INDIRECT2__SHIFT 0x0 +#define CP_ROQ_IB2_STAT__ROQ_WPTR_INDIRECT2__SHIFT 0x10 +#define CP_ROQ_IB2_STAT__ROQ_RPTR_INDIRECT2_MASK 0x000003FFL +#define CP_ROQ_IB2_STAT__ROQ_WPTR_INDIRECT2_MASK 0x03FF0000L +//CP_STQ_STAT +#define CP_STQ_STAT__STQ_RPTR__SHIFT 0x0 +#define CP_STQ_STAT__STQ_RPTR_MASK 0x000003FFL +//CP_STQ_WR_STAT +#define CP_STQ_WR_STAT__STQ_WPTR__SHIFT 0x0 +#define CP_STQ_WR_STAT__STQ_WPTR_MASK 0x000003FFL +//CP_MEQ_STAT +#define CP_MEQ_STAT__MEQ_RPTR__SHIFT 0x0 +#define CP_MEQ_STAT__MEQ_WPTR__SHIFT 0x10 +#define CP_MEQ_STAT__MEQ_RPTR_MASK 0x000003FFL +#define CP_MEQ_STAT__MEQ_WPTR_MASK 0x03FF0000L +//CP_CEQ1_AVAIL +#define CP_CEQ1_AVAIL__CEQ_CNT_RING__SHIFT 0x0 +#define CP_CEQ1_AVAIL__CEQ_CNT_IB1__SHIFT 0x10 +#define CP_CEQ1_AVAIL__CEQ_CNT_RING_MASK 0x000007FFL +#define CP_CEQ1_AVAIL__CEQ_CNT_IB1_MASK 0x07FF0000L +//CP_CEQ2_AVAIL +#define CP_CEQ2_AVAIL__CEQ_CNT_IB2__SHIFT 0x0 +#define CP_CEQ2_AVAIL__CEQ_CNT_IB2_MASK 0x000007FFL +//CP_CE_ROQ_RB_STAT +#define CP_CE_ROQ_RB_STAT__CEQ_RPTR_PRIMARY__SHIFT 0x0 +#define CP_CE_ROQ_RB_STAT__CEQ_WPTR_PRIMARY__SHIFT 0x10 +#define CP_CE_ROQ_RB_STAT__CEQ_RPTR_PRIMARY_MASK 0x000003FFL +#define CP_CE_ROQ_RB_STAT__CEQ_WPTR_PRIMARY_MASK 0x03FF0000L +//CP_CE_ROQ_IB1_STAT +#define CP_CE_ROQ_IB1_STAT__CEQ_RPTR_INDIRECT1__SHIFT 0x0 +#define CP_CE_ROQ_IB1_STAT__CEQ_WPTR_INDIRECT1__SHIFT 0x10 +#define CP_CE_ROQ_IB1_STAT__CEQ_RPTR_INDIRECT1_MASK 0x000003FFL +#define CP_CE_ROQ_IB1_STAT__CEQ_WPTR_INDIRECT1_MASK 0x03FF0000L +//CP_CE_ROQ_IB2_STAT +#define CP_CE_ROQ_IB2_STAT__CEQ_RPTR_INDIRECT2__SHIFT 0x0 +#define CP_CE_ROQ_IB2_STAT__CEQ_WPTR_INDIRECT2__SHIFT 0x10 +#define CP_CE_ROQ_IB2_STAT__CEQ_RPTR_INDIRECT2_MASK 0x000003FFL +#define CP_CE_ROQ_IB2_STAT__CEQ_WPTR_INDIRECT2_MASK 0x03FF0000L +#define CP_INT_STAT_DEBUG__PRIV_INSTR_INT_ASSERTED__SHIFT 0x16 +#define CP_INT_STAT_DEBUG__PRIV_REG_INT_ASSERTED__SHIFT 0x17 +#define CP_INT_STAT_DEBUG__PRIV_INSTR_INT_ASSERTED_MASK 0x00400000L +#define CP_INT_STAT_DEBUG__PRIV_REG_INT_ASSERTED_MASK 0x00800000L + + +// addressBlock: gc_padec +//VGT_VTX_VECT_EJECT_REG +#define VGT_VTX_VECT_EJECT_REG__PRIM_COUNT__SHIFT 0x0 +#define VGT_VTX_VECT_EJECT_REG__PRIM_COUNT_MASK 0x0000007FL +//VGT_DMA_DATA_FIFO_DEPTH +#define VGT_DMA_DATA_FIFO_DEPTH__DMA_DATA_FIFO_DEPTH__SHIFT 0x0 +#define VGT_DMA_DATA_FIFO_DEPTH__DMA2DRAW_FIFO_DEPTH__SHIFT 0x9 +#define VGT_DMA_DATA_FIFO_DEPTH__DMA_DATA_FIFO_DEPTH_MASK 0x000001FFL +#define VGT_DMA_DATA_FIFO_DEPTH__DMA2DRAW_FIFO_DEPTH_MASK 0x0007FE00L +//VGT_DMA_REQ_FIFO_DEPTH +#define VGT_DMA_REQ_FIFO_DEPTH__DMA_REQ_FIFO_DEPTH__SHIFT 0x0 +#define VGT_DMA_REQ_FIFO_DEPTH__DMA_REQ_FIFO_DEPTH_MASK 0x0000003FL +//VGT_DRAW_INIT_FIFO_DEPTH +#define VGT_DRAW_INIT_FIFO_DEPTH__DRAW_INIT_FIFO_DEPTH__SHIFT 0x0 +#define VGT_DRAW_INIT_FIFO_DEPTH__DRAW_INIT_FIFO_DEPTH_MASK 0x0000003FL +//VGT_LAST_COPY_STATE +#define VGT_LAST_COPY_STATE__SRC_STATE_ID__SHIFT 0x0 +#define VGT_LAST_COPY_STATE__DST_STATE_ID__SHIFT 0x10 +#define VGT_LAST_COPY_STATE__SRC_STATE_ID_MASK 0x00000007L +#define VGT_LAST_COPY_STATE__DST_STATE_ID_MASK 0x00070000L +//VGT_CACHE_INVALIDATION +#define VGT_CACHE_INVALIDATION__CACHE_INVALIDATION__SHIFT 0x0 +#define VGT_CACHE_INVALIDATION__DIS_INSTANCING_OPT__SHIFT 0x4 +#define VGT_CACHE_INVALIDATION__VS_NO_EXTRA_BUFFER__SHIFT 0x5 +#define VGT_CACHE_INVALIDATION__AUTO_INVLD_EN__SHIFT 0x6 +#define VGT_CACHE_INVALIDATION__USE_GS_DONE__SHIFT 0x9 +#define VGT_CACHE_INVALIDATION__DIS_RANGE_FULL_INVLD__SHIFT 0xb +#define VGT_CACHE_INVALIDATION__GS_LATE_ALLOC_EN__SHIFT 0xc +#define VGT_CACHE_INVALIDATION__STREAMOUT_FULL_FLUSH__SHIFT 0xd +#define VGT_CACHE_INVALIDATION__ES_LIMIT__SHIFT 0x10 +#define VGT_CACHE_INVALIDATION__ENABLE_PING_PONG__SHIFT 0x15 +#define VGT_CACHE_INVALIDATION__OPT_FLOW_CNTL_1__SHIFT 0x16 +#define VGT_CACHE_INVALIDATION__OPT_FLOW_CNTL_2__SHIFT 0x19 +#define VGT_CACHE_INVALIDATION__EN_WAVE_MERGE__SHIFT 0x1c +#define VGT_CACHE_INVALIDATION__ENABLE_PING_PONG_EOI__SHIFT 0x1d +#define VGT_CACHE_INVALIDATION__CACHE_INVALIDATION_MASK 0x00000003L +#define VGT_CACHE_INVALIDATION__DIS_INSTANCING_OPT_MASK 0x00000010L +#define VGT_CACHE_INVALIDATION__VS_NO_EXTRA_BUFFER_MASK 0x00000020L +#define VGT_CACHE_INVALIDATION__AUTO_INVLD_EN_MASK 0x000000C0L +#define VGT_CACHE_INVALIDATION__USE_GS_DONE_MASK 0x00000200L +#define VGT_CACHE_INVALIDATION__DIS_RANGE_FULL_INVLD_MASK 0x00000800L +#define VGT_CACHE_INVALIDATION__GS_LATE_ALLOC_EN_MASK 0x00001000L +#define VGT_CACHE_INVALIDATION__STREAMOUT_FULL_FLUSH_MASK 0x00002000L +#define VGT_CACHE_INVALIDATION__ES_LIMIT_MASK 0x001F0000L +#define VGT_CACHE_INVALIDATION__ENABLE_PING_PONG_MASK 0x00200000L +#define VGT_CACHE_INVALIDATION__OPT_FLOW_CNTL_1_MASK 0x01C00000L +#define VGT_CACHE_INVALIDATION__OPT_FLOW_CNTL_2_MASK 0x0E000000L +#define VGT_CACHE_INVALIDATION__EN_WAVE_MERGE_MASK 0x10000000L +#define VGT_CACHE_INVALIDATION__ENABLE_PING_PONG_EOI_MASK 0x20000000L +//VGT_STRMOUT_DELAY +#define VGT_STRMOUT_DELAY__SKIP_DELAY__SHIFT 0x0 +#define VGT_STRMOUT_DELAY__SE0_WD_DELAY__SHIFT 0x8 +#define VGT_STRMOUT_DELAY__SE1_WD_DELAY__SHIFT 0xb +#define VGT_STRMOUT_DELAY__SE2_WD_DELAY__SHIFT 0xe +#define VGT_STRMOUT_DELAY__SE3_WD_DELAY__SHIFT 0x11 +#define VGT_STRMOUT_DELAY__SKIP_DELAY_MASK 0x000000FFL +#define VGT_STRMOUT_DELAY__SE0_WD_DELAY_MASK 0x00000700L +#define VGT_STRMOUT_DELAY__SE1_WD_DELAY_MASK 0x00003800L +#define VGT_STRMOUT_DELAY__SE2_WD_DELAY_MASK 0x0001C000L +#define VGT_STRMOUT_DELAY__SE3_WD_DELAY_MASK 0x000E0000L +//VGT_FIFO_DEPTHS +#define VGT_FIFO_DEPTHS__VS_DEALLOC_TBL_DEPTH__SHIFT 0x0 +#define VGT_FIFO_DEPTHS__RESERVED_0__SHIFT 0x7 +#define VGT_FIFO_DEPTHS__CLIPP_FIFO_DEPTH__SHIFT 0x8 +#define VGT_FIFO_DEPTHS__HSINPUT_FIFO_DEPTH__SHIFT 0x16 +#define VGT_FIFO_DEPTHS__VS_DEALLOC_TBL_DEPTH_MASK 0x0000007FL +#define VGT_FIFO_DEPTHS__RESERVED_0_MASK 0x00000080L +#define VGT_FIFO_DEPTHS__CLIPP_FIFO_DEPTH_MASK 0x003FFF00L +#define VGT_FIFO_DEPTHS__HSINPUT_FIFO_DEPTH_MASK 0x0FC00000L +//VGT_GS_VERTEX_REUSE +#define VGT_GS_VERTEX_REUSE__VERT_REUSE__SHIFT 0x0 +#define VGT_GS_VERTEX_REUSE__VERT_REUSE_MASK 0x0000001FL +//VGT_MC_LAT_CNTL +#define VGT_MC_LAT_CNTL__MC_TIME_STAMP_RES__SHIFT 0x0 +#define VGT_MC_LAT_CNTL__MC_TIME_STAMP_RES_MASK 0x0000000FL +//IA_CNTL_STATUS +#define IA_CNTL_STATUS__IA_BUSY__SHIFT 0x0 +#define IA_CNTL_STATUS__IA_DMA_BUSY__SHIFT 0x1 +#define IA_CNTL_STATUS__IA_DMA_REQ_BUSY__SHIFT 0x2 +#define IA_CNTL_STATUS__IA_GRP_BUSY__SHIFT 0x3 +#define IA_CNTL_STATUS__IA_ADC_BUSY__SHIFT 0x4 +#define IA_CNTL_STATUS__IA_BUSY_MASK 0x00000001L +#define IA_CNTL_STATUS__IA_DMA_BUSY_MASK 0x00000002L +#define IA_CNTL_STATUS__IA_DMA_REQ_BUSY_MASK 0x00000004L +#define IA_CNTL_STATUS__IA_GRP_BUSY_MASK 0x00000008L +#define IA_CNTL_STATUS__IA_ADC_BUSY_MASK 0x00000010L +//VGT_CNTL_STATUS +#define VGT_CNTL_STATUS__VGT_BUSY__SHIFT 0x0 +#define VGT_CNTL_STATUS__VGT_OUT_INDX_BUSY__SHIFT 0x1 +#define VGT_CNTL_STATUS__VGT_OUT_BUSY__SHIFT 0x2 +#define VGT_CNTL_STATUS__VGT_PT_BUSY__SHIFT 0x3 +#define VGT_CNTL_STATUS__VGT_TE_BUSY__SHIFT 0x4 +#define VGT_CNTL_STATUS__VGT_VR_BUSY__SHIFT 0x5 +#define VGT_CNTL_STATUS__VGT_PI_BUSY__SHIFT 0x6 +#define VGT_CNTL_STATUS__VGT_GS_BUSY__SHIFT 0x7 +#define VGT_CNTL_STATUS__VGT_HS_BUSY__SHIFT 0x8 +#define VGT_CNTL_STATUS__VGT_TE11_BUSY__SHIFT 0x9 +#define VGT_CNTL_STATUS__VGT_PRIMGEN_BUSY__SHIFT 0xa +#define VGT_CNTL_STATUS__VGT_BUSY_MASK 0x00000001L +#define VGT_CNTL_STATUS__VGT_OUT_INDX_BUSY_MASK 0x00000002L +#define VGT_CNTL_STATUS__VGT_OUT_BUSY_MASK 0x00000004L +#define VGT_CNTL_STATUS__VGT_PT_BUSY_MASK 0x00000008L +#define VGT_CNTL_STATUS__VGT_TE_BUSY_MASK 0x00000010L +#define VGT_CNTL_STATUS__VGT_VR_BUSY_MASK 0x00000020L +#define VGT_CNTL_STATUS__VGT_PI_BUSY_MASK 0x00000040L +#define VGT_CNTL_STATUS__VGT_GS_BUSY_MASK 0x00000080L +#define VGT_CNTL_STATUS__VGT_HS_BUSY_MASK 0x00000100L +#define VGT_CNTL_STATUS__VGT_TE11_BUSY_MASK 0x00000200L +#define VGT_CNTL_STATUS__VGT_PRIMGEN_BUSY_MASK 0x00000400L +//WD_CNTL_STATUS +#define WD_CNTL_STATUS__WD_BUSY__SHIFT 0x0 +#define WD_CNTL_STATUS__WD_SPL_DMA_BUSY__SHIFT 0x1 +#define WD_CNTL_STATUS__WD_SPL_DI_BUSY__SHIFT 0x2 +#define WD_CNTL_STATUS__WD_ADC_BUSY__SHIFT 0x3 +#define WD_CNTL_STATUS__WD_BUSY_MASK 0x00000001L +#define WD_CNTL_STATUS__WD_SPL_DMA_BUSY_MASK 0x00000002L +#define WD_CNTL_STATUS__WD_SPL_DI_BUSY_MASK 0x00000004L +#define WD_CNTL_STATUS__WD_ADC_BUSY_MASK 0x00000008L +//CC_GC_PRIM_CONFIG +#define CC_GC_PRIM_CONFIG__INACTIVE_IA__SHIFT 0x10 +#define CC_GC_PRIM_CONFIG__INACTIVE_VGT_PA__SHIFT 0x18 +#define CC_GC_PRIM_CONFIG__INACTIVE_IA_MASK 0x00030000L +#define CC_GC_PRIM_CONFIG__INACTIVE_VGT_PA_MASK 0x0F000000L +//GC_USER_PRIM_CONFIG +#define GC_USER_PRIM_CONFIG__INACTIVE_IA__SHIFT 0x10 +#define GC_USER_PRIM_CONFIG__INACTIVE_VGT_PA__SHIFT 0x18 +#define GC_USER_PRIM_CONFIG__INACTIVE_IA_MASK 0x00030000L +#define GC_USER_PRIM_CONFIG__INACTIVE_VGT_PA_MASK 0x0F000000L +//WD_QOS +#define WD_QOS__DRAW_STALL__SHIFT 0x0 +#define WD_QOS__DRAW_STALL_MASK 0x00000001L +//WD_UTCL1_CNTL +#define WD_UTCL1_CNTL__XNACK_REDO_TIMER_CNT__SHIFT 0x0 +#define WD_UTCL1_CNTL__VMID_RESET_MODE__SHIFT 0x17 +#define WD_UTCL1_CNTL__DROP_MODE__SHIFT 0x18 +#define WD_UTCL1_CNTL__BYPASS__SHIFT 0x19 +#define WD_UTCL1_CNTL__INVALIDATE__SHIFT 0x1a +#define WD_UTCL1_CNTL__FRAG_LIMIT_MODE__SHIFT 0x1b +#define WD_UTCL1_CNTL__FORCE_SNOOP__SHIFT 0x1c +#define WD_UTCL1_CNTL__FORCE_SD_VMID_DIRTY__SHIFT 0x1d +#define WD_UTCL1_CNTL__XNACK_REDO_TIMER_CNT_MASK 0x000FFFFFL +#define WD_UTCL1_CNTL__VMID_RESET_MODE_MASK 0x00800000L +#define WD_UTCL1_CNTL__DROP_MODE_MASK 0x01000000L +#define WD_UTCL1_CNTL__BYPASS_MASK 0x02000000L +#define WD_UTCL1_CNTL__INVALIDATE_MASK 0x04000000L +#define WD_UTCL1_CNTL__FRAG_LIMIT_MODE_MASK 0x08000000L +#define WD_UTCL1_CNTL__FORCE_SNOOP_MASK 0x10000000L +#define WD_UTCL1_CNTL__FORCE_SD_VMID_DIRTY_MASK 0x20000000L +//WD_UTCL1_STATUS +#define WD_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define WD_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define WD_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define WD_UTCL1_STATUS__FAULT_UTCL1ID__SHIFT 0x8 +#define WD_UTCL1_STATUS__RETRY_UTCL1ID__SHIFT 0x10 +#define WD_UTCL1_STATUS__PRT_UTCL1ID__SHIFT 0x18 +#define WD_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define WD_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define WD_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L +#define WD_UTCL1_STATUS__FAULT_UTCL1ID_MASK 0x00003F00L +#define WD_UTCL1_STATUS__RETRY_UTCL1ID_MASK 0x003F0000L +#define WD_UTCL1_STATUS__PRT_UTCL1ID_MASK 0x3F000000L +//IA_UTCL1_CNTL +#define IA_UTCL1_CNTL__XNACK_REDO_TIMER_CNT__SHIFT 0x0 +#define IA_UTCL1_CNTL__VMID_RESET_MODE__SHIFT 0x17 +#define IA_UTCL1_CNTL__DROP_MODE__SHIFT 0x18 +#define IA_UTCL1_CNTL__BYPASS__SHIFT 0x19 +#define IA_UTCL1_CNTL__INVALIDATE__SHIFT 0x1a +#define IA_UTCL1_CNTL__FRAG_LIMIT_MODE__SHIFT 0x1b +#define IA_UTCL1_CNTL__FORCE_SNOOP__SHIFT 0x1c +#define IA_UTCL1_CNTL__FORCE_SD_VMID_DIRTY__SHIFT 0x1d +#define IA_UTCL1_CNTL__XNACK_REDO_TIMER_CNT_MASK 0x000FFFFFL +#define IA_UTCL1_CNTL__VMID_RESET_MODE_MASK 0x00800000L +#define IA_UTCL1_CNTL__DROP_MODE_MASK 0x01000000L +#define IA_UTCL1_CNTL__BYPASS_MASK 0x02000000L +#define IA_UTCL1_CNTL__INVALIDATE_MASK 0x04000000L +#define IA_UTCL1_CNTL__FRAG_LIMIT_MODE_MASK 0x08000000L +#define IA_UTCL1_CNTL__FORCE_SNOOP_MASK 0x10000000L +#define IA_UTCL1_CNTL__FORCE_SD_VMID_DIRTY_MASK 0x20000000L +//IA_UTCL1_STATUS +#define IA_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define IA_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define IA_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define IA_UTCL1_STATUS__FAULT_UTCL1ID__SHIFT 0x8 +#define IA_UTCL1_STATUS__RETRY_UTCL1ID__SHIFT 0x10 +#define IA_UTCL1_STATUS__PRT_UTCL1ID__SHIFT 0x18 +#define IA_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define IA_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define IA_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L +#define IA_UTCL1_STATUS__FAULT_UTCL1ID_MASK 0x00003F00L +#define IA_UTCL1_STATUS__RETRY_UTCL1ID_MASK 0x003F0000L +#define IA_UTCL1_STATUS__PRT_UTCL1ID_MASK 0x3F000000L +//VGT_SYS_CONFIG +#define VGT_SYS_CONFIG__DUAL_CORE_EN__SHIFT 0x0 +#define VGT_SYS_CONFIG__MAX_LS_HS_THDGRP__SHIFT 0x1 +#define VGT_SYS_CONFIG__ADC_EVENT_FILTER_DISABLE__SHIFT 0x7 +#define VGT_SYS_CONFIG__DUAL_CORE_EN_MASK 0x00000001L +#define VGT_SYS_CONFIG__MAX_LS_HS_THDGRP_MASK 0x0000007EL +#define VGT_SYS_CONFIG__ADC_EVENT_FILTER_DISABLE_MASK 0x00000080L +//VGT_VS_MAX_WAVE_ID +#define VGT_VS_MAX_WAVE_ID__MAX_WAVE_ID__SHIFT 0x0 +#define VGT_VS_MAX_WAVE_ID__MAX_WAVE_ID_MASK 0x00000FFFL +//VGT_GS_MAX_WAVE_ID +#define VGT_GS_MAX_WAVE_ID__MAX_WAVE_ID__SHIFT 0x0 +#define VGT_GS_MAX_WAVE_ID__MAX_WAVE_ID_MASK 0x00000FFFL +//GFX_PIPE_CONTROL +#define GFX_PIPE_CONTROL__HYSTERESIS_CNT__SHIFT 0x0 +#define GFX_PIPE_CONTROL__RESERVED__SHIFT 0xd +#define GFX_PIPE_CONTROL__CONTEXT_SUSPEND_EN__SHIFT 0x10 +#define GFX_PIPE_CONTROL__HYSTERESIS_CNT_MASK 0x00001FFFL +#define GFX_PIPE_CONTROL__RESERVED_MASK 0x0000E000L +#define GFX_PIPE_CONTROL__CONTEXT_SUSPEND_EN_MASK 0x00010000L +//CC_GC_SHADER_ARRAY_CONFIG +#define CC_GC_SHADER_ARRAY_CONFIG__INACTIVE_CUS__SHIFT 0x10 +#define CC_GC_SHADER_ARRAY_CONFIG__INACTIVE_CUS_MASK 0xFFFF0000L +//GC_USER_SHADER_ARRAY_CONFIG +#define GC_USER_SHADER_ARRAY_CONFIG__INACTIVE_CUS__SHIFT 0x10 +#define GC_USER_SHADER_ARRAY_CONFIG__INACTIVE_CUS_MASK 0xFFFF0000L +//VGT_DMA_PRIMITIVE_TYPE +#define VGT_DMA_PRIMITIVE_TYPE__PRIM_TYPE__SHIFT 0x0 +#define VGT_DMA_PRIMITIVE_TYPE__PRIM_TYPE_MASK 0x0000003FL +//VGT_DMA_CONTROL +#define VGT_DMA_CONTROL__PRIMGROUP_SIZE__SHIFT 0x0 +#define VGT_DMA_CONTROL__IA_SWITCH_ON_EOP__SHIFT 0x11 +#define VGT_DMA_CONTROL__SWITCH_ON_EOI__SHIFT 0x13 +#define VGT_DMA_CONTROL__WD_SWITCH_ON_EOP__SHIFT 0x14 +#define VGT_DMA_CONTROL__EN_INST_OPT_BASIC__SHIFT 0x15 +#define VGT_DMA_CONTROL__EN_INST_OPT_ADV__SHIFT 0x16 +#define VGT_DMA_CONTROL__HW_USE_ONLY__SHIFT 0x17 +#define VGT_DMA_CONTROL__PRIMGROUP_SIZE_MASK 0x0000FFFFL +#define VGT_DMA_CONTROL__IA_SWITCH_ON_EOP_MASK 0x00020000L +#define VGT_DMA_CONTROL__SWITCH_ON_EOI_MASK 0x00080000L +#define VGT_DMA_CONTROL__WD_SWITCH_ON_EOP_MASK 0x00100000L +#define VGT_DMA_CONTROL__EN_INST_OPT_BASIC_MASK 0x00200000L +#define VGT_DMA_CONTROL__EN_INST_OPT_ADV_MASK 0x00400000L +#define VGT_DMA_CONTROL__HW_USE_ONLY_MASK 0x00800000L +//VGT_DMA_LS_HS_CONFIG +#define VGT_DMA_LS_HS_CONFIG__HS_NUM_INPUT_CP__SHIFT 0x8 +#define VGT_DMA_LS_HS_CONFIG__HS_NUM_INPUT_CP_MASK 0x00003F00L +//WD_BUF_RESOURCE_1 +#define WD_BUF_RESOURCE_1__POS_BUF_SIZE__SHIFT 0x0 +#define WD_BUF_RESOURCE_1__INDEX_BUF_SIZE__SHIFT 0x10 +#define WD_BUF_RESOURCE_1__POS_BUF_SIZE_MASK 0x0000FFFFL +#define WD_BUF_RESOURCE_1__INDEX_BUF_SIZE_MASK 0xFFFF0000L +//WD_BUF_RESOURCE_2 +#define WD_BUF_RESOURCE_2__PARAM_BUF_SIZE__SHIFT 0x0 +#define WD_BUF_RESOURCE_2__ADDR_MODE__SHIFT 0xf +#define WD_BUF_RESOURCE_2__CNTL_SB_BUF_SIZE__SHIFT 0x10 +#define WD_BUF_RESOURCE_2__PARAM_BUF_SIZE_MASK 0x00001FFFL +#define WD_BUF_RESOURCE_2__ADDR_MODE_MASK 0x00008000L +#define WD_BUF_RESOURCE_2__CNTL_SB_BUF_SIZE_MASK 0xFFFF0000L +//PA_CL_CNTL_STATUS +#define PA_CL_CNTL_STATUS__UTC_FAULT_DETECTED__SHIFT 0x0 +#define PA_CL_CNTL_STATUS__UTC_RETRY_DETECTED__SHIFT 0x1 +#define PA_CL_CNTL_STATUS__UTC_PRT_DETECTED__SHIFT 0x2 +#define PA_CL_CNTL_STATUS__UTC_FAULT_DETECTED_MASK 0x00000001L +#define PA_CL_CNTL_STATUS__UTC_RETRY_DETECTED_MASK 0x00000002L +#define PA_CL_CNTL_STATUS__UTC_PRT_DETECTED_MASK 0x00000004L +//PA_CL_ENHANCE +#define PA_CL_ENHANCE__CLIP_VTX_REORDER_ENA__SHIFT 0x0 +#define PA_CL_ENHANCE__NUM_CLIP_SEQ__SHIFT 0x1 +#define PA_CL_ENHANCE__CLIPPED_PRIM_SEQ_STALL__SHIFT 0x3 +#define PA_CL_ENHANCE__VE_NAN_PROC_DISABLE__SHIFT 0x4 +#define PA_CL_ENHANCE__IGNORE_PIPELINE_RESET__SHIFT 0x6 +#define PA_CL_ENHANCE__KILL_INNER_EDGE_FLAGS__SHIFT 0x7 +#define PA_CL_ENHANCE__NGG_PA_TO_ALL_SC__SHIFT 0x8 +#define PA_CL_ENHANCE__TC_LATENCY_TIME_STAMP_RESOLUTION__SHIFT 0x9 +#define PA_CL_ENHANCE__NGG_BYPASS_PRIM_FILTER__SHIFT 0xb +#define PA_CL_ENHANCE__NGG_SIDEBAND_MEMORY_DEPTH__SHIFT 0xc +#define PA_CL_ENHANCE__NGG_PRIM_INDICES_FIFO_DEPTH__SHIFT 0xe +#define PA_CL_ENHANCE__ECO_SPARE3__SHIFT 0x1c +#define PA_CL_ENHANCE__ECO_SPARE2__SHIFT 0x1d +#define PA_CL_ENHANCE__ECO_SPARE1__SHIFT 0x1e +#define PA_CL_ENHANCE__ECO_SPARE0__SHIFT 0x1f +#define PA_CL_ENHANCE__CLIP_VTX_REORDER_ENA_MASK 0x00000001L +#define PA_CL_ENHANCE__NUM_CLIP_SEQ_MASK 0x00000006L +#define PA_CL_ENHANCE__CLIPPED_PRIM_SEQ_STALL_MASK 0x00000008L +#define PA_CL_ENHANCE__VE_NAN_PROC_DISABLE_MASK 0x00000010L +#define PA_CL_ENHANCE__IGNORE_PIPELINE_RESET_MASK 0x00000040L +#define PA_CL_ENHANCE__KILL_INNER_EDGE_FLAGS_MASK 0x00000080L +#define PA_CL_ENHANCE__NGG_PA_TO_ALL_SC_MASK 0x00000100L +#define PA_CL_ENHANCE__TC_LATENCY_TIME_STAMP_RESOLUTION_MASK 0x00000600L +#define PA_CL_ENHANCE__NGG_BYPASS_PRIM_FILTER_MASK 0x00000800L +#define PA_CL_ENHANCE__NGG_SIDEBAND_MEMORY_DEPTH_MASK 0x00003000L +#define PA_CL_ENHANCE__NGG_PRIM_INDICES_FIFO_DEPTH_MASK 0x0001C000L +#define PA_CL_ENHANCE__ECO_SPARE3_MASK 0x10000000L +#define PA_CL_ENHANCE__ECO_SPARE2_MASK 0x20000000L +#define PA_CL_ENHANCE__ECO_SPARE1_MASK 0x40000000L +#define PA_CL_ENHANCE__ECO_SPARE0_MASK 0x80000000L +//PA_SU_CNTL_STATUS +#define PA_SU_CNTL_STATUS__SU_BUSY__SHIFT 0x1f +#define PA_SU_CNTL_STATUS__SU_BUSY_MASK 0x80000000L +//PA_SC_FIFO_DEPTH_CNTL +#define PA_SC_FIFO_DEPTH_CNTL__DEPTH__SHIFT 0x0 +#define PA_SC_FIFO_DEPTH_CNTL__DEPTH_MASK 0x000003FFL +//PA_SC_P3D_TRAP_SCREEN_HV_LOCK +#define PA_SC_P3D_TRAP_SCREEN_HV_LOCK__DISABLE_NON_PRIV_WRITES__SHIFT 0x0 +#define PA_SC_P3D_TRAP_SCREEN_HV_LOCK__DISABLE_NON_PRIV_WRITES_MASK 0x00000001L +//PA_SC_HP3D_TRAP_SCREEN_HV_LOCK +#define PA_SC_HP3D_TRAP_SCREEN_HV_LOCK__DISABLE_NON_PRIV_WRITES__SHIFT 0x0 +#define PA_SC_HP3D_TRAP_SCREEN_HV_LOCK__DISABLE_NON_PRIV_WRITES_MASK 0x00000001L +//PA_SC_TRAP_SCREEN_HV_LOCK +#define PA_SC_TRAP_SCREEN_HV_LOCK__DISABLE_NON_PRIV_WRITES__SHIFT 0x0 +#define PA_SC_TRAP_SCREEN_HV_LOCK__DISABLE_NON_PRIV_WRITES_MASK 0x00000001L +//PA_SC_FORCE_EOV_MAX_CNTS +#define PA_SC_FORCE_EOV_MAX_CNTS__FORCE_EOV_MAX_CLK_CNT__SHIFT 0x0 +#define PA_SC_FORCE_EOV_MAX_CNTS__FORCE_EOV_MAX_REZ_CNT__SHIFT 0x10 +#define PA_SC_FORCE_EOV_MAX_CNTS__FORCE_EOV_MAX_CLK_CNT_MASK 0x0000FFFFL +#define PA_SC_FORCE_EOV_MAX_CNTS__FORCE_EOV_MAX_REZ_CNT_MASK 0xFFFF0000L +//PA_SC_BINNER_EVENT_CNTL_0 +#define PA_SC_BINNER_EVENT_CNTL_0__RESERVED_0__SHIFT 0x0 +#define PA_SC_BINNER_EVENT_CNTL_0__SAMPLE_STREAMOUTSTATS1__SHIFT 0x2 +#define PA_SC_BINNER_EVENT_CNTL_0__SAMPLE_STREAMOUTSTATS2__SHIFT 0x4 +#define PA_SC_BINNER_EVENT_CNTL_0__SAMPLE_STREAMOUTSTATS3__SHIFT 0x6 +#define PA_SC_BINNER_EVENT_CNTL_0__CACHE_FLUSH_TS__SHIFT 0x8 +#define PA_SC_BINNER_EVENT_CNTL_0__CONTEXT_DONE__SHIFT 0xa +#define PA_SC_BINNER_EVENT_CNTL_0__CACHE_FLUSH__SHIFT 0xc +#define PA_SC_BINNER_EVENT_CNTL_0__CS_PARTIAL_FLUSH__SHIFT 0xe +#define PA_SC_BINNER_EVENT_CNTL_0__VGT_STREAMOUT_SYNC__SHIFT 0x10 +#define PA_SC_BINNER_EVENT_CNTL_0__RESERVED_9__SHIFT 0x12 +#define PA_SC_BINNER_EVENT_CNTL_0__VGT_STREAMOUT_RESET__SHIFT 0x14 +#define PA_SC_BINNER_EVENT_CNTL_0__END_OF_PIPE_INCR_DE__SHIFT 0x16 +#define PA_SC_BINNER_EVENT_CNTL_0__END_OF_PIPE_IB_END__SHIFT 0x18 +#define PA_SC_BINNER_EVENT_CNTL_0__RST_PIX_CNT__SHIFT 0x1a +#define PA_SC_BINNER_EVENT_CNTL_0__BREAK_BATCH__SHIFT 0x1c +#define PA_SC_BINNER_EVENT_CNTL_0__VS_PARTIAL_FLUSH__SHIFT 0x1e +#define PA_SC_BINNER_EVENT_CNTL_0__RESERVED_0_MASK 0x00000003L +#define PA_SC_BINNER_EVENT_CNTL_0__SAMPLE_STREAMOUTSTATS1_MASK 0x0000000CL +#define PA_SC_BINNER_EVENT_CNTL_0__SAMPLE_STREAMOUTSTATS2_MASK 0x00000030L +#define PA_SC_BINNER_EVENT_CNTL_0__SAMPLE_STREAMOUTSTATS3_MASK 0x000000C0L +#define PA_SC_BINNER_EVENT_CNTL_0__CACHE_FLUSH_TS_MASK 0x00000300L +#define PA_SC_BINNER_EVENT_CNTL_0__CONTEXT_DONE_MASK 0x00000C00L +#define PA_SC_BINNER_EVENT_CNTL_0__CACHE_FLUSH_MASK 0x00003000L +#define PA_SC_BINNER_EVENT_CNTL_0__CS_PARTIAL_FLUSH_MASK 0x0000C000L +#define PA_SC_BINNER_EVENT_CNTL_0__VGT_STREAMOUT_SYNC_MASK 0x00030000L +#define PA_SC_BINNER_EVENT_CNTL_0__RESERVED_9_MASK 0x000C0000L +#define PA_SC_BINNER_EVENT_CNTL_0__VGT_STREAMOUT_RESET_MASK 0x00300000L +#define PA_SC_BINNER_EVENT_CNTL_0__END_OF_PIPE_INCR_DE_MASK 0x00C00000L +#define PA_SC_BINNER_EVENT_CNTL_0__END_OF_PIPE_IB_END_MASK 0x03000000L +#define PA_SC_BINNER_EVENT_CNTL_0__RST_PIX_CNT_MASK 0x0C000000L +#define PA_SC_BINNER_EVENT_CNTL_0__BREAK_BATCH_MASK 0x30000000L +#define PA_SC_BINNER_EVENT_CNTL_0__VS_PARTIAL_FLUSH_MASK 0xC0000000L +//PA_SC_BINNER_EVENT_CNTL_1 +#define PA_SC_BINNER_EVENT_CNTL_1__PS_PARTIAL_FLUSH__SHIFT 0x0 +#define PA_SC_BINNER_EVENT_CNTL_1__FLUSH_HS_OUTPUT__SHIFT 0x2 +#define PA_SC_BINNER_EVENT_CNTL_1__FLUSH_DFSM__SHIFT 0x4 +#define PA_SC_BINNER_EVENT_CNTL_1__RESET_TO_LOWEST_VGT__SHIFT 0x6 +#define PA_SC_BINNER_EVENT_CNTL_1__CACHE_FLUSH_AND_INV_TS_EVENT__SHIFT 0x8 +#define PA_SC_BINNER_EVENT_CNTL_1__ZPASS_DONE__SHIFT 0xa +#define PA_SC_BINNER_EVENT_CNTL_1__CACHE_FLUSH_AND_INV_EVENT__SHIFT 0xc +#define PA_SC_BINNER_EVENT_CNTL_1__PERFCOUNTER_START__SHIFT 0xe +#define PA_SC_BINNER_EVENT_CNTL_1__PERFCOUNTER_STOP__SHIFT 0x10 +#define PA_SC_BINNER_EVENT_CNTL_1__PIPELINESTAT_START__SHIFT 0x12 +#define PA_SC_BINNER_EVENT_CNTL_1__PIPELINESTAT_STOP__SHIFT 0x14 +#define PA_SC_BINNER_EVENT_CNTL_1__PERFCOUNTER_SAMPLE__SHIFT 0x16 +#define PA_SC_BINNER_EVENT_CNTL_1__FLUSH_ES_OUTPUT__SHIFT 0x18 +#define PA_SC_BINNER_EVENT_CNTL_1__FLUSH_GS_OUTPUT__SHIFT 0x1a +#define PA_SC_BINNER_EVENT_CNTL_1__SAMPLE_PIPELINESTAT__SHIFT 0x1c +#define PA_SC_BINNER_EVENT_CNTL_1__SO_VGTSTREAMOUT_FLUSH__SHIFT 0x1e +#define PA_SC_BINNER_EVENT_CNTL_1__PS_PARTIAL_FLUSH_MASK 0x00000003L +#define PA_SC_BINNER_EVENT_CNTL_1__FLUSH_HS_OUTPUT_MASK 0x0000000CL +#define PA_SC_BINNER_EVENT_CNTL_1__FLUSH_DFSM_MASK 0x00000030L +#define PA_SC_BINNER_EVENT_CNTL_1__RESET_TO_LOWEST_VGT_MASK 0x000000C0L +#define PA_SC_BINNER_EVENT_CNTL_1__CACHE_FLUSH_AND_INV_TS_EVENT_MASK 0x00000300L +#define PA_SC_BINNER_EVENT_CNTL_1__ZPASS_DONE_MASK 0x00000C00L +#define PA_SC_BINNER_EVENT_CNTL_1__CACHE_FLUSH_AND_INV_EVENT_MASK 0x00003000L +#define PA_SC_BINNER_EVENT_CNTL_1__PERFCOUNTER_START_MASK 0x0000C000L +#define PA_SC_BINNER_EVENT_CNTL_1__PERFCOUNTER_STOP_MASK 0x00030000L +#define PA_SC_BINNER_EVENT_CNTL_1__PIPELINESTAT_START_MASK 0x000C0000L +#define PA_SC_BINNER_EVENT_CNTL_1__PIPELINESTAT_STOP_MASK 0x00300000L +#define PA_SC_BINNER_EVENT_CNTL_1__PERFCOUNTER_SAMPLE_MASK 0x00C00000L +#define PA_SC_BINNER_EVENT_CNTL_1__FLUSH_ES_OUTPUT_MASK 0x03000000L +#define PA_SC_BINNER_EVENT_CNTL_1__FLUSH_GS_OUTPUT_MASK 0x0C000000L +#define PA_SC_BINNER_EVENT_CNTL_1__SAMPLE_PIPELINESTAT_MASK 0x30000000L +#define PA_SC_BINNER_EVENT_CNTL_1__SO_VGTSTREAMOUT_FLUSH_MASK 0xC0000000L +//PA_SC_BINNER_EVENT_CNTL_2 +#define PA_SC_BINNER_EVENT_CNTL_2__SAMPLE_STREAMOUTSTATS__SHIFT 0x0 +#define PA_SC_BINNER_EVENT_CNTL_2__RESET_VTX_CNT__SHIFT 0x2 +#define PA_SC_BINNER_EVENT_CNTL_2__BLOCK_CONTEXT_DONE__SHIFT 0x4 +#define PA_SC_BINNER_EVENT_CNTL_2__CS_CONTEXT_DONE__SHIFT 0x6 +#define PA_SC_BINNER_EVENT_CNTL_2__VGT_FLUSH__SHIFT 0x8 +#define PA_SC_BINNER_EVENT_CNTL_2__TGID_ROLLOVER__SHIFT 0xa +#define PA_SC_BINNER_EVENT_CNTL_2__SQ_NON_EVENT__SHIFT 0xc +#define PA_SC_BINNER_EVENT_CNTL_2__SC_SEND_DB_VPZ__SHIFT 0xe +#define PA_SC_BINNER_EVENT_CNTL_2__BOTTOM_OF_PIPE_TS__SHIFT 0x10 +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_SX_TS__SHIFT 0x12 +#define PA_SC_BINNER_EVENT_CNTL_2__DB_CACHE_FLUSH_AND_INV__SHIFT 0x14 +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_AND_INV_DB_DATA_TS__SHIFT 0x16 +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_AND_INV_DB_META__SHIFT 0x18 +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_AND_INV_CB_DATA_TS__SHIFT 0x1a +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_AND_INV_CB_META__SHIFT 0x1c +#define PA_SC_BINNER_EVENT_CNTL_2__CS_DONE__SHIFT 0x1e +#define PA_SC_BINNER_EVENT_CNTL_2__SAMPLE_STREAMOUTSTATS_MASK 0x00000003L +#define PA_SC_BINNER_EVENT_CNTL_2__RESET_VTX_CNT_MASK 0x0000000CL +#define PA_SC_BINNER_EVENT_CNTL_2__BLOCK_CONTEXT_DONE_MASK 0x00000030L +#define PA_SC_BINNER_EVENT_CNTL_2__CS_CONTEXT_DONE_MASK 0x000000C0L +#define PA_SC_BINNER_EVENT_CNTL_2__VGT_FLUSH_MASK 0x00000300L +#define PA_SC_BINNER_EVENT_CNTL_2__TGID_ROLLOVER_MASK 0x00000C00L +#define PA_SC_BINNER_EVENT_CNTL_2__SQ_NON_EVENT_MASK 0x00003000L +#define PA_SC_BINNER_EVENT_CNTL_2__SC_SEND_DB_VPZ_MASK 0x0000C000L +#define PA_SC_BINNER_EVENT_CNTL_2__BOTTOM_OF_PIPE_TS_MASK 0x00030000L +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_SX_TS_MASK 0x000C0000L +#define PA_SC_BINNER_EVENT_CNTL_2__DB_CACHE_FLUSH_AND_INV_MASK 0x00300000L +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_AND_INV_DB_DATA_TS_MASK 0x00C00000L +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_AND_INV_DB_META_MASK 0x03000000L +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_AND_INV_CB_DATA_TS_MASK 0x0C000000L +#define PA_SC_BINNER_EVENT_CNTL_2__FLUSH_AND_INV_CB_META_MASK 0x30000000L +#define PA_SC_BINNER_EVENT_CNTL_2__CS_DONE_MASK 0xC0000000L +//PA_SC_BINNER_EVENT_CNTL_3 +#define PA_SC_BINNER_EVENT_CNTL_3__PS_DONE__SHIFT 0x0 +#define PA_SC_BINNER_EVENT_CNTL_3__FLUSH_AND_INV_CB_PIXEL_DATA__SHIFT 0x2 +#define PA_SC_BINNER_EVENT_CNTL_3__SX_CB_RAT_ACK_REQUEST__SHIFT 0x4 +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_START__SHIFT 0x6 +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_STOP__SHIFT 0x8 +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_MARKER__SHIFT 0xa +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_FLUSH__SHIFT 0xc +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_FINISH__SHIFT 0xe +#define PA_SC_BINNER_EVENT_CNTL_3__PIXEL_PIPE_STAT_CONTROL__SHIFT 0x10 +#define PA_SC_BINNER_EVENT_CNTL_3__PIXEL_PIPE_STAT_DUMP__SHIFT 0x12 +#define PA_SC_BINNER_EVENT_CNTL_3__PIXEL_PIPE_STAT_RESET__SHIFT 0x14 +#define PA_SC_BINNER_EVENT_CNTL_3__CONTEXT_SUSPEND__SHIFT 0x16 +#define PA_SC_BINNER_EVENT_CNTL_3__OFFCHIP_HS_DEALLOC__SHIFT 0x18 +#define PA_SC_BINNER_EVENT_CNTL_3__ENABLE_NGG_PIPELINE__SHIFT 0x1a +#define PA_SC_BINNER_EVENT_CNTL_3__ENABLE_LEGACY_PIPELINE__SHIFT 0x1c +#define PA_SC_BINNER_EVENT_CNTL_3__RESERVED_63__SHIFT 0x1e +#define PA_SC_BINNER_EVENT_CNTL_3__PS_DONE_MASK 0x00000003L +#define PA_SC_BINNER_EVENT_CNTL_3__FLUSH_AND_INV_CB_PIXEL_DATA_MASK 0x0000000CL +#define PA_SC_BINNER_EVENT_CNTL_3__SX_CB_RAT_ACK_REQUEST_MASK 0x00000030L +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_START_MASK 0x000000C0L +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_STOP_MASK 0x00000300L +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_MARKER_MASK 0x00000C00L +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_FLUSH_MASK 0x00003000L +#define PA_SC_BINNER_EVENT_CNTL_3__THREAD_TRACE_FINISH_MASK 0x0000C000L +#define PA_SC_BINNER_EVENT_CNTL_3__PIXEL_PIPE_STAT_CONTROL_MASK 0x00030000L +#define PA_SC_BINNER_EVENT_CNTL_3__PIXEL_PIPE_STAT_DUMP_MASK 0x000C0000L +#define PA_SC_BINNER_EVENT_CNTL_3__PIXEL_PIPE_STAT_RESET_MASK 0x00300000L +#define PA_SC_BINNER_EVENT_CNTL_3__CONTEXT_SUSPEND_MASK 0x00C00000L +#define PA_SC_BINNER_EVENT_CNTL_3__OFFCHIP_HS_DEALLOC_MASK 0x03000000L +#define PA_SC_BINNER_EVENT_CNTL_3__ENABLE_NGG_PIPELINE_MASK 0x0C000000L +#define PA_SC_BINNER_EVENT_CNTL_3__ENABLE_LEGACY_PIPELINE_MASK 0x30000000L +#define PA_SC_BINNER_EVENT_CNTL_3__RESERVED_63_MASK 0xC0000000L +//PA_SC_BINNER_TIMEOUT_COUNTER +#define PA_SC_BINNER_TIMEOUT_COUNTER__THRESHOLD__SHIFT 0x0 +#define PA_SC_BINNER_TIMEOUT_COUNTER__THRESHOLD_MASK 0xFFFFFFFFL +//PA_SC_BINNER_PERF_CNTL_0 +#define PA_SC_BINNER_PERF_CNTL_0__BIN_HIST_NUM_PRIMS_THRESHOLD__SHIFT 0x0 +#define PA_SC_BINNER_PERF_CNTL_0__BATCH_HIST_NUM_PRIMS_THRESHOLD__SHIFT 0xa +#define PA_SC_BINNER_PERF_CNTL_0__BIN_HIST_NUM_CONTEXT_THRESHOLD__SHIFT 0x14 +#define PA_SC_BINNER_PERF_CNTL_0__BATCH_HIST_NUM_CONTEXT_THRESHOLD__SHIFT 0x17 +#define PA_SC_BINNER_PERF_CNTL_0__BIN_HIST_NUM_PRIMS_THRESHOLD_MASK 0x000003FFL +#define PA_SC_BINNER_PERF_CNTL_0__BATCH_HIST_NUM_PRIMS_THRESHOLD_MASK 0x000FFC00L +#define PA_SC_BINNER_PERF_CNTL_0__BIN_HIST_NUM_CONTEXT_THRESHOLD_MASK 0x00700000L +#define PA_SC_BINNER_PERF_CNTL_0__BATCH_HIST_NUM_CONTEXT_THRESHOLD_MASK 0x03800000L +//PA_SC_BINNER_PERF_CNTL_1 +#define PA_SC_BINNER_PERF_CNTL_1__BIN_HIST_NUM_PERSISTENT_STATE_THRESHOLD__SHIFT 0x0 +#define PA_SC_BINNER_PERF_CNTL_1__BATCH_HIST_NUM_PERSISTENT_STATE_THRESHOLD__SHIFT 0x5 +#define PA_SC_BINNER_PERF_CNTL_1__BATCH_HIST_NUM_TRIV_REJECTED_PRIMS_THRESHOLD__SHIFT 0xa +#define PA_SC_BINNER_PERF_CNTL_1__BIN_HIST_NUM_PERSISTENT_STATE_THRESHOLD_MASK 0x0000001FL +#define PA_SC_BINNER_PERF_CNTL_1__BATCH_HIST_NUM_PERSISTENT_STATE_THRESHOLD_MASK 0x000003E0L +#define PA_SC_BINNER_PERF_CNTL_1__BATCH_HIST_NUM_TRIV_REJECTED_PRIMS_THRESHOLD_MASK 0x03FFFC00L +//PA_SC_BINNER_PERF_CNTL_2 +#define PA_SC_BINNER_PERF_CNTL_2__BATCH_HIST_NUM_ROWS_PER_PRIM_THRESHOLD__SHIFT 0x0 +#define PA_SC_BINNER_PERF_CNTL_2__BATCH_HIST_NUM_COLUMNS_PER_ROW_THRESHOLD__SHIFT 0xb +#define PA_SC_BINNER_PERF_CNTL_2__BATCH_HIST_NUM_ROWS_PER_PRIM_THRESHOLD_MASK 0x000007FFL +#define PA_SC_BINNER_PERF_CNTL_2__BATCH_HIST_NUM_COLUMNS_PER_ROW_THRESHOLD_MASK 0x003FF800L +//PA_SC_BINNER_PERF_CNTL_3 +#define PA_SC_BINNER_PERF_CNTL_3__BATCH_HIST_NUM_PS_WAVE_BREAKS_THRESHOLD__SHIFT 0x0 +#define PA_SC_BINNER_PERF_CNTL_3__BATCH_HIST_NUM_PS_WAVE_BREAKS_THRESHOLD_MASK 0xFFFFFFFFL +//PA_SC_FIFO_SIZE +#define PA_SC_FIFO_SIZE__SC_FRONTEND_PRIM_FIFO_SIZE__SHIFT 0x0 +#define PA_SC_FIFO_SIZE__SC_BACKEND_PRIM_FIFO_SIZE__SHIFT 0x6 +#define PA_SC_FIFO_SIZE__SC_HIZ_TILE_FIFO_SIZE__SHIFT 0xf +#define PA_SC_FIFO_SIZE__SC_EARLYZ_TILE_FIFO_SIZE__SHIFT 0x15 +#define PA_SC_FIFO_SIZE__SC_FRONTEND_PRIM_FIFO_SIZE_MASK 0x0000003FL +#define PA_SC_FIFO_SIZE__SC_BACKEND_PRIM_FIFO_SIZE_MASK 0x00007FC0L +#define PA_SC_FIFO_SIZE__SC_HIZ_TILE_FIFO_SIZE_MASK 0x001F8000L +#define PA_SC_FIFO_SIZE__SC_EARLYZ_TILE_FIFO_SIZE_MASK 0xFFE00000L +//PA_SC_IF_FIFO_SIZE +#define PA_SC_IF_FIFO_SIZE__SC_DB_TILE_IF_FIFO_SIZE__SHIFT 0x0 +#define PA_SC_IF_FIFO_SIZE__SC_DB_QUAD_IF_FIFO_SIZE__SHIFT 0x6 +#define PA_SC_IF_FIFO_SIZE__SC_SPI_IF_FIFO_SIZE__SHIFT 0xc +#define PA_SC_IF_FIFO_SIZE__SC_BCI_IF_FIFO_SIZE__SHIFT 0x12 +#define PA_SC_IF_FIFO_SIZE__SC_DB_TILE_IF_FIFO_SIZE_MASK 0x0000003FL +#define PA_SC_IF_FIFO_SIZE__SC_DB_QUAD_IF_FIFO_SIZE_MASK 0x00000FC0L +#define PA_SC_IF_FIFO_SIZE__SC_SPI_IF_FIFO_SIZE_MASK 0x0003F000L +#define PA_SC_IF_FIFO_SIZE__SC_BCI_IF_FIFO_SIZE_MASK 0x00FC0000L +//PA_SC_PKR_WAVE_TABLE_CNTL +#define PA_SC_PKR_WAVE_TABLE_CNTL__SIZE__SHIFT 0x0 +#define PA_SC_PKR_WAVE_TABLE_CNTL__SIZE_MASK 0x0000003FL +//PA_UTCL1_CNTL1 +#define PA_UTCL1_CNTL1__FORCE_4K_L2_RESP__SHIFT 0x0 +#define PA_UTCL1_CNTL1__GPUVM_64K_DEFAULT__SHIFT 0x1 +#define PA_UTCL1_CNTL1__GPUVM_PERM_MODE__SHIFT 0x2 +#define PA_UTCL1_CNTL1__RESP_MODE__SHIFT 0x3 +#define PA_UTCL1_CNTL1__RESP_FAULT_MODE__SHIFT 0x5 +#define PA_UTCL1_CNTL1__CLIENTID__SHIFT 0x7 +#define PA_UTCL1_CNTL1__SPARE__SHIFT 0x10 +#define PA_UTCL1_CNTL1__ENABLE_PUSH_LFIFO__SHIFT 0x11 +#define PA_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB__SHIFT 0x12 +#define PA_UTCL1_CNTL1__REG_INV_VMID__SHIFT 0x13 +#define PA_UTCL1_CNTL1__REG_INV_ALL_VMID__SHIFT 0x17 +#define PA_UTCL1_CNTL1__REG_INV_TOGGLE__SHIFT 0x18 +#define PA_UTCL1_CNTL1__INVALIDATE_ALL_VMID__SHIFT 0x19 +#define PA_UTCL1_CNTL1__FORCE_MISS__SHIFT 0x1a +#define PA_UTCL1_CNTL1__FORCE_IN_ORDER__SHIFT 0x1b +#define PA_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2__SHIFT 0x1c +#define PA_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2__SHIFT 0x1e +#define PA_UTCL1_CNTL1__FORCE_4K_L2_RESP_MASK 0x00000001L +#define PA_UTCL1_CNTL1__GPUVM_64K_DEFAULT_MASK 0x00000002L +#define PA_UTCL1_CNTL1__GPUVM_PERM_MODE_MASK 0x00000004L +#define PA_UTCL1_CNTL1__RESP_MODE_MASK 0x00000018L +#define PA_UTCL1_CNTL1__RESP_FAULT_MODE_MASK 0x00000060L +#define PA_UTCL1_CNTL1__CLIENTID_MASK 0x0000FF80L +#define PA_UTCL1_CNTL1__SPARE_MASK 0x00010000L +#define PA_UTCL1_CNTL1__ENABLE_PUSH_LFIFO_MASK 0x00020000L +#define PA_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB_MASK 0x00040000L +#define PA_UTCL1_CNTL1__REG_INV_VMID_MASK 0x00780000L +#define PA_UTCL1_CNTL1__REG_INV_ALL_VMID_MASK 0x00800000L +#define PA_UTCL1_CNTL1__REG_INV_TOGGLE_MASK 0x01000000L +#define PA_UTCL1_CNTL1__INVALIDATE_ALL_VMID_MASK 0x02000000L +#define PA_UTCL1_CNTL1__FORCE_MISS_MASK 0x04000000L +#define PA_UTCL1_CNTL1__FORCE_IN_ORDER_MASK 0x08000000L +#define PA_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2_MASK 0x30000000L +#define PA_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2_MASK 0xC0000000L +//PA_UTCL1_CNTL2 +#define PA_UTCL1_CNTL2__SPARE1__SHIFT 0x0 +#define PA_UTCL1_CNTL2__SPARE2__SHIFT 0x8 +#define PA_UTCL1_CNTL2__MTYPE_OVRD_DIS__SHIFT 0x9 +#define PA_UTCL1_CNTL2__LINE_VALID__SHIFT 0xa +#define PA_UTCL1_CNTL2__SPARE3__SHIFT 0xb +#define PA_UTCL1_CNTL2__GPUVM_INV_MODE__SHIFT 0xc +#define PA_UTCL1_CNTL2__ENABLE_SHOOTDOWN_OPT__SHIFT 0xd +#define PA_UTCL1_CNTL2__FORCE_SNOOP__SHIFT 0xe +#define PA_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK__SHIFT 0xf +#define PA_UTCL1_CNTL2__SPARE4__SHIFT 0x10 +#define PA_UTCL1_CNTL2__ENABLE_PERF_EVENT_RD_WR__SHIFT 0x12 +#define PA_UTCL1_CNTL2__PERF_EVENT_RD_WR__SHIFT 0x13 +#define PA_UTCL1_CNTL2__ENABLE_PERF_EVENT_VMID__SHIFT 0x14 +#define PA_UTCL1_CNTL2__PERF_EVENT_VMID__SHIFT 0x15 +#define PA_UTCL1_CNTL2__SPARE5__SHIFT 0x19 +#define PA_UTCL1_CNTL2__FORCE_FRAG_2M_TO_64K__SHIFT 0x1a +#define PA_UTCL1_CNTL2__RESERVED__SHIFT 0x1b +#define PA_UTCL1_CNTL2__SPARE1_MASK 0x000000FFL +#define PA_UTCL1_CNTL2__SPARE2_MASK 0x00000100L +#define PA_UTCL1_CNTL2__MTYPE_OVRD_DIS_MASK 0x00000200L +#define PA_UTCL1_CNTL2__LINE_VALID_MASK 0x00000400L +#define PA_UTCL1_CNTL2__SPARE3_MASK 0x00000800L +#define PA_UTCL1_CNTL2__GPUVM_INV_MODE_MASK 0x00001000L +#define PA_UTCL1_CNTL2__ENABLE_SHOOTDOWN_OPT_MASK 0x00002000L +#define PA_UTCL1_CNTL2__FORCE_SNOOP_MASK 0x00004000L +#define PA_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK_MASK 0x00008000L +#define PA_UTCL1_CNTL2__SPARE4_MASK 0x00030000L +#define PA_UTCL1_CNTL2__ENABLE_PERF_EVENT_RD_WR_MASK 0x00040000L +#define PA_UTCL1_CNTL2__PERF_EVENT_RD_WR_MASK 0x00080000L +#define PA_UTCL1_CNTL2__ENABLE_PERF_EVENT_VMID_MASK 0x00100000L +#define PA_UTCL1_CNTL2__PERF_EVENT_VMID_MASK 0x01E00000L +#define PA_UTCL1_CNTL2__SPARE5_MASK 0x02000000L +#define PA_UTCL1_CNTL2__FORCE_FRAG_2M_TO_64K_MASK 0x04000000L +#define PA_UTCL1_CNTL2__RESERVED_MASK 0xF8000000L +//PA_SIDEBAND_REQUEST_DELAYS +#define PA_SIDEBAND_REQUEST_DELAYS__RETRY_DELAY__SHIFT 0x0 +#define PA_SIDEBAND_REQUEST_DELAYS__INITIAL_DELAY__SHIFT 0x10 +#define PA_SIDEBAND_REQUEST_DELAYS__RETRY_DELAY_MASK 0x0000FFFFL +#define PA_SIDEBAND_REQUEST_DELAYS__INITIAL_DELAY_MASK 0xFFFF0000L +//PA_SC_ENHANCE +#define PA_SC_ENHANCE__ENABLE_PA_SC_OUT_OF_ORDER__SHIFT 0x0 +#define PA_SC_ENHANCE__DISABLE_SC_DB_TILE_FIX__SHIFT 0x1 +#define PA_SC_ENHANCE__DISABLE_AA_MASK_FULL_FIX__SHIFT 0x2 +#define PA_SC_ENHANCE__ENABLE_1XMSAA_SAMPLE_LOCATIONS__SHIFT 0x3 +#define PA_SC_ENHANCE__ENABLE_1XMSAA_SAMPLE_LOC_CENTROID__SHIFT 0x4 +#define PA_SC_ENHANCE__DISABLE_SCISSOR_FIX__SHIFT 0x5 +#define PA_SC_ENHANCE__SEND_UNLIT_STILES_TO_PACKER__SHIFT 0x6 +#define PA_SC_ENHANCE__DISABLE_DUALGRAD_PERF_OPTIMIZATION__SHIFT 0x7 +#define PA_SC_ENHANCE__DISABLE_SC_PROCESS_RESET_PRIM__SHIFT 0x8 +#define PA_SC_ENHANCE__DISABLE_SC_PROCESS_RESET_SUPERTILE__SHIFT 0x9 +#define PA_SC_ENHANCE__DISABLE_SC_PROCESS_RESET_TILE__SHIFT 0xa +#define PA_SC_ENHANCE__DISABLE_PA_SC_GUIDANCE__SHIFT 0xb +#define PA_SC_ENHANCE__DISABLE_EOV_ALL_CTRL_ONLY_COMBINATIONS__SHIFT 0xc +#define PA_SC_ENHANCE__ENABLE_MULTICYCLE_BUBBLE_FREEZE__SHIFT 0xd +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_PA_SC_GUIDANCE__SHIFT 0xe +#define PA_SC_ENHANCE__ENABLE_OUT_OF_ORDER_POLY_MODE__SHIFT 0xf +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_EOP_SYNC_NULL_PRIMS_LAST__SHIFT 0x10 +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_THRESHOLD_SWITCHING__SHIFT 0x11 +#define PA_SC_ENHANCE__ENABLE_OUT_OF_ORDER_THRESHOLD_SWITCH_AT_EOPG_ONLY__SHIFT 0x12 +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_DESIRED_FIFO_EMPTY_SWITCHING__SHIFT 0x13 +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_SELECTED_FIFO_EMPTY_SWITCHING__SHIFT 0x14 +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_EMPTY_SWITCHING_HYSTERYSIS__SHIFT 0x15 +#define PA_SC_ENHANCE__ENABLE_OUT_OF_ORDER_DESIRED_FIFO_IS_NEXT_FEID__SHIFT 0x16 +#define PA_SC_ENHANCE__DISABLE_OOO_NO_EOPG_SKEW_DESIRED_FIFO_IS_CURRENT_FIFO__SHIFT 0x17 +#define PA_SC_ENHANCE__OOO_DISABLE_EOP_ON_FIRST_LIVE_PRIM_HIT__SHIFT 0x18 +#define PA_SC_ENHANCE__OOO_DISABLE_EOPG_SKEW_THRESHOLD_SWITCHING__SHIFT 0x19 +#define PA_SC_ENHANCE__DISABLE_EOP_LINE_STIPPLE_RESET__SHIFT 0x1a +#define PA_SC_ENHANCE__DISABLE_VPZ_EOP_LINE_STIPPLE_RESET__SHIFT 0x1b +#define PA_SC_ENHANCE__IOO_DISABLE_SCAN_UNSELECTED_FIFOS_FOR_DUAL_GFX_RING_CHANGE__SHIFT 0x1c +#define PA_SC_ENHANCE__OOO_USE_ABSOLUTE_FIFO_COUNT_IN_THRESHOLD_SWITCHING__SHIFT 0x1d +#define PA_SC_ENHANCE__ENABLE_PA_SC_OUT_OF_ORDER_MASK 0x00000001L +#define PA_SC_ENHANCE__DISABLE_SC_DB_TILE_FIX_MASK 0x00000002L +#define PA_SC_ENHANCE__DISABLE_AA_MASK_FULL_FIX_MASK 0x00000004L +#define PA_SC_ENHANCE__ENABLE_1XMSAA_SAMPLE_LOCATIONS_MASK 0x00000008L +#define PA_SC_ENHANCE__ENABLE_1XMSAA_SAMPLE_LOC_CENTROID_MASK 0x00000010L +#define PA_SC_ENHANCE__DISABLE_SCISSOR_FIX_MASK 0x00000020L +#define PA_SC_ENHANCE__SEND_UNLIT_STILES_TO_PACKER_MASK 0x00000040L +#define PA_SC_ENHANCE__DISABLE_DUALGRAD_PERF_OPTIMIZATION_MASK 0x00000080L +#define PA_SC_ENHANCE__DISABLE_SC_PROCESS_RESET_PRIM_MASK 0x00000100L +#define PA_SC_ENHANCE__DISABLE_SC_PROCESS_RESET_SUPERTILE_MASK 0x00000200L +#define PA_SC_ENHANCE__DISABLE_SC_PROCESS_RESET_TILE_MASK 0x00000400L +#define PA_SC_ENHANCE__DISABLE_PA_SC_GUIDANCE_MASK 0x00000800L +#define PA_SC_ENHANCE__DISABLE_EOV_ALL_CTRL_ONLY_COMBINATIONS_MASK 0x00001000L +#define PA_SC_ENHANCE__ENABLE_MULTICYCLE_BUBBLE_FREEZE_MASK 0x00002000L +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_PA_SC_GUIDANCE_MASK 0x00004000L +#define PA_SC_ENHANCE__ENABLE_OUT_OF_ORDER_POLY_MODE_MASK 0x00008000L +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_EOP_SYNC_NULL_PRIMS_LAST_MASK 0x00010000L +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_THRESHOLD_SWITCHING_MASK 0x00020000L +#define PA_SC_ENHANCE__ENABLE_OUT_OF_ORDER_THRESHOLD_SWITCH_AT_EOPG_ONLY_MASK 0x00040000L +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_DESIRED_FIFO_EMPTY_SWITCHING_MASK 0x00080000L +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_SELECTED_FIFO_EMPTY_SWITCHING_MASK 0x00100000L +#define PA_SC_ENHANCE__DISABLE_OUT_OF_ORDER_EMPTY_SWITCHING_HYSTERYSIS_MASK 0x00200000L +#define PA_SC_ENHANCE__ENABLE_OUT_OF_ORDER_DESIRED_FIFO_IS_NEXT_FEID_MASK 0x00400000L +#define PA_SC_ENHANCE__DISABLE_OOO_NO_EOPG_SKEW_DESIRED_FIFO_IS_CURRENT_FIFO_MASK 0x00800000L +#define PA_SC_ENHANCE__OOO_DISABLE_EOP_ON_FIRST_LIVE_PRIM_HIT_MASK 0x01000000L +#define PA_SC_ENHANCE__OOO_DISABLE_EOPG_SKEW_THRESHOLD_SWITCHING_MASK 0x02000000L +#define PA_SC_ENHANCE__DISABLE_EOP_LINE_STIPPLE_RESET_MASK 0x04000000L +#define PA_SC_ENHANCE__DISABLE_VPZ_EOP_LINE_STIPPLE_RESET_MASK 0x08000000L +#define PA_SC_ENHANCE__IOO_DISABLE_SCAN_UNSELECTED_FIFOS_FOR_DUAL_GFX_RING_CHANGE_MASK 0x10000000L +#define PA_SC_ENHANCE__OOO_USE_ABSOLUTE_FIFO_COUNT_IN_THRESHOLD_SWITCHING_MASK 0x20000000L +//PA_SC_ENHANCE_1 +#define PA_SC_ENHANCE_1__REALIGN_DQUADS_OVERRIDE_ENABLE__SHIFT 0x0 +#define PA_SC_ENHANCE_1__REALIGN_DQUADS_OVERRIDE__SHIFT 0x1 +#define PA_SC_ENHANCE_1__DISABLE_SC_BINNING__SHIFT 0x3 +#define PA_SC_ENHANCE_1__BYPASS_PBB__SHIFT 0x4 +#define PA_SC_ENHANCE_1__ECO_SPARE0__SHIFT 0x5 +#define PA_SC_ENHANCE_1__ECO_SPARE1__SHIFT 0x6 +#define PA_SC_ENHANCE_1__ECO_SPARE2__SHIFT 0x7 +#define PA_SC_ENHANCE_1__ECO_SPARE3__SHIFT 0x8 +#define PA_SC_ENHANCE_1__DISABLE_SC_PROCESS_RESET_PBB__SHIFT 0x9 +#define PA_SC_ENHANCE_1__DISABLE_PBB_SCISSOR_OPT__SHIFT 0xa +#define PA_SC_ENHANCE_1__ENABLE_DFSM_FLUSH_EVENT_TO_FLUSH_POPS_CAM__SHIFT 0xb +#define PA_SC_ENHANCE_1__DISABLE_PACKER_GRAD_FDCE_ENHANCE__SHIFT 0xd +#define PA_SC_ENHANCE_1__DISABLE_SC_DB_TILE_INTF_FINE_CLOCK_GATE__SHIFT 0xe +#define PA_SC_ENHANCE_1__DISABLE_SC_PIPELINE_RESET_LEGACY_MODE_TRANSITION__SHIFT 0xf +#define PA_SC_ENHANCE_1__DISABLE_PACKER_ODC_ENHANCE__SHIFT 0x10 +#define PA_SC_ENHANCE_1__ALLOW_SCALE_LINE_WIDTH_PAD_WITH_BINNING__SHIFT 0x11 +#define PA_SC_ENHANCE_1__OPTIMAL_BIN_SELECTION__SHIFT 0x12 +#define PA_SC_ENHANCE_1__DISABLE_FORCE_SOP_ALL_EVENTS__SHIFT 0x13 +#define PA_SC_ENHANCE_1__DISABLE_PBB_CLK_OPTIMIZATION__SHIFT 0x14 +#define PA_SC_ENHANCE_1__DISABLE_PBB_SCISSOR_CLK_OPTIMIZATION__SHIFT 0x15 +#define PA_SC_ENHANCE_1__DISABLE_PBB_BINNING_CLK_OPTIMIZATION__SHIFT 0x16 +#define PA_SC_ENHANCE_1__RSVD__SHIFT 0x17 +#define PA_SC_ENHANCE_1__REALIGN_DQUADS_OVERRIDE_ENABLE_MASK 0x00000001L +#define PA_SC_ENHANCE_1__REALIGN_DQUADS_OVERRIDE_MASK 0x00000006L +#define PA_SC_ENHANCE_1__DISABLE_SC_BINNING_MASK 0x00000008L +#define PA_SC_ENHANCE_1__BYPASS_PBB_MASK 0x00000010L +#define PA_SC_ENHANCE_1__ECO_SPARE0_MASK 0x00000020L +#define PA_SC_ENHANCE_1__ECO_SPARE1_MASK 0x00000040L +#define PA_SC_ENHANCE_1__ECO_SPARE2_MASK 0x00000080L +#define PA_SC_ENHANCE_1__ECO_SPARE3_MASK 0x00000100L +#define PA_SC_ENHANCE_1__DISABLE_SC_PROCESS_RESET_PBB_MASK 0x00000200L +#define PA_SC_ENHANCE_1__DISABLE_PBB_SCISSOR_OPT_MASK 0x00000400L +#define PA_SC_ENHANCE_1__ENABLE_DFSM_FLUSH_EVENT_TO_FLUSH_POPS_CAM_MASK 0x00000800L +#define PA_SC_ENHANCE_1__DISABLE_PACKER_GRAD_FDCE_ENHANCE_MASK 0x00002000L +#define PA_SC_ENHANCE_1__DISABLE_SC_DB_TILE_INTF_FINE_CLOCK_GATE_MASK 0x00004000L +#define PA_SC_ENHANCE_1__DISABLE_SC_PIPELINE_RESET_LEGACY_MODE_TRANSITION_MASK 0x00008000L +#define PA_SC_ENHANCE_1__DISABLE_PACKER_ODC_ENHANCE_MASK 0x00010000L +#define PA_SC_ENHANCE_1__ALLOW_SCALE_LINE_WIDTH_PAD_WITH_BINNING_MASK 0x00020000L +#define PA_SC_ENHANCE_1__OPTIMAL_BIN_SELECTION_MASK 0x00040000L +#define PA_SC_ENHANCE_1__DISABLE_FORCE_SOP_ALL_EVENTS_MASK 0x00080000L +#define PA_SC_ENHANCE_1__DISABLE_PBB_CLK_OPTIMIZATION_MASK 0x00100000L +#define PA_SC_ENHANCE_1__DISABLE_PBB_SCISSOR_CLK_OPTIMIZATION_MASK 0x00200000L +#define PA_SC_ENHANCE_1__DISABLE_PBB_BINNING_CLK_OPTIMIZATION_MASK 0x00400000L +#define PA_SC_ENHANCE_1__RSVD_MASK 0xFF800000L +//PA_SC_DSM_CNTL +#define PA_SC_DSM_CNTL__FORCE_EOV_REZ_0__SHIFT 0x0 +#define PA_SC_DSM_CNTL__FORCE_EOV_REZ_1__SHIFT 0x1 +#define PA_SC_DSM_CNTL__FORCE_EOV_REZ_0_MASK 0x00000001L +#define PA_SC_DSM_CNTL__FORCE_EOV_REZ_1_MASK 0x00000002L +//PA_SC_TILE_STEERING_CREST_OVERRIDE +#define PA_SC_TILE_STEERING_CREST_OVERRIDE__ONE_RB_MODE_ENABLE__SHIFT 0x0 +#define PA_SC_TILE_STEERING_CREST_OVERRIDE__SE_SELECT__SHIFT 0x1 +#define PA_SC_TILE_STEERING_CREST_OVERRIDE__RB_SELECT__SHIFT 0x5 +#define PA_SC_TILE_STEERING_CREST_OVERRIDE__ONE_RB_MODE_ENABLE_MASK 0x00000001L +#define PA_SC_TILE_STEERING_CREST_OVERRIDE__SE_SELECT_MASK 0x00000006L +#define PA_SC_TILE_STEERING_CREST_OVERRIDE__RB_SELECT_MASK 0x00000060L + + +// addressBlock: gc_sqdec +//SQ_CONFIG +#define SQ_CONFIG__UNUSED__SHIFT 0x0 +#define SQ_CONFIG__OVERRIDE_ALU_BUSY__SHIFT 0x7 +#define SQ_CONFIG__OVERRIDE_LDS_IDX_BUSY__SHIFT 0xb +#define SQ_CONFIG__EARLY_TA_DONE_DISABLE__SHIFT 0xc +#define SQ_CONFIG__DUA_FLAT_LOCK_ENABLE__SHIFT 0xd +#define SQ_CONFIG__DUA_LDS_BYPASS_DISABLE__SHIFT 0xe +#define SQ_CONFIG__DUA_FLAT_LDS_PINGPONG_DISABLE__SHIFT 0xf +#define SQ_CONFIG__DISABLE_VMEM_SOFT_CLAUSE__SHIFT 0x10 +#define SQ_CONFIG__DISABLE_SMEM_SOFT_CLAUSE__SHIFT 0x11 +#define SQ_CONFIG__ENABLE_HIPRIO_ON_EXP_RDY_VS__SHIFT 0x12 +#define SQ_CONFIG__PRIO_VAL_ON_EXP_RDY_VS__SHIFT 0x13 +#define SQ_CONFIG__REPLAY_SLEEP_CNT__SHIFT 0x15 +#define SQ_CONFIG__DISABLE_SP_VGPR_WRITE_SKIP__SHIFT 0x1c +#define SQ_CONFIG__DISABLE_SP_REDUNDANT_THREAD_GATING__SHIFT 0x1d +#define SQ_CONFIG__DISABLE_FLAT_SOFT_CLAUSE__SHIFT 0x1e +#define SQ_CONFIG__DISABLE_MIMG_SOFT_CLAUSE__SHIFT 0x1f +#define SQ_CONFIG__UNUSED_MASK 0x0000007FL +#define SQ_CONFIG__OVERRIDE_ALU_BUSY_MASK 0x00000080L +#define SQ_CONFIG__OVERRIDE_LDS_IDX_BUSY_MASK 0x00000800L +#define SQ_CONFIG__EARLY_TA_DONE_DISABLE_MASK 0x00001000L +#define SQ_CONFIG__DUA_FLAT_LOCK_ENABLE_MASK 0x00002000L +#define SQ_CONFIG__DUA_LDS_BYPASS_DISABLE_MASK 0x00004000L +#define SQ_CONFIG__DUA_FLAT_LDS_PINGPONG_DISABLE_MASK 0x00008000L +#define SQ_CONFIG__DISABLE_VMEM_SOFT_CLAUSE_MASK 0x00010000L +#define SQ_CONFIG__DISABLE_SMEM_SOFT_CLAUSE_MASK 0x00020000L +#define SQ_CONFIG__ENABLE_HIPRIO_ON_EXP_RDY_VS_MASK 0x00040000L +#define SQ_CONFIG__PRIO_VAL_ON_EXP_RDY_VS_MASK 0x00180000L +#define SQ_CONFIG__REPLAY_SLEEP_CNT_MASK 0x0FE00000L +#define SQ_CONFIG__DISABLE_SP_VGPR_WRITE_SKIP_MASK 0x10000000L +#define SQ_CONFIG__DISABLE_SP_REDUNDANT_THREAD_GATING_MASK 0x20000000L +#define SQ_CONFIG__DISABLE_FLAT_SOFT_CLAUSE_MASK 0x40000000L +#define SQ_CONFIG__DISABLE_MIMG_SOFT_CLAUSE_MASK 0x80000000L +//SQC_CONFIG +#define SQC_CONFIG__INST_CACHE_SIZE__SHIFT 0x0 +#define SQC_CONFIG__DATA_CACHE_SIZE__SHIFT 0x2 +#define SQC_CONFIG__MISS_FIFO_DEPTH__SHIFT 0x4 +#define SQC_CONFIG__HIT_FIFO_DEPTH__SHIFT 0x6 +#define SQC_CONFIG__FORCE_ALWAYS_MISS__SHIFT 0x7 +#define SQC_CONFIG__FORCE_IN_ORDER__SHIFT 0x8 +#define SQC_CONFIG__IDENTITY_HASH_BANK__SHIFT 0x9 +#define SQC_CONFIG__IDENTITY_HASH_SET__SHIFT 0xa +#define SQC_CONFIG__PER_VMID_INV_DISABLE__SHIFT 0xb +#define SQC_CONFIG__EVICT_LRU__SHIFT 0xc +#define SQC_CONFIG__FORCE_2_BANK__SHIFT 0xe +#define SQC_CONFIG__FORCE_1_BANK__SHIFT 0xf +#define SQC_CONFIG__LS_DISABLE_CLOCKS__SHIFT 0x10 +#define SQC_CONFIG__INST_PRF_COUNT__SHIFT 0x18 +#define SQC_CONFIG__INST_PRF_FILTER_DIS__SHIFT 0x1a +#define SQC_CONFIG__INST_CACHE_SIZE_MASK 0x00000003L +#define SQC_CONFIG__DATA_CACHE_SIZE_MASK 0x0000000CL +#define SQC_CONFIG__MISS_FIFO_DEPTH_MASK 0x00000030L +#define SQC_CONFIG__HIT_FIFO_DEPTH_MASK 0x00000040L +#define SQC_CONFIG__FORCE_ALWAYS_MISS_MASK 0x00000080L +#define SQC_CONFIG__FORCE_IN_ORDER_MASK 0x00000100L +#define SQC_CONFIG__IDENTITY_HASH_BANK_MASK 0x00000200L +#define SQC_CONFIG__IDENTITY_HASH_SET_MASK 0x00000400L +#define SQC_CONFIG__PER_VMID_INV_DISABLE_MASK 0x00000800L +#define SQC_CONFIG__EVICT_LRU_MASK 0x00003000L +#define SQC_CONFIG__FORCE_2_BANK_MASK 0x00004000L +#define SQC_CONFIG__FORCE_1_BANK_MASK 0x00008000L +#define SQC_CONFIG__LS_DISABLE_CLOCKS_MASK 0x00FF0000L +#define SQC_CONFIG__INST_PRF_COUNT_MASK 0x03000000L +#define SQC_CONFIG__INST_PRF_FILTER_DIS_MASK 0x04000000L +//LDS_CONFIG +#define LDS_CONFIG__ADDR_OUT_OF_RANGE_REPORTING__SHIFT 0x0 +#define LDS_CONFIG__ADDR_OUT_OF_RANGE_REPORTING_MASK 0x00000001L +//SQ_RANDOM_WAVE_PRI +#define SQ_RANDOM_WAVE_PRI__RET__SHIFT 0x0 +#define SQ_RANDOM_WAVE_PRI__RUI__SHIFT 0x7 +#define SQ_RANDOM_WAVE_PRI__RNG__SHIFT 0xa +#define SQ_RANDOM_WAVE_PRI__RET_MASK 0x0000007FL +#define SQ_RANDOM_WAVE_PRI__RUI_MASK 0x00000380L +#define SQ_RANDOM_WAVE_PRI__RNG_MASK 0x007FFC00L +//SQ_REG_CREDITS +#define SQ_REG_CREDITS__SRBM_CREDITS__SHIFT 0x0 +#define SQ_REG_CREDITS__CMD_CREDITS__SHIFT 0x8 +#define SQ_REG_CREDITS__REG_BUSY__SHIFT 0x1c +#define SQ_REG_CREDITS__SRBM_OVERFLOW__SHIFT 0x1d +#define SQ_REG_CREDITS__IMMED_OVERFLOW__SHIFT 0x1e +#define SQ_REG_CREDITS__CMD_OVERFLOW__SHIFT 0x1f +#define SQ_REG_CREDITS__SRBM_CREDITS_MASK 0x0000003FL +#define SQ_REG_CREDITS__CMD_CREDITS_MASK 0x00000F00L +#define SQ_REG_CREDITS__REG_BUSY_MASK 0x10000000L +#define SQ_REG_CREDITS__SRBM_OVERFLOW_MASK 0x20000000L +#define SQ_REG_CREDITS__IMMED_OVERFLOW_MASK 0x40000000L +#define SQ_REG_CREDITS__CMD_OVERFLOW_MASK 0x80000000L +//SQ_FIFO_SIZES +#define SQ_FIFO_SIZES__INTERRUPT_FIFO_SIZE__SHIFT 0x0 +#define SQ_FIFO_SIZES__TTRACE_FIFO_SIZE__SHIFT 0x8 +#define SQ_FIFO_SIZES__EXPORT_BUF_SIZE__SHIFT 0x10 +#define SQ_FIFO_SIZES__VMEM_DATA_FIFO_SIZE__SHIFT 0x12 +#define SQ_FIFO_SIZES__INTERRUPT_FIFO_SIZE_MASK 0x0000000FL +#define SQ_FIFO_SIZES__TTRACE_FIFO_SIZE_MASK 0x00000F00L +#define SQ_FIFO_SIZES__EXPORT_BUF_SIZE_MASK 0x00030000L +#define SQ_FIFO_SIZES__VMEM_DATA_FIFO_SIZE_MASK 0x000C0000L +//SQ_DSM_CNTL +#define SQ_DSM_CNTL__WAVEFRONT_STALL_0__SHIFT 0x0 +#define SQ_DSM_CNTL__WAVEFRONT_STALL_1__SHIFT 0x1 +#define SQ_DSM_CNTL__SPI_BACKPRESSURE_0__SHIFT 0x2 +#define SQ_DSM_CNTL__SPI_BACKPRESSURE_1__SHIFT 0x3 +#define SQ_DSM_CNTL__SEL_DSM_SGPR_IRRITATOR_DATA0__SHIFT 0x8 +#define SQ_DSM_CNTL__SEL_DSM_SGPR_IRRITATOR_DATA1__SHIFT 0x9 +#define SQ_DSM_CNTL__SGPR_ENABLE_SINGLE_WRITE__SHIFT 0xa +#define SQ_DSM_CNTL__SEL_DSM_LDS_IRRITATOR_DATA0__SHIFT 0x10 +#define SQ_DSM_CNTL__SEL_DSM_LDS_IRRITATOR_DATA1__SHIFT 0x11 +#define SQ_DSM_CNTL__LDS_ENABLE_SINGLE_WRITE01__SHIFT 0x12 +#define SQ_DSM_CNTL__SEL_DSM_LDS_IRRITATOR_DATA2__SHIFT 0x13 +#define SQ_DSM_CNTL__SEL_DSM_LDS_IRRITATOR_DATA3__SHIFT 0x14 +#define SQ_DSM_CNTL__LDS_ENABLE_SINGLE_WRITE23__SHIFT 0x15 +#define SQ_DSM_CNTL__SEL_DSM_SP_IRRITATOR_DATA0__SHIFT 0x18 +#define SQ_DSM_CNTL__SEL_DSM_SP_IRRITATOR_DATA1__SHIFT 0x19 +#define SQ_DSM_CNTL__SP_ENABLE_SINGLE_WRITE__SHIFT 0x1a +#define SQ_DSM_CNTL__WAVEFRONT_STALL_0_MASK 0x00000001L +#define SQ_DSM_CNTL__WAVEFRONT_STALL_1_MASK 0x00000002L +#define SQ_DSM_CNTL__SPI_BACKPRESSURE_0_MASK 0x00000004L +#define SQ_DSM_CNTL__SPI_BACKPRESSURE_1_MASK 0x00000008L +#define SQ_DSM_CNTL__SEL_DSM_SGPR_IRRITATOR_DATA0_MASK 0x00000100L +#define SQ_DSM_CNTL__SEL_DSM_SGPR_IRRITATOR_DATA1_MASK 0x00000200L +#define SQ_DSM_CNTL__SGPR_ENABLE_SINGLE_WRITE_MASK 0x00000400L +#define SQ_DSM_CNTL__SEL_DSM_LDS_IRRITATOR_DATA0_MASK 0x00010000L +#define SQ_DSM_CNTL__SEL_DSM_LDS_IRRITATOR_DATA1_MASK 0x00020000L +#define SQ_DSM_CNTL__LDS_ENABLE_SINGLE_WRITE01_MASK 0x00040000L +#define SQ_DSM_CNTL__SEL_DSM_LDS_IRRITATOR_DATA2_MASK 0x00080000L +#define SQ_DSM_CNTL__SEL_DSM_LDS_IRRITATOR_DATA3_MASK 0x00100000L +#define SQ_DSM_CNTL__LDS_ENABLE_SINGLE_WRITE23_MASK 0x00200000L +#define SQ_DSM_CNTL__SEL_DSM_SP_IRRITATOR_DATA0_MASK 0x01000000L +#define SQ_DSM_CNTL__SEL_DSM_SP_IRRITATOR_DATA1_MASK 0x02000000L +#define SQ_DSM_CNTL__SP_ENABLE_SINGLE_WRITE_MASK 0x04000000L +//SQ_DSM_CNTL2 +#define SQ_DSM_CNTL2__SGPR_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define SQ_DSM_CNTL2__SGPR_SELECT_INJECT_DELAY__SHIFT 0x2 +#define SQ_DSM_CNTL2__LDS_D_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define SQ_DSM_CNTL2__LDS_D_SELECT_INJECT_DELAY__SHIFT 0x5 +#define SQ_DSM_CNTL2__LDS_I_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define SQ_DSM_CNTL2__LDS_I_SELECT_INJECT_DELAY__SHIFT 0x8 +#define SQ_DSM_CNTL2__SP_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define SQ_DSM_CNTL2__SP_SELECT_INJECT_DELAY__SHIFT 0xb +#define SQ_DSM_CNTL2__LDS_INJECT_DELAY__SHIFT 0xe +#define SQ_DSM_CNTL2__SP_INJECT_DELAY__SHIFT 0x14 +#define SQ_DSM_CNTL2__SQ_INJECT_DELAY__SHIFT 0x1a +#define SQ_DSM_CNTL2__SGPR_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define SQ_DSM_CNTL2__SGPR_SELECT_INJECT_DELAY_MASK 0x00000004L +#define SQ_DSM_CNTL2__LDS_D_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define SQ_DSM_CNTL2__LDS_D_SELECT_INJECT_DELAY_MASK 0x00000020L +#define SQ_DSM_CNTL2__LDS_I_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define SQ_DSM_CNTL2__LDS_I_SELECT_INJECT_DELAY_MASK 0x00000100L +#define SQ_DSM_CNTL2__SP_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define SQ_DSM_CNTL2__SP_SELECT_INJECT_DELAY_MASK 0x00000800L +#define SQ_DSM_CNTL2__LDS_INJECT_DELAY_MASK 0x000FC000L +#define SQ_DSM_CNTL2__SP_INJECT_DELAY_MASK 0x03F00000L +#define SQ_DSM_CNTL2__SQ_INJECT_DELAY_MASK 0xFC000000L +//SQ_RUNTIME_CONFIG +#define SQ_RUNTIME_CONFIG__ENABLE_TEX_ARB_OLDEST__SHIFT 0x0 +#define SQ_RUNTIME_CONFIG__ENABLE_TEX_ARB_OLDEST_MASK 0x00000001L +//SH_MEM_BASES +#define SH_MEM_BASES__PRIVATE_BASE__SHIFT 0x0 +#define SH_MEM_BASES__SHARED_BASE__SHIFT 0x10 +#define SH_MEM_BASES__PRIVATE_BASE_MASK 0x0000FFFFL +#define SH_MEM_BASES__SHARED_BASE_MASK 0xFFFF0000L +//SH_MEM_CONFIG +#define SH_MEM_CONFIG__ADDRESS_MODE__SHIFT 0x0 +#define SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT 0x3 +#define SH_MEM_CONFIG__RETRY_DISABLE__SHIFT 0xc +#define SH_MEM_CONFIG__PRIVATE_NV__SHIFT 0xd +#define SH_MEM_CONFIG__ADDRESS_MODE_MASK 0x00000001L +#define SH_MEM_CONFIG__ALIGNMENT_MODE_MASK 0x00000018L +#define SH_MEM_CONFIG__RETRY_DISABLE_MASK 0x00001000L +#define SH_MEM_CONFIG__PRIVATE_NV_MASK 0x00002000L +//CC_GC_SHADER_RATE_CONFIG +#define CC_GC_SHADER_RATE_CONFIG__DPFP_RATE__SHIFT 0x1 +#define CC_GC_SHADER_RATE_CONFIG__SQC_BALANCE_DISABLE__SHIFT 0x3 +#define CC_GC_SHADER_RATE_CONFIG__HALF_LDS__SHIFT 0x4 +#define CC_GC_SHADER_RATE_CONFIG__DPFP_RATE_MASK 0x00000006L +#define CC_GC_SHADER_RATE_CONFIG__SQC_BALANCE_DISABLE_MASK 0x00000008L +#define CC_GC_SHADER_RATE_CONFIG__HALF_LDS_MASK 0x00000010L +//GC_USER_SHADER_RATE_CONFIG +#define GC_USER_SHADER_RATE_CONFIG__DPFP_RATE__SHIFT 0x1 +#define GC_USER_SHADER_RATE_CONFIG__SQC_BALANCE_DISABLE__SHIFT 0x3 +#define GC_USER_SHADER_RATE_CONFIG__HALF_LDS__SHIFT 0x4 +#define GC_USER_SHADER_RATE_CONFIG__DPFP_RATE_MASK 0x00000006L +#define GC_USER_SHADER_RATE_CONFIG__SQC_BALANCE_DISABLE_MASK 0x00000008L +#define GC_USER_SHADER_RATE_CONFIG__HALF_LDS_MASK 0x00000010L +//SQ_INTERRUPT_AUTO_MASK +#define SQ_INTERRUPT_AUTO_MASK__MASK__SHIFT 0x0 +#define SQ_INTERRUPT_AUTO_MASK__MASK_MASK 0x00FFFFFFL +//SQ_INTERRUPT_MSG_CTRL +#define SQ_INTERRUPT_MSG_CTRL__STALL__SHIFT 0x0 +#define SQ_INTERRUPT_MSG_CTRL__STALL_MASK 0x00000001L +//SQ_UTCL1_CNTL1 +#define SQ_UTCL1_CNTL1__FORCE_4K_L2_RESP__SHIFT 0x0 +#define SQ_UTCL1_CNTL1__GPUVM_64K_DEF__SHIFT 0x1 +#define SQ_UTCL1_CNTL1__GPUVM_PERM_MODE__SHIFT 0x2 +#define SQ_UTCL1_CNTL1__RESP_MODE__SHIFT 0x3 +#define SQ_UTCL1_CNTL1__RESP_FAULT_MODE__SHIFT 0x5 +#define SQ_UTCL1_CNTL1__CLIENTID__SHIFT 0x7 +#define SQ_UTCL1_CNTL1__USERVM_DIS__SHIFT 0x10 +#define SQ_UTCL1_CNTL1__ENABLE_PUSH_LFIFO__SHIFT 0x11 +#define SQ_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB__SHIFT 0x12 +#define SQ_UTCL1_CNTL1__REG_INVALIDATE_VMID__SHIFT 0x13 +#define SQ_UTCL1_CNTL1__REG_INVALIDATE_ALL_VMID__SHIFT 0x17 +#define SQ_UTCL1_CNTL1__REG_INVALIDATE_TOGGLE__SHIFT 0x18 +#define SQ_UTCL1_CNTL1__REG_INVALIDATE_ALL__SHIFT 0x19 +#define SQ_UTCL1_CNTL1__FORCE_MISS__SHIFT 0x1a +#define SQ_UTCL1_CNTL1__FORCE_IN_ORDER__SHIFT 0x1b +#define SQ_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2__SHIFT 0x1c +#define SQ_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2__SHIFT 0x1e +#define SQ_UTCL1_CNTL1__FORCE_4K_L2_RESP_MASK 0x00000001L +#define SQ_UTCL1_CNTL1__GPUVM_64K_DEF_MASK 0x00000002L +#define SQ_UTCL1_CNTL1__GPUVM_PERM_MODE_MASK 0x00000004L +#define SQ_UTCL1_CNTL1__RESP_MODE_MASK 0x00000018L +#define SQ_UTCL1_CNTL1__RESP_FAULT_MODE_MASK 0x00000060L +#define SQ_UTCL1_CNTL1__CLIENTID_MASK 0x0000FF80L +#define SQ_UTCL1_CNTL1__USERVM_DIS_MASK 0x00010000L +#define SQ_UTCL1_CNTL1__ENABLE_PUSH_LFIFO_MASK 0x00020000L +#define SQ_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB_MASK 0x00040000L +#define SQ_UTCL1_CNTL1__REG_INVALIDATE_VMID_MASK 0x00780000L +#define SQ_UTCL1_CNTL1__REG_INVALIDATE_ALL_VMID_MASK 0x00800000L +#define SQ_UTCL1_CNTL1__REG_INVALIDATE_TOGGLE_MASK 0x01000000L +#define SQ_UTCL1_CNTL1__REG_INVALIDATE_ALL_MASK 0x02000000L +#define SQ_UTCL1_CNTL1__FORCE_MISS_MASK 0x04000000L +#define SQ_UTCL1_CNTL1__FORCE_IN_ORDER_MASK 0x08000000L +#define SQ_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2_MASK 0x30000000L +#define SQ_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2_MASK 0xC0000000L +//SQ_UTCL1_CNTL2 +#define SQ_UTCL1_CNTL2__SPARE__SHIFT 0x0 +#define SQ_UTCL1_CNTL2__LFIFO_SCAN_DISABLE__SHIFT 0x8 +#define SQ_UTCL1_CNTL2__MTYPE_OVRD_DIS__SHIFT 0x9 +#define SQ_UTCL1_CNTL2__LINE_VALID__SHIFT 0xa +#define SQ_UTCL1_CNTL2__DIS_EDC__SHIFT 0xb +#define SQ_UTCL1_CNTL2__GPUVM_INV_MODE__SHIFT 0xc +#define SQ_UTCL1_CNTL2__SHOOTDOWN_OPT__SHIFT 0xd +#define SQ_UTCL1_CNTL2__FORCE_SNOOP__SHIFT 0xe +#define SQ_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK__SHIFT 0xf +#define SQ_UTCL1_CNTL2__RETRY_TIMER__SHIFT 0x10 +#define SQ_UTCL1_CNTL2__FORCE_FRAG_2M_TO_64K__SHIFT 0x1a +#define SQ_UTCL1_CNTL2__PREFETCH_PAGE__SHIFT 0x1c +#define SQ_UTCL1_CNTL2__SPARE_MASK 0x000000FFL +#define SQ_UTCL1_CNTL2__LFIFO_SCAN_DISABLE_MASK 0x00000100L +#define SQ_UTCL1_CNTL2__MTYPE_OVRD_DIS_MASK 0x00000200L +#define SQ_UTCL1_CNTL2__LINE_VALID_MASK 0x00000400L +#define SQ_UTCL1_CNTL2__DIS_EDC_MASK 0x00000800L +#define SQ_UTCL1_CNTL2__GPUVM_INV_MODE_MASK 0x00001000L +#define SQ_UTCL1_CNTL2__SHOOTDOWN_OPT_MASK 0x00002000L +#define SQ_UTCL1_CNTL2__FORCE_SNOOP_MASK 0x00004000L +#define SQ_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK_MASK 0x00008000L +#define SQ_UTCL1_CNTL2__RETRY_TIMER_MASK 0x007F0000L +#define SQ_UTCL1_CNTL2__FORCE_FRAG_2M_TO_64K_MASK 0x04000000L +#define SQ_UTCL1_CNTL2__PREFETCH_PAGE_MASK 0xF0000000L +//SQ_UTCL1_STATUS +#define SQ_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define SQ_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define SQ_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define SQ_UTCL1_STATUS__RESERVED__SHIFT 0x3 +#define SQ_UTCL1_STATUS__UNUSED__SHIFT 0x10 +#define SQ_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define SQ_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define SQ_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L +#define SQ_UTCL1_STATUS__RESERVED_MASK 0x0000FFF8L +#define SQ_UTCL1_STATUS__UNUSED_MASK 0xFFFF0000L +//SQ_SHADER_TBA_LO +#define SQ_SHADER_TBA_LO__ADDR_LO__SHIFT 0x0 +#define SQ_SHADER_TBA_LO__ADDR_LO_MASK 0xFFFFFFFFL +//SQ_SHADER_TBA_HI +#define SQ_SHADER_TBA_HI__ADDR_HI__SHIFT 0x0 +#define SQ_SHADER_TBA_HI__ADDR_HI_MASK 0x000000FFL +//SQ_SHADER_TMA_LO +#define SQ_SHADER_TMA_LO__ADDR_LO__SHIFT 0x0 +#define SQ_SHADER_TMA_LO__ADDR_LO_MASK 0xFFFFFFFFL +//SQ_SHADER_TMA_HI +#define SQ_SHADER_TMA_HI__ADDR_HI__SHIFT 0x0 +#define SQ_SHADER_TMA_HI__ADDR_HI_MASK 0x000000FFL +//SQC_DSM_CNTL +#define SQC_DSM_CNTL__INST_UTCL1_LFIFO_DSM_IRRITATOR_DATA__SHIFT 0x0 +#define SQC_DSM_CNTL__INST_UTCL1_LFIFO_ENABLE_SINGLE_WRITE__SHIFT 0x2 +#define SQC_DSM_CNTL__DATA_CU0_WRITE_DATA_BUF_DSM_IRRITATOR_DATA__SHIFT 0x3 +#define SQC_DSM_CNTL__DATA_CU0_WRITE_DATA_BUF_ENABLE_SINGLE_WRITE__SHIFT 0x5 +#define SQC_DSM_CNTL__DATA_CU0_UTCL1_LFIFO_DSM_IRRITATOR_DATA__SHIFT 0x6 +#define SQC_DSM_CNTL__DATA_CU0_UTCL1_LFIFO_ENABLE_SINGLE_WRITE__SHIFT 0x8 +#define SQC_DSM_CNTL__DATA_CU1_WRITE_DATA_BUF_DSM_IRRITATOR_DATA__SHIFT 0x9 +#define SQC_DSM_CNTL__DATA_CU1_WRITE_DATA_BUF_ENABLE_SINGLE_WRITE__SHIFT 0xb +#define SQC_DSM_CNTL__DATA_CU1_UTCL1_LFIFO_DSM_IRRITATOR_DATA__SHIFT 0xc +#define SQC_DSM_CNTL__DATA_CU1_UTCL1_LFIFO_ENABLE_SINGLE_WRITE__SHIFT 0xe +#define SQC_DSM_CNTL__DATA_CU2_WRITE_DATA_BUF_DSM_IRRITATOR_DATA__SHIFT 0xf +#define SQC_DSM_CNTL__DATA_CU2_WRITE_DATA_BUF_ENABLE_SINGLE_WRITE__SHIFT 0x11 +#define SQC_DSM_CNTL__DATA_CU2_UTCL1_LFIFO_DSM_IRRITATOR_DATA__SHIFT 0x12 +#define SQC_DSM_CNTL__DATA_CU2_UTCL1_LFIFO_ENABLE_SINGLE_WRITE__SHIFT 0x14 +#define SQC_DSM_CNTL__INST_UTCL1_LFIFO_DSM_IRRITATOR_DATA_MASK 0x00000003L +#define SQC_DSM_CNTL__INST_UTCL1_LFIFO_ENABLE_SINGLE_WRITE_MASK 0x00000004L +#define SQC_DSM_CNTL__DATA_CU0_WRITE_DATA_BUF_DSM_IRRITATOR_DATA_MASK 0x00000018L +#define SQC_DSM_CNTL__DATA_CU0_WRITE_DATA_BUF_ENABLE_SINGLE_WRITE_MASK 0x00000020L +#define SQC_DSM_CNTL__DATA_CU0_UTCL1_LFIFO_DSM_IRRITATOR_DATA_MASK 0x000000C0L +#define SQC_DSM_CNTL__DATA_CU0_UTCL1_LFIFO_ENABLE_SINGLE_WRITE_MASK 0x00000100L +#define SQC_DSM_CNTL__DATA_CU1_WRITE_DATA_BUF_DSM_IRRITATOR_DATA_MASK 0x00000600L +#define SQC_DSM_CNTL__DATA_CU1_WRITE_DATA_BUF_ENABLE_SINGLE_WRITE_MASK 0x00000800L +#define SQC_DSM_CNTL__DATA_CU1_UTCL1_LFIFO_DSM_IRRITATOR_DATA_MASK 0x00003000L +#define SQC_DSM_CNTL__DATA_CU1_UTCL1_LFIFO_ENABLE_SINGLE_WRITE_MASK 0x00004000L +#define SQC_DSM_CNTL__DATA_CU2_WRITE_DATA_BUF_DSM_IRRITATOR_DATA_MASK 0x00018000L +#define SQC_DSM_CNTL__DATA_CU2_WRITE_DATA_BUF_ENABLE_SINGLE_WRITE_MASK 0x00020000L +#define SQC_DSM_CNTL__DATA_CU2_UTCL1_LFIFO_DSM_IRRITATOR_DATA_MASK 0x000C0000L +#define SQC_DSM_CNTL__DATA_CU2_UTCL1_LFIFO_ENABLE_SINGLE_WRITE_MASK 0x00100000L +//SQC_DSM_CNTLA +#define SQC_DSM_CNTLA__INST_TAG_RAM_DSM_IRRITATOR_DATA__SHIFT 0x0 +#define SQC_DSM_CNTLA__INST_TAG_RAM_ENABLE_SINGLE_WRITE__SHIFT 0x2 +#define SQC_DSM_CNTLA__INST_UTCL1_MISS_FIFO_DSM_IRRITATOR_DATA__SHIFT 0x3 +#define SQC_DSM_CNTLA__INST_UTCL1_MISS_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x5 +#define SQC_DSM_CNTLA__INST_MISS_FIFO_DSM_IRRITATOR_DATA__SHIFT 0x6 +#define SQC_DSM_CNTLA__INST_MISS_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x8 +#define SQC_DSM_CNTLA__INST_BANK_RAM_DSM_IRRITATOR_DATA__SHIFT 0x9 +#define SQC_DSM_CNTLA__INST_BANK_RAM_ENABLE_SINGLE_WRITE__SHIFT 0xb +#define SQC_DSM_CNTLA__DATA_TAG_RAM_DSM_IRRITATOR_DATA__SHIFT 0xc +#define SQC_DSM_CNTLA__DATA_TAG_RAM_ENABLE_SINGLE_WRITE__SHIFT 0xe +#define SQC_DSM_CNTLA__DATA_HIT_FIFO_DSM_IRRITATOR_DATA__SHIFT 0xf +#define SQC_DSM_CNTLA__DATA_HIT_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x11 +#define SQC_DSM_CNTLA__DATA_MISS_FIFO_DSM_IRRITATOR_DATA__SHIFT 0x12 +#define SQC_DSM_CNTLA__DATA_MISS_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x14 +#define SQC_DSM_CNTLA__DATA_DIRTY_BIT_RAM_DSM_IRRITATOR_DATA__SHIFT 0x15 +#define SQC_DSM_CNTLA__DATA_DIRTY_BIT_RAM_ENABLE_SINGLE_WRITE__SHIFT 0x17 +#define SQC_DSM_CNTLA__DATA_BANK_RAM_DSM_IRRITATOR_DATA__SHIFT 0x18 +#define SQC_DSM_CNTLA__DATA_BANK_RAM_ENABLE_SINGLE_WRITE__SHIFT 0x1a +#define SQC_DSM_CNTLA__INST_TAG_RAM_DSM_IRRITATOR_DATA_MASK 0x00000003L +#define SQC_DSM_CNTLA__INST_TAG_RAM_ENABLE_SINGLE_WRITE_MASK 0x00000004L +#define SQC_DSM_CNTLA__INST_UTCL1_MISS_FIFO_DSM_IRRITATOR_DATA_MASK 0x00000018L +#define SQC_DSM_CNTLA__INST_UTCL1_MISS_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00000020L +#define SQC_DSM_CNTLA__INST_MISS_FIFO_DSM_IRRITATOR_DATA_MASK 0x000000C0L +#define SQC_DSM_CNTLA__INST_MISS_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00000100L +#define SQC_DSM_CNTLA__INST_BANK_RAM_DSM_IRRITATOR_DATA_MASK 0x00000600L +#define SQC_DSM_CNTLA__INST_BANK_RAM_ENABLE_SINGLE_WRITE_MASK 0x00000800L +#define SQC_DSM_CNTLA__DATA_TAG_RAM_DSM_IRRITATOR_DATA_MASK 0x00003000L +#define SQC_DSM_CNTLA__DATA_TAG_RAM_ENABLE_SINGLE_WRITE_MASK 0x00004000L +#define SQC_DSM_CNTLA__DATA_HIT_FIFO_DSM_IRRITATOR_DATA_MASK 0x00018000L +#define SQC_DSM_CNTLA__DATA_HIT_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00020000L +#define SQC_DSM_CNTLA__DATA_MISS_FIFO_DSM_IRRITATOR_DATA_MASK 0x000C0000L +#define SQC_DSM_CNTLA__DATA_MISS_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00100000L +#define SQC_DSM_CNTLA__DATA_DIRTY_BIT_RAM_DSM_IRRITATOR_DATA_MASK 0x00600000L +#define SQC_DSM_CNTLA__DATA_DIRTY_BIT_RAM_ENABLE_SINGLE_WRITE_MASK 0x00800000L +#define SQC_DSM_CNTLA__DATA_BANK_RAM_DSM_IRRITATOR_DATA_MASK 0x03000000L +#define SQC_DSM_CNTLA__DATA_BANK_RAM_ENABLE_SINGLE_WRITE_MASK 0x04000000L +//SQC_DSM_CNTLB +#define SQC_DSM_CNTLB__INST_TAG_RAM_DSM_IRRITATOR_DATA__SHIFT 0x0 +#define SQC_DSM_CNTLB__INST_TAG_RAM_ENABLE_SINGLE_WRITE__SHIFT 0x2 +#define SQC_DSM_CNTLB__INST_UTCL1_MISS_FIFO_DSM_IRRITATOR_DATA__SHIFT 0x3 +#define SQC_DSM_CNTLB__INST_UTCL1_MISS_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x5 +#define SQC_DSM_CNTLB__INST_MISS_FIFO_DSM_IRRITATOR_DATA__SHIFT 0x6 +#define SQC_DSM_CNTLB__INST_MISS_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x8 +#define SQC_DSM_CNTLB__INST_BANK_RAM_DSM_IRRITATOR_DATA__SHIFT 0x9 +#define SQC_DSM_CNTLB__INST_BANK_RAM_ENABLE_SINGLE_WRITE__SHIFT 0xb +#define SQC_DSM_CNTLB__DATA_TAG_RAM_DSM_IRRITATOR_DATA__SHIFT 0xc +#define SQC_DSM_CNTLB__DATA_TAG_RAM_ENABLE_SINGLE_WRITE__SHIFT 0xe +#define SQC_DSM_CNTLB__DATA_HIT_FIFO_DSM_IRRITATOR_DATA__SHIFT 0xf +#define SQC_DSM_CNTLB__DATA_HIT_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x11 +#define SQC_DSM_CNTLB__DATA_MISS_FIFO_DSM_IRRITATOR_DATA__SHIFT 0x12 +#define SQC_DSM_CNTLB__DATA_MISS_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x14 +#define SQC_DSM_CNTLB__DATA_DIRTY_BIT_RAM_DSM_IRRITATOR_DATA__SHIFT 0x15 +#define SQC_DSM_CNTLB__DATA_DIRTY_BIT_RAM_ENABLE_SINGLE_WRITE__SHIFT 0x17 +#define SQC_DSM_CNTLB__DATA_BANK_RAM_DSM_IRRITATOR_DATA__SHIFT 0x18 +#define SQC_DSM_CNTLB__DATA_BANK_RAM_ENABLE_SINGLE_WRITE__SHIFT 0x1a +#define SQC_DSM_CNTLB__INST_TAG_RAM_DSM_IRRITATOR_DATA_MASK 0x00000003L +#define SQC_DSM_CNTLB__INST_TAG_RAM_ENABLE_SINGLE_WRITE_MASK 0x00000004L +#define SQC_DSM_CNTLB__INST_UTCL1_MISS_FIFO_DSM_IRRITATOR_DATA_MASK 0x00000018L +#define SQC_DSM_CNTLB__INST_UTCL1_MISS_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00000020L +#define SQC_DSM_CNTLB__INST_MISS_FIFO_DSM_IRRITATOR_DATA_MASK 0x000000C0L +#define SQC_DSM_CNTLB__INST_MISS_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00000100L +#define SQC_DSM_CNTLB__INST_BANK_RAM_DSM_IRRITATOR_DATA_MASK 0x00000600L +#define SQC_DSM_CNTLB__INST_BANK_RAM_ENABLE_SINGLE_WRITE_MASK 0x00000800L +#define SQC_DSM_CNTLB__DATA_TAG_RAM_DSM_IRRITATOR_DATA_MASK 0x00003000L +#define SQC_DSM_CNTLB__DATA_TAG_RAM_ENABLE_SINGLE_WRITE_MASK 0x00004000L +#define SQC_DSM_CNTLB__DATA_HIT_FIFO_DSM_IRRITATOR_DATA_MASK 0x00018000L +#define SQC_DSM_CNTLB__DATA_HIT_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00020000L +#define SQC_DSM_CNTLB__DATA_MISS_FIFO_DSM_IRRITATOR_DATA_MASK 0x000C0000L +#define SQC_DSM_CNTLB__DATA_MISS_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00100000L +#define SQC_DSM_CNTLB__DATA_DIRTY_BIT_RAM_DSM_IRRITATOR_DATA_MASK 0x00600000L +#define SQC_DSM_CNTLB__DATA_DIRTY_BIT_RAM_ENABLE_SINGLE_WRITE_MASK 0x00800000L +#define SQC_DSM_CNTLB__DATA_BANK_RAM_DSM_IRRITATOR_DATA_MASK 0x03000000L +#define SQC_DSM_CNTLB__DATA_BANK_RAM_ENABLE_SINGLE_WRITE_MASK 0x04000000L +//SQC_DSM_CNTL2 +#define SQC_DSM_CNTL2__INST_UTCL1_LFIFO_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define SQC_DSM_CNTL2__INST_UTCL1_LFIFO_SELECT_INJECT_DELAY__SHIFT 0x2 +#define SQC_DSM_CNTL2__DATA_CU0_WRITE_DATA_BUF_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define SQC_DSM_CNTL2__DATA_CU0_WRITE_DATA_BUF_SELECT_INJECT_DELAY__SHIFT 0x5 +#define SQC_DSM_CNTL2__DATA_CU0_UTCL1_LFIFO_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define SQC_DSM_CNTL2__DATA_CU0_UTCL1_LFIFO_SELECT_INJECT_DELAY__SHIFT 0x8 +#define SQC_DSM_CNTL2__DATA_CU1_WRITE_DATA_BUF_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define SQC_DSM_CNTL2__DATA_CU1_WRITE_DATA_BUF_SELECT_INJECT_DELAY__SHIFT 0xb +#define SQC_DSM_CNTL2__DATA_CU1_UTCL1_LFIFO_ENABLE_ERROR_INJECT__SHIFT 0xc +#define SQC_DSM_CNTL2__DATA_CU1_UTCL1_LFIFO_SELECT_INJECT_DELAY__SHIFT 0xe +#define SQC_DSM_CNTL2__DATA_CU2_WRITE_DATA_BUF_ENABLE_ERROR_INJECT__SHIFT 0xf +#define SQC_DSM_CNTL2__DATA_CU2_WRITE_DATA_BUF_SELECT_INJECT_DELAY__SHIFT 0x11 +#define SQC_DSM_CNTL2__DATA_CU2_UTCL1_LFIFO_ENABLE_ERROR_INJECT__SHIFT 0x12 +#define SQC_DSM_CNTL2__DATA_CU2_UTCL1_LFIFO_SELECT_INJECT_DELAY__SHIFT 0x14 +#define SQC_DSM_CNTL2__INJECT_DELAY__SHIFT 0x1a +#define SQC_DSM_CNTL2__INST_UTCL1_LFIFO_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define SQC_DSM_CNTL2__INST_UTCL1_LFIFO_SELECT_INJECT_DELAY_MASK 0x00000004L +#define SQC_DSM_CNTL2__DATA_CU0_WRITE_DATA_BUF_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define SQC_DSM_CNTL2__DATA_CU0_WRITE_DATA_BUF_SELECT_INJECT_DELAY_MASK 0x00000020L +#define SQC_DSM_CNTL2__DATA_CU0_UTCL1_LFIFO_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define SQC_DSM_CNTL2__DATA_CU0_UTCL1_LFIFO_SELECT_INJECT_DELAY_MASK 0x00000100L +#define SQC_DSM_CNTL2__DATA_CU1_WRITE_DATA_BUF_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define SQC_DSM_CNTL2__DATA_CU1_WRITE_DATA_BUF_SELECT_INJECT_DELAY_MASK 0x00000800L +#define SQC_DSM_CNTL2__DATA_CU1_UTCL1_LFIFO_ENABLE_ERROR_INJECT_MASK 0x00003000L +#define SQC_DSM_CNTL2__DATA_CU1_UTCL1_LFIFO_SELECT_INJECT_DELAY_MASK 0x00004000L +#define SQC_DSM_CNTL2__DATA_CU2_WRITE_DATA_BUF_ENABLE_ERROR_INJECT_MASK 0x00018000L +#define SQC_DSM_CNTL2__DATA_CU2_WRITE_DATA_BUF_SELECT_INJECT_DELAY_MASK 0x00020000L +#define SQC_DSM_CNTL2__DATA_CU2_UTCL1_LFIFO_ENABLE_ERROR_INJECT_MASK 0x000C0000L +#define SQC_DSM_CNTL2__DATA_CU2_UTCL1_LFIFO_SELECT_INJECT_DELAY_MASK 0x00100000L +#define SQC_DSM_CNTL2__INJECT_DELAY_MASK 0xFC000000L +//SQC_DSM_CNTL2A +#define SQC_DSM_CNTL2A__INST_TAG_RAM_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define SQC_DSM_CNTL2A__INST_TAG_RAM_SELECT_INJECT_DELAY__SHIFT 0x2 +#define SQC_DSM_CNTL2A__INST_UTCL1_MISS_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define SQC_DSM_CNTL2A__INST_UTCL1_MISS_FIFO_SELECT_INJECT_DELAY__SHIFT 0x5 +#define SQC_DSM_CNTL2A__INST_MISS_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define SQC_DSM_CNTL2A__INST_MISS_FIFO_SELECT_INJECT_DELAY__SHIFT 0x8 +#define SQC_DSM_CNTL2A__INST_BANK_RAM_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define SQC_DSM_CNTL2A__INST_BANK_RAM_SELECT_INJECT_DELAY__SHIFT 0xb +#define SQC_DSM_CNTL2A__DATA_TAG_RAM_ENABLE_ERROR_INJECT__SHIFT 0xc +#define SQC_DSM_CNTL2A__DATA_TAG_RAM_SELECT_INJECT_DELAY__SHIFT 0xe +#define SQC_DSM_CNTL2A__DATA_HIT_FIFO_ENABLE_ERROR_INJECT__SHIFT 0xf +#define SQC_DSM_CNTL2A__DATA_HIT_FIFO_SELECT_INJECT_DELAY__SHIFT 0x11 +#define SQC_DSM_CNTL2A__DATA_MISS_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x12 +#define SQC_DSM_CNTL2A__DATA_MISS_FIFO_SELECT_INJECT_DELAY__SHIFT 0x14 +#define SQC_DSM_CNTL2A__DATA_DIRTY_BIT_RAM_ENABLE_ERROR_INJECT__SHIFT 0x15 +#define SQC_DSM_CNTL2A__DATA_DIRTY_BIT_RAM_SELECT_INJECT_DELAY__SHIFT 0x17 +#define SQC_DSM_CNTL2A__DATA_BANK_RAM_ENABLE_ERROR_INJECT__SHIFT 0x18 +#define SQC_DSM_CNTL2A__DATA_BANK_RAM_SELECT_INJECT_DELAY__SHIFT 0x1a +#define SQC_DSM_CNTL2A__INST_TAG_RAM_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define SQC_DSM_CNTL2A__INST_TAG_RAM_SELECT_INJECT_DELAY_MASK 0x00000004L +#define SQC_DSM_CNTL2A__INST_UTCL1_MISS_FIFO_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define SQC_DSM_CNTL2A__INST_UTCL1_MISS_FIFO_SELECT_INJECT_DELAY_MASK 0x00000020L +#define SQC_DSM_CNTL2A__INST_MISS_FIFO_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define SQC_DSM_CNTL2A__INST_MISS_FIFO_SELECT_INJECT_DELAY_MASK 0x00000100L +#define SQC_DSM_CNTL2A__INST_BANK_RAM_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define SQC_DSM_CNTL2A__INST_BANK_RAM_SELECT_INJECT_DELAY_MASK 0x00000800L +#define SQC_DSM_CNTL2A__DATA_TAG_RAM_ENABLE_ERROR_INJECT_MASK 0x00003000L +#define SQC_DSM_CNTL2A__DATA_TAG_RAM_SELECT_INJECT_DELAY_MASK 0x00004000L +#define SQC_DSM_CNTL2A__DATA_HIT_FIFO_ENABLE_ERROR_INJECT_MASK 0x00018000L +#define SQC_DSM_CNTL2A__DATA_HIT_FIFO_SELECT_INJECT_DELAY_MASK 0x00020000L +#define SQC_DSM_CNTL2A__DATA_MISS_FIFO_ENABLE_ERROR_INJECT_MASK 0x000C0000L +#define SQC_DSM_CNTL2A__DATA_MISS_FIFO_SELECT_INJECT_DELAY_MASK 0x00100000L +#define SQC_DSM_CNTL2A__DATA_DIRTY_BIT_RAM_ENABLE_ERROR_INJECT_MASK 0x00600000L +#define SQC_DSM_CNTL2A__DATA_DIRTY_BIT_RAM_SELECT_INJECT_DELAY_MASK 0x00800000L +#define SQC_DSM_CNTL2A__DATA_BANK_RAM_ENABLE_ERROR_INJECT_MASK 0x03000000L +#define SQC_DSM_CNTL2A__DATA_BANK_RAM_SELECT_INJECT_DELAY_MASK 0x04000000L +//SQC_DSM_CNTL2B +#define SQC_DSM_CNTL2B__INST_TAG_RAM_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define SQC_DSM_CNTL2B__INST_TAG_RAM_SELECT_INJECT_DELAY__SHIFT 0x2 +#define SQC_DSM_CNTL2B__INST_UTCL1_MISS_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define SQC_DSM_CNTL2B__INST_UTCL1_MISS_FIFO_SELECT_INJECT_DELAY__SHIFT 0x5 +#define SQC_DSM_CNTL2B__INST_MISS_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define SQC_DSM_CNTL2B__INST_MISS_FIFO_SELECT_INJECT_DELAY__SHIFT 0x8 +#define SQC_DSM_CNTL2B__INST_BANK_RAM_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define SQC_DSM_CNTL2B__INST_BANK_RAM_SELECT_INJECT_DELAY__SHIFT 0xb +#define SQC_DSM_CNTL2B__DATA_TAG_RAM_ENABLE_ERROR_INJECT__SHIFT 0xc +#define SQC_DSM_CNTL2B__DATA_TAG_RAM_SELECT_INJECT_DELAY__SHIFT 0xe +#define SQC_DSM_CNTL2B__DATA_HIT_FIFO_ENABLE_ERROR_INJECT__SHIFT 0xf +#define SQC_DSM_CNTL2B__DATA_HIT_FIFO_SELECT_INJECT_DELAY__SHIFT 0x11 +#define SQC_DSM_CNTL2B__DATA_MISS_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x12 +#define SQC_DSM_CNTL2B__DATA_MISS_FIFO_SELECT_INJECT_DELAY__SHIFT 0x14 +#define SQC_DSM_CNTL2B__DATA_DIRTY_BIT_RAM_ENABLE_ERROR_INJECT__SHIFT 0x15 +#define SQC_DSM_CNTL2B__DATA_DIRTY_BIT_RAM_SELECT_INJECT_DELAY__SHIFT 0x17 +#define SQC_DSM_CNTL2B__DATA_BANK_RAM_ENABLE_ERROR_INJECT__SHIFT 0x18 +#define SQC_DSM_CNTL2B__DATA_BANK_RAM_SELECT_INJECT_DELAY__SHIFT 0x1a +#define SQC_DSM_CNTL2B__INST_TAG_RAM_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define SQC_DSM_CNTL2B__INST_TAG_RAM_SELECT_INJECT_DELAY_MASK 0x00000004L +#define SQC_DSM_CNTL2B__INST_UTCL1_MISS_FIFO_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define SQC_DSM_CNTL2B__INST_UTCL1_MISS_FIFO_SELECT_INJECT_DELAY_MASK 0x00000020L +#define SQC_DSM_CNTL2B__INST_MISS_FIFO_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define SQC_DSM_CNTL2B__INST_MISS_FIFO_SELECT_INJECT_DELAY_MASK 0x00000100L +#define SQC_DSM_CNTL2B__INST_BANK_RAM_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define SQC_DSM_CNTL2B__INST_BANK_RAM_SELECT_INJECT_DELAY_MASK 0x00000800L +#define SQC_DSM_CNTL2B__DATA_TAG_RAM_ENABLE_ERROR_INJECT_MASK 0x00003000L +#define SQC_DSM_CNTL2B__DATA_TAG_RAM_SELECT_INJECT_DELAY_MASK 0x00004000L +#define SQC_DSM_CNTL2B__DATA_HIT_FIFO_ENABLE_ERROR_INJECT_MASK 0x00018000L +#define SQC_DSM_CNTL2B__DATA_HIT_FIFO_SELECT_INJECT_DELAY_MASK 0x00020000L +#define SQC_DSM_CNTL2B__DATA_MISS_FIFO_ENABLE_ERROR_INJECT_MASK 0x000C0000L +#define SQC_DSM_CNTL2B__DATA_MISS_FIFO_SELECT_INJECT_DELAY_MASK 0x00100000L +#define SQC_DSM_CNTL2B__DATA_DIRTY_BIT_RAM_ENABLE_ERROR_INJECT_MASK 0x00600000L +#define SQC_DSM_CNTL2B__DATA_DIRTY_BIT_RAM_SELECT_INJECT_DELAY_MASK 0x00800000L +#define SQC_DSM_CNTL2B__DATA_BANK_RAM_ENABLE_ERROR_INJECT_MASK 0x03000000L +#define SQC_DSM_CNTL2B__DATA_BANK_RAM_SELECT_INJECT_DELAY_MASK 0x04000000L +//SQC_EDC_FUE_CNTL +#define SQC_EDC_FUE_CNTL__BLOCK_FUE_FLAGS__SHIFT 0x0 +#define SQC_EDC_FUE_CNTL__FUE_INTERRUPT_ENABLES__SHIFT 0x10 +#define SQC_EDC_FUE_CNTL__BLOCK_FUE_FLAGS_MASK 0x0000FFFFL +#define SQC_EDC_FUE_CNTL__FUE_INTERRUPT_ENABLES_MASK 0xFFFF0000L +//SQC_EDC_CNT2 +#define SQC_EDC_CNT2__INST_BANKA_TAG_RAM_SEC_COUNT__SHIFT 0x0 +#define SQC_EDC_CNT2__INST_BANKA_TAG_RAM_DED_COUNT__SHIFT 0x2 +#define SQC_EDC_CNT2__INST_BANKA_BANK_RAM_SEC_COUNT__SHIFT 0x4 +#define SQC_EDC_CNT2__INST_BANKA_BANK_RAM_DED_COUNT__SHIFT 0x6 +#define SQC_EDC_CNT2__DATA_BANKA_TAG_RAM_SEC_COUNT__SHIFT 0x8 +#define SQC_EDC_CNT2__DATA_BANKA_TAG_RAM_DED_COUNT__SHIFT 0xa +#define SQC_EDC_CNT2__DATA_BANKA_BANK_RAM_SEC_COUNT__SHIFT 0xc +#define SQC_EDC_CNT2__DATA_BANKA_BANK_RAM_DED_COUNT__SHIFT 0xe +#define SQC_EDC_CNT2__INST_BANKA_UTCL1_MISS_FIFO_SED_COUNT__SHIFT 0x10 +#define SQC_EDC_CNT2__INST_BANKA_MISS_FIFO_SED_COUNT__SHIFT 0x12 +#define SQC_EDC_CNT2__DATA_BANKA_HIT_FIFO_SED_COUNT__SHIFT 0x14 +#define SQC_EDC_CNT2__DATA_BANKA_MISS_FIFO_SED_COUNT__SHIFT 0x16 +#define SQC_EDC_CNT2__DATA_BANKA_DIRTY_BIT_RAM_SED_COUNT__SHIFT 0x18 +#define SQC_EDC_CNT2__INST_UTCL1_LFIFO_SEC_COUNT__SHIFT 0x1a +#define SQC_EDC_CNT2__INST_UTCL1_LFIFO_DED_COUNT__SHIFT 0x1c +#define SQC_EDC_CNT2__INST_BANKA_TAG_RAM_SEC_COUNT_MASK 0x00000003L +#define SQC_EDC_CNT2__INST_BANKA_TAG_RAM_DED_COUNT_MASK 0x0000000CL +#define SQC_EDC_CNT2__INST_BANKA_BANK_RAM_SEC_COUNT_MASK 0x00000030L +#define SQC_EDC_CNT2__INST_BANKA_BANK_RAM_DED_COUNT_MASK 0x000000C0L +#define SQC_EDC_CNT2__DATA_BANKA_TAG_RAM_SEC_COUNT_MASK 0x00000300L +#define SQC_EDC_CNT2__DATA_BANKA_TAG_RAM_DED_COUNT_MASK 0x00000C00L +#define SQC_EDC_CNT2__DATA_BANKA_BANK_RAM_SEC_COUNT_MASK 0x00003000L +#define SQC_EDC_CNT2__DATA_BANKA_BANK_RAM_DED_COUNT_MASK 0x0000C000L +#define SQC_EDC_CNT2__INST_BANKA_UTCL1_MISS_FIFO_SED_COUNT_MASK 0x00030000L +#define SQC_EDC_CNT2__INST_BANKA_MISS_FIFO_SED_COUNT_MASK 0x000C0000L +#define SQC_EDC_CNT2__DATA_BANKA_HIT_FIFO_SED_COUNT_MASK 0x00300000L +#define SQC_EDC_CNT2__DATA_BANKA_MISS_FIFO_SED_COUNT_MASK 0x00C00000L +#define SQC_EDC_CNT2__DATA_BANKA_DIRTY_BIT_RAM_SED_COUNT_MASK 0x03000000L +#define SQC_EDC_CNT2__INST_UTCL1_LFIFO_SEC_COUNT_MASK 0x0C000000L +#define SQC_EDC_CNT2__INST_UTCL1_LFIFO_DED_COUNT_MASK 0x30000000L +//SQC_EDC_CNT3 +#define SQC_EDC_CNT3__INST_BANKB_TAG_RAM_SEC_COUNT__SHIFT 0x0 +#define SQC_EDC_CNT3__INST_BANKB_TAG_RAM_DED_COUNT__SHIFT 0x2 +#define SQC_EDC_CNT3__INST_BANKB_BANK_RAM_SEC_COUNT__SHIFT 0x4 +#define SQC_EDC_CNT3__INST_BANKB_BANK_RAM_DED_COUNT__SHIFT 0x6 +#define SQC_EDC_CNT3__DATA_BANKB_TAG_RAM_SEC_COUNT__SHIFT 0x8 +#define SQC_EDC_CNT3__DATA_BANKB_TAG_RAM_DED_COUNT__SHIFT 0xa +#define SQC_EDC_CNT3__DATA_BANKB_BANK_RAM_SEC_COUNT__SHIFT 0xc +#define SQC_EDC_CNT3__DATA_BANKB_BANK_RAM_DED_COUNT__SHIFT 0xe +#define SQC_EDC_CNT3__INST_BANKB_UTCL1_MISS_FIFO_SED_COUNT__SHIFT 0x10 +#define SQC_EDC_CNT3__INST_BANKB_MISS_FIFO_SED_COUNT__SHIFT 0x12 +#define SQC_EDC_CNT3__DATA_BANKB_HIT_FIFO_SED_COUNT__SHIFT 0x14 +#define SQC_EDC_CNT3__DATA_BANKB_MISS_FIFO_SED_COUNT__SHIFT 0x16 +#define SQC_EDC_CNT3__DATA_BANKB_DIRTY_BIT_RAM_SED_COUNT__SHIFT 0x18 +#define SQC_EDC_CNT3__INST_BANKB_TAG_RAM_SEC_COUNT_MASK 0x00000003L +#define SQC_EDC_CNT3__INST_BANKB_TAG_RAM_DED_COUNT_MASK 0x0000000CL +#define SQC_EDC_CNT3__INST_BANKB_BANK_RAM_SEC_COUNT_MASK 0x00000030L +#define SQC_EDC_CNT3__INST_BANKB_BANK_RAM_DED_COUNT_MASK 0x000000C0L +#define SQC_EDC_CNT3__DATA_BANKB_TAG_RAM_SEC_COUNT_MASK 0x00000300L +#define SQC_EDC_CNT3__DATA_BANKB_TAG_RAM_DED_COUNT_MASK 0x00000C00L +#define SQC_EDC_CNT3__DATA_BANKB_BANK_RAM_SEC_COUNT_MASK 0x00003000L +#define SQC_EDC_CNT3__DATA_BANKB_BANK_RAM_DED_COUNT_MASK 0x0000C000L +#define SQC_EDC_CNT3__INST_BANKB_UTCL1_MISS_FIFO_SED_COUNT_MASK 0x00030000L +#define SQC_EDC_CNT3__INST_BANKB_MISS_FIFO_SED_COUNT_MASK 0x000C0000L +#define SQC_EDC_CNT3__DATA_BANKB_HIT_FIFO_SED_COUNT_MASK 0x00300000L +#define SQC_EDC_CNT3__DATA_BANKB_MISS_FIFO_SED_COUNT_MASK 0x00C00000L +#define SQC_EDC_CNT3__DATA_BANKB_DIRTY_BIT_RAM_SED_COUNT_MASK 0x03000000L +//SQ_REG_TIMESTAMP +#define SQ_REG_TIMESTAMP__TIMESTAMP__SHIFT 0x0 +#define SQ_REG_TIMESTAMP__TIMESTAMP_MASK 0x000000FFL +//SQ_CMD_TIMESTAMP +#define SQ_CMD_TIMESTAMP__TIMESTAMP__SHIFT 0x0 +#define SQ_CMD_TIMESTAMP__TIMESTAMP_MASK 0x000000FFL +//SQ_IND_INDEX +#define SQ_IND_INDEX__WAVE_ID__SHIFT 0x0 +#define SQ_IND_INDEX__SIMD_ID__SHIFT 0x4 +#define SQ_IND_INDEX__THREAD_ID__SHIFT 0x6 +#define SQ_IND_INDEX__AUTO_INCR__SHIFT 0xc +#define SQ_IND_INDEX__FORCE_READ__SHIFT 0xd +#define SQ_IND_INDEX__READ_TIMEOUT__SHIFT 0xe +#define SQ_IND_INDEX__UNINDEXED__SHIFT 0xf +#define SQ_IND_INDEX__INDEX__SHIFT 0x10 +#define SQ_IND_INDEX__WAVE_ID_MASK 0x0000000FL +#define SQ_IND_INDEX__SIMD_ID_MASK 0x00000030L +#define SQ_IND_INDEX__THREAD_ID_MASK 0x00000FC0L +#define SQ_IND_INDEX__AUTO_INCR_MASK 0x00001000L +#define SQ_IND_INDEX__FORCE_READ_MASK 0x00002000L +#define SQ_IND_INDEX__READ_TIMEOUT_MASK 0x00004000L +#define SQ_IND_INDEX__UNINDEXED_MASK 0x00008000L +#define SQ_IND_INDEX__INDEX_MASK 0xFFFF0000L +//SQ_IND_DATA +#define SQ_IND_DATA__DATA__SHIFT 0x0 +#define SQ_IND_DATA__DATA_MASK 0xFFFFFFFFL +//SQ_CMD +#define SQ_CMD__CMD__SHIFT 0x0 +#define SQ_CMD__MODE__SHIFT 0x4 +#define SQ_CMD__CHECK_VMID__SHIFT 0x7 +#define SQ_CMD__DATA__SHIFT 0x8 +#define SQ_CMD__WAVE_ID__SHIFT 0x10 +#define SQ_CMD__SIMD_ID__SHIFT 0x14 +#define SQ_CMD__QUEUE_ID__SHIFT 0x18 +#define SQ_CMD__VM_ID__SHIFT 0x1c +#define SQ_CMD__CMD_MASK 0x00000007L +#define SQ_CMD__MODE_MASK 0x00000070L +#define SQ_CMD__CHECK_VMID_MASK 0x00000080L +#define SQ_CMD__DATA_MASK 0x00000F00L +#define SQ_CMD__WAVE_ID_MASK 0x000F0000L +#define SQ_CMD__SIMD_ID_MASK 0x00300000L +#define SQ_CMD__QUEUE_ID_MASK 0x07000000L +#define SQ_CMD__VM_ID_MASK 0xF0000000L +//SQ_TIME_HI +#define SQ_TIME_HI__TIME__SHIFT 0x0 +#define SQ_TIME_HI__TIME_MASK 0xFFFFFFFFL +//SQ_TIME_LO +#define SQ_TIME_LO__TIME__SHIFT 0x0 +#define SQ_TIME_LO__TIME_MASK 0xFFFFFFFFL +//SQ_DS_0 +#define SQ_DS_0__OFFSET0__SHIFT 0x0 +#define SQ_DS_0__OFFSET1__SHIFT 0x8 +#define SQ_DS_0__GDS__SHIFT 0x10 +#define SQ_DS_0__OP__SHIFT 0x11 +#define SQ_DS_0__ENCODING__SHIFT 0x1a +#define SQ_DS_0__OFFSET0_MASK 0x000000FFL +#define SQ_DS_0__OFFSET1_MASK 0x0000FF00L +#define SQ_DS_0__GDS_MASK 0x00010000L +#define SQ_DS_0__OP_MASK 0x01FE0000L +#define SQ_DS_0__ENCODING_MASK 0xFC000000L +//SQ_DS_1 +#define SQ_DS_1__ADDR__SHIFT 0x0 +#define SQ_DS_1__DATA0__SHIFT 0x8 +#define SQ_DS_1__DATA1__SHIFT 0x10 +#define SQ_DS_1__VDST__SHIFT 0x18 +#define SQ_DS_1__ADDR_MASK 0x000000FFL +#define SQ_DS_1__DATA0_MASK 0x0000FF00L +#define SQ_DS_1__DATA1_MASK 0x00FF0000L +#define SQ_DS_1__VDST_MASK 0xFF000000L +//SQ_EXP_0 +#define SQ_EXP_0__EN__SHIFT 0x0 +#define SQ_EXP_0__TGT__SHIFT 0x4 +#define SQ_EXP_0__COMPR__SHIFT 0xa +#define SQ_EXP_0__DONE__SHIFT 0xb +#define SQ_EXP_0__VM__SHIFT 0xc +#define SQ_EXP_0__ENCODING__SHIFT 0x1a +#define SQ_EXP_0__EN_MASK 0x0000000FL +#define SQ_EXP_0__TGT_MASK 0x000003F0L +#define SQ_EXP_0__COMPR_MASK 0x00000400L +#define SQ_EXP_0__DONE_MASK 0x00000800L +#define SQ_EXP_0__VM_MASK 0x00001000L +#define SQ_EXP_0__ENCODING_MASK 0xFC000000L +//SQ_EXP_1 +#define SQ_EXP_1__VSRC0__SHIFT 0x0 +#define SQ_EXP_1__VSRC1__SHIFT 0x8 +#define SQ_EXP_1__VSRC2__SHIFT 0x10 +#define SQ_EXP_1__VSRC3__SHIFT 0x18 +#define SQ_EXP_1__VSRC0_MASK 0x000000FFL +#define SQ_EXP_1__VSRC1_MASK 0x0000FF00L +#define SQ_EXP_1__VSRC2_MASK 0x00FF0000L +#define SQ_EXP_1__VSRC3_MASK 0xFF000000L +//SQ_FLAT_0 +#define SQ_FLAT_0__OFFSET__SHIFT 0x0 +#define SQ_FLAT_0__LDS__SHIFT 0xd +#define SQ_FLAT_0__SEG__SHIFT 0xe +#define SQ_FLAT_0__GLC__SHIFT 0x10 +#define SQ_FLAT_0__SLC__SHIFT 0x11 +#define SQ_FLAT_0__OP__SHIFT 0x12 +#define SQ_FLAT_0__ENCODING__SHIFT 0x1a +#define SQ_FLAT_0__OFFSET_MASK 0x00000FFFL +#define SQ_FLAT_0__LDS_MASK 0x00002000L +#define SQ_FLAT_0__SEG_MASK 0x0000C000L +#define SQ_FLAT_0__GLC_MASK 0x00010000L +#define SQ_FLAT_0__SLC_MASK 0x00020000L +#define SQ_FLAT_0__OP_MASK 0x01FC0000L +#define SQ_FLAT_0__ENCODING_MASK 0xFC000000L +//SQ_FLAT_1 +#define SQ_FLAT_1__ADDR__SHIFT 0x0 +#define SQ_FLAT_1__DATA__SHIFT 0x8 +#define SQ_FLAT_1__SADDR__SHIFT 0x10 +#define SQ_FLAT_1__NV__SHIFT 0x17 +#define SQ_FLAT_1__VDST__SHIFT 0x18 +#define SQ_FLAT_1__ADDR_MASK 0x000000FFL +#define SQ_FLAT_1__DATA_MASK 0x0000FF00L +#define SQ_FLAT_1__SADDR_MASK 0x007F0000L +#define SQ_FLAT_1__NV_MASK 0x00800000L +#define SQ_FLAT_1__VDST_MASK 0xFF000000L +//SQ_GLBL_0 +#define SQ_GLBL_0__OFFSET__SHIFT 0x0 +#define SQ_GLBL_0__LDS__SHIFT 0xd +#define SQ_GLBL_0__SEG__SHIFT 0xe +#define SQ_GLBL_0__GLC__SHIFT 0x10 +#define SQ_GLBL_0__SLC__SHIFT 0x11 +#define SQ_GLBL_0__OP__SHIFT 0x12 +#define SQ_GLBL_0__ENCODING__SHIFT 0x1a +#define SQ_GLBL_0__OFFSET_MASK 0x00001FFFL +#define SQ_GLBL_0__LDS_MASK 0x00002000L +#define SQ_GLBL_0__SEG_MASK 0x0000C000L +#define SQ_GLBL_0__GLC_MASK 0x00010000L +#define SQ_GLBL_0__SLC_MASK 0x00020000L +#define SQ_GLBL_0__OP_MASK 0x01FC0000L +#define SQ_GLBL_0__ENCODING_MASK 0xFC000000L +//SQ_GLBL_1 +#define SQ_GLBL_1__ADDR__SHIFT 0x0 +#define SQ_GLBL_1__DATA__SHIFT 0x8 +#define SQ_GLBL_1__SADDR__SHIFT 0x10 +#define SQ_GLBL_1__NV__SHIFT 0x17 +#define SQ_GLBL_1__VDST__SHIFT 0x18 +#define SQ_GLBL_1__ADDR_MASK 0x000000FFL +#define SQ_GLBL_1__DATA_MASK 0x0000FF00L +#define SQ_GLBL_1__SADDR_MASK 0x007F0000L +#define SQ_GLBL_1__NV_MASK 0x00800000L +#define SQ_GLBL_1__VDST_MASK 0xFF000000L +//SQ_INST +#define SQ_INST__ENCODING__SHIFT 0x0 +#define SQ_INST__ENCODING_MASK 0xFFFFFFFFL +//SQ_MIMG_0 +#define SQ_MIMG_0__OPM__SHIFT 0x0 +#define SQ_MIMG_0__DMASK__SHIFT 0x8 +#define SQ_MIMG_0__UNORM__SHIFT 0xc +#define SQ_MIMG_0__GLC__SHIFT 0xd +#define SQ_MIMG_0__DA__SHIFT 0xe +#define SQ_MIMG_0__A16__SHIFT 0xf +#define SQ_MIMG_0__TFE__SHIFT 0x10 +#define SQ_MIMG_0__LWE__SHIFT 0x11 +#define SQ_MIMG_0__OP__SHIFT 0x12 +#define SQ_MIMG_0__SLC__SHIFT 0x19 +#define SQ_MIMG_0__ENCODING__SHIFT 0x1a +#define SQ_MIMG_0__OPM_MASK 0x00000001L +#define SQ_MIMG_0__DMASK_MASK 0x00000F00L +#define SQ_MIMG_0__UNORM_MASK 0x00001000L +#define SQ_MIMG_0__GLC_MASK 0x00002000L +#define SQ_MIMG_0__DA_MASK 0x00004000L +#define SQ_MIMG_0__A16_MASK 0x00008000L +#define SQ_MIMG_0__TFE_MASK 0x00010000L +#define SQ_MIMG_0__LWE_MASK 0x00020000L +#define SQ_MIMG_0__OP_MASK 0x01FC0000L +#define SQ_MIMG_0__SLC_MASK 0x02000000L +#define SQ_MIMG_0__ENCODING_MASK 0xFC000000L +//SQ_MIMG_1 +#define SQ_MIMG_1__VADDR__SHIFT 0x0 +#define SQ_MIMG_1__VDATA__SHIFT 0x8 +#define SQ_MIMG_1__SRSRC__SHIFT 0x10 +#define SQ_MIMG_1__SSAMP__SHIFT 0x15 +#define SQ_MIMG_1__D16__SHIFT 0x1f +#define SQ_MIMG_1__VADDR_MASK 0x000000FFL +#define SQ_MIMG_1__VDATA_MASK 0x0000FF00L +#define SQ_MIMG_1__SRSRC_MASK 0x001F0000L +#define SQ_MIMG_1__SSAMP_MASK 0x03E00000L +#define SQ_MIMG_1__D16_MASK 0x80000000L +//SQ_MTBUF_0 +#define SQ_MTBUF_0__OFFSET__SHIFT 0x0 +#define SQ_MTBUF_0__OFFEN__SHIFT 0xc +#define SQ_MTBUF_0__IDXEN__SHIFT 0xd +#define SQ_MTBUF_0__GLC__SHIFT 0xe +#define SQ_MTBUF_0__OP__SHIFT 0xf +#define SQ_MTBUF_0__DFMT__SHIFT 0x13 +#define SQ_MTBUF_0__NFMT__SHIFT 0x17 +#define SQ_MTBUF_0__ENCODING__SHIFT 0x1a +#define SQ_MTBUF_0__OFFSET_MASK 0x00000FFFL +#define SQ_MTBUF_0__OFFEN_MASK 0x00001000L +#define SQ_MTBUF_0__IDXEN_MASK 0x00002000L +#define SQ_MTBUF_0__GLC_MASK 0x00004000L +#define SQ_MTBUF_0__OP_MASK 0x00078000L +#define SQ_MTBUF_0__DFMT_MASK 0x00780000L +#define SQ_MTBUF_0__NFMT_MASK 0x03800000L +#define SQ_MTBUF_0__ENCODING_MASK 0xFC000000L +//SQ_MTBUF_1 +#define SQ_MTBUF_1__VADDR__SHIFT 0x0 +#define SQ_MTBUF_1__VDATA__SHIFT 0x8 +#define SQ_MTBUF_1__SRSRC__SHIFT 0x10 +#define SQ_MTBUF_1__SLC__SHIFT 0x16 +#define SQ_MTBUF_1__TFE__SHIFT 0x17 +#define SQ_MTBUF_1__SOFFSET__SHIFT 0x18 +#define SQ_MTBUF_1__VADDR_MASK 0x000000FFL +#define SQ_MTBUF_1__VDATA_MASK 0x0000FF00L +#define SQ_MTBUF_1__SRSRC_MASK 0x001F0000L +#define SQ_MTBUF_1__SLC_MASK 0x00400000L +#define SQ_MTBUF_1__TFE_MASK 0x00800000L +#define SQ_MTBUF_1__SOFFSET_MASK 0xFF000000L +//SQ_MUBUF_0 +#define SQ_MUBUF_0__OFFSET__SHIFT 0x0 +#define SQ_MUBUF_0__OFFEN__SHIFT 0xc +#define SQ_MUBUF_0__IDXEN__SHIFT 0xd +#define SQ_MUBUF_0__GLC__SHIFT 0xe +#define SQ_MUBUF_0__LDS__SHIFT 0x10 +#define SQ_MUBUF_0__SLC__SHIFT 0x11 +#define SQ_MUBUF_0__OP__SHIFT 0x12 +#define SQ_MUBUF_0__ENCODING__SHIFT 0x1a +#define SQ_MUBUF_0__OFFSET_MASK 0x00000FFFL +#define SQ_MUBUF_0__OFFEN_MASK 0x00001000L +#define SQ_MUBUF_0__IDXEN_MASK 0x00002000L +#define SQ_MUBUF_0__GLC_MASK 0x00004000L +#define SQ_MUBUF_0__LDS_MASK 0x00010000L +#define SQ_MUBUF_0__SLC_MASK 0x00020000L +#define SQ_MUBUF_0__OP_MASK 0x01FC0000L +#define SQ_MUBUF_0__ENCODING_MASK 0xFC000000L +//SQ_MUBUF_1 +#define SQ_MUBUF_1__VADDR__SHIFT 0x0 +#define SQ_MUBUF_1__VDATA__SHIFT 0x8 +#define SQ_MUBUF_1__SRSRC__SHIFT 0x10 +#define SQ_MUBUF_1__TFE__SHIFT 0x17 +#define SQ_MUBUF_1__SOFFSET__SHIFT 0x18 +#define SQ_MUBUF_1__VADDR_MASK 0x000000FFL +#define SQ_MUBUF_1__VDATA_MASK 0x0000FF00L +#define SQ_MUBUF_1__SRSRC_MASK 0x001F0000L +#define SQ_MUBUF_1__TFE_MASK 0x00800000L +#define SQ_MUBUF_1__SOFFSET_MASK 0xFF000000L +//SQ_SCRATCH_0 +#define SQ_SCRATCH_0__OFFSET__SHIFT 0x0 +#define SQ_SCRATCH_0__LDS__SHIFT 0xd +#define SQ_SCRATCH_0__SEG__SHIFT 0xe +#define SQ_SCRATCH_0__GLC__SHIFT 0x10 +#define SQ_SCRATCH_0__SLC__SHIFT 0x11 +#define SQ_SCRATCH_0__OP__SHIFT 0x12 +#define SQ_SCRATCH_0__ENCODING__SHIFT 0x1a +#define SQ_SCRATCH_0__OFFSET_MASK 0x00001FFFL +#define SQ_SCRATCH_0__LDS_MASK 0x00002000L +#define SQ_SCRATCH_0__SEG_MASK 0x0000C000L +#define SQ_SCRATCH_0__GLC_MASK 0x00010000L +#define SQ_SCRATCH_0__SLC_MASK 0x00020000L +#define SQ_SCRATCH_0__OP_MASK 0x01FC0000L +#define SQ_SCRATCH_0__ENCODING_MASK 0xFC000000L +//SQ_SCRATCH_1 +#define SQ_SCRATCH_1__ADDR__SHIFT 0x0 +#define SQ_SCRATCH_1__DATA__SHIFT 0x8 +#define SQ_SCRATCH_1__SADDR__SHIFT 0x10 +#define SQ_SCRATCH_1__NV__SHIFT 0x17 +#define SQ_SCRATCH_1__VDST__SHIFT 0x18 +#define SQ_SCRATCH_1__ADDR_MASK 0x000000FFL +#define SQ_SCRATCH_1__DATA_MASK 0x0000FF00L +#define SQ_SCRATCH_1__SADDR_MASK 0x007F0000L +#define SQ_SCRATCH_1__NV_MASK 0x00800000L +#define SQ_SCRATCH_1__VDST_MASK 0xFF000000L +//SQ_SMEM_0 +#define SQ_SMEM_0__SBASE__SHIFT 0x0 +#define SQ_SMEM_0__SDATA__SHIFT 0x6 +#define SQ_SMEM_0__SOFFSET_EN__SHIFT 0xe +#define SQ_SMEM_0__NV__SHIFT 0xf +#define SQ_SMEM_0__GLC__SHIFT 0x10 +#define SQ_SMEM_0__IMM__SHIFT 0x11 +#define SQ_SMEM_0__OP__SHIFT 0x12 +#define SQ_SMEM_0__ENCODING__SHIFT 0x1a +#define SQ_SMEM_0__SBASE_MASK 0x0000003FL +#define SQ_SMEM_0__SDATA_MASK 0x00001FC0L +#define SQ_SMEM_0__SOFFSET_EN_MASK 0x00004000L +#define SQ_SMEM_0__NV_MASK 0x00008000L +#define SQ_SMEM_0__GLC_MASK 0x00010000L +#define SQ_SMEM_0__IMM_MASK 0x00020000L +#define SQ_SMEM_0__OP_MASK 0x03FC0000L +#define SQ_SMEM_0__ENCODING_MASK 0xFC000000L +//SQ_SMEM_1 +#define SQ_SMEM_1__OFFSET__SHIFT 0x0 +#define SQ_SMEM_1__SOFFSET__SHIFT 0x19 +#define SQ_SMEM_1__OFFSET_MASK 0x001FFFFFL +#define SQ_SMEM_1__SOFFSET_MASK 0xFE000000L +//SQ_SOP1 +#define SQ_SOP1__SSRC0__SHIFT 0x0 +#define SQ_SOP1__OP__SHIFT 0x8 +#define SQ_SOP1__SDST__SHIFT 0x10 +#define SQ_SOP1__ENCODING__SHIFT 0x17 +#define SQ_SOP1__SSRC0_MASK 0x000000FFL +#define SQ_SOP1__OP_MASK 0x0000FF00L +#define SQ_SOP1__SDST_MASK 0x007F0000L +#define SQ_SOP1__ENCODING_MASK 0xFF800000L +//SQ_SOP2 +#define SQ_SOP2__SSRC0__SHIFT 0x0 +#define SQ_SOP2__SSRC1__SHIFT 0x8 +#define SQ_SOP2__SDST__SHIFT 0x10 +#define SQ_SOP2__OP__SHIFT 0x17 +#define SQ_SOP2__ENCODING__SHIFT 0x1e +#define SQ_SOP2__SSRC0_MASK 0x000000FFL +#define SQ_SOP2__SSRC1_MASK 0x0000FF00L +#define SQ_SOP2__SDST_MASK 0x007F0000L +#define SQ_SOP2__OP_MASK 0x3F800000L +#define SQ_SOP2__ENCODING_MASK 0xC0000000L +//SQ_SOPC +#define SQ_SOPC__SSRC0__SHIFT 0x0 +#define SQ_SOPC__SSRC1__SHIFT 0x8 +#define SQ_SOPC__OP__SHIFT 0x10 +#define SQ_SOPC__ENCODING__SHIFT 0x17 +#define SQ_SOPC__SSRC0_MASK 0x000000FFL +#define SQ_SOPC__SSRC1_MASK 0x0000FF00L +#define SQ_SOPC__OP_MASK 0x007F0000L +#define SQ_SOPC__ENCODING_MASK 0xFF800000L +//SQ_SOPK +#define SQ_SOPK__SIMM16__SHIFT 0x0 +#define SQ_SOPK__SDST__SHIFT 0x10 +#define SQ_SOPK__OP__SHIFT 0x17 +#define SQ_SOPK__ENCODING__SHIFT 0x1c +#define SQ_SOPK__SIMM16_MASK 0x0000FFFFL +#define SQ_SOPK__SDST_MASK 0x007F0000L +#define SQ_SOPK__OP_MASK 0x0F800000L +#define SQ_SOPK__ENCODING_MASK 0xF0000000L +//SQ_SOPP +#define SQ_SOPP__SIMM16__SHIFT 0x0 +#define SQ_SOPP__OP__SHIFT 0x10 +#define SQ_SOPP__ENCODING__SHIFT 0x17 +#define SQ_SOPP__SIMM16_MASK 0x0000FFFFL +#define SQ_SOPP__OP_MASK 0x007F0000L +#define SQ_SOPP__ENCODING_MASK 0xFF800000L +//SQ_VINTRP +#define SQ_VINTRP__VSRC__SHIFT 0x0 +#define SQ_VINTRP__ATTRCHAN__SHIFT 0x8 +#define SQ_VINTRP__ATTR__SHIFT 0xa +#define SQ_VINTRP__OP__SHIFT 0x10 +#define SQ_VINTRP__VDST__SHIFT 0x12 +#define SQ_VINTRP__ENCODING__SHIFT 0x1a +#define SQ_VINTRP__VSRC_MASK 0x000000FFL +#define SQ_VINTRP__ATTRCHAN_MASK 0x00000300L +#define SQ_VINTRP__ATTR_MASK 0x0000FC00L +#define SQ_VINTRP__OP_MASK 0x00030000L +#define SQ_VINTRP__VDST_MASK 0x03FC0000L +#define SQ_VINTRP__ENCODING_MASK 0xFC000000L +//SQ_VOP1 +#define SQ_VOP1__SRC0__SHIFT 0x0 +#define SQ_VOP1__OP__SHIFT 0x9 +#define SQ_VOP1__VDST__SHIFT 0x11 +#define SQ_VOP1__ENCODING__SHIFT 0x19 +#define SQ_VOP1__SRC0_MASK 0x000001FFL +#define SQ_VOP1__OP_MASK 0x0001FE00L +#define SQ_VOP1__VDST_MASK 0x01FE0000L +#define SQ_VOP1__ENCODING_MASK 0xFE000000L +//SQ_VOP2 +#define SQ_VOP2__SRC0__SHIFT 0x0 +#define SQ_VOP2__VSRC1__SHIFT 0x9 +#define SQ_VOP2__VDST__SHIFT 0x11 +#define SQ_VOP2__OP__SHIFT 0x19 +#define SQ_VOP2__ENCODING__SHIFT 0x1f +#define SQ_VOP2__SRC0_MASK 0x000001FFL +#define SQ_VOP2__VSRC1_MASK 0x0001FE00L +#define SQ_VOP2__VDST_MASK 0x01FE0000L +#define SQ_VOP2__OP_MASK 0x7E000000L +#define SQ_VOP2__ENCODING_MASK 0x80000000L +//SQ_VOP3P_0 +#define SQ_VOP3P_0__VDST__SHIFT 0x0 +#define SQ_VOP3P_0__NEG_HI__SHIFT 0x8 +#define SQ_VOP3P_0__OP_SEL__SHIFT 0xb +#define SQ_VOP3P_0__OP_SEL_HI_2__SHIFT 0xe +#define SQ_VOP3P_0__CLAMP__SHIFT 0xf +#define SQ_VOP3P_0__OP__SHIFT 0x10 +#define SQ_VOP3P_0__ENCODING__SHIFT 0x17 +#define SQ_VOP3P_0__VDST_MASK 0x000000FFL +#define SQ_VOP3P_0__NEG_HI_MASK 0x00000700L +#define SQ_VOP3P_0__OP_SEL_MASK 0x00003800L +#define SQ_VOP3P_0__OP_SEL_HI_2_MASK 0x00004000L +#define SQ_VOP3P_0__CLAMP_MASK 0x00008000L +#define SQ_VOP3P_0__OP_MASK 0x007F0000L +#define SQ_VOP3P_0__ENCODING_MASK 0xFF800000L +//SQ_VOP3P_1 +#define SQ_VOP3P_1__SRC0__SHIFT 0x0 +#define SQ_VOP3P_1__SRC1__SHIFT 0x9 +#define SQ_VOP3P_1__SRC2__SHIFT 0x12 +#define SQ_VOP3P_1__OP_SEL_HI__SHIFT 0x1b +#define SQ_VOP3P_1__NEG__SHIFT 0x1d +#define SQ_VOP3P_1__SRC0_MASK 0x000001FFL +#define SQ_VOP3P_1__SRC1_MASK 0x0003FE00L +#define SQ_VOP3P_1__SRC2_MASK 0x07FC0000L +#define SQ_VOP3P_1__OP_SEL_HI_MASK 0x18000000L +#define SQ_VOP3P_1__NEG_MASK 0xE0000000L +//SQ_VOP3_0 +#define SQ_VOP3_0__VDST__SHIFT 0x0 +#define SQ_VOP3_0__ABS__SHIFT 0x8 +#define SQ_VOP3_0__OP_SEL__SHIFT 0xb +#define SQ_VOP3_0__CLAMP__SHIFT 0xf +#define SQ_VOP3_0__OP__SHIFT 0x10 +#define SQ_VOP3_0__ENCODING__SHIFT 0x1a +#define SQ_VOP3_0__VDST_MASK 0x000000FFL +#define SQ_VOP3_0__ABS_MASK 0x00000700L +#define SQ_VOP3_0__OP_SEL_MASK 0x00007800L +#define SQ_VOP3_0__CLAMP_MASK 0x00008000L +#define SQ_VOP3_0__OP_MASK 0x03FF0000L +#define SQ_VOP3_0__ENCODING_MASK 0xFC000000L +//SQ_VOP3_0_SDST_ENC +#define SQ_VOP3_0_SDST_ENC__VDST__SHIFT 0x0 +#define SQ_VOP3_0_SDST_ENC__SDST__SHIFT 0x8 +#define SQ_VOP3_0_SDST_ENC__CLAMP__SHIFT 0xf +#define SQ_VOP3_0_SDST_ENC__OP__SHIFT 0x10 +#define SQ_VOP3_0_SDST_ENC__ENCODING__SHIFT 0x1a +#define SQ_VOP3_0_SDST_ENC__VDST_MASK 0x000000FFL +#define SQ_VOP3_0_SDST_ENC__SDST_MASK 0x00007F00L +#define SQ_VOP3_0_SDST_ENC__CLAMP_MASK 0x00008000L +#define SQ_VOP3_0_SDST_ENC__OP_MASK 0x03FF0000L +#define SQ_VOP3_0_SDST_ENC__ENCODING_MASK 0xFC000000L +//SQ_VOP3_1 +#define SQ_VOP3_1__SRC0__SHIFT 0x0 +#define SQ_VOP3_1__SRC1__SHIFT 0x9 +#define SQ_VOP3_1__SRC2__SHIFT 0x12 +#define SQ_VOP3_1__OMOD__SHIFT 0x1b +#define SQ_VOP3_1__NEG__SHIFT 0x1d +#define SQ_VOP3_1__SRC0_MASK 0x000001FFL +#define SQ_VOP3_1__SRC1_MASK 0x0003FE00L +#define SQ_VOP3_1__SRC2_MASK 0x07FC0000L +#define SQ_VOP3_1__OMOD_MASK 0x18000000L +#define SQ_VOP3_1__NEG_MASK 0xE0000000L +//SQ_VOPC +#define SQ_VOPC__SRC0__SHIFT 0x0 +#define SQ_VOPC__VSRC1__SHIFT 0x9 +#define SQ_VOPC__OP__SHIFT 0x11 +#define SQ_VOPC__ENCODING__SHIFT 0x19 +#define SQ_VOPC__SRC0_MASK 0x000001FFL +#define SQ_VOPC__VSRC1_MASK 0x0001FE00L +#define SQ_VOPC__OP_MASK 0x01FE0000L +#define SQ_VOPC__ENCODING_MASK 0xFE000000L +//SQ_VOP_DPP +#define SQ_VOP_DPP__SRC0__SHIFT 0x0 +#define SQ_VOP_DPP__DPP_CTRL__SHIFT 0x8 +#define SQ_VOP_DPP__BOUND_CTRL__SHIFT 0x13 +#define SQ_VOP_DPP__SRC0_NEG__SHIFT 0x14 +#define SQ_VOP_DPP__SRC0_ABS__SHIFT 0x15 +#define SQ_VOP_DPP__SRC1_NEG__SHIFT 0x16 +#define SQ_VOP_DPP__SRC1_ABS__SHIFT 0x17 +#define SQ_VOP_DPP__BANK_MASK__SHIFT 0x18 +#define SQ_VOP_DPP__ROW_MASK__SHIFT 0x1c +#define SQ_VOP_DPP__SRC0_MASK 0x000000FFL +#define SQ_VOP_DPP__DPP_CTRL_MASK 0x0001FF00L +#define SQ_VOP_DPP__BOUND_CTRL_MASK 0x00080000L +#define SQ_VOP_DPP__SRC0_NEG_MASK 0x00100000L +#define SQ_VOP_DPP__SRC0_ABS_MASK 0x00200000L +#define SQ_VOP_DPP__SRC1_NEG_MASK 0x00400000L +#define SQ_VOP_DPP__SRC1_ABS_MASK 0x00800000L +#define SQ_VOP_DPP__BANK_MASK_MASK 0x0F000000L +#define SQ_VOP_DPP__ROW_MASK_MASK 0xF0000000L +//SQ_VOP_SDWA +#define SQ_VOP_SDWA__SRC0__SHIFT 0x0 +#define SQ_VOP_SDWA__DST_SEL__SHIFT 0x8 +#define SQ_VOP_SDWA__DST_UNUSED__SHIFT 0xb +#define SQ_VOP_SDWA__CLAMP__SHIFT 0xd +#define SQ_VOP_SDWA__OMOD__SHIFT 0xe +#define SQ_VOP_SDWA__SRC0_SEL__SHIFT 0x10 +#define SQ_VOP_SDWA__SRC0_SEXT__SHIFT 0x13 +#define SQ_VOP_SDWA__SRC0_NEG__SHIFT 0x14 +#define SQ_VOP_SDWA__SRC0_ABS__SHIFT 0x15 +#define SQ_VOP_SDWA__S0__SHIFT 0x17 +#define SQ_VOP_SDWA__SRC1_SEL__SHIFT 0x18 +#define SQ_VOP_SDWA__SRC1_SEXT__SHIFT 0x1b +#define SQ_VOP_SDWA__SRC1_NEG__SHIFT 0x1c +#define SQ_VOP_SDWA__SRC1_ABS__SHIFT 0x1d +#define SQ_VOP_SDWA__S1__SHIFT 0x1f +#define SQ_VOP_SDWA__SRC0_MASK 0x000000FFL +#define SQ_VOP_SDWA__DST_SEL_MASK 0x00000700L +#define SQ_VOP_SDWA__DST_UNUSED_MASK 0x00001800L +#define SQ_VOP_SDWA__CLAMP_MASK 0x00002000L +#define SQ_VOP_SDWA__OMOD_MASK 0x0000C000L +#define SQ_VOP_SDWA__SRC0_SEL_MASK 0x00070000L +#define SQ_VOP_SDWA__SRC0_SEXT_MASK 0x00080000L +#define SQ_VOP_SDWA__SRC0_NEG_MASK 0x00100000L +#define SQ_VOP_SDWA__SRC0_ABS_MASK 0x00200000L +#define SQ_VOP_SDWA__S0_MASK 0x00800000L +#define SQ_VOP_SDWA__SRC1_SEL_MASK 0x07000000L +#define SQ_VOP_SDWA__SRC1_SEXT_MASK 0x08000000L +#define SQ_VOP_SDWA__SRC1_NEG_MASK 0x10000000L +#define SQ_VOP_SDWA__SRC1_ABS_MASK 0x20000000L +#define SQ_VOP_SDWA__S1_MASK 0x80000000L +//SQ_VOP_SDWA_SDST_ENC +#define SQ_VOP_SDWA_SDST_ENC__SRC0__SHIFT 0x0 +#define SQ_VOP_SDWA_SDST_ENC__SDST__SHIFT 0x8 +#define SQ_VOP_SDWA_SDST_ENC__SD__SHIFT 0xf +#define SQ_VOP_SDWA_SDST_ENC__SRC0_SEL__SHIFT 0x10 +#define SQ_VOP_SDWA_SDST_ENC__SRC0_SEXT__SHIFT 0x13 +#define SQ_VOP_SDWA_SDST_ENC__SRC0_NEG__SHIFT 0x14 +#define SQ_VOP_SDWA_SDST_ENC__SRC0_ABS__SHIFT 0x15 +#define SQ_VOP_SDWA_SDST_ENC__S0__SHIFT 0x17 +#define SQ_VOP_SDWA_SDST_ENC__SRC1_SEL__SHIFT 0x18 +#define SQ_VOP_SDWA_SDST_ENC__SRC1_SEXT__SHIFT 0x1b +#define SQ_VOP_SDWA_SDST_ENC__SRC1_NEG__SHIFT 0x1c +#define SQ_VOP_SDWA_SDST_ENC__SRC1_ABS__SHIFT 0x1d +#define SQ_VOP_SDWA_SDST_ENC__S1__SHIFT 0x1f +#define SQ_VOP_SDWA_SDST_ENC__SRC0_MASK 0x000000FFL +#define SQ_VOP_SDWA_SDST_ENC__SDST_MASK 0x00007F00L +#define SQ_VOP_SDWA_SDST_ENC__SD_MASK 0x00008000L +#define SQ_VOP_SDWA_SDST_ENC__SRC0_SEL_MASK 0x00070000L +#define SQ_VOP_SDWA_SDST_ENC__SRC0_SEXT_MASK 0x00080000L +#define SQ_VOP_SDWA_SDST_ENC__SRC0_NEG_MASK 0x00100000L +#define SQ_VOP_SDWA_SDST_ENC__SRC0_ABS_MASK 0x00200000L +#define SQ_VOP_SDWA_SDST_ENC__S0_MASK 0x00800000L +#define SQ_VOP_SDWA_SDST_ENC__SRC1_SEL_MASK 0x07000000L +#define SQ_VOP_SDWA_SDST_ENC__SRC1_SEXT_MASK 0x08000000L +#define SQ_VOP_SDWA_SDST_ENC__SRC1_NEG_MASK 0x10000000L +#define SQ_VOP_SDWA_SDST_ENC__SRC1_ABS_MASK 0x20000000L +#define SQ_VOP_SDWA_SDST_ENC__S1_MASK 0x80000000L +//SQ_LB_CTR_CTRL +#define SQ_LB_CTR_CTRL__START__SHIFT 0x0 +#define SQ_LB_CTR_CTRL__LOAD__SHIFT 0x1 +#define SQ_LB_CTR_CTRL__CLEAR__SHIFT 0x2 +#define SQ_LB_CTR_CTRL__START_MASK 0x00000001L +#define SQ_LB_CTR_CTRL__LOAD_MASK 0x00000002L +#define SQ_LB_CTR_CTRL__CLEAR_MASK 0x00000004L +//SQ_LB_DATA0 +#define SQ_LB_DATA0__DATA__SHIFT 0x0 +#define SQ_LB_DATA0__DATA_MASK 0xFFFFFFFFL +//SQ_LB_DATA1 +#define SQ_LB_DATA1__DATA__SHIFT 0x0 +#define SQ_LB_DATA1__DATA_MASK 0xFFFFFFFFL +//SQ_LB_DATA2 +#define SQ_LB_DATA2__DATA__SHIFT 0x0 +#define SQ_LB_DATA2__DATA_MASK 0xFFFFFFFFL +//SQ_LB_DATA3 +#define SQ_LB_DATA3__DATA__SHIFT 0x0 +#define SQ_LB_DATA3__DATA_MASK 0xFFFFFFFFL +//SQ_LB_CTR_SEL +#define SQ_LB_CTR_SEL__SEL0__SHIFT 0x0 +#define SQ_LB_CTR_SEL__SEL1__SHIFT 0x4 +#define SQ_LB_CTR_SEL__SEL2__SHIFT 0x8 +#define SQ_LB_CTR_SEL__SEL3__SHIFT 0xc +#define SQ_LB_CTR_SEL__SEL0_MASK 0x0000000FL +#define SQ_LB_CTR_SEL__SEL1_MASK 0x000000F0L +#define SQ_LB_CTR_SEL__SEL2_MASK 0x00000F00L +#define SQ_LB_CTR_SEL__SEL3_MASK 0x0000F000L +//SQ_LB_CTR0_CU +#define SQ_LB_CTR0_CU__SH0_MASK__SHIFT 0x0 +#define SQ_LB_CTR0_CU__SH1_MASK__SHIFT 0x10 +#define SQ_LB_CTR0_CU__SH0_MASK_MASK 0x0000FFFFL +#define SQ_LB_CTR0_CU__SH1_MASK_MASK 0xFFFF0000L +//SQ_LB_CTR1_CU +#define SQ_LB_CTR1_CU__SH0_MASK__SHIFT 0x0 +#define SQ_LB_CTR1_CU__SH1_MASK__SHIFT 0x10 +#define SQ_LB_CTR1_CU__SH0_MASK_MASK 0x0000FFFFL +#define SQ_LB_CTR1_CU__SH1_MASK_MASK 0xFFFF0000L +//SQ_LB_CTR2_CU +#define SQ_LB_CTR2_CU__SH0_MASK__SHIFT 0x0 +#define SQ_LB_CTR2_CU__SH1_MASK__SHIFT 0x10 +#define SQ_LB_CTR2_CU__SH0_MASK_MASK 0x0000FFFFL +#define SQ_LB_CTR2_CU__SH1_MASK_MASK 0xFFFF0000L +//SQ_LB_CTR3_CU +#define SQ_LB_CTR3_CU__SH0_MASK__SHIFT 0x0 +#define SQ_LB_CTR3_CU__SH1_MASK__SHIFT 0x10 +#define SQ_LB_CTR3_CU__SH0_MASK_MASK 0x0000FFFFL +#define SQ_LB_CTR3_CU__SH1_MASK_MASK 0xFFFF0000L +//SQC_EDC_CNT +#define SQC_EDC_CNT__DATA_CU0_WRITE_DATA_BUF_SEC_COUNT__SHIFT 0x0 +#define SQC_EDC_CNT__DATA_CU0_WRITE_DATA_BUF_DED_COUNT__SHIFT 0x2 +#define SQC_EDC_CNT__DATA_CU0_UTCL1_LFIFO_SEC_COUNT__SHIFT 0x4 +#define SQC_EDC_CNT__DATA_CU0_UTCL1_LFIFO_DED_COUNT__SHIFT 0x6 +#define SQC_EDC_CNT__DATA_CU1_WRITE_DATA_BUF_SEC_COUNT__SHIFT 0x8 +#define SQC_EDC_CNT__DATA_CU1_WRITE_DATA_BUF_DED_COUNT__SHIFT 0xa +#define SQC_EDC_CNT__DATA_CU1_UTCL1_LFIFO_SEC_COUNT__SHIFT 0xc +#define SQC_EDC_CNT__DATA_CU1_UTCL1_LFIFO_DED_COUNT__SHIFT 0xe +#define SQC_EDC_CNT__DATA_CU2_WRITE_DATA_BUF_SEC_COUNT__SHIFT 0x10 +#define SQC_EDC_CNT__DATA_CU2_WRITE_DATA_BUF_DED_COUNT__SHIFT 0x12 +#define SQC_EDC_CNT__DATA_CU2_UTCL1_LFIFO_SEC_COUNT__SHIFT 0x14 +#define SQC_EDC_CNT__DATA_CU2_UTCL1_LFIFO_DED_COUNT__SHIFT 0x16 +#define SQC_EDC_CNT__DATA_CU3_WRITE_DATA_BUF_SEC_COUNT__SHIFT 0x18 +#define SQC_EDC_CNT__DATA_CU3_WRITE_DATA_BUF_DED_COUNT__SHIFT 0x1a +#define SQC_EDC_CNT__DATA_CU3_UTCL1_LFIFO_SEC_COUNT__SHIFT 0x1c +#define SQC_EDC_CNT__DATA_CU3_UTCL1_LFIFO_DED_COUNT__SHIFT 0x1e +#define SQC_EDC_CNT__DATA_CU0_WRITE_DATA_BUF_SEC_COUNT_MASK 0x00000003L +#define SQC_EDC_CNT__DATA_CU0_WRITE_DATA_BUF_DED_COUNT_MASK 0x0000000CL +#define SQC_EDC_CNT__DATA_CU0_UTCL1_LFIFO_SEC_COUNT_MASK 0x00000030L +#define SQC_EDC_CNT__DATA_CU0_UTCL1_LFIFO_DED_COUNT_MASK 0x000000C0L +#define SQC_EDC_CNT__DATA_CU1_WRITE_DATA_BUF_SEC_COUNT_MASK 0x00000300L +#define SQC_EDC_CNT__DATA_CU1_WRITE_DATA_BUF_DED_COUNT_MASK 0x00000C00L +#define SQC_EDC_CNT__DATA_CU1_UTCL1_LFIFO_SEC_COUNT_MASK 0x00003000L +#define SQC_EDC_CNT__DATA_CU1_UTCL1_LFIFO_DED_COUNT_MASK 0x0000C000L +#define SQC_EDC_CNT__DATA_CU2_WRITE_DATA_BUF_SEC_COUNT_MASK 0x00030000L +#define SQC_EDC_CNT__DATA_CU2_WRITE_DATA_BUF_DED_COUNT_MASK 0x000C0000L +#define SQC_EDC_CNT__DATA_CU2_UTCL1_LFIFO_SEC_COUNT_MASK 0x00300000L +#define SQC_EDC_CNT__DATA_CU2_UTCL1_LFIFO_DED_COUNT_MASK 0x00C00000L +#define SQC_EDC_CNT__DATA_CU3_WRITE_DATA_BUF_SEC_COUNT_MASK 0x03000000L +#define SQC_EDC_CNT__DATA_CU3_WRITE_DATA_BUF_DED_COUNT_MASK 0x0C000000L +#define SQC_EDC_CNT__DATA_CU3_UTCL1_LFIFO_SEC_COUNT_MASK 0x30000000L +#define SQC_EDC_CNT__DATA_CU3_UTCL1_LFIFO_DED_COUNT_MASK 0xC0000000L +//SQ_EDC_SEC_CNT +#define SQ_EDC_SEC_CNT__LDS_SEC__SHIFT 0x0 +#define SQ_EDC_SEC_CNT__SGPR_SEC__SHIFT 0x8 +#define SQ_EDC_SEC_CNT__VGPR_SEC__SHIFT 0x10 +#define SQ_EDC_SEC_CNT__LDS_SEC_MASK 0x000000FFL +#define SQ_EDC_SEC_CNT__SGPR_SEC_MASK 0x0000FF00L +#define SQ_EDC_SEC_CNT__VGPR_SEC_MASK 0x00FF0000L +//SQ_EDC_DED_CNT +#define SQ_EDC_DED_CNT__LDS_DED__SHIFT 0x0 +#define SQ_EDC_DED_CNT__SGPR_DED__SHIFT 0x8 +#define SQ_EDC_DED_CNT__VGPR_DED__SHIFT 0x10 +#define SQ_EDC_DED_CNT__LDS_DED_MASK 0x000000FFL +#define SQ_EDC_DED_CNT__SGPR_DED_MASK 0x0000FF00L +#define SQ_EDC_DED_CNT__VGPR_DED_MASK 0x00FF0000L +//SQ_EDC_INFO +#define SQ_EDC_INFO__WAVE_ID__SHIFT 0x0 +#define SQ_EDC_INFO__SIMD_ID__SHIFT 0x4 +#define SQ_EDC_INFO__SOURCE__SHIFT 0x6 +#define SQ_EDC_INFO__VM_ID__SHIFT 0x9 +#define SQ_EDC_INFO__WAVE_ID_MASK 0x0000000FL +#define SQ_EDC_INFO__SIMD_ID_MASK 0x00000030L +#define SQ_EDC_INFO__SOURCE_MASK 0x000001C0L +#define SQ_EDC_INFO__VM_ID_MASK 0x00001E00L +//SQ_EDC_CNT +#define SQ_EDC_CNT__LDS_D_SEC_COUNT__SHIFT 0x0 +#define SQ_EDC_CNT__LDS_D_DED_COUNT__SHIFT 0x2 +#define SQ_EDC_CNT__LDS_I_SEC_COUNT__SHIFT 0x4 +#define SQ_EDC_CNT__LDS_I_DED_COUNT__SHIFT 0x6 +#define SQ_EDC_CNT__SGPR_SEC_COUNT__SHIFT 0x8 +#define SQ_EDC_CNT__SGPR_DED_COUNT__SHIFT 0xa +#define SQ_EDC_CNT__VGPR0_SEC_COUNT__SHIFT 0xc +#define SQ_EDC_CNT__VGPR0_DED_COUNT__SHIFT 0xe +#define SQ_EDC_CNT__VGPR1_SEC_COUNT__SHIFT 0x10 +#define SQ_EDC_CNT__VGPR1_DED_COUNT__SHIFT 0x12 +#define SQ_EDC_CNT__VGPR2_SEC_COUNT__SHIFT 0x14 +#define SQ_EDC_CNT__VGPR2_DED_COUNT__SHIFT 0x16 +#define SQ_EDC_CNT__VGPR3_SEC_COUNT__SHIFT 0x18 +#define SQ_EDC_CNT__VGPR3_DED_COUNT__SHIFT 0x1a +#define SQ_EDC_CNT__LDS_D_SEC_COUNT_MASK 0x00000003L +#define SQ_EDC_CNT__LDS_D_DED_COUNT_MASK 0x0000000CL +#define SQ_EDC_CNT__LDS_I_SEC_COUNT_MASK 0x00000030L +#define SQ_EDC_CNT__LDS_I_DED_COUNT_MASK 0x000000C0L +#define SQ_EDC_CNT__SGPR_SEC_COUNT_MASK 0x00000300L +#define SQ_EDC_CNT__SGPR_DED_COUNT_MASK 0x00000C00L +#define SQ_EDC_CNT__VGPR0_SEC_COUNT_MASK 0x00003000L +#define SQ_EDC_CNT__VGPR0_DED_COUNT_MASK 0x0000C000L +#define SQ_EDC_CNT__VGPR1_SEC_COUNT_MASK 0x00030000L +#define SQ_EDC_CNT__VGPR1_DED_COUNT_MASK 0x000C0000L +#define SQ_EDC_CNT__VGPR2_SEC_COUNT_MASK 0x00300000L +#define SQ_EDC_CNT__VGPR2_DED_COUNT_MASK 0x00C00000L +#define SQ_EDC_CNT__VGPR3_SEC_COUNT_MASK 0x03000000L +#define SQ_EDC_CNT__VGPR3_DED_COUNT_MASK 0x0C000000L +//SQ_EDC_FUE_CNTL +#define SQ_EDC_FUE_CNTL__BLOCK_FUE_FLAGS__SHIFT 0x0 +#define SQ_EDC_FUE_CNTL__FUE_INTERRUPT_ENABLES__SHIFT 0x10 +#define SQ_EDC_FUE_CNTL__BLOCK_FUE_FLAGS_MASK 0x0000FFFFL +#define SQ_EDC_FUE_CNTL__FUE_INTERRUPT_ENABLES_MASK 0xFFFF0000L +//SQ_THREAD_TRACE_WORD_CMN +#define SQ_THREAD_TRACE_WORD_CMN__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_CMN__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_CMN__TOKEN_TYPE_MASK 0x000FL +#define SQ_THREAD_TRACE_WORD_CMN__TIME_DELTA_MASK 0x0010L +//SQ_THREAD_TRACE_WORD_EVENT +#define SQ_THREAD_TRACE_WORD_EVENT__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_EVENT__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_EVENT__SH_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_EVENT__STAGE__SHIFT 0x6 +#define SQ_THREAD_TRACE_WORD_EVENT__EVENT_TYPE__SHIFT 0xa +#define SQ_THREAD_TRACE_WORD_EVENT__TOKEN_TYPE_MASK 0x000FL +#define SQ_THREAD_TRACE_WORD_EVENT__TIME_DELTA_MASK 0x0010L +#define SQ_THREAD_TRACE_WORD_EVENT__SH_ID_MASK 0x0020L +#define SQ_THREAD_TRACE_WORD_EVENT__STAGE_MASK 0x01C0L +#define SQ_THREAD_TRACE_WORD_EVENT__EVENT_TYPE_MASK 0xFC00L +//SQ_THREAD_TRACE_WORD_INST +#define SQ_THREAD_TRACE_WORD_INST__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_INST__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_INST__WAVE_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_INST__SIMD_ID__SHIFT 0x9 +#define SQ_THREAD_TRACE_WORD_INST__INST_TYPE__SHIFT 0xb +#define SQ_THREAD_TRACE_WORD_INST__TOKEN_TYPE_MASK 0x000FL +#define SQ_THREAD_TRACE_WORD_INST__TIME_DELTA_MASK 0x0010L +#define SQ_THREAD_TRACE_WORD_INST__WAVE_ID_MASK 0x01E0L +#define SQ_THREAD_TRACE_WORD_INST__SIMD_ID_MASK 0x0600L +#define SQ_THREAD_TRACE_WORD_INST__INST_TYPE_MASK 0xF800L +//SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2 +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__WAVE_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__SIMD_ID__SHIFT 0x9 +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__TRAP_ERROR__SHIFT 0xf +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__PC_LO__SHIFT 0x10 +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__TOKEN_TYPE_MASK 0x0000000FL +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__TIME_DELTA_MASK 0x00000010L +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__WAVE_ID_MASK 0x000001E0L +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__SIMD_ID_MASK 0x00000600L +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__TRAP_ERROR_MASK 0x00008000L +#define SQ_THREAD_TRACE_WORD_INST_PC_1_OF_2__PC_LO_MASK 0xFFFF0000L +//SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2 +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__SH_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__CU_ID__SHIFT 0x6 +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__WAVE_ID__SHIFT 0xa +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__SIMD_ID__SHIFT 0xe +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__DATA_LO__SHIFT 0x10 +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__TOKEN_TYPE_MASK 0x0000000FL +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__TIME_DELTA_MASK 0x00000010L +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__SH_ID_MASK 0x00000020L +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__CU_ID_MASK 0x000003C0L +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__WAVE_ID_MASK 0x00003C00L +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__SIMD_ID_MASK 0x0000C000L +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_1_OF_2__DATA_LO_MASK 0xFFFF0000L +//SQ_THREAD_TRACE_WORD_ISSUE +#define SQ_THREAD_TRACE_WORD_ISSUE__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_ISSUE__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_ISSUE__SIMD_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_ISSUE__INST0__SHIFT 0x8 +#define SQ_THREAD_TRACE_WORD_ISSUE__INST1__SHIFT 0xa +#define SQ_THREAD_TRACE_WORD_ISSUE__INST2__SHIFT 0xc +#define SQ_THREAD_TRACE_WORD_ISSUE__INST3__SHIFT 0xe +#define SQ_THREAD_TRACE_WORD_ISSUE__INST4__SHIFT 0x10 +#define SQ_THREAD_TRACE_WORD_ISSUE__INST5__SHIFT 0x12 +#define SQ_THREAD_TRACE_WORD_ISSUE__INST6__SHIFT 0x14 +#define SQ_THREAD_TRACE_WORD_ISSUE__INST7__SHIFT 0x16 +#define SQ_THREAD_TRACE_WORD_ISSUE__INST8__SHIFT 0x18 +#define SQ_THREAD_TRACE_WORD_ISSUE__INST9__SHIFT 0x1a +#define SQ_THREAD_TRACE_WORD_ISSUE__TOKEN_TYPE_MASK 0x0000000FL +#define SQ_THREAD_TRACE_WORD_ISSUE__TIME_DELTA_MASK 0x00000010L +#define SQ_THREAD_TRACE_WORD_ISSUE__SIMD_ID_MASK 0x00000060L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST0_MASK 0x00000300L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST1_MASK 0x00000C00L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST2_MASK 0x00003000L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST3_MASK 0x0000C000L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST4_MASK 0x00030000L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST5_MASK 0x000C0000L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST6_MASK 0x00300000L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST7_MASK 0x00C00000L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST8_MASK 0x03000000L +#define SQ_THREAD_TRACE_WORD_ISSUE__INST9_MASK 0x0C000000L +//SQ_THREAD_TRACE_WORD_MISC +#define SQ_THREAD_TRACE_WORD_MISC__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_MISC__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_MISC__SH_ID__SHIFT 0xc +#define SQ_THREAD_TRACE_WORD_MISC__MISC_TOKEN_TYPE__SHIFT 0xd +#define SQ_THREAD_TRACE_WORD_MISC__TOKEN_TYPE_MASK 0x000FL +#define SQ_THREAD_TRACE_WORD_MISC__TIME_DELTA_MASK 0x0FF0L +#define SQ_THREAD_TRACE_WORD_MISC__SH_ID_MASK 0x1000L +#define SQ_THREAD_TRACE_WORD_MISC__MISC_TOKEN_TYPE_MASK 0xE000L +//SQ_THREAD_TRACE_WORD_PERF_1_OF_2 +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__SH_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__CU_ID__SHIFT 0x6 +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__CNTR_BANK__SHIFT 0xa +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__CNTR0__SHIFT 0xc +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__CNTR1_LO__SHIFT 0x19 +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__TOKEN_TYPE_MASK 0x0000000FL +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__TIME_DELTA_MASK 0x00000010L +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__SH_ID_MASK 0x00000020L +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__CU_ID_MASK 0x000003C0L +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__CNTR_BANK_MASK 0x00000C00L +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__CNTR0_MASK 0x01FFF000L +#define SQ_THREAD_TRACE_WORD_PERF_1_OF_2__CNTR1_LO_MASK 0xFE000000L +//SQ_THREAD_TRACE_WORD_REG_1_OF_2 +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__PIPE_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__ME_ID__SHIFT 0x7 +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_DROPPED_PREV__SHIFT 0x9 +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_TYPE__SHIFT 0xa +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_PRIV__SHIFT 0xe +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_OP__SHIFT 0xf +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_ADDR__SHIFT 0x10 +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__TOKEN_TYPE_MASK 0x0000000FL +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__TIME_DELTA_MASK 0x00000010L +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__PIPE_ID_MASK 0x00000060L +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__ME_ID_MASK 0x00000180L +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_DROPPED_PREV_MASK 0x00000200L +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_TYPE_MASK 0x00001C00L +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_PRIV_MASK 0x00004000L +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_OP_MASK 0x00008000L +#define SQ_THREAD_TRACE_WORD_REG_1_OF_2__REG_ADDR_MASK 0xFFFF0000L +//SQ_THREAD_TRACE_WORD_REG_2_OF_2 +#define SQ_THREAD_TRACE_WORD_REG_2_OF_2__DATA__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_REG_2_OF_2__DATA_MASK 0xFFFFFFFFL +//SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2 +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__PIPE_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__ME_ID__SHIFT 0x7 +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__REG_ADDR__SHIFT 0x9 +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__DATA_LO__SHIFT 0x10 +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__TOKEN_TYPE_MASK 0x0000000FL +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__TIME_DELTA_MASK 0x00000010L +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__PIPE_ID_MASK 0x00000060L +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__ME_ID_MASK 0x00000180L +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__REG_ADDR_MASK 0x0000FE00L +#define SQ_THREAD_TRACE_WORD_REG_CS_1_OF_2__DATA_LO_MASK 0xFFFF0000L +//SQ_THREAD_TRACE_WORD_REG_CS_2_OF_2 +#define SQ_THREAD_TRACE_WORD_REG_CS_2_OF_2__DATA_HI__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_REG_CS_2_OF_2__DATA_HI_MASK 0x0000FFFFL +//SQ_THREAD_TRACE_WORD_TIMESTAMP_1_OF_2 +#define SQ_THREAD_TRACE_WORD_TIMESTAMP_1_OF_2__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_TIMESTAMP_1_OF_2__TIME_LO__SHIFT 0x10 +#define SQ_THREAD_TRACE_WORD_TIMESTAMP_1_OF_2__TOKEN_TYPE_MASK 0x0000000FL +#define SQ_THREAD_TRACE_WORD_TIMESTAMP_1_OF_2__TIME_LO_MASK 0xFFFF0000L +//SQ_THREAD_TRACE_WORD_WAVE +#define SQ_THREAD_TRACE_WORD_WAVE__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_WAVE__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_WAVE__SH_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_WAVE__CU_ID__SHIFT 0x6 +#define SQ_THREAD_TRACE_WORD_WAVE__WAVE_ID__SHIFT 0xa +#define SQ_THREAD_TRACE_WORD_WAVE__SIMD_ID__SHIFT 0xe +#define SQ_THREAD_TRACE_WORD_WAVE__TOKEN_TYPE_MASK 0x000FL +#define SQ_THREAD_TRACE_WORD_WAVE__TIME_DELTA_MASK 0x0010L +#define SQ_THREAD_TRACE_WORD_WAVE__SH_ID_MASK 0x0020L +#define SQ_THREAD_TRACE_WORD_WAVE__CU_ID_MASK 0x03C0L +#define SQ_THREAD_TRACE_WORD_WAVE__WAVE_ID_MASK 0x3C00L +#define SQ_THREAD_TRACE_WORD_WAVE__SIMD_ID_MASK 0xC000L +//SQ_THREAD_TRACE_WORD_WAVE_START +#define SQ_THREAD_TRACE_WORD_WAVE_START__TOKEN_TYPE__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_WAVE_START__TIME_DELTA__SHIFT 0x4 +#define SQ_THREAD_TRACE_WORD_WAVE_START__SH_ID__SHIFT 0x5 +#define SQ_THREAD_TRACE_WORD_WAVE_START__CU_ID__SHIFT 0x6 +#define SQ_THREAD_TRACE_WORD_WAVE_START__WAVE_ID__SHIFT 0xa +#define SQ_THREAD_TRACE_WORD_WAVE_START__SIMD_ID__SHIFT 0xe +#define SQ_THREAD_TRACE_WORD_WAVE_START__DISPATCHER__SHIFT 0x10 +#define SQ_THREAD_TRACE_WORD_WAVE_START__VS_NO_ALLOC_OR_GROUPED__SHIFT 0x15 +#define SQ_THREAD_TRACE_WORD_WAVE_START__COUNT__SHIFT 0x16 +#define SQ_THREAD_TRACE_WORD_WAVE_START__TG_ID__SHIFT 0x1d +#define SQ_THREAD_TRACE_WORD_WAVE_START__TOKEN_TYPE_MASK 0x0000000FL +#define SQ_THREAD_TRACE_WORD_WAVE_START__TIME_DELTA_MASK 0x00000010L +#define SQ_THREAD_TRACE_WORD_WAVE_START__SH_ID_MASK 0x00000020L +#define SQ_THREAD_TRACE_WORD_WAVE_START__CU_ID_MASK 0x000003C0L +#define SQ_THREAD_TRACE_WORD_WAVE_START__WAVE_ID_MASK 0x00003C00L +#define SQ_THREAD_TRACE_WORD_WAVE_START__SIMD_ID_MASK 0x0000C000L +#define SQ_THREAD_TRACE_WORD_WAVE_START__DISPATCHER_MASK 0x001F0000L +#define SQ_THREAD_TRACE_WORD_WAVE_START__VS_NO_ALLOC_OR_GROUPED_MASK 0x00200000L +#define SQ_THREAD_TRACE_WORD_WAVE_START__COUNT_MASK 0x1FC00000L +#define SQ_THREAD_TRACE_WORD_WAVE_START__TG_ID_MASK 0xE0000000L +//SQ_THREAD_TRACE_WORD_INST_PC_2_OF_2 +#define SQ_THREAD_TRACE_WORD_INST_PC_2_OF_2__PC_HI__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_INST_PC_2_OF_2__PC_HI_MASK 0x00FFFFFFL +//SQ_THREAD_TRACE_WORD_INST_USERDATA_2_OF_2 +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_2_OF_2__DATA_HI__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_INST_USERDATA_2_OF_2__DATA_HI_MASK 0xFFFFL +//SQ_THREAD_TRACE_WORD_PERF_2_OF_2 +#define SQ_THREAD_TRACE_WORD_PERF_2_OF_2__CNTR1_HI__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_PERF_2_OF_2__CNTR2__SHIFT 0x6 +#define SQ_THREAD_TRACE_WORD_PERF_2_OF_2__CNTR3__SHIFT 0x13 +#define SQ_THREAD_TRACE_WORD_PERF_2_OF_2__CNTR1_HI_MASK 0x0000003FL +#define SQ_THREAD_TRACE_WORD_PERF_2_OF_2__CNTR2_MASK 0x0007FFC0L +#define SQ_THREAD_TRACE_WORD_PERF_2_OF_2__CNTR3_MASK 0xFFF80000L +//SQ_THREAD_TRACE_WORD_TIMESTAMP_2_OF_2 +#define SQ_THREAD_TRACE_WORD_TIMESTAMP_2_OF_2__TIME_HI__SHIFT 0x0 +#define SQ_THREAD_TRACE_WORD_TIMESTAMP_2_OF_2__TIME_HI_MASK 0xFFFFFFFFL +//SQ_WREXEC_EXEC_HI +#define SQ_WREXEC_EXEC_HI__ADDR_HI__SHIFT 0x0 +#define SQ_WREXEC_EXEC_HI__FIRST_WAVE__SHIFT 0x1a +#define SQ_WREXEC_EXEC_HI__ATC__SHIFT 0x1b +#define SQ_WREXEC_EXEC_HI__MTYPE__SHIFT 0x1c +#define SQ_WREXEC_EXEC_HI__MSB__SHIFT 0x1f +#define SQ_WREXEC_EXEC_HI__ADDR_HI_MASK 0x0000FFFFL +#define SQ_WREXEC_EXEC_HI__FIRST_WAVE_MASK 0x04000000L +#define SQ_WREXEC_EXEC_HI__ATC_MASK 0x08000000L +#define SQ_WREXEC_EXEC_HI__MTYPE_MASK 0x70000000L +#define SQ_WREXEC_EXEC_HI__MSB_MASK 0x80000000L +//SQ_WREXEC_EXEC_LO +#define SQ_WREXEC_EXEC_LO__ADDR_LO__SHIFT 0x0 +#define SQ_WREXEC_EXEC_LO__ADDR_LO_MASK 0xFFFFFFFFL +//SQ_BUF_RSRC_WORD0 +#define SQ_BUF_RSRC_WORD0__BASE_ADDRESS__SHIFT 0x0 +#define SQ_BUF_RSRC_WORD0__BASE_ADDRESS_MASK 0xFFFFFFFFL +//SQ_BUF_RSRC_WORD1 +#define SQ_BUF_RSRC_WORD1__BASE_ADDRESS_HI__SHIFT 0x0 +#define SQ_BUF_RSRC_WORD1__STRIDE__SHIFT 0x10 +#define SQ_BUF_RSRC_WORD1__CACHE_SWIZZLE__SHIFT 0x1e +#define SQ_BUF_RSRC_WORD1__SWIZZLE_ENABLE__SHIFT 0x1f +#define SQ_BUF_RSRC_WORD1__BASE_ADDRESS_HI_MASK 0x0000FFFFL +#define SQ_BUF_RSRC_WORD1__STRIDE_MASK 0x3FFF0000L +#define SQ_BUF_RSRC_WORD1__CACHE_SWIZZLE_MASK 0x40000000L +#define SQ_BUF_RSRC_WORD1__SWIZZLE_ENABLE_MASK 0x80000000L +//SQ_BUF_RSRC_WORD2 +#define SQ_BUF_RSRC_WORD2__NUM_RECORDS__SHIFT 0x0 +#define SQ_BUF_RSRC_WORD2__NUM_RECORDS_MASK 0xFFFFFFFFL +//SQ_BUF_RSRC_WORD3 +#define SQ_BUF_RSRC_WORD3__DST_SEL_X__SHIFT 0x0 +#define SQ_BUF_RSRC_WORD3__DST_SEL_Y__SHIFT 0x3 +#define SQ_BUF_RSRC_WORD3__DST_SEL_Z__SHIFT 0x6 +#define SQ_BUF_RSRC_WORD3__DST_SEL_W__SHIFT 0x9 +#define SQ_BUF_RSRC_WORD3__NUM_FORMAT__SHIFT 0xc +#define SQ_BUF_RSRC_WORD3__DATA_FORMAT__SHIFT 0xf +#define SQ_BUF_RSRC_WORD3__USER_VM_ENABLE__SHIFT 0x13 +#define SQ_BUF_RSRC_WORD3__USER_VM_MODE__SHIFT 0x14 +#define SQ_BUF_RSRC_WORD3__INDEX_STRIDE__SHIFT 0x15 +#define SQ_BUF_RSRC_WORD3__ADD_TID_ENABLE__SHIFT 0x17 +#define SQ_BUF_RSRC_WORD3__NV__SHIFT 0x1b +#define SQ_BUF_RSRC_WORD3__TYPE__SHIFT 0x1e +#define SQ_BUF_RSRC_WORD3__DST_SEL_X_MASK 0x00000007L +#define SQ_BUF_RSRC_WORD3__DST_SEL_Y_MASK 0x00000038L +#define SQ_BUF_RSRC_WORD3__DST_SEL_Z_MASK 0x000001C0L +#define SQ_BUF_RSRC_WORD3__DST_SEL_W_MASK 0x00000E00L +#define SQ_BUF_RSRC_WORD3__NUM_FORMAT_MASK 0x00007000L +#define SQ_BUF_RSRC_WORD3__DATA_FORMAT_MASK 0x00078000L +#define SQ_BUF_RSRC_WORD3__USER_VM_ENABLE_MASK 0x00080000L +#define SQ_BUF_RSRC_WORD3__USER_VM_MODE_MASK 0x00100000L +#define SQ_BUF_RSRC_WORD3__INDEX_STRIDE_MASK 0x00600000L +#define SQ_BUF_RSRC_WORD3__ADD_TID_ENABLE_MASK 0x00800000L +#define SQ_BUF_RSRC_WORD3__NV_MASK 0x08000000L +#define SQ_BUF_RSRC_WORD3__TYPE_MASK 0xC0000000L +//SQ_IMG_RSRC_WORD0 +#define SQ_IMG_RSRC_WORD0__BASE_ADDRESS__SHIFT 0x0 +#define SQ_IMG_RSRC_WORD0__BASE_ADDRESS_MASK 0xFFFFFFFFL +//SQ_IMG_RSRC_WORD1 +#define SQ_IMG_RSRC_WORD1__BASE_ADDRESS_HI__SHIFT 0x0 +#define SQ_IMG_RSRC_WORD1__MIN_LOD__SHIFT 0x8 +#define SQ_IMG_RSRC_WORD1__DATA_FORMAT__SHIFT 0x14 +#define SQ_IMG_RSRC_WORD1__NUM_FORMAT__SHIFT 0x1a +#define SQ_IMG_RSRC_WORD1__NV__SHIFT 0x1e +#define SQ_IMG_RSRC_WORD1__META_DIRECT__SHIFT 0x1f +#define SQ_IMG_RSRC_WORD1__BASE_ADDRESS_HI_MASK 0x000000FFL +#define SQ_IMG_RSRC_WORD1__MIN_LOD_MASK 0x000FFF00L +#define SQ_IMG_RSRC_WORD1__DATA_FORMAT_MASK 0x03F00000L +#define SQ_IMG_RSRC_WORD1__NUM_FORMAT_MASK 0x3C000000L +#define SQ_IMG_RSRC_WORD1__NV_MASK 0x40000000L +#define SQ_IMG_RSRC_WORD1__META_DIRECT_MASK 0x80000000L +//SQ_IMG_RSRC_WORD2 +#define SQ_IMG_RSRC_WORD2__WIDTH__SHIFT 0x0 +#define SQ_IMG_RSRC_WORD2__HEIGHT__SHIFT 0xe +#define SQ_IMG_RSRC_WORD2__PERF_MOD__SHIFT 0x1c +#define SQ_IMG_RSRC_WORD2__WIDTH_MASK 0x00003FFFL +#define SQ_IMG_RSRC_WORD2__HEIGHT_MASK 0x0FFFC000L +#define SQ_IMG_RSRC_WORD2__PERF_MOD_MASK 0x70000000L +//SQ_IMG_RSRC_WORD3 +#define SQ_IMG_RSRC_WORD3__DST_SEL_X__SHIFT 0x0 +#define SQ_IMG_RSRC_WORD3__DST_SEL_Y__SHIFT 0x3 +#define SQ_IMG_RSRC_WORD3__DST_SEL_Z__SHIFT 0x6 +#define SQ_IMG_RSRC_WORD3__DST_SEL_W__SHIFT 0x9 +#define SQ_IMG_RSRC_WORD3__BASE_LEVEL__SHIFT 0xc +#define SQ_IMG_RSRC_WORD3__LAST_LEVEL__SHIFT 0x10 +#define SQ_IMG_RSRC_WORD3__SW_MODE__SHIFT 0x14 +#define SQ_IMG_RSRC_WORD3__TYPE__SHIFT 0x1c +#define SQ_IMG_RSRC_WORD3__DST_SEL_X_MASK 0x00000007L +#define SQ_IMG_RSRC_WORD3__DST_SEL_Y_MASK 0x00000038L +#define SQ_IMG_RSRC_WORD3__DST_SEL_Z_MASK 0x000001C0L +#define SQ_IMG_RSRC_WORD3__DST_SEL_W_MASK 0x00000E00L +#define SQ_IMG_RSRC_WORD3__BASE_LEVEL_MASK 0x0000F000L +#define SQ_IMG_RSRC_WORD3__LAST_LEVEL_MASK 0x000F0000L +#define SQ_IMG_RSRC_WORD3__SW_MODE_MASK 0x01F00000L +#define SQ_IMG_RSRC_WORD3__TYPE_MASK 0xF0000000L +//SQ_IMG_RSRC_WORD4 +#define SQ_IMG_RSRC_WORD4__DEPTH__SHIFT 0x0 +#define SQ_IMG_RSRC_WORD4__PITCH__SHIFT 0xd +#define SQ_IMG_RSRC_WORD4__BC_SWIZZLE__SHIFT 0x1d +#define SQ_IMG_RSRC_WORD4__DEPTH_MASK 0x00001FFFL +#define SQ_IMG_RSRC_WORD4__PITCH_MASK 0x1FFFE000L +#define SQ_IMG_RSRC_WORD4__BC_SWIZZLE_MASK 0xE0000000L +//SQ_IMG_RSRC_WORD5 +#define SQ_IMG_RSRC_WORD5__BASE_ARRAY__SHIFT 0x0 +#define SQ_IMG_RSRC_WORD5__ARRAY_PITCH__SHIFT 0xd +#define SQ_IMG_RSRC_WORD5__META_DATA_ADDRESS__SHIFT 0x11 +#define SQ_IMG_RSRC_WORD5__META_LINEAR__SHIFT 0x19 +#define SQ_IMG_RSRC_WORD5__META_PIPE_ALIGNED__SHIFT 0x1a +#define SQ_IMG_RSRC_WORD5__META_RB_ALIGNED__SHIFT 0x1b +#define SQ_IMG_RSRC_WORD5__MAX_MIP__SHIFT 0x1c +#define SQ_IMG_RSRC_WORD5__BASE_ARRAY_MASK 0x00001FFFL +#define SQ_IMG_RSRC_WORD5__ARRAY_PITCH_MASK 0x0001E000L +#define SQ_IMG_RSRC_WORD5__META_DATA_ADDRESS_MASK 0x01FE0000L +#define SQ_IMG_RSRC_WORD5__META_LINEAR_MASK 0x02000000L +#define SQ_IMG_RSRC_WORD5__META_PIPE_ALIGNED_MASK 0x04000000L +#define SQ_IMG_RSRC_WORD5__META_RB_ALIGNED_MASK 0x08000000L +#define SQ_IMG_RSRC_WORD5__MAX_MIP_MASK 0xF0000000L +//SQ_IMG_RSRC_WORD6 +#define SQ_IMG_RSRC_WORD6__MIN_LOD_WARN__SHIFT 0x0 +#define SQ_IMG_RSRC_WORD6__COUNTER_BANK_ID__SHIFT 0xc +#define SQ_IMG_RSRC_WORD6__LOD_HDW_CNT_EN__SHIFT 0x14 +#define SQ_IMG_RSRC_WORD6__COMPRESSION_EN__SHIFT 0x15 +#define SQ_IMG_RSRC_WORD6__ALPHA_IS_ON_MSB__SHIFT 0x16 +#define SQ_IMG_RSRC_WORD6__COLOR_TRANSFORM__SHIFT 0x17 +#define SQ_IMG_RSRC_WORD6__LOST_ALPHA_BITS__SHIFT 0x18 +#define SQ_IMG_RSRC_WORD6__LOST_COLOR_BITS__SHIFT 0x1c +#define SQ_IMG_RSRC_WORD6__MIN_LOD_WARN_MASK 0x00000FFFL +#define SQ_IMG_RSRC_WORD6__COUNTER_BANK_ID_MASK 0x000FF000L +#define SQ_IMG_RSRC_WORD6__LOD_HDW_CNT_EN_MASK 0x00100000L +#define SQ_IMG_RSRC_WORD6__COMPRESSION_EN_MASK 0x00200000L +#define SQ_IMG_RSRC_WORD6__ALPHA_IS_ON_MSB_MASK 0x00400000L +#define SQ_IMG_RSRC_WORD6__COLOR_TRANSFORM_MASK 0x00800000L +#define SQ_IMG_RSRC_WORD6__LOST_ALPHA_BITS_MASK 0x0F000000L +#define SQ_IMG_RSRC_WORD6__LOST_COLOR_BITS_MASK 0xF0000000L +//SQ_IMG_RSRC_WORD7 +#define SQ_IMG_RSRC_WORD7__META_DATA_ADDRESS__SHIFT 0x0 +#define SQ_IMG_RSRC_WORD7__META_DATA_ADDRESS_MASK 0xFFFFFFFFL +//SQ_IMG_SAMP_WORD0 +#define SQ_IMG_SAMP_WORD0__CLAMP_X__SHIFT 0x0 +#define SQ_IMG_SAMP_WORD0__CLAMP_Y__SHIFT 0x3 +#define SQ_IMG_SAMP_WORD0__CLAMP_Z__SHIFT 0x6 +#define SQ_IMG_SAMP_WORD0__MAX_ANISO_RATIO__SHIFT 0x9 +#define SQ_IMG_SAMP_WORD0__DEPTH_COMPARE_FUNC__SHIFT 0xc +#define SQ_IMG_SAMP_WORD0__FORCE_UNNORMALIZED__SHIFT 0xf +#define SQ_IMG_SAMP_WORD0__ANISO_THRESHOLD__SHIFT 0x10 +#define SQ_IMG_SAMP_WORD0__MC_COORD_TRUNC__SHIFT 0x13 +#define SQ_IMG_SAMP_WORD0__FORCE_DEGAMMA__SHIFT 0x14 +#define SQ_IMG_SAMP_WORD0__ANISO_BIAS__SHIFT 0x15 +#define SQ_IMG_SAMP_WORD0__TRUNC_COORD__SHIFT 0x1b +#define SQ_IMG_SAMP_WORD0__DISABLE_CUBE_WRAP__SHIFT 0x1c +#define SQ_IMG_SAMP_WORD0__FILTER_MODE__SHIFT 0x1d +#define SQ_IMG_SAMP_WORD0__COMPAT_MODE__SHIFT 0x1f +#define SQ_IMG_SAMP_WORD0__CLAMP_X_MASK 0x00000007L +#define SQ_IMG_SAMP_WORD0__CLAMP_Y_MASK 0x00000038L +#define SQ_IMG_SAMP_WORD0__CLAMP_Z_MASK 0x000001C0L +#define SQ_IMG_SAMP_WORD0__MAX_ANISO_RATIO_MASK 0x00000E00L +#define SQ_IMG_SAMP_WORD0__DEPTH_COMPARE_FUNC_MASK 0x00007000L +#define SQ_IMG_SAMP_WORD0__FORCE_UNNORMALIZED_MASK 0x00008000L +#define SQ_IMG_SAMP_WORD0__ANISO_THRESHOLD_MASK 0x00070000L +#define SQ_IMG_SAMP_WORD0__MC_COORD_TRUNC_MASK 0x00080000L +#define SQ_IMG_SAMP_WORD0__FORCE_DEGAMMA_MASK 0x00100000L +#define SQ_IMG_SAMP_WORD0__ANISO_BIAS_MASK 0x07E00000L +#define SQ_IMG_SAMP_WORD0__TRUNC_COORD_MASK 0x08000000L +#define SQ_IMG_SAMP_WORD0__DISABLE_CUBE_WRAP_MASK 0x10000000L +#define SQ_IMG_SAMP_WORD0__FILTER_MODE_MASK 0x60000000L +#define SQ_IMG_SAMP_WORD0__COMPAT_MODE_MASK 0x80000000L +//SQ_IMG_SAMP_WORD1 +#define SQ_IMG_SAMP_WORD1__MIN_LOD__SHIFT 0x0 +#define SQ_IMG_SAMP_WORD1__MAX_LOD__SHIFT 0xc +#define SQ_IMG_SAMP_WORD1__PERF_MIP__SHIFT 0x18 +#define SQ_IMG_SAMP_WORD1__PERF_Z__SHIFT 0x1c +#define SQ_IMG_SAMP_WORD1__MIN_LOD_MASK 0x00000FFFL +#define SQ_IMG_SAMP_WORD1__MAX_LOD_MASK 0x00FFF000L +#define SQ_IMG_SAMP_WORD1__PERF_MIP_MASK 0x0F000000L +#define SQ_IMG_SAMP_WORD1__PERF_Z_MASK 0xF0000000L +//SQ_IMG_SAMP_WORD2 +#define SQ_IMG_SAMP_WORD2__LOD_BIAS__SHIFT 0x0 +#define SQ_IMG_SAMP_WORD2__LOD_BIAS_SEC__SHIFT 0xe +#define SQ_IMG_SAMP_WORD2__XY_MAG_FILTER__SHIFT 0x14 +#define SQ_IMG_SAMP_WORD2__XY_MIN_FILTER__SHIFT 0x16 +#define SQ_IMG_SAMP_WORD2__Z_FILTER__SHIFT 0x18 +#define SQ_IMG_SAMP_WORD2__MIP_FILTER__SHIFT 0x1a +#define SQ_IMG_SAMP_WORD2__MIP_POINT_PRECLAMP__SHIFT 0x1c +#define SQ_IMG_SAMP_WORD2__BLEND_ZERO_PRT__SHIFT 0x1d +#define SQ_IMG_SAMP_WORD2__FILTER_PREC_FIX__SHIFT 0x1e +#define SQ_IMG_SAMP_WORD2__ANISO_OVERRIDE__SHIFT 0x1f +#define SQ_IMG_SAMP_WORD2__LOD_BIAS_MASK 0x00003FFFL +#define SQ_IMG_SAMP_WORD2__LOD_BIAS_SEC_MASK 0x000FC000L +#define SQ_IMG_SAMP_WORD2__XY_MAG_FILTER_MASK 0x00300000L +#define SQ_IMG_SAMP_WORD2__XY_MIN_FILTER_MASK 0x00C00000L +#define SQ_IMG_SAMP_WORD2__Z_FILTER_MASK 0x03000000L +#define SQ_IMG_SAMP_WORD2__MIP_FILTER_MASK 0x0C000000L +#define SQ_IMG_SAMP_WORD2__MIP_POINT_PRECLAMP_MASK 0x10000000L +#define SQ_IMG_SAMP_WORD2__BLEND_ZERO_PRT_MASK 0x20000000L +#define SQ_IMG_SAMP_WORD2__FILTER_PREC_FIX_MASK 0x40000000L +#define SQ_IMG_SAMP_WORD2__ANISO_OVERRIDE_MASK 0x80000000L +//SQ_IMG_SAMP_WORD3 +#define SQ_IMG_SAMP_WORD3__BORDER_COLOR_PTR__SHIFT 0x0 +#define SQ_IMG_SAMP_WORD3__SKIP_DEGAMMA__SHIFT 0xc +#define SQ_IMG_SAMP_WORD3__BORDER_COLOR_TYPE__SHIFT 0x1e +#define SQ_IMG_SAMP_WORD3__BORDER_COLOR_PTR_MASK 0x00000FFFL +#define SQ_IMG_SAMP_WORD3__SKIP_DEGAMMA_MASK 0x00001000L +#define SQ_IMG_SAMP_WORD3__BORDER_COLOR_TYPE_MASK 0xC0000000L +//SQ_FLAT_SCRATCH_WORD0 +#define SQ_FLAT_SCRATCH_WORD0__SIZE__SHIFT 0x0 +#define SQ_FLAT_SCRATCH_WORD0__SIZE_MASK 0x0007FFFFL +//SQ_FLAT_SCRATCH_WORD1 +#define SQ_FLAT_SCRATCH_WORD1__OFFSET__SHIFT 0x0 +#define SQ_FLAT_SCRATCH_WORD1__OFFSET_MASK 0x00FFFFFFL +//SQ_M0_GPR_IDX_WORD +#define SQ_M0_GPR_IDX_WORD__INDEX__SHIFT 0x0 +#define SQ_M0_GPR_IDX_WORD__VSRC0_REL__SHIFT 0xc +#define SQ_M0_GPR_IDX_WORD__VSRC1_REL__SHIFT 0xd +#define SQ_M0_GPR_IDX_WORD__VSRC2_REL__SHIFT 0xe +#define SQ_M0_GPR_IDX_WORD__VDST_REL__SHIFT 0xf +#define SQ_M0_GPR_IDX_WORD__INDEX_MASK 0x000000FFL +#define SQ_M0_GPR_IDX_WORD__VSRC0_REL_MASK 0x00001000L +#define SQ_M0_GPR_IDX_WORD__VSRC1_REL_MASK 0x00002000L +#define SQ_M0_GPR_IDX_WORD__VSRC2_REL_MASK 0x00004000L +#define SQ_M0_GPR_IDX_WORD__VDST_REL_MASK 0x00008000L +//SQC_ICACHE_UTCL1_CNTL1 +#define SQC_ICACHE_UTCL1_CNTL1__FORCE_4K_L2_RESP__SHIFT 0x0 +#define SQC_ICACHE_UTCL1_CNTL1__GPUVM_64K_DEF__SHIFT 0x1 +#define SQC_ICACHE_UTCL1_CNTL1__GPUVM_PERM_MODE__SHIFT 0x2 +#define SQC_ICACHE_UTCL1_CNTL1__RESP_MODE__SHIFT 0x3 +#define SQC_ICACHE_UTCL1_CNTL1__RESP_FAULT_MODE__SHIFT 0x5 +#define SQC_ICACHE_UTCL1_CNTL1__CLIENTID__SHIFT 0x7 +#define SQC_ICACHE_UTCL1_CNTL1__ENABLE_PUSH_LFIFO__SHIFT 0x11 +#define SQC_ICACHE_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB__SHIFT 0x12 +#define SQC_ICACHE_UTCL1_CNTL1__REG_INVALIDATE_VMID__SHIFT 0x13 +#define SQC_ICACHE_UTCL1_CNTL1__REG_INVALIDATE_ALL_VMID__SHIFT 0x17 +#define SQC_ICACHE_UTCL1_CNTL1__REG_INVALIDATE_TOGGLE__SHIFT 0x18 +#define SQC_ICACHE_UTCL1_CNTL1__CLIENT_INVALIDATE_ALL_VMID__SHIFT 0x19 +#define SQC_ICACHE_UTCL1_CNTL1__FORCE_MISS__SHIFT 0x1a +#define SQC_ICACHE_UTCL1_CNTL1__FORCE_IN_ORDER__SHIFT 0x1b +#define SQC_ICACHE_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2__SHIFT 0x1c +#define SQC_ICACHE_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2__SHIFT 0x1e +#define SQC_ICACHE_UTCL1_CNTL1__FORCE_4K_L2_RESP_MASK 0x00000001L +#define SQC_ICACHE_UTCL1_CNTL1__GPUVM_64K_DEF_MASK 0x00000002L +#define SQC_ICACHE_UTCL1_CNTL1__GPUVM_PERM_MODE_MASK 0x00000004L +#define SQC_ICACHE_UTCL1_CNTL1__RESP_MODE_MASK 0x00000018L +#define SQC_ICACHE_UTCL1_CNTL1__RESP_FAULT_MODE_MASK 0x00000060L +#define SQC_ICACHE_UTCL1_CNTL1__CLIENTID_MASK 0x0000FF80L +#define SQC_ICACHE_UTCL1_CNTL1__ENABLE_PUSH_LFIFO_MASK 0x00020000L +#define SQC_ICACHE_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB_MASK 0x00040000L +#define SQC_ICACHE_UTCL1_CNTL1__REG_INVALIDATE_VMID_MASK 0x00780000L +#define SQC_ICACHE_UTCL1_CNTL1__REG_INVALIDATE_ALL_VMID_MASK 0x00800000L +#define SQC_ICACHE_UTCL1_CNTL1__REG_INVALIDATE_TOGGLE_MASK 0x01000000L +#define SQC_ICACHE_UTCL1_CNTL1__CLIENT_INVALIDATE_ALL_VMID_MASK 0x02000000L +#define SQC_ICACHE_UTCL1_CNTL1__FORCE_MISS_MASK 0x04000000L +#define SQC_ICACHE_UTCL1_CNTL1__FORCE_IN_ORDER_MASK 0x08000000L +#define SQC_ICACHE_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2_MASK 0x30000000L +#define SQC_ICACHE_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2_MASK 0xC0000000L +//SQC_ICACHE_UTCL1_CNTL2 +#define SQC_ICACHE_UTCL1_CNTL2__SPARE__SHIFT 0x0 +#define SQC_ICACHE_UTCL1_CNTL2__LFIFO_SCAN_DISABLE__SHIFT 0x8 +#define SQC_ICACHE_UTCL1_CNTL2__MTYPE_OVRD_DIS__SHIFT 0x9 +#define SQC_ICACHE_UTCL1_CNTL2__LINE_VALID__SHIFT 0xa +#define SQC_ICACHE_UTCL1_CNTL2__DIS_EDC__SHIFT 0xb +#define SQC_ICACHE_UTCL1_CNTL2__GPUVM_INV_MODE__SHIFT 0xc +#define SQC_ICACHE_UTCL1_CNTL2__SHOOTDOWN_OPT__SHIFT 0xd +#define SQC_ICACHE_UTCL1_CNTL2__FORCE_SNOOP__SHIFT 0xe +#define SQC_ICACHE_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK__SHIFT 0xf +#define SQC_ICACHE_UTCL1_CNTL2__ARB_BURST_MODE__SHIFT 0x10 +#define SQC_ICACHE_UTCL1_CNTL2__ENABLE_PERF_EVENT_RD_WR__SHIFT 0x12 +#define SQC_ICACHE_UTCL1_CNTL2__PERF_EVENT_RD_WR__SHIFT 0x13 +#define SQC_ICACHE_UTCL1_CNTL2__ENABLE_PERF_EVENT_VMID__SHIFT 0x14 +#define SQC_ICACHE_UTCL1_CNTL2__PERF_EVENT_VMID__SHIFT 0x15 +#define SQC_ICACHE_UTCL1_CNTL2__FORCE_FRAG_2M_TO_64K__SHIFT 0x1a +#define SQC_ICACHE_UTCL1_CNTL2__SPARE_MASK 0x000000FFL +#define SQC_ICACHE_UTCL1_CNTL2__LFIFO_SCAN_DISABLE_MASK 0x00000100L +#define SQC_ICACHE_UTCL1_CNTL2__MTYPE_OVRD_DIS_MASK 0x00000200L +#define SQC_ICACHE_UTCL1_CNTL2__LINE_VALID_MASK 0x00000400L +#define SQC_ICACHE_UTCL1_CNTL2__DIS_EDC_MASK 0x00000800L +#define SQC_ICACHE_UTCL1_CNTL2__GPUVM_INV_MODE_MASK 0x00001000L +#define SQC_ICACHE_UTCL1_CNTL2__SHOOTDOWN_OPT_MASK 0x00002000L +#define SQC_ICACHE_UTCL1_CNTL2__FORCE_SNOOP_MASK 0x00004000L +#define SQC_ICACHE_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK_MASK 0x00008000L +#define SQC_ICACHE_UTCL1_CNTL2__ARB_BURST_MODE_MASK 0x00030000L +#define SQC_ICACHE_UTCL1_CNTL2__ENABLE_PERF_EVENT_RD_WR_MASK 0x00040000L +#define SQC_ICACHE_UTCL1_CNTL2__PERF_EVENT_RD_WR_MASK 0x00080000L +#define SQC_ICACHE_UTCL1_CNTL2__ENABLE_PERF_EVENT_VMID_MASK 0x00100000L +#define SQC_ICACHE_UTCL1_CNTL2__PERF_EVENT_VMID_MASK 0x01E00000L +#define SQC_ICACHE_UTCL1_CNTL2__FORCE_FRAG_2M_TO_64K_MASK 0x04000000L +//SQC_DCACHE_UTCL1_CNTL1 +#define SQC_DCACHE_UTCL1_CNTL1__FORCE_4K_L2_RESP__SHIFT 0x0 +#define SQC_DCACHE_UTCL1_CNTL1__GPUVM_64K_DEF__SHIFT 0x1 +#define SQC_DCACHE_UTCL1_CNTL1__GPUVM_PERM_MODE__SHIFT 0x2 +#define SQC_DCACHE_UTCL1_CNTL1__RESP_MODE__SHIFT 0x3 +#define SQC_DCACHE_UTCL1_CNTL1__RESP_FAULT_MODE__SHIFT 0x5 +#define SQC_DCACHE_UTCL1_CNTL1__CLIENTID__SHIFT 0x7 +#define SQC_DCACHE_UTCL1_CNTL1__ENABLE_PUSH_LFIFO__SHIFT 0x11 +#define SQC_DCACHE_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB__SHIFT 0x12 +#define SQC_DCACHE_UTCL1_CNTL1__REG_INVALIDATE_VMID__SHIFT 0x13 +#define SQC_DCACHE_UTCL1_CNTL1__REG_INVALIDATE_ALL_VMID__SHIFT 0x17 +#define SQC_DCACHE_UTCL1_CNTL1__REG_INVALIDATE_TOGGLE__SHIFT 0x18 +#define SQC_DCACHE_UTCL1_CNTL1__CLIENT_INVALIDATE_ALL_VMID__SHIFT 0x19 +#define SQC_DCACHE_UTCL1_CNTL1__FORCE_MISS__SHIFT 0x1a +#define SQC_DCACHE_UTCL1_CNTL1__FORCE_IN_ORDER__SHIFT 0x1b +#define SQC_DCACHE_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2__SHIFT 0x1c +#define SQC_DCACHE_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2__SHIFT 0x1e +#define SQC_DCACHE_UTCL1_CNTL1__FORCE_4K_L2_RESP_MASK 0x00000001L +#define SQC_DCACHE_UTCL1_CNTL1__GPUVM_64K_DEF_MASK 0x00000002L +#define SQC_DCACHE_UTCL1_CNTL1__GPUVM_PERM_MODE_MASK 0x00000004L +#define SQC_DCACHE_UTCL1_CNTL1__RESP_MODE_MASK 0x00000018L +#define SQC_DCACHE_UTCL1_CNTL1__RESP_FAULT_MODE_MASK 0x00000060L +#define SQC_DCACHE_UTCL1_CNTL1__CLIENTID_MASK 0x0000FF80L +#define SQC_DCACHE_UTCL1_CNTL1__ENABLE_PUSH_LFIFO_MASK 0x00020000L +#define SQC_DCACHE_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB_MASK 0x00040000L +#define SQC_DCACHE_UTCL1_CNTL1__REG_INVALIDATE_VMID_MASK 0x00780000L +#define SQC_DCACHE_UTCL1_CNTL1__REG_INVALIDATE_ALL_VMID_MASK 0x00800000L +#define SQC_DCACHE_UTCL1_CNTL1__REG_INVALIDATE_TOGGLE_MASK 0x01000000L +#define SQC_DCACHE_UTCL1_CNTL1__CLIENT_INVALIDATE_ALL_VMID_MASK 0x02000000L +#define SQC_DCACHE_UTCL1_CNTL1__FORCE_MISS_MASK 0x04000000L +#define SQC_DCACHE_UTCL1_CNTL1__FORCE_IN_ORDER_MASK 0x08000000L +#define SQC_DCACHE_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2_MASK 0x30000000L +#define SQC_DCACHE_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2_MASK 0xC0000000L +//SQC_DCACHE_UTCL1_CNTL2 +#define SQC_DCACHE_UTCL1_CNTL2__SPARE__SHIFT 0x0 +#define SQC_DCACHE_UTCL1_CNTL2__LFIFO_SCAN_DISABLE__SHIFT 0x8 +#define SQC_DCACHE_UTCL1_CNTL2__MTYPE_OVRD_DIS__SHIFT 0x9 +#define SQC_DCACHE_UTCL1_CNTL2__LINE_VALID__SHIFT 0xa +#define SQC_DCACHE_UTCL1_CNTL2__DIS_EDC__SHIFT 0xb +#define SQC_DCACHE_UTCL1_CNTL2__GPUVM_INV_MODE__SHIFT 0xc +#define SQC_DCACHE_UTCL1_CNTL2__SHOOTDOWN_OPT__SHIFT 0xd +#define SQC_DCACHE_UTCL1_CNTL2__FORCE_SNOOP__SHIFT 0xe +#define SQC_DCACHE_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK__SHIFT 0xf +#define SQC_DCACHE_UTCL1_CNTL2__ARB_BURST_MODE__SHIFT 0x10 +#define SQC_DCACHE_UTCL1_CNTL2__ENABLE_PERF_EVENT_RD_WR__SHIFT 0x12 +#define SQC_DCACHE_UTCL1_CNTL2__PERF_EVENT_RD_WR__SHIFT 0x13 +#define SQC_DCACHE_UTCL1_CNTL2__ENABLE_PERF_EVENT_VMID__SHIFT 0x14 +#define SQC_DCACHE_UTCL1_CNTL2__PERF_EVENT_VMID__SHIFT 0x15 +#define SQC_DCACHE_UTCL1_CNTL2__FORCE_FRAG_2M_TO_64K__SHIFT 0x1a +#define SQC_DCACHE_UTCL1_CNTL2__SPARE_MASK 0x000000FFL +#define SQC_DCACHE_UTCL1_CNTL2__LFIFO_SCAN_DISABLE_MASK 0x00000100L +#define SQC_DCACHE_UTCL1_CNTL2__MTYPE_OVRD_DIS_MASK 0x00000200L +#define SQC_DCACHE_UTCL1_CNTL2__LINE_VALID_MASK 0x00000400L +#define SQC_DCACHE_UTCL1_CNTL2__DIS_EDC_MASK 0x00000800L +#define SQC_DCACHE_UTCL1_CNTL2__GPUVM_INV_MODE_MASK 0x00001000L +#define SQC_DCACHE_UTCL1_CNTL2__SHOOTDOWN_OPT_MASK 0x00002000L +#define SQC_DCACHE_UTCL1_CNTL2__FORCE_SNOOP_MASK 0x00004000L +#define SQC_DCACHE_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK_MASK 0x00008000L +#define SQC_DCACHE_UTCL1_CNTL2__ARB_BURST_MODE_MASK 0x00030000L +#define SQC_DCACHE_UTCL1_CNTL2__ENABLE_PERF_EVENT_RD_WR_MASK 0x00040000L +#define SQC_DCACHE_UTCL1_CNTL2__PERF_EVENT_RD_WR_MASK 0x00080000L +#define SQC_DCACHE_UTCL1_CNTL2__ENABLE_PERF_EVENT_VMID_MASK 0x00100000L +#define SQC_DCACHE_UTCL1_CNTL2__PERF_EVENT_VMID_MASK 0x01E00000L +#define SQC_DCACHE_UTCL1_CNTL2__FORCE_FRAG_2M_TO_64K_MASK 0x04000000L +//SQC_ICACHE_UTCL1_STATUS +#define SQC_ICACHE_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define SQC_ICACHE_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define SQC_ICACHE_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define SQC_ICACHE_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define SQC_ICACHE_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define SQC_ICACHE_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L +//SQC_DCACHE_UTCL1_STATUS +#define SQC_DCACHE_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define SQC_DCACHE_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define SQC_DCACHE_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define SQC_DCACHE_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define SQC_DCACHE_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define SQC_DCACHE_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L + + +// addressBlock: gc_shsdec +//SX_DEBUG_1 +#define SX_DEBUG_1__SX_DB_QUAD_CREDIT__SHIFT 0x0 +#define SX_DEBUG_1__DISABLE_BLEND_OPT_DONT_RD_DST__SHIFT 0x8 +#define SX_DEBUG_1__DISABLE_BLEND_OPT_BYPASS__SHIFT 0x9 +#define SX_DEBUG_1__DISABLE_BLEND_OPT_DISCARD_PIXEL__SHIFT 0xa +#define SX_DEBUG_1__DISABLE_QUAD_PAIR_OPT__SHIFT 0xb +#define SX_DEBUG_1__DISABLE_PIX_EN_ZERO_OPT__SHIFT 0xc +#define SX_DEBUG_1__PC_CFG__SHIFT 0xd +#define SX_DEBUG_1__DEBUG_DATA__SHIFT 0xe +#define SX_DEBUG_1__SX_DB_QUAD_CREDIT_MASK 0x0000007FL +#define SX_DEBUG_1__DISABLE_BLEND_OPT_DONT_RD_DST_MASK 0x00000100L +#define SX_DEBUG_1__DISABLE_BLEND_OPT_BYPASS_MASK 0x00000200L +#define SX_DEBUG_1__DISABLE_BLEND_OPT_DISCARD_PIXEL_MASK 0x00000400L +#define SX_DEBUG_1__DISABLE_QUAD_PAIR_OPT_MASK 0x00000800L +#define SX_DEBUG_1__DISABLE_PIX_EN_ZERO_OPT_MASK 0x00001000L +#define SX_DEBUG_1__PC_CFG_MASK 0x00002000L +#define SX_DEBUG_1__DEBUG_DATA_MASK 0xFFFFC000L +//SPI_PS_MAX_WAVE_ID +#define SPI_PS_MAX_WAVE_ID__MAX_WAVE_ID__SHIFT 0x0 +#define SPI_PS_MAX_WAVE_ID__MAX_COLLISION_WAVE_ID__SHIFT 0x10 +#define SPI_PS_MAX_WAVE_ID__MAX_WAVE_ID_MASK 0x00000FFFL +#define SPI_PS_MAX_WAVE_ID__MAX_COLLISION_WAVE_ID_MASK 0x03FF0000L +//SPI_START_PHASE +#define SPI_START_PHASE__VGPR_START_PHASE__SHIFT 0x0 +#define SPI_START_PHASE__SGPR_START_PHASE__SHIFT 0x2 +#define SPI_START_PHASE__WAVE_START_PHASE__SHIFT 0x4 +#define SPI_START_PHASE__VGPR_START_PHASE_MASK 0x00000003L +#define SPI_START_PHASE__SGPR_START_PHASE_MASK 0x0000000CL +#define SPI_START_PHASE__WAVE_START_PHASE_MASK 0x00000030L +//SPI_GFX_CNTL +#define SPI_GFX_CNTL__RESET_COUNTS__SHIFT 0x0 +#define SPI_GFX_CNTL__RESET_COUNTS_MASK 0x00000001L +//SPI_DSM_CNTL +#define SPI_DSM_CNTL__SPI_SR_MEM_DSM_IRRITATOR_DATA__SHIFT 0x0 +#define SPI_DSM_CNTL__SPI_SR_MEM_ENABLE_SINGLE_WRITE__SHIFT 0x2 +#define SPI_DSM_CNTL__UNUSED__SHIFT 0x3 +#define SPI_DSM_CNTL__SPI_SR_MEM_DSM_IRRITATOR_DATA_MASK 0x00000003L +#define SPI_DSM_CNTL__SPI_SR_MEM_ENABLE_SINGLE_WRITE_MASK 0x00000004L +#define SPI_DSM_CNTL__UNUSED_MASK 0xFFFFFFF8L +//SPI_DSM_CNTL2 +#define SPI_DSM_CNTL2__SPI_SR_MEM_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define SPI_DSM_CNTL2__SPI_SR_MEM_SELECT_INJECT_DELAY__SHIFT 0x2 +#define SPI_DSM_CNTL2__SPI_SR_MEM_INJECT_DELAY__SHIFT 0x4 +#define SPI_DSM_CNTL2__UNUSED__SHIFT 0xa +#define SPI_DSM_CNTL2__SPI_SR_MEM_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define SPI_DSM_CNTL2__SPI_SR_MEM_SELECT_INJECT_DELAY_MASK 0x00000004L +#define SPI_DSM_CNTL2__SPI_SR_MEM_INJECT_DELAY_MASK 0x000003F0L +#define SPI_DSM_CNTL2__UNUSED_MASK 0xFFFFFC00L +//SPI_EDC_CNT +#define SPI_EDC_CNT__SPI_SR_MEM_SED_COUNT__SHIFT 0x0 +#define SPI_EDC_CNT__SPI_SR_MEM_SED_COUNT_MASK 0x00000003L +//SPI_CONFIG_PS_CU_EN +#define SPI_CONFIG_PS_CU_EN__ENABLE__SHIFT 0x0 +#define SPI_CONFIG_PS_CU_EN__PKR0_CU_EN__SHIFT 0x1 +#define SPI_CONFIG_PS_CU_EN__PKR1_CU_EN__SHIFT 0x10 +#define SPI_CONFIG_PS_CU_EN__ENABLE_MASK 0x00000001L +#define SPI_CONFIG_PS_CU_EN__PKR0_CU_EN_MASK 0x0000FFFEL +#define SPI_CONFIG_PS_CU_EN__PKR1_CU_EN_MASK 0xFFFF0000L +//SPI_WF_LIFETIME_CNTL +#define SPI_WF_LIFETIME_CNTL__SAMPLE_PERIOD__SHIFT 0x0 +#define SPI_WF_LIFETIME_CNTL__EN__SHIFT 0x4 +#define SPI_WF_LIFETIME_CNTL__SAMPLE_PERIOD_MASK 0x0000000FL +#define SPI_WF_LIFETIME_CNTL__EN_MASK 0x00000010L +//SPI_WF_LIFETIME_LIMIT_0 +#define SPI_WF_LIFETIME_LIMIT_0__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_0__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_0__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_0__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_1 +#define SPI_WF_LIFETIME_LIMIT_1__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_1__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_1__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_1__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_2 +#define SPI_WF_LIFETIME_LIMIT_2__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_2__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_2__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_2__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_3 +#define SPI_WF_LIFETIME_LIMIT_3__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_3__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_3__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_3__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_4 +#define SPI_WF_LIFETIME_LIMIT_4__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_4__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_4__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_4__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_5 +#define SPI_WF_LIFETIME_LIMIT_5__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_5__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_5__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_5__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_6 +#define SPI_WF_LIFETIME_LIMIT_6__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_6__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_6__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_6__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_7 +#define SPI_WF_LIFETIME_LIMIT_7__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_7__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_7__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_7__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_8 +#define SPI_WF_LIFETIME_LIMIT_8__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_8__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_8__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_8__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_LIMIT_9 +#define SPI_WF_LIFETIME_LIMIT_9__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_LIMIT_9__EN_WARN__SHIFT 0x1f +#define SPI_WF_LIFETIME_LIMIT_9__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_LIMIT_9__EN_WARN_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_0 +#define SPI_WF_LIFETIME_STATUS_0__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_0__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_0__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_0__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_1 +#define SPI_WF_LIFETIME_STATUS_1__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_1__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_1__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_1__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_2 +#define SPI_WF_LIFETIME_STATUS_2__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_2__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_2__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_2__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_3 +#define SPI_WF_LIFETIME_STATUS_3__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_3__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_3__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_3__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_4 +#define SPI_WF_LIFETIME_STATUS_4__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_4__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_4__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_4__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_5 +#define SPI_WF_LIFETIME_STATUS_5__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_5__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_5__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_5__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_6 +#define SPI_WF_LIFETIME_STATUS_6__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_6__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_6__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_6__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_7 +#define SPI_WF_LIFETIME_STATUS_7__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_7__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_7__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_7__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_8 +#define SPI_WF_LIFETIME_STATUS_8__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_8__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_8__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_8__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_9 +#define SPI_WF_LIFETIME_STATUS_9__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_9__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_9__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_9__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_10 +#define SPI_WF_LIFETIME_STATUS_10__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_10__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_10__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_10__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_11 +#define SPI_WF_LIFETIME_STATUS_11__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_11__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_11__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_11__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_12 +#define SPI_WF_LIFETIME_STATUS_12__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_12__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_12__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_12__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_13 +#define SPI_WF_LIFETIME_STATUS_13__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_13__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_13__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_13__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_14 +#define SPI_WF_LIFETIME_STATUS_14__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_14__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_14__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_14__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_15 +#define SPI_WF_LIFETIME_STATUS_15__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_15__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_15__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_15__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_16 +#define SPI_WF_LIFETIME_STATUS_16__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_16__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_16__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_16__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_17 +#define SPI_WF_LIFETIME_STATUS_17__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_17__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_17__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_17__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_18 +#define SPI_WF_LIFETIME_STATUS_18__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_18__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_18__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_18__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_19 +#define SPI_WF_LIFETIME_STATUS_19__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_19__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_19__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_19__INT_SENT_MASK 0x80000000L +//SPI_WF_LIFETIME_STATUS_20 +#define SPI_WF_LIFETIME_STATUS_20__MAX_CNT__SHIFT 0x0 +#define SPI_WF_LIFETIME_STATUS_20__INT_SENT__SHIFT 0x1f +#define SPI_WF_LIFETIME_STATUS_20__MAX_CNT_MASK 0x7FFFFFFFL +#define SPI_WF_LIFETIME_STATUS_20__INT_SENT_MASK 0x80000000L +//SPI_LB_CTR_CTRL +#define SPI_LB_CTR_CTRL__LOAD__SHIFT 0x0 +#define SPI_LB_CTR_CTRL__WAVES_SELECT__SHIFT 0x1 +#define SPI_LB_CTR_CTRL__CLEAR_ON_READ__SHIFT 0x3 +#define SPI_LB_CTR_CTRL__RESET_COUNTS__SHIFT 0x4 +#define SPI_LB_CTR_CTRL__LOAD_MASK 0x00000001L +#define SPI_LB_CTR_CTRL__WAVES_SELECT_MASK 0x00000006L +#define SPI_LB_CTR_CTRL__CLEAR_ON_READ_MASK 0x00000008L +#define SPI_LB_CTR_CTRL__RESET_COUNTS_MASK 0x00000010L +//SPI_LB_CU_MASK +#define SPI_LB_CU_MASK__CU_MASK__SHIFT 0x0 +#define SPI_LB_CU_MASK__CU_MASK_MASK 0xFFFFL +//SPI_LB_DATA_REG +#define SPI_LB_DATA_REG__CNT_DATA__SHIFT 0x0 +#define SPI_LB_DATA_REG__CNT_DATA_MASK 0xFFFFFFFFL +//SPI_PG_ENABLE_STATIC_CU_MASK +#define SPI_PG_ENABLE_STATIC_CU_MASK__CU_MASK__SHIFT 0x0 +#define SPI_PG_ENABLE_STATIC_CU_MASK__CU_MASK_MASK 0xFFFFL +//SPI_GDS_CREDITS +#define SPI_GDS_CREDITS__DS_DATA_CREDITS__SHIFT 0x0 +#define SPI_GDS_CREDITS__DS_CMD_CREDITS__SHIFT 0x8 +#define SPI_GDS_CREDITS__UNUSED__SHIFT 0x10 +#define SPI_GDS_CREDITS__DS_DATA_CREDITS_MASK 0x000000FFL +#define SPI_GDS_CREDITS__DS_CMD_CREDITS_MASK 0x0000FF00L +#define SPI_GDS_CREDITS__UNUSED_MASK 0xFFFF0000L +//SPI_SX_EXPORT_BUFFER_SIZES +#define SPI_SX_EXPORT_BUFFER_SIZES__COLOR_BUFFER_SIZE__SHIFT 0x0 +#define SPI_SX_EXPORT_BUFFER_SIZES__POSITION_BUFFER_SIZE__SHIFT 0x10 +#define SPI_SX_EXPORT_BUFFER_SIZES__COLOR_BUFFER_SIZE_MASK 0x0000FFFFL +#define SPI_SX_EXPORT_BUFFER_SIZES__POSITION_BUFFER_SIZE_MASK 0xFFFF0000L +//SPI_SX_SCOREBOARD_BUFFER_SIZES +#define SPI_SX_SCOREBOARD_BUFFER_SIZES__COLOR_SCOREBOARD_SIZE__SHIFT 0x0 +#define SPI_SX_SCOREBOARD_BUFFER_SIZES__POSITION_SCOREBOARD_SIZE__SHIFT 0x10 +#define SPI_SX_SCOREBOARD_BUFFER_SIZES__COLOR_SCOREBOARD_SIZE_MASK 0x0000FFFFL +#define SPI_SX_SCOREBOARD_BUFFER_SIZES__POSITION_SCOREBOARD_SIZE_MASK 0xFFFF0000L +//SPI_CSQ_WF_ACTIVE_STATUS +#define SPI_CSQ_WF_ACTIVE_STATUS__ACTIVE__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_STATUS__ACTIVE_MASK 0xFFFFFFFFL +//SPI_CSQ_WF_ACTIVE_COUNT_0 +#define SPI_CSQ_WF_ACTIVE_COUNT_0__COUNT__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_COUNT_0__EVENTS__SHIFT 0x10 +#define SPI_CSQ_WF_ACTIVE_COUNT_0__COUNT_MASK 0x000007FFL +#define SPI_CSQ_WF_ACTIVE_COUNT_0__EVENTS_MASK 0x07FF0000L +//SPI_CSQ_WF_ACTIVE_COUNT_1 +#define SPI_CSQ_WF_ACTIVE_COUNT_1__COUNT__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_COUNT_1__EVENTS__SHIFT 0x10 +#define SPI_CSQ_WF_ACTIVE_COUNT_1__COUNT_MASK 0x000007FFL +#define SPI_CSQ_WF_ACTIVE_COUNT_1__EVENTS_MASK 0x07FF0000L +//SPI_CSQ_WF_ACTIVE_COUNT_2 +#define SPI_CSQ_WF_ACTIVE_COUNT_2__COUNT__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_COUNT_2__EVENTS__SHIFT 0x10 +#define SPI_CSQ_WF_ACTIVE_COUNT_2__COUNT_MASK 0x000007FFL +#define SPI_CSQ_WF_ACTIVE_COUNT_2__EVENTS_MASK 0x07FF0000L +//SPI_CSQ_WF_ACTIVE_COUNT_3 +#define SPI_CSQ_WF_ACTIVE_COUNT_3__COUNT__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_COUNT_3__EVENTS__SHIFT 0x10 +#define SPI_CSQ_WF_ACTIVE_COUNT_3__COUNT_MASK 0x000007FFL +#define SPI_CSQ_WF_ACTIVE_COUNT_3__EVENTS_MASK 0x07FF0000L +//SPI_CSQ_WF_ACTIVE_COUNT_4 +#define SPI_CSQ_WF_ACTIVE_COUNT_4__COUNT__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_COUNT_4__EVENTS__SHIFT 0x10 +#define SPI_CSQ_WF_ACTIVE_COUNT_4__COUNT_MASK 0x000007FFL +#define SPI_CSQ_WF_ACTIVE_COUNT_4__EVENTS_MASK 0x07FF0000L +//SPI_CSQ_WF_ACTIVE_COUNT_5 +#define SPI_CSQ_WF_ACTIVE_COUNT_5__COUNT__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_COUNT_5__EVENTS__SHIFT 0x10 +#define SPI_CSQ_WF_ACTIVE_COUNT_5__COUNT_MASK 0x000007FFL +#define SPI_CSQ_WF_ACTIVE_COUNT_5__EVENTS_MASK 0x07FF0000L +//SPI_CSQ_WF_ACTIVE_COUNT_6 +#define SPI_CSQ_WF_ACTIVE_COUNT_6__COUNT__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_COUNT_6__EVENTS__SHIFT 0x10 +#define SPI_CSQ_WF_ACTIVE_COUNT_6__COUNT_MASK 0x000007FFL +#define SPI_CSQ_WF_ACTIVE_COUNT_6__EVENTS_MASK 0x07FF0000L +//SPI_CSQ_WF_ACTIVE_COUNT_7 +#define SPI_CSQ_WF_ACTIVE_COUNT_7__COUNT__SHIFT 0x0 +#define SPI_CSQ_WF_ACTIVE_COUNT_7__EVENTS__SHIFT 0x10 +#define SPI_CSQ_WF_ACTIVE_COUNT_7__COUNT_MASK 0x000007FFL +#define SPI_CSQ_WF_ACTIVE_COUNT_7__EVENTS_MASK 0x07FF0000L +//SPI_LB_DATA_WAVES +#define SPI_LB_DATA_WAVES__COUNT0__SHIFT 0x0 +#define SPI_LB_DATA_WAVES__COUNT1__SHIFT 0x10 +#define SPI_LB_DATA_WAVES__COUNT0_MASK 0x0000FFFFL +#define SPI_LB_DATA_WAVES__COUNT1_MASK 0xFFFF0000L +//SPI_LB_DATA_PERCU_WAVE_HSGS +#define SPI_LB_DATA_PERCU_WAVE_HSGS__CU_USED_HS__SHIFT 0x0 +#define SPI_LB_DATA_PERCU_WAVE_HSGS__CU_USED_GS__SHIFT 0x10 +#define SPI_LB_DATA_PERCU_WAVE_HSGS__CU_USED_HS_MASK 0x0000FFFFL +#define SPI_LB_DATA_PERCU_WAVE_HSGS__CU_USED_GS_MASK 0xFFFF0000L +//SPI_LB_DATA_PERCU_WAVE_VSPS +#define SPI_LB_DATA_PERCU_WAVE_VSPS__CU_USED_VS__SHIFT 0x0 +#define SPI_LB_DATA_PERCU_WAVE_VSPS__CU_USED_PS__SHIFT 0x10 +#define SPI_LB_DATA_PERCU_WAVE_VSPS__CU_USED_VS_MASK 0x0000FFFFL +#define SPI_LB_DATA_PERCU_WAVE_VSPS__CU_USED_PS_MASK 0xFFFF0000L +//SPI_LB_DATA_PERCU_WAVE_CS +#define SPI_LB_DATA_PERCU_WAVE_CS__ACTIVE__SHIFT 0x0 +#define SPI_LB_DATA_PERCU_WAVE_CS__ACTIVE_MASK 0xFFFFL +//SPI_P0_TRAP_SCREEN_PSBA_LO +#define SPI_P0_TRAP_SCREEN_PSBA_LO__MEM_BASE__SHIFT 0x0 +#define SPI_P0_TRAP_SCREEN_PSBA_LO__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_P0_TRAP_SCREEN_PSBA_HI +#define SPI_P0_TRAP_SCREEN_PSBA_HI__MEM_BASE__SHIFT 0x0 +#define SPI_P0_TRAP_SCREEN_PSBA_HI__MEM_BASE_MASK 0xFFL +//SPI_P0_TRAP_SCREEN_PSMA_LO +#define SPI_P0_TRAP_SCREEN_PSMA_LO__MEM_BASE__SHIFT 0x0 +#define SPI_P0_TRAP_SCREEN_PSMA_LO__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_P0_TRAP_SCREEN_PSMA_HI +#define SPI_P0_TRAP_SCREEN_PSMA_HI__MEM_BASE__SHIFT 0x0 +#define SPI_P0_TRAP_SCREEN_PSMA_HI__MEM_BASE_MASK 0xFFL +//SPI_P0_TRAP_SCREEN_GPR_MIN +#define SPI_P0_TRAP_SCREEN_GPR_MIN__VGPR_MIN__SHIFT 0x0 +#define SPI_P0_TRAP_SCREEN_GPR_MIN__SGPR_MIN__SHIFT 0x6 +#define SPI_P0_TRAP_SCREEN_GPR_MIN__VGPR_MIN_MASK 0x003FL +#define SPI_P0_TRAP_SCREEN_GPR_MIN__SGPR_MIN_MASK 0x03C0L +//SPI_P1_TRAP_SCREEN_PSBA_LO +#define SPI_P1_TRAP_SCREEN_PSBA_LO__MEM_BASE__SHIFT 0x0 +#define SPI_P1_TRAP_SCREEN_PSBA_LO__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_P1_TRAP_SCREEN_PSBA_HI +#define SPI_P1_TRAP_SCREEN_PSBA_HI__MEM_BASE__SHIFT 0x0 +#define SPI_P1_TRAP_SCREEN_PSBA_HI__MEM_BASE_MASK 0xFFL +//SPI_P1_TRAP_SCREEN_PSMA_LO +#define SPI_P1_TRAP_SCREEN_PSMA_LO__MEM_BASE__SHIFT 0x0 +#define SPI_P1_TRAP_SCREEN_PSMA_LO__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_P1_TRAP_SCREEN_PSMA_HI +#define SPI_P1_TRAP_SCREEN_PSMA_HI__MEM_BASE__SHIFT 0x0 +#define SPI_P1_TRAP_SCREEN_PSMA_HI__MEM_BASE_MASK 0xFFL +//SPI_P1_TRAP_SCREEN_GPR_MIN +#define SPI_P1_TRAP_SCREEN_GPR_MIN__VGPR_MIN__SHIFT 0x0 +#define SPI_P1_TRAP_SCREEN_GPR_MIN__SGPR_MIN__SHIFT 0x6 +#define SPI_P1_TRAP_SCREEN_GPR_MIN__VGPR_MIN_MASK 0x003FL +#define SPI_P1_TRAP_SCREEN_GPR_MIN__SGPR_MIN_MASK 0x03C0L + + +// addressBlock: gc_tpdec +//TD_CNTL +#define TD_CNTL__SYNC_PHASE_SH__SHIFT 0x0 +#define TD_CNTL__SYNC_PHASE_VC_SMX__SHIFT 0x4 +#define TD_CNTL__PAD_STALL_EN__SHIFT 0x8 +#define TD_CNTL__EXTEND_LDS_STALL__SHIFT 0x9 +#define TD_CNTL__LDS_STALL_PHASE_ADJUST__SHIFT 0xb +#define TD_CNTL__PRECISION_COMPATIBILITY__SHIFT 0xf +#define TD_CNTL__GATHER4_FLOAT_MODE__SHIFT 0x10 +#define TD_CNTL__LD_FLOAT_MODE__SHIFT 0x12 +#define TD_CNTL__GATHER4_DX9_MODE__SHIFT 0x13 +#define TD_CNTL__DISABLE_POWER_THROTTLE__SHIFT 0x14 +#define TD_CNTL__ENABLE_ROUND_TO_ZERO__SHIFT 0x15 +#define TD_CNTL__DISABLE_2BIT_SIGNED_FORMAT__SHIFT 0x17 +#define TD_CNTL__DISABLE_MM_QNAN_COMPARE_RESULT__SHIFT 0x18 +#define TD_CNTL__SYNC_PHASE_SH_MASK 0x00000003L +#define TD_CNTL__SYNC_PHASE_VC_SMX_MASK 0x00000030L +#define TD_CNTL__PAD_STALL_EN_MASK 0x00000100L +#define TD_CNTL__EXTEND_LDS_STALL_MASK 0x00000600L +#define TD_CNTL__LDS_STALL_PHASE_ADJUST_MASK 0x00001800L +#define TD_CNTL__PRECISION_COMPATIBILITY_MASK 0x00008000L +#define TD_CNTL__GATHER4_FLOAT_MODE_MASK 0x00010000L +#define TD_CNTL__LD_FLOAT_MODE_MASK 0x00040000L +#define TD_CNTL__GATHER4_DX9_MODE_MASK 0x00080000L +#define TD_CNTL__DISABLE_POWER_THROTTLE_MASK 0x00100000L +#define TD_CNTL__ENABLE_ROUND_TO_ZERO_MASK 0x00200000L +#define TD_CNTL__DISABLE_2BIT_SIGNED_FORMAT_MASK 0x00800000L +#define TD_CNTL__DISABLE_MM_QNAN_COMPARE_RESULT_MASK 0x01000000L +//TD_STATUS +#define TD_STATUS__BUSY__SHIFT 0x1f +#define TD_STATUS__BUSY_MASK 0x80000000L +//TD_DSM_CNTL +#define TD_DSM_CNTL__TD_SS_FIFO_LO_DSM_IRRITATOR_DATA__SHIFT 0x0 +#define TD_DSM_CNTL__TD_SS_FIFO_LO_ENABLE_SINGLE_WRITE__SHIFT 0x2 +#define TD_DSM_CNTL__TD_SS_FIFO_HI_DSM_IRRITATOR_DATA__SHIFT 0x3 +#define TD_DSM_CNTL__TD_SS_FIFO_HI_ENABLE_SINGLE_WRITE__SHIFT 0x5 +#define TD_DSM_CNTL__TD_CS_FIFO_DSM_IRRITATOR_DATA__SHIFT 0x6 +#define TD_DSM_CNTL__TD_CS_FIFO_ENABLE_SINGLE_WRITE__SHIFT 0x8 +#define TD_DSM_CNTL__TD_SS_FIFO_LO_DSM_IRRITATOR_DATA_MASK 0x00000003L +#define TD_DSM_CNTL__TD_SS_FIFO_LO_ENABLE_SINGLE_WRITE_MASK 0x00000004L +#define TD_DSM_CNTL__TD_SS_FIFO_HI_DSM_IRRITATOR_DATA_MASK 0x00000018L +#define TD_DSM_CNTL__TD_SS_FIFO_HI_ENABLE_SINGLE_WRITE_MASK 0x00000020L +#define TD_DSM_CNTL__TD_CS_FIFO_DSM_IRRITATOR_DATA_MASK 0x000000C0L +#define TD_DSM_CNTL__TD_CS_FIFO_ENABLE_SINGLE_WRITE_MASK 0x00000100L +//TD_DSM_CNTL2 +#define TD_DSM_CNTL2__TD_SS_FIFO_LO_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define TD_DSM_CNTL2__TD_SS_FIFO_LO_SELECT_INJECT_DELAY__SHIFT 0x2 +#define TD_DSM_CNTL2__TD_SS_FIFO_HI_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define TD_DSM_CNTL2__TD_SS_FIFO_HI_SELECT_INJECT_DELAY__SHIFT 0x5 +#define TD_DSM_CNTL2__TD_CS_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define TD_DSM_CNTL2__TD_CS_FIFO_SELECT_INJECT_DELAY__SHIFT 0x8 +#define TD_DSM_CNTL2__TD_INJECT_DELAY__SHIFT 0x1a +#define TD_DSM_CNTL2__TD_SS_FIFO_LO_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define TD_DSM_CNTL2__TD_SS_FIFO_LO_SELECT_INJECT_DELAY_MASK 0x00000004L +#define TD_DSM_CNTL2__TD_SS_FIFO_HI_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define TD_DSM_CNTL2__TD_SS_FIFO_HI_SELECT_INJECT_DELAY_MASK 0x00000020L +#define TD_DSM_CNTL2__TD_CS_FIFO_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define TD_DSM_CNTL2__TD_CS_FIFO_SELECT_INJECT_DELAY_MASK 0x00000100L +#define TD_DSM_CNTL2__TD_INJECT_DELAY_MASK 0xFC000000L +//TD_SCRATCH +#define TD_SCRATCH__SCRATCH__SHIFT 0x0 +#define TD_SCRATCH__SCRATCH_MASK 0xFFFFFFFFL +//TA_CNTL +#define TA_CNTL__FX_XNACK_CREDIT__SHIFT 0x0 +#define TA_CNTL__SQ_XNACK_CREDIT__SHIFT 0x9 +#define TA_CNTL__TC_DATA_CREDIT__SHIFT 0xd +#define TA_CNTL__ALIGNER_CREDIT__SHIFT 0x10 +#define TA_CNTL__TD_FIFO_CREDIT__SHIFT 0x16 +#define TA_CNTL__FX_XNACK_CREDIT_MASK 0x0000007FL +#define TA_CNTL__SQ_XNACK_CREDIT_MASK 0x00001E00L +#define TA_CNTL__TC_DATA_CREDIT_MASK 0x0000E000L +#define TA_CNTL__ALIGNER_CREDIT_MASK 0x001F0000L +#define TA_CNTL__TD_FIFO_CREDIT_MASK 0xFFC00000L +//TA_CNTL_AUX +#define TA_CNTL_AUX__SCOAL_DSWIZZLE_N__SHIFT 0x0 +#define TA_CNTL_AUX__RESERVED__SHIFT 0x1 +#define TA_CNTL_AUX__TFAULT_EN_OVERRIDE__SHIFT 0x5 +#define TA_CNTL_AUX__GATHERH_DST_SEL__SHIFT 0x6 +#define TA_CNTL_AUX__DISABLE_GATHER4_BC_SWIZZLE__SHIFT 0x7 +#define TA_CNTL_AUX__NONIMG_ANISO_BYPASS__SHIFT 0x9 +#define TA_CNTL_AUX__ANISO_HALF_THRESH__SHIFT 0xa +#define TA_CNTL_AUX__ANISO_ERROR_FP_VBIAS__SHIFT 0xc +#define TA_CNTL_AUX__ANISO_STEP_ORDER__SHIFT 0xd +#define TA_CNTL_AUX__ANISO_STEP__SHIFT 0xe +#define TA_CNTL_AUX__MINMAG_UNNORM__SHIFT 0xf +#define TA_CNTL_AUX__ANISO_WEIGHT_MODE__SHIFT 0x10 +#define TA_CNTL_AUX__ANISO_RATIO_LUT__SHIFT 0x11 +#define TA_CNTL_AUX__ANISO_TAP__SHIFT 0x12 +#define TA_CNTL_AUX__ANISO_MIP_ADJ_MODE__SHIFT 0x13 +#define TA_CNTL_AUX__DETERMINISM_RESERVED_DISABLE__SHIFT 0x14 +#define TA_CNTL_AUX__DETERMINISM_OPCODE_STRICT_DISABLE__SHIFT 0x15 +#define TA_CNTL_AUX__DETERMINISM_MISC_DISABLE__SHIFT 0x16 +#define TA_CNTL_AUX__DETERMINISM_SAMPLE_C_DFMT_DISABLE__SHIFT 0x17 +#define TA_CNTL_AUX__DETERMINISM_SAMPLER_MSAA_DISABLE__SHIFT 0x18 +#define TA_CNTL_AUX__DETERMINISM_WRITEOP_READFMT_DISABLE__SHIFT 0x19 +#define TA_CNTL_AUX__DETERMINISM_DFMT_NFMT_DISABLE__SHIFT 0x1a +#define TA_CNTL_AUX__DISABLE_DWORD_X2_COALESCE__SHIFT 0x1b +#define TA_CNTL_AUX__CUBEMAP_SLICE_CLAMP__SHIFT 0x1c +#define TA_CNTL_AUX__TRUNC_SMALL_NEG__SHIFT 0x1d +#define TA_CNTL_AUX__ARRAY_ROUND_MODE__SHIFT 0x1e +#define TA_CNTL_AUX__SCOAL_DSWIZZLE_N_MASK 0x00000001L +#define TA_CNTL_AUX__RESERVED_MASK 0x0000000EL +#define TA_CNTL_AUX__TFAULT_EN_OVERRIDE_MASK 0x00000020L +#define TA_CNTL_AUX__GATHERH_DST_SEL_MASK 0x00000040L +#define TA_CNTL_AUX__DISABLE_GATHER4_BC_SWIZZLE_MASK 0x00000080L +#define TA_CNTL_AUX__NONIMG_ANISO_BYPASS_MASK 0x00000200L +#define TA_CNTL_AUX__ANISO_HALF_THRESH_MASK 0x00000C00L +#define TA_CNTL_AUX__ANISO_ERROR_FP_VBIAS_MASK 0x00001000L +#define TA_CNTL_AUX__ANISO_STEP_ORDER_MASK 0x00002000L +#define TA_CNTL_AUX__ANISO_STEP_MASK 0x00004000L +#define TA_CNTL_AUX__MINMAG_UNNORM_MASK 0x00008000L +#define TA_CNTL_AUX__ANISO_WEIGHT_MODE_MASK 0x00010000L +#define TA_CNTL_AUX__ANISO_RATIO_LUT_MASK 0x00020000L +#define TA_CNTL_AUX__ANISO_TAP_MASK 0x00040000L +#define TA_CNTL_AUX__ANISO_MIP_ADJ_MODE_MASK 0x00080000L +#define TA_CNTL_AUX__DETERMINISM_RESERVED_DISABLE_MASK 0x00100000L +#define TA_CNTL_AUX__DETERMINISM_OPCODE_STRICT_DISABLE_MASK 0x00200000L +#define TA_CNTL_AUX__DETERMINISM_MISC_DISABLE_MASK 0x00400000L +#define TA_CNTL_AUX__DETERMINISM_SAMPLE_C_DFMT_DISABLE_MASK 0x00800000L +#define TA_CNTL_AUX__DETERMINISM_SAMPLER_MSAA_DISABLE_MASK 0x01000000L +#define TA_CNTL_AUX__DETERMINISM_WRITEOP_READFMT_DISABLE_MASK 0x02000000L +#define TA_CNTL_AUX__DETERMINISM_DFMT_NFMT_DISABLE_MASK 0x04000000L +#define TA_CNTL_AUX__DISABLE_DWORD_X2_COALESCE_MASK 0x08000000L +#define TA_CNTL_AUX__CUBEMAP_SLICE_CLAMP_MASK 0x10000000L +#define TA_CNTL_AUX__TRUNC_SMALL_NEG_MASK 0x20000000L +#define TA_CNTL_AUX__ARRAY_ROUND_MODE_MASK 0xC0000000L +//TA_RESERVED_010C +#define TA_RESERVED_010C__Unused__SHIFT 0x0 +#define TA_RESERVED_010C__Unused_MASK 0xFFFFFFFFL +//TA_STATUS +#define TA_STATUS__FG_PFIFO_EMPTYB__SHIFT 0xc +#define TA_STATUS__FG_LFIFO_EMPTYB__SHIFT 0xd +#define TA_STATUS__FG_SFIFO_EMPTYB__SHIFT 0xe +#define TA_STATUS__FL_PFIFO_EMPTYB__SHIFT 0x10 +#define TA_STATUS__FL_LFIFO_EMPTYB__SHIFT 0x11 +#define TA_STATUS__FL_SFIFO_EMPTYB__SHIFT 0x12 +#define TA_STATUS__FA_PFIFO_EMPTYB__SHIFT 0x14 +#define TA_STATUS__FA_LFIFO_EMPTYB__SHIFT 0x15 +#define TA_STATUS__FA_SFIFO_EMPTYB__SHIFT 0x16 +#define TA_STATUS__IN_BUSY__SHIFT 0x18 +#define TA_STATUS__FG_BUSY__SHIFT 0x19 +#define TA_STATUS__LA_BUSY__SHIFT 0x1a +#define TA_STATUS__FL_BUSY__SHIFT 0x1b +#define TA_STATUS__TA_BUSY__SHIFT 0x1c +#define TA_STATUS__FA_BUSY__SHIFT 0x1d +#define TA_STATUS__AL_BUSY__SHIFT 0x1e +#define TA_STATUS__BUSY__SHIFT 0x1f +#define TA_STATUS__FG_PFIFO_EMPTYB_MASK 0x00001000L +#define TA_STATUS__FG_LFIFO_EMPTYB_MASK 0x00002000L +#define TA_STATUS__FG_SFIFO_EMPTYB_MASK 0x00004000L +#define TA_STATUS__FL_PFIFO_EMPTYB_MASK 0x00010000L +#define TA_STATUS__FL_LFIFO_EMPTYB_MASK 0x00020000L +#define TA_STATUS__FL_SFIFO_EMPTYB_MASK 0x00040000L +#define TA_STATUS__FA_PFIFO_EMPTYB_MASK 0x00100000L +#define TA_STATUS__FA_LFIFO_EMPTYB_MASK 0x00200000L +#define TA_STATUS__FA_SFIFO_EMPTYB_MASK 0x00400000L +#define TA_STATUS__IN_BUSY_MASK 0x01000000L +#define TA_STATUS__FG_BUSY_MASK 0x02000000L +#define TA_STATUS__LA_BUSY_MASK 0x04000000L +#define TA_STATUS__FL_BUSY_MASK 0x08000000L +#define TA_STATUS__TA_BUSY_MASK 0x10000000L +#define TA_STATUS__FA_BUSY_MASK 0x20000000L +#define TA_STATUS__AL_BUSY_MASK 0x40000000L +#define TA_STATUS__BUSY_MASK 0x80000000L +//TA_SCRATCH +#define TA_SCRATCH__SCRATCH__SHIFT 0x0 +#define TA_SCRATCH__SCRATCH_MASK 0xFFFFFFFFL + + +// addressBlock: gc_gdsdec +//GDS_CONFIG +#define GDS_CONFIG__SH0_GPR_PHASE_SEL__SHIFT 0x1 +#define GDS_CONFIG__SH1_GPR_PHASE_SEL__SHIFT 0x3 +#define GDS_CONFIG__SH2_GPR_PHASE_SEL__SHIFT 0x5 +#define GDS_CONFIG__SH3_GPR_PHASE_SEL__SHIFT 0x7 +#define GDS_CONFIG__SH0_GPR_PHASE_SEL_MASK 0x00000006L +#define GDS_CONFIG__SH1_GPR_PHASE_SEL_MASK 0x00000018L +#define GDS_CONFIG__SH2_GPR_PHASE_SEL_MASK 0x00000060L +#define GDS_CONFIG__SH3_GPR_PHASE_SEL_MASK 0x00000180L +//GDS_CNTL_STATUS +#define GDS_CNTL_STATUS__GDS_BUSY__SHIFT 0x0 +#define GDS_CNTL_STATUS__GRBM_WBUF_BUSY__SHIFT 0x1 +#define GDS_CNTL_STATUS__ORD_APP_BUSY__SHIFT 0x2 +#define GDS_CNTL_STATUS__DS_BANK_CONFLICT__SHIFT 0x3 +#define GDS_CNTL_STATUS__DS_ADDR_CONFLICT__SHIFT 0x4 +#define GDS_CNTL_STATUS__DS_WR_CLAMP__SHIFT 0x5 +#define GDS_CNTL_STATUS__DS_RD_CLAMP__SHIFT 0x6 +#define GDS_CNTL_STATUS__GRBM_RBUF_BUSY__SHIFT 0x7 +#define GDS_CNTL_STATUS__DS_BUSY__SHIFT 0x8 +#define GDS_CNTL_STATUS__GWS_BUSY__SHIFT 0x9 +#define GDS_CNTL_STATUS__ORD_FIFO_BUSY__SHIFT 0xa +#define GDS_CNTL_STATUS__CREDIT_BUSY0__SHIFT 0xb +#define GDS_CNTL_STATUS__CREDIT_BUSY1__SHIFT 0xc +#define GDS_CNTL_STATUS__CREDIT_BUSY2__SHIFT 0xd +#define GDS_CNTL_STATUS__CREDIT_BUSY3__SHIFT 0xe +#define GDS_CNTL_STATUS__GDS_BUSY_MASK 0x00000001L +#define GDS_CNTL_STATUS__GRBM_WBUF_BUSY_MASK 0x00000002L +#define GDS_CNTL_STATUS__ORD_APP_BUSY_MASK 0x00000004L +#define GDS_CNTL_STATUS__DS_BANK_CONFLICT_MASK 0x00000008L +#define GDS_CNTL_STATUS__DS_ADDR_CONFLICT_MASK 0x00000010L +#define GDS_CNTL_STATUS__DS_WR_CLAMP_MASK 0x00000020L +#define GDS_CNTL_STATUS__DS_RD_CLAMP_MASK 0x00000040L +#define GDS_CNTL_STATUS__GRBM_RBUF_BUSY_MASK 0x00000080L +#define GDS_CNTL_STATUS__DS_BUSY_MASK 0x00000100L +#define GDS_CNTL_STATUS__GWS_BUSY_MASK 0x00000200L +#define GDS_CNTL_STATUS__ORD_FIFO_BUSY_MASK 0x00000400L +#define GDS_CNTL_STATUS__CREDIT_BUSY0_MASK 0x00000800L +#define GDS_CNTL_STATUS__CREDIT_BUSY1_MASK 0x00001000L +#define GDS_CNTL_STATUS__CREDIT_BUSY2_MASK 0x00002000L +#define GDS_CNTL_STATUS__CREDIT_BUSY3_MASK 0x00004000L +//GDS_ENHANCE2 +#define GDS_ENHANCE2__MISC__SHIFT 0x0 +#define GDS_ENHANCE2__UNUSED__SHIFT 0x10 +#define GDS_ENHANCE2__MISC_MASK 0x0000FFFFL +#define GDS_ENHANCE2__UNUSED_MASK 0xFFFF0000L +//GDS_PROTECTION_FAULT +#define GDS_PROTECTION_FAULT__WRITE_DIS__SHIFT 0x0 +#define GDS_PROTECTION_FAULT__FAULT_DETECTED__SHIFT 0x1 +#define GDS_PROTECTION_FAULT__GRBM__SHIFT 0x2 +#define GDS_PROTECTION_FAULT__SH_ID__SHIFT 0x3 +#define GDS_PROTECTION_FAULT__CU_ID__SHIFT 0x6 +#define GDS_PROTECTION_FAULT__SIMD_ID__SHIFT 0xa +#define GDS_PROTECTION_FAULT__WAVE_ID__SHIFT 0xc +#define GDS_PROTECTION_FAULT__ADDRESS__SHIFT 0x10 +#define GDS_PROTECTION_FAULT__WRITE_DIS_MASK 0x00000001L +#define GDS_PROTECTION_FAULT__FAULT_DETECTED_MASK 0x00000002L +#define GDS_PROTECTION_FAULT__GRBM_MASK 0x00000004L +#define GDS_PROTECTION_FAULT__SH_ID_MASK 0x00000038L +#define GDS_PROTECTION_FAULT__CU_ID_MASK 0x000003C0L +#define GDS_PROTECTION_FAULT__SIMD_ID_MASK 0x00000C00L +#define GDS_PROTECTION_FAULT__WAVE_ID_MASK 0x0000F000L +#define GDS_PROTECTION_FAULT__ADDRESS_MASK 0xFFFF0000L +//GDS_VM_PROTECTION_FAULT +#define GDS_VM_PROTECTION_FAULT__WRITE_DIS__SHIFT 0x0 +#define GDS_VM_PROTECTION_FAULT__FAULT_DETECTED__SHIFT 0x1 +#define GDS_VM_PROTECTION_FAULT__GWS__SHIFT 0x2 +#define GDS_VM_PROTECTION_FAULT__OA__SHIFT 0x3 +#define GDS_VM_PROTECTION_FAULT__GRBM__SHIFT 0x4 +#define GDS_VM_PROTECTION_FAULT__TMZ__SHIFT 0x5 +#define GDS_VM_PROTECTION_FAULT__VMID__SHIFT 0x8 +#define GDS_VM_PROTECTION_FAULT__ADDRESS__SHIFT 0x10 +#define GDS_VM_PROTECTION_FAULT__WRITE_DIS_MASK 0x00000001L +#define GDS_VM_PROTECTION_FAULT__FAULT_DETECTED_MASK 0x00000002L +#define GDS_VM_PROTECTION_FAULT__GWS_MASK 0x00000004L +#define GDS_VM_PROTECTION_FAULT__OA_MASK 0x00000008L +#define GDS_VM_PROTECTION_FAULT__GRBM_MASK 0x00000010L +#define GDS_VM_PROTECTION_FAULT__TMZ_MASK 0x00000020L +#define GDS_VM_PROTECTION_FAULT__VMID_MASK 0x00000F00L +#define GDS_VM_PROTECTION_FAULT__ADDRESS_MASK 0xFFFF0000L +//GDS_EDC_CNT +#define GDS_EDC_CNT__GDS_MEM_DED__SHIFT 0x0 +#define GDS_EDC_CNT__GDS_INPUT_QUEUE_SED__SHIFT 0x2 +#define GDS_EDC_CNT__GDS_MEM_SEC__SHIFT 0x4 +#define GDS_EDC_CNT__UNUSED__SHIFT 0x6 +#define GDS_EDC_CNT__GDS_MEM_DED_MASK 0x00000003L +#define GDS_EDC_CNT__GDS_INPUT_QUEUE_SED_MASK 0x0000000CL +#define GDS_EDC_CNT__GDS_MEM_SEC_MASK 0x00000030L +#define GDS_EDC_CNT__UNUSED_MASK 0xFFFFFFC0L +//GDS_EDC_GRBM_CNT +#define GDS_EDC_GRBM_CNT__DED__SHIFT 0x0 +#define GDS_EDC_GRBM_CNT__SEC__SHIFT 0x2 +#define GDS_EDC_GRBM_CNT__UNUSED__SHIFT 0x4 +#define GDS_EDC_GRBM_CNT__DED_MASK 0x00000003L +#define GDS_EDC_GRBM_CNT__SEC_MASK 0x0000000CL +#define GDS_EDC_GRBM_CNT__UNUSED_MASK 0xFFFFFFF0L +//GDS_EDC_OA_DED +#define GDS_EDC_OA_DED__ME0_GFXHP3D_PIX_DED__SHIFT 0x0 +#define GDS_EDC_OA_DED__ME0_GFXHP3D_VTX_DED__SHIFT 0x1 +#define GDS_EDC_OA_DED__ME0_CS_DED__SHIFT 0x2 +#define GDS_EDC_OA_DED__ME0_GFXHP3D_GS_DED__SHIFT 0x3 +#define GDS_EDC_OA_DED__ME1_PIPE0_DED__SHIFT 0x4 +#define GDS_EDC_OA_DED__ME1_PIPE1_DED__SHIFT 0x5 +#define GDS_EDC_OA_DED__ME1_PIPE2_DED__SHIFT 0x6 +#define GDS_EDC_OA_DED__ME1_PIPE3_DED__SHIFT 0x7 +#define GDS_EDC_OA_DED__ME2_PIPE0_DED__SHIFT 0x8 +#define GDS_EDC_OA_DED__ME2_PIPE1_DED__SHIFT 0x9 +#define GDS_EDC_OA_DED__ME2_PIPE2_DED__SHIFT 0xa +#define GDS_EDC_OA_DED__ME2_PIPE3_DED__SHIFT 0xb +#define GDS_EDC_OA_DED__UNUSED1__SHIFT 0xc +#define GDS_EDC_OA_DED__ME0_GFXHP3D_PIX_DED_MASK 0x00000001L +#define GDS_EDC_OA_DED__ME0_GFXHP3D_VTX_DED_MASK 0x00000002L +#define GDS_EDC_OA_DED__ME0_CS_DED_MASK 0x00000004L +#define GDS_EDC_OA_DED__ME0_GFXHP3D_GS_DED_MASK 0x00000008L +#define GDS_EDC_OA_DED__ME1_PIPE0_DED_MASK 0x00000010L +#define GDS_EDC_OA_DED__ME1_PIPE1_DED_MASK 0x00000020L +#define GDS_EDC_OA_DED__ME1_PIPE2_DED_MASK 0x00000040L +#define GDS_EDC_OA_DED__ME1_PIPE3_DED_MASK 0x00000080L +#define GDS_EDC_OA_DED__ME2_PIPE0_DED_MASK 0x00000100L +#define GDS_EDC_OA_DED__ME2_PIPE1_DED_MASK 0x00000200L +#define GDS_EDC_OA_DED__ME2_PIPE2_DED_MASK 0x00000400L +#define GDS_EDC_OA_DED__ME2_PIPE3_DED_MASK 0x00000800L +#define GDS_EDC_OA_DED__UNUSED1_MASK 0xFFFFF000L +//GDS_DSM_CNTL +#define GDS_DSM_CNTL__SEL_DSM_GDS_MEM_IRRITATOR_DATA_0__SHIFT 0x0 +#define GDS_DSM_CNTL__SEL_DSM_GDS_MEM_IRRITATOR_DATA_1__SHIFT 0x1 +#define GDS_DSM_CNTL__GDS_MEM_ENABLE_SINGLE_WRITE__SHIFT 0x2 +#define GDS_DSM_CNTL__SEL_DSM_GDS_INPUT_QUEUE_IRRITATOR_DATA_0__SHIFT 0x3 +#define GDS_DSM_CNTL__SEL_DSM_GDS_INPUT_QUEUE_IRRITATOR_DATA_1__SHIFT 0x4 +#define GDS_DSM_CNTL__GDS_INPUT_QUEUE_ENABLE_SINGLE_WRITE__SHIFT 0x5 +#define GDS_DSM_CNTL__SEL_DSM_GDS_PHY_CMD_RAM_IRRITATOR_DATA_0__SHIFT 0x6 +#define GDS_DSM_CNTL__SEL_DSM_GDS_PHY_CMD_RAM_IRRITATOR_DATA_1__SHIFT 0x7 +#define GDS_DSM_CNTL__GDS_PHY_CMD_RAM_ENABLE_SINGLE_WRITE__SHIFT 0x8 +#define GDS_DSM_CNTL__SEL_DSM_GDS_PHY_DATA_RAM_IRRITATOR_DATA_0__SHIFT 0x9 +#define GDS_DSM_CNTL__SEL_DSM_GDS_PHY_DATA_RAM_IRRITATOR_DATA_1__SHIFT 0xa +#define GDS_DSM_CNTL__GDS_PHY_DATA_RAM_ENABLE_SINGLE_WRITE__SHIFT 0xb +#define GDS_DSM_CNTL__SEL_DSM_GDS_PIPE_MEM_IRRITATOR_DATA_0__SHIFT 0xc +#define GDS_DSM_CNTL__SEL_DSM_GDS_PIPE_MEM_IRRITATOR_DATA_1__SHIFT 0xd +#define GDS_DSM_CNTL__GDS_PIPE_MEM_ENABLE_SINGLE_WRITE__SHIFT 0xe +#define GDS_DSM_CNTL__UNUSED__SHIFT 0xf +#define GDS_DSM_CNTL__SEL_DSM_GDS_MEM_IRRITATOR_DATA_0_MASK 0x00000001L +#define GDS_DSM_CNTL__SEL_DSM_GDS_MEM_IRRITATOR_DATA_1_MASK 0x00000002L +#define GDS_DSM_CNTL__GDS_MEM_ENABLE_SINGLE_WRITE_MASK 0x00000004L +#define GDS_DSM_CNTL__SEL_DSM_GDS_INPUT_QUEUE_IRRITATOR_DATA_0_MASK 0x00000008L +#define GDS_DSM_CNTL__SEL_DSM_GDS_INPUT_QUEUE_IRRITATOR_DATA_1_MASK 0x00000010L +#define GDS_DSM_CNTL__GDS_INPUT_QUEUE_ENABLE_SINGLE_WRITE_MASK 0x00000020L +#define GDS_DSM_CNTL__SEL_DSM_GDS_PHY_CMD_RAM_IRRITATOR_DATA_0_MASK 0x00000040L +#define GDS_DSM_CNTL__SEL_DSM_GDS_PHY_CMD_RAM_IRRITATOR_DATA_1_MASK 0x00000080L +#define GDS_DSM_CNTL__GDS_PHY_CMD_RAM_ENABLE_SINGLE_WRITE_MASK 0x00000100L +#define GDS_DSM_CNTL__SEL_DSM_GDS_PHY_DATA_RAM_IRRITATOR_DATA_0_MASK 0x00000200L +#define GDS_DSM_CNTL__SEL_DSM_GDS_PHY_DATA_RAM_IRRITATOR_DATA_1_MASK 0x00000400L +#define GDS_DSM_CNTL__GDS_PHY_DATA_RAM_ENABLE_SINGLE_WRITE_MASK 0x00000800L +#define GDS_DSM_CNTL__SEL_DSM_GDS_PIPE_MEM_IRRITATOR_DATA_0_MASK 0x00001000L +#define GDS_DSM_CNTL__SEL_DSM_GDS_PIPE_MEM_IRRITATOR_DATA_1_MASK 0x00002000L +#define GDS_DSM_CNTL__GDS_PIPE_MEM_ENABLE_SINGLE_WRITE_MASK 0x00004000L +#define GDS_DSM_CNTL__UNUSED_MASK 0xFFFF8000L +//GDS_EDC_OA_PHY_CNT +#define GDS_EDC_OA_PHY_CNT__ME0_CS_PIPE_MEM_SEC__SHIFT 0x0 +#define GDS_EDC_OA_PHY_CNT__ME0_CS_PIPE_MEM_DED__SHIFT 0x2 +#define GDS_EDC_OA_PHY_CNT__PHY_CMD_RAM_MEM_SEC__SHIFT 0x4 +#define GDS_EDC_OA_PHY_CNT__PHY_CMD_RAM_MEM_DED__SHIFT 0x6 +#define GDS_EDC_OA_PHY_CNT__PHY_DATA_RAM_MEM_SED__SHIFT 0x8 +#define GDS_EDC_OA_PHY_CNT__UNUSED1__SHIFT 0xa +#define GDS_EDC_OA_PHY_CNT__ME0_CS_PIPE_MEM_SEC_MASK 0x00000003L +#define GDS_EDC_OA_PHY_CNT__ME0_CS_PIPE_MEM_DED_MASK 0x0000000CL +#define GDS_EDC_OA_PHY_CNT__PHY_CMD_RAM_MEM_SEC_MASK 0x00000030L +#define GDS_EDC_OA_PHY_CNT__PHY_CMD_RAM_MEM_DED_MASK 0x000000C0L +#define GDS_EDC_OA_PHY_CNT__PHY_DATA_RAM_MEM_SED_MASK 0x00000300L +#define GDS_EDC_OA_PHY_CNT__UNUSED1_MASK 0xFFFFFC00L +//GDS_EDC_OA_PIPE_CNT +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE0_PIPE_MEM_SEC__SHIFT 0x0 +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE0_PIPE_MEM_DED__SHIFT 0x2 +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE1_PIPE_MEM_SEC__SHIFT 0x4 +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE1_PIPE_MEM_DED__SHIFT 0x6 +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE2_PIPE_MEM_SEC__SHIFT 0x8 +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE2_PIPE_MEM_DED__SHIFT 0xa +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE3_PIPE_MEM_SEC__SHIFT 0xc +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE3_PIPE_MEM_DED__SHIFT 0xe +#define GDS_EDC_OA_PIPE_CNT__UNUSED__SHIFT 0x10 +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE0_PIPE_MEM_SEC_MASK 0x00000003L +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE0_PIPE_MEM_DED_MASK 0x0000000CL +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE1_PIPE_MEM_SEC_MASK 0x00000030L +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE1_PIPE_MEM_DED_MASK 0x000000C0L +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE2_PIPE_MEM_SEC_MASK 0x00000300L +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE2_PIPE_MEM_DED_MASK 0x00000C00L +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE3_PIPE_MEM_SEC_MASK 0x00003000L +#define GDS_EDC_OA_PIPE_CNT__ME1_PIPE3_PIPE_MEM_DED_MASK 0x0000C000L +#define GDS_EDC_OA_PIPE_CNT__UNUSED_MASK 0xFFFF0000L +//GDS_DSM_CNTL2 +#define GDS_DSM_CNTL2__GDS_MEM_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define GDS_DSM_CNTL2__GDS_MEM_SELECT_INJECT_DELAY__SHIFT 0x2 +#define GDS_DSM_CNTL2__GDS_INPUT_QUEUE_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define GDS_DSM_CNTL2__GDS_INPUT_QUEUE_SELECT_INJECT_DELAY__SHIFT 0x5 +#define GDS_DSM_CNTL2__GDS_PHY_CMD_RAM_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define GDS_DSM_CNTL2__GDS_PHY_CMD_RAM_SELECT_INJECT_DELAY__SHIFT 0x8 +#define GDS_DSM_CNTL2__GDS_PHY_DATA_RAM_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define GDS_DSM_CNTL2__GDS_PHY_DATA_RAM_SELECT_INJECT_DELAY__SHIFT 0xb +#define GDS_DSM_CNTL2__GDS_PIPE_MEM_ENABLE_ERROR_INJECT__SHIFT 0xc +#define GDS_DSM_CNTL2__GDS_PIPE_MEM_SELECT_INJECT_DELAY__SHIFT 0xe +#define GDS_DSM_CNTL2__UNUSED__SHIFT 0xf +#define GDS_DSM_CNTL2__GDS_INJECT_DELAY__SHIFT 0x1a +#define GDS_DSM_CNTL2__GDS_MEM_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define GDS_DSM_CNTL2__GDS_MEM_SELECT_INJECT_DELAY_MASK 0x00000004L +#define GDS_DSM_CNTL2__GDS_INPUT_QUEUE_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define GDS_DSM_CNTL2__GDS_INPUT_QUEUE_SELECT_INJECT_DELAY_MASK 0x00000020L +#define GDS_DSM_CNTL2__GDS_PHY_CMD_RAM_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define GDS_DSM_CNTL2__GDS_PHY_CMD_RAM_SELECT_INJECT_DELAY_MASK 0x00000100L +#define GDS_DSM_CNTL2__GDS_PHY_DATA_RAM_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define GDS_DSM_CNTL2__GDS_PHY_DATA_RAM_SELECT_INJECT_DELAY_MASK 0x00000800L +#define GDS_DSM_CNTL2__GDS_PIPE_MEM_ENABLE_ERROR_INJECT_MASK 0x00003000L +#define GDS_DSM_CNTL2__GDS_PIPE_MEM_SELECT_INJECT_DELAY_MASK 0x00004000L +#define GDS_DSM_CNTL2__UNUSED_MASK 0x03FF8000L +#define GDS_DSM_CNTL2__GDS_INJECT_DELAY_MASK 0xFC000000L +//GDS_WD_GDS_CSB +#define GDS_WD_GDS_CSB__COUNTER__SHIFT 0x0 +#define GDS_WD_GDS_CSB__UNUSED__SHIFT 0xd +#define GDS_WD_GDS_CSB__COUNTER_MASK 0x00001FFFL +#define GDS_WD_GDS_CSB__UNUSED_MASK 0xFFFFE000L + + +// addressBlock: gc_rbdec +//DB_DEBUG +#define DB_DEBUG__DEBUG_STENCIL_COMPRESS_DISABLE__SHIFT 0x0 +#define DB_DEBUG__DEBUG_DEPTH_COMPRESS_DISABLE__SHIFT 0x1 +#define DB_DEBUG__FETCH_FULL_Z_TILE__SHIFT 0x2 +#define DB_DEBUG__FETCH_FULL_STENCIL_TILE__SHIFT 0x3 +#define DB_DEBUG__FORCE_Z_MODE__SHIFT 0x4 +#define DB_DEBUG__DEBUG_FORCE_DEPTH_READ__SHIFT 0x6 +#define DB_DEBUG__DEBUG_FORCE_STENCIL_READ__SHIFT 0x7 +#define DB_DEBUG__DEBUG_FORCE_HIZ_ENABLE__SHIFT 0x8 +#define DB_DEBUG__DEBUG_FORCE_HIS_ENABLE0__SHIFT 0xa +#define DB_DEBUG__DEBUG_FORCE_HIS_ENABLE1__SHIFT 0xc +#define DB_DEBUG__DEBUG_FAST_Z_DISABLE__SHIFT 0xe +#define DB_DEBUG__DEBUG_FAST_STENCIL_DISABLE__SHIFT 0xf +#define DB_DEBUG__DEBUG_NOOP_CULL_DISABLE__SHIFT 0x10 +#define DB_DEBUG__DISABLE_SUMM_SQUADS__SHIFT 0x11 +#define DB_DEBUG__DEPTH_CACHE_FORCE_MISS__SHIFT 0x12 +#define DB_DEBUG__DEBUG_FORCE_FULL_Z_RANGE__SHIFT 0x13 +#define DB_DEBUG__NEVER_FREE_Z_ONLY__SHIFT 0x15 +#define DB_DEBUG__ZPASS_COUNTS_LOOK_AT_PIPE_STAT_EVENTS__SHIFT 0x16 +#define DB_DEBUG__DISABLE_VPORT_ZPLANE_OPTIMIZATION__SHIFT 0x17 +#define DB_DEBUG__DECOMPRESS_AFTER_N_ZPLANES__SHIFT 0x18 +#define DB_DEBUG__ONE_FREE_IN_FLIGHT__SHIFT 0x1c +#define DB_DEBUG__FORCE_MISS_IF_NOT_INFLIGHT__SHIFT 0x1d +#define DB_DEBUG__DISABLE_DEPTH_SURFACE_SYNC__SHIFT 0x1e +#define DB_DEBUG__DISABLE_HTILE_SURFACE_SYNC__SHIFT 0x1f +#define DB_DEBUG__DEBUG_STENCIL_COMPRESS_DISABLE_MASK 0x00000001L +#define DB_DEBUG__DEBUG_DEPTH_COMPRESS_DISABLE_MASK 0x00000002L +#define DB_DEBUG__FETCH_FULL_Z_TILE_MASK 0x00000004L +#define DB_DEBUG__FETCH_FULL_STENCIL_TILE_MASK 0x00000008L +#define DB_DEBUG__FORCE_Z_MODE_MASK 0x00000030L +#define DB_DEBUG__DEBUG_FORCE_DEPTH_READ_MASK 0x00000040L +#define DB_DEBUG__DEBUG_FORCE_STENCIL_READ_MASK 0x00000080L +#define DB_DEBUG__DEBUG_FORCE_HIZ_ENABLE_MASK 0x00000300L +#define DB_DEBUG__DEBUG_FORCE_HIS_ENABLE0_MASK 0x00000C00L +#define DB_DEBUG__DEBUG_FORCE_HIS_ENABLE1_MASK 0x00003000L +#define DB_DEBUG__DEBUG_FAST_Z_DISABLE_MASK 0x00004000L +#define DB_DEBUG__DEBUG_FAST_STENCIL_DISABLE_MASK 0x00008000L +#define DB_DEBUG__DEBUG_NOOP_CULL_DISABLE_MASK 0x00010000L +#define DB_DEBUG__DISABLE_SUMM_SQUADS_MASK 0x00020000L +#define DB_DEBUG__DEPTH_CACHE_FORCE_MISS_MASK 0x00040000L +#define DB_DEBUG__DEBUG_FORCE_FULL_Z_RANGE_MASK 0x00180000L +#define DB_DEBUG__NEVER_FREE_Z_ONLY_MASK 0x00200000L +#define DB_DEBUG__ZPASS_COUNTS_LOOK_AT_PIPE_STAT_EVENTS_MASK 0x00400000L +#define DB_DEBUG__DISABLE_VPORT_ZPLANE_OPTIMIZATION_MASK 0x00800000L +#define DB_DEBUG__DECOMPRESS_AFTER_N_ZPLANES_MASK 0x0F000000L +#define DB_DEBUG__ONE_FREE_IN_FLIGHT_MASK 0x10000000L +#define DB_DEBUG__FORCE_MISS_IF_NOT_INFLIGHT_MASK 0x20000000L +#define DB_DEBUG__DISABLE_DEPTH_SURFACE_SYNC_MASK 0x40000000L +#define DB_DEBUG__DISABLE_HTILE_SURFACE_SYNC_MASK 0x80000000L +//DB_DEBUG2 +#define DB_DEBUG2__ALLOW_COMPZ_BYTE_MASKING__SHIFT 0x0 +#define DB_DEBUG2__DISABLE_TC_ZRANGE_L0_CACHE__SHIFT 0x1 +#define DB_DEBUG2__DISABLE_TC_MASK_L0_CACHE__SHIFT 0x2 +#define DB_DEBUG2__DTR_ROUND_ROBIN_ARB__SHIFT 0x3 +#define DB_DEBUG2__DTR_PREZ_STALLS_FOR_ETF_ROOM__SHIFT 0x4 +#define DB_DEBUG2__DISABLE_PREZL_FIFO_STALL__SHIFT 0x5 +#define DB_DEBUG2__DISABLE_PREZL_FIFO_STALL_REZ__SHIFT 0x6 +#define DB_DEBUG2__ENABLE_VIEWPORT_STALL_ON_ALL__SHIFT 0x7 +#define DB_DEBUG2__OPTIMIZE_HIZ_MATCHES_FB_DISABLE__SHIFT 0x8 +#define DB_DEBUG2__CLK_OFF_DELAY__SHIFT 0x9 +#define DB_DEBUG2__DISABLE_TILE_COVERED_FOR_PS_ITER__SHIFT 0xe +#define DB_DEBUG2__ENABLE_SUBTILE_GROUPING__SHIFT 0xf +#define DB_DEBUG2__RESERVED__SHIFT 0x10 +#define DB_DEBUG2__DISABLE_NULL_EOT_FORWARDING__SHIFT 0x11 +#define DB_DEBUG2__DISABLE_DTT_DATA_FORWARDING__SHIFT 0x12 +#define DB_DEBUG2__DISABLE_QUAD_COHERENCY_STALL__SHIFT 0x13 +#define DB_DEBUG2__ENABLE_PREZ_OF_REZ_SUMM__SHIFT 0x1c +#define DB_DEBUG2__DISABLE_PREZL_VIEWPORT_STALL__SHIFT 0x1d +#define DB_DEBUG2__DISABLE_SINGLE_STENCIL_QUAD_SUMM__SHIFT 0x1e +#define DB_DEBUG2__DISABLE_WRITE_STALL_ON_RDWR_CONFLICT__SHIFT 0x1f +#define DB_DEBUG2__ALLOW_COMPZ_BYTE_MASKING_MASK 0x00000001L +#define DB_DEBUG2__DISABLE_TC_ZRANGE_L0_CACHE_MASK 0x00000002L +#define DB_DEBUG2__DISABLE_TC_MASK_L0_CACHE_MASK 0x00000004L +#define DB_DEBUG2__DTR_ROUND_ROBIN_ARB_MASK 0x00000008L +#define DB_DEBUG2__DTR_PREZ_STALLS_FOR_ETF_ROOM_MASK 0x00000010L +#define DB_DEBUG2__DISABLE_PREZL_FIFO_STALL_MASK 0x00000020L +#define DB_DEBUG2__DISABLE_PREZL_FIFO_STALL_REZ_MASK 0x00000040L +#define DB_DEBUG2__ENABLE_VIEWPORT_STALL_ON_ALL_MASK 0x00000080L +#define DB_DEBUG2__OPTIMIZE_HIZ_MATCHES_FB_DISABLE_MASK 0x00000100L +#define DB_DEBUG2__CLK_OFF_DELAY_MASK 0x00003E00L +#define DB_DEBUG2__DISABLE_TILE_COVERED_FOR_PS_ITER_MASK 0x00004000L +#define DB_DEBUG2__ENABLE_SUBTILE_GROUPING_MASK 0x00008000L +#define DB_DEBUG2__RESERVED_MASK 0x00010000L +#define DB_DEBUG2__DISABLE_NULL_EOT_FORWARDING_MASK 0x00020000L +#define DB_DEBUG2__DISABLE_DTT_DATA_FORWARDING_MASK 0x00040000L +#define DB_DEBUG2__DISABLE_QUAD_COHERENCY_STALL_MASK 0x00080000L +#define DB_DEBUG2__ENABLE_PREZ_OF_REZ_SUMM_MASK 0x10000000L +#define DB_DEBUG2__DISABLE_PREZL_VIEWPORT_STALL_MASK 0x20000000L +#define DB_DEBUG2__DISABLE_SINGLE_STENCIL_QUAD_SUMM_MASK 0x40000000L +#define DB_DEBUG2__DISABLE_WRITE_STALL_ON_RDWR_CONFLICT_MASK 0x80000000L +//DB_DEBUG3 +#define DB_DEBUG3__DISABLE_CLEAR_ZRANGE_CORRECTION__SHIFT 0x0 +#define DB_DEBUG3__ROUND_ZRANGE_CORRECTION__SHIFT 0x1 +#define DB_DEBUG3__FORCE_DB_IS_GOOD__SHIFT 0x2 +#define DB_DEBUG3__DISABLE_TL_SSO_NULL_SUPPRESSION__SHIFT 0x3 +#define DB_DEBUG3__DISABLE_HIZ_ON_VPORT_CLAMP__SHIFT 0x4 +#define DB_DEBUG3__EQAA_INTERPOLATE_COMP_Z__SHIFT 0x5 +#define DB_DEBUG3__EQAA_INTERPOLATE_SRC_Z__SHIFT 0x6 +#define DB_DEBUG3__DISABLE_TCP_CAM_BYPASS__SHIFT 0x7 +#define DB_DEBUG3__DISABLE_ZCMP_DIRTY_SUPPRESSION__SHIFT 0x8 +#define DB_DEBUG3__DISABLE_REDUNDANT_PLANE_FLUSHES_OPT__SHIFT 0x9 +#define DB_DEBUG3__DISABLE_RECOMP_TO_1ZPLANE_WITHOUT_FASTOP__SHIFT 0xa +#define DB_DEBUG3__ENABLE_INCOHERENT_EQAA_READS__SHIFT 0xb +#define DB_DEBUG3__DISABLE_OP_Z_DATA_FORWARDING__SHIFT 0xc +#define DB_DEBUG3__DISABLE_OP_DF_BYPASS__SHIFT 0xd +#define DB_DEBUG3__DISABLE_OP_DF_WRITE_COMBINE__SHIFT 0xe +#define DB_DEBUG3__DISABLE_OP_DF_DIRECT_FEEDBACK__SHIFT 0xf +#define DB_DEBUG3__ALLOW_RF2P_RW_COLLISION__SHIFT 0x10 +#define DB_DEBUG3__SLOW_PREZ_TO_A2M_OMASK_RATE__SHIFT 0x11 +#define DB_DEBUG3__DISABLE_OP_S_DATA_FORWARDING__SHIFT 0x12 +#define DB_DEBUG3__DISABLE_TC_UPDATE_WRITE_COMBINE__SHIFT 0x13 +#define DB_DEBUG3__DISABLE_HZ_TC_WRITE_COMBINE__SHIFT 0x14 +#define DB_DEBUG3__ENABLE_RECOMP_ZDIRTY_SUPPRESSION_OPT__SHIFT 0x15 +#define DB_DEBUG3__ENABLE_TC_MA_ROUND_ROBIN_ARB__SHIFT 0x16 +#define DB_DEBUG3__DISABLE_RAM_READ_SUPPRESION_ON_FWD__SHIFT 0x17 +#define DB_DEBUG3__DISABLE_EQAA_A2M_PERF_OPT__SHIFT 0x18 +#define DB_DEBUG3__DISABLE_DI_DT_STALL__SHIFT 0x19 +#define DB_DEBUG3__ENABLE_DB_PROCESS_RESET__SHIFT 0x1a +#define DB_DEBUG3__DISABLE_OVERRASTERIZATION_FIX__SHIFT 0x1b +#define DB_DEBUG3__DONT_INSERT_CONTEXT_SUSPEND__SHIFT 0x1c +#define DB_DEBUG3__DONT_DELETE_CONTEXT_SUSPEND__SHIFT 0x1d +#define DB_DEBUG3__DISABLE_4XAA_2P_DELAYED_WRITE__SHIFT 0x1e +#define DB_DEBUG3__DISABLE_4XAA_2P_INTERLEAVED_PMASK__SHIFT 0x1f +#define DB_DEBUG3__DISABLE_CLEAR_ZRANGE_CORRECTION_MASK 0x00000001L +#define DB_DEBUG3__ROUND_ZRANGE_CORRECTION_MASK 0x00000002L +#define DB_DEBUG3__FORCE_DB_IS_GOOD_MASK 0x00000004L +#define DB_DEBUG3__DISABLE_TL_SSO_NULL_SUPPRESSION_MASK 0x00000008L +#define DB_DEBUG3__DISABLE_HIZ_ON_VPORT_CLAMP_MASK 0x00000010L +#define DB_DEBUG3__EQAA_INTERPOLATE_COMP_Z_MASK 0x00000020L +#define DB_DEBUG3__EQAA_INTERPOLATE_SRC_Z_MASK 0x00000040L +#define DB_DEBUG3__DISABLE_TCP_CAM_BYPASS_MASK 0x00000080L +#define DB_DEBUG3__DISABLE_ZCMP_DIRTY_SUPPRESSION_MASK 0x00000100L +#define DB_DEBUG3__DISABLE_REDUNDANT_PLANE_FLUSHES_OPT_MASK 0x00000200L +#define DB_DEBUG3__DISABLE_RECOMP_TO_1ZPLANE_WITHOUT_FASTOP_MASK 0x00000400L +#define DB_DEBUG3__ENABLE_INCOHERENT_EQAA_READS_MASK 0x00000800L +#define DB_DEBUG3__DISABLE_OP_Z_DATA_FORWARDING_MASK 0x00001000L +#define DB_DEBUG3__DISABLE_OP_DF_BYPASS_MASK 0x00002000L +#define DB_DEBUG3__DISABLE_OP_DF_WRITE_COMBINE_MASK 0x00004000L +#define DB_DEBUG3__DISABLE_OP_DF_DIRECT_FEEDBACK_MASK 0x00008000L +#define DB_DEBUG3__ALLOW_RF2P_RW_COLLISION_MASK 0x00010000L +#define DB_DEBUG3__SLOW_PREZ_TO_A2M_OMASK_RATE_MASK 0x00020000L +#define DB_DEBUG3__DISABLE_OP_S_DATA_FORWARDING_MASK 0x00040000L +#define DB_DEBUG3__DISABLE_TC_UPDATE_WRITE_COMBINE_MASK 0x00080000L +#define DB_DEBUG3__DISABLE_HZ_TC_WRITE_COMBINE_MASK 0x00100000L +#define DB_DEBUG3__ENABLE_RECOMP_ZDIRTY_SUPPRESSION_OPT_MASK 0x00200000L +#define DB_DEBUG3__ENABLE_TC_MA_ROUND_ROBIN_ARB_MASK 0x00400000L +#define DB_DEBUG3__DISABLE_RAM_READ_SUPPRESION_ON_FWD_MASK 0x00800000L +#define DB_DEBUG3__DISABLE_EQAA_A2M_PERF_OPT_MASK 0x01000000L +#define DB_DEBUG3__DISABLE_DI_DT_STALL_MASK 0x02000000L +#define DB_DEBUG3__ENABLE_DB_PROCESS_RESET_MASK 0x04000000L +#define DB_DEBUG3__DISABLE_OVERRASTERIZATION_FIX_MASK 0x08000000L +#define DB_DEBUG3__DONT_INSERT_CONTEXT_SUSPEND_MASK 0x10000000L +#define DB_DEBUG3__DONT_DELETE_CONTEXT_SUSPEND_MASK 0x20000000L +#define DB_DEBUG3__DISABLE_4XAA_2P_DELAYED_WRITE_MASK 0x40000000L +#define DB_DEBUG3__DISABLE_4XAA_2P_INTERLEAVED_PMASK_MASK 0x80000000L +//DB_DEBUG4 +#define DB_DEBUG4__DISABLE_QC_Z_MASK_SUMMATION__SHIFT 0x0 +#define DB_DEBUG4__DISABLE_QC_STENCIL_MASK_SUMMATION__SHIFT 0x1 +#define DB_DEBUG4__DISABLE_RESUMM_TO_SINGLE_STENCIL__SHIFT 0x2 +#define DB_DEBUG4__DISABLE_PREZ_POSTZ_DTILE_CONFLICT_STALL__SHIFT 0x3 +#define DB_DEBUG4__DISABLE_4XAA_2P_ZD_HOLDOFF__SHIFT 0x4 +#define DB_DEBUG4__ENABLE_A2M_DQUAD_OPTIMIZATION__SHIFT 0x5 +#define DB_DEBUG4__ENABLE_DBCB_SLOW_FORMAT_COLLAPSE__SHIFT 0x6 +#define DB_DEBUG4__ALWAYS_ON_RMI_CLK_EN__SHIFT 0x7 +#define DB_DEBUG4__DFSM_CONVERT_PASSTHROUGH_TO_BYPASS__SHIFT 0x8 +#define DB_DEBUG4__DISABLE_UNMAPPED_Z_INDICATOR__SHIFT 0x9 +#define DB_DEBUG4__DISABLE_UNMAPPED_S_INDICATOR__SHIFT 0xa +#define DB_DEBUG4__DISABLE_UNMAPPED_H_INDICATOR__SHIFT 0xb +#define DB_DEBUG4__DISABLE_SEPARATE_DFSM_CLK__SHIFT 0xc +#define DB_DEBUG4__DISABLE_DTT_FAST_HTILENACK_LOOKUP__SHIFT 0xd +#define DB_DEBUG4__DISABLE_RESCHECK_MEMCOHER_OPTIMIZATION__SHIFT 0xe +#define DB_DEBUG4__DISABLE_TS_WRITE_L0__SHIFT 0xf +#define DB_DEBUG4__DISABLE_DYNAMIC_RAM_LIGHT_SLEEP_MODE__SHIFT 0x10 +#define DB_DEBUG4__DISABLE_HIZ_Q1_TS_COLLISION_DETECT__SHIFT 0x11 +#define DB_DEBUG4__DISABLE_HIZ_Q2_TS_COLLISION_DETECT__SHIFT 0x12 +#define DB_DEBUG4__DB_EXTRA_DEBUG4__SHIFT 0x13 +#define DB_DEBUG4__DISABLE_QC_Z_MASK_SUMMATION_MASK 0x00000001L +#define DB_DEBUG4__DISABLE_QC_STENCIL_MASK_SUMMATION_MASK 0x00000002L +#define DB_DEBUG4__DISABLE_RESUMM_TO_SINGLE_STENCIL_MASK 0x00000004L +#define DB_DEBUG4__DISABLE_PREZ_POSTZ_DTILE_CONFLICT_STALL_MASK 0x00000008L +#define DB_DEBUG4__DISABLE_4XAA_2P_ZD_HOLDOFF_MASK 0x00000010L +#define DB_DEBUG4__ENABLE_A2M_DQUAD_OPTIMIZATION_MASK 0x00000020L +#define DB_DEBUG4__ENABLE_DBCB_SLOW_FORMAT_COLLAPSE_MASK 0x00000040L +#define DB_DEBUG4__ALWAYS_ON_RMI_CLK_EN_MASK 0x00000080L +#define DB_DEBUG4__DFSM_CONVERT_PASSTHROUGH_TO_BYPASS_MASK 0x00000100L +#define DB_DEBUG4__DISABLE_UNMAPPED_Z_INDICATOR_MASK 0x00000200L +#define DB_DEBUG4__DISABLE_UNMAPPED_S_INDICATOR_MASK 0x00000400L +#define DB_DEBUG4__DISABLE_UNMAPPED_H_INDICATOR_MASK 0x00000800L +#define DB_DEBUG4__DISABLE_SEPARATE_DFSM_CLK_MASK 0x00001000L +#define DB_DEBUG4__DISABLE_DTT_FAST_HTILENACK_LOOKUP_MASK 0x00002000L +#define DB_DEBUG4__DISABLE_RESCHECK_MEMCOHER_OPTIMIZATION_MASK 0x00004000L +#define DB_DEBUG4__DISABLE_TS_WRITE_L0_MASK 0x00008000L +#define DB_DEBUG4__DISABLE_DYNAMIC_RAM_LIGHT_SLEEP_MODE_MASK 0x00010000L +#define DB_DEBUG4__DISABLE_HIZ_Q1_TS_COLLISION_DETECT_MASK 0x00020000L +#define DB_DEBUG4__DISABLE_HIZ_Q2_TS_COLLISION_DETECT_MASK 0x00040000L +#define DB_DEBUG4__DB_EXTRA_DEBUG4_MASK 0xFFF80000L +//DB_CREDIT_LIMIT +#define DB_CREDIT_LIMIT__DB_SC_TILE_CREDITS__SHIFT 0x0 +#define DB_CREDIT_LIMIT__DB_SC_QUAD_CREDITS__SHIFT 0x5 +#define DB_CREDIT_LIMIT__DB_CB_LQUAD_CREDITS__SHIFT 0xa +#define DB_CREDIT_LIMIT__DB_CB_TILE_CREDITS__SHIFT 0x18 +#define DB_CREDIT_LIMIT__DB_SC_TILE_CREDITS_MASK 0x0000001FL +#define DB_CREDIT_LIMIT__DB_SC_QUAD_CREDITS_MASK 0x000003E0L +#define DB_CREDIT_LIMIT__DB_CB_LQUAD_CREDITS_MASK 0x00001C00L +#define DB_CREDIT_LIMIT__DB_CB_TILE_CREDITS_MASK 0x7F000000L +//DB_WATERMARKS +#define DB_WATERMARKS__DEPTH_FREE__SHIFT 0x0 +#define DB_WATERMARKS__DEPTH_FLUSH__SHIFT 0x5 +#define DB_WATERMARKS__FORCE_SUMMARIZE__SHIFT 0xb +#define DB_WATERMARKS__DEPTH_PENDING_FREE__SHIFT 0xf +#define DB_WATERMARKS__DEPTH_CACHELINE_FREE__SHIFT 0x14 +#define DB_WATERMARKS__AUTO_FLUSH_HTILE__SHIFT 0x1e +#define DB_WATERMARKS__AUTO_FLUSH_QUAD__SHIFT 0x1f +#define DB_WATERMARKS__DEPTH_FREE_MASK 0x0000001FL +#define DB_WATERMARKS__DEPTH_FLUSH_MASK 0x000007E0L +#define DB_WATERMARKS__FORCE_SUMMARIZE_MASK 0x00007800L +#define DB_WATERMARKS__DEPTH_PENDING_FREE_MASK 0x000F8000L +#define DB_WATERMARKS__DEPTH_CACHELINE_FREE_MASK 0x0FF00000L +#define DB_WATERMARKS__AUTO_FLUSH_HTILE_MASK 0x40000000L +#define DB_WATERMARKS__AUTO_FLUSH_QUAD_MASK 0x80000000L +//DB_SUBTILE_CONTROL +#define DB_SUBTILE_CONTROL__MSAA1_X__SHIFT 0x0 +#define DB_SUBTILE_CONTROL__MSAA1_Y__SHIFT 0x2 +#define DB_SUBTILE_CONTROL__MSAA2_X__SHIFT 0x4 +#define DB_SUBTILE_CONTROL__MSAA2_Y__SHIFT 0x6 +#define DB_SUBTILE_CONTROL__MSAA4_X__SHIFT 0x8 +#define DB_SUBTILE_CONTROL__MSAA4_Y__SHIFT 0xa +#define DB_SUBTILE_CONTROL__MSAA8_X__SHIFT 0xc +#define DB_SUBTILE_CONTROL__MSAA8_Y__SHIFT 0xe +#define DB_SUBTILE_CONTROL__MSAA16_X__SHIFT 0x10 +#define DB_SUBTILE_CONTROL__MSAA16_Y__SHIFT 0x12 +#define DB_SUBTILE_CONTROL__MSAA1_X_MASK 0x00000003L +#define DB_SUBTILE_CONTROL__MSAA1_Y_MASK 0x0000000CL +#define DB_SUBTILE_CONTROL__MSAA2_X_MASK 0x00000030L +#define DB_SUBTILE_CONTROL__MSAA2_Y_MASK 0x000000C0L +#define DB_SUBTILE_CONTROL__MSAA4_X_MASK 0x00000300L +#define DB_SUBTILE_CONTROL__MSAA4_Y_MASK 0x00000C00L +#define DB_SUBTILE_CONTROL__MSAA8_X_MASK 0x00003000L +#define DB_SUBTILE_CONTROL__MSAA8_Y_MASK 0x0000C000L +#define DB_SUBTILE_CONTROL__MSAA16_X_MASK 0x00030000L +#define DB_SUBTILE_CONTROL__MSAA16_Y_MASK 0x000C0000L +//DB_FREE_CACHELINES +#define DB_FREE_CACHELINES__FREE_DTILE_DEPTH__SHIFT 0x0 +#define DB_FREE_CACHELINES__FREE_PLANE_DEPTH__SHIFT 0x7 +#define DB_FREE_CACHELINES__FREE_Z_DEPTH__SHIFT 0xe +#define DB_FREE_CACHELINES__FREE_HTILE_DEPTH__SHIFT 0x14 +#define DB_FREE_CACHELINES__QUAD_READ_REQS__SHIFT 0x18 +#define DB_FREE_CACHELINES__FREE_DTILE_DEPTH_MASK 0x0000007FL +#define DB_FREE_CACHELINES__FREE_PLANE_DEPTH_MASK 0x00003F80L +#define DB_FREE_CACHELINES__FREE_Z_DEPTH_MASK 0x000FC000L +#define DB_FREE_CACHELINES__FREE_HTILE_DEPTH_MASK 0x00F00000L +#define DB_FREE_CACHELINES__QUAD_READ_REQS_MASK 0xFF000000L +//DB_FIFO_DEPTH1 +#define DB_FIFO_DEPTH1__DB_RMI_RDREQ_CREDITS__SHIFT 0x0 +#define DB_FIFO_DEPTH1__DB_RMI_WRREQ_CREDITS__SHIFT 0x5 +#define DB_FIFO_DEPTH1__MCC_DEPTH__SHIFT 0xa +#define DB_FIFO_DEPTH1__QC_DEPTH__SHIFT 0x10 +#define DB_FIFO_DEPTH1__LTILE_PROBE_FIFO_DEPTH__SHIFT 0x15 +#define DB_FIFO_DEPTH1__DB_RMI_RDREQ_CREDITS_MASK 0x0000001FL +#define DB_FIFO_DEPTH1__DB_RMI_WRREQ_CREDITS_MASK 0x000003E0L +#define DB_FIFO_DEPTH1__MCC_DEPTH_MASK 0x0000FC00L +#define DB_FIFO_DEPTH1__QC_DEPTH_MASK 0x001F0000L +#define DB_FIFO_DEPTH1__LTILE_PROBE_FIFO_DEPTH_MASK 0x1FE00000L +//DB_FIFO_DEPTH2 +#define DB_FIFO_DEPTH2__EQUAD_FIFO_DEPTH__SHIFT 0x0 +#define DB_FIFO_DEPTH2__ETILE_OP_FIFO_DEPTH__SHIFT 0x8 +#define DB_FIFO_DEPTH2__LQUAD_FIFO_DEPTH__SHIFT 0xf +#define DB_FIFO_DEPTH2__LTILE_OP_FIFO_DEPTH__SHIFT 0x19 +#define DB_FIFO_DEPTH2__EQUAD_FIFO_DEPTH_MASK 0x000000FFL +#define DB_FIFO_DEPTH2__ETILE_OP_FIFO_DEPTH_MASK 0x00007F00L +#define DB_FIFO_DEPTH2__LQUAD_FIFO_DEPTH_MASK 0x01FF8000L +#define DB_FIFO_DEPTH2__LTILE_OP_FIFO_DEPTH_MASK 0xFE000000L +//DB_EXCEPTION_CONTROL +#define DB_EXCEPTION_CONTROL__EARLY_Z_PANIC_DISABLE__SHIFT 0x0 +#define DB_EXCEPTION_CONTROL__LATE_Z_PANIC_DISABLE__SHIFT 0x1 +#define DB_EXCEPTION_CONTROL__RE_Z_PANIC_DISABLE__SHIFT 0x2 +#define DB_EXCEPTION_CONTROL__EARLY_Z_PANIC_DISABLE_MASK 0x00000001L +#define DB_EXCEPTION_CONTROL__LATE_Z_PANIC_DISABLE_MASK 0x00000002L +#define DB_EXCEPTION_CONTROL__RE_Z_PANIC_DISABLE_MASK 0x00000004L +//DB_RING_CONTROL +#define DB_RING_CONTROL__COUNTER_CONTROL__SHIFT 0x0 +#define DB_RING_CONTROL__COUNTER_CONTROL_MASK 0x00000003L +//DB_MEM_ARB_WATERMARKS +#define DB_MEM_ARB_WATERMARKS__CLIENT0_WATERMARK__SHIFT 0x0 +#define DB_MEM_ARB_WATERMARKS__CLIENT1_WATERMARK__SHIFT 0x8 +#define DB_MEM_ARB_WATERMARKS__CLIENT2_WATERMARK__SHIFT 0x10 +#define DB_MEM_ARB_WATERMARKS__CLIENT3_WATERMARK__SHIFT 0x18 +#define DB_MEM_ARB_WATERMARKS__CLIENT0_WATERMARK_MASK 0x00000007L +#define DB_MEM_ARB_WATERMARKS__CLIENT1_WATERMARK_MASK 0x00000700L +#define DB_MEM_ARB_WATERMARKS__CLIENT2_WATERMARK_MASK 0x00070000L +#define DB_MEM_ARB_WATERMARKS__CLIENT3_WATERMARK_MASK 0x07000000L +//DB_RMI_CACHE_POLICY +#define DB_RMI_CACHE_POLICY__Z_RD__SHIFT 0x0 +#define DB_RMI_CACHE_POLICY__S_RD__SHIFT 0x1 +#define DB_RMI_CACHE_POLICY__HTILE_RD__SHIFT 0x2 +#define DB_RMI_CACHE_POLICY__Z_WR__SHIFT 0x8 +#define DB_RMI_CACHE_POLICY__S_WR__SHIFT 0x9 +#define DB_RMI_CACHE_POLICY__HTILE_WR__SHIFT 0xa +#define DB_RMI_CACHE_POLICY__ZPCPSD_WR__SHIFT 0xb +#define DB_RMI_CACHE_POLICY__CC_RD__SHIFT 0x10 +#define DB_RMI_CACHE_POLICY__FMASK_RD__SHIFT 0x11 +#define DB_RMI_CACHE_POLICY__CMASK_RD__SHIFT 0x12 +#define DB_RMI_CACHE_POLICY__DCC_RD__SHIFT 0x13 +#define DB_RMI_CACHE_POLICY__CC_WR__SHIFT 0x18 +#define DB_RMI_CACHE_POLICY__FMASK_WR__SHIFT 0x19 +#define DB_RMI_CACHE_POLICY__CMASK_WR__SHIFT 0x1a +#define DB_RMI_CACHE_POLICY__DCC_WR__SHIFT 0x1b +#define DB_RMI_CACHE_POLICY__Z_RD_MASK 0x00000001L +#define DB_RMI_CACHE_POLICY__S_RD_MASK 0x00000002L +#define DB_RMI_CACHE_POLICY__HTILE_RD_MASK 0x00000004L +#define DB_RMI_CACHE_POLICY__Z_WR_MASK 0x00000100L +#define DB_RMI_CACHE_POLICY__S_WR_MASK 0x00000200L +#define DB_RMI_CACHE_POLICY__HTILE_WR_MASK 0x00000400L +#define DB_RMI_CACHE_POLICY__ZPCPSD_WR_MASK 0x00000800L +#define DB_RMI_CACHE_POLICY__CC_RD_MASK 0x00010000L +#define DB_RMI_CACHE_POLICY__FMASK_RD_MASK 0x00020000L +#define DB_RMI_CACHE_POLICY__CMASK_RD_MASK 0x00040000L +#define DB_RMI_CACHE_POLICY__DCC_RD_MASK 0x00080000L +#define DB_RMI_CACHE_POLICY__CC_WR_MASK 0x01000000L +#define DB_RMI_CACHE_POLICY__FMASK_WR_MASK 0x02000000L +#define DB_RMI_CACHE_POLICY__CMASK_WR_MASK 0x04000000L +#define DB_RMI_CACHE_POLICY__DCC_WR_MASK 0x08000000L +//DB_DFSM_CONFIG +#define DB_DFSM_CONFIG__BYPASS_DFSM__SHIFT 0x0 +#define DB_DFSM_CONFIG__DISABLE_PUNCHOUT__SHIFT 0x1 +#define DB_DFSM_CONFIG__DISABLE_POPS__SHIFT 0x2 +#define DB_DFSM_CONFIG__FORCE_FLUSH__SHIFT 0x3 +#define DB_DFSM_CONFIG__MIDDLE_PIPE_MAX_DEPTH__SHIFT 0x8 +#define DB_DFSM_CONFIG__BYPASS_DFSM_MASK 0x00000001L +#define DB_DFSM_CONFIG__DISABLE_PUNCHOUT_MASK 0x00000002L +#define DB_DFSM_CONFIG__DISABLE_POPS_MASK 0x00000004L +#define DB_DFSM_CONFIG__FORCE_FLUSH_MASK 0x00000008L +#define DB_DFSM_CONFIG__MIDDLE_PIPE_MAX_DEPTH_MASK 0x00007F00L +//DB_DFSM_WATERMARK +#define DB_DFSM_WATERMARK__DFSM_HIGH_WATERMARK__SHIFT 0x0 +#define DB_DFSM_WATERMARK__POPS_HIGH_WATERMARK__SHIFT 0x10 +#define DB_DFSM_WATERMARK__DFSM_HIGH_WATERMARK_MASK 0x0000FFFFL +#define DB_DFSM_WATERMARK__POPS_HIGH_WATERMARK_MASK 0xFFFF0000L +//DB_DFSM_TILES_IN_FLIGHT +#define DB_DFSM_TILES_IN_FLIGHT__HIGH_WATERMARK__SHIFT 0x0 +#define DB_DFSM_TILES_IN_FLIGHT__HARD_LIMIT__SHIFT 0x10 +#define DB_DFSM_TILES_IN_FLIGHT__HIGH_WATERMARK_MASK 0x0000FFFFL +#define DB_DFSM_TILES_IN_FLIGHT__HARD_LIMIT_MASK 0xFFFF0000L +//DB_DFSM_PRIMS_IN_FLIGHT +#define DB_DFSM_PRIMS_IN_FLIGHT__HIGH_WATERMARK__SHIFT 0x0 +#define DB_DFSM_PRIMS_IN_FLIGHT__HARD_LIMIT__SHIFT 0x10 +#define DB_DFSM_PRIMS_IN_FLIGHT__HIGH_WATERMARK_MASK 0x0000FFFFL +#define DB_DFSM_PRIMS_IN_FLIGHT__HARD_LIMIT_MASK 0xFFFF0000L +//DB_DFSM_WATCHDOG +#define DB_DFSM_WATCHDOG__TIMER_TARGET__SHIFT 0x0 +#define DB_DFSM_WATCHDOG__TIMER_TARGET_MASK 0xFFFFFFFFL +//DB_DFSM_FLUSH_ENABLE +#define DB_DFSM_FLUSH_ENABLE__PRIMARY_EVENTS__SHIFT 0x0 +#define DB_DFSM_FLUSH_ENABLE__AUX_FORCE_PASSTHRU__SHIFT 0x18 +#define DB_DFSM_FLUSH_ENABLE__AUX_EVENTS__SHIFT 0x1c +#define DB_DFSM_FLUSH_ENABLE__PRIMARY_EVENTS_MASK 0x000003FFL +#define DB_DFSM_FLUSH_ENABLE__AUX_FORCE_PASSTHRU_MASK 0x0F000000L +#define DB_DFSM_FLUSH_ENABLE__AUX_EVENTS_MASK 0xF0000000L +//DB_DFSM_FLUSH_AUX_EVENT +#define DB_DFSM_FLUSH_AUX_EVENT__EVENT_A__SHIFT 0x0 +#define DB_DFSM_FLUSH_AUX_EVENT__EVENT_B__SHIFT 0x8 +#define DB_DFSM_FLUSH_AUX_EVENT__EVENT_C__SHIFT 0x10 +#define DB_DFSM_FLUSH_AUX_EVENT__EVENT_D__SHIFT 0x18 +#define DB_DFSM_FLUSH_AUX_EVENT__EVENT_A_MASK 0x000000FFL +#define DB_DFSM_FLUSH_AUX_EVENT__EVENT_B_MASK 0x0000FF00L +#define DB_DFSM_FLUSH_AUX_EVENT__EVENT_C_MASK 0x00FF0000L +#define DB_DFSM_FLUSH_AUX_EVENT__EVENT_D_MASK 0xFF000000L +//CC_RB_REDUNDANCY +#define CC_RB_REDUNDANCY__FAILED_RB0__SHIFT 0x8 +#define CC_RB_REDUNDANCY__EN_REDUNDANCY0__SHIFT 0xc +#define CC_RB_REDUNDANCY__FAILED_RB1__SHIFT 0x10 +#define CC_RB_REDUNDANCY__EN_REDUNDANCY1__SHIFT 0x14 +#define CC_RB_REDUNDANCY__FAILED_RB0_MASK 0x00000F00L +#define CC_RB_REDUNDANCY__EN_REDUNDANCY0_MASK 0x00001000L +#define CC_RB_REDUNDANCY__FAILED_RB1_MASK 0x000F0000L +#define CC_RB_REDUNDANCY__EN_REDUNDANCY1_MASK 0x00100000L +//CC_RB_BACKEND_DISABLE +#define CC_RB_BACKEND_DISABLE__BACKEND_DISABLE__SHIFT 0x10 +#define CC_RB_BACKEND_DISABLE__BACKEND_DISABLE_MASK 0x00FF0000L +//GB_ADDR_CONFIG +#define GB_ADDR_CONFIG__NUM_PIPES__SHIFT 0x0 +#define GB_ADDR_CONFIG__PIPE_INTERLEAVE_SIZE__SHIFT 0x3 +#define GB_ADDR_CONFIG__MAX_COMPRESSED_FRAGS__SHIFT 0x6 +#define GB_ADDR_CONFIG__BANK_INTERLEAVE_SIZE__SHIFT 0x8 +#define GB_ADDR_CONFIG__NUM_BANKS__SHIFT 0xc +#define GB_ADDR_CONFIG__SHADER_ENGINE_TILE_SIZE__SHIFT 0x10 +#define GB_ADDR_CONFIG__NUM_SHADER_ENGINES__SHIFT 0x13 +#define GB_ADDR_CONFIG__NUM_GPUS__SHIFT 0x15 +#define GB_ADDR_CONFIG__MULTI_GPU_TILE_SIZE__SHIFT 0x18 +#define GB_ADDR_CONFIG__NUM_RB_PER_SE__SHIFT 0x1a +#define GB_ADDR_CONFIG__ROW_SIZE__SHIFT 0x1c +#define GB_ADDR_CONFIG__NUM_LOWER_PIPES__SHIFT 0x1e +#define GB_ADDR_CONFIG__SE_ENABLE__SHIFT 0x1f +#define GB_ADDR_CONFIG__NUM_PIPES_MASK 0x00000007L +#define GB_ADDR_CONFIG__PIPE_INTERLEAVE_SIZE_MASK 0x00000038L +#define GB_ADDR_CONFIG__MAX_COMPRESSED_FRAGS_MASK 0x000000C0L +#define GB_ADDR_CONFIG__BANK_INTERLEAVE_SIZE_MASK 0x00000700L +#define GB_ADDR_CONFIG__NUM_BANKS_MASK 0x00007000L +#define GB_ADDR_CONFIG__SHADER_ENGINE_TILE_SIZE_MASK 0x00070000L +#define GB_ADDR_CONFIG__NUM_SHADER_ENGINES_MASK 0x00180000L +#define GB_ADDR_CONFIG__NUM_GPUS_MASK 0x00E00000L +#define GB_ADDR_CONFIG__MULTI_GPU_TILE_SIZE_MASK 0x03000000L +#define GB_ADDR_CONFIG__NUM_RB_PER_SE_MASK 0x0C000000L +#define GB_ADDR_CONFIG__ROW_SIZE_MASK 0x30000000L +#define GB_ADDR_CONFIG__NUM_LOWER_PIPES_MASK 0x40000000L +#define GB_ADDR_CONFIG__SE_ENABLE_MASK 0x80000000L +//GB_BACKEND_MAP +#define GB_BACKEND_MAP__BACKEND_MAP__SHIFT 0x0 +#define GB_BACKEND_MAP__BACKEND_MAP_MASK 0xFFFFFFFFL +//GB_GPU_ID +#define GB_GPU_ID__GPU_ID__SHIFT 0x0 +#define GB_GPU_ID__GPU_ID_MASK 0x0000000FL +//CC_RB_DAISY_CHAIN +#define CC_RB_DAISY_CHAIN__RB_0__SHIFT 0x0 +#define CC_RB_DAISY_CHAIN__RB_1__SHIFT 0x4 +#define CC_RB_DAISY_CHAIN__RB_2__SHIFT 0x8 +#define CC_RB_DAISY_CHAIN__RB_3__SHIFT 0xc +#define CC_RB_DAISY_CHAIN__RB_4__SHIFT 0x10 +#define CC_RB_DAISY_CHAIN__RB_5__SHIFT 0x14 +#define CC_RB_DAISY_CHAIN__RB_6__SHIFT 0x18 +#define CC_RB_DAISY_CHAIN__RB_7__SHIFT 0x1c +#define CC_RB_DAISY_CHAIN__RB_0_MASK 0x0000000FL +#define CC_RB_DAISY_CHAIN__RB_1_MASK 0x000000F0L +#define CC_RB_DAISY_CHAIN__RB_2_MASK 0x00000F00L +#define CC_RB_DAISY_CHAIN__RB_3_MASK 0x0000F000L +#define CC_RB_DAISY_CHAIN__RB_4_MASK 0x000F0000L +#define CC_RB_DAISY_CHAIN__RB_5_MASK 0x00F00000L +#define CC_RB_DAISY_CHAIN__RB_6_MASK 0x0F000000L +#define CC_RB_DAISY_CHAIN__RB_7_MASK 0xF0000000L +//GB_ADDR_CONFIG_READ +#define GB_ADDR_CONFIG_READ__NUM_PIPES__SHIFT 0x0 +#define GB_ADDR_CONFIG_READ__PIPE_INTERLEAVE_SIZE__SHIFT 0x3 +#define GB_ADDR_CONFIG_READ__MAX_COMPRESSED_FRAGS__SHIFT 0x6 +#define GB_ADDR_CONFIG_READ__BANK_INTERLEAVE_SIZE__SHIFT 0x8 +#define GB_ADDR_CONFIG_READ__NUM_BANKS__SHIFT 0xc +#define GB_ADDR_CONFIG_READ__SHADER_ENGINE_TILE_SIZE__SHIFT 0x10 +#define GB_ADDR_CONFIG_READ__NUM_SHADER_ENGINES__SHIFT 0x13 +#define GB_ADDR_CONFIG_READ__NUM_GPUS__SHIFT 0x15 +#define GB_ADDR_CONFIG_READ__MULTI_GPU_TILE_SIZE__SHIFT 0x18 +#define GB_ADDR_CONFIG_READ__NUM_RB_PER_SE__SHIFT 0x1a +#define GB_ADDR_CONFIG_READ__ROW_SIZE__SHIFT 0x1c +#define GB_ADDR_CONFIG_READ__NUM_LOWER_PIPES__SHIFT 0x1e +#define GB_ADDR_CONFIG_READ__SE_ENABLE__SHIFT 0x1f +#define GB_ADDR_CONFIG_READ__NUM_PIPES_MASK 0x00000007L +#define GB_ADDR_CONFIG_READ__PIPE_INTERLEAVE_SIZE_MASK 0x00000038L +#define GB_ADDR_CONFIG_READ__MAX_COMPRESSED_FRAGS_MASK 0x000000C0L +#define GB_ADDR_CONFIG_READ__BANK_INTERLEAVE_SIZE_MASK 0x00000700L +#define GB_ADDR_CONFIG_READ__NUM_BANKS_MASK 0x00007000L +#define GB_ADDR_CONFIG_READ__SHADER_ENGINE_TILE_SIZE_MASK 0x00070000L +#define GB_ADDR_CONFIG_READ__NUM_SHADER_ENGINES_MASK 0x00180000L +#define GB_ADDR_CONFIG_READ__NUM_GPUS_MASK 0x00E00000L +#define GB_ADDR_CONFIG_READ__MULTI_GPU_TILE_SIZE_MASK 0x03000000L +#define GB_ADDR_CONFIG_READ__NUM_RB_PER_SE_MASK 0x0C000000L +#define GB_ADDR_CONFIG_READ__ROW_SIZE_MASK 0x30000000L +#define GB_ADDR_CONFIG_READ__NUM_LOWER_PIPES_MASK 0x40000000L +#define GB_ADDR_CONFIG_READ__SE_ENABLE_MASK 0x80000000L +//GB_TILE_MODE0 +#define GB_TILE_MODE0__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE0__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE0__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE0__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE0__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE0__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE0__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE0__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE0__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE0__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE1 +#define GB_TILE_MODE1__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE1__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE1__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE1__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE1__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE1__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE1__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE1__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE1__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE1__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE2 +#define GB_TILE_MODE2__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE2__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE2__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE2__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE2__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE2__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE2__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE2__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE2__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE2__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE3 +#define GB_TILE_MODE3__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE3__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE3__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE3__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE3__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE3__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE3__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE3__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE3__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE3__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE4 +#define GB_TILE_MODE4__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE4__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE4__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE4__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE4__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE4__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE4__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE4__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE4__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE4__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE5 +#define GB_TILE_MODE5__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE5__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE5__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE5__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE5__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE5__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE5__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE5__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE5__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE5__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE6 +#define GB_TILE_MODE6__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE6__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE6__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE6__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE6__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE6__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE6__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE6__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE6__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE6__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE7 +#define GB_TILE_MODE7__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE7__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE7__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE7__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE7__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE7__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE7__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE7__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE7__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE7__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE8 +#define GB_TILE_MODE8__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE8__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE8__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE8__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE8__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE8__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE8__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE8__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE8__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE8__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE9 +#define GB_TILE_MODE9__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE9__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE9__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE9__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE9__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE9__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE9__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE9__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE9__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE9__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE10 +#define GB_TILE_MODE10__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE10__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE10__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE10__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE10__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE10__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE10__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE10__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE10__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE10__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE11 +#define GB_TILE_MODE11__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE11__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE11__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE11__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE11__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE11__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE11__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE11__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE11__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE11__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE12 +#define GB_TILE_MODE12__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE12__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE12__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE12__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE12__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE12__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE12__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE12__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE12__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE12__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE13 +#define GB_TILE_MODE13__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE13__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE13__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE13__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE13__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE13__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE13__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE13__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE13__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE13__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE14 +#define GB_TILE_MODE14__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE14__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE14__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE14__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE14__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE14__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE14__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE14__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE14__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE14__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE15 +#define GB_TILE_MODE15__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE15__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE15__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE15__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE15__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE15__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE15__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE15__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE15__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE15__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE16 +#define GB_TILE_MODE16__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE16__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE16__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE16__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE16__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE16__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE16__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE16__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE16__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE16__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE17 +#define GB_TILE_MODE17__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE17__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE17__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE17__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE17__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE17__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE17__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE17__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE17__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE17__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE18 +#define GB_TILE_MODE18__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE18__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE18__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE18__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE18__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE18__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE18__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE18__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE18__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE18__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE19 +#define GB_TILE_MODE19__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE19__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE19__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE19__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE19__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE19__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE19__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE19__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE19__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE19__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE20 +#define GB_TILE_MODE20__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE20__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE20__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE20__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE20__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE20__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE20__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE20__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE20__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE20__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE21 +#define GB_TILE_MODE21__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE21__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE21__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE21__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE21__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE21__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE21__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE21__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE21__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE21__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE22 +#define GB_TILE_MODE22__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE22__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE22__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE22__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE22__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE22__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE22__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE22__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE22__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE22__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE23 +#define GB_TILE_MODE23__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE23__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE23__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE23__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE23__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE23__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE23__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE23__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE23__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE23__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE24 +#define GB_TILE_MODE24__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE24__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE24__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE24__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE24__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE24__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE24__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE24__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE24__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE24__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE25 +#define GB_TILE_MODE25__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE25__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE25__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE25__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE25__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE25__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE25__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE25__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE25__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE25__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE26 +#define GB_TILE_MODE26__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE26__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE26__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE26__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE26__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE26__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE26__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE26__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE26__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE26__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE27 +#define GB_TILE_MODE27__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE27__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE27__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE27__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE27__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE27__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE27__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE27__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE27__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE27__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE28 +#define GB_TILE_MODE28__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE28__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE28__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE28__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE28__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE28__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE28__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE28__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE28__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE28__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE29 +#define GB_TILE_MODE29__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE29__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE29__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE29__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE29__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE29__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE29__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE29__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE29__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE29__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE30 +#define GB_TILE_MODE30__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE30__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE30__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE30__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE30__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE30__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE30__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE30__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE30__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE30__SAMPLE_SPLIT_MASK 0x06000000L +//GB_TILE_MODE31 +#define GB_TILE_MODE31__ARRAY_MODE__SHIFT 0x2 +#define GB_TILE_MODE31__PIPE_CONFIG__SHIFT 0x6 +#define GB_TILE_MODE31__TILE_SPLIT__SHIFT 0xb +#define GB_TILE_MODE31__MICRO_TILE_MODE_NEW__SHIFT 0x16 +#define GB_TILE_MODE31__SAMPLE_SPLIT__SHIFT 0x19 +#define GB_TILE_MODE31__ARRAY_MODE_MASK 0x0000003CL +#define GB_TILE_MODE31__PIPE_CONFIG_MASK 0x000007C0L +#define GB_TILE_MODE31__TILE_SPLIT_MASK 0x00003800L +#define GB_TILE_MODE31__MICRO_TILE_MODE_NEW_MASK 0x01C00000L +#define GB_TILE_MODE31__SAMPLE_SPLIT_MASK 0x06000000L +//GB_MACROTILE_MODE0 +#define GB_MACROTILE_MODE0__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE0__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE0__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE0__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE0__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE0__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE0__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE0__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE1 +#define GB_MACROTILE_MODE1__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE1__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE1__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE1__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE1__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE1__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE1__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE1__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE2 +#define GB_MACROTILE_MODE2__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE2__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE2__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE2__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE2__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE2__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE2__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE2__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE3 +#define GB_MACROTILE_MODE3__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE3__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE3__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE3__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE3__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE3__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE3__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE3__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE4 +#define GB_MACROTILE_MODE4__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE4__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE4__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE4__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE4__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE4__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE4__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE4__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE5 +#define GB_MACROTILE_MODE5__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE5__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE5__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE5__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE5__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE5__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE5__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE5__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE6 +#define GB_MACROTILE_MODE6__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE6__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE6__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE6__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE6__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE6__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE6__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE6__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE7 +#define GB_MACROTILE_MODE7__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE7__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE7__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE7__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE7__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE7__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE7__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE7__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE8 +#define GB_MACROTILE_MODE8__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE8__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE8__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE8__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE8__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE8__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE8__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE8__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE9 +#define GB_MACROTILE_MODE9__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE9__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE9__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE9__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE9__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE9__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE9__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE9__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE10 +#define GB_MACROTILE_MODE10__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE10__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE10__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE10__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE10__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE10__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE10__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE10__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE11 +#define GB_MACROTILE_MODE11__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE11__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE11__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE11__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE11__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE11__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE11__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE11__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE12 +#define GB_MACROTILE_MODE12__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE12__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE12__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE12__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE12__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE12__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE12__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE12__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE13 +#define GB_MACROTILE_MODE13__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE13__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE13__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE13__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE13__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE13__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE13__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE13__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE14 +#define GB_MACROTILE_MODE14__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE14__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE14__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE14__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE14__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE14__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE14__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE14__NUM_BANKS_MASK 0x000000C0L +//GB_MACROTILE_MODE15 +#define GB_MACROTILE_MODE15__BANK_WIDTH__SHIFT 0x0 +#define GB_MACROTILE_MODE15__BANK_HEIGHT__SHIFT 0x2 +#define GB_MACROTILE_MODE15__MACRO_TILE_ASPECT__SHIFT 0x4 +#define GB_MACROTILE_MODE15__NUM_BANKS__SHIFT 0x6 +#define GB_MACROTILE_MODE15__BANK_WIDTH_MASK 0x00000003L +#define GB_MACROTILE_MODE15__BANK_HEIGHT_MASK 0x0000000CL +#define GB_MACROTILE_MODE15__MACRO_TILE_ASPECT_MASK 0x00000030L +#define GB_MACROTILE_MODE15__NUM_BANKS_MASK 0x000000C0L +//CB_HW_CONTROL +#define CB_HW_CONTROL__CM_CACHE_EVICT_POINT__SHIFT 0x0 +#define CB_HW_CONTROL__FC_CACHE_EVICT_POINT__SHIFT 0x6 +#define CB_HW_CONTROL__CC_CACHE_EVICT_POINT__SHIFT 0xc +#define CB_HW_CONTROL__ALLOW_MRT_WITH_DUAL_SOURCE__SHIFT 0x10 +#define CB_HW_CONTROL__DISABLE_INTNORM_LE11BPC_CLAMPING__SHIFT 0x12 +#define CB_HW_CONTROL__FORCE_NEEDS_DST__SHIFT 0x13 +#define CB_HW_CONTROL__FORCE_ALWAYS_TOGGLE__SHIFT 0x14 +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_RESULT_EQ_DEST__SHIFT 0x15 +#define CB_HW_CONTROL__DISABLE_FULL_WRITE_MASK__SHIFT 0x16 +#define CB_HW_CONTROL__DISABLE_RESOLVE_OPT_FOR_SINGLE_FRAG__SHIFT 0x17 +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_DONT_RD_DST__SHIFT 0x18 +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_BYPASS__SHIFT 0x19 +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_DISCARD_PIXEL__SHIFT 0x1a +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_WHEN_DISABLED_SRCALPHA_IS_USED__SHIFT 0x1b +#define CB_HW_CONTROL__PRIORITIZE_FC_WR_OVER_FC_RD_ON_CMASK_CONFLICT__SHIFT 0x1c +#define CB_HW_CONTROL__PRIORITIZE_FC_EVICT_OVER_FOP_RD_ON_BANK_CONFLICT__SHIFT 0x1d +#define CB_HW_CONTROL__DISABLE_CC_IB_SERIALIZER_STATE_OPT__SHIFT 0x1e +#define CB_HW_CONTROL__DISABLE_PIXEL_IN_QUAD_FIX_FOR_LINEAR_SURFACE__SHIFT 0x1f +#define CB_HW_CONTROL__CM_CACHE_EVICT_POINT_MASK 0x0000000FL +#define CB_HW_CONTROL__FC_CACHE_EVICT_POINT_MASK 0x000003C0L +#define CB_HW_CONTROL__CC_CACHE_EVICT_POINT_MASK 0x0000F000L +#define CB_HW_CONTROL__ALLOW_MRT_WITH_DUAL_SOURCE_MASK 0x00010000L +#define CB_HW_CONTROL__DISABLE_INTNORM_LE11BPC_CLAMPING_MASK 0x00040000L +#define CB_HW_CONTROL__FORCE_NEEDS_DST_MASK 0x00080000L +#define CB_HW_CONTROL__FORCE_ALWAYS_TOGGLE_MASK 0x00100000L +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_RESULT_EQ_DEST_MASK 0x00200000L +#define CB_HW_CONTROL__DISABLE_FULL_WRITE_MASK_MASK 0x00400000L +#define CB_HW_CONTROL__DISABLE_RESOLVE_OPT_FOR_SINGLE_FRAG_MASK 0x00800000L +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_DONT_RD_DST_MASK 0x01000000L +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_BYPASS_MASK 0x02000000L +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_DISCARD_PIXEL_MASK 0x04000000L +#define CB_HW_CONTROL__DISABLE_BLEND_OPT_WHEN_DISABLED_SRCALPHA_IS_USED_MASK 0x08000000L +#define CB_HW_CONTROL__PRIORITIZE_FC_WR_OVER_FC_RD_ON_CMASK_CONFLICT_MASK 0x10000000L +#define CB_HW_CONTROL__PRIORITIZE_FC_EVICT_OVER_FOP_RD_ON_BANK_CONFLICT_MASK 0x20000000L +#define CB_HW_CONTROL__DISABLE_CC_IB_SERIALIZER_STATE_OPT_MASK 0x40000000L +#define CB_HW_CONTROL__DISABLE_PIXEL_IN_QUAD_FIX_FOR_LINEAR_SURFACE_MASK 0x80000000L +//CB_HW_CONTROL_1 +#define CB_HW_CONTROL_1__CM_CACHE_NUM_TAGS__SHIFT 0x0 +#define CB_HW_CONTROL_1__FC_CACHE_NUM_TAGS__SHIFT 0x5 +#define CB_HW_CONTROL_1__CC_CACHE_NUM_TAGS__SHIFT 0xb +#define CB_HW_CONTROL_1__CM_TILE_FIFO_DEPTH__SHIFT 0x11 +#define CB_HW_CONTROL_1__RMI_CREDITS__SHIFT 0x1a +#define CB_HW_CONTROL_1__CM_CACHE_NUM_TAGS_MASK 0x0000001FL +#define CB_HW_CONTROL_1__FC_CACHE_NUM_TAGS_MASK 0x000007E0L +#define CB_HW_CONTROL_1__CC_CACHE_NUM_TAGS_MASK 0x0001F800L +#define CB_HW_CONTROL_1__CM_TILE_FIFO_DEPTH_MASK 0x03FE0000L +#define CB_HW_CONTROL_1__RMI_CREDITS_MASK 0xFC000000L +//CB_HW_CONTROL_2 +#define CB_HW_CONTROL_2__CC_EVEN_ODD_FIFO_DEPTH__SHIFT 0x0 +#define CB_HW_CONTROL_2__FC_RDLAT_TILE_FIFO_DEPTH__SHIFT 0x8 +#define CB_HW_CONTROL_2__FC_RDLAT_QUAD_FIFO_DEPTH__SHIFT 0xf +#define CB_HW_CONTROL_2__DRR_ASSUMED_FIFO_DEPTH_DIV8__SHIFT 0x18 +#define CB_HW_CONTROL_2__CHICKEN_BITS__SHIFT 0x1c +#define CB_HW_CONTROL_2__CC_EVEN_ODD_FIFO_DEPTH_MASK 0x000000FFL +#define CB_HW_CONTROL_2__FC_RDLAT_TILE_FIFO_DEPTH_MASK 0x00007F00L +#define CB_HW_CONTROL_2__FC_RDLAT_QUAD_FIFO_DEPTH_MASK 0x007F8000L +#define CB_HW_CONTROL_2__DRR_ASSUMED_FIFO_DEPTH_DIV8_MASK 0x0F000000L +#define CB_HW_CONTROL_2__CHICKEN_BITS_MASK 0xF0000000L +//CB_HW_CONTROL_3 +#define CB_HW_CONTROL_3__DISABLE_SLOW_MODE_EMPTY_HALF_QUAD_KILL__SHIFT 0x0 +#define CB_HW_CONTROL_3__RAM_ADDRESS_CONFLICTS_DISALLOWED__SHIFT 0x1 +#define CB_HW_CONTROL_3__DISABLE_FAST_CLEAR_FETCH_OPT__SHIFT 0x2 +#define CB_HW_CONTROL_3__DISABLE_QUAD_MARKER_DROP_STOP__SHIFT 0x3 +#define CB_HW_CONTROL_3__DISABLE_OVERWRITE_COMBINER_CAM_CLR__SHIFT 0x4 +#define CB_HW_CONTROL_3__DISABLE_CC_CACHE_OVWR_STATUS_ACCUM__SHIFT 0x5 +#define CB_HW_CONTROL_3__DISABLE_CC_CACHE_OVWR_KEY_MOD__SHIFT 0x6 +#define CB_HW_CONTROL_3__DISABLE_CC_CACHE_PANIC_GATING__SHIFT 0x7 +#define CB_HW_CONTROL_3__DISABLE_OVERWRITE_COMBINER_TARGET_MASK_VALIDATION__SHIFT 0x8 +#define CB_HW_CONTROL_3__SPLIT_ALL_FAST_MODE_TRANSFERS__SHIFT 0x9 +#define CB_HW_CONTROL_3__DISABLE_SHADER_BLEND_OPTS__SHIFT 0xa +#define CB_HW_CONTROL_3__DISABLE_CMASK_LAST_QUAD_INSERTION__SHIFT 0xb +#define CB_HW_CONTROL_3__DISABLE_ROP3_FIXES_OF_BUG_511967__SHIFT 0xc +#define CB_HW_CONTROL_3__DISABLE_ROP3_FIXES_OF_BUG_520657__SHIFT 0xd +#define CB_HW_CONTROL_3__DISABLE_OC_FIXES_OF_BUG_522542__SHIFT 0xe +#define CB_HW_CONTROL_3__FORCE_RMI_LAST_HIGH__SHIFT 0xf +#define CB_HW_CONTROL_3__FORCE_RMI_CLKEN_HIGH__SHIFT 0x10 +#define CB_HW_CONTROL_3__DISABLE_EARLY_WRACKS_CC__SHIFT 0x11 +#define CB_HW_CONTROL_3__DISABLE_EARLY_WRACKS_FC__SHIFT 0x12 +#define CB_HW_CONTROL_3__DISABLE_EARLY_WRACKS_DC__SHIFT 0x13 +#define CB_HW_CONTROL_3__DISABLE_EARLY_WRACKS_CM__SHIFT 0x14 +#define CB_HW_CONTROL_3__DISABLE_NACK_PROCESSING_CC__SHIFT 0x15 +#define CB_HW_CONTROL_3__DISABLE_NACK_PROCESSING_FC__SHIFT 0x16 +#define CB_HW_CONTROL_3__DISABLE_NACK_PROCESSING_DC__SHIFT 0x17 +#define CB_HW_CONTROL_3__DISABLE_NACK_PROCESSING_CM__SHIFT 0x18 +#define CB_HW_CONTROL_3__DISABLE_NACK_COLOR_RD_WR_OPT__SHIFT 0x19 +#define CB_HW_CONTROL_3__DISABLE_BLENDER_CLOCK_GATING__SHIFT 0x1a +#define CB_HW_CONTROL_3__DISABLE_DUALSRC_WITH_OBJPRIMID_FIX__SHIFT 0x1b +#define CB_HW_CONTROL_3__COLOR_CACHE_PREFETCH_NUM_CLS__SHIFT 0x1c +#define CB_HW_CONTROL_3__DISABLE_SLOW_MODE_EMPTY_HALF_QUAD_KILL_MASK 0x00000001L +#define CB_HW_CONTROL_3__RAM_ADDRESS_CONFLICTS_DISALLOWED_MASK 0x00000002L +#define CB_HW_CONTROL_3__DISABLE_FAST_CLEAR_FETCH_OPT_MASK 0x00000004L +#define CB_HW_CONTROL_3__DISABLE_QUAD_MARKER_DROP_STOP_MASK 0x00000008L +#define CB_HW_CONTROL_3__DISABLE_OVERWRITE_COMBINER_CAM_CLR_MASK 0x00000010L +#define CB_HW_CONTROL_3__DISABLE_CC_CACHE_OVWR_STATUS_ACCUM_MASK 0x00000020L +#define CB_HW_CONTROL_3__DISABLE_CC_CACHE_OVWR_KEY_MOD_MASK 0x00000040L +#define CB_HW_CONTROL_3__DISABLE_CC_CACHE_PANIC_GATING_MASK 0x00000080L +#define CB_HW_CONTROL_3__DISABLE_OVERWRITE_COMBINER_TARGET_MASK_VALIDATION_MASK 0x00000100L +#define CB_HW_CONTROL_3__SPLIT_ALL_FAST_MODE_TRANSFERS_MASK 0x00000200L +#define CB_HW_CONTROL_3__DISABLE_SHADER_BLEND_OPTS_MASK 0x00000400L +#define CB_HW_CONTROL_3__DISABLE_CMASK_LAST_QUAD_INSERTION_MASK 0x00000800L +#define CB_HW_CONTROL_3__DISABLE_ROP3_FIXES_OF_BUG_511967_MASK 0x00001000L +#define CB_HW_CONTROL_3__DISABLE_ROP3_FIXES_OF_BUG_520657_MASK 0x00002000L +#define CB_HW_CONTROL_3__DISABLE_OC_FIXES_OF_BUG_522542_MASK 0x00004000L +#define CB_HW_CONTROL_3__FORCE_RMI_LAST_HIGH_MASK 0x00008000L +#define CB_HW_CONTROL_3__FORCE_RMI_CLKEN_HIGH_MASK 0x00010000L +#define CB_HW_CONTROL_3__DISABLE_EARLY_WRACKS_CC_MASK 0x00020000L +#define CB_HW_CONTROL_3__DISABLE_EARLY_WRACKS_FC_MASK 0x00040000L +#define CB_HW_CONTROL_3__DISABLE_EARLY_WRACKS_DC_MASK 0x00080000L +#define CB_HW_CONTROL_3__DISABLE_EARLY_WRACKS_CM_MASK 0x00100000L +#define CB_HW_CONTROL_3__DISABLE_NACK_PROCESSING_CC_MASK 0x00200000L +#define CB_HW_CONTROL_3__DISABLE_NACK_PROCESSING_FC_MASK 0x00400000L +#define CB_HW_CONTROL_3__DISABLE_NACK_PROCESSING_DC_MASK 0x00800000L +#define CB_HW_CONTROL_3__DISABLE_NACK_PROCESSING_CM_MASK 0x01000000L +#define CB_HW_CONTROL_3__DISABLE_NACK_COLOR_RD_WR_OPT_MASK 0x02000000L +#define CB_HW_CONTROL_3__DISABLE_BLENDER_CLOCK_GATING_MASK 0x04000000L +#define CB_HW_CONTROL_3__DISABLE_DUALSRC_WITH_OBJPRIMID_FIX_MASK 0x08000000L +#define CB_HW_CONTROL_3__COLOR_CACHE_PREFETCH_NUM_CLS_MASK 0x30000000L +//CB_HW_MEM_ARBITER_RD +#define CB_HW_MEM_ARBITER_RD__MODE__SHIFT 0x0 +#define CB_HW_MEM_ARBITER_RD__IGNORE_URGENT_AGE__SHIFT 0x2 +#define CB_HW_MEM_ARBITER_RD__BREAK_GROUP_AGE__SHIFT 0x6 +#define CB_HW_MEM_ARBITER_RD__WEIGHT_CC__SHIFT 0xa +#define CB_HW_MEM_ARBITER_RD__WEIGHT_FC__SHIFT 0xc +#define CB_HW_MEM_ARBITER_RD__WEIGHT_CM__SHIFT 0xe +#define CB_HW_MEM_ARBITER_RD__WEIGHT_DC__SHIFT 0x10 +#define CB_HW_MEM_ARBITER_RD__WEIGHT_DECAY_REQS__SHIFT 0x12 +#define CB_HW_MEM_ARBITER_RD__WEIGHT_DECAY_NOREQS__SHIFT 0x14 +#define CB_HW_MEM_ARBITER_RD__WEIGHT_IGNORE_NUM_TIDS__SHIFT 0x16 +#define CB_HW_MEM_ARBITER_RD__SCALE_AGE__SHIFT 0x17 +#define CB_HW_MEM_ARBITER_RD__SCALE_WEIGHT__SHIFT 0x1a +#define CB_HW_MEM_ARBITER_RD__SEND_LASTS_WITHIN_GROUPS__SHIFT 0x1d +#define CB_HW_MEM_ARBITER_RD__MODE_MASK 0x00000003L +#define CB_HW_MEM_ARBITER_RD__IGNORE_URGENT_AGE_MASK 0x0000003CL +#define CB_HW_MEM_ARBITER_RD__BREAK_GROUP_AGE_MASK 0x000003C0L +#define CB_HW_MEM_ARBITER_RD__WEIGHT_CC_MASK 0x00000C00L +#define CB_HW_MEM_ARBITER_RD__WEIGHT_FC_MASK 0x00003000L +#define CB_HW_MEM_ARBITER_RD__WEIGHT_CM_MASK 0x0000C000L +#define CB_HW_MEM_ARBITER_RD__WEIGHT_DC_MASK 0x00030000L +#define CB_HW_MEM_ARBITER_RD__WEIGHT_DECAY_REQS_MASK 0x000C0000L +#define CB_HW_MEM_ARBITER_RD__WEIGHT_DECAY_NOREQS_MASK 0x00300000L +#define CB_HW_MEM_ARBITER_RD__WEIGHT_IGNORE_NUM_TIDS_MASK 0x00400000L +#define CB_HW_MEM_ARBITER_RD__SCALE_AGE_MASK 0x03800000L +#define CB_HW_MEM_ARBITER_RD__SCALE_WEIGHT_MASK 0x1C000000L +#define CB_HW_MEM_ARBITER_RD__SEND_LASTS_WITHIN_GROUPS_MASK 0x20000000L +//CB_HW_MEM_ARBITER_WR +#define CB_HW_MEM_ARBITER_WR__MODE__SHIFT 0x0 +#define CB_HW_MEM_ARBITER_WR__IGNORE_URGENT_AGE__SHIFT 0x2 +#define CB_HW_MEM_ARBITER_WR__BREAK_GROUP_AGE__SHIFT 0x6 +#define CB_HW_MEM_ARBITER_WR__WEIGHT_CC__SHIFT 0xa +#define CB_HW_MEM_ARBITER_WR__WEIGHT_FC__SHIFT 0xc +#define CB_HW_MEM_ARBITER_WR__WEIGHT_CM__SHIFT 0xe +#define CB_HW_MEM_ARBITER_WR__WEIGHT_DC__SHIFT 0x10 +#define CB_HW_MEM_ARBITER_WR__WEIGHT_DECAY_REQS__SHIFT 0x12 +#define CB_HW_MEM_ARBITER_WR__WEIGHT_DECAY_NOREQS__SHIFT 0x14 +#define CB_HW_MEM_ARBITER_WR__WEIGHT_IGNORE_BYTE_MASK__SHIFT 0x16 +#define CB_HW_MEM_ARBITER_WR__SCALE_AGE__SHIFT 0x17 +#define CB_HW_MEM_ARBITER_WR__SCALE_WEIGHT__SHIFT 0x1a +#define CB_HW_MEM_ARBITER_WR__SEND_LASTS_WITHIN_GROUPS__SHIFT 0x1d +#define CB_HW_MEM_ARBITER_WR__MODE_MASK 0x00000003L +#define CB_HW_MEM_ARBITER_WR__IGNORE_URGENT_AGE_MASK 0x0000003CL +#define CB_HW_MEM_ARBITER_WR__BREAK_GROUP_AGE_MASK 0x000003C0L +#define CB_HW_MEM_ARBITER_WR__WEIGHT_CC_MASK 0x00000C00L +#define CB_HW_MEM_ARBITER_WR__WEIGHT_FC_MASK 0x00003000L +#define CB_HW_MEM_ARBITER_WR__WEIGHT_CM_MASK 0x0000C000L +#define CB_HW_MEM_ARBITER_WR__WEIGHT_DC_MASK 0x00030000L +#define CB_HW_MEM_ARBITER_WR__WEIGHT_DECAY_REQS_MASK 0x000C0000L +#define CB_HW_MEM_ARBITER_WR__WEIGHT_DECAY_NOREQS_MASK 0x00300000L +#define CB_HW_MEM_ARBITER_WR__WEIGHT_IGNORE_BYTE_MASK_MASK 0x00400000L +#define CB_HW_MEM_ARBITER_WR__SCALE_AGE_MASK 0x03800000L +#define CB_HW_MEM_ARBITER_WR__SCALE_WEIGHT_MASK 0x1C000000L +#define CB_HW_MEM_ARBITER_WR__SEND_LASTS_WITHIN_GROUPS_MASK 0x20000000L +//CB_DCC_CONFIG +#define CB_DCC_CONFIG__OVERWRITE_COMBINER_DEPTH__SHIFT 0x0 +#define CB_DCC_CONFIG__OVERWRITE_COMBINER_DISABLE__SHIFT 0x5 +#define CB_DCC_CONFIG__OVERWRITE_COMBINER_CC_POP_DISABLE__SHIFT 0x6 +#define CB_DCC_CONFIG__FC_RDLAT_KEYID_FIFO_DEPTH__SHIFT 0x8 +#define CB_DCC_CONFIG__READ_RETURN_SKID_FIFO_DEPTH__SHIFT 0x10 +#define CB_DCC_CONFIG__DCC_CACHE_EVICT_POINT__SHIFT 0x18 +#define CB_DCC_CONFIG__DCC_CACHE_NUM_TAGS__SHIFT 0x1c +#define CB_DCC_CONFIG__OVERWRITE_COMBINER_DEPTH_MASK 0x0000001FL +#define CB_DCC_CONFIG__OVERWRITE_COMBINER_DISABLE_MASK 0x00000020L +#define CB_DCC_CONFIG__OVERWRITE_COMBINER_CC_POP_DISABLE_MASK 0x00000040L +#define CB_DCC_CONFIG__FC_RDLAT_KEYID_FIFO_DEPTH_MASK 0x0000FF00L +#define CB_DCC_CONFIG__READ_RETURN_SKID_FIFO_DEPTH_MASK 0x007F0000L +#define CB_DCC_CONFIG__DCC_CACHE_EVICT_POINT_MASK 0x0F000000L +#define CB_DCC_CONFIG__DCC_CACHE_NUM_TAGS_MASK 0xF0000000L +//GC_USER_RB_REDUNDANCY +#define GC_USER_RB_REDUNDANCY__FAILED_RB0__SHIFT 0x8 +#define GC_USER_RB_REDUNDANCY__EN_REDUNDANCY0__SHIFT 0xc +#define GC_USER_RB_REDUNDANCY__FAILED_RB1__SHIFT 0x10 +#define GC_USER_RB_REDUNDANCY__EN_REDUNDANCY1__SHIFT 0x14 +#define GC_USER_RB_REDUNDANCY__FAILED_RB0_MASK 0x00000F00L +#define GC_USER_RB_REDUNDANCY__EN_REDUNDANCY0_MASK 0x00001000L +#define GC_USER_RB_REDUNDANCY__FAILED_RB1_MASK 0x000F0000L +#define GC_USER_RB_REDUNDANCY__EN_REDUNDANCY1_MASK 0x00100000L +//GC_USER_RB_BACKEND_DISABLE +#define GC_USER_RB_BACKEND_DISABLE__BACKEND_DISABLE__SHIFT 0x10 +#define GC_USER_RB_BACKEND_DISABLE__BACKEND_DISABLE_MASK 0x00FF0000L + + +// addressBlock: gc_ea_gceadec2 +//GCEA_EDC_CNT +#define GCEA_EDC_CNT__DRAMRD_CMDMEM_SEC_COUNT__SHIFT 0x0 +#define GCEA_EDC_CNT__DRAMRD_CMDMEM_DED_COUNT__SHIFT 0x2 +#define GCEA_EDC_CNT__DRAMWR_CMDMEM_SEC_COUNT__SHIFT 0x4 +#define GCEA_EDC_CNT__DRAMWR_CMDMEM_DED_COUNT__SHIFT 0x6 +#define GCEA_EDC_CNT__DRAMWR_DATAMEM_SEC_COUNT__SHIFT 0x8 +#define GCEA_EDC_CNT__DRAMWR_DATAMEM_DED_COUNT__SHIFT 0xa +#define GCEA_EDC_CNT__RRET_TAGMEM_SEC_COUNT__SHIFT 0xc +#define GCEA_EDC_CNT__RRET_TAGMEM_DED_COUNT__SHIFT 0xe +#define GCEA_EDC_CNT__WRET_TAGMEM_SEC_COUNT__SHIFT 0x10 +#define GCEA_EDC_CNT__WRET_TAGMEM_DED_COUNT__SHIFT 0x12 +#define GCEA_EDC_CNT__DRAMRD_PAGEMEM_SED_COUNT__SHIFT 0x14 +#define GCEA_EDC_CNT__DRAMWR_PAGEMEM_SED_COUNT__SHIFT 0x16 +#define GCEA_EDC_CNT__IORD_CMDMEM_SED_COUNT__SHIFT 0x18 +#define GCEA_EDC_CNT__IOWR_CMDMEM_SED_COUNT__SHIFT 0x1a +#define GCEA_EDC_CNT__IOWR_DATAMEM_SED_COUNT__SHIFT 0x1c +#define GCEA_EDC_CNT__DRAMRD_CMDMEM_SEC_COUNT_MASK 0x00000003L +#define GCEA_EDC_CNT__DRAMRD_CMDMEM_DED_COUNT_MASK 0x0000000CL +#define GCEA_EDC_CNT__DRAMWR_CMDMEM_SEC_COUNT_MASK 0x00000030L +#define GCEA_EDC_CNT__DRAMWR_CMDMEM_DED_COUNT_MASK 0x000000C0L +#define GCEA_EDC_CNT__DRAMWR_DATAMEM_SEC_COUNT_MASK 0x00000300L +#define GCEA_EDC_CNT__DRAMWR_DATAMEM_DED_COUNT_MASK 0x00000C00L +#define GCEA_EDC_CNT__RRET_TAGMEM_SEC_COUNT_MASK 0x00003000L +#define GCEA_EDC_CNT__RRET_TAGMEM_DED_COUNT_MASK 0x0000C000L +#define GCEA_EDC_CNT__WRET_TAGMEM_SEC_COUNT_MASK 0x00030000L +#define GCEA_EDC_CNT__WRET_TAGMEM_DED_COUNT_MASK 0x000C0000L +#define GCEA_EDC_CNT__DRAMRD_PAGEMEM_SED_COUNT_MASK 0x00300000L +#define GCEA_EDC_CNT__DRAMWR_PAGEMEM_SED_COUNT_MASK 0x00C00000L +#define GCEA_EDC_CNT__IORD_CMDMEM_SED_COUNT_MASK 0x03000000L +#define GCEA_EDC_CNT__IOWR_CMDMEM_SED_COUNT_MASK 0x0C000000L +#define GCEA_EDC_CNT__IOWR_DATAMEM_SED_COUNT_MASK 0x30000000L +//GCEA_EDC_CNT2 +#define GCEA_EDC_CNT2__GMIRD_CMDMEM_SEC_COUNT__SHIFT 0x0 +#define GCEA_EDC_CNT2__GMIRD_CMDMEM_DED_COUNT__SHIFT 0x2 +#define GCEA_EDC_CNT2__GMIWR_CMDMEM_SEC_COUNT__SHIFT 0x4 +#define GCEA_EDC_CNT2__GMIWR_CMDMEM_DED_COUNT__SHIFT 0x6 +#define GCEA_EDC_CNT2__GMIWR_DATAMEM_SEC_COUNT__SHIFT 0x8 +#define GCEA_EDC_CNT2__GMIWR_DATAMEM_DED_COUNT__SHIFT 0xa +#define GCEA_EDC_CNT2__GMIRD_PAGEMEM_SED_COUNT__SHIFT 0xc +#define GCEA_EDC_CNT2__GMIWR_PAGEMEM_SED_COUNT__SHIFT 0xe +#define GCEA_EDC_CNT2__GMIRD_CMDMEM_SEC_COUNT_MASK 0x00000003L +#define GCEA_EDC_CNT2__GMIRD_CMDMEM_DED_COUNT_MASK 0x0000000CL +#define GCEA_EDC_CNT2__GMIWR_CMDMEM_SEC_COUNT_MASK 0x00000030L +#define GCEA_EDC_CNT2__GMIWR_CMDMEM_DED_COUNT_MASK 0x000000C0L +#define GCEA_EDC_CNT2__GMIWR_DATAMEM_SEC_COUNT_MASK 0x00000300L +#define GCEA_EDC_CNT2__GMIWR_DATAMEM_DED_COUNT_MASK 0x00000C00L +#define GCEA_EDC_CNT2__GMIRD_PAGEMEM_SED_COUNT_MASK 0x00003000L +#define GCEA_EDC_CNT2__GMIWR_PAGEMEM_SED_COUNT_MASK 0x0000C000L +//GCEA_DSM_CNTL +#define GCEA_DSM_CNTL__DRAMRD_CMDMEM_DSM_IRRITATOR_DATA__SHIFT 0x0 +#define GCEA_DSM_CNTL__DRAMRD_CMDMEM_ENABLE_SINGLE_WRITE__SHIFT 0x2 +#define GCEA_DSM_CNTL__DRAMWR_CMDMEM_DSM_IRRITATOR_DATA__SHIFT 0x3 +#define GCEA_DSM_CNTL__DRAMWR_CMDMEM_ENABLE_SINGLE_WRITE__SHIFT 0x5 +#define GCEA_DSM_CNTL__DRAMWR_DATAMEM_DSM_IRRITATOR_DATA__SHIFT 0x6 +#define GCEA_DSM_CNTL__DRAMWR_DATAMEM_ENABLE_SINGLE_WRITE__SHIFT 0x8 +#define GCEA_DSM_CNTL__RRET_TAGMEM_DSM_IRRITATOR_DATA__SHIFT 0x9 +#define GCEA_DSM_CNTL__RRET_TAGMEM_ENABLE_SINGLE_WRITE__SHIFT 0xb +#define GCEA_DSM_CNTL__WRET_TAGMEM_DSM_IRRITATOR_DATA__SHIFT 0xc +#define GCEA_DSM_CNTL__WRET_TAGMEM_ENABLE_SINGLE_WRITE__SHIFT 0xe +#define GCEA_DSM_CNTL__GMIRD_CMDMEM_DSM_IRRITATOR_DATA__SHIFT 0xf +#define GCEA_DSM_CNTL__GMIRD_CMDMEM_ENABLE_SINGLE_WRITE__SHIFT 0x11 +#define GCEA_DSM_CNTL__GMIWR_CMDMEM_DSM_IRRITATOR_DATA__SHIFT 0x12 +#define GCEA_DSM_CNTL__GMIWR_CMDMEM_ENABLE_SINGLE_WRITE__SHIFT 0x14 +#define GCEA_DSM_CNTL__GMIWR_DATAMEM_DSM_IRRITATOR_DATA__SHIFT 0x15 +#define GCEA_DSM_CNTL__GMIWR_DATAMEM_ENABLE_SINGLE_WRITE__SHIFT 0x17 +#define GCEA_DSM_CNTL__DRAMRD_CMDMEM_DSM_IRRITATOR_DATA_MASK 0x00000003L +#define GCEA_DSM_CNTL__DRAMRD_CMDMEM_ENABLE_SINGLE_WRITE_MASK 0x00000004L +#define GCEA_DSM_CNTL__DRAMWR_CMDMEM_DSM_IRRITATOR_DATA_MASK 0x00000018L +#define GCEA_DSM_CNTL__DRAMWR_CMDMEM_ENABLE_SINGLE_WRITE_MASK 0x00000020L +#define GCEA_DSM_CNTL__DRAMWR_DATAMEM_DSM_IRRITATOR_DATA_MASK 0x000000C0L +#define GCEA_DSM_CNTL__DRAMWR_DATAMEM_ENABLE_SINGLE_WRITE_MASK 0x00000100L +#define GCEA_DSM_CNTL__RRET_TAGMEM_DSM_IRRITATOR_DATA_MASK 0x00000600L +#define GCEA_DSM_CNTL__RRET_TAGMEM_ENABLE_SINGLE_WRITE_MASK 0x00000800L +#define GCEA_DSM_CNTL__WRET_TAGMEM_DSM_IRRITATOR_DATA_MASK 0x00003000L +#define GCEA_DSM_CNTL__WRET_TAGMEM_ENABLE_SINGLE_WRITE_MASK 0x00004000L +#define GCEA_DSM_CNTL__GMIRD_CMDMEM_DSM_IRRITATOR_DATA_MASK 0x00018000L +#define GCEA_DSM_CNTL__GMIRD_CMDMEM_ENABLE_SINGLE_WRITE_MASK 0x00020000L +#define GCEA_DSM_CNTL__GMIWR_CMDMEM_DSM_IRRITATOR_DATA_MASK 0x000C0000L +#define GCEA_DSM_CNTL__GMIWR_CMDMEM_ENABLE_SINGLE_WRITE_MASK 0x00100000L +#define GCEA_DSM_CNTL__GMIWR_DATAMEM_DSM_IRRITATOR_DATA_MASK 0x00600000L +#define GCEA_DSM_CNTL__GMIWR_DATAMEM_ENABLE_SINGLE_WRITE_MASK 0x00800000L +//GCEA_DSM_CNTLA +#define GCEA_DSM_CNTLA__DRAMRD_PAGEMEM_DSM_IRRITATOR_DATA__SHIFT 0x0 +#define GCEA_DSM_CNTLA__DRAMRD_PAGEMEM_ENABLE_SINGLE_WRITE__SHIFT 0x2 +#define GCEA_DSM_CNTLA__DRAMWR_PAGEMEM_DSM_IRRITATOR_DATA__SHIFT 0x3 +#define GCEA_DSM_CNTLA__DRAMWR_PAGEMEM_ENABLE_SINGLE_WRITE__SHIFT 0x5 +#define GCEA_DSM_CNTLA__IORD_CMDMEM_DSM_IRRITATOR_DATA__SHIFT 0x6 +#define GCEA_DSM_CNTLA__IORD_CMDMEM_ENABLE_SINGLE_WRITE__SHIFT 0x8 +#define GCEA_DSM_CNTLA__IOWR_CMDMEM_DSM_IRRITATOR_DATA__SHIFT 0x9 +#define GCEA_DSM_CNTLA__IOWR_CMDMEM_ENABLE_SINGLE_WRITE__SHIFT 0xb +#define GCEA_DSM_CNTLA__IOWR_DATAMEM_DSM_IRRITATOR_DATA__SHIFT 0xc +#define GCEA_DSM_CNTLA__IOWR_DATAMEM_ENABLE_SINGLE_WRITE__SHIFT 0xe +#define GCEA_DSM_CNTLA__GMIRD_PAGEMEM_DSM_IRRITATOR_DATA__SHIFT 0xf +#define GCEA_DSM_CNTLA__GMIRD_PAGEMEM_ENABLE_SINGLE_WRITE__SHIFT 0x11 +#define GCEA_DSM_CNTLA__GMIWR_PAGEMEM_DSM_IRRITATOR_DATA__SHIFT 0x12 +#define GCEA_DSM_CNTLA__GMIWR_PAGEMEM_ENABLE_SINGLE_WRITE__SHIFT 0x14 +#define GCEA_DSM_CNTLA__DRAMRD_PAGEMEM_DSM_IRRITATOR_DATA_MASK 0x00000003L +#define GCEA_DSM_CNTLA__DRAMRD_PAGEMEM_ENABLE_SINGLE_WRITE_MASK 0x00000004L +#define GCEA_DSM_CNTLA__DRAMWR_PAGEMEM_DSM_IRRITATOR_DATA_MASK 0x00000018L +#define GCEA_DSM_CNTLA__DRAMWR_PAGEMEM_ENABLE_SINGLE_WRITE_MASK 0x00000020L +#define GCEA_DSM_CNTLA__IORD_CMDMEM_DSM_IRRITATOR_DATA_MASK 0x000000C0L +#define GCEA_DSM_CNTLA__IORD_CMDMEM_ENABLE_SINGLE_WRITE_MASK 0x00000100L +#define GCEA_DSM_CNTLA__IOWR_CMDMEM_DSM_IRRITATOR_DATA_MASK 0x00000600L +#define GCEA_DSM_CNTLA__IOWR_CMDMEM_ENABLE_SINGLE_WRITE_MASK 0x00000800L +#define GCEA_DSM_CNTLA__IOWR_DATAMEM_DSM_IRRITATOR_DATA_MASK 0x00003000L +#define GCEA_DSM_CNTLA__IOWR_DATAMEM_ENABLE_SINGLE_WRITE_MASK 0x00004000L +#define GCEA_DSM_CNTLA__GMIRD_PAGEMEM_DSM_IRRITATOR_DATA_MASK 0x00018000L +#define GCEA_DSM_CNTLA__GMIRD_PAGEMEM_ENABLE_SINGLE_WRITE_MASK 0x00020000L +#define GCEA_DSM_CNTLA__GMIWR_PAGEMEM_DSM_IRRITATOR_DATA_MASK 0x000C0000L +#define GCEA_DSM_CNTLA__GMIWR_PAGEMEM_ENABLE_SINGLE_WRITE_MASK 0x00100000L +//GCEA_DSM_CNTLB +//GCEA_DSM_CNTL2 +#define GCEA_DSM_CNTL2__DRAMRD_CMDMEM_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define GCEA_DSM_CNTL2__DRAMRD_CMDMEM_SELECT_INJECT_DELAY__SHIFT 0x2 +#define GCEA_DSM_CNTL2__DRAMWR_CMDMEM_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define GCEA_DSM_CNTL2__DRAMWR_CMDMEM_SELECT_INJECT_DELAY__SHIFT 0x5 +#define GCEA_DSM_CNTL2__DRAMWR_DATAMEM_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define GCEA_DSM_CNTL2__DRAMWR_DATAMEM_SELECT_INJECT_DELAY__SHIFT 0x8 +#define GCEA_DSM_CNTL2__RRET_TAGMEM_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define GCEA_DSM_CNTL2__RRET_TAGMEM_SELECT_INJECT_DELAY__SHIFT 0xb +#define GCEA_DSM_CNTL2__WRET_TAGMEM_ENABLE_ERROR_INJECT__SHIFT 0xc +#define GCEA_DSM_CNTL2__WRET_TAGMEM_SELECT_INJECT_DELAY__SHIFT 0xe +#define GCEA_DSM_CNTL2__GMIRD_CMDMEM_ENABLE_ERROR_INJECT__SHIFT 0xf +#define GCEA_DSM_CNTL2__GMIRD_CMDMEM_SELECT_INJECT_DELAY__SHIFT 0x11 +#define GCEA_DSM_CNTL2__GMIWR_CMDMEM_ENABLE_ERROR_INJECT__SHIFT 0x12 +#define GCEA_DSM_CNTL2__GMIWR_CMDMEM_SELECT_INJECT_DELAY__SHIFT 0x14 +#define GCEA_DSM_CNTL2__GMIWR_DATAMEM_ENABLE_ERROR_INJECT__SHIFT 0x15 +#define GCEA_DSM_CNTL2__GMIWR_DATAMEM_SELECT_INJECT_DELAY__SHIFT 0x17 +#define GCEA_DSM_CNTL2__INJECT_DELAY__SHIFT 0x1a +#define GCEA_DSM_CNTL2__DRAMRD_CMDMEM_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define GCEA_DSM_CNTL2__DRAMRD_CMDMEM_SELECT_INJECT_DELAY_MASK 0x00000004L +#define GCEA_DSM_CNTL2__DRAMWR_CMDMEM_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define GCEA_DSM_CNTL2__DRAMWR_CMDMEM_SELECT_INJECT_DELAY_MASK 0x00000020L +#define GCEA_DSM_CNTL2__DRAMWR_DATAMEM_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define GCEA_DSM_CNTL2__DRAMWR_DATAMEM_SELECT_INJECT_DELAY_MASK 0x00000100L +#define GCEA_DSM_CNTL2__RRET_TAGMEM_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define GCEA_DSM_CNTL2__RRET_TAGMEM_SELECT_INJECT_DELAY_MASK 0x00000800L +#define GCEA_DSM_CNTL2__WRET_TAGMEM_ENABLE_ERROR_INJECT_MASK 0x00003000L +#define GCEA_DSM_CNTL2__WRET_TAGMEM_SELECT_INJECT_DELAY_MASK 0x00004000L +#define GCEA_DSM_CNTL2__GMIRD_CMDMEM_ENABLE_ERROR_INJECT_MASK 0x00018000L +#define GCEA_DSM_CNTL2__GMIRD_CMDMEM_SELECT_INJECT_DELAY_MASK 0x00020000L +#define GCEA_DSM_CNTL2__GMIWR_CMDMEM_ENABLE_ERROR_INJECT_MASK 0x000C0000L +#define GCEA_DSM_CNTL2__GMIWR_CMDMEM_SELECT_INJECT_DELAY_MASK 0x00100000L +#define GCEA_DSM_CNTL2__GMIWR_DATAMEM_ENABLE_ERROR_INJECT_MASK 0x00600000L +#define GCEA_DSM_CNTL2__GMIWR_DATAMEM_SELECT_INJECT_DELAY_MASK 0x00800000L +#define GCEA_DSM_CNTL2__INJECT_DELAY_MASK 0xFC000000L +//GCEA_DSM_CNTL2A +#define GCEA_DSM_CNTL2A__DRAMRD_PAGEMEM_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define GCEA_DSM_CNTL2A__DRAMRD_PAGEMEM_SELECT_INJECT_DELAY__SHIFT 0x2 +#define GCEA_DSM_CNTL2A__DRAMWR_PAGEMEM_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define GCEA_DSM_CNTL2A__DRAMWR_PAGEMEM_SELECT_INJECT_DELAY__SHIFT 0x5 +#define GCEA_DSM_CNTL2A__IORD_CMDMEM_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define GCEA_DSM_CNTL2A__IORD_CMDMEM_SELECT_INJECT_DELAY__SHIFT 0x8 +#define GCEA_DSM_CNTL2A__IOWR_CMDMEM_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define GCEA_DSM_CNTL2A__IOWR_CMDMEM_SELECT_INJECT_DELAY__SHIFT 0xb +#define GCEA_DSM_CNTL2A__IOWR_DATAMEM_ENABLE_ERROR_INJECT__SHIFT 0xc +#define GCEA_DSM_CNTL2A__IOWR_DATAMEM_SELECT_INJECT_DELAY__SHIFT 0xe +#define GCEA_DSM_CNTL2A__GMIRD_PAGEMEM_ENABLE_ERROR_INJECT__SHIFT 0xf +#define GCEA_DSM_CNTL2A__GMIRD_PAGEMEM_SELECT_INJECT_DELAY__SHIFT 0x11 +#define GCEA_DSM_CNTL2A__GMIWR_PAGEMEM_ENABLE_ERROR_INJECT__SHIFT 0x12 +#define GCEA_DSM_CNTL2A__GMIWR_PAGEMEM_SELECT_INJECT_DELAY__SHIFT 0x14 +#define GCEA_DSM_CNTL2A__DRAMRD_PAGEMEM_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define GCEA_DSM_CNTL2A__DRAMRD_PAGEMEM_SELECT_INJECT_DELAY_MASK 0x00000004L +#define GCEA_DSM_CNTL2A__DRAMWR_PAGEMEM_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define GCEA_DSM_CNTL2A__DRAMWR_PAGEMEM_SELECT_INJECT_DELAY_MASK 0x00000020L +#define GCEA_DSM_CNTL2A__IORD_CMDMEM_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define GCEA_DSM_CNTL2A__IORD_CMDMEM_SELECT_INJECT_DELAY_MASK 0x00000100L +#define GCEA_DSM_CNTL2A__IOWR_CMDMEM_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define GCEA_DSM_CNTL2A__IOWR_CMDMEM_SELECT_INJECT_DELAY_MASK 0x00000800L +#define GCEA_DSM_CNTL2A__IOWR_DATAMEM_ENABLE_ERROR_INJECT_MASK 0x00003000L +#define GCEA_DSM_CNTL2A__IOWR_DATAMEM_SELECT_INJECT_DELAY_MASK 0x00004000L +#define GCEA_DSM_CNTL2A__GMIRD_PAGEMEM_ENABLE_ERROR_INJECT_MASK 0x00018000L +#define GCEA_DSM_CNTL2A__GMIRD_PAGEMEM_SELECT_INJECT_DELAY_MASK 0x00020000L +#define GCEA_DSM_CNTL2A__GMIWR_PAGEMEM_ENABLE_ERROR_INJECT_MASK 0x000C0000L +#define GCEA_DSM_CNTL2A__GMIWR_PAGEMEM_SELECT_INJECT_DELAY_MASK 0x00100000L +//GCEA_DSM_CNTL2B +//GCEA_TCC_XBR_CREDITS +#define GCEA_TCC_XBR_CREDITS__DRAM_RD_LIMIT__SHIFT 0x0 +#define GCEA_TCC_XBR_CREDITS__DRAM_RD_RESERVE__SHIFT 0x6 +#define GCEA_TCC_XBR_CREDITS__IO_RD_LIMIT__SHIFT 0x8 +#define GCEA_TCC_XBR_CREDITS__IO_RD_RESERVE__SHIFT 0xe +#define GCEA_TCC_XBR_CREDITS__DRAM_WR_LIMIT__SHIFT 0x10 +#define GCEA_TCC_XBR_CREDITS__DRAM_WR_RESERVE__SHIFT 0x16 +#define GCEA_TCC_XBR_CREDITS__IO_WR_LIMIT__SHIFT 0x18 +#define GCEA_TCC_XBR_CREDITS__IO_WR_RESERVE__SHIFT 0x1e +#define GCEA_TCC_XBR_CREDITS__DRAM_RD_LIMIT_MASK 0x0000003FL +#define GCEA_TCC_XBR_CREDITS__DRAM_RD_RESERVE_MASK 0x000000C0L +#define GCEA_TCC_XBR_CREDITS__IO_RD_LIMIT_MASK 0x00003F00L +#define GCEA_TCC_XBR_CREDITS__IO_RD_RESERVE_MASK 0x0000C000L +#define GCEA_TCC_XBR_CREDITS__DRAM_WR_LIMIT_MASK 0x003F0000L +#define GCEA_TCC_XBR_CREDITS__DRAM_WR_RESERVE_MASK 0x00C00000L +#define GCEA_TCC_XBR_CREDITS__IO_WR_LIMIT_MASK 0x3F000000L +#define GCEA_TCC_XBR_CREDITS__IO_WR_RESERVE_MASK 0xC0000000L +//GCEA_TCC_XBR_MAXBURST +#define GCEA_TCC_XBR_MAXBURST__DRAM_RD__SHIFT 0x0 +#define GCEA_TCC_XBR_MAXBURST__IO_RD__SHIFT 0x4 +#define GCEA_TCC_XBR_MAXBURST__DRAM_WR__SHIFT 0x8 +#define GCEA_TCC_XBR_MAXBURST__IO_WR__SHIFT 0xc +#define GCEA_TCC_XBR_MAXBURST__DRAM_RD_MASK 0x0000000FL +#define GCEA_TCC_XBR_MAXBURST__IO_RD_MASK 0x000000F0L +#define GCEA_TCC_XBR_MAXBURST__DRAM_WR_MASK 0x00000F00L +#define GCEA_TCC_XBR_MAXBURST__IO_WR_MASK 0x0000F000L +//GCEA_PROBE_CNTL +#define GCEA_PROBE_CNTL__REQ2RSP_DELAY__SHIFT 0x0 +#define GCEA_PROBE_CNTL__PRB_FILTER_DISABLE__SHIFT 0x5 +#define GCEA_PROBE_CNTL__REQ2RSP_DELAY_MASK 0x0000001FL +#define GCEA_PROBE_CNTL__PRB_FILTER_DISABLE_MASK 0x00000020L +//GCEA_PROBE_MAP +#define GCEA_PROBE_MAP__CHADDR0_TO_RIGHTTCC__SHIFT 0x0 +#define GCEA_PROBE_MAP__CHADDR1_TO_RIGHTTCC__SHIFT 0x1 +#define GCEA_PROBE_MAP__CHADDR2_TO_RIGHTTCC__SHIFT 0x2 +#define GCEA_PROBE_MAP__CHADDR3_TO_RIGHTTCC__SHIFT 0x3 +#define GCEA_PROBE_MAP__CHADDR4_TO_RIGHTTCC__SHIFT 0x4 +#define GCEA_PROBE_MAP__CHADDR5_TO_RIGHTTCC__SHIFT 0x5 +#define GCEA_PROBE_MAP__CHADDR6_TO_RIGHTTCC__SHIFT 0x6 +#define GCEA_PROBE_MAP__CHADDR7_TO_RIGHTTCC__SHIFT 0x7 +#define GCEA_PROBE_MAP__CHADDR8_TO_RIGHTTCC__SHIFT 0x8 +#define GCEA_PROBE_MAP__CHADDR9_TO_RIGHTTCC__SHIFT 0x9 +#define GCEA_PROBE_MAP__CHADDR10_TO_RIGHTTCC__SHIFT 0xa +#define GCEA_PROBE_MAP__CHADDR11_TO_RIGHTTCC__SHIFT 0xb +#define GCEA_PROBE_MAP__CHADDR12_TO_RIGHTTCC__SHIFT 0xc +#define GCEA_PROBE_MAP__CHADDR13_TO_RIGHTTCC__SHIFT 0xd +#define GCEA_PROBE_MAP__CHADDR14_TO_RIGHTTCC__SHIFT 0xe +#define GCEA_PROBE_MAP__CHADDR15_TO_RIGHTTCC__SHIFT 0xf +#define GCEA_PROBE_MAP__INTLV_SIZE__SHIFT 0x10 +#define GCEA_PROBE_MAP__CHADDR0_TO_RIGHTTCC_MASK 0x00000001L +#define GCEA_PROBE_MAP__CHADDR1_TO_RIGHTTCC_MASK 0x00000002L +#define GCEA_PROBE_MAP__CHADDR2_TO_RIGHTTCC_MASK 0x00000004L +#define GCEA_PROBE_MAP__CHADDR3_TO_RIGHTTCC_MASK 0x00000008L +#define GCEA_PROBE_MAP__CHADDR4_TO_RIGHTTCC_MASK 0x00000010L +#define GCEA_PROBE_MAP__CHADDR5_TO_RIGHTTCC_MASK 0x00000020L +#define GCEA_PROBE_MAP__CHADDR6_TO_RIGHTTCC_MASK 0x00000040L +#define GCEA_PROBE_MAP__CHADDR7_TO_RIGHTTCC_MASK 0x00000080L +#define GCEA_PROBE_MAP__CHADDR8_TO_RIGHTTCC_MASK 0x00000100L +#define GCEA_PROBE_MAP__CHADDR9_TO_RIGHTTCC_MASK 0x00000200L +#define GCEA_PROBE_MAP__CHADDR10_TO_RIGHTTCC_MASK 0x00000400L +#define GCEA_PROBE_MAP__CHADDR11_TO_RIGHTTCC_MASK 0x00000800L +#define GCEA_PROBE_MAP__CHADDR12_TO_RIGHTTCC_MASK 0x00001000L +#define GCEA_PROBE_MAP__CHADDR13_TO_RIGHTTCC_MASK 0x00002000L +#define GCEA_PROBE_MAP__CHADDR14_TO_RIGHTTCC_MASK 0x00004000L +#define GCEA_PROBE_MAP__CHADDR15_TO_RIGHTTCC_MASK 0x00008000L +#define GCEA_PROBE_MAP__INTLV_SIZE_MASK 0x00030000L +//GCEA_ERR_STATUS +#define GCEA_ERR_STATUS__SDP_RDRSP_STATUS__SHIFT 0x0 +#define GCEA_ERR_STATUS__SDP_WRRSP_STATUS__SHIFT 0x4 +#define GCEA_ERR_STATUS__SDP_RDRSP_DATAPARITY_ERROR__SHIFT 0x8 +#define GCEA_ERR_STATUS__CLEAR_ERROR_STATUS__SHIFT 0x9 +#define GCEA_ERR_STATUS__BUSY_ON_ERROR__SHIFT 0xa +#define GCEA_ERR_STATUS__SDP_RDRSP_STATUS_MASK 0x0000000FL +#define GCEA_ERR_STATUS__SDP_WRRSP_STATUS_MASK 0x000000F0L +#define GCEA_ERR_STATUS__SDP_RDRSP_DATAPARITY_ERROR_MASK 0x00000100L +#define GCEA_ERR_STATUS__CLEAR_ERROR_STATUS_MASK 0x00000200L +#define GCEA_ERR_STATUS__BUSY_ON_ERROR_MASK 0x00000400L +//GCEA_MISC2 +#define GCEA_MISC2__CSGROUP_SWAP_IN_DRAM_ARB__SHIFT 0x0 +#define GCEA_MISC2__CSGROUP_SWAP_IN_GMI_ARB__SHIFT 0x1 +#define GCEA_MISC2__CSGRP_BURST_LIMIT_DATA_DRAM__SHIFT 0x2 +#define GCEA_MISC2__CSGRP_BURST_LIMIT_DATA_GMI__SHIFT 0x7 +#define GCEA_MISC2__CSGROUP_SWAP_IN_DRAM_ARB_MASK 0x00000001L +#define GCEA_MISC2__CSGROUP_SWAP_IN_GMI_ARB_MASK 0x00000002L +#define GCEA_MISC2__CSGRP_BURST_LIMIT_DATA_DRAM_MASK 0x0000007CL +#define GCEA_MISC2__CSGRP_BURST_LIMIT_DATA_GMI_MASK 0x00000F80L +//GCEA_SDP_BACKDOOR_CMDCREDITS0 +#define GCEA_SDP_BACKDOOR_CMDCREDITS0__CREDITS_RECEIVED__SHIFT 0x0 +#define GCEA_SDP_BACKDOOR_CMDCREDITS0__CREDITS_RECEIVED_MASK 0xFFFFFFFFL +//GCEA_SDP_BACKDOOR_CMDCREDITS1 +#define GCEA_SDP_BACKDOOR_CMDCREDITS1__CREDITS_RECEIVED__SHIFT 0x0 +#define GCEA_SDP_BACKDOOR_CMDCREDITS1__CREDITS_RECEIVED_MASK 0x7FFFFFFFL +//GCEA_SDP_BACKDOOR_DATACREDITS0 +#define GCEA_SDP_BACKDOOR_DATACREDITS0__CREDITS_RECEIVED__SHIFT 0x0 +#define GCEA_SDP_BACKDOOR_DATACREDITS0__CREDITS_RECEIVED_MASK 0xFFFFFFFFL +//GCEA_SDP_BACKDOOR_DATACREDITS1 +#define GCEA_SDP_BACKDOOR_DATACREDITS1__CREDITS_RECEIVED__SHIFT 0x0 +#define GCEA_SDP_BACKDOOR_DATACREDITS1__CREDITS_RECEIVED_MASK 0x7FFFFFFFL +//GCEA_SDP_BACKDOOR_MISCCREDITS +#define GCEA_SDP_BACKDOOR_MISCCREDITS__RDRSP_CREDITS_RELEASED__SHIFT 0x0 +#define GCEA_SDP_BACKDOOR_MISCCREDITS__WRRSP_CREDITS_RELEASED__SHIFT 0x8 +#define GCEA_SDP_BACKDOOR_MISCCREDITS__PRB_REQ_CREDITS_RELEASED__SHIFT 0x10 +#define GCEA_SDP_BACKDOOR_MISCCREDITS__PRB_RSP_CREDITS_RECEIVED__SHIFT 0x17 +#define GCEA_SDP_BACKDOOR_MISCCREDITS__RDRSP_CREDITS_RELEASED_MASK 0x000000FFL +#define GCEA_SDP_BACKDOOR_MISCCREDITS__WRRSP_CREDITS_RELEASED_MASK 0x0000FF00L +#define GCEA_SDP_BACKDOOR_MISCCREDITS__PRB_REQ_CREDITS_RELEASED_MASK 0x007F0000L +#define GCEA_SDP_BACKDOOR_MISCCREDITS__PRB_RSP_CREDITS_RECEIVED_MASK 0x3F800000L +//GCEA_SDP_ENABLE +#define GCEA_SDP_ENABLE__ENABLE__SHIFT 0x0 +#define GCEA_SDP_ENABLE__ENABLE_MASK 0x00000001L + + +// addressBlock: gc_rmi_rmidec +//RMI_GENERAL_CNTL +#define RMI_GENERAL_CNTL__BURST_DISABLE__SHIFT 0x0 +#define RMI_GENERAL_CNTL__VMID_BYPASS_ENABLE__SHIFT 0x1 +#define RMI_GENERAL_CNTL__XBAR_MUX_CONFIG__SHIFT 0x11 +#define RMI_GENERAL_CNTL__RB0_HARVEST_EN__SHIFT 0x13 +#define RMI_GENERAL_CNTL__RB1_HARVEST_EN__SHIFT 0x14 +#define RMI_GENERAL_CNTL__LOOPBACK_DIS_BY_REQ_TYPE__SHIFT 0x15 +#define RMI_GENERAL_CNTL__XBAR_MUX_CONFIG_UPDATE__SHIFT 0x19 +#define RMI_GENERAL_CNTL__SKID_FIFO_0_OVERFLOW_ERROR_MASK__SHIFT 0x1a +#define RMI_GENERAL_CNTL__SKID_FIFO_0_UNDERFLOW_ERROR_MASK__SHIFT 0x1b +#define RMI_GENERAL_CNTL__SKID_FIFO_1_OVERFLOW_ERROR_MASK__SHIFT 0x1c +#define RMI_GENERAL_CNTL__SKID_FIFO_1_UNDERFLOW_ERROR_MASK__SHIFT 0x1d +#define RMI_GENERAL_CNTL__SKID_FIFO_FREESPACE_IS_ZERO_ERROR_MASK__SHIFT 0x1e +#define RMI_GENERAL_CNTL__BURST_DISABLE_MASK 0x00000001L +#define RMI_GENERAL_CNTL__VMID_BYPASS_ENABLE_MASK 0x0001FFFEL +#define RMI_GENERAL_CNTL__XBAR_MUX_CONFIG_MASK 0x00060000L +#define RMI_GENERAL_CNTL__RB0_HARVEST_EN_MASK 0x00080000L +#define RMI_GENERAL_CNTL__RB1_HARVEST_EN_MASK 0x00100000L +#define RMI_GENERAL_CNTL__LOOPBACK_DIS_BY_REQ_TYPE_MASK 0x01E00000L +#define RMI_GENERAL_CNTL__XBAR_MUX_CONFIG_UPDATE_MASK 0x02000000L +#define RMI_GENERAL_CNTL__SKID_FIFO_0_OVERFLOW_ERROR_MASK_MASK 0x04000000L +#define RMI_GENERAL_CNTL__SKID_FIFO_0_UNDERFLOW_ERROR_MASK_MASK 0x08000000L +#define RMI_GENERAL_CNTL__SKID_FIFO_1_OVERFLOW_ERROR_MASK_MASK 0x10000000L +#define RMI_GENERAL_CNTL__SKID_FIFO_1_UNDERFLOW_ERROR_MASK_MASK 0x20000000L +#define RMI_GENERAL_CNTL__SKID_FIFO_FREESPACE_IS_ZERO_ERROR_MASK_MASK 0x40000000L +//RMI_GENERAL_CNTL1 +#define RMI_GENERAL_CNTL1__EARLY_WRACK_ENABLE_PER_MTYPE__SHIFT 0x0 +#define RMI_GENERAL_CNTL1__TCIW0_64B_RD_STALL_MODE__SHIFT 0x4 +#define RMI_GENERAL_CNTL1__TCIW1_64B_RD_STALL_MODE__SHIFT 0x6 +#define RMI_GENERAL_CNTL1__EARLY_WRACK_DISABLE_FOR_LOOPBACK__SHIFT 0x8 +#define RMI_GENERAL_CNTL1__POLICY_OVERRIDE_VALUE__SHIFT 0x9 +#define RMI_GENERAL_CNTL1__POLICY_OVERRIDE__SHIFT 0xa +#define RMI_GENERAL_CNTL1__UTCL1_PROBE0_RR_ARB_BURST_HINT_EN__SHIFT 0xb +#define RMI_GENERAL_CNTL1__UTCL1_PROBE1_RR_ARB_BURST_HINT_EN__SHIFT 0xc +#define RMI_GENERAL_CNTL1__EARLY_WRACK_ENABLE_PER_MTYPE_MASK 0x0000000FL +#define RMI_GENERAL_CNTL1__TCIW0_64B_RD_STALL_MODE_MASK 0x00000030L +#define RMI_GENERAL_CNTL1__TCIW1_64B_RD_STALL_MODE_MASK 0x000000C0L +#define RMI_GENERAL_CNTL1__EARLY_WRACK_DISABLE_FOR_LOOPBACK_MASK 0x00000100L +#define RMI_GENERAL_CNTL1__POLICY_OVERRIDE_VALUE_MASK 0x00000200L +#define RMI_GENERAL_CNTL1__POLICY_OVERRIDE_MASK 0x00000400L +#define RMI_GENERAL_CNTL1__UTCL1_PROBE0_RR_ARB_BURST_HINT_EN_MASK 0x00000800L +#define RMI_GENERAL_CNTL1__UTCL1_PROBE1_RR_ARB_BURST_HINT_EN_MASK 0x00001000L +//RMI_GENERAL_STATUS +#define RMI_GENERAL_STATUS__GENERAL_RMI_ERRORS_COMBINED__SHIFT 0x0 +#define RMI_GENERAL_STATUS__SKID_FIFO_0_OVERFLOW_ERROR__SHIFT 0x1 +#define RMI_GENERAL_STATUS__SKID_FIFO_0_UNDERFLOW_ERROR__SHIFT 0x2 +#define RMI_GENERAL_STATUS__SKID_FIFO_1_OVERFLOW_ERROR__SHIFT 0x3 +#define RMI_GENERAL_STATUS__SKID_FIFO_1_UNDERFLOW_ERROR__SHIFT 0x4 +#define RMI_GENERAL_STATUS__RMI_XBAR_BUSY__SHIFT 0x5 +#define RMI_GENERAL_STATUS__RMI_UTCL1_BUSY__SHIFT 0x6 +#define RMI_GENERAL_STATUS__RMI_SCOREBOARD_BUSY__SHIFT 0x7 +#define RMI_GENERAL_STATUS__TCIW0_PRT_FIFO_BUSY__SHIFT 0x8 +#define RMI_GENERAL_STATUS__TCIW_FRMTR0_BUSY__SHIFT 0x9 +#define RMI_GENERAL_STATUS__TCIW_RTN_FRMTR0_BUSY__SHIFT 0xa +#define RMI_GENERAL_STATUS__WRREQ_CONSUMER_FIFO_0_BUSY__SHIFT 0xb +#define RMI_GENERAL_STATUS__RDREQ_CONSUMER_FIFO_0_BUSY__SHIFT 0xc +#define RMI_GENERAL_STATUS__TCIW1_PRT_FIFO_BUSY__SHIFT 0xd +#define RMI_GENERAL_STATUS__TCIW_FRMTR1_BUSY__SHIFT 0xe +#define RMI_GENERAL_STATUS__TCIW_RTN_FRMTR1_BUSY__SHIFT 0xf +#define RMI_GENERAL_STATUS__WRREQ_CONSUMER_FIFO_1_BUSY__SHIFT 0x10 +#define RMI_GENERAL_STATUS__RDREQ_CONSUMER_FIFO_1_BUSY__SHIFT 0x11 +#define RMI_GENERAL_STATUS__UTC_PROBE1_BUSY__SHIFT 0x12 +#define RMI_GENERAL_STATUS__UTC_PROBE0_BUSY__SHIFT 0x13 +#define RMI_GENERAL_STATUS__RMI_XNACK_BUSY__SHIFT 0x14 +#define RMI_GENERAL_STATUS__XNACK_FIFO_NUM_USED__SHIFT 0x15 +#define RMI_GENERAL_STATUS__XNACK_FIFO_EMPTY__SHIFT 0x1d +#define RMI_GENERAL_STATUS__XNACK_FIFO_FULL__SHIFT 0x1e +#define RMI_GENERAL_STATUS__SKID_FIFO_FREESPACE_IS_ZERO_ERROR__SHIFT 0x1f +#define RMI_GENERAL_STATUS__GENERAL_RMI_ERRORS_COMBINED_MASK 0x00000001L +#define RMI_GENERAL_STATUS__SKID_FIFO_0_OVERFLOW_ERROR_MASK 0x00000002L +#define RMI_GENERAL_STATUS__SKID_FIFO_0_UNDERFLOW_ERROR_MASK 0x00000004L +#define RMI_GENERAL_STATUS__SKID_FIFO_1_OVERFLOW_ERROR_MASK 0x00000008L +#define RMI_GENERAL_STATUS__SKID_FIFO_1_UNDERFLOW_ERROR_MASK 0x00000010L +#define RMI_GENERAL_STATUS__RMI_XBAR_BUSY_MASK 0x00000020L +#define RMI_GENERAL_STATUS__RMI_UTCL1_BUSY_MASK 0x00000040L +#define RMI_GENERAL_STATUS__RMI_SCOREBOARD_BUSY_MASK 0x00000080L +#define RMI_GENERAL_STATUS__TCIW0_PRT_FIFO_BUSY_MASK 0x00000100L +#define RMI_GENERAL_STATUS__TCIW_FRMTR0_BUSY_MASK 0x00000200L +#define RMI_GENERAL_STATUS__TCIW_RTN_FRMTR0_BUSY_MASK 0x00000400L +#define RMI_GENERAL_STATUS__WRREQ_CONSUMER_FIFO_0_BUSY_MASK 0x00000800L +#define RMI_GENERAL_STATUS__RDREQ_CONSUMER_FIFO_0_BUSY_MASK 0x00001000L +#define RMI_GENERAL_STATUS__TCIW1_PRT_FIFO_BUSY_MASK 0x00002000L +#define RMI_GENERAL_STATUS__TCIW_FRMTR1_BUSY_MASK 0x00004000L +#define RMI_GENERAL_STATUS__TCIW_RTN_FRMTR1_BUSY_MASK 0x00008000L +#define RMI_GENERAL_STATUS__WRREQ_CONSUMER_FIFO_1_BUSY_MASK 0x00010000L +#define RMI_GENERAL_STATUS__RDREQ_CONSUMER_FIFO_1_BUSY_MASK 0x00020000L +#define RMI_GENERAL_STATUS__UTC_PROBE1_BUSY_MASK 0x00040000L +#define RMI_GENERAL_STATUS__UTC_PROBE0_BUSY_MASK 0x00080000L +#define RMI_GENERAL_STATUS__RMI_XNACK_BUSY_MASK 0x00100000L +#define RMI_GENERAL_STATUS__XNACK_FIFO_NUM_USED_MASK 0x1FE00000L +#define RMI_GENERAL_STATUS__XNACK_FIFO_EMPTY_MASK 0x20000000L +#define RMI_GENERAL_STATUS__XNACK_FIFO_FULL_MASK 0x40000000L +#define RMI_GENERAL_STATUS__SKID_FIFO_FREESPACE_IS_ZERO_ERROR_MASK 0x80000000L +//RMI_SUBBLOCK_STATUS0 +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_NUM_USED_PROBE0__SHIFT 0x0 +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_FULL_PROBE0__SHIFT 0x7 +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_EMPTY_PROBE0__SHIFT 0x8 +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_NUM_USED_PROBE1__SHIFT 0x9 +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_FULL_PROBE1__SHIFT 0x10 +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_EMPTY_PROBE1__SHIFT 0x11 +#define RMI_SUBBLOCK_STATUS0__TCIW0_INFLIGHT_CNT__SHIFT 0x12 +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_NUM_USED_PROBE0_MASK 0x0000007FL +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_FULL_PROBE0_MASK 0x00000080L +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_EMPTY_PROBE0_MASK 0x00000100L +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_NUM_USED_PROBE1_MASK 0x0000FE00L +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_FULL_PROBE1_MASK 0x00010000L +#define RMI_SUBBLOCK_STATUS0__UTC_EXT_LAT_HID_FIFO_EMPTY_PROBE1_MASK 0x00020000L +#define RMI_SUBBLOCK_STATUS0__TCIW0_INFLIGHT_CNT_MASK 0x0FFC0000L +//RMI_SUBBLOCK_STATUS1 +#define RMI_SUBBLOCK_STATUS1__SKID_FIFO_0_FREE_SPACE__SHIFT 0x0 +#define RMI_SUBBLOCK_STATUS1__SKID_FIFO_1_FREE_SPACE__SHIFT 0xa +#define RMI_SUBBLOCK_STATUS1__TCIW1_INFLIGHT_CNT__SHIFT 0x14 +#define RMI_SUBBLOCK_STATUS1__SKID_FIFO_0_FREE_SPACE_MASK 0x000003FFL +#define RMI_SUBBLOCK_STATUS1__SKID_FIFO_1_FREE_SPACE_MASK 0x000FFC00L +#define RMI_SUBBLOCK_STATUS1__TCIW1_INFLIGHT_CNT_MASK 0x3FF00000L +//RMI_SUBBLOCK_STATUS2 +#define RMI_SUBBLOCK_STATUS2__PRT_FIFO_0_NUM_USED__SHIFT 0x0 +#define RMI_SUBBLOCK_STATUS2__PRT_FIFO_1_NUM_USED__SHIFT 0x9 +#define RMI_SUBBLOCK_STATUS2__PRT_FIFO_0_NUM_USED_MASK 0x000001FFL +#define RMI_SUBBLOCK_STATUS2__PRT_FIFO_1_NUM_USED_MASK 0x0003FE00L +//RMI_SUBBLOCK_STATUS3 +#define RMI_SUBBLOCK_STATUS3__SKID_FIFO_0_FREE_SPACE_TOTAL__SHIFT 0x0 +#define RMI_SUBBLOCK_STATUS3__SKID_FIFO_1_FREE_SPACE_TOTAL__SHIFT 0xa +#define RMI_SUBBLOCK_STATUS3__SKID_FIFO_0_FREE_SPACE_TOTAL_MASK 0x000003FFL +#define RMI_SUBBLOCK_STATUS3__SKID_FIFO_1_FREE_SPACE_TOTAL_MASK 0x000FFC00L +//RMI_XBAR_CONFIG +#define RMI_XBAR_CONFIG__XBAR_MUX_CONFIG_OVERRIDE__SHIFT 0x0 +#define RMI_XBAR_CONFIG__XBAR_MUX_CONFIG_REQ_TYPE_OVERRIDE__SHIFT 0x2 +#define RMI_XBAR_CONFIG__XBAR_MUX_CONFIG_CB_DB_OVERRIDE__SHIFT 0x6 +#define RMI_XBAR_CONFIG__ARBITER_DIS__SHIFT 0x7 +#define RMI_XBAR_CONFIG__XBAR_EN_IN_REQ__SHIFT 0x8 +#define RMI_XBAR_CONFIG__XBAR_EN_IN_REQ_OVERRIDE__SHIFT 0xc +#define RMI_XBAR_CONFIG__XBAR_EN_IN_RB0__SHIFT 0xd +#define RMI_XBAR_CONFIG__XBAR_EN_IN_RB1__SHIFT 0xe +#define RMI_XBAR_CONFIG__XBAR_MUX_CONFIG_OVERRIDE_MASK 0x00000003L +#define RMI_XBAR_CONFIG__XBAR_MUX_CONFIG_REQ_TYPE_OVERRIDE_MASK 0x0000003CL +#define RMI_XBAR_CONFIG__XBAR_MUX_CONFIG_CB_DB_OVERRIDE_MASK 0x00000040L +#define RMI_XBAR_CONFIG__ARBITER_DIS_MASK 0x00000080L +#define RMI_XBAR_CONFIG__XBAR_EN_IN_REQ_MASK 0x00000F00L +#define RMI_XBAR_CONFIG__XBAR_EN_IN_REQ_OVERRIDE_MASK 0x00001000L +#define RMI_XBAR_CONFIG__XBAR_EN_IN_RB0_MASK 0x00002000L +#define RMI_XBAR_CONFIG__XBAR_EN_IN_RB1_MASK 0x00004000L +//RMI_PROBE_POP_LOGIC_CNTL +#define RMI_PROBE_POP_LOGIC_CNTL__EXT_LAT_FIFO_0_MAX_DEPTH__SHIFT 0x0 +#define RMI_PROBE_POP_LOGIC_CNTL__XLAT_COMBINE0_DIS__SHIFT 0x7 +#define RMI_PROBE_POP_LOGIC_CNTL__REDUCE_MAX_XLAT_CHAIN_SIZE_BY_2__SHIFT 0x8 +#define RMI_PROBE_POP_LOGIC_CNTL__EXT_LAT_FIFO_1_MAX_DEPTH__SHIFT 0xa +#define RMI_PROBE_POP_LOGIC_CNTL__XLAT_COMBINE1_DIS__SHIFT 0x11 +#define RMI_PROBE_POP_LOGIC_CNTL__EXT_LAT_FIFO_0_MAX_DEPTH_MASK 0x0000007FL +#define RMI_PROBE_POP_LOGIC_CNTL__XLAT_COMBINE0_DIS_MASK 0x00000080L +#define RMI_PROBE_POP_LOGIC_CNTL__REDUCE_MAX_XLAT_CHAIN_SIZE_BY_2_MASK 0x00000300L +#define RMI_PROBE_POP_LOGIC_CNTL__EXT_LAT_FIFO_1_MAX_DEPTH_MASK 0x0001FC00L +#define RMI_PROBE_POP_LOGIC_CNTL__XLAT_COMBINE1_DIS_MASK 0x00020000L +//RMI_UTC_XNACK_N_MISC_CNTL +#define RMI_UTC_XNACK_N_MISC_CNTL__MASTER_XNACK_TIMER_INC__SHIFT 0x0 +#define RMI_UTC_XNACK_N_MISC_CNTL__IND_XNACK_TIMER_START_VALUE__SHIFT 0x8 +#define RMI_UTC_XNACK_N_MISC_CNTL__UTCL1_PERM_MODE__SHIFT 0xc +#define RMI_UTC_XNACK_N_MISC_CNTL__CP_VMID_RESET_REQUEST_DISABLE__SHIFT 0xd +#define RMI_UTC_XNACK_N_MISC_CNTL__MASTER_XNACK_TIMER_INC_MASK 0x000000FFL +#define RMI_UTC_XNACK_N_MISC_CNTL__IND_XNACK_TIMER_START_VALUE_MASK 0x00000F00L +#define RMI_UTC_XNACK_N_MISC_CNTL__UTCL1_PERM_MODE_MASK 0x00001000L +#define RMI_UTC_XNACK_N_MISC_CNTL__CP_VMID_RESET_REQUEST_DISABLE_MASK 0x00002000L +//RMI_DEMUX_CNTL +#define RMI_DEMUX_CNTL__DEMUX_ARB0_STALL__SHIFT 0x0 +#define RMI_DEMUX_CNTL__DEMUX_ARB0_BREAK_LOB_ON_IDLEIN__SHIFT 0x1 +#define RMI_DEMUX_CNTL__DEMUX_ARB0_STALL_TIMER_OVERRIDE__SHIFT 0x4 +#define RMI_DEMUX_CNTL__DEMUX_ARB0_STALL_TIMER_START_VALUE__SHIFT 0x6 +#define RMI_DEMUX_CNTL__DEMUX_ARB0_MODE__SHIFT 0xe +#define RMI_DEMUX_CNTL__DEMUX_ARB1_STALL__SHIFT 0x10 +#define RMI_DEMUX_CNTL__DEMUX_ARB1_BREAK_LOB_ON_IDLEIN__SHIFT 0x11 +#define RMI_DEMUX_CNTL__DEMUX_ARB1_STALL_TIMER_OVERRIDE__SHIFT 0x14 +#define RMI_DEMUX_CNTL__DEMUX_ARB1_STALL_TIMER_START_VALUE__SHIFT 0x16 +#define RMI_DEMUX_CNTL__DEMUX_ARB1_MODE__SHIFT 0x1e +#define RMI_DEMUX_CNTL__DEMUX_ARB0_STALL_MASK 0x00000001L +#define RMI_DEMUX_CNTL__DEMUX_ARB0_BREAK_LOB_ON_IDLEIN_MASK 0x00000002L +#define RMI_DEMUX_CNTL__DEMUX_ARB0_STALL_TIMER_OVERRIDE_MASK 0x00000030L +#define RMI_DEMUX_CNTL__DEMUX_ARB0_STALL_TIMER_START_VALUE_MASK 0x00003FC0L +#define RMI_DEMUX_CNTL__DEMUX_ARB0_MODE_MASK 0x0000C000L +#define RMI_DEMUX_CNTL__DEMUX_ARB1_STALL_MASK 0x00010000L +#define RMI_DEMUX_CNTL__DEMUX_ARB1_BREAK_LOB_ON_IDLEIN_MASK 0x00020000L +#define RMI_DEMUX_CNTL__DEMUX_ARB1_STALL_TIMER_OVERRIDE_MASK 0x00300000L +#define RMI_DEMUX_CNTL__DEMUX_ARB1_STALL_TIMER_START_VALUE_MASK 0x3FC00000L +#define RMI_DEMUX_CNTL__DEMUX_ARB1_MODE_MASK 0xC0000000L +//RMI_UTCL1_CNTL1 +#define RMI_UTCL1_CNTL1__FORCE_4K_L2_RESP__SHIFT 0x0 +#define RMI_UTCL1_CNTL1__GPUVM_64K_DEF__SHIFT 0x1 +#define RMI_UTCL1_CNTL1__GPUVM_PERM_MODE__SHIFT 0x2 +#define RMI_UTCL1_CNTL1__RESP_MODE__SHIFT 0x3 +#define RMI_UTCL1_CNTL1__RESP_FAULT_MODE__SHIFT 0x5 +#define RMI_UTCL1_CNTL1__CLIENTID__SHIFT 0x7 +#define RMI_UTCL1_CNTL1__USERVM_DIS__SHIFT 0x10 +#define RMI_UTCL1_CNTL1__ENABLE_PUSH_LFIFO__SHIFT 0x11 +#define RMI_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB__SHIFT 0x12 +#define RMI_UTCL1_CNTL1__REG_INV_VMID__SHIFT 0x13 +#define RMI_UTCL1_CNTL1__REG_INV_ALL_VMID__SHIFT 0x17 +#define RMI_UTCL1_CNTL1__REG_INV_TOGGLE__SHIFT 0x18 +#define RMI_UTCL1_CNTL1__CLIENT_INVALIDATE_ALL_VMID__SHIFT 0x19 +#define RMI_UTCL1_CNTL1__FORCE_MISS__SHIFT 0x1a +#define RMI_UTCL1_CNTL1__FORCE_IN_ORDER__SHIFT 0x1b +#define RMI_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2__SHIFT 0x1c +#define RMI_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2__SHIFT 0x1e +#define RMI_UTCL1_CNTL1__FORCE_4K_L2_RESP_MASK 0x00000001L +#define RMI_UTCL1_CNTL1__GPUVM_64K_DEF_MASK 0x00000002L +#define RMI_UTCL1_CNTL1__GPUVM_PERM_MODE_MASK 0x00000004L +#define RMI_UTCL1_CNTL1__RESP_MODE_MASK 0x00000018L +#define RMI_UTCL1_CNTL1__RESP_FAULT_MODE_MASK 0x00000060L +#define RMI_UTCL1_CNTL1__CLIENTID_MASK 0x0000FF80L +#define RMI_UTCL1_CNTL1__USERVM_DIS_MASK 0x00010000L +#define RMI_UTCL1_CNTL1__ENABLE_PUSH_LFIFO_MASK 0x00020000L +#define RMI_UTCL1_CNTL1__ENABLE_LFIFO_PRI_ARB_MASK 0x00040000L +#define RMI_UTCL1_CNTL1__REG_INV_VMID_MASK 0x00780000L +#define RMI_UTCL1_CNTL1__REG_INV_ALL_VMID_MASK 0x00800000L +#define RMI_UTCL1_CNTL1__REG_INV_TOGGLE_MASK 0x01000000L +#define RMI_UTCL1_CNTL1__CLIENT_INVALIDATE_ALL_VMID_MASK 0x02000000L +#define RMI_UTCL1_CNTL1__FORCE_MISS_MASK 0x04000000L +#define RMI_UTCL1_CNTL1__FORCE_IN_ORDER_MASK 0x08000000L +#define RMI_UTCL1_CNTL1__REDUCE_FIFO_DEPTH_BY_2_MASK 0x30000000L +#define RMI_UTCL1_CNTL1__REDUCE_CACHE_SIZE_BY_2_MASK 0xC0000000L +//RMI_UTCL1_CNTL2 +#define RMI_UTCL1_CNTL2__UTC_SPARE__SHIFT 0x0 +#define RMI_UTCL1_CNTL2__MTYPE_OVRD_DIS__SHIFT 0x9 +#define RMI_UTCL1_CNTL2__LINE_VALID__SHIFT 0xa +#define RMI_UTCL1_CNTL2__DIS_EDC__SHIFT 0xb +#define RMI_UTCL1_CNTL2__GPUVM_INV_MODE__SHIFT 0xc +#define RMI_UTCL1_CNTL2__SHOOTDOWN_OPT__SHIFT 0xd +#define RMI_UTCL1_CNTL2__FORCE_SNOOP__SHIFT 0xe +#define RMI_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK__SHIFT 0xf +#define RMI_UTCL1_CNTL2__UTCL1_ARB_BURST_MODE__SHIFT 0x10 +#define RMI_UTCL1_CNTL2__UTCL1_ENABLE_PERF_EVENT_RD_WR__SHIFT 0x12 +#define RMI_UTCL1_CNTL2__UTCL1_PERF_EVENT_RD_WR__SHIFT 0x13 +#define RMI_UTCL1_CNTL2__UTCL1_ENABLE_PERF_EVENT_VMID__SHIFT 0x14 +#define RMI_UTCL1_CNTL2__UTCL1_PERF_EVENT_VMID__SHIFT 0x15 +#define RMI_UTCL1_CNTL2__UTCL1_DIS_DUAL_L2_REQ__SHIFT 0x19 +#define RMI_UTCL1_CNTL2__UTCL1_FORCE_FRAG_2M_TO_64K__SHIFT 0x1a +#define RMI_UTCL1_CNTL2__UTC_SPARE_MASK 0x000000FFL +#define RMI_UTCL1_CNTL2__MTYPE_OVRD_DIS_MASK 0x00000200L +#define RMI_UTCL1_CNTL2__LINE_VALID_MASK 0x00000400L +#define RMI_UTCL1_CNTL2__DIS_EDC_MASK 0x00000800L +#define RMI_UTCL1_CNTL2__GPUVM_INV_MODE_MASK 0x00001000L +#define RMI_UTCL1_CNTL2__SHOOTDOWN_OPT_MASK 0x00002000L +#define RMI_UTCL1_CNTL2__FORCE_SNOOP_MASK 0x00004000L +#define RMI_UTCL1_CNTL2__FORCE_GPUVM_INV_ACK_MASK 0x00008000L +#define RMI_UTCL1_CNTL2__UTCL1_ARB_BURST_MODE_MASK 0x00030000L +#define RMI_UTCL1_CNTL2__UTCL1_ENABLE_PERF_EVENT_RD_WR_MASK 0x00040000L +#define RMI_UTCL1_CNTL2__UTCL1_PERF_EVENT_RD_WR_MASK 0x00080000L +#define RMI_UTCL1_CNTL2__UTCL1_ENABLE_PERF_EVENT_VMID_MASK 0x00100000L +#define RMI_UTCL1_CNTL2__UTCL1_PERF_EVENT_VMID_MASK 0x01E00000L +#define RMI_UTCL1_CNTL2__UTCL1_DIS_DUAL_L2_REQ_MASK 0x02000000L +#define RMI_UTCL1_CNTL2__UTCL1_FORCE_FRAG_2M_TO_64K_MASK 0x04000000L +//RMI_UTC_UNIT_CONFIG +//RMI_TCIW_FORMATTER0_CNTL +#define RMI_TCIW_FORMATTER0_CNTL__WR_COMBINE0_DIS_OVERRIDE__SHIFT 0x0 +#define RMI_TCIW_FORMATTER0_CNTL__WR_COMBINE0_TIME_OUT_WINDOW__SHIFT 0x1 +#define RMI_TCIW_FORMATTER0_CNTL__TCIW0_MAX_ALLOWED_INFLIGHT_REQ__SHIFT 0x9 +#define RMI_TCIW_FORMATTER0_CNTL__SKID_FIFO_0_FREE_SPACE_DELTA__SHIFT 0x13 +#define RMI_TCIW_FORMATTER0_CNTL__SKID_FIFO_0_FREE_SPACE_DELTA_UPDATE__SHIFT 0x1b +#define RMI_TCIW_FORMATTER0_CNTL__TCIW0_REQ_SAFE_MODE__SHIFT 0x1c +#define RMI_TCIW_FORMATTER0_CNTL__RMI_IN0_REORDER_DIS__SHIFT 0x1d +#define RMI_TCIW_FORMATTER0_CNTL__WR_COMBINE0_DIS_AT_LAST_OF_BURST__SHIFT 0x1e +#define RMI_TCIW_FORMATTER0_CNTL__ALL_FAULT_RET0_DATA__SHIFT 0x1f +#define RMI_TCIW_FORMATTER0_CNTL__WR_COMBINE0_DIS_OVERRIDE_MASK 0x00000001L +#define RMI_TCIW_FORMATTER0_CNTL__WR_COMBINE0_TIME_OUT_WINDOW_MASK 0x000001FEL +#define RMI_TCIW_FORMATTER0_CNTL__TCIW0_MAX_ALLOWED_INFLIGHT_REQ_MASK 0x0007FE00L +#define RMI_TCIW_FORMATTER0_CNTL__SKID_FIFO_0_FREE_SPACE_DELTA_MASK 0x07F80000L +#define RMI_TCIW_FORMATTER0_CNTL__SKID_FIFO_0_FREE_SPACE_DELTA_UPDATE_MASK 0x08000000L +#define RMI_TCIW_FORMATTER0_CNTL__TCIW0_REQ_SAFE_MODE_MASK 0x10000000L +#define RMI_TCIW_FORMATTER0_CNTL__RMI_IN0_REORDER_DIS_MASK 0x20000000L +#define RMI_TCIW_FORMATTER0_CNTL__WR_COMBINE0_DIS_AT_LAST_OF_BURST_MASK 0x40000000L +#define RMI_TCIW_FORMATTER0_CNTL__ALL_FAULT_RET0_DATA_MASK 0x80000000L +//RMI_TCIW_FORMATTER1_CNTL +#define RMI_TCIW_FORMATTER1_CNTL__WR_COMBINE1_DIS_OVERRIDE__SHIFT 0x0 +#define RMI_TCIW_FORMATTER1_CNTL__WR_COMBINE1_TIME_OUT_WINDOW__SHIFT 0x1 +#define RMI_TCIW_FORMATTER1_CNTL__TCIW1_MAX_ALLOWED_INFLIGHT_REQ__SHIFT 0x9 +#define RMI_TCIW_FORMATTER1_CNTL__SKID_FIFO_1_FREE_SPACE_DELTA__SHIFT 0x13 +#define RMI_TCIW_FORMATTER1_CNTL__SKID_FIFO_1_FREE_SPACE_DELTA_UPDATE__SHIFT 0x1b +#define RMI_TCIW_FORMATTER1_CNTL__TCIW1_REQ_SAFE_MODE__SHIFT 0x1c +#define RMI_TCIW_FORMATTER1_CNTL__RMI_IN1_REORDER_DIS__SHIFT 0x1d +#define RMI_TCIW_FORMATTER1_CNTL__WR_COMBINE1_DIS_AT_LAST_OF_BURST__SHIFT 0x1e +#define RMI_TCIW_FORMATTER1_CNTL__ALL_FAULT_RET1_DATA__SHIFT 0x1f +#define RMI_TCIW_FORMATTER1_CNTL__WR_COMBINE1_DIS_OVERRIDE_MASK 0x00000001L +#define RMI_TCIW_FORMATTER1_CNTL__WR_COMBINE1_TIME_OUT_WINDOW_MASK 0x000001FEL +#define RMI_TCIW_FORMATTER1_CNTL__TCIW1_MAX_ALLOWED_INFLIGHT_REQ_MASK 0x0007FE00L +#define RMI_TCIW_FORMATTER1_CNTL__SKID_FIFO_1_FREE_SPACE_DELTA_MASK 0x07F80000L +#define RMI_TCIW_FORMATTER1_CNTL__SKID_FIFO_1_FREE_SPACE_DELTA_UPDATE_MASK 0x08000000L +#define RMI_TCIW_FORMATTER1_CNTL__TCIW1_REQ_SAFE_MODE_MASK 0x10000000L +#define RMI_TCIW_FORMATTER1_CNTL__RMI_IN1_REORDER_DIS_MASK 0x20000000L +#define RMI_TCIW_FORMATTER1_CNTL__WR_COMBINE1_DIS_AT_LAST_OF_BURST_MASK 0x40000000L +#define RMI_TCIW_FORMATTER1_CNTL__ALL_FAULT_RET1_DATA_MASK 0x80000000L +//RMI_SCOREBOARD_CNTL +#define RMI_SCOREBOARD_CNTL__COMPLETE_RB0_FLUSH__SHIFT 0x0 +#define RMI_SCOREBOARD_CNTL__REQ_IN_RE_EN_AFTER_FLUSH_RB0__SHIFT 0x1 +#define RMI_SCOREBOARD_CNTL__COMPLETE_RB1_FLUSH__SHIFT 0x2 +#define RMI_SCOREBOARD_CNTL__REQ_IN_RE_EN_AFTER_FLUSH_RB1__SHIFT 0x3 +#define RMI_SCOREBOARD_CNTL__TIME_STAMP_FLUSH_RB1__SHIFT 0x4 +#define RMI_SCOREBOARD_CNTL__VMID_INVAL_FLUSH_TYPE_OVERRIDE_EN__SHIFT 0x5 +#define RMI_SCOREBOARD_CNTL__VMID_INVAL_FLUSH_TYPE_OVERRIDE_VALUE__SHIFT 0x6 +#define RMI_SCOREBOARD_CNTL__TIME_STAMP_FLUSH_RB0__SHIFT 0x7 +#define RMI_SCOREBOARD_CNTL__FORCE_VMID_INVAL_DONE_EN__SHIFT 0x8 +#define RMI_SCOREBOARD_CNTL__FORCE_VMID_INVAL_DONE_TIMER_START_VALUE__SHIFT 0x9 +#define RMI_SCOREBOARD_CNTL__COMPLETE_RB0_FLUSH_MASK 0x00000001L +#define RMI_SCOREBOARD_CNTL__REQ_IN_RE_EN_AFTER_FLUSH_RB0_MASK 0x00000002L +#define RMI_SCOREBOARD_CNTL__COMPLETE_RB1_FLUSH_MASK 0x00000004L +#define RMI_SCOREBOARD_CNTL__REQ_IN_RE_EN_AFTER_FLUSH_RB1_MASK 0x00000008L +#define RMI_SCOREBOARD_CNTL__TIME_STAMP_FLUSH_RB1_MASK 0x00000010L +#define RMI_SCOREBOARD_CNTL__VMID_INVAL_FLUSH_TYPE_OVERRIDE_EN_MASK 0x00000020L +#define RMI_SCOREBOARD_CNTL__VMID_INVAL_FLUSH_TYPE_OVERRIDE_VALUE_MASK 0x00000040L +#define RMI_SCOREBOARD_CNTL__TIME_STAMP_FLUSH_RB0_MASK 0x00000080L +#define RMI_SCOREBOARD_CNTL__FORCE_VMID_INVAL_DONE_EN_MASK 0x00000100L +#define RMI_SCOREBOARD_CNTL__FORCE_VMID_INVAL_DONE_TIMER_START_VALUE_MASK 0x001FFE00L +//RMI_SCOREBOARD_STATUS0 +#define RMI_SCOREBOARD_STATUS0__CURRENT_SESSION_ID__SHIFT 0x0 +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_IN_PROG__SHIFT 0x1 +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_REQ_VMID__SHIFT 0x2 +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_UTC_DONE__SHIFT 0x12 +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_DONE__SHIFT 0x13 +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_FLUSH_TYPE__SHIFT 0x14 +#define RMI_SCOREBOARD_STATUS0__FORCE_VMID_INV_DONE__SHIFT 0x15 +#define RMI_SCOREBOARD_STATUS0__CURRENT_SESSION_ID_MASK 0x00000001L +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_IN_PROG_MASK 0x00000002L +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_REQ_VMID_MASK 0x0003FFFCL +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_UTC_DONE_MASK 0x00040000L +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_DONE_MASK 0x00080000L +#define RMI_SCOREBOARD_STATUS0__CP_VMID_INV_FLUSH_TYPE_MASK 0x00100000L +#define RMI_SCOREBOARD_STATUS0__FORCE_VMID_INV_DONE_MASK 0x00200000L +//RMI_SCOREBOARD_STATUS1 +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_RB0__SHIFT 0x0 +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_UNDERFLOW_RB0__SHIFT 0xc +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_OVERFLOW_RB0__SHIFT 0xd +#define RMI_SCOREBOARD_STATUS1__MULTI_VMID_INVAL_FROM_CP_DETECTED__SHIFT 0xe +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_RB1__SHIFT 0xf +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_UNDERFLOW_RB1__SHIFT 0x1b +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_OVERFLOW_RB1__SHIFT 0x1c +#define RMI_SCOREBOARD_STATUS1__COM_FLUSH_IN_PROG_RB1__SHIFT 0x1d +#define RMI_SCOREBOARD_STATUS1__COM_FLUSH_IN_PROG_RB0__SHIFT 0x1e +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_RB0_MASK 0x00000FFFL +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_UNDERFLOW_RB0_MASK 0x00001000L +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_OVERFLOW_RB0_MASK 0x00002000L +#define RMI_SCOREBOARD_STATUS1__MULTI_VMID_INVAL_FROM_CP_DETECTED_MASK 0x00004000L +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_RB1_MASK 0x07FF8000L +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_UNDERFLOW_RB1_MASK 0x08000000L +#define RMI_SCOREBOARD_STATUS1__RUNNING_CNT_OVERFLOW_RB1_MASK 0x10000000L +#define RMI_SCOREBOARD_STATUS1__COM_FLUSH_IN_PROG_RB1_MASK 0x20000000L +#define RMI_SCOREBOARD_STATUS1__COM_FLUSH_IN_PROG_RB0_MASK 0x40000000L +//RMI_SCOREBOARD_STATUS2 +#define RMI_SCOREBOARD_STATUS2__SNAPSHOT_CNT_RB0__SHIFT 0x0 +#define RMI_SCOREBOARD_STATUS2__SNAPSHOT_CNT_UNDERFLOW_RB0__SHIFT 0xc +#define RMI_SCOREBOARD_STATUS2__SNAPSHOT_CNT_RB1__SHIFT 0xd +#define RMI_SCOREBOARD_STATUS2__SNAPSHOT_CNT_UNDERFLOW_RB1__SHIFT 0x19 +#define RMI_SCOREBOARD_STATUS2__COM_FLUSH_DONE_RB1__SHIFT 0x1a +#define RMI_SCOREBOARD_STATUS2__COM_FLUSH_DONE_RB0__SHIFT 0x1b +#define RMI_SCOREBOARD_STATUS2__TIME_STAMP_FLUSH_IN_PROG_RB0__SHIFT 0x1c +#define RMI_SCOREBOARD_STATUS2__TIME_STAMP_FLUSH_IN_PROG_RB1__SHIFT 0x1d +#define RMI_SCOREBOARD_STATUS2__TIME_STAMP_FLUSH_DONE_RB0__SHIFT 0x1e +#define RMI_SCOREBOARD_STATUS2__TIME_STAMP_FLUSH_DONE_RB1__SHIFT 0x1f +#define RMI_SCOREBOARD_STATUS2__SNAPSHOT_CNT_RB0_MASK 0x00000FFFL +#define RMI_SCOREBOARD_STATUS2__SNAPSHOT_CNT_UNDERFLOW_RB0_MASK 0x00001000L +#define RMI_SCOREBOARD_STATUS2__SNAPSHOT_CNT_RB1_MASK 0x01FFE000L +#define RMI_SCOREBOARD_STATUS2__SNAPSHOT_CNT_UNDERFLOW_RB1_MASK 0x02000000L +#define RMI_SCOREBOARD_STATUS2__COM_FLUSH_DONE_RB1_MASK 0x04000000L +#define RMI_SCOREBOARD_STATUS2__COM_FLUSH_DONE_RB0_MASK 0x08000000L +#define RMI_SCOREBOARD_STATUS2__TIME_STAMP_FLUSH_IN_PROG_RB0_MASK 0x10000000L +#define RMI_SCOREBOARD_STATUS2__TIME_STAMP_FLUSH_IN_PROG_RB1_MASK 0x20000000L +#define RMI_SCOREBOARD_STATUS2__TIME_STAMP_FLUSH_DONE_RB0_MASK 0x40000000L +#define RMI_SCOREBOARD_STATUS2__TIME_STAMP_FLUSH_DONE_RB1_MASK 0x80000000L +//RMI_XBAR_ARBITER_CONFIG +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_MODE__SHIFT 0x0 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_BREAK_LOB_ON_WEIGHTEDRR__SHIFT 0x2 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_STALL__SHIFT 0x3 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_BREAK_LOB_ON_IDLEIN__SHIFT 0x4 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_STALL_TIMER_OVERRIDE__SHIFT 0x6 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_STALL_TIMER_START_VALUE__SHIFT 0x8 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_MODE__SHIFT 0x10 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_BREAK_LOB_ON_WEIGHTEDRR__SHIFT 0x12 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_STALL__SHIFT 0x13 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_BREAK_LOB_ON_IDLEIN__SHIFT 0x14 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_STALL_TIMER_OVERRIDE__SHIFT 0x16 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_STALL_TIMER_START_VALUE__SHIFT 0x18 +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_MODE_MASK 0x00000003L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_BREAK_LOB_ON_WEIGHTEDRR_MASK 0x00000004L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_STALL_MASK 0x00000008L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_BREAK_LOB_ON_IDLEIN_MASK 0x00000010L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_STALL_TIMER_OVERRIDE_MASK 0x000000C0L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB0_STALL_TIMER_START_VALUE_MASK 0x0000FF00L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_MODE_MASK 0x00030000L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_BREAK_LOB_ON_WEIGHTEDRR_MASK 0x00040000L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_STALL_MASK 0x00080000L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_BREAK_LOB_ON_IDLEIN_MASK 0x00100000L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_STALL_TIMER_OVERRIDE_MASK 0x00C00000L +#define RMI_XBAR_ARBITER_CONFIG__XBAR_ARB1_STALL_TIMER_START_VALUE_MASK 0xFF000000L +//RMI_XBAR_ARBITER_CONFIG_1 +#define RMI_XBAR_ARBITER_CONFIG_1__XBAR_ARB_ROUND_ROBIN_WEIGHT_RB0_RD__SHIFT 0x0 +#define RMI_XBAR_ARBITER_CONFIG_1__XBAR_ARB_ROUND_ROBIN_WEIGHT_RB0_WR__SHIFT 0x8 +#define RMI_XBAR_ARBITER_CONFIG_1__XBAR_ARB_ROUND_ROBIN_WEIGHT_RB1_RD__SHIFT 0x10 +#define RMI_XBAR_ARBITER_CONFIG_1__XBAR_ARB_ROUND_ROBIN_WEIGHT_RB1_WR__SHIFT 0x18 +#define RMI_XBAR_ARBITER_CONFIG_1__XBAR_ARB_ROUND_ROBIN_WEIGHT_RB0_RD_MASK 0x000000FFL +#define RMI_XBAR_ARBITER_CONFIG_1__XBAR_ARB_ROUND_ROBIN_WEIGHT_RB0_WR_MASK 0x0000FF00L +#define RMI_XBAR_ARBITER_CONFIG_1__XBAR_ARB_ROUND_ROBIN_WEIGHT_RB1_RD_MASK 0x00FF0000L +#define RMI_XBAR_ARBITER_CONFIG_1__XBAR_ARB_ROUND_ROBIN_WEIGHT_RB1_WR_MASK 0xFF000000L +//RMI_CLOCK_CNTRL +#define RMI_CLOCK_CNTRL__DYN_CLK_RB0_BUSY_MASK__SHIFT 0x0 +#define RMI_CLOCK_CNTRL__DYN_CLK_CMN_BUSY_MASK__SHIFT 0x5 +#define RMI_CLOCK_CNTRL__DYN_CLK_RB0_WAKEUP_MASK__SHIFT 0xa +#define RMI_CLOCK_CNTRL__DYN_CLK_CMN_WAKEUP_MASK__SHIFT 0xf +#define RMI_CLOCK_CNTRL__DYN_CLK_RB1_BUSY_MASK__SHIFT 0x14 +#define RMI_CLOCK_CNTRL__DYN_CLK_RB1_WAKEUP_MASK__SHIFT 0x19 +#define RMI_CLOCK_CNTRL__DYN_CLK_RB0_BUSY_MASK_MASK 0x0000001FL +#define RMI_CLOCK_CNTRL__DYN_CLK_CMN_BUSY_MASK_MASK 0x000003E0L +#define RMI_CLOCK_CNTRL__DYN_CLK_RB0_WAKEUP_MASK_MASK 0x00007C00L +#define RMI_CLOCK_CNTRL__DYN_CLK_CMN_WAKEUP_MASK_MASK 0x000F8000L +#define RMI_CLOCK_CNTRL__DYN_CLK_RB1_BUSY_MASK_MASK 0x01F00000L +#define RMI_CLOCK_CNTRL__DYN_CLK_RB1_WAKEUP_MASK_MASK 0x3E000000L +//RMI_UTCL1_STATUS +#define RMI_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define RMI_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define RMI_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define RMI_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define RMI_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define RMI_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L +//RMI_SPARE +#define RMI_SPARE__RMI_ARBITER_STALL_TIMER_ENABLED_ALLOW_STREAMING__SHIFT 0x0 +#define RMI_SPARE__SPARE_BIT_1__SHIFT 0x1 +#define RMI_SPARE__SPARE_BIT_2__SHIFT 0x2 +#define RMI_SPARE__SPARE_BIT_3__SHIFT 0x3 +#define RMI_SPARE__SPARE_BIT_4__SHIFT 0x4 +#define RMI_SPARE__SPARE_BIT_5__SHIFT 0x5 +#define RMI_SPARE__SPARE_BIT_6__SHIFT 0x6 +#define RMI_SPARE__SPARE_BIT_7__SHIFT 0x7 +#define RMI_SPARE__SPARE_BIT_8_0__SHIFT 0x8 +#define RMI_SPARE__SPARE_BIT_16_0__SHIFT 0x10 +#define RMI_SPARE__RMI_ARBITER_STALL_TIMER_ENABLED_ALLOW_STREAMING_MASK 0x00000001L +#define RMI_SPARE__SPARE_BIT_1_MASK 0x00000002L +#define RMI_SPARE__SPARE_BIT_2_MASK 0x00000004L +#define RMI_SPARE__SPARE_BIT_3_MASK 0x00000008L +#define RMI_SPARE__SPARE_BIT_4_MASK 0x00000010L +#define RMI_SPARE__SPARE_BIT_5_MASK 0x00000020L +#define RMI_SPARE__SPARE_BIT_6_MASK 0x00000040L +#define RMI_SPARE__SPARE_BIT_7_MASK 0x00000080L +#define RMI_SPARE__SPARE_BIT_8_0_MASK 0x0000FF00L +#define RMI_SPARE__SPARE_BIT_16_0_MASK 0xFFFF0000L +//RMI_SPARE_1 +#define RMI_SPARE_1__SPARE_BIT_8__SHIFT 0x0 +#define RMI_SPARE_1__SPARE_BIT_9__SHIFT 0x1 +#define RMI_SPARE_1__SPARE_BIT_10__SHIFT 0x2 +#define RMI_SPARE_1__SPARE_BIT_11__SHIFT 0x3 +#define RMI_SPARE_1__SPARE_BIT_12__SHIFT 0x4 +#define RMI_SPARE_1__SPARE_BIT_13__SHIFT 0x5 +#define RMI_SPARE_1__SPARE_BIT_14__SHIFT 0x6 +#define RMI_SPARE_1__SPARE_BIT_15__SHIFT 0x7 +#define RMI_SPARE_1__SPARE_BIT_8_1__SHIFT 0x8 +#define RMI_SPARE_1__SPARE_BIT_16_1__SHIFT 0x10 +#define RMI_SPARE_1__SPARE_BIT_8_MASK 0x00000001L +#define RMI_SPARE_1__SPARE_BIT_9_MASK 0x00000002L +#define RMI_SPARE_1__SPARE_BIT_10_MASK 0x00000004L +#define RMI_SPARE_1__SPARE_BIT_11_MASK 0x00000008L +#define RMI_SPARE_1__SPARE_BIT_12_MASK 0x00000010L +#define RMI_SPARE_1__SPARE_BIT_13_MASK 0x00000020L +#define RMI_SPARE_1__SPARE_BIT_14_MASK 0x00000040L +#define RMI_SPARE_1__SPARE_BIT_15_MASK 0x00000080L +#define RMI_SPARE_1__SPARE_BIT_8_1_MASK 0x0000FF00L +#define RMI_SPARE_1__SPARE_BIT_16_1_MASK 0xFFFF0000L +//RMI_SPARE_2 +#define RMI_SPARE_2__SPARE_BIT_16__SHIFT 0x0 +#define RMI_SPARE_2__SPARE_BIT_17__SHIFT 0x1 +#define RMI_SPARE_2__SPARE_BIT_18__SHIFT 0x2 +#define RMI_SPARE_2__SPARE_BIT_19__SHIFT 0x3 +#define RMI_SPARE_2__SPARE_BIT_20__SHIFT 0x4 +#define RMI_SPARE_2__SPARE_BIT_21__SHIFT 0x5 +#define RMI_SPARE_2__SPARE_BIT_22__SHIFT 0x6 +#define RMI_SPARE_2__SPARE_BIT_23__SHIFT 0x7 +#define RMI_SPARE_2__SPARE_BIT_4_0__SHIFT 0x8 +#define RMI_SPARE_2__SPARE_BIT_4_1__SHIFT 0xc +#define RMI_SPARE_2__SPARE_BIT_8_2__SHIFT 0x10 +#define RMI_SPARE_2__SPARE_BIT_8_3__SHIFT 0x18 +#define RMI_SPARE_2__SPARE_BIT_16_MASK 0x00000001L +#define RMI_SPARE_2__SPARE_BIT_17_MASK 0x00000002L +#define RMI_SPARE_2__SPARE_BIT_18_MASK 0x00000004L +#define RMI_SPARE_2__SPARE_BIT_19_MASK 0x00000008L +#define RMI_SPARE_2__SPARE_BIT_20_MASK 0x00000010L +#define RMI_SPARE_2__SPARE_BIT_21_MASK 0x00000020L +#define RMI_SPARE_2__SPARE_BIT_22_MASK 0x00000040L +#define RMI_SPARE_2__SPARE_BIT_23_MASK 0x00000080L +#define RMI_SPARE_2__SPARE_BIT_4_0_MASK 0x00000F00L +#define RMI_SPARE_2__SPARE_BIT_4_1_MASK 0x0000F000L +#define RMI_SPARE_2__SPARE_BIT_8_2_MASK 0x00FF0000L +#define RMI_SPARE_2__SPARE_BIT_8_3_MASK 0xFF000000L + + +// addressBlock: gc_dbgu_gfx_dbgudec +//port_a_addr +#define port_a_addr__Index__SHIFT 0x0 +#define port_a_addr__Reserved__SHIFT 0x8 +#define port_a_addr__ReadEnable__SHIFT 0x1f +#define port_a_addr__Index_MASK 0x000000FFL +#define port_a_addr__Reserved_MASK 0x7FFFFF00L +#define port_a_addr__ReadEnable_MASK 0x80000000L +//port_a_data_lo +#define port_a_data_lo__Data__SHIFT 0x0 +#define port_a_data_lo__Data_MASK 0xFFFFFFFFL +//port_a_data_hi +#define port_a_data_hi__Data__SHIFT 0x0 +#define port_a_data_hi__Data_MASK 0xFFFFFFFFL +//port_b_addr +#define port_b_addr__Index__SHIFT 0x0 +#define port_b_addr__Reserved__SHIFT 0x8 +#define port_b_addr__ReadEnable__SHIFT 0x1f +#define port_b_addr__Index_MASK 0x000000FFL +#define port_b_addr__Reserved_MASK 0x7FFFFF00L +#define port_b_addr__ReadEnable_MASK 0x80000000L +//port_b_data_lo +#define port_b_data_lo__Data__SHIFT 0x0 +#define port_b_data_lo__Data_MASK 0xFFFFFFFFL +//port_b_data_hi +#define port_b_data_hi__Data__SHIFT 0x0 +#define port_b_data_hi__Data_MASK 0xFFFFFFFFL +//port_c_addr +#define port_c_addr__Index__SHIFT 0x0 +#define port_c_addr__Reserved__SHIFT 0x8 +#define port_c_addr__ReadEnable__SHIFT 0x1f +#define port_c_addr__Index_MASK 0x000000FFL +#define port_c_addr__Reserved_MASK 0x7FFFFF00L +#define port_c_addr__ReadEnable_MASK 0x80000000L +//port_c_data_lo +#define port_c_data_lo__Data__SHIFT 0x0 +#define port_c_data_lo__Data_MASK 0xFFFFFFFFL +//port_c_data_hi +#define port_c_data_hi__Data__SHIFT 0x0 +#define port_c_data_hi__Data_MASK 0xFFFFFFFFL +//port_d_addr +#define port_d_addr__Index__SHIFT 0x0 +#define port_d_addr__Reserved__SHIFT 0x8 +#define port_d_addr__ReadEnable__SHIFT 0x1f +#define port_d_addr__Index_MASK 0x000000FFL +#define port_d_addr__Reserved_MASK 0x7FFFFF00L +#define port_d_addr__ReadEnable_MASK 0x80000000L +//port_d_data_lo +#define port_d_data_lo__Data__SHIFT 0x0 +#define port_d_data_lo__Data_MASK 0xFFFFFFFFL +//port_d_data_hi +#define port_d_data_hi__Data__SHIFT 0x0 +#define port_d_data_hi__Data_MASK 0xFFFFFFFFL + + +// addressBlock: gc_utcl2_atcl2dec +//ATC_L2_CNTL +#define ATC_L2_CNTL__NUMBER_OF_TRANSLATION_READ_REQUESTS__SHIFT 0x0 +#define ATC_L2_CNTL__NUMBER_OF_TRANSLATION_WRITE_REQUESTS__SHIFT 0x3 +#define ATC_L2_CNTL__NUMBER_OF_TRANSLATION_READS_DEPENDS_ON_ADDR_MOD__SHIFT 0x6 +#define ATC_L2_CNTL__NUMBER_OF_TRANSLATION_WRITES_DEPENDS_ON_ADDR_MOD__SHIFT 0x7 +#define ATC_L2_CNTL__CACHE_INVALIDATE_MODE__SHIFT 0x8 +#define ATC_L2_CNTL__ENABLE_DEFAULT_PAGE_OUT_TO_SYSTEM_MEMORY__SHIFT 0xb +#define ATC_L2_CNTL__NUMBER_OF_TRANSLATION_READ_REQUESTS_MASK 0x00000003L +#define ATC_L2_CNTL__NUMBER_OF_TRANSLATION_WRITE_REQUESTS_MASK 0x00000018L +#define ATC_L2_CNTL__NUMBER_OF_TRANSLATION_READS_DEPENDS_ON_ADDR_MOD_MASK 0x00000040L +#define ATC_L2_CNTL__NUMBER_OF_TRANSLATION_WRITES_DEPENDS_ON_ADDR_MOD_MASK 0x00000080L +#define ATC_L2_CNTL__CACHE_INVALIDATE_MODE_MASK 0x00000700L +#define ATC_L2_CNTL__ENABLE_DEFAULT_PAGE_OUT_TO_SYSTEM_MEMORY_MASK 0x00000800L +//ATC_L2_CNTL2 +#define ATC_L2_CNTL2__BANK_SELECT__SHIFT 0x0 +#define ATC_L2_CNTL2__L2_CACHE_UPDATE_MODE__SHIFT 0x6 +#define ATC_L2_CNTL2__ENABLE_L2_CACHE_LRU_UPDATE_BY_WRITE__SHIFT 0x8 +#define ATC_L2_CNTL2__L2_CACHE_SWAP_TAG_INDEX_LSBS__SHIFT 0x9 +#define ATC_L2_CNTL2__L2_CACHE_VMID_MODE__SHIFT 0xc +#define ATC_L2_CNTL2__L2_CACHE_UPDATE_WILDCARD_REFERENCE_VALUE__SHIFT 0xf +#define ATC_L2_CNTL2__BANK_SELECT_MASK 0x0000003FL +#define ATC_L2_CNTL2__L2_CACHE_UPDATE_MODE_MASK 0x000000C0L +#define ATC_L2_CNTL2__ENABLE_L2_CACHE_LRU_UPDATE_BY_WRITE_MASK 0x00000100L +#define ATC_L2_CNTL2__L2_CACHE_SWAP_TAG_INDEX_LSBS_MASK 0x00000E00L +#define ATC_L2_CNTL2__L2_CACHE_VMID_MODE_MASK 0x00007000L +#define ATC_L2_CNTL2__L2_CACHE_UPDATE_WILDCARD_REFERENCE_VALUE_MASK 0x001F8000L +//ATC_L2_CACHE_DATA0 +#define ATC_L2_CACHE_DATA0__DATA_REGISTER_VALID__SHIFT 0x0 +#define ATC_L2_CACHE_DATA0__CACHE_ENTRY_VALID__SHIFT 0x1 +#define ATC_L2_CACHE_DATA0__CACHED_ATTRIBUTES__SHIFT 0x2 +#define ATC_L2_CACHE_DATA0__VIRTUAL_PAGE_ADDRESS_HIGH__SHIFT 0x17 +#define ATC_L2_CACHE_DATA0__DATA_REGISTER_VALID_MASK 0x00000001L +#define ATC_L2_CACHE_DATA0__CACHE_ENTRY_VALID_MASK 0x00000002L +#define ATC_L2_CACHE_DATA0__CACHED_ATTRIBUTES_MASK 0x007FFFFCL +#define ATC_L2_CACHE_DATA0__VIRTUAL_PAGE_ADDRESS_HIGH_MASK 0x07800000L +//ATC_L2_CACHE_DATA1 +#define ATC_L2_CACHE_DATA1__VIRTUAL_PAGE_ADDRESS_LOW__SHIFT 0x0 +#define ATC_L2_CACHE_DATA1__VIRTUAL_PAGE_ADDRESS_LOW_MASK 0xFFFFFFFFL +//ATC_L2_CACHE_DATA2 +#define ATC_L2_CACHE_DATA2__PHYSICAL_PAGE_ADDRESS__SHIFT 0x0 +#define ATC_L2_CACHE_DATA2__PHYSICAL_PAGE_ADDRESS_MASK 0xFFFFFFFFL +//ATC_L2_CNTL3 +#define ATC_L2_CNTL3__DELAY_SEND_INVALIDATION_REQUEST__SHIFT 0x0 +#define ATC_L2_CNTL3__ATS_REQUEST_CREDIT_MINUS1__SHIFT 0x3 +#define ATC_L2_CNTL3__DELAY_SEND_INVALIDATION_REQUEST_MASK 0x00000007L +#define ATC_L2_CNTL3__ATS_REQUEST_CREDIT_MINUS1_MASK 0x000001F8L +//ATC_L2_STATUS +#define ATC_L2_STATUS__BUSY__SHIFT 0x0 +#define ATC_L2_STATUS__PARITY_ERROR_INFO__SHIFT 0x1 +#define ATC_L2_STATUS__BUSY_MASK 0x00000001L +#define ATC_L2_STATUS__PARITY_ERROR_INFO_MASK 0x3FFFFFFEL +//ATC_L2_STATUS2 +#define ATC_L2_STATUS2__IFIFO_NON_FATAL_PARITY_ERROR_INFO__SHIFT 0x0 +#define ATC_L2_STATUS2__IFIFO_FATAL_PARITY_ERROR_INFO__SHIFT 0x8 +#define ATC_L2_STATUS2__IFIFO_NON_FATAL_PARITY_ERROR_INFO_MASK 0x000000FFL +#define ATC_L2_STATUS2__IFIFO_FATAL_PARITY_ERROR_INFO_MASK 0x0000FF00L +//ATC_L2_MISC_CG +#define ATC_L2_MISC_CG__OFFDLY__SHIFT 0x6 +#define ATC_L2_MISC_CG__ENABLE__SHIFT 0x12 +#define ATC_L2_MISC_CG__MEM_LS_ENABLE__SHIFT 0x13 +#define ATC_L2_MISC_CG__OFFDLY_MASK 0x00000FC0L +#define ATC_L2_MISC_CG__ENABLE_MASK 0x00040000L +#define ATC_L2_MISC_CG__MEM_LS_ENABLE_MASK 0x00080000L +//ATC_L2_MEM_POWER_LS +#define ATC_L2_MEM_POWER_LS__LS_SETUP__SHIFT 0x0 +#define ATC_L2_MEM_POWER_LS__LS_HOLD__SHIFT 0x6 +#define ATC_L2_MEM_POWER_LS__LS_SETUP_MASK 0x0000003FL +#define ATC_L2_MEM_POWER_LS__LS_HOLD_MASK 0x00000FC0L +//ATC_L2_CGTT_CLK_CTRL +#define ATC_L2_CGTT_CLK_CTRL__ON_DELAY__SHIFT 0x0 +#define ATC_L2_CGTT_CLK_CTRL__OFF_HYSTERESIS__SHIFT 0x4 +#define ATC_L2_CGTT_CLK_CTRL__MGLS_OVERRIDE__SHIFT 0xf +#define ATC_L2_CGTT_CLK_CTRL__SOFT_STALL_OVERRIDE__SHIFT 0x10 +#define ATC_L2_CGTT_CLK_CTRL__SOFT_OVERRIDE__SHIFT 0x18 +#define ATC_L2_CGTT_CLK_CTRL__ON_DELAY_MASK 0x0000000FL +#define ATC_L2_CGTT_CLK_CTRL__OFF_HYSTERESIS_MASK 0x00000FF0L +#define ATC_L2_CGTT_CLK_CTRL__MGLS_OVERRIDE_MASK 0x00008000L +#define ATC_L2_CGTT_CLK_CTRL__SOFT_STALL_OVERRIDE_MASK 0x00FF0000L +#define ATC_L2_CGTT_CLK_CTRL__SOFT_OVERRIDE_MASK 0xFF000000L + + +// addressBlock: gc_utcl2_vml2pfdec +//VM_L2_CNTL +#define VM_L2_CNTL__ENABLE_L2_CACHE__SHIFT 0x0 +#define VM_L2_CNTL__ENABLE_L2_FRAGMENT_PROCESSING__SHIFT 0x1 +#define VM_L2_CNTL__L2_CACHE_PTE_ENDIAN_SWAP_MODE__SHIFT 0x2 +#define VM_L2_CNTL__L2_CACHE_PDE_ENDIAN_SWAP_MODE__SHIFT 0x4 +#define VM_L2_CNTL__L2_PDE0_CACHE_TAG_GENERATION_MODE__SHIFT 0x8 +#define VM_L2_CNTL__ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE__SHIFT 0x9 +#define VM_L2_CNTL__ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE__SHIFT 0xa +#define VM_L2_CNTL__ENABLE_DEFAULT_PAGE_OUT_TO_SYSTEM_MEMORY__SHIFT 0xb +#define VM_L2_CNTL__L2_PDE0_CACHE_SPLIT_MODE__SHIFT 0xc +#define VM_L2_CNTL__EFFECTIVE_L2_QUEUE_SIZE__SHIFT 0xf +#define VM_L2_CNTL__PDE_FAULT_CLASSIFICATION__SHIFT 0x12 +#define VM_L2_CNTL__CONTEXT1_IDENTITY_ACCESS_MODE__SHIFT 0x13 +#define VM_L2_CNTL__IDENTITY_MODE_FRAGMENT_SIZE__SHIFT 0x15 +#define VM_L2_CNTL__L2_PTE_CACHE_ADDR_MODE__SHIFT 0x1a +#define VM_L2_CNTL__ENABLE_L2_CACHE_MASK 0x00000001L +#define VM_L2_CNTL__ENABLE_L2_FRAGMENT_PROCESSING_MASK 0x00000002L +#define VM_L2_CNTL__L2_CACHE_PTE_ENDIAN_SWAP_MODE_MASK 0x0000000CL +#define VM_L2_CNTL__L2_CACHE_PDE_ENDIAN_SWAP_MODE_MASK 0x00000030L +#define VM_L2_CNTL__L2_PDE0_CACHE_TAG_GENERATION_MODE_MASK 0x00000100L +#define VM_L2_CNTL__ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE_MASK 0x00000200L +#define VM_L2_CNTL__ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE_MASK 0x00000400L +#define VM_L2_CNTL__ENABLE_DEFAULT_PAGE_OUT_TO_SYSTEM_MEMORY_MASK 0x00000800L +#define VM_L2_CNTL__L2_PDE0_CACHE_SPLIT_MODE_MASK 0x00007000L +#define VM_L2_CNTL__EFFECTIVE_L2_QUEUE_SIZE_MASK 0x00038000L +#define VM_L2_CNTL__PDE_FAULT_CLASSIFICATION_MASK 0x00040000L +#define VM_L2_CNTL__CONTEXT1_IDENTITY_ACCESS_MODE_MASK 0x00180000L +#define VM_L2_CNTL__IDENTITY_MODE_FRAGMENT_SIZE_MASK 0x03E00000L +#define VM_L2_CNTL__L2_PTE_CACHE_ADDR_MODE_MASK 0x0C000000L +//VM_L2_CNTL2 +#define VM_L2_CNTL2__INVALIDATE_ALL_L1_TLBS__SHIFT 0x0 +#define VM_L2_CNTL2__INVALIDATE_L2_CACHE__SHIFT 0x1 +#define VM_L2_CNTL2__DISABLE_INVALIDATE_PER_DOMAIN__SHIFT 0x15 +#define VM_L2_CNTL2__DISABLE_BIGK_CACHE_OPTIMIZATION__SHIFT 0x16 +#define VM_L2_CNTL2__L2_PTE_CACHE_VMID_MODE__SHIFT 0x17 +#define VM_L2_CNTL2__INVALIDATE_CACHE_MODE__SHIFT 0x1a +#define VM_L2_CNTL2__PDE_CACHE_EFFECTIVE_SIZE__SHIFT 0x1c +#define VM_L2_CNTL2__INVALIDATE_ALL_L1_TLBS_MASK 0x00000001L +#define VM_L2_CNTL2__INVALIDATE_L2_CACHE_MASK 0x00000002L +#define VM_L2_CNTL2__DISABLE_INVALIDATE_PER_DOMAIN_MASK 0x00200000L +#define VM_L2_CNTL2__DISABLE_BIGK_CACHE_OPTIMIZATION_MASK 0x00400000L +#define VM_L2_CNTL2__L2_PTE_CACHE_VMID_MODE_MASK 0x03800000L +#define VM_L2_CNTL2__INVALIDATE_CACHE_MODE_MASK 0x0C000000L +#define VM_L2_CNTL2__PDE_CACHE_EFFECTIVE_SIZE_MASK 0x70000000L +//VM_L2_CNTL3 +#define VM_L2_CNTL3__BANK_SELECT__SHIFT 0x0 +#define VM_L2_CNTL3__L2_CACHE_UPDATE_MODE__SHIFT 0x6 +#define VM_L2_CNTL3__L2_CACHE_UPDATE_WILDCARD_REFERENCE_VALUE__SHIFT 0x8 +#define VM_L2_CNTL3__L2_CACHE_BIGK_FRAGMENT_SIZE__SHIFT 0xf +#define VM_L2_CNTL3__L2_CACHE_BIGK_ASSOCIATIVITY__SHIFT 0x14 +#define VM_L2_CNTL3__L2_CACHE_4K_EFFECTIVE_SIZE__SHIFT 0x15 +#define VM_L2_CNTL3__L2_CACHE_BIGK_EFFECTIVE_SIZE__SHIFT 0x18 +#define VM_L2_CNTL3__L2_CACHE_4K_FORCE_MISS__SHIFT 0x1c +#define VM_L2_CNTL3__L2_CACHE_BIGK_FORCE_MISS__SHIFT 0x1d +#define VM_L2_CNTL3__PDE_CACHE_FORCE_MISS__SHIFT 0x1e +#define VM_L2_CNTL3__L2_CACHE_4K_ASSOCIATIVITY__SHIFT 0x1f +#define VM_L2_CNTL3__BANK_SELECT_MASK 0x0000003FL +#define VM_L2_CNTL3__L2_CACHE_UPDATE_MODE_MASK 0x000000C0L +#define VM_L2_CNTL3__L2_CACHE_UPDATE_WILDCARD_REFERENCE_VALUE_MASK 0x00001F00L +#define VM_L2_CNTL3__L2_CACHE_BIGK_FRAGMENT_SIZE_MASK 0x000F8000L +#define VM_L2_CNTL3__L2_CACHE_BIGK_ASSOCIATIVITY_MASK 0x00100000L +#define VM_L2_CNTL3__L2_CACHE_4K_EFFECTIVE_SIZE_MASK 0x00E00000L +#define VM_L2_CNTL3__L2_CACHE_BIGK_EFFECTIVE_SIZE_MASK 0x0F000000L +#define VM_L2_CNTL3__L2_CACHE_4K_FORCE_MISS_MASK 0x10000000L +#define VM_L2_CNTL3__L2_CACHE_BIGK_FORCE_MISS_MASK 0x20000000L +#define VM_L2_CNTL3__PDE_CACHE_FORCE_MISS_MASK 0x40000000L +#define VM_L2_CNTL3__L2_CACHE_4K_ASSOCIATIVITY_MASK 0x80000000L +//VM_L2_STATUS +#define VM_L2_STATUS__L2_BUSY__SHIFT 0x0 +#define VM_L2_STATUS__CONTEXT_DOMAIN_BUSY__SHIFT 0x1 +#define VM_L2_STATUS__FOUND_4K_PTE_CACHE_PARITY_ERRORS__SHIFT 0x11 +#define VM_L2_STATUS__FOUND_BIGK_PTE_CACHE_PARITY_ERRORS__SHIFT 0x12 +#define VM_L2_STATUS__FOUND_PDE0_CACHE_PARITY_ERRORS__SHIFT 0x13 +#define VM_L2_STATUS__FOUND_PDE1_CACHE_PARITY_ERRORS__SHIFT 0x14 +#define VM_L2_STATUS__FOUND_PDE2_CACHE_PARITY_ERRORS__SHIFT 0x15 +#define VM_L2_STATUS__L2_BUSY_MASK 0x00000001L +#define VM_L2_STATUS__CONTEXT_DOMAIN_BUSY_MASK 0x0001FFFEL +#define VM_L2_STATUS__FOUND_4K_PTE_CACHE_PARITY_ERRORS_MASK 0x00020000L +#define VM_L2_STATUS__FOUND_BIGK_PTE_CACHE_PARITY_ERRORS_MASK 0x00040000L +#define VM_L2_STATUS__FOUND_PDE0_CACHE_PARITY_ERRORS_MASK 0x00080000L +#define VM_L2_STATUS__FOUND_PDE1_CACHE_PARITY_ERRORS_MASK 0x00100000L +#define VM_L2_STATUS__FOUND_PDE2_CACHE_PARITY_ERRORS_MASK 0x00200000L +//VM_DUMMY_PAGE_FAULT_CNTL +#define VM_DUMMY_PAGE_FAULT_CNTL__DUMMY_PAGE_FAULT_ENABLE__SHIFT 0x0 +#define VM_DUMMY_PAGE_FAULT_CNTL__DUMMY_PAGE_ADDRESS_LOGICAL__SHIFT 0x1 +#define VM_DUMMY_PAGE_FAULT_CNTL__DUMMY_PAGE_COMPARE_MSBS__SHIFT 0x2 +#define VM_DUMMY_PAGE_FAULT_CNTL__DUMMY_PAGE_FAULT_ENABLE_MASK 0x00000001L +#define VM_DUMMY_PAGE_FAULT_CNTL__DUMMY_PAGE_ADDRESS_LOGICAL_MASK 0x00000002L +#define VM_DUMMY_PAGE_FAULT_CNTL__DUMMY_PAGE_COMPARE_MSBS_MASK 0x000000FCL +//VM_DUMMY_PAGE_FAULT_ADDR_LO32 +#define VM_DUMMY_PAGE_FAULT_ADDR_LO32__DUMMY_PAGE_ADDR_LO32__SHIFT 0x0 +#define VM_DUMMY_PAGE_FAULT_ADDR_LO32__DUMMY_PAGE_ADDR_LO32_MASK 0xFFFFFFFFL +//VM_DUMMY_PAGE_FAULT_ADDR_HI32 +#define VM_DUMMY_PAGE_FAULT_ADDR_HI32__DUMMY_PAGE_ADDR_HI4__SHIFT 0x0 +#define VM_DUMMY_PAGE_FAULT_ADDR_HI32__DUMMY_PAGE_ADDR_HI4_MASK 0x0000000FL +//VM_L2_PROTECTION_FAULT_CNTL +#define VM_L2_PROTECTION_FAULT_CNTL__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_CNTL__ALLOW_SUBSEQUENT_PROTECTION_FAULT_STATUS_ADDR_UPDATES__SHIFT 0x1 +#define VM_L2_PROTECTION_FAULT_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x2 +#define VM_L2_PROTECTION_FAULT_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x3 +#define VM_L2_PROTECTION_FAULT_CNTL__PDE1_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x4 +#define VM_L2_PROTECTION_FAULT_CNTL__PDE2_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x5 +#define VM_L2_PROTECTION_FAULT_CNTL__TRANSLATE_FURTHER_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x6 +#define VM_L2_PROTECTION_FAULT_CNTL__NACK_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x7 +#define VM_L2_PROTECTION_FAULT_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x8 +#define VM_L2_PROTECTION_FAULT_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x9 +#define VM_L2_PROTECTION_FAULT_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_L2_PROTECTION_FAULT_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xb +#define VM_L2_PROTECTION_FAULT_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_L2_PROTECTION_FAULT_CNTL__CLIENT_ID_NO_RETRY_FAULT_INTERRUPT__SHIFT 0xd +#define VM_L2_PROTECTION_FAULT_CNTL__OTHER_CLIENT_ID_NO_RETRY_FAULT_INTERRUPT__SHIFT 0x1d +#define VM_L2_PROTECTION_FAULT_CNTL__CRASH_ON_NO_RETRY_FAULT__SHIFT 0x1e +#define VM_L2_PROTECTION_FAULT_CNTL__CRASH_ON_RETRY_FAULT__SHIFT 0x1f +#define VM_L2_PROTECTION_FAULT_CNTL__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00000001L +#define VM_L2_PROTECTION_FAULT_CNTL__ALLOW_SUBSEQUENT_PROTECTION_FAULT_STATUS_ADDR_UPDATES_MASK 0x00000002L +#define VM_L2_PROTECTION_FAULT_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000004L +#define VM_L2_PROTECTION_FAULT_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000008L +#define VM_L2_PROTECTION_FAULT_CNTL__PDE1_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000010L +#define VM_L2_PROTECTION_FAULT_CNTL__PDE2_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000020L +#define VM_L2_PROTECTION_FAULT_CNTL__TRANSLATE_FURTHER_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000040L +#define VM_L2_PROTECTION_FAULT_CNTL__NACK_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000080L +#define VM_L2_PROTECTION_FAULT_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000100L +#define VM_L2_PROTECTION_FAULT_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000200L +#define VM_L2_PROTECTION_FAULT_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_L2_PROTECTION_FAULT_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000800L +#define VM_L2_PROTECTION_FAULT_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_L2_PROTECTION_FAULT_CNTL__CLIENT_ID_NO_RETRY_FAULT_INTERRUPT_MASK 0x1FFFE000L +#define VM_L2_PROTECTION_FAULT_CNTL__OTHER_CLIENT_ID_NO_RETRY_FAULT_INTERRUPT_MASK 0x20000000L +#define VM_L2_PROTECTION_FAULT_CNTL__CRASH_ON_NO_RETRY_FAULT_MASK 0x40000000L +#define VM_L2_PROTECTION_FAULT_CNTL__CRASH_ON_RETRY_FAULT_MASK 0x80000000L +//VM_L2_PROTECTION_FAULT_CNTL2 +#define VM_L2_PROTECTION_FAULT_CNTL2__CLIENT_ID_PRT_FAULT_INTERRUPT__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_CNTL2__OTHER_CLIENT_ID_PRT_FAULT_INTERRUPT__SHIFT 0x10 +#define VM_L2_PROTECTION_FAULT_CNTL2__ACTIVE_PAGE_MIGRATION_PTE__SHIFT 0x11 +#define VM_L2_PROTECTION_FAULT_CNTL2__ACTIVE_PAGE_MIGRATION_PTE_READ_RETRY__SHIFT 0x12 +#define VM_L2_PROTECTION_FAULT_CNTL2__ENABLE_RETRY_FAULT_INTERRUPT__SHIFT 0x13 +#define VM_L2_PROTECTION_FAULT_CNTL2__CLIENT_ID_PRT_FAULT_INTERRUPT_MASK 0x0000FFFFL +#define VM_L2_PROTECTION_FAULT_CNTL2__OTHER_CLIENT_ID_PRT_FAULT_INTERRUPT_MASK 0x00010000L +#define VM_L2_PROTECTION_FAULT_CNTL2__ACTIVE_PAGE_MIGRATION_PTE_MASK 0x00020000L +#define VM_L2_PROTECTION_FAULT_CNTL2__ACTIVE_PAGE_MIGRATION_PTE_READ_RETRY_MASK 0x00040000L +#define VM_L2_PROTECTION_FAULT_CNTL2__ENABLE_RETRY_FAULT_INTERRUPT_MASK 0x00080000L +//VM_L2_PROTECTION_FAULT_MM_CNTL3 +#define VM_L2_PROTECTION_FAULT_MM_CNTL3__VML1_READ_CLIENT_ID_NO_RETRY_FAULT_INTERRUPT__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_MM_CNTL3__VML1_READ_CLIENT_ID_NO_RETRY_FAULT_INTERRUPT_MASK 0xFFFFFFFFL +//VM_L2_PROTECTION_FAULT_MM_CNTL4 +#define VM_L2_PROTECTION_FAULT_MM_CNTL4__VML1_WRITE_CLIENT_ID_NO_RETRY_FAULT_INTERRUPT__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_MM_CNTL4__VML1_WRITE_CLIENT_ID_NO_RETRY_FAULT_INTERRUPT_MASK 0xFFFFFFFFL +//VM_L2_PROTECTION_FAULT_STATUS +#define VM_L2_PROTECTION_FAULT_STATUS__MORE_FAULTS__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_STATUS__WALKER_ERROR__SHIFT 0x1 +#define VM_L2_PROTECTION_FAULT_STATUS__PERMISSION_FAULTS__SHIFT 0x4 +#define VM_L2_PROTECTION_FAULT_STATUS__MAPPING_ERROR__SHIFT 0x8 +#define VM_L2_PROTECTION_FAULT_STATUS__CID__SHIFT 0x9 +#define VM_L2_PROTECTION_FAULT_STATUS__RW__SHIFT 0x12 +#define VM_L2_PROTECTION_FAULT_STATUS__ATOMIC__SHIFT 0x13 +#define VM_L2_PROTECTION_FAULT_STATUS__VMID__SHIFT 0x14 +#define VM_L2_PROTECTION_FAULT_STATUS__VF__SHIFT 0x18 +#define VM_L2_PROTECTION_FAULT_STATUS__VFID__SHIFT 0x19 +#define VM_L2_PROTECTION_FAULT_STATUS__MORE_FAULTS_MASK 0x00000001L +#define VM_L2_PROTECTION_FAULT_STATUS__WALKER_ERROR_MASK 0x0000000EL +#define VM_L2_PROTECTION_FAULT_STATUS__PERMISSION_FAULTS_MASK 0x000000F0L +#define VM_L2_PROTECTION_FAULT_STATUS__MAPPING_ERROR_MASK 0x00000100L +#define VM_L2_PROTECTION_FAULT_STATUS__CID_MASK 0x0003FE00L +#define VM_L2_PROTECTION_FAULT_STATUS__RW_MASK 0x00040000L +#define VM_L2_PROTECTION_FAULT_STATUS__ATOMIC_MASK 0x00080000L +#define VM_L2_PROTECTION_FAULT_STATUS__VMID_MASK 0x00F00000L +#define VM_L2_PROTECTION_FAULT_STATUS__VF_MASK 0x01000000L +#define VM_L2_PROTECTION_FAULT_STATUS__VFID_MASK 0x1E000000L +//VM_L2_PROTECTION_FAULT_ADDR_LO32 +#define VM_L2_PROTECTION_FAULT_ADDR_LO32__LOGICAL_PAGE_ADDR_LO32__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_ADDR_LO32__LOGICAL_PAGE_ADDR_LO32_MASK 0xFFFFFFFFL +//VM_L2_PROTECTION_FAULT_ADDR_HI32 +#define VM_L2_PROTECTION_FAULT_ADDR_HI32__LOGICAL_PAGE_ADDR_HI4__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_ADDR_HI32__LOGICAL_PAGE_ADDR_HI4_MASK 0x0000000FL +//VM_L2_PROTECTION_FAULT_DEFAULT_ADDR_LO32 +#define VM_L2_PROTECTION_FAULT_DEFAULT_ADDR_LO32__PHYSICAL_PAGE_ADDR_LO32__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_DEFAULT_ADDR_LO32__PHYSICAL_PAGE_ADDR_LO32_MASK 0xFFFFFFFFL +//VM_L2_PROTECTION_FAULT_DEFAULT_ADDR_HI32 +#define VM_L2_PROTECTION_FAULT_DEFAULT_ADDR_HI32__PHYSICAL_PAGE_ADDR_HI4__SHIFT 0x0 +#define VM_L2_PROTECTION_FAULT_DEFAULT_ADDR_HI32__PHYSICAL_PAGE_ADDR_HI4_MASK 0x0000000FL +//VM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_LO32 +#define VM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_HI32 +#define VM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_L2_CONTEXT1_IDENTITY_APERTURE_HIGH_ADDR_LO32 +#define VM_L2_CONTEXT1_IDENTITY_APERTURE_HIGH_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_L2_CONTEXT1_IDENTITY_APERTURE_HIGH_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_L2_CONTEXT1_IDENTITY_APERTURE_HIGH_ADDR_HI32 +#define VM_L2_CONTEXT1_IDENTITY_APERTURE_HIGH_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_L2_CONTEXT1_IDENTITY_APERTURE_HIGH_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_L2_CONTEXT_IDENTITY_PHYSICAL_OFFSET_LO32 +#define VM_L2_CONTEXT_IDENTITY_PHYSICAL_OFFSET_LO32__PHYSICAL_PAGE_OFFSET_LO32__SHIFT 0x0 +#define VM_L2_CONTEXT_IDENTITY_PHYSICAL_OFFSET_LO32__PHYSICAL_PAGE_OFFSET_LO32_MASK 0xFFFFFFFFL +//VM_L2_CONTEXT_IDENTITY_PHYSICAL_OFFSET_HI32 +#define VM_L2_CONTEXT_IDENTITY_PHYSICAL_OFFSET_HI32__PHYSICAL_PAGE_OFFSET_HI4__SHIFT 0x0 +#define VM_L2_CONTEXT_IDENTITY_PHYSICAL_OFFSET_HI32__PHYSICAL_PAGE_OFFSET_HI4_MASK 0x0000000FL +//VM_L2_CNTL4 +#define VM_L2_CNTL4__L2_CACHE_4K_PARTITION_COUNT__SHIFT 0x0 +#define VM_L2_CNTL4__VMC_TAP_PDE_REQUEST_PHYSICAL__SHIFT 0x6 +#define VM_L2_CNTL4__VMC_TAP_PTE_REQUEST_PHYSICAL__SHIFT 0x7 +#define VM_L2_CNTL4__MM_NONRT_IFIFO_ACTIVE_TRANSACTION_LIMIT__SHIFT 0x8 +#define VM_L2_CNTL4__MM_SOFTRT_IFIFO_ACTIVE_TRANSACTION_LIMIT__SHIFT 0x12 +#define VM_L2_CNTL4__BPM_CGCGLS_OVERRIDE__SHIFT 0x1c +#define VM_L2_CNTL4__L2_CACHE_4K_PARTITION_COUNT_MASK 0x0000003FL +#define VM_L2_CNTL4__VMC_TAP_PDE_REQUEST_PHYSICAL_MASK 0x00000040L +#define VM_L2_CNTL4__VMC_TAP_PTE_REQUEST_PHYSICAL_MASK 0x00000080L +#define VM_L2_CNTL4__MM_NONRT_IFIFO_ACTIVE_TRANSACTION_LIMIT_MASK 0x0003FF00L +#define VM_L2_CNTL4__MM_SOFTRT_IFIFO_ACTIVE_TRANSACTION_LIMIT_MASK 0x0FFC0000L +#define VM_L2_CNTL4__BPM_CGCGLS_OVERRIDE_MASK 0x10000000L +//VM_L2_MM_GROUP_RT_CLASSES +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_0_RT_CLASS__SHIFT 0x0 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_1_RT_CLASS__SHIFT 0x1 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_2_RT_CLASS__SHIFT 0x2 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_3_RT_CLASS__SHIFT 0x3 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_4_RT_CLASS__SHIFT 0x4 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_5_RT_CLASS__SHIFT 0x5 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_6_RT_CLASS__SHIFT 0x6 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_7_RT_CLASS__SHIFT 0x7 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_8_RT_CLASS__SHIFT 0x8 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_9_RT_CLASS__SHIFT 0x9 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_10_RT_CLASS__SHIFT 0xa +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_11_RT_CLASS__SHIFT 0xb +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_12_RT_CLASS__SHIFT 0xc +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_13_RT_CLASS__SHIFT 0xd +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_14_RT_CLASS__SHIFT 0xe +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_15_RT_CLASS__SHIFT 0xf +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_16_RT_CLASS__SHIFT 0x10 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_17_RT_CLASS__SHIFT 0x11 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_18_RT_CLASS__SHIFT 0x12 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_19_RT_CLASS__SHIFT 0x13 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_20_RT_CLASS__SHIFT 0x14 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_21_RT_CLASS__SHIFT 0x15 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_22_RT_CLASS__SHIFT 0x16 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_23_RT_CLASS__SHIFT 0x17 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_24_RT_CLASS__SHIFT 0x18 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_25_RT_CLASS__SHIFT 0x19 +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_26_RT_CLASS__SHIFT 0x1a +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_27_RT_CLASS__SHIFT 0x1b +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_28_RT_CLASS__SHIFT 0x1c +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_29_RT_CLASS__SHIFT 0x1d +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_30_RT_CLASS__SHIFT 0x1e +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_31_RT_CLASS__SHIFT 0x1f +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_0_RT_CLASS_MASK 0x00000001L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_1_RT_CLASS_MASK 0x00000002L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_2_RT_CLASS_MASK 0x00000004L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_3_RT_CLASS_MASK 0x00000008L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_4_RT_CLASS_MASK 0x00000010L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_5_RT_CLASS_MASK 0x00000020L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_6_RT_CLASS_MASK 0x00000040L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_7_RT_CLASS_MASK 0x00000080L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_8_RT_CLASS_MASK 0x00000100L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_9_RT_CLASS_MASK 0x00000200L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_10_RT_CLASS_MASK 0x00000400L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_11_RT_CLASS_MASK 0x00000800L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_12_RT_CLASS_MASK 0x00001000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_13_RT_CLASS_MASK 0x00002000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_14_RT_CLASS_MASK 0x00004000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_15_RT_CLASS_MASK 0x00008000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_16_RT_CLASS_MASK 0x00010000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_17_RT_CLASS_MASK 0x00020000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_18_RT_CLASS_MASK 0x00040000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_19_RT_CLASS_MASK 0x00080000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_20_RT_CLASS_MASK 0x00100000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_21_RT_CLASS_MASK 0x00200000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_22_RT_CLASS_MASK 0x00400000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_23_RT_CLASS_MASK 0x00800000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_24_RT_CLASS_MASK 0x01000000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_25_RT_CLASS_MASK 0x02000000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_26_RT_CLASS_MASK 0x04000000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_27_RT_CLASS_MASK 0x08000000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_28_RT_CLASS_MASK 0x10000000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_29_RT_CLASS_MASK 0x20000000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_30_RT_CLASS_MASK 0x40000000L +#define VM_L2_MM_GROUP_RT_CLASSES__GROUP_31_RT_CLASS_MASK 0x80000000L +//VM_L2_BANK_SELECT_RESERVED_CID +#define VM_L2_BANK_SELECT_RESERVED_CID__RESERVED_READ_CLIENT_ID__SHIFT 0x0 +#define VM_L2_BANK_SELECT_RESERVED_CID__RESERVED_WRITE_CLIENT_ID__SHIFT 0xa +#define VM_L2_BANK_SELECT_RESERVED_CID__ENABLE__SHIFT 0x14 +#define VM_L2_BANK_SELECT_RESERVED_CID__RESERVED_CACHE_INVALIDATION_MODE__SHIFT 0x18 +#define VM_L2_BANK_SELECT_RESERVED_CID__RESERVED_CACHE_PRIVATE_INVALIDATION__SHIFT 0x19 +#define VM_L2_BANK_SELECT_RESERVED_CID__RESERVED_READ_CLIENT_ID_MASK 0x000001FFL +#define VM_L2_BANK_SELECT_RESERVED_CID__RESERVED_WRITE_CLIENT_ID_MASK 0x0007FC00L +#define VM_L2_BANK_SELECT_RESERVED_CID__ENABLE_MASK 0x00100000L +#define VM_L2_BANK_SELECT_RESERVED_CID__RESERVED_CACHE_INVALIDATION_MODE_MASK 0x01000000L +#define VM_L2_BANK_SELECT_RESERVED_CID__RESERVED_CACHE_PRIVATE_INVALIDATION_MASK 0x02000000L +//VM_L2_BANK_SELECT_RESERVED_CID2 +#define VM_L2_BANK_SELECT_RESERVED_CID2__RESERVED_READ_CLIENT_ID__SHIFT 0x0 +#define VM_L2_BANK_SELECT_RESERVED_CID2__RESERVED_WRITE_CLIENT_ID__SHIFT 0xa +#define VM_L2_BANK_SELECT_RESERVED_CID2__ENABLE__SHIFT 0x14 +#define VM_L2_BANK_SELECT_RESERVED_CID2__RESERVED_CACHE_INVALIDATION_MODE__SHIFT 0x18 +#define VM_L2_BANK_SELECT_RESERVED_CID2__RESERVED_CACHE_PRIVATE_INVALIDATION__SHIFT 0x19 +#define VM_L2_BANK_SELECT_RESERVED_CID2__RESERVED_READ_CLIENT_ID_MASK 0x000001FFL +#define VM_L2_BANK_SELECT_RESERVED_CID2__RESERVED_WRITE_CLIENT_ID_MASK 0x0007FC00L +#define VM_L2_BANK_SELECT_RESERVED_CID2__ENABLE_MASK 0x00100000L +#define VM_L2_BANK_SELECT_RESERVED_CID2__RESERVED_CACHE_INVALIDATION_MODE_MASK 0x01000000L +#define VM_L2_BANK_SELECT_RESERVED_CID2__RESERVED_CACHE_PRIVATE_INVALIDATION_MASK 0x02000000L +//VM_L2_CACHE_PARITY_CNTL +#define VM_L2_CACHE_PARITY_CNTL__ENABLE_PARITY_CHECKS_IN_4K_PTE_CACHES__SHIFT 0x0 +#define VM_L2_CACHE_PARITY_CNTL__ENABLE_PARITY_CHECKS_IN_BIGK_PTE_CACHES__SHIFT 0x1 +#define VM_L2_CACHE_PARITY_CNTL__ENABLE_PARITY_CHECKS_IN_PDE_CACHES__SHIFT 0x2 +#define VM_L2_CACHE_PARITY_CNTL__FORCE_PARITY_MISMATCH_IN_4K_PTE_CACHE__SHIFT 0x3 +#define VM_L2_CACHE_PARITY_CNTL__FORCE_PARITY_MISMATCH_IN_BIGK_PTE_CACHE__SHIFT 0x4 +#define VM_L2_CACHE_PARITY_CNTL__FORCE_PARITY_MISMATCH_IN_PDE_CACHE__SHIFT 0x5 +#define VM_L2_CACHE_PARITY_CNTL__FORCE_CACHE_BANK__SHIFT 0x6 +#define VM_L2_CACHE_PARITY_CNTL__FORCE_CACHE_NUMBER__SHIFT 0x9 +#define VM_L2_CACHE_PARITY_CNTL__FORCE_CACHE_ASSOC__SHIFT 0xc +#define VM_L2_CACHE_PARITY_CNTL__ENABLE_PARITY_CHECKS_IN_4K_PTE_CACHES_MASK 0x00000001L +#define VM_L2_CACHE_PARITY_CNTL__ENABLE_PARITY_CHECKS_IN_BIGK_PTE_CACHES_MASK 0x00000002L +#define VM_L2_CACHE_PARITY_CNTL__ENABLE_PARITY_CHECKS_IN_PDE_CACHES_MASK 0x00000004L +#define VM_L2_CACHE_PARITY_CNTL__FORCE_PARITY_MISMATCH_IN_4K_PTE_CACHE_MASK 0x00000008L +#define VM_L2_CACHE_PARITY_CNTL__FORCE_PARITY_MISMATCH_IN_BIGK_PTE_CACHE_MASK 0x00000010L +#define VM_L2_CACHE_PARITY_CNTL__FORCE_PARITY_MISMATCH_IN_PDE_CACHE_MASK 0x00000020L +#define VM_L2_CACHE_PARITY_CNTL__FORCE_CACHE_BANK_MASK 0x000001C0L +#define VM_L2_CACHE_PARITY_CNTL__FORCE_CACHE_NUMBER_MASK 0x00000E00L +#define VM_L2_CACHE_PARITY_CNTL__FORCE_CACHE_ASSOC_MASK 0x0000F000L +//VM_L2_CGTT_CLK_CTRL +#define VM_L2_CGTT_CLK_CTRL__ON_DELAY__SHIFT 0x0 +#define VM_L2_CGTT_CLK_CTRL__OFF_HYSTERESIS__SHIFT 0x4 +#define VM_L2_CGTT_CLK_CTRL__MGLS_OVERRIDE__SHIFT 0xf +#define VM_L2_CGTT_CLK_CTRL__SOFT_STALL_OVERRIDE__SHIFT 0x10 +#define VM_L2_CGTT_CLK_CTRL__SOFT_OVERRIDE__SHIFT 0x18 +#define VM_L2_CGTT_CLK_CTRL__ON_DELAY_MASK 0x0000000FL +#define VM_L2_CGTT_CLK_CTRL__OFF_HYSTERESIS_MASK 0x00000FF0L +#define VM_L2_CGTT_CLK_CTRL__MGLS_OVERRIDE_MASK 0x00008000L +#define VM_L2_CGTT_CLK_CTRL__SOFT_STALL_OVERRIDE_MASK 0x00FF0000L +#define VM_L2_CGTT_CLK_CTRL__SOFT_OVERRIDE_MASK 0xFF000000L + + +// addressBlock: gc_utcl2_vml2vcdec +//VM_CONTEXT0_CNTL +#define VM_CONTEXT0_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT0_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT0_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT0_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT0_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT0_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT0_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT0_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT0_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT0_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT0_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT0_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT0_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT0_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT0_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT0_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT0_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT0_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT0_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT0_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT0_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT0_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT0_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT0_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT0_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT0_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT0_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT0_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT0_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT0_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT0_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT0_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT0_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT0_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT0_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT0_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT0_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT0_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT1_CNTL +#define VM_CONTEXT1_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT1_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT1_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT1_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT1_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT1_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT1_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT1_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT1_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT1_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT2_CNTL +#define VM_CONTEXT2_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT2_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT2_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT2_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT2_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT2_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT2_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT2_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT2_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT2_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT2_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT2_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT2_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT2_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT2_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT2_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT2_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT2_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT2_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT2_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT2_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT2_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT2_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT2_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT2_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT2_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT2_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT2_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT2_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT2_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT2_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT2_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT2_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT2_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT2_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT2_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT2_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT2_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT3_CNTL +#define VM_CONTEXT3_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT3_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT3_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT3_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT3_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT3_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT3_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT3_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT3_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT3_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT3_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT3_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT3_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT3_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT3_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT3_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT3_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT3_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT3_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT3_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT3_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT3_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT3_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT3_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT3_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT3_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT3_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT3_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT3_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT3_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT3_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT3_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT3_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT3_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT3_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT3_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT3_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT3_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT4_CNTL +#define VM_CONTEXT4_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT4_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT4_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT4_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT4_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT4_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT4_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT4_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT4_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT4_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT4_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT4_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT4_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT4_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT4_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT4_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT4_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT4_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT4_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT4_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT4_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT4_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT4_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT4_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT4_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT4_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT4_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT4_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT4_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT4_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT4_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT4_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT4_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT4_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT4_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT4_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT4_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT4_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT5_CNTL +#define VM_CONTEXT5_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT5_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT5_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT5_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT5_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT5_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT5_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT5_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT5_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT5_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT5_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT5_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT5_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT5_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT5_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT5_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT5_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT5_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT5_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT5_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT5_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT5_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT5_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT5_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT5_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT5_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT5_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT5_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT5_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT5_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT5_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT5_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT5_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT5_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT5_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT5_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT5_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT5_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT6_CNTL +#define VM_CONTEXT6_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT6_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT6_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT6_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT6_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT6_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT6_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT6_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT6_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT6_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT6_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT6_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT6_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT6_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT6_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT6_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT6_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT6_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT6_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT6_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT6_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT6_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT6_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT6_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT6_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT6_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT6_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT6_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT6_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT6_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT6_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT6_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT6_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT6_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT6_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT6_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT6_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT6_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT7_CNTL +#define VM_CONTEXT7_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT7_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT7_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT7_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT7_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT7_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT7_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT7_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT7_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT7_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT7_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT7_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT7_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT7_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT7_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT7_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT7_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT7_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT7_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT7_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT7_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT7_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT7_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT7_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT7_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT7_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT7_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT7_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT7_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT7_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT7_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT7_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT7_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT7_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT7_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT7_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT7_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT7_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT8_CNTL +#define VM_CONTEXT8_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT8_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT8_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT8_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT8_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT8_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT8_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT8_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT8_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT8_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT8_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT8_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT8_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT8_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT8_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT8_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT8_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT8_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT8_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT8_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT8_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT8_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT8_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT8_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT8_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT8_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT8_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT8_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT8_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT8_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT8_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT8_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT8_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT8_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT8_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT8_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT8_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT8_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT9_CNTL +#define VM_CONTEXT9_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT9_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT9_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT9_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT9_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT9_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT9_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT9_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT9_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT9_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT9_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT9_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT9_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT9_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT9_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT9_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT9_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT9_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT9_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT9_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT9_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT9_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT9_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT9_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT9_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT9_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT9_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT9_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT9_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT9_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT9_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT9_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT9_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT9_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT9_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT9_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT9_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT9_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT10_CNTL +#define VM_CONTEXT10_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT10_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT10_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT10_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT10_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT10_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT10_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT10_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT10_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT10_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT10_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT10_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT10_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT10_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT10_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT10_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT10_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT10_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT10_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT10_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT10_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT10_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT10_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT10_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT10_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT10_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT10_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT10_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT10_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT10_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT10_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT10_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT10_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT10_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT10_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT10_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT10_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT10_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT11_CNTL +#define VM_CONTEXT11_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT11_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT11_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT11_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT11_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT11_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT11_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT11_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT11_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT11_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT11_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT11_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT11_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT11_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT11_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT11_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT11_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT11_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT11_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT11_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT11_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT11_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT11_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT11_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT11_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT11_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT11_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT11_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT11_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT11_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT11_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT11_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT11_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT11_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT11_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT11_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT11_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT11_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT12_CNTL +#define VM_CONTEXT12_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT12_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT12_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT12_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT12_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT12_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT12_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT12_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT12_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT12_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT12_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT12_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT12_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT12_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT12_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT12_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT12_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT12_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT12_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT12_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT12_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT12_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT12_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT12_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT12_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT12_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT12_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT12_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT12_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT12_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT12_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT12_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT12_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT12_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT12_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT12_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT12_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT12_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT13_CNTL +#define VM_CONTEXT13_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT13_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT13_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT13_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT13_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT13_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT13_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT13_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT13_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT13_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT13_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT13_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT13_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT13_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT13_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT13_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT13_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT13_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT13_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT13_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT13_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT13_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT13_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT13_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT13_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT13_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT13_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT13_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT13_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT13_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT13_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT13_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT13_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT13_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT13_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT13_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT13_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT13_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT14_CNTL +#define VM_CONTEXT14_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT14_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT14_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT14_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT14_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT14_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT14_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT14_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT14_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT14_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT14_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT14_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT14_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT14_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT14_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT14_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT14_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT14_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT14_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT14_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT14_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT14_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT14_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT14_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT14_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT14_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT14_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT14_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT14_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT14_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT14_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT14_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT14_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT14_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT14_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT14_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT14_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT14_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXT15_CNTL +#define VM_CONTEXT15_CNTL__ENABLE_CONTEXT__SHIFT 0x0 +#define VM_CONTEXT15_CNTL__PAGE_TABLE_DEPTH__SHIFT 0x1 +#define VM_CONTEXT15_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT 0x3 +#define VM_CONTEXT15_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT__SHIFT 0x7 +#define VM_CONTEXT15_CNTL__RETRY_OTHER_FAULT__SHIFT 0x8 +#define VM_CONTEXT15_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT15_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define VM_CONTEXT15_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xb +#define VM_CONTEXT15_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xc +#define VM_CONTEXT15_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xd +#define VM_CONTEXT15_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xe +#define VM_CONTEXT15_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT15_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define VM_CONTEXT15_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x11 +#define VM_CONTEXT15_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x12 +#define VM_CONTEXT15_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x13 +#define VM_CONTEXT15_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x14 +#define VM_CONTEXT15_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x15 +#define VM_CONTEXT15_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x16 +#define VM_CONTEXT15_CNTL__ENABLE_CONTEXT_MASK 0x00000001L +#define VM_CONTEXT15_CNTL__PAGE_TABLE_DEPTH_MASK 0x00000006L +#define VM_CONTEXT15_CNTL__PAGE_TABLE_BLOCK_SIZE_MASK 0x00000078L +#define VM_CONTEXT15_CNTL__RETRY_PERMISSION_OR_INVALID_PAGE_FAULT_MASK 0x00000080L +#define VM_CONTEXT15_CNTL__RETRY_OTHER_FAULT_MASK 0x00000100L +#define VM_CONTEXT15_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000200L +#define VM_CONTEXT15_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00000400L +#define VM_CONTEXT15_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00000800L +#define VM_CONTEXT15_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00001000L +#define VM_CONTEXT15_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00002000L +#define VM_CONTEXT15_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00004000L +#define VM_CONTEXT15_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00008000L +#define VM_CONTEXT15_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00010000L +#define VM_CONTEXT15_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00020000L +#define VM_CONTEXT15_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00040000L +#define VM_CONTEXT15_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00080000L +#define VM_CONTEXT15_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00100000L +#define VM_CONTEXT15_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x00200000L +#define VM_CONTEXT15_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x00400000L +//VM_CONTEXTS_DISABLE +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_0__SHIFT 0x0 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_1__SHIFT 0x1 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_2__SHIFT 0x2 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_3__SHIFT 0x3 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_4__SHIFT 0x4 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_5__SHIFT 0x5 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_6__SHIFT 0x6 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_7__SHIFT 0x7 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_8__SHIFT 0x8 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_9__SHIFT 0x9 +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_10__SHIFT 0xa +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_11__SHIFT 0xb +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_12__SHIFT 0xc +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_13__SHIFT 0xd +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_14__SHIFT 0xe +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_15__SHIFT 0xf +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_0_MASK 0x00000001L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_1_MASK 0x00000002L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_2_MASK 0x00000004L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_3_MASK 0x00000008L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_4_MASK 0x00000010L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_5_MASK 0x00000020L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_6_MASK 0x00000040L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_7_MASK 0x00000080L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_8_MASK 0x00000100L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_9_MASK 0x00000200L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_10_MASK 0x00000400L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_11_MASK 0x00000800L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_12_MASK 0x00001000L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_13_MASK 0x00002000L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_14_MASK 0x00004000L +#define VM_CONTEXTS_DISABLE__DISABLE_CONTEXT_15_MASK 0x00008000L +//VM_INVALIDATE_ENG0_SEM +#define VM_INVALIDATE_ENG0_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG0_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG1_SEM +#define VM_INVALIDATE_ENG1_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG1_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG2_SEM +#define VM_INVALIDATE_ENG2_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG2_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG3_SEM +#define VM_INVALIDATE_ENG3_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG3_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG4_SEM +#define VM_INVALIDATE_ENG4_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG4_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG5_SEM +#define VM_INVALIDATE_ENG5_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG5_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG6_SEM +#define VM_INVALIDATE_ENG6_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG6_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG7_SEM +#define VM_INVALIDATE_ENG7_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG7_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG8_SEM +#define VM_INVALIDATE_ENG8_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG8_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG9_SEM +#define VM_INVALIDATE_ENG9_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG9_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG10_SEM +#define VM_INVALIDATE_ENG10_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG10_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG11_SEM +#define VM_INVALIDATE_ENG11_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG11_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG12_SEM +#define VM_INVALIDATE_ENG12_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG12_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG13_SEM +#define VM_INVALIDATE_ENG13_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG13_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG14_SEM +#define VM_INVALIDATE_ENG14_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG14_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG15_SEM +#define VM_INVALIDATE_ENG15_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG15_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG16_SEM +#define VM_INVALIDATE_ENG16_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG16_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG17_SEM +#define VM_INVALIDATE_ENG17_SEM__SEMAPHORE__SHIFT 0x0 +#define VM_INVALIDATE_ENG17_SEM__SEMAPHORE_MASK 0x00000001L +//VM_INVALIDATE_ENG0_REQ +#define VM_INVALIDATE_ENG0_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG0_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG0_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG0_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG0_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG0_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG0_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG1_REQ +#define VM_INVALIDATE_ENG1_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG1_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG1_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG1_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG1_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG1_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG1_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG2_REQ +#define VM_INVALIDATE_ENG2_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG2_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG2_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG2_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG2_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG2_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG2_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG3_REQ +#define VM_INVALIDATE_ENG3_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG3_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG3_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG3_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG3_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG3_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG3_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG4_REQ +#define VM_INVALIDATE_ENG4_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG4_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG4_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG4_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG4_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG4_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG4_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG5_REQ +#define VM_INVALIDATE_ENG5_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG5_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG5_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG5_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG5_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG5_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG5_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG6_REQ +#define VM_INVALIDATE_ENG6_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG6_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG6_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG6_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG6_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG6_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG6_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG7_REQ +#define VM_INVALIDATE_ENG7_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG7_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG7_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG7_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG7_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG7_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG7_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG8_REQ +#define VM_INVALIDATE_ENG8_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG8_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG8_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG8_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG8_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG8_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG8_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG9_REQ +#define VM_INVALIDATE_ENG9_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG9_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG9_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG9_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG9_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG9_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG9_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG10_REQ +#define VM_INVALIDATE_ENG10_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG10_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG10_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG10_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG10_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG10_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG10_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG11_REQ +#define VM_INVALIDATE_ENG11_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG11_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG11_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG11_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG11_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG11_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG11_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG12_REQ +#define VM_INVALIDATE_ENG12_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG12_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG12_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG12_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG12_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG12_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG12_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG13_REQ +#define VM_INVALIDATE_ENG13_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG13_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG13_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG13_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG13_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG13_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG13_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG14_REQ +#define VM_INVALIDATE_ENG14_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG14_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG14_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG14_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG14_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG14_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG14_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG15_REQ +#define VM_INVALIDATE_ENG15_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG15_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG15_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG15_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG15_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG15_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG15_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG16_REQ +#define VM_INVALIDATE_ENG16_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG16_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG16_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG16_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG16_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG16_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG16_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG17_REQ +#define VM_INVALIDATE_ENG17_REQ__PER_VMID_INVALIDATE_REQ__SHIFT 0x0 +#define VM_INVALIDATE_ENG17_REQ__FLUSH_TYPE__SHIFT 0x10 +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L2_PTES__SHIFT 0x12 +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L2_PDE0__SHIFT 0x13 +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L2_PDE1__SHIFT 0x14 +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L2_PDE2__SHIFT 0x15 +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L1_PTES__SHIFT 0x16 +#define VM_INVALIDATE_ENG17_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR__SHIFT 0x17 +#define VM_INVALIDATE_ENG17_REQ__PER_VMID_INVALIDATE_REQ_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG17_REQ__FLUSH_TYPE_MASK 0x00030000L +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L2_PTES_MASK 0x00040000L +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L2_PDE0_MASK 0x00080000L +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L2_PDE1_MASK 0x00100000L +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L2_PDE2_MASK 0x00200000L +#define VM_INVALIDATE_ENG17_REQ__INVALIDATE_L1_PTES_MASK 0x00400000L +#define VM_INVALIDATE_ENG17_REQ__CLEAR_PROTECTION_FAULT_STATUS_ADDR_MASK 0x00800000L +//VM_INVALIDATE_ENG0_ACK +#define VM_INVALIDATE_ENG0_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG0_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG0_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG0_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG1_ACK +#define VM_INVALIDATE_ENG1_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG1_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG1_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG1_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG2_ACK +#define VM_INVALIDATE_ENG2_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG2_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG2_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG2_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG3_ACK +#define VM_INVALIDATE_ENG3_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG3_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG3_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG3_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG4_ACK +#define VM_INVALIDATE_ENG4_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG4_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG4_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG4_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG5_ACK +#define VM_INVALIDATE_ENG5_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG5_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG5_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG5_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG6_ACK +#define VM_INVALIDATE_ENG6_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG6_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG6_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG6_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG7_ACK +#define VM_INVALIDATE_ENG7_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG7_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG7_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG7_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG8_ACK +#define VM_INVALIDATE_ENG8_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG8_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG8_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG8_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG9_ACK +#define VM_INVALIDATE_ENG9_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG9_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG9_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG9_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG10_ACK +#define VM_INVALIDATE_ENG10_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG10_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG10_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG10_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG11_ACK +#define VM_INVALIDATE_ENG11_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG11_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG11_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG11_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG12_ACK +#define VM_INVALIDATE_ENG12_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG12_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG12_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG12_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG13_ACK +#define VM_INVALIDATE_ENG13_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG13_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG13_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG13_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG14_ACK +#define VM_INVALIDATE_ENG14_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG14_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG14_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG14_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG15_ACK +#define VM_INVALIDATE_ENG15_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG15_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG15_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG15_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG16_ACK +#define VM_INVALIDATE_ENG16_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG16_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG16_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG16_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG17_ACK +#define VM_INVALIDATE_ENG17_ACK__PER_VMID_INVALIDATE_ACK__SHIFT 0x0 +#define VM_INVALIDATE_ENG17_ACK__SEMAPHORE__SHIFT 0x10 +#define VM_INVALIDATE_ENG17_ACK__PER_VMID_INVALIDATE_ACK_MASK 0x0000FFFFL +#define VM_INVALIDATE_ENG17_ACK__SEMAPHORE_MASK 0x00010000L +//VM_INVALIDATE_ENG0_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG0_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG0_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG0_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG0_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG0_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG0_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG0_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG1_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG1_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG1_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG1_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG1_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG1_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG1_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG1_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG2_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG2_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG2_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG2_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG2_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG2_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG2_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG2_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG3_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG3_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG3_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG3_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG3_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG3_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG3_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG3_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG4_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG4_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG4_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG4_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG4_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG4_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG4_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG4_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG5_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG5_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG5_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG5_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG5_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG5_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG5_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG5_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG6_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG6_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG6_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG6_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG6_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG6_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG6_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG6_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG7_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG7_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG7_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG7_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG7_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG7_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG7_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG7_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG8_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG8_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG8_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG8_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG8_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG8_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG8_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG8_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG9_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG9_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG9_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG9_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG9_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG9_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG9_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG9_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG10_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG10_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG10_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG10_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG10_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG10_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG10_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG10_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG11_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG11_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG11_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG11_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG11_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG11_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG11_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG11_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG12_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG12_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG12_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG12_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG12_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG12_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG12_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG12_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG13_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG13_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG13_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG13_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG13_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG13_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG13_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG13_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG14_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG14_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG14_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG14_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG14_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG14_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG14_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG14_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG15_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG15_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG15_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG15_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG15_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG15_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG15_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG15_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG16_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG16_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG16_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG16_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG16_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG16_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG16_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG16_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_INVALIDATE_ENG17_ADDR_RANGE_LO32 +#define VM_INVALIDATE_ENG17_ADDR_RANGE_LO32__S_BIT__SHIFT 0x0 +#define VM_INVALIDATE_ENG17_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31__SHIFT 0x1 +#define VM_INVALIDATE_ENG17_ADDR_RANGE_LO32__S_BIT_MASK 0x00000001L +#define VM_INVALIDATE_ENG17_ADDR_RANGE_LO32__LOGI_PAGE_ADDR_RANGE_LO31_MASK 0xFFFFFFFEL +//VM_INVALIDATE_ENG17_ADDR_RANGE_HI32 +#define VM_INVALIDATE_ENG17_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5__SHIFT 0x0 +#define VM_INVALIDATE_ENG17_ADDR_RANGE_HI32__LOGI_PAGE_ADDR_RANGE_HI5_MASK 0x0000001FL +//VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT1_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT1_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT1_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT1_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT1_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT1_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT2_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT2_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT2_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT2_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT2_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT2_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT3_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT3_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT3_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT3_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT3_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT3_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT4_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT4_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT4_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT4_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT4_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT4_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT5_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT5_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT5_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT5_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT5_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT5_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT6_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT6_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT6_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT6_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT6_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT6_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT7_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT7_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT7_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT7_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT7_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT7_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT8_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT8_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT8_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT8_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT8_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT8_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT9_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT9_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT9_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT9_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT9_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT9_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT10_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT10_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT10_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT10_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT10_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT10_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT11_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT11_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT11_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT11_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT11_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT11_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT12_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT12_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT12_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT12_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT12_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT12_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT13_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT13_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT13_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT13_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT13_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT13_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT14_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT14_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT14_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT14_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT14_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT14_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT15_PAGE_TABLE_BASE_ADDR_LO32 +#define VM_CONTEXT15_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32__SHIFT 0x0 +#define VM_CONTEXT15_PAGE_TABLE_BASE_ADDR_LO32__PAGE_DIRECTORY_ENTRY_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT15_PAGE_TABLE_BASE_ADDR_HI32 +#define VM_CONTEXT15_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32__SHIFT 0x0 +#define VM_CONTEXT15_PAGE_TABLE_BASE_ADDR_HI32__PAGE_DIRECTORY_ENTRY_HI32_MASK 0xFFFFFFFFL +//VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT2_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT2_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT2_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT2_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT2_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT2_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT3_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT3_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT3_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT3_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT3_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT3_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT4_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT4_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT4_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT4_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT4_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT4_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT5_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT5_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT5_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT5_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT5_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT5_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT6_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT6_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT6_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT6_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT6_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT6_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT7_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT7_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT7_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT7_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT7_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT7_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT8_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT8_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT8_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT8_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT8_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT8_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT9_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT9_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT9_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT9_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT9_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT9_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT10_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT10_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT10_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT10_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT10_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT10_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT11_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT11_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT11_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT11_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT11_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT11_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT12_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT12_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT12_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT12_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT12_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT12_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT13_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT13_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT13_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT13_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT13_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT13_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT14_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT14_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT14_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT14_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT14_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT14_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT15_PAGE_TABLE_START_ADDR_LO32 +#define VM_CONTEXT15_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT15_PAGE_TABLE_START_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT15_PAGE_TABLE_START_ADDR_HI32 +#define VM_CONTEXT15_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT15_PAGE_TABLE_START_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT1_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT1_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT1_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT1_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT1_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT1_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT2_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT2_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT2_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT2_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT2_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT2_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT3_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT3_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT3_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT3_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT3_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT3_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT4_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT4_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT4_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT4_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT4_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT4_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT5_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT5_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT5_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT5_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT5_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT5_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT6_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT6_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT6_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT6_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT6_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT6_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT7_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT7_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT7_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT7_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT7_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT7_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT8_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT8_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT8_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT8_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT8_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT8_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT9_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT9_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT9_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT9_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT9_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT9_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT10_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT10_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT10_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT10_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT10_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT10_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT11_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT11_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT11_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT11_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT11_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT11_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT12_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT12_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT12_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT12_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT12_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT12_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT13_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT13_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT13_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT13_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT13_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT13_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT14_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT14_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT14_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT14_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT14_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT14_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL +//VM_CONTEXT15_PAGE_TABLE_END_ADDR_LO32 +#define VM_CONTEXT15_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32__SHIFT 0x0 +#define VM_CONTEXT15_PAGE_TABLE_END_ADDR_LO32__LOGICAL_PAGE_NUMBER_LO32_MASK 0xFFFFFFFFL +//VM_CONTEXT15_PAGE_TABLE_END_ADDR_HI32 +#define VM_CONTEXT15_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4__SHIFT 0x0 +#define VM_CONTEXT15_PAGE_TABLE_END_ADDR_HI32__LOGICAL_PAGE_NUMBER_HI4_MASK 0x0000000FL + + +// addressBlock: gc_utcl2_vmsharedpfdec +//MC_VM_NB_MMIOBASE +#define MC_VM_NB_MMIOBASE__MMIOBASE__SHIFT 0x0 +#define MC_VM_NB_MMIOBASE__MMIOBASE_MASK 0xFFFFFFFFL +//MC_VM_NB_MMIOLIMIT +#define MC_VM_NB_MMIOLIMIT__MMIOLIMIT__SHIFT 0x0 +#define MC_VM_NB_MMIOLIMIT__MMIOLIMIT_MASK 0xFFFFFFFFL +//MC_VM_NB_PCI_CTRL +#define MC_VM_NB_PCI_CTRL__MMIOENABLE__SHIFT 0x17 +#define MC_VM_NB_PCI_CTRL__MMIOENABLE_MASK 0x00800000L +//MC_VM_NB_PCI_ARB +#define MC_VM_NB_PCI_ARB__VGA_HOLE__SHIFT 0x3 +#define MC_VM_NB_PCI_ARB__VGA_HOLE_MASK 0x00000008L +//MC_VM_NB_TOP_OF_DRAM_SLOT1 +#define MC_VM_NB_TOP_OF_DRAM_SLOT1__TOP_OF_DRAM__SHIFT 0x17 +#define MC_VM_NB_TOP_OF_DRAM_SLOT1__TOP_OF_DRAM_MASK 0xFF800000L +//MC_VM_NB_LOWER_TOP_OF_DRAM2 +#define MC_VM_NB_LOWER_TOP_OF_DRAM2__ENABLE__SHIFT 0x0 +#define MC_VM_NB_LOWER_TOP_OF_DRAM2__LOWER_TOM2__SHIFT 0x17 +#define MC_VM_NB_LOWER_TOP_OF_DRAM2__ENABLE_MASK 0x00000001L +#define MC_VM_NB_LOWER_TOP_OF_DRAM2__LOWER_TOM2_MASK 0xFF800000L +//MC_VM_NB_UPPER_TOP_OF_DRAM2 +#define MC_VM_NB_UPPER_TOP_OF_DRAM2__UPPER_TOM2__SHIFT 0x0 +#define MC_VM_NB_UPPER_TOP_OF_DRAM2__UPPER_TOM2_MASK 0x00000FFFL +//MC_VM_FB_OFFSET +#define MC_VM_FB_OFFSET__FB_OFFSET__SHIFT 0x0 +#define MC_VM_FB_OFFSET__FB_OFFSET_MASK 0x00FFFFFFL +//MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB +#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB__PHYSICAL_PAGE_NUMBER_LSB__SHIFT 0x0 +#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB__PHYSICAL_PAGE_NUMBER_LSB_MASK 0xFFFFFFFFL +//MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_MSB +#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_MSB__PHYSICAL_PAGE_NUMBER_MSB__SHIFT 0x0 +#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_MSB__PHYSICAL_PAGE_NUMBER_MSB_MASK 0x0000000FL +//MC_VM_STEERING +#define MC_VM_STEERING__DEFAULT_STEERING__SHIFT 0x0 +#define MC_VM_STEERING__DEFAULT_STEERING_MASK 0x00000003L +//MC_SHARED_VIRT_RESET_REQ +#define MC_SHARED_VIRT_RESET_REQ__VF__SHIFT 0x0 +#define MC_SHARED_VIRT_RESET_REQ__PF__SHIFT 0x1f +#define MC_SHARED_VIRT_RESET_REQ__VF_MASK 0x0000FFFFL +#define MC_SHARED_VIRT_RESET_REQ__PF_MASK 0x80000000L +//MC_MEM_POWER_LS +#define MC_MEM_POWER_LS__LS_SETUP__SHIFT 0x0 +#define MC_MEM_POWER_LS__LS_HOLD__SHIFT 0x6 +#define MC_MEM_POWER_LS__LS_SETUP_MASK 0x0000003FL +#define MC_MEM_POWER_LS__LS_HOLD_MASK 0x00000FC0L +//MC_VM_CACHEABLE_DRAM_ADDRESS_START +#define MC_VM_CACHEABLE_DRAM_ADDRESS_START__ADDRESS__SHIFT 0x0 +#define MC_VM_CACHEABLE_DRAM_ADDRESS_START__ADDRESS_MASK 0x000FFFFFL +//MC_VM_CACHEABLE_DRAM_ADDRESS_END +#define MC_VM_CACHEABLE_DRAM_ADDRESS_END__ADDRESS__SHIFT 0x0 +#define MC_VM_CACHEABLE_DRAM_ADDRESS_END__ADDRESS_MASK 0x000FFFFFL +//MC_VM_APT_CNTL +#define MC_VM_APT_CNTL__FORCE_MTYPE_UC__SHIFT 0x0 +#define MC_VM_APT_CNTL__DIRECT_SYSTEM_EN__SHIFT 0x1 +#define MC_VM_APT_CNTL__FORCE_MTYPE_UC_MASK 0x00000001L +#define MC_VM_APT_CNTL__DIRECT_SYSTEM_EN_MASK 0x00000002L +//MC_VM_LOCAL_HBM_ADDRESS_START +#define MC_VM_LOCAL_HBM_ADDRESS_START__ADDRESS__SHIFT 0x0 +#define MC_VM_LOCAL_HBM_ADDRESS_START__ADDRESS_MASK 0x000FFFFFL +//MC_VM_LOCAL_HBM_ADDRESS_END +#define MC_VM_LOCAL_HBM_ADDRESS_END__ADDRESS__SHIFT 0x0 +#define MC_VM_LOCAL_HBM_ADDRESS_END__ADDRESS_MASK 0x000FFFFFL +//MC_VM_LOCAL_HBM_ADDRESS_LOCK_CNTL +#define MC_VM_LOCAL_HBM_ADDRESS_LOCK_CNTL__LOCK__SHIFT 0x0 +#define MC_VM_LOCAL_HBM_ADDRESS_LOCK_CNTL__LOCK_MASK 0x00000001L + + +// addressBlock: gc_utcl2_vmsharedvcdec +//MC_VM_FB_LOCATION_BASE +#define MC_VM_FB_LOCATION_BASE__FB_BASE__SHIFT 0x0 +#define MC_VM_FB_LOCATION_BASE__FB_BASE_MASK 0x00FFFFFFL +//MC_VM_FB_LOCATION_TOP +#define MC_VM_FB_LOCATION_TOP__FB_TOP__SHIFT 0x0 +#define MC_VM_FB_LOCATION_TOP__FB_TOP_MASK 0x00FFFFFFL +//MC_VM_AGP_TOP +#define MC_VM_AGP_TOP__AGP_TOP__SHIFT 0x0 +#define MC_VM_AGP_TOP__AGP_TOP_MASK 0x00FFFFFFL +//MC_VM_AGP_BOT +#define MC_VM_AGP_BOT__AGP_BOT__SHIFT 0x0 +#define MC_VM_AGP_BOT__AGP_BOT_MASK 0x00FFFFFFL +//MC_VM_AGP_BASE +#define MC_VM_AGP_BASE__AGP_BASE__SHIFT 0x0 +#define MC_VM_AGP_BASE__AGP_BASE_MASK 0x00FFFFFFL +//MC_VM_SYSTEM_APERTURE_LOW_ADDR +#define MC_VM_SYSTEM_APERTURE_LOW_ADDR__LOGICAL_ADDR__SHIFT 0x0 +#define MC_VM_SYSTEM_APERTURE_LOW_ADDR__LOGICAL_ADDR_MASK 0x3FFFFFFFL +//MC_VM_SYSTEM_APERTURE_HIGH_ADDR +#define MC_VM_SYSTEM_APERTURE_HIGH_ADDR__LOGICAL_ADDR__SHIFT 0x0 +#define MC_VM_SYSTEM_APERTURE_HIGH_ADDR__LOGICAL_ADDR_MASK 0x3FFFFFFFL +//MC_VM_MX_L1_TLB_CNTL +#define MC_VM_MX_L1_TLB_CNTL__ENABLE_L1_TLB__SHIFT 0x0 +#define MC_VM_MX_L1_TLB_CNTL__SYSTEM_ACCESS_MODE__SHIFT 0x3 +#define MC_VM_MX_L1_TLB_CNTL__SYSTEM_APERTURE_UNMAPPED_ACCESS__SHIFT 0x5 +#define MC_VM_MX_L1_TLB_CNTL__ENABLE_ADVANCED_DRIVER_MODEL__SHIFT 0x6 +#define MC_VM_MX_L1_TLB_CNTL__ECO_BITS__SHIFT 0x7 +#define MC_VM_MX_L1_TLB_CNTL__MTYPE__SHIFT 0xb +#define MC_VM_MX_L1_TLB_CNTL__ATC_EN__SHIFT 0xd +#define MC_VM_MX_L1_TLB_CNTL__ENABLE_L1_TLB_MASK 0x00000001L +#define MC_VM_MX_L1_TLB_CNTL__SYSTEM_ACCESS_MODE_MASK 0x00000018L +#define MC_VM_MX_L1_TLB_CNTL__SYSTEM_APERTURE_UNMAPPED_ACCESS_MASK 0x00000020L +#define MC_VM_MX_L1_TLB_CNTL__ENABLE_ADVANCED_DRIVER_MODEL_MASK 0x00000040L +#define MC_VM_MX_L1_TLB_CNTL__ECO_BITS_MASK 0x00000780L +#define MC_VM_MX_L1_TLB_CNTL__MTYPE_MASK 0x00001800L +#define MC_VM_MX_L1_TLB_CNTL__ATC_EN_MASK 0x00002000L + + +// addressBlock: gc_ea_gceadec +//GCEA_DRAM_RD_CLI2GRP_MAP0 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID0_GROUP__SHIFT 0x0 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID1_GROUP__SHIFT 0x2 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID2_GROUP__SHIFT 0x4 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID3_GROUP__SHIFT 0x6 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID4_GROUP__SHIFT 0x8 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID5_GROUP__SHIFT 0xa +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID6_GROUP__SHIFT 0xc +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID7_GROUP__SHIFT 0xe +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID8_GROUP__SHIFT 0x10 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID9_GROUP__SHIFT 0x12 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID10_GROUP__SHIFT 0x14 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID11_GROUP__SHIFT 0x16 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID12_GROUP__SHIFT 0x18 +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID13_GROUP__SHIFT 0x1a +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID14_GROUP__SHIFT 0x1c +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID15_GROUP__SHIFT 0x1e +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID0_GROUP_MASK 0x00000003L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID1_GROUP_MASK 0x0000000CL +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID2_GROUP_MASK 0x00000030L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID3_GROUP_MASK 0x000000C0L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID4_GROUP_MASK 0x00000300L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID5_GROUP_MASK 0x00000C00L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID6_GROUP_MASK 0x00003000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID7_GROUP_MASK 0x0000C000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID8_GROUP_MASK 0x00030000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID9_GROUP_MASK 0x000C0000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID10_GROUP_MASK 0x00300000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID11_GROUP_MASK 0x00C00000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID12_GROUP_MASK 0x03000000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID13_GROUP_MASK 0x0C000000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID14_GROUP_MASK 0x30000000L +#define GCEA_DRAM_RD_CLI2GRP_MAP0__CID15_GROUP_MASK 0xC0000000L +//GCEA_DRAM_RD_CLI2GRP_MAP1 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID16_GROUP__SHIFT 0x0 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID17_GROUP__SHIFT 0x2 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID18_GROUP__SHIFT 0x4 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID19_GROUP__SHIFT 0x6 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID20_GROUP__SHIFT 0x8 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID21_GROUP__SHIFT 0xa +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID22_GROUP__SHIFT 0xc +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID23_GROUP__SHIFT 0xe +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID24_GROUP__SHIFT 0x10 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID25_GROUP__SHIFT 0x12 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID26_GROUP__SHIFT 0x14 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID27_GROUP__SHIFT 0x16 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID28_GROUP__SHIFT 0x18 +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID29_GROUP__SHIFT 0x1a +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID30_GROUP__SHIFT 0x1c +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID31_GROUP__SHIFT 0x1e +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID16_GROUP_MASK 0x00000003L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID17_GROUP_MASK 0x0000000CL +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID18_GROUP_MASK 0x00000030L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID19_GROUP_MASK 0x000000C0L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID20_GROUP_MASK 0x00000300L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID21_GROUP_MASK 0x00000C00L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID22_GROUP_MASK 0x00003000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID23_GROUP_MASK 0x0000C000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID24_GROUP_MASK 0x00030000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID25_GROUP_MASK 0x000C0000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID26_GROUP_MASK 0x00300000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID27_GROUP_MASK 0x00C00000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID28_GROUP_MASK 0x03000000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID29_GROUP_MASK 0x0C000000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID30_GROUP_MASK 0x30000000L +#define GCEA_DRAM_RD_CLI2GRP_MAP1__CID31_GROUP_MASK 0xC0000000L +//GCEA_DRAM_WR_CLI2GRP_MAP0 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID0_GROUP__SHIFT 0x0 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID1_GROUP__SHIFT 0x2 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID2_GROUP__SHIFT 0x4 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID3_GROUP__SHIFT 0x6 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID4_GROUP__SHIFT 0x8 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID5_GROUP__SHIFT 0xa +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID6_GROUP__SHIFT 0xc +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID7_GROUP__SHIFT 0xe +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID8_GROUP__SHIFT 0x10 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID9_GROUP__SHIFT 0x12 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID10_GROUP__SHIFT 0x14 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID11_GROUP__SHIFT 0x16 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID12_GROUP__SHIFT 0x18 +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID13_GROUP__SHIFT 0x1a +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID14_GROUP__SHIFT 0x1c +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID15_GROUP__SHIFT 0x1e +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID0_GROUP_MASK 0x00000003L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID1_GROUP_MASK 0x0000000CL +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID2_GROUP_MASK 0x00000030L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID3_GROUP_MASK 0x000000C0L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID4_GROUP_MASK 0x00000300L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID5_GROUP_MASK 0x00000C00L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID6_GROUP_MASK 0x00003000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID7_GROUP_MASK 0x0000C000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID8_GROUP_MASK 0x00030000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID9_GROUP_MASK 0x000C0000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID10_GROUP_MASK 0x00300000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID11_GROUP_MASK 0x00C00000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID12_GROUP_MASK 0x03000000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID13_GROUP_MASK 0x0C000000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID14_GROUP_MASK 0x30000000L +#define GCEA_DRAM_WR_CLI2GRP_MAP0__CID15_GROUP_MASK 0xC0000000L +//GCEA_DRAM_WR_CLI2GRP_MAP1 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID16_GROUP__SHIFT 0x0 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID17_GROUP__SHIFT 0x2 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID18_GROUP__SHIFT 0x4 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID19_GROUP__SHIFT 0x6 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID20_GROUP__SHIFT 0x8 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID21_GROUP__SHIFT 0xa +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID22_GROUP__SHIFT 0xc +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID23_GROUP__SHIFT 0xe +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID24_GROUP__SHIFT 0x10 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID25_GROUP__SHIFT 0x12 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID26_GROUP__SHIFT 0x14 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID27_GROUP__SHIFT 0x16 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID28_GROUP__SHIFT 0x18 +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID29_GROUP__SHIFT 0x1a +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID30_GROUP__SHIFT 0x1c +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID31_GROUP__SHIFT 0x1e +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID16_GROUP_MASK 0x00000003L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID17_GROUP_MASK 0x0000000CL +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID18_GROUP_MASK 0x00000030L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID19_GROUP_MASK 0x000000C0L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID20_GROUP_MASK 0x00000300L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID21_GROUP_MASK 0x00000C00L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID22_GROUP_MASK 0x00003000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID23_GROUP_MASK 0x0000C000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID24_GROUP_MASK 0x00030000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID25_GROUP_MASK 0x000C0000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID26_GROUP_MASK 0x00300000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID27_GROUP_MASK 0x00C00000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID28_GROUP_MASK 0x03000000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID29_GROUP_MASK 0x0C000000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID30_GROUP_MASK 0x30000000L +#define GCEA_DRAM_WR_CLI2GRP_MAP1__CID31_GROUP_MASK 0xC0000000L +//GCEA_DRAM_RD_GRP2VC_MAP +#define GCEA_DRAM_RD_GRP2VC_MAP__GROUP0_VC__SHIFT 0x0 +#define GCEA_DRAM_RD_GRP2VC_MAP__GROUP1_VC__SHIFT 0x3 +#define GCEA_DRAM_RD_GRP2VC_MAP__GROUP2_VC__SHIFT 0x6 +#define GCEA_DRAM_RD_GRP2VC_MAP__GROUP3_VC__SHIFT 0x9 +#define GCEA_DRAM_RD_GRP2VC_MAP__GROUP0_VC_MASK 0x00000007L +#define GCEA_DRAM_RD_GRP2VC_MAP__GROUP1_VC_MASK 0x00000038L +#define GCEA_DRAM_RD_GRP2VC_MAP__GROUP2_VC_MASK 0x000001C0L +#define GCEA_DRAM_RD_GRP2VC_MAP__GROUP3_VC_MASK 0x00000E00L +//GCEA_DRAM_WR_GRP2VC_MAP +#define GCEA_DRAM_WR_GRP2VC_MAP__GROUP0_VC__SHIFT 0x0 +#define GCEA_DRAM_WR_GRP2VC_MAP__GROUP1_VC__SHIFT 0x3 +#define GCEA_DRAM_WR_GRP2VC_MAP__GROUP2_VC__SHIFT 0x6 +#define GCEA_DRAM_WR_GRP2VC_MAP__GROUP3_VC__SHIFT 0x9 +#define GCEA_DRAM_WR_GRP2VC_MAP__GROUP0_VC_MASK 0x00000007L +#define GCEA_DRAM_WR_GRP2VC_MAP__GROUP1_VC_MASK 0x00000038L +#define GCEA_DRAM_WR_GRP2VC_MAP__GROUP2_VC_MASK 0x000001C0L +#define GCEA_DRAM_WR_GRP2VC_MAP__GROUP3_VC_MASK 0x00000E00L +//GCEA_DRAM_RD_LAZY +#define GCEA_DRAM_RD_LAZY__GROUP0_DELAY__SHIFT 0x0 +#define GCEA_DRAM_RD_LAZY__GROUP1_DELAY__SHIFT 0x3 +#define GCEA_DRAM_RD_LAZY__GROUP2_DELAY__SHIFT 0x6 +#define GCEA_DRAM_RD_LAZY__GROUP3_DELAY__SHIFT 0x9 +#define GCEA_DRAM_RD_LAZY__GROUP0_DELAY_MASK 0x00000007L +#define GCEA_DRAM_RD_LAZY__GROUP1_DELAY_MASK 0x00000038L +#define GCEA_DRAM_RD_LAZY__GROUP2_DELAY_MASK 0x000001C0L +#define GCEA_DRAM_RD_LAZY__GROUP3_DELAY_MASK 0x00000E00L +//GCEA_DRAM_WR_LAZY +#define GCEA_DRAM_WR_LAZY__GROUP0_DELAY__SHIFT 0x0 +#define GCEA_DRAM_WR_LAZY__GROUP1_DELAY__SHIFT 0x3 +#define GCEA_DRAM_WR_LAZY__GROUP2_DELAY__SHIFT 0x6 +#define GCEA_DRAM_WR_LAZY__GROUP3_DELAY__SHIFT 0x9 +#define GCEA_DRAM_WR_LAZY__GROUP0_DELAY_MASK 0x00000007L +#define GCEA_DRAM_WR_LAZY__GROUP1_DELAY_MASK 0x00000038L +#define GCEA_DRAM_WR_LAZY__GROUP2_DELAY_MASK 0x000001C0L +#define GCEA_DRAM_WR_LAZY__GROUP3_DELAY_MASK 0x00000E00L +//GCEA_DRAM_RD_CAM_CNTL +#define GCEA_DRAM_RD_CAM_CNTL__DEPTH_GROUP0__SHIFT 0x0 +#define GCEA_DRAM_RD_CAM_CNTL__DEPTH_GROUP1__SHIFT 0x4 +#define GCEA_DRAM_RD_CAM_CNTL__DEPTH_GROUP2__SHIFT 0x8 +#define GCEA_DRAM_RD_CAM_CNTL__DEPTH_GROUP3__SHIFT 0xc +#define GCEA_DRAM_RD_CAM_CNTL__REORDER_LIMIT_GROUP0__SHIFT 0x10 +#define GCEA_DRAM_RD_CAM_CNTL__REORDER_LIMIT_GROUP1__SHIFT 0x13 +#define GCEA_DRAM_RD_CAM_CNTL__REORDER_LIMIT_GROUP2__SHIFT 0x16 +#define GCEA_DRAM_RD_CAM_CNTL__REORDER_LIMIT_GROUP3__SHIFT 0x19 +#define GCEA_DRAM_RD_CAM_CNTL__DEPTH_GROUP0_MASK 0x0000000FL +#define GCEA_DRAM_RD_CAM_CNTL__DEPTH_GROUP1_MASK 0x000000F0L +#define GCEA_DRAM_RD_CAM_CNTL__DEPTH_GROUP2_MASK 0x00000F00L +#define GCEA_DRAM_RD_CAM_CNTL__DEPTH_GROUP3_MASK 0x0000F000L +#define GCEA_DRAM_RD_CAM_CNTL__REORDER_LIMIT_GROUP0_MASK 0x00070000L +#define GCEA_DRAM_RD_CAM_CNTL__REORDER_LIMIT_GROUP1_MASK 0x00380000L +#define GCEA_DRAM_RD_CAM_CNTL__REORDER_LIMIT_GROUP2_MASK 0x01C00000L +#define GCEA_DRAM_RD_CAM_CNTL__REORDER_LIMIT_GROUP3_MASK 0x0E000000L +//GCEA_DRAM_WR_CAM_CNTL +#define GCEA_DRAM_WR_CAM_CNTL__DEPTH_GROUP0__SHIFT 0x0 +#define GCEA_DRAM_WR_CAM_CNTL__DEPTH_GROUP1__SHIFT 0x4 +#define GCEA_DRAM_WR_CAM_CNTL__DEPTH_GROUP2__SHIFT 0x8 +#define GCEA_DRAM_WR_CAM_CNTL__DEPTH_GROUP3__SHIFT 0xc +#define GCEA_DRAM_WR_CAM_CNTL__REORDER_LIMIT_GROUP0__SHIFT 0x10 +#define GCEA_DRAM_WR_CAM_CNTL__REORDER_LIMIT_GROUP1__SHIFT 0x13 +#define GCEA_DRAM_WR_CAM_CNTL__REORDER_LIMIT_GROUP2__SHIFT 0x16 +#define GCEA_DRAM_WR_CAM_CNTL__REORDER_LIMIT_GROUP3__SHIFT 0x19 +#define GCEA_DRAM_WR_CAM_CNTL__DEPTH_GROUP0_MASK 0x0000000FL +#define GCEA_DRAM_WR_CAM_CNTL__DEPTH_GROUP1_MASK 0x000000F0L +#define GCEA_DRAM_WR_CAM_CNTL__DEPTH_GROUP2_MASK 0x00000F00L +#define GCEA_DRAM_WR_CAM_CNTL__DEPTH_GROUP3_MASK 0x0000F000L +#define GCEA_DRAM_WR_CAM_CNTL__REORDER_LIMIT_GROUP0_MASK 0x00070000L +#define GCEA_DRAM_WR_CAM_CNTL__REORDER_LIMIT_GROUP1_MASK 0x00380000L +#define GCEA_DRAM_WR_CAM_CNTL__REORDER_LIMIT_GROUP2_MASK 0x01C00000L +#define GCEA_DRAM_WR_CAM_CNTL__REORDER_LIMIT_GROUP3_MASK 0x0E000000L +//GCEA_DRAM_PAGE_BURST +#define GCEA_DRAM_PAGE_BURST__RD_LIMIT_LO__SHIFT 0x0 +#define GCEA_DRAM_PAGE_BURST__RD_LIMIT_HI__SHIFT 0x8 +#define GCEA_DRAM_PAGE_BURST__WR_LIMIT_LO__SHIFT 0x10 +#define GCEA_DRAM_PAGE_BURST__WR_LIMIT_HI__SHIFT 0x18 +#define GCEA_DRAM_PAGE_BURST__RD_LIMIT_LO_MASK 0x000000FFL +#define GCEA_DRAM_PAGE_BURST__RD_LIMIT_HI_MASK 0x0000FF00L +#define GCEA_DRAM_PAGE_BURST__WR_LIMIT_LO_MASK 0x00FF0000L +#define GCEA_DRAM_PAGE_BURST__WR_LIMIT_HI_MASK 0xFF000000L +//GCEA_DRAM_RD_PRI_AGE +#define GCEA_DRAM_RD_PRI_AGE__GROUP0_AGING_RATE__SHIFT 0x0 +#define GCEA_DRAM_RD_PRI_AGE__GROUP1_AGING_RATE__SHIFT 0x3 +#define GCEA_DRAM_RD_PRI_AGE__GROUP2_AGING_RATE__SHIFT 0x6 +#define GCEA_DRAM_RD_PRI_AGE__GROUP3_AGING_RATE__SHIFT 0x9 +#define GCEA_DRAM_RD_PRI_AGE__GROUP0_AGE_COEFFICIENT__SHIFT 0xc +#define GCEA_DRAM_RD_PRI_AGE__GROUP1_AGE_COEFFICIENT__SHIFT 0xf +#define GCEA_DRAM_RD_PRI_AGE__GROUP2_AGE_COEFFICIENT__SHIFT 0x12 +#define GCEA_DRAM_RD_PRI_AGE__GROUP3_AGE_COEFFICIENT__SHIFT 0x15 +#define GCEA_DRAM_RD_PRI_AGE__GROUP0_AGING_RATE_MASK 0x00000007L +#define GCEA_DRAM_RD_PRI_AGE__GROUP1_AGING_RATE_MASK 0x00000038L +#define GCEA_DRAM_RD_PRI_AGE__GROUP2_AGING_RATE_MASK 0x000001C0L +#define GCEA_DRAM_RD_PRI_AGE__GROUP3_AGING_RATE_MASK 0x00000E00L +#define GCEA_DRAM_RD_PRI_AGE__GROUP0_AGE_COEFFICIENT_MASK 0x00007000L +#define GCEA_DRAM_RD_PRI_AGE__GROUP1_AGE_COEFFICIENT_MASK 0x00038000L +#define GCEA_DRAM_RD_PRI_AGE__GROUP2_AGE_COEFFICIENT_MASK 0x001C0000L +#define GCEA_DRAM_RD_PRI_AGE__GROUP3_AGE_COEFFICIENT_MASK 0x00E00000L +//GCEA_DRAM_WR_PRI_AGE +#define GCEA_DRAM_WR_PRI_AGE__GROUP0_AGING_RATE__SHIFT 0x0 +#define GCEA_DRAM_WR_PRI_AGE__GROUP1_AGING_RATE__SHIFT 0x3 +#define GCEA_DRAM_WR_PRI_AGE__GROUP2_AGING_RATE__SHIFT 0x6 +#define GCEA_DRAM_WR_PRI_AGE__GROUP3_AGING_RATE__SHIFT 0x9 +#define GCEA_DRAM_WR_PRI_AGE__GROUP0_AGE_COEFFICIENT__SHIFT 0xc +#define GCEA_DRAM_WR_PRI_AGE__GROUP1_AGE_COEFFICIENT__SHIFT 0xf +#define GCEA_DRAM_WR_PRI_AGE__GROUP2_AGE_COEFFICIENT__SHIFT 0x12 +#define GCEA_DRAM_WR_PRI_AGE__GROUP3_AGE_COEFFICIENT__SHIFT 0x15 +#define GCEA_DRAM_WR_PRI_AGE__GROUP0_AGING_RATE_MASK 0x00000007L +#define GCEA_DRAM_WR_PRI_AGE__GROUP1_AGING_RATE_MASK 0x00000038L +#define GCEA_DRAM_WR_PRI_AGE__GROUP2_AGING_RATE_MASK 0x000001C0L +#define GCEA_DRAM_WR_PRI_AGE__GROUP3_AGING_RATE_MASK 0x00000E00L +#define GCEA_DRAM_WR_PRI_AGE__GROUP0_AGE_COEFFICIENT_MASK 0x00007000L +#define GCEA_DRAM_WR_PRI_AGE__GROUP1_AGE_COEFFICIENT_MASK 0x00038000L +#define GCEA_DRAM_WR_PRI_AGE__GROUP2_AGE_COEFFICIENT_MASK 0x001C0000L +#define GCEA_DRAM_WR_PRI_AGE__GROUP3_AGE_COEFFICIENT_MASK 0x00E00000L +//GCEA_DRAM_RD_PRI_QUEUING +#define GCEA_DRAM_RD_PRI_QUEUING__GROUP0_QUEUING_COEFFICIENT__SHIFT 0x0 +#define GCEA_DRAM_RD_PRI_QUEUING__GROUP1_QUEUING_COEFFICIENT__SHIFT 0x3 +#define GCEA_DRAM_RD_PRI_QUEUING__GROUP2_QUEUING_COEFFICIENT__SHIFT 0x6 +#define GCEA_DRAM_RD_PRI_QUEUING__GROUP3_QUEUING_COEFFICIENT__SHIFT 0x9 +#define GCEA_DRAM_RD_PRI_QUEUING__GROUP0_QUEUING_COEFFICIENT_MASK 0x00000007L +#define GCEA_DRAM_RD_PRI_QUEUING__GROUP1_QUEUING_COEFFICIENT_MASK 0x00000038L +#define GCEA_DRAM_RD_PRI_QUEUING__GROUP2_QUEUING_COEFFICIENT_MASK 0x000001C0L +#define GCEA_DRAM_RD_PRI_QUEUING__GROUP3_QUEUING_COEFFICIENT_MASK 0x00000E00L +//GCEA_DRAM_WR_PRI_QUEUING +#define GCEA_DRAM_WR_PRI_QUEUING__GROUP0_QUEUING_COEFFICIENT__SHIFT 0x0 +#define GCEA_DRAM_WR_PRI_QUEUING__GROUP1_QUEUING_COEFFICIENT__SHIFT 0x3 +#define GCEA_DRAM_WR_PRI_QUEUING__GROUP2_QUEUING_COEFFICIENT__SHIFT 0x6 +#define GCEA_DRAM_WR_PRI_QUEUING__GROUP3_QUEUING_COEFFICIENT__SHIFT 0x9 +#define GCEA_DRAM_WR_PRI_QUEUING__GROUP0_QUEUING_COEFFICIENT_MASK 0x00000007L +#define GCEA_DRAM_WR_PRI_QUEUING__GROUP1_QUEUING_COEFFICIENT_MASK 0x00000038L +#define GCEA_DRAM_WR_PRI_QUEUING__GROUP2_QUEUING_COEFFICIENT_MASK 0x000001C0L +#define GCEA_DRAM_WR_PRI_QUEUING__GROUP3_QUEUING_COEFFICIENT_MASK 0x00000E00L +//GCEA_DRAM_RD_PRI_FIXED +#define GCEA_DRAM_RD_PRI_FIXED__GROUP0_FIXED_COEFFICIENT__SHIFT 0x0 +#define GCEA_DRAM_RD_PRI_FIXED__GROUP1_FIXED_COEFFICIENT__SHIFT 0x3 +#define GCEA_DRAM_RD_PRI_FIXED__GROUP2_FIXED_COEFFICIENT__SHIFT 0x6 +#define GCEA_DRAM_RD_PRI_FIXED__GROUP3_FIXED_COEFFICIENT__SHIFT 0x9 +#define GCEA_DRAM_RD_PRI_FIXED__GROUP0_FIXED_COEFFICIENT_MASK 0x00000007L +#define GCEA_DRAM_RD_PRI_FIXED__GROUP1_FIXED_COEFFICIENT_MASK 0x00000038L +#define GCEA_DRAM_RD_PRI_FIXED__GROUP2_FIXED_COEFFICIENT_MASK 0x000001C0L +#define GCEA_DRAM_RD_PRI_FIXED__GROUP3_FIXED_COEFFICIENT_MASK 0x00000E00L +//GCEA_DRAM_WR_PRI_FIXED +#define GCEA_DRAM_WR_PRI_FIXED__GROUP0_FIXED_COEFFICIENT__SHIFT 0x0 +#define GCEA_DRAM_WR_PRI_FIXED__GROUP1_FIXED_COEFFICIENT__SHIFT 0x3 +#define GCEA_DRAM_WR_PRI_FIXED__GROUP2_FIXED_COEFFICIENT__SHIFT 0x6 +#define GCEA_DRAM_WR_PRI_FIXED__GROUP3_FIXED_COEFFICIENT__SHIFT 0x9 +#define GCEA_DRAM_WR_PRI_FIXED__GROUP0_FIXED_COEFFICIENT_MASK 0x00000007L +#define GCEA_DRAM_WR_PRI_FIXED__GROUP1_FIXED_COEFFICIENT_MASK 0x00000038L +#define GCEA_DRAM_WR_PRI_FIXED__GROUP2_FIXED_COEFFICIENT_MASK 0x000001C0L +#define GCEA_DRAM_WR_PRI_FIXED__GROUP3_FIXED_COEFFICIENT_MASK 0x00000E00L +//GCEA_DRAM_RD_PRI_URGENCY +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP0_URGENCY_COEFFICIENT__SHIFT 0x0 +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP1_URGENCY_COEFFICIENT__SHIFT 0x3 +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP2_URGENCY_COEFFICIENT__SHIFT 0x6 +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP3_URGENCY_COEFFICIENT__SHIFT 0x9 +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP0_URGENCY_MODE__SHIFT 0xc +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP1_URGENCY_MODE__SHIFT 0xd +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP2_URGENCY_MODE__SHIFT 0xe +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP3_URGENCY_MODE__SHIFT 0xf +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP0_URGENCY_COEFFICIENT_MASK 0x00000007L +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP1_URGENCY_COEFFICIENT_MASK 0x00000038L +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP2_URGENCY_COEFFICIENT_MASK 0x000001C0L +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP3_URGENCY_COEFFICIENT_MASK 0x00000E00L +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP0_URGENCY_MODE_MASK 0x00001000L +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP1_URGENCY_MODE_MASK 0x00002000L +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP2_URGENCY_MODE_MASK 0x00004000L +#define GCEA_DRAM_RD_PRI_URGENCY__GROUP3_URGENCY_MODE_MASK 0x00008000L +//GCEA_DRAM_WR_PRI_URGENCY +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP0_URGENCY_COEFFICIENT__SHIFT 0x0 +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP1_URGENCY_COEFFICIENT__SHIFT 0x3 +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP2_URGENCY_COEFFICIENT__SHIFT 0x6 +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP3_URGENCY_COEFFICIENT__SHIFT 0x9 +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP0_URGENCY_MODE__SHIFT 0xc +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP1_URGENCY_MODE__SHIFT 0xd +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP2_URGENCY_MODE__SHIFT 0xe +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP3_URGENCY_MODE__SHIFT 0xf +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP0_URGENCY_COEFFICIENT_MASK 0x00000007L +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP1_URGENCY_COEFFICIENT_MASK 0x00000038L +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP2_URGENCY_COEFFICIENT_MASK 0x000001C0L +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP3_URGENCY_COEFFICIENT_MASK 0x00000E00L +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP0_URGENCY_MODE_MASK 0x00001000L +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP1_URGENCY_MODE_MASK 0x00002000L +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP2_URGENCY_MODE_MASK 0x00004000L +#define GCEA_DRAM_WR_PRI_URGENCY__GROUP3_URGENCY_MODE_MASK 0x00008000L +//GCEA_DRAM_RD_PRI_QUANT_PRI1 +#define GCEA_DRAM_RD_PRI_QUANT_PRI1__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_DRAM_RD_PRI_QUANT_PRI1__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_DRAM_RD_PRI_QUANT_PRI1__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_DRAM_RD_PRI_QUANT_PRI1__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_DRAM_RD_PRI_QUANT_PRI1__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_DRAM_RD_PRI_QUANT_PRI1__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_DRAM_RD_PRI_QUANT_PRI1__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_DRAM_RD_PRI_QUANT_PRI1__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_DRAM_RD_PRI_QUANT_PRI2 +#define GCEA_DRAM_RD_PRI_QUANT_PRI2__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_DRAM_RD_PRI_QUANT_PRI2__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_DRAM_RD_PRI_QUANT_PRI2__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_DRAM_RD_PRI_QUANT_PRI2__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_DRAM_RD_PRI_QUANT_PRI2__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_DRAM_RD_PRI_QUANT_PRI2__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_DRAM_RD_PRI_QUANT_PRI2__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_DRAM_RD_PRI_QUANT_PRI2__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_DRAM_RD_PRI_QUANT_PRI3 +#define GCEA_DRAM_RD_PRI_QUANT_PRI3__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_DRAM_RD_PRI_QUANT_PRI3__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_DRAM_RD_PRI_QUANT_PRI3__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_DRAM_RD_PRI_QUANT_PRI3__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_DRAM_RD_PRI_QUANT_PRI3__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_DRAM_RD_PRI_QUANT_PRI3__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_DRAM_RD_PRI_QUANT_PRI3__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_DRAM_RD_PRI_QUANT_PRI3__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_DRAM_WR_PRI_QUANT_PRI1 +#define GCEA_DRAM_WR_PRI_QUANT_PRI1__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_DRAM_WR_PRI_QUANT_PRI1__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_DRAM_WR_PRI_QUANT_PRI1__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_DRAM_WR_PRI_QUANT_PRI1__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_DRAM_WR_PRI_QUANT_PRI1__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_DRAM_WR_PRI_QUANT_PRI1__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_DRAM_WR_PRI_QUANT_PRI1__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_DRAM_WR_PRI_QUANT_PRI1__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_DRAM_WR_PRI_QUANT_PRI2 +#define GCEA_DRAM_WR_PRI_QUANT_PRI2__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_DRAM_WR_PRI_QUANT_PRI2__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_DRAM_WR_PRI_QUANT_PRI2__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_DRAM_WR_PRI_QUANT_PRI2__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_DRAM_WR_PRI_QUANT_PRI2__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_DRAM_WR_PRI_QUANT_PRI2__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_DRAM_WR_PRI_QUANT_PRI2__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_DRAM_WR_PRI_QUANT_PRI2__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_DRAM_WR_PRI_QUANT_PRI3 +#define GCEA_DRAM_WR_PRI_QUANT_PRI3__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_DRAM_WR_PRI_QUANT_PRI3__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_DRAM_WR_PRI_QUANT_PRI3__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_DRAM_WR_PRI_QUANT_PRI3__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_DRAM_WR_PRI_QUANT_PRI3__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_DRAM_WR_PRI_QUANT_PRI3__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_DRAM_WR_PRI_QUANT_PRI3__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_DRAM_WR_PRI_QUANT_PRI3__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_ADDRNORM_BASE_ADDR0 +#define GCEA_ADDRNORM_BASE_ADDR0__ADDR_RNG_VAL__SHIFT 0x0 +#define GCEA_ADDRNORM_BASE_ADDR0__LGCY_MMIO_HOLE_EN__SHIFT 0x1 +#define GCEA_ADDRNORM_BASE_ADDR0__INTLV_NUM_CHAN__SHIFT 0x4 +#define GCEA_ADDRNORM_BASE_ADDR0__INTLV_ADDR_SEL__SHIFT 0x8 +#define GCEA_ADDRNORM_BASE_ADDR0__BASE_ADDR__SHIFT 0xc +#define GCEA_ADDRNORM_BASE_ADDR0__ADDR_RNG_VAL_MASK 0x00000001L +#define GCEA_ADDRNORM_BASE_ADDR0__LGCY_MMIO_HOLE_EN_MASK 0x00000002L +#define GCEA_ADDRNORM_BASE_ADDR0__INTLV_NUM_CHAN_MASK 0x000000F0L +#define GCEA_ADDRNORM_BASE_ADDR0__INTLV_ADDR_SEL_MASK 0x00000700L +#define GCEA_ADDRNORM_BASE_ADDR0__BASE_ADDR_MASK 0xFFFFF000L +//GCEA_ADDRNORM_LIMIT_ADDR0 +#define GCEA_ADDRNORM_LIMIT_ADDR0__DST_FABRIC_ID__SHIFT 0x0 +#define GCEA_ADDRNORM_LIMIT_ADDR0__INTLV_NUM_SOCKETS__SHIFT 0x8 +#define GCEA_ADDRNORM_LIMIT_ADDR0__INTLV_NUM_DIES__SHIFT 0xa +#define GCEA_ADDRNORM_LIMIT_ADDR0__LIMIT_ADDR__SHIFT 0xc +#define GCEA_ADDRNORM_LIMIT_ADDR0__DST_FABRIC_ID_MASK 0x0000000FL +#define GCEA_ADDRNORM_LIMIT_ADDR0__INTLV_NUM_SOCKETS_MASK 0x00000100L +#define GCEA_ADDRNORM_LIMIT_ADDR0__INTLV_NUM_DIES_MASK 0x00000C00L +#define GCEA_ADDRNORM_LIMIT_ADDR0__LIMIT_ADDR_MASK 0xFFFFF000L +//GCEA_ADDRNORM_BASE_ADDR1 +#define GCEA_ADDRNORM_BASE_ADDR1__ADDR_RNG_VAL__SHIFT 0x0 +#define GCEA_ADDRNORM_BASE_ADDR1__LGCY_MMIO_HOLE_EN__SHIFT 0x1 +#define GCEA_ADDRNORM_BASE_ADDR1__INTLV_NUM_CHAN__SHIFT 0x4 +#define GCEA_ADDRNORM_BASE_ADDR1__INTLV_ADDR_SEL__SHIFT 0x8 +#define GCEA_ADDRNORM_BASE_ADDR1__BASE_ADDR__SHIFT 0xc +#define GCEA_ADDRNORM_BASE_ADDR1__ADDR_RNG_VAL_MASK 0x00000001L +#define GCEA_ADDRNORM_BASE_ADDR1__LGCY_MMIO_HOLE_EN_MASK 0x00000002L +#define GCEA_ADDRNORM_BASE_ADDR1__INTLV_NUM_CHAN_MASK 0x000000F0L +#define GCEA_ADDRNORM_BASE_ADDR1__INTLV_ADDR_SEL_MASK 0x00000700L +#define GCEA_ADDRNORM_BASE_ADDR1__BASE_ADDR_MASK 0xFFFFF000L +//GCEA_ADDRNORM_LIMIT_ADDR1 +#define GCEA_ADDRNORM_LIMIT_ADDR1__DST_FABRIC_ID__SHIFT 0x0 +#define GCEA_ADDRNORM_LIMIT_ADDR1__INTLV_NUM_SOCKETS__SHIFT 0x8 +#define GCEA_ADDRNORM_LIMIT_ADDR1__INTLV_NUM_DIES__SHIFT 0xa +#define GCEA_ADDRNORM_LIMIT_ADDR1__LIMIT_ADDR__SHIFT 0xc +#define GCEA_ADDRNORM_LIMIT_ADDR1__DST_FABRIC_ID_MASK 0x0000000FL +#define GCEA_ADDRNORM_LIMIT_ADDR1__INTLV_NUM_SOCKETS_MASK 0x00000100L +#define GCEA_ADDRNORM_LIMIT_ADDR1__INTLV_NUM_DIES_MASK 0x00000C00L +#define GCEA_ADDRNORM_LIMIT_ADDR1__LIMIT_ADDR_MASK 0xFFFFF000L +//GCEA_ADDRNORM_OFFSET_ADDR1 +#define GCEA_ADDRNORM_OFFSET_ADDR1__HI_ADDR_OFFSET_EN__SHIFT 0x0 +#define GCEA_ADDRNORM_OFFSET_ADDR1__HI_ADDR_OFFSET__SHIFT 0x14 +#define GCEA_ADDRNORM_OFFSET_ADDR1__HI_ADDR_OFFSET_EN_MASK 0x00000001L +#define GCEA_ADDRNORM_OFFSET_ADDR1__HI_ADDR_OFFSET_MASK 0xFFF00000L +//GCEA_ADDRNORM_HOLE_CNTL +#define GCEA_ADDRNORM_HOLE_CNTL__DRAM_HOLE_VALID__SHIFT 0x0 +#define GCEA_ADDRNORM_HOLE_CNTL__DRAM_HOLE_OFFSET__SHIFT 0x7 +#define GCEA_ADDRNORM_HOLE_CNTL__DRAM_HOLE_VALID_MASK 0x00000001L +#define GCEA_ADDRNORM_HOLE_CNTL__DRAM_HOLE_OFFSET_MASK 0x0000FF80L +//GCEA_ADDRDEC_BANK_CFG +#define GCEA_ADDRDEC_BANK_CFG__BANK_MASK_DRAM__SHIFT 0x0 +#define GCEA_ADDRDEC_BANK_CFG__BANK_MASK_GMI__SHIFT 0x5 +#define GCEA_ADDRDEC_BANK_CFG__BANKGROUP_SEL_DRAM__SHIFT 0xa +#define GCEA_ADDRDEC_BANK_CFG__BANKGROUP_SEL_GMI__SHIFT 0xd +#define GCEA_ADDRDEC_BANK_CFG__BANKGROUP_INTERLEAVE_DRAM__SHIFT 0x10 +#define GCEA_ADDRDEC_BANK_CFG__BANKGROUP_INTERLEAVE_GMI__SHIFT 0x11 +#define GCEA_ADDRDEC_BANK_CFG__BANK_MASK_DRAM_MASK 0x0000001FL +#define GCEA_ADDRDEC_BANK_CFG__BANK_MASK_GMI_MASK 0x000003E0L +#define GCEA_ADDRDEC_BANK_CFG__BANKGROUP_SEL_DRAM_MASK 0x00001C00L +#define GCEA_ADDRDEC_BANK_CFG__BANKGROUP_SEL_GMI_MASK 0x0000E000L +#define GCEA_ADDRDEC_BANK_CFG__BANKGROUP_INTERLEAVE_DRAM_MASK 0x00010000L +#define GCEA_ADDRDEC_BANK_CFG__BANKGROUP_INTERLEAVE_GMI_MASK 0x00020000L +//GCEA_ADDRDEC_MISC_CFG +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN0__SHIFT 0x0 +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN1__SHIFT 0x1 +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN2__SHIFT 0x2 +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN3__SHIFT 0x3 +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN4__SHIFT 0x4 +#define GCEA_ADDRDEC_MISC_CFG__PCH_MASK_DRAM__SHIFT 0x8 +#define GCEA_ADDRDEC_MISC_CFG__PCH_MASK_GMI__SHIFT 0x9 +#define GCEA_ADDRDEC_MISC_CFG__CH_MASK_DRAM__SHIFT 0xc +#define GCEA_ADDRDEC_MISC_CFG__CH_MASK_GMI__SHIFT 0x10 +#define GCEA_ADDRDEC_MISC_CFG__CS_MASK_DRAM__SHIFT 0x14 +#define GCEA_ADDRDEC_MISC_CFG__CS_MASK_GMI__SHIFT 0x16 +#define GCEA_ADDRDEC_MISC_CFG__RM_MASK_DRAM__SHIFT 0x18 +#define GCEA_ADDRDEC_MISC_CFG__RM_MASK_GMI__SHIFT 0x1b +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN0_MASK 0x00000001L +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN1_MASK 0x00000002L +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN2_MASK 0x00000004L +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN3_MASK 0x00000008L +#define GCEA_ADDRDEC_MISC_CFG__VCM_EN4_MASK 0x00000010L +#define GCEA_ADDRDEC_MISC_CFG__PCH_MASK_DRAM_MASK 0x00000100L +#define GCEA_ADDRDEC_MISC_CFG__PCH_MASK_GMI_MASK 0x00000200L +#define GCEA_ADDRDEC_MISC_CFG__CH_MASK_DRAM_MASK 0x0000F000L +#define GCEA_ADDRDEC_MISC_CFG__CH_MASK_GMI_MASK 0x000F0000L +#define GCEA_ADDRDEC_MISC_CFG__CS_MASK_DRAM_MASK 0x00300000L +#define GCEA_ADDRDEC_MISC_CFG__CS_MASK_GMI_MASK 0x00C00000L +#define GCEA_ADDRDEC_MISC_CFG__RM_MASK_DRAM_MASK 0x07000000L +#define GCEA_ADDRDEC_MISC_CFG__RM_MASK_GMI_MASK 0x38000000L +//GCEA_ADDRDECDRAM_ADDR_HASH_BANK0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK0__XOR_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK0__COL_XOR__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK0__ROW_XOR__SHIFT 0xe +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK0__XOR_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK0__COL_XOR_MASK 0x00003FFEL +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK0__ROW_XOR_MASK 0xFFFFC000L +//GCEA_ADDRDECDRAM_ADDR_HASH_BANK1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK1__XOR_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK1__COL_XOR__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK1__ROW_XOR__SHIFT 0xe +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK1__XOR_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK1__COL_XOR_MASK 0x00003FFEL +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK1__ROW_XOR_MASK 0xFFFFC000L +//GCEA_ADDRDECDRAM_ADDR_HASH_BANK2 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK2__XOR_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK2__COL_XOR__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK2__ROW_XOR__SHIFT 0xe +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK2__XOR_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK2__COL_XOR_MASK 0x00003FFEL +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK2__ROW_XOR_MASK 0xFFFFC000L +//GCEA_ADDRDECDRAM_ADDR_HASH_BANK3 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK3__XOR_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK3__COL_XOR__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK3__ROW_XOR__SHIFT 0xe +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK3__XOR_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK3__COL_XOR_MASK 0x00003FFEL +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK3__ROW_XOR_MASK 0xFFFFC000L +//GCEA_ADDRDECDRAM_ADDR_HASH_BANK4 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK4__XOR_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK4__COL_XOR__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK4__ROW_XOR__SHIFT 0xe +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK4__XOR_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK4__COL_XOR_MASK 0x00003FFEL +#define GCEA_ADDRDECDRAM_ADDR_HASH_BANK4__ROW_XOR_MASK 0xFFFFC000L +//GCEA_ADDRDECDRAM_ADDR_HASH_PC +#define GCEA_ADDRDECDRAM_ADDR_HASH_PC__XOR_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_PC__COL_XOR__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_PC__ROW_XOR__SHIFT 0xe +#define GCEA_ADDRDECDRAM_ADDR_HASH_PC__XOR_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_ADDR_HASH_PC__COL_XOR_MASK 0x00003FFEL +#define GCEA_ADDRDECDRAM_ADDR_HASH_PC__ROW_XOR_MASK 0xFFFFC000L +//GCEA_ADDRDECDRAM_ADDR_HASH_PC2 +#define GCEA_ADDRDECDRAM_ADDR_HASH_PC2__BANK_XOR__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_PC2__BANK_XOR_MASK 0x0000001FL +//GCEA_ADDRDECDRAM_ADDR_HASH_CS0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_CS0__XOR_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_CS0__NA_XOR__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_CS0__XOR_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_ADDR_HASH_CS0__NA_XOR_MASK 0xFFFFFFFEL +//GCEA_ADDRDECDRAM_ADDR_HASH_CS1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_CS1__XOR_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_ADDR_HASH_CS1__NA_XOR__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_ADDR_HASH_CS1__XOR_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_ADDR_HASH_CS1__NA_XOR_MASK 0xFFFFFFFEL +//GCEA_ADDRDECDRAM_HARVEST_ENABLE +#define GCEA_ADDRDECDRAM_HARVEST_ENABLE__FORCE_B3_EN__SHIFT 0x0 +#define GCEA_ADDRDECDRAM_HARVEST_ENABLE__FORCE_B3_VAL__SHIFT 0x1 +#define GCEA_ADDRDECDRAM_HARVEST_ENABLE__FORCE_B4_EN__SHIFT 0x2 +#define GCEA_ADDRDECDRAM_HARVEST_ENABLE__FORCE_B4_VAL__SHIFT 0x3 +#define GCEA_ADDRDECDRAM_HARVEST_ENABLE__FORCE_B3_EN_MASK 0x00000001L +#define GCEA_ADDRDECDRAM_HARVEST_ENABLE__FORCE_B3_VAL_MASK 0x00000002L +#define GCEA_ADDRDECDRAM_HARVEST_ENABLE__FORCE_B4_EN_MASK 0x00000004L +#define GCEA_ADDRDECDRAM_HARVEST_ENABLE__FORCE_B4_VAL_MASK 0x00000008L +//GCEA_ADDRDEC0_BASE_ADDR_CS0 +#define GCEA_ADDRDEC0_BASE_ADDR_CS0__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC0_BASE_ADDR_CS0__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC0_BASE_ADDR_CS0__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC0_BASE_ADDR_CS0__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_BASE_ADDR_CS1 +#define GCEA_ADDRDEC0_BASE_ADDR_CS1__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC0_BASE_ADDR_CS1__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC0_BASE_ADDR_CS1__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC0_BASE_ADDR_CS1__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_BASE_ADDR_CS2 +#define GCEA_ADDRDEC0_BASE_ADDR_CS2__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC0_BASE_ADDR_CS2__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC0_BASE_ADDR_CS2__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC0_BASE_ADDR_CS2__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_BASE_ADDR_CS3 +#define GCEA_ADDRDEC0_BASE_ADDR_CS3__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC0_BASE_ADDR_CS3__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC0_BASE_ADDR_CS3__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC0_BASE_ADDR_CS3__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_BASE_ADDR_SECCS0 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS0__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS0__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS0__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS0__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_BASE_ADDR_SECCS1 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS1__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS1__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS1__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS1__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_BASE_ADDR_SECCS2 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS2__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS2__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS2__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS2__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_BASE_ADDR_SECCS3 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS3__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS3__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS3__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC0_BASE_ADDR_SECCS3__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_ADDR_MASK_CS01 +#define GCEA_ADDRDEC0_ADDR_MASK_CS01__ADDR_MASK__SHIFT 0x1 +#define GCEA_ADDRDEC0_ADDR_MASK_CS01__ADDR_MASK_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_ADDR_MASK_CS23 +#define GCEA_ADDRDEC0_ADDR_MASK_CS23__ADDR_MASK__SHIFT 0x1 +#define GCEA_ADDRDEC0_ADDR_MASK_CS23__ADDR_MASK_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_ADDR_MASK_SECCS01 +#define GCEA_ADDRDEC0_ADDR_MASK_SECCS01__ADDR_MASK__SHIFT 0x1 +#define GCEA_ADDRDEC0_ADDR_MASK_SECCS01__ADDR_MASK_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_ADDR_MASK_SECCS23 +#define GCEA_ADDRDEC0_ADDR_MASK_SECCS23__ADDR_MASK__SHIFT 0x1 +#define GCEA_ADDRDEC0_ADDR_MASK_SECCS23__ADDR_MASK_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC0_ADDR_CFG_CS01 +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_BANK_GROUPS__SHIFT 0x2 +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_RM__SHIFT 0x4 +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_ROW_LO__SHIFT 0x8 +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_ROW_HI__SHIFT 0xc +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_COL__SHIFT 0x10 +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_BANKS__SHIFT 0x14 +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_BANK_GROUPS_MASK 0x0000000CL +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_RM_MASK 0x00000030L +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_ROW_LO_MASK 0x00000F00L +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_ROW_HI_MASK 0x0000F000L +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_COL_MASK 0x000F0000L +#define GCEA_ADDRDEC0_ADDR_CFG_CS01__NUM_BANKS_MASK 0x00300000L +//GCEA_ADDRDEC0_ADDR_CFG_CS23 +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_BANK_GROUPS__SHIFT 0x2 +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_RM__SHIFT 0x4 +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_ROW_LO__SHIFT 0x8 +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_ROW_HI__SHIFT 0xc +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_COL__SHIFT 0x10 +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_BANKS__SHIFT 0x14 +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_BANK_GROUPS_MASK 0x0000000CL +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_RM_MASK 0x00000030L +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_ROW_LO_MASK 0x00000F00L +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_ROW_HI_MASK 0x0000F000L +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_COL_MASK 0x000F0000L +#define GCEA_ADDRDEC0_ADDR_CFG_CS23__NUM_BANKS_MASK 0x00300000L +//GCEA_ADDRDEC0_ADDR_SEL_CS01 +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK0__SHIFT 0x0 +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK1__SHIFT 0x4 +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK2__SHIFT 0x8 +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK3__SHIFT 0xc +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK4__SHIFT 0x10 +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__ROW_LO__SHIFT 0x18 +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__ROW_HI__SHIFT 0x1c +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK0_MASK 0x0000000FL +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK1_MASK 0x000000F0L +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK2_MASK 0x00000F00L +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK3_MASK 0x0000F000L +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__BANK4_MASK 0x000F0000L +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__ROW_LO_MASK 0x0F000000L +#define GCEA_ADDRDEC0_ADDR_SEL_CS01__ROW_HI_MASK 0xF0000000L +//GCEA_ADDRDEC0_ADDR_SEL_CS23 +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK0__SHIFT 0x0 +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK1__SHIFT 0x4 +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK2__SHIFT 0x8 +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK3__SHIFT 0xc +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK4__SHIFT 0x10 +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__ROW_LO__SHIFT 0x18 +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__ROW_HI__SHIFT 0x1c +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK0_MASK 0x0000000FL +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK1_MASK 0x000000F0L +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK2_MASK 0x00000F00L +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK3_MASK 0x0000F000L +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__BANK4_MASK 0x000F0000L +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__ROW_LO_MASK 0x0F000000L +#define GCEA_ADDRDEC0_ADDR_SEL_CS23__ROW_HI_MASK 0xF0000000L +//GCEA_ADDRDEC0_COL_SEL_LO_CS01 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL0__SHIFT 0x0 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL1__SHIFT 0x4 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL2__SHIFT 0x8 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL3__SHIFT 0xc +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL4__SHIFT 0x10 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL5__SHIFT 0x14 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL6__SHIFT 0x18 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL7__SHIFT 0x1c +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL0_MASK 0x0000000FL +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL1_MASK 0x000000F0L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL2_MASK 0x00000F00L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL3_MASK 0x0000F000L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL4_MASK 0x000F0000L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL5_MASK 0x00F00000L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL6_MASK 0x0F000000L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS01__COL7_MASK 0xF0000000L +//GCEA_ADDRDEC0_COL_SEL_LO_CS23 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL0__SHIFT 0x0 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL1__SHIFT 0x4 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL2__SHIFT 0x8 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL3__SHIFT 0xc +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL4__SHIFT 0x10 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL5__SHIFT 0x14 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL6__SHIFT 0x18 +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL7__SHIFT 0x1c +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL0_MASK 0x0000000FL +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL1_MASK 0x000000F0L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL2_MASK 0x00000F00L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL3_MASK 0x0000F000L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL4_MASK 0x000F0000L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL5_MASK 0x00F00000L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL6_MASK 0x0F000000L +#define GCEA_ADDRDEC0_COL_SEL_LO_CS23__COL7_MASK 0xF0000000L +//GCEA_ADDRDEC0_COL_SEL_HI_CS01 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL8__SHIFT 0x0 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL9__SHIFT 0x4 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL10__SHIFT 0x8 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL11__SHIFT 0xc +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL12__SHIFT 0x10 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL13__SHIFT 0x14 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL14__SHIFT 0x18 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL15__SHIFT 0x1c +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL8_MASK 0x0000000FL +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL9_MASK 0x000000F0L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL10_MASK 0x00000F00L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL11_MASK 0x0000F000L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL12_MASK 0x000F0000L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL13_MASK 0x00F00000L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL14_MASK 0x0F000000L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS01__COL15_MASK 0xF0000000L +//GCEA_ADDRDEC0_COL_SEL_HI_CS23 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL8__SHIFT 0x0 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL9__SHIFT 0x4 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL10__SHIFT 0x8 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL11__SHIFT 0xc +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL12__SHIFT 0x10 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL13__SHIFT 0x14 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL14__SHIFT 0x18 +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL15__SHIFT 0x1c +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL8_MASK 0x0000000FL +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL9_MASK 0x000000F0L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL10_MASK 0x00000F00L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL11_MASK 0x0000F000L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL12_MASK 0x000F0000L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL13_MASK 0x00F00000L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL14_MASK 0x0F000000L +#define GCEA_ADDRDEC0_COL_SEL_HI_CS23__COL15_MASK 0xF0000000L +//GCEA_ADDRDEC0_RM_SEL_CS01 +#define GCEA_ADDRDEC0_RM_SEL_CS01__RM0__SHIFT 0x0 +#define GCEA_ADDRDEC0_RM_SEL_CS01__RM1__SHIFT 0x4 +#define GCEA_ADDRDEC0_RM_SEL_CS01__RM2__SHIFT 0x8 +#define GCEA_ADDRDEC0_RM_SEL_CS01__CHAN_BIT__SHIFT 0xc +#define GCEA_ADDRDEC0_RM_SEL_CS01__INVERT_ROW_MSBS_EVEN__SHIFT 0x10 +#define GCEA_ADDRDEC0_RM_SEL_CS01__INVERT_ROW_MSBS_ODD__SHIFT 0x12 +#define GCEA_ADDRDEC0_RM_SEL_CS01__RM0_MASK 0x0000000FL +#define GCEA_ADDRDEC0_RM_SEL_CS01__RM1_MASK 0x000000F0L +#define GCEA_ADDRDEC0_RM_SEL_CS01__RM2_MASK 0x00000F00L +#define GCEA_ADDRDEC0_RM_SEL_CS01__CHAN_BIT_MASK 0x0000F000L +#define GCEA_ADDRDEC0_RM_SEL_CS01__INVERT_ROW_MSBS_EVEN_MASK 0x00030000L +#define GCEA_ADDRDEC0_RM_SEL_CS01__INVERT_ROW_MSBS_ODD_MASK 0x000C0000L +//GCEA_ADDRDEC0_RM_SEL_CS23 +#define GCEA_ADDRDEC0_RM_SEL_CS23__RM0__SHIFT 0x0 +#define GCEA_ADDRDEC0_RM_SEL_CS23__RM1__SHIFT 0x4 +#define GCEA_ADDRDEC0_RM_SEL_CS23__RM2__SHIFT 0x8 +#define GCEA_ADDRDEC0_RM_SEL_CS23__CHAN_BIT__SHIFT 0xc +#define GCEA_ADDRDEC0_RM_SEL_CS23__INVERT_ROW_MSBS_EVEN__SHIFT 0x10 +#define GCEA_ADDRDEC0_RM_SEL_CS23__INVERT_ROW_MSBS_ODD__SHIFT 0x12 +#define GCEA_ADDRDEC0_RM_SEL_CS23__RM0_MASK 0x0000000FL +#define GCEA_ADDRDEC0_RM_SEL_CS23__RM1_MASK 0x000000F0L +#define GCEA_ADDRDEC0_RM_SEL_CS23__RM2_MASK 0x00000F00L +#define GCEA_ADDRDEC0_RM_SEL_CS23__CHAN_BIT_MASK 0x0000F000L +#define GCEA_ADDRDEC0_RM_SEL_CS23__INVERT_ROW_MSBS_EVEN_MASK 0x00030000L +#define GCEA_ADDRDEC0_RM_SEL_CS23__INVERT_ROW_MSBS_ODD_MASK 0x000C0000L +//GCEA_ADDRDEC0_RM_SEL_SECCS01 +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__RM0__SHIFT 0x0 +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__RM1__SHIFT 0x4 +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__RM2__SHIFT 0x8 +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__CHAN_BIT__SHIFT 0xc +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__INVERT_ROW_MSBS_EVEN__SHIFT 0x10 +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__INVERT_ROW_MSBS_ODD__SHIFT 0x12 +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__RM0_MASK 0x0000000FL +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__RM1_MASK 0x000000F0L +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__RM2_MASK 0x00000F00L +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__CHAN_BIT_MASK 0x0000F000L +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__INVERT_ROW_MSBS_EVEN_MASK 0x00030000L +#define GCEA_ADDRDEC0_RM_SEL_SECCS01__INVERT_ROW_MSBS_ODD_MASK 0x000C0000L +//GCEA_ADDRDEC0_RM_SEL_SECCS23 +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__RM0__SHIFT 0x0 +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__RM1__SHIFT 0x4 +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__RM2__SHIFT 0x8 +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__CHAN_BIT__SHIFT 0xc +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__INVERT_ROW_MSBS_EVEN__SHIFT 0x10 +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__INVERT_ROW_MSBS_ODD__SHIFT 0x12 +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__RM0_MASK 0x0000000FL +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__RM1_MASK 0x000000F0L +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__RM2_MASK 0x00000F00L +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__CHAN_BIT_MASK 0x0000F000L +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__INVERT_ROW_MSBS_EVEN_MASK 0x00030000L +#define GCEA_ADDRDEC0_RM_SEL_SECCS23__INVERT_ROW_MSBS_ODD_MASK 0x000C0000L +//GCEA_ADDRDEC1_BASE_ADDR_CS0 +#define GCEA_ADDRDEC1_BASE_ADDR_CS0__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC1_BASE_ADDR_CS0__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC1_BASE_ADDR_CS0__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC1_BASE_ADDR_CS0__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_BASE_ADDR_CS1 +#define GCEA_ADDRDEC1_BASE_ADDR_CS1__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC1_BASE_ADDR_CS1__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC1_BASE_ADDR_CS1__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC1_BASE_ADDR_CS1__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_BASE_ADDR_CS2 +#define GCEA_ADDRDEC1_BASE_ADDR_CS2__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC1_BASE_ADDR_CS2__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC1_BASE_ADDR_CS2__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC1_BASE_ADDR_CS2__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_BASE_ADDR_CS3 +#define GCEA_ADDRDEC1_BASE_ADDR_CS3__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC1_BASE_ADDR_CS3__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC1_BASE_ADDR_CS3__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC1_BASE_ADDR_CS3__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_BASE_ADDR_SECCS0 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS0__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS0__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS0__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS0__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_BASE_ADDR_SECCS1 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS1__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS1__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS1__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS1__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_BASE_ADDR_SECCS2 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS2__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS2__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS2__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS2__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_BASE_ADDR_SECCS3 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS3__CS_ENABLE__SHIFT 0x0 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS3__BASE_ADDR__SHIFT 0x1 +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS3__CS_ENABLE_MASK 0x00000001L +#define GCEA_ADDRDEC1_BASE_ADDR_SECCS3__BASE_ADDR_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_ADDR_MASK_CS01 +#define GCEA_ADDRDEC1_ADDR_MASK_CS01__ADDR_MASK__SHIFT 0x1 +#define GCEA_ADDRDEC1_ADDR_MASK_CS01__ADDR_MASK_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_ADDR_MASK_CS23 +#define GCEA_ADDRDEC1_ADDR_MASK_CS23__ADDR_MASK__SHIFT 0x1 +#define GCEA_ADDRDEC1_ADDR_MASK_CS23__ADDR_MASK_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_ADDR_MASK_SECCS01 +#define GCEA_ADDRDEC1_ADDR_MASK_SECCS01__ADDR_MASK__SHIFT 0x1 +#define GCEA_ADDRDEC1_ADDR_MASK_SECCS01__ADDR_MASK_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_ADDR_MASK_SECCS23 +#define GCEA_ADDRDEC1_ADDR_MASK_SECCS23__ADDR_MASK__SHIFT 0x1 +#define GCEA_ADDRDEC1_ADDR_MASK_SECCS23__ADDR_MASK_MASK 0xFFFFFFFEL +//GCEA_ADDRDEC1_ADDR_CFG_CS01 +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_BANK_GROUPS__SHIFT 0x2 +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_RM__SHIFT 0x4 +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_ROW_LO__SHIFT 0x8 +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_ROW_HI__SHIFT 0xc +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_COL__SHIFT 0x10 +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_BANKS__SHIFT 0x14 +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_BANK_GROUPS_MASK 0x0000000CL +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_RM_MASK 0x00000030L +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_ROW_LO_MASK 0x00000F00L +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_ROW_HI_MASK 0x0000F000L +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_COL_MASK 0x000F0000L +#define GCEA_ADDRDEC1_ADDR_CFG_CS01__NUM_BANKS_MASK 0x00300000L +//GCEA_ADDRDEC1_ADDR_CFG_CS23 +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_BANK_GROUPS__SHIFT 0x2 +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_RM__SHIFT 0x4 +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_ROW_LO__SHIFT 0x8 +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_ROW_HI__SHIFT 0xc +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_COL__SHIFT 0x10 +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_BANKS__SHIFT 0x14 +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_BANK_GROUPS_MASK 0x0000000CL +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_RM_MASK 0x00000030L +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_ROW_LO_MASK 0x00000F00L +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_ROW_HI_MASK 0x0000F000L +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_COL_MASK 0x000F0000L +#define GCEA_ADDRDEC1_ADDR_CFG_CS23__NUM_BANKS_MASK 0x00300000L +//GCEA_ADDRDEC1_ADDR_SEL_CS01 +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK0__SHIFT 0x0 +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK1__SHIFT 0x4 +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK2__SHIFT 0x8 +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK3__SHIFT 0xc +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK4__SHIFT 0x10 +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__ROW_LO__SHIFT 0x18 +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__ROW_HI__SHIFT 0x1c +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK0_MASK 0x0000000FL +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK1_MASK 0x000000F0L +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK2_MASK 0x00000F00L +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK3_MASK 0x0000F000L +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__BANK4_MASK 0x000F0000L +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__ROW_LO_MASK 0x0F000000L +#define GCEA_ADDRDEC1_ADDR_SEL_CS01__ROW_HI_MASK 0xF0000000L +//GCEA_ADDRDEC1_ADDR_SEL_CS23 +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK0__SHIFT 0x0 +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK1__SHIFT 0x4 +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK2__SHIFT 0x8 +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK3__SHIFT 0xc +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK4__SHIFT 0x10 +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__ROW_LO__SHIFT 0x18 +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__ROW_HI__SHIFT 0x1c +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK0_MASK 0x0000000FL +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK1_MASK 0x000000F0L +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK2_MASK 0x00000F00L +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK3_MASK 0x0000F000L +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__BANK4_MASK 0x000F0000L +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__ROW_LO_MASK 0x0F000000L +#define GCEA_ADDRDEC1_ADDR_SEL_CS23__ROW_HI_MASK 0xF0000000L +//GCEA_ADDRDEC1_COL_SEL_LO_CS01 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL0__SHIFT 0x0 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL1__SHIFT 0x4 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL2__SHIFT 0x8 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL3__SHIFT 0xc +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL4__SHIFT 0x10 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL5__SHIFT 0x14 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL6__SHIFT 0x18 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL7__SHIFT 0x1c +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL0_MASK 0x0000000FL +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL1_MASK 0x000000F0L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL2_MASK 0x00000F00L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL3_MASK 0x0000F000L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL4_MASK 0x000F0000L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL5_MASK 0x00F00000L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL6_MASK 0x0F000000L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS01__COL7_MASK 0xF0000000L +//GCEA_ADDRDEC1_COL_SEL_LO_CS23 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL0__SHIFT 0x0 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL1__SHIFT 0x4 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL2__SHIFT 0x8 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL3__SHIFT 0xc +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL4__SHIFT 0x10 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL5__SHIFT 0x14 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL6__SHIFT 0x18 +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL7__SHIFT 0x1c +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL0_MASK 0x0000000FL +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL1_MASK 0x000000F0L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL2_MASK 0x00000F00L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL3_MASK 0x0000F000L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL4_MASK 0x000F0000L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL5_MASK 0x00F00000L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL6_MASK 0x0F000000L +#define GCEA_ADDRDEC1_COL_SEL_LO_CS23__COL7_MASK 0xF0000000L +//GCEA_ADDRDEC1_COL_SEL_HI_CS01 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL8__SHIFT 0x0 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL9__SHIFT 0x4 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL10__SHIFT 0x8 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL11__SHIFT 0xc +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL12__SHIFT 0x10 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL13__SHIFT 0x14 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL14__SHIFT 0x18 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL15__SHIFT 0x1c +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL8_MASK 0x0000000FL +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL9_MASK 0x000000F0L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL10_MASK 0x00000F00L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL11_MASK 0x0000F000L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL12_MASK 0x000F0000L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL13_MASK 0x00F00000L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL14_MASK 0x0F000000L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS01__COL15_MASK 0xF0000000L +//GCEA_ADDRDEC1_COL_SEL_HI_CS23 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL8__SHIFT 0x0 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL9__SHIFT 0x4 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL10__SHIFT 0x8 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL11__SHIFT 0xc +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL12__SHIFT 0x10 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL13__SHIFT 0x14 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL14__SHIFT 0x18 +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL15__SHIFT 0x1c +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL8_MASK 0x0000000FL +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL9_MASK 0x000000F0L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL10_MASK 0x00000F00L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL11_MASK 0x0000F000L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL12_MASK 0x000F0000L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL13_MASK 0x00F00000L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL14_MASK 0x0F000000L +#define GCEA_ADDRDEC1_COL_SEL_HI_CS23__COL15_MASK 0xF0000000L +//GCEA_ADDRDEC1_RM_SEL_CS01 +#define GCEA_ADDRDEC1_RM_SEL_CS01__RM0__SHIFT 0x0 +#define GCEA_ADDRDEC1_RM_SEL_CS01__RM1__SHIFT 0x4 +#define GCEA_ADDRDEC1_RM_SEL_CS01__RM2__SHIFT 0x8 +#define GCEA_ADDRDEC1_RM_SEL_CS01__CHAN_BIT__SHIFT 0xc +#define GCEA_ADDRDEC1_RM_SEL_CS01__INVERT_ROW_MSBS_EVEN__SHIFT 0x10 +#define GCEA_ADDRDEC1_RM_SEL_CS01__INVERT_ROW_MSBS_ODD__SHIFT 0x12 +#define GCEA_ADDRDEC1_RM_SEL_CS01__RM0_MASK 0x0000000FL +#define GCEA_ADDRDEC1_RM_SEL_CS01__RM1_MASK 0x000000F0L +#define GCEA_ADDRDEC1_RM_SEL_CS01__RM2_MASK 0x00000F00L +#define GCEA_ADDRDEC1_RM_SEL_CS01__CHAN_BIT_MASK 0x0000F000L +#define GCEA_ADDRDEC1_RM_SEL_CS01__INVERT_ROW_MSBS_EVEN_MASK 0x00030000L +#define GCEA_ADDRDEC1_RM_SEL_CS01__INVERT_ROW_MSBS_ODD_MASK 0x000C0000L +//GCEA_ADDRDEC1_RM_SEL_CS23 +#define GCEA_ADDRDEC1_RM_SEL_CS23__RM0__SHIFT 0x0 +#define GCEA_ADDRDEC1_RM_SEL_CS23__RM1__SHIFT 0x4 +#define GCEA_ADDRDEC1_RM_SEL_CS23__RM2__SHIFT 0x8 +#define GCEA_ADDRDEC1_RM_SEL_CS23__CHAN_BIT__SHIFT 0xc +#define GCEA_ADDRDEC1_RM_SEL_CS23__INVERT_ROW_MSBS_EVEN__SHIFT 0x10 +#define GCEA_ADDRDEC1_RM_SEL_CS23__INVERT_ROW_MSBS_ODD__SHIFT 0x12 +#define GCEA_ADDRDEC1_RM_SEL_CS23__RM0_MASK 0x0000000FL +#define GCEA_ADDRDEC1_RM_SEL_CS23__RM1_MASK 0x000000F0L +#define GCEA_ADDRDEC1_RM_SEL_CS23__RM2_MASK 0x00000F00L +#define GCEA_ADDRDEC1_RM_SEL_CS23__CHAN_BIT_MASK 0x0000F000L +#define GCEA_ADDRDEC1_RM_SEL_CS23__INVERT_ROW_MSBS_EVEN_MASK 0x00030000L +#define GCEA_ADDRDEC1_RM_SEL_CS23__INVERT_ROW_MSBS_ODD_MASK 0x000C0000L +//GCEA_ADDRDEC1_RM_SEL_SECCS01 +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__RM0__SHIFT 0x0 +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__RM1__SHIFT 0x4 +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__RM2__SHIFT 0x8 +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__CHAN_BIT__SHIFT 0xc +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__INVERT_ROW_MSBS_EVEN__SHIFT 0x10 +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__INVERT_ROW_MSBS_ODD__SHIFT 0x12 +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__RM0_MASK 0x0000000FL +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__RM1_MASK 0x000000F0L +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__RM2_MASK 0x00000F00L +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__CHAN_BIT_MASK 0x0000F000L +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__INVERT_ROW_MSBS_EVEN_MASK 0x00030000L +#define GCEA_ADDRDEC1_RM_SEL_SECCS01__INVERT_ROW_MSBS_ODD_MASK 0x000C0000L +//GCEA_ADDRDEC1_RM_SEL_SECCS23 +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__RM0__SHIFT 0x0 +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__RM1__SHIFT 0x4 +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__RM2__SHIFT 0x8 +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__CHAN_BIT__SHIFT 0xc +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__INVERT_ROW_MSBS_EVEN__SHIFT 0x10 +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__INVERT_ROW_MSBS_ODD__SHIFT 0x12 +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__RM0_MASK 0x0000000FL +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__RM1_MASK 0x000000F0L +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__RM2_MASK 0x00000F00L +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__CHAN_BIT_MASK 0x0000F000L +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__INVERT_ROW_MSBS_EVEN_MASK 0x00030000L +#define GCEA_ADDRDEC1_RM_SEL_SECCS23__INVERT_ROW_MSBS_ODD_MASK 0x000C0000L +//GCEA_IO_RD_CLI2GRP_MAP0 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID0_GROUP__SHIFT 0x0 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID1_GROUP__SHIFT 0x2 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID2_GROUP__SHIFT 0x4 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID3_GROUP__SHIFT 0x6 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID4_GROUP__SHIFT 0x8 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID5_GROUP__SHIFT 0xa +#define GCEA_IO_RD_CLI2GRP_MAP0__CID6_GROUP__SHIFT 0xc +#define GCEA_IO_RD_CLI2GRP_MAP0__CID7_GROUP__SHIFT 0xe +#define GCEA_IO_RD_CLI2GRP_MAP0__CID8_GROUP__SHIFT 0x10 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID9_GROUP__SHIFT 0x12 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID10_GROUP__SHIFT 0x14 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID11_GROUP__SHIFT 0x16 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID12_GROUP__SHIFT 0x18 +#define GCEA_IO_RD_CLI2GRP_MAP0__CID13_GROUP__SHIFT 0x1a +#define GCEA_IO_RD_CLI2GRP_MAP0__CID14_GROUP__SHIFT 0x1c +#define GCEA_IO_RD_CLI2GRP_MAP0__CID15_GROUP__SHIFT 0x1e +#define GCEA_IO_RD_CLI2GRP_MAP0__CID0_GROUP_MASK 0x00000003L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID1_GROUP_MASK 0x0000000CL +#define GCEA_IO_RD_CLI2GRP_MAP0__CID2_GROUP_MASK 0x00000030L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID3_GROUP_MASK 0x000000C0L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID4_GROUP_MASK 0x00000300L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID5_GROUP_MASK 0x00000C00L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID6_GROUP_MASK 0x00003000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID7_GROUP_MASK 0x0000C000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID8_GROUP_MASK 0x00030000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID9_GROUP_MASK 0x000C0000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID10_GROUP_MASK 0x00300000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID11_GROUP_MASK 0x00C00000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID12_GROUP_MASK 0x03000000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID13_GROUP_MASK 0x0C000000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID14_GROUP_MASK 0x30000000L +#define GCEA_IO_RD_CLI2GRP_MAP0__CID15_GROUP_MASK 0xC0000000L +//GCEA_IO_RD_CLI2GRP_MAP1 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID16_GROUP__SHIFT 0x0 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID17_GROUP__SHIFT 0x2 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID18_GROUP__SHIFT 0x4 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID19_GROUP__SHIFT 0x6 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID20_GROUP__SHIFT 0x8 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID21_GROUP__SHIFT 0xa +#define GCEA_IO_RD_CLI2GRP_MAP1__CID22_GROUP__SHIFT 0xc +#define GCEA_IO_RD_CLI2GRP_MAP1__CID23_GROUP__SHIFT 0xe +#define GCEA_IO_RD_CLI2GRP_MAP1__CID24_GROUP__SHIFT 0x10 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID25_GROUP__SHIFT 0x12 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID26_GROUP__SHIFT 0x14 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID27_GROUP__SHIFT 0x16 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID28_GROUP__SHIFT 0x18 +#define GCEA_IO_RD_CLI2GRP_MAP1__CID29_GROUP__SHIFT 0x1a +#define GCEA_IO_RD_CLI2GRP_MAP1__CID30_GROUP__SHIFT 0x1c +#define GCEA_IO_RD_CLI2GRP_MAP1__CID31_GROUP__SHIFT 0x1e +#define GCEA_IO_RD_CLI2GRP_MAP1__CID16_GROUP_MASK 0x00000003L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID17_GROUP_MASK 0x0000000CL +#define GCEA_IO_RD_CLI2GRP_MAP1__CID18_GROUP_MASK 0x00000030L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID19_GROUP_MASK 0x000000C0L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID20_GROUP_MASK 0x00000300L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID21_GROUP_MASK 0x00000C00L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID22_GROUP_MASK 0x00003000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID23_GROUP_MASK 0x0000C000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID24_GROUP_MASK 0x00030000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID25_GROUP_MASK 0x000C0000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID26_GROUP_MASK 0x00300000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID27_GROUP_MASK 0x00C00000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID28_GROUP_MASK 0x03000000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID29_GROUP_MASK 0x0C000000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID30_GROUP_MASK 0x30000000L +#define GCEA_IO_RD_CLI2GRP_MAP1__CID31_GROUP_MASK 0xC0000000L +//GCEA_IO_WR_CLI2GRP_MAP0 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID0_GROUP__SHIFT 0x0 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID1_GROUP__SHIFT 0x2 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID2_GROUP__SHIFT 0x4 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID3_GROUP__SHIFT 0x6 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID4_GROUP__SHIFT 0x8 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID5_GROUP__SHIFT 0xa +#define GCEA_IO_WR_CLI2GRP_MAP0__CID6_GROUP__SHIFT 0xc +#define GCEA_IO_WR_CLI2GRP_MAP0__CID7_GROUP__SHIFT 0xe +#define GCEA_IO_WR_CLI2GRP_MAP0__CID8_GROUP__SHIFT 0x10 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID9_GROUP__SHIFT 0x12 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID10_GROUP__SHIFT 0x14 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID11_GROUP__SHIFT 0x16 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID12_GROUP__SHIFT 0x18 +#define GCEA_IO_WR_CLI2GRP_MAP0__CID13_GROUP__SHIFT 0x1a +#define GCEA_IO_WR_CLI2GRP_MAP0__CID14_GROUP__SHIFT 0x1c +#define GCEA_IO_WR_CLI2GRP_MAP0__CID15_GROUP__SHIFT 0x1e +#define GCEA_IO_WR_CLI2GRP_MAP0__CID0_GROUP_MASK 0x00000003L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID1_GROUP_MASK 0x0000000CL +#define GCEA_IO_WR_CLI2GRP_MAP0__CID2_GROUP_MASK 0x00000030L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID3_GROUP_MASK 0x000000C0L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID4_GROUP_MASK 0x00000300L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID5_GROUP_MASK 0x00000C00L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID6_GROUP_MASK 0x00003000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID7_GROUP_MASK 0x0000C000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID8_GROUP_MASK 0x00030000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID9_GROUP_MASK 0x000C0000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID10_GROUP_MASK 0x00300000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID11_GROUP_MASK 0x00C00000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID12_GROUP_MASK 0x03000000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID13_GROUP_MASK 0x0C000000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID14_GROUP_MASK 0x30000000L +#define GCEA_IO_WR_CLI2GRP_MAP0__CID15_GROUP_MASK 0xC0000000L +//GCEA_IO_WR_CLI2GRP_MAP1 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID16_GROUP__SHIFT 0x0 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID17_GROUP__SHIFT 0x2 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID18_GROUP__SHIFT 0x4 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID19_GROUP__SHIFT 0x6 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID20_GROUP__SHIFT 0x8 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID21_GROUP__SHIFT 0xa +#define GCEA_IO_WR_CLI2GRP_MAP1__CID22_GROUP__SHIFT 0xc +#define GCEA_IO_WR_CLI2GRP_MAP1__CID23_GROUP__SHIFT 0xe +#define GCEA_IO_WR_CLI2GRP_MAP1__CID24_GROUP__SHIFT 0x10 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID25_GROUP__SHIFT 0x12 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID26_GROUP__SHIFT 0x14 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID27_GROUP__SHIFT 0x16 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID28_GROUP__SHIFT 0x18 +#define GCEA_IO_WR_CLI2GRP_MAP1__CID29_GROUP__SHIFT 0x1a +#define GCEA_IO_WR_CLI2GRP_MAP1__CID30_GROUP__SHIFT 0x1c +#define GCEA_IO_WR_CLI2GRP_MAP1__CID31_GROUP__SHIFT 0x1e +#define GCEA_IO_WR_CLI2GRP_MAP1__CID16_GROUP_MASK 0x00000003L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID17_GROUP_MASK 0x0000000CL +#define GCEA_IO_WR_CLI2GRP_MAP1__CID18_GROUP_MASK 0x00000030L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID19_GROUP_MASK 0x000000C0L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID20_GROUP_MASK 0x00000300L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID21_GROUP_MASK 0x00000C00L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID22_GROUP_MASK 0x00003000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID23_GROUP_MASK 0x0000C000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID24_GROUP_MASK 0x00030000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID25_GROUP_MASK 0x000C0000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID26_GROUP_MASK 0x00300000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID27_GROUP_MASK 0x00C00000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID28_GROUP_MASK 0x03000000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID29_GROUP_MASK 0x0C000000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID30_GROUP_MASK 0x30000000L +#define GCEA_IO_WR_CLI2GRP_MAP1__CID31_GROUP_MASK 0xC0000000L +//GCEA_IO_RD_COMBINE_FLUSH +#define GCEA_IO_RD_COMBINE_FLUSH__GROUP0_TIMER__SHIFT 0x0 +#define GCEA_IO_RD_COMBINE_FLUSH__GROUP1_TIMER__SHIFT 0x4 +#define GCEA_IO_RD_COMBINE_FLUSH__GROUP2_TIMER__SHIFT 0x8 +#define GCEA_IO_RD_COMBINE_FLUSH__GROUP3_TIMER__SHIFT 0xc +#define GCEA_IO_RD_COMBINE_FLUSH__GROUP0_TIMER_MASK 0x0000000FL +#define GCEA_IO_RD_COMBINE_FLUSH__GROUP1_TIMER_MASK 0x000000F0L +#define GCEA_IO_RD_COMBINE_FLUSH__GROUP2_TIMER_MASK 0x00000F00L +#define GCEA_IO_RD_COMBINE_FLUSH__GROUP3_TIMER_MASK 0x0000F000L +//GCEA_IO_WR_COMBINE_FLUSH +#define GCEA_IO_WR_COMBINE_FLUSH__GROUP0_TIMER__SHIFT 0x0 +#define GCEA_IO_WR_COMBINE_FLUSH__GROUP1_TIMER__SHIFT 0x4 +#define GCEA_IO_WR_COMBINE_FLUSH__GROUP2_TIMER__SHIFT 0x8 +#define GCEA_IO_WR_COMBINE_FLUSH__GROUP3_TIMER__SHIFT 0xc +#define GCEA_IO_WR_COMBINE_FLUSH__GROUP0_TIMER_MASK 0x0000000FL +#define GCEA_IO_WR_COMBINE_FLUSH__GROUP1_TIMER_MASK 0x000000F0L +#define GCEA_IO_WR_COMBINE_FLUSH__GROUP2_TIMER_MASK 0x00000F00L +#define GCEA_IO_WR_COMBINE_FLUSH__GROUP3_TIMER_MASK 0x0000F000L +//GCEA_IO_GROUP_BURST +#define GCEA_IO_GROUP_BURST__RD_LIMIT_LO__SHIFT 0x0 +#define GCEA_IO_GROUP_BURST__RD_LIMIT_HI__SHIFT 0x8 +#define GCEA_IO_GROUP_BURST__WR_LIMIT_LO__SHIFT 0x10 +#define GCEA_IO_GROUP_BURST__WR_LIMIT_HI__SHIFT 0x18 +#define GCEA_IO_GROUP_BURST__RD_LIMIT_LO_MASK 0x000000FFL +#define GCEA_IO_GROUP_BURST__RD_LIMIT_HI_MASK 0x0000FF00L +#define GCEA_IO_GROUP_BURST__WR_LIMIT_LO_MASK 0x00FF0000L +#define GCEA_IO_GROUP_BURST__WR_LIMIT_HI_MASK 0xFF000000L +//GCEA_IO_RD_PRI_AGE +#define GCEA_IO_RD_PRI_AGE__GROUP0_AGING_RATE__SHIFT 0x0 +#define GCEA_IO_RD_PRI_AGE__GROUP1_AGING_RATE__SHIFT 0x3 +#define GCEA_IO_RD_PRI_AGE__GROUP2_AGING_RATE__SHIFT 0x6 +#define GCEA_IO_RD_PRI_AGE__GROUP3_AGING_RATE__SHIFT 0x9 +#define GCEA_IO_RD_PRI_AGE__GROUP0_AGE_COEFFICIENT__SHIFT 0xc +#define GCEA_IO_RD_PRI_AGE__GROUP1_AGE_COEFFICIENT__SHIFT 0xf +#define GCEA_IO_RD_PRI_AGE__GROUP2_AGE_COEFFICIENT__SHIFT 0x12 +#define GCEA_IO_RD_PRI_AGE__GROUP3_AGE_COEFFICIENT__SHIFT 0x15 +#define GCEA_IO_RD_PRI_AGE__GROUP0_AGING_RATE_MASK 0x00000007L +#define GCEA_IO_RD_PRI_AGE__GROUP1_AGING_RATE_MASK 0x00000038L +#define GCEA_IO_RD_PRI_AGE__GROUP2_AGING_RATE_MASK 0x000001C0L +#define GCEA_IO_RD_PRI_AGE__GROUP3_AGING_RATE_MASK 0x00000E00L +#define GCEA_IO_RD_PRI_AGE__GROUP0_AGE_COEFFICIENT_MASK 0x00007000L +#define GCEA_IO_RD_PRI_AGE__GROUP1_AGE_COEFFICIENT_MASK 0x00038000L +#define GCEA_IO_RD_PRI_AGE__GROUP2_AGE_COEFFICIENT_MASK 0x001C0000L +#define GCEA_IO_RD_PRI_AGE__GROUP3_AGE_COEFFICIENT_MASK 0x00E00000L +//GCEA_IO_WR_PRI_AGE +#define GCEA_IO_WR_PRI_AGE__GROUP0_AGING_RATE__SHIFT 0x0 +#define GCEA_IO_WR_PRI_AGE__GROUP1_AGING_RATE__SHIFT 0x3 +#define GCEA_IO_WR_PRI_AGE__GROUP2_AGING_RATE__SHIFT 0x6 +#define GCEA_IO_WR_PRI_AGE__GROUP3_AGING_RATE__SHIFT 0x9 +#define GCEA_IO_WR_PRI_AGE__GROUP0_AGE_COEFFICIENT__SHIFT 0xc +#define GCEA_IO_WR_PRI_AGE__GROUP1_AGE_COEFFICIENT__SHIFT 0xf +#define GCEA_IO_WR_PRI_AGE__GROUP2_AGE_COEFFICIENT__SHIFT 0x12 +#define GCEA_IO_WR_PRI_AGE__GROUP3_AGE_COEFFICIENT__SHIFT 0x15 +#define GCEA_IO_WR_PRI_AGE__GROUP0_AGING_RATE_MASK 0x00000007L +#define GCEA_IO_WR_PRI_AGE__GROUP1_AGING_RATE_MASK 0x00000038L +#define GCEA_IO_WR_PRI_AGE__GROUP2_AGING_RATE_MASK 0x000001C0L +#define GCEA_IO_WR_PRI_AGE__GROUP3_AGING_RATE_MASK 0x00000E00L +#define GCEA_IO_WR_PRI_AGE__GROUP0_AGE_COEFFICIENT_MASK 0x00007000L +#define GCEA_IO_WR_PRI_AGE__GROUP1_AGE_COEFFICIENT_MASK 0x00038000L +#define GCEA_IO_WR_PRI_AGE__GROUP2_AGE_COEFFICIENT_MASK 0x001C0000L +#define GCEA_IO_WR_PRI_AGE__GROUP3_AGE_COEFFICIENT_MASK 0x00E00000L +//GCEA_IO_RD_PRI_QUEUING +#define GCEA_IO_RD_PRI_QUEUING__GROUP0_QUEUING_COEFFICIENT__SHIFT 0x0 +#define GCEA_IO_RD_PRI_QUEUING__GROUP1_QUEUING_COEFFICIENT__SHIFT 0x3 +#define GCEA_IO_RD_PRI_QUEUING__GROUP2_QUEUING_COEFFICIENT__SHIFT 0x6 +#define GCEA_IO_RD_PRI_QUEUING__GROUP3_QUEUING_COEFFICIENT__SHIFT 0x9 +#define GCEA_IO_RD_PRI_QUEUING__GROUP0_QUEUING_COEFFICIENT_MASK 0x00000007L +#define GCEA_IO_RD_PRI_QUEUING__GROUP1_QUEUING_COEFFICIENT_MASK 0x00000038L +#define GCEA_IO_RD_PRI_QUEUING__GROUP2_QUEUING_COEFFICIENT_MASK 0x000001C0L +#define GCEA_IO_RD_PRI_QUEUING__GROUP3_QUEUING_COEFFICIENT_MASK 0x00000E00L +//GCEA_IO_WR_PRI_QUEUING +#define GCEA_IO_WR_PRI_QUEUING__GROUP0_QUEUING_COEFFICIENT__SHIFT 0x0 +#define GCEA_IO_WR_PRI_QUEUING__GROUP1_QUEUING_COEFFICIENT__SHIFT 0x3 +#define GCEA_IO_WR_PRI_QUEUING__GROUP2_QUEUING_COEFFICIENT__SHIFT 0x6 +#define GCEA_IO_WR_PRI_QUEUING__GROUP3_QUEUING_COEFFICIENT__SHIFT 0x9 +#define GCEA_IO_WR_PRI_QUEUING__GROUP0_QUEUING_COEFFICIENT_MASK 0x00000007L +#define GCEA_IO_WR_PRI_QUEUING__GROUP1_QUEUING_COEFFICIENT_MASK 0x00000038L +#define GCEA_IO_WR_PRI_QUEUING__GROUP2_QUEUING_COEFFICIENT_MASK 0x000001C0L +#define GCEA_IO_WR_PRI_QUEUING__GROUP3_QUEUING_COEFFICIENT_MASK 0x00000E00L +//GCEA_IO_RD_PRI_FIXED +#define GCEA_IO_RD_PRI_FIXED__GROUP0_FIXED_COEFFICIENT__SHIFT 0x0 +#define GCEA_IO_RD_PRI_FIXED__GROUP1_FIXED_COEFFICIENT__SHIFT 0x3 +#define GCEA_IO_RD_PRI_FIXED__GROUP2_FIXED_COEFFICIENT__SHIFT 0x6 +#define GCEA_IO_RD_PRI_FIXED__GROUP3_FIXED_COEFFICIENT__SHIFT 0x9 +#define GCEA_IO_RD_PRI_FIXED__GROUP0_FIXED_COEFFICIENT_MASK 0x00000007L +#define GCEA_IO_RD_PRI_FIXED__GROUP1_FIXED_COEFFICIENT_MASK 0x00000038L +#define GCEA_IO_RD_PRI_FIXED__GROUP2_FIXED_COEFFICIENT_MASK 0x000001C0L +#define GCEA_IO_RD_PRI_FIXED__GROUP3_FIXED_COEFFICIENT_MASK 0x00000E00L +//GCEA_IO_WR_PRI_FIXED +#define GCEA_IO_WR_PRI_FIXED__GROUP0_FIXED_COEFFICIENT__SHIFT 0x0 +#define GCEA_IO_WR_PRI_FIXED__GROUP1_FIXED_COEFFICIENT__SHIFT 0x3 +#define GCEA_IO_WR_PRI_FIXED__GROUP2_FIXED_COEFFICIENT__SHIFT 0x6 +#define GCEA_IO_WR_PRI_FIXED__GROUP3_FIXED_COEFFICIENT__SHIFT 0x9 +#define GCEA_IO_WR_PRI_FIXED__GROUP0_FIXED_COEFFICIENT_MASK 0x00000007L +#define GCEA_IO_WR_PRI_FIXED__GROUP1_FIXED_COEFFICIENT_MASK 0x00000038L +#define GCEA_IO_WR_PRI_FIXED__GROUP2_FIXED_COEFFICIENT_MASK 0x000001C0L +#define GCEA_IO_WR_PRI_FIXED__GROUP3_FIXED_COEFFICIENT_MASK 0x00000E00L +//GCEA_IO_RD_PRI_URGENCY +#define GCEA_IO_RD_PRI_URGENCY__GROUP0_URGENCY_COEFFICIENT__SHIFT 0x0 +#define GCEA_IO_RD_PRI_URGENCY__GROUP1_URGENCY_COEFFICIENT__SHIFT 0x3 +#define GCEA_IO_RD_PRI_URGENCY__GROUP2_URGENCY_COEFFICIENT__SHIFT 0x6 +#define GCEA_IO_RD_PRI_URGENCY__GROUP3_URGENCY_COEFFICIENT__SHIFT 0x9 +#define GCEA_IO_RD_PRI_URGENCY__GROUP0_URGENCY_MODE__SHIFT 0xc +#define GCEA_IO_RD_PRI_URGENCY__GROUP1_URGENCY_MODE__SHIFT 0xd +#define GCEA_IO_RD_PRI_URGENCY__GROUP2_URGENCY_MODE__SHIFT 0xe +#define GCEA_IO_RD_PRI_URGENCY__GROUP3_URGENCY_MODE__SHIFT 0xf +#define GCEA_IO_RD_PRI_URGENCY__GROUP0_URGENCY_COEFFICIENT_MASK 0x00000007L +#define GCEA_IO_RD_PRI_URGENCY__GROUP1_URGENCY_COEFFICIENT_MASK 0x00000038L +#define GCEA_IO_RD_PRI_URGENCY__GROUP2_URGENCY_COEFFICIENT_MASK 0x000001C0L +#define GCEA_IO_RD_PRI_URGENCY__GROUP3_URGENCY_COEFFICIENT_MASK 0x00000E00L +#define GCEA_IO_RD_PRI_URGENCY__GROUP0_URGENCY_MODE_MASK 0x00001000L +#define GCEA_IO_RD_PRI_URGENCY__GROUP1_URGENCY_MODE_MASK 0x00002000L +#define GCEA_IO_RD_PRI_URGENCY__GROUP2_URGENCY_MODE_MASK 0x00004000L +#define GCEA_IO_RD_PRI_URGENCY__GROUP3_URGENCY_MODE_MASK 0x00008000L +//GCEA_IO_WR_PRI_URGENCY +#define GCEA_IO_WR_PRI_URGENCY__GROUP0_URGENCY_COEFFICIENT__SHIFT 0x0 +#define GCEA_IO_WR_PRI_URGENCY__GROUP1_URGENCY_COEFFICIENT__SHIFT 0x3 +#define GCEA_IO_WR_PRI_URGENCY__GROUP2_URGENCY_COEFFICIENT__SHIFT 0x6 +#define GCEA_IO_WR_PRI_URGENCY__GROUP3_URGENCY_COEFFICIENT__SHIFT 0x9 +#define GCEA_IO_WR_PRI_URGENCY__GROUP0_URGENCY_MODE__SHIFT 0xc +#define GCEA_IO_WR_PRI_URGENCY__GROUP1_URGENCY_MODE__SHIFT 0xd +#define GCEA_IO_WR_PRI_URGENCY__GROUP2_URGENCY_MODE__SHIFT 0xe +#define GCEA_IO_WR_PRI_URGENCY__GROUP3_URGENCY_MODE__SHIFT 0xf +#define GCEA_IO_WR_PRI_URGENCY__GROUP0_URGENCY_COEFFICIENT_MASK 0x00000007L +#define GCEA_IO_WR_PRI_URGENCY__GROUP1_URGENCY_COEFFICIENT_MASK 0x00000038L +#define GCEA_IO_WR_PRI_URGENCY__GROUP2_URGENCY_COEFFICIENT_MASK 0x000001C0L +#define GCEA_IO_WR_PRI_URGENCY__GROUP3_URGENCY_COEFFICIENT_MASK 0x00000E00L +#define GCEA_IO_WR_PRI_URGENCY__GROUP0_URGENCY_MODE_MASK 0x00001000L +#define GCEA_IO_WR_PRI_URGENCY__GROUP1_URGENCY_MODE_MASK 0x00002000L +#define GCEA_IO_WR_PRI_URGENCY__GROUP2_URGENCY_MODE_MASK 0x00004000L +#define GCEA_IO_WR_PRI_URGENCY__GROUP3_URGENCY_MODE_MASK 0x00008000L +//GCEA_IO_RD_PRI_URGENCY_MASK +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID0_MASK__SHIFT 0x0 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID1_MASK__SHIFT 0x1 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID2_MASK__SHIFT 0x2 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID3_MASK__SHIFT 0x3 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID4_MASK__SHIFT 0x4 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID5_MASK__SHIFT 0x5 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID6_MASK__SHIFT 0x6 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID7_MASK__SHIFT 0x7 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID8_MASK__SHIFT 0x8 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID9_MASK__SHIFT 0x9 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID10_MASK__SHIFT 0xa +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID11_MASK__SHIFT 0xb +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID12_MASK__SHIFT 0xc +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID13_MASK__SHIFT 0xd +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID14_MASK__SHIFT 0xe +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID15_MASK__SHIFT 0xf +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID16_MASK__SHIFT 0x10 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID17_MASK__SHIFT 0x11 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID18_MASK__SHIFT 0x12 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID19_MASK__SHIFT 0x13 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID20_MASK__SHIFT 0x14 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID21_MASK__SHIFT 0x15 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID22_MASK__SHIFT 0x16 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID23_MASK__SHIFT 0x17 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID24_MASK__SHIFT 0x18 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID25_MASK__SHIFT 0x19 +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID26_MASK__SHIFT 0x1a +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID27_MASK__SHIFT 0x1b +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID28_MASK__SHIFT 0x1c +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID29_MASK__SHIFT 0x1d +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID30_MASK__SHIFT 0x1e +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID31_MASK__SHIFT 0x1f +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID0_MASK_MASK 0x00000001L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID1_MASK_MASK 0x00000002L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID2_MASK_MASK 0x00000004L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID3_MASK_MASK 0x00000008L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID4_MASK_MASK 0x00000010L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID5_MASK_MASK 0x00000020L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID6_MASK_MASK 0x00000040L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID7_MASK_MASK 0x00000080L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID8_MASK_MASK 0x00000100L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID9_MASK_MASK 0x00000200L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID10_MASK_MASK 0x00000400L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID11_MASK_MASK 0x00000800L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID12_MASK_MASK 0x00001000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID13_MASK_MASK 0x00002000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID14_MASK_MASK 0x00004000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID15_MASK_MASK 0x00008000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID16_MASK_MASK 0x00010000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID17_MASK_MASK 0x00020000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID18_MASK_MASK 0x00040000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID19_MASK_MASK 0x00080000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID20_MASK_MASK 0x00100000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID21_MASK_MASK 0x00200000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID22_MASK_MASK 0x00400000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID23_MASK_MASK 0x00800000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID24_MASK_MASK 0x01000000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID25_MASK_MASK 0x02000000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID26_MASK_MASK 0x04000000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID27_MASK_MASK 0x08000000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID28_MASK_MASK 0x10000000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID29_MASK_MASK 0x20000000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID30_MASK_MASK 0x40000000L +#define GCEA_IO_RD_PRI_URGENCY_MASK__CID31_MASK_MASK 0x80000000L +//GCEA_IO_WR_PRI_URGENCY_MASK +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID0_MASK__SHIFT 0x0 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID1_MASK__SHIFT 0x1 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID2_MASK__SHIFT 0x2 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID3_MASK__SHIFT 0x3 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID4_MASK__SHIFT 0x4 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID5_MASK__SHIFT 0x5 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID6_MASK__SHIFT 0x6 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID7_MASK__SHIFT 0x7 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID8_MASK__SHIFT 0x8 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID9_MASK__SHIFT 0x9 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID10_MASK__SHIFT 0xa +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID11_MASK__SHIFT 0xb +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID12_MASK__SHIFT 0xc +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID13_MASK__SHIFT 0xd +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID14_MASK__SHIFT 0xe +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID15_MASK__SHIFT 0xf +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID16_MASK__SHIFT 0x10 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID17_MASK__SHIFT 0x11 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID18_MASK__SHIFT 0x12 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID19_MASK__SHIFT 0x13 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID20_MASK__SHIFT 0x14 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID21_MASK__SHIFT 0x15 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID22_MASK__SHIFT 0x16 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID23_MASK__SHIFT 0x17 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID24_MASK__SHIFT 0x18 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID25_MASK__SHIFT 0x19 +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID26_MASK__SHIFT 0x1a +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID27_MASK__SHIFT 0x1b +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID28_MASK__SHIFT 0x1c +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID29_MASK__SHIFT 0x1d +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID30_MASK__SHIFT 0x1e +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID31_MASK__SHIFT 0x1f +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID0_MASK_MASK 0x00000001L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID1_MASK_MASK 0x00000002L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID2_MASK_MASK 0x00000004L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID3_MASK_MASK 0x00000008L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID4_MASK_MASK 0x00000010L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID5_MASK_MASK 0x00000020L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID6_MASK_MASK 0x00000040L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID7_MASK_MASK 0x00000080L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID8_MASK_MASK 0x00000100L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID9_MASK_MASK 0x00000200L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID10_MASK_MASK 0x00000400L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID11_MASK_MASK 0x00000800L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID12_MASK_MASK 0x00001000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID13_MASK_MASK 0x00002000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID14_MASK_MASK 0x00004000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID15_MASK_MASK 0x00008000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID16_MASK_MASK 0x00010000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID17_MASK_MASK 0x00020000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID18_MASK_MASK 0x00040000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID19_MASK_MASK 0x00080000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID20_MASK_MASK 0x00100000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID21_MASK_MASK 0x00200000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID22_MASK_MASK 0x00400000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID23_MASK_MASK 0x00800000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID24_MASK_MASK 0x01000000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID25_MASK_MASK 0x02000000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID26_MASK_MASK 0x04000000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID27_MASK_MASK 0x08000000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID28_MASK_MASK 0x10000000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID29_MASK_MASK 0x20000000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID30_MASK_MASK 0x40000000L +#define GCEA_IO_WR_PRI_URGENCY_MASK__CID31_MASK_MASK 0x80000000L +//GCEA_IO_RD_PRI_QUANT_PRI1 +#define GCEA_IO_RD_PRI_QUANT_PRI1__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_IO_RD_PRI_QUANT_PRI1__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_IO_RD_PRI_QUANT_PRI1__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_IO_RD_PRI_QUANT_PRI1__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_IO_RD_PRI_QUANT_PRI1__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_IO_RD_PRI_QUANT_PRI1__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_IO_RD_PRI_QUANT_PRI1__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_IO_RD_PRI_QUANT_PRI1__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_IO_RD_PRI_QUANT_PRI2 +#define GCEA_IO_RD_PRI_QUANT_PRI2__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_IO_RD_PRI_QUANT_PRI2__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_IO_RD_PRI_QUANT_PRI2__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_IO_RD_PRI_QUANT_PRI2__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_IO_RD_PRI_QUANT_PRI2__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_IO_RD_PRI_QUANT_PRI2__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_IO_RD_PRI_QUANT_PRI2__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_IO_RD_PRI_QUANT_PRI2__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_IO_RD_PRI_QUANT_PRI3 +#define GCEA_IO_RD_PRI_QUANT_PRI3__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_IO_RD_PRI_QUANT_PRI3__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_IO_RD_PRI_QUANT_PRI3__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_IO_RD_PRI_QUANT_PRI3__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_IO_RD_PRI_QUANT_PRI3__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_IO_RD_PRI_QUANT_PRI3__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_IO_RD_PRI_QUANT_PRI3__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_IO_RD_PRI_QUANT_PRI3__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_IO_WR_PRI_QUANT_PRI1 +#define GCEA_IO_WR_PRI_QUANT_PRI1__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_IO_WR_PRI_QUANT_PRI1__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_IO_WR_PRI_QUANT_PRI1__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_IO_WR_PRI_QUANT_PRI1__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_IO_WR_PRI_QUANT_PRI1__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_IO_WR_PRI_QUANT_PRI1__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_IO_WR_PRI_QUANT_PRI1__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_IO_WR_PRI_QUANT_PRI1__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_IO_WR_PRI_QUANT_PRI2 +#define GCEA_IO_WR_PRI_QUANT_PRI2__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_IO_WR_PRI_QUANT_PRI2__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_IO_WR_PRI_QUANT_PRI2__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_IO_WR_PRI_QUANT_PRI2__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_IO_WR_PRI_QUANT_PRI2__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_IO_WR_PRI_QUANT_PRI2__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_IO_WR_PRI_QUANT_PRI2__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_IO_WR_PRI_QUANT_PRI2__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_IO_WR_PRI_QUANT_PRI3 +#define GCEA_IO_WR_PRI_QUANT_PRI3__GROUP0_THRESHOLD__SHIFT 0x0 +#define GCEA_IO_WR_PRI_QUANT_PRI3__GROUP1_THRESHOLD__SHIFT 0x8 +#define GCEA_IO_WR_PRI_QUANT_PRI3__GROUP2_THRESHOLD__SHIFT 0x10 +#define GCEA_IO_WR_PRI_QUANT_PRI3__GROUP3_THRESHOLD__SHIFT 0x18 +#define GCEA_IO_WR_PRI_QUANT_PRI3__GROUP0_THRESHOLD_MASK 0x000000FFL +#define GCEA_IO_WR_PRI_QUANT_PRI3__GROUP1_THRESHOLD_MASK 0x0000FF00L +#define GCEA_IO_WR_PRI_QUANT_PRI3__GROUP2_THRESHOLD_MASK 0x00FF0000L +#define GCEA_IO_WR_PRI_QUANT_PRI3__GROUP3_THRESHOLD_MASK 0xFF000000L +//GCEA_SDP_ARB_DRAM +#define GCEA_SDP_ARB_DRAM__RDWR_BURST_LIMIT_CYCL__SHIFT 0x0 +#define GCEA_SDP_ARB_DRAM__RDWR_BURST_LIMIT_DATA__SHIFT 0x8 +#define GCEA_SDP_ARB_DRAM__EARLY_SW2RD_ON_PRI__SHIFT 0x10 +#define GCEA_SDP_ARB_DRAM__EARLY_SW2WR_ON_PRI__SHIFT 0x11 +#define GCEA_SDP_ARB_DRAM__EARLY_SW2RD_ON_RES__SHIFT 0x12 +#define GCEA_SDP_ARB_DRAM__EARLY_SW2WR_ON_RES__SHIFT 0x13 +#define GCEA_SDP_ARB_DRAM__EOB_ON_EXPIRE__SHIFT 0x14 +#define GCEA_SDP_ARB_DRAM__RDWR_BURST_LIMIT_CYCL_MASK 0x0000007FL +#define GCEA_SDP_ARB_DRAM__RDWR_BURST_LIMIT_DATA_MASK 0x00007F00L +#define GCEA_SDP_ARB_DRAM__EARLY_SW2RD_ON_PRI_MASK 0x00010000L +#define GCEA_SDP_ARB_DRAM__EARLY_SW2WR_ON_PRI_MASK 0x00020000L +#define GCEA_SDP_ARB_DRAM__EARLY_SW2RD_ON_RES_MASK 0x00040000L +#define GCEA_SDP_ARB_DRAM__EARLY_SW2WR_ON_RES_MASK 0x00080000L +#define GCEA_SDP_ARB_DRAM__EOB_ON_EXPIRE_MASK 0x00100000L +//GCEA_SDP_ARB_FINAL +#define GCEA_SDP_ARB_FINAL__DRAM_BURST_LIMIT__SHIFT 0x0 +#define GCEA_SDP_ARB_FINAL__GMI_BURST_LIMIT__SHIFT 0x5 +#define GCEA_SDP_ARB_FINAL__IO_BURST_LIMIT__SHIFT 0xa +#define GCEA_SDP_ARB_FINAL__BURST_LIMIT_MULTIPLIER__SHIFT 0xf +#define GCEA_SDP_ARB_FINAL__RDONLY_VC0__SHIFT 0x11 +#define GCEA_SDP_ARB_FINAL__RDONLY_VC1__SHIFT 0x12 +#define GCEA_SDP_ARB_FINAL__RDONLY_VC2__SHIFT 0x13 +#define GCEA_SDP_ARB_FINAL__RDONLY_VC3__SHIFT 0x14 +#define GCEA_SDP_ARB_FINAL__RDONLY_VC4__SHIFT 0x15 +#define GCEA_SDP_ARB_FINAL__RDONLY_VC5__SHIFT 0x16 +#define GCEA_SDP_ARB_FINAL__RDONLY_VC6__SHIFT 0x17 +#define GCEA_SDP_ARB_FINAL__RDONLY_VC7__SHIFT 0x18 +#define GCEA_SDP_ARB_FINAL__ERREVENT_ON_ERROR__SHIFT 0x19 +#define GCEA_SDP_ARB_FINAL__HALTREQ_ON_ERROR__SHIFT 0x1a +#define GCEA_SDP_ARB_FINAL__DRAM_BURST_LIMIT_MASK 0x0000001FL +#define GCEA_SDP_ARB_FINAL__GMI_BURST_LIMIT_MASK 0x000003E0L +#define GCEA_SDP_ARB_FINAL__IO_BURST_LIMIT_MASK 0x00007C00L +#define GCEA_SDP_ARB_FINAL__BURST_LIMIT_MULTIPLIER_MASK 0x00018000L +#define GCEA_SDP_ARB_FINAL__RDONLY_VC0_MASK 0x00020000L +#define GCEA_SDP_ARB_FINAL__RDONLY_VC1_MASK 0x00040000L +#define GCEA_SDP_ARB_FINAL__RDONLY_VC2_MASK 0x00080000L +#define GCEA_SDP_ARB_FINAL__RDONLY_VC3_MASK 0x00100000L +#define GCEA_SDP_ARB_FINAL__RDONLY_VC4_MASK 0x00200000L +#define GCEA_SDP_ARB_FINAL__RDONLY_VC5_MASK 0x00400000L +#define GCEA_SDP_ARB_FINAL__RDONLY_VC6_MASK 0x00800000L +#define GCEA_SDP_ARB_FINAL__RDONLY_VC7_MASK 0x01000000L +#define GCEA_SDP_ARB_FINAL__ERREVENT_ON_ERROR_MASK 0x02000000L +#define GCEA_SDP_ARB_FINAL__HALTREQ_ON_ERROR_MASK 0x04000000L +//GCEA_SDP_DRAM_PRIORITY +#define GCEA_SDP_DRAM_PRIORITY__RD_GROUP0_PRIORITY__SHIFT 0x0 +#define GCEA_SDP_DRAM_PRIORITY__RD_GROUP1_PRIORITY__SHIFT 0x4 +#define GCEA_SDP_DRAM_PRIORITY__RD_GROUP2_PRIORITY__SHIFT 0x8 +#define GCEA_SDP_DRAM_PRIORITY__RD_GROUP3_PRIORITY__SHIFT 0xc +#define GCEA_SDP_DRAM_PRIORITY__WR_GROUP0_PRIORITY__SHIFT 0x10 +#define GCEA_SDP_DRAM_PRIORITY__WR_GROUP1_PRIORITY__SHIFT 0x14 +#define GCEA_SDP_DRAM_PRIORITY__WR_GROUP2_PRIORITY__SHIFT 0x18 +#define GCEA_SDP_DRAM_PRIORITY__WR_GROUP3_PRIORITY__SHIFT 0x1c +#define GCEA_SDP_DRAM_PRIORITY__RD_GROUP0_PRIORITY_MASK 0x0000000FL +#define GCEA_SDP_DRAM_PRIORITY__RD_GROUP1_PRIORITY_MASK 0x000000F0L +#define GCEA_SDP_DRAM_PRIORITY__RD_GROUP2_PRIORITY_MASK 0x00000F00L +#define GCEA_SDP_DRAM_PRIORITY__RD_GROUP3_PRIORITY_MASK 0x0000F000L +#define GCEA_SDP_DRAM_PRIORITY__WR_GROUP0_PRIORITY_MASK 0x000F0000L +#define GCEA_SDP_DRAM_PRIORITY__WR_GROUP1_PRIORITY_MASK 0x00F00000L +#define GCEA_SDP_DRAM_PRIORITY__WR_GROUP2_PRIORITY_MASK 0x0F000000L +#define GCEA_SDP_DRAM_PRIORITY__WR_GROUP3_PRIORITY_MASK 0xF0000000L +//GCEA_SDP_IO_PRIORITY +#define GCEA_SDP_IO_PRIORITY__RD_GROUP0_PRIORITY__SHIFT 0x0 +#define GCEA_SDP_IO_PRIORITY__RD_GROUP1_PRIORITY__SHIFT 0x4 +#define GCEA_SDP_IO_PRIORITY__RD_GROUP2_PRIORITY__SHIFT 0x8 +#define GCEA_SDP_IO_PRIORITY__RD_GROUP3_PRIORITY__SHIFT 0xc +#define GCEA_SDP_IO_PRIORITY__WR_GROUP0_PRIORITY__SHIFT 0x10 +#define GCEA_SDP_IO_PRIORITY__WR_GROUP1_PRIORITY__SHIFT 0x14 +#define GCEA_SDP_IO_PRIORITY__WR_GROUP2_PRIORITY__SHIFT 0x18 +#define GCEA_SDP_IO_PRIORITY__WR_GROUP3_PRIORITY__SHIFT 0x1c +#define GCEA_SDP_IO_PRIORITY__RD_GROUP0_PRIORITY_MASK 0x0000000FL +#define GCEA_SDP_IO_PRIORITY__RD_GROUP1_PRIORITY_MASK 0x000000F0L +#define GCEA_SDP_IO_PRIORITY__RD_GROUP2_PRIORITY_MASK 0x00000F00L +#define GCEA_SDP_IO_PRIORITY__RD_GROUP3_PRIORITY_MASK 0x0000F000L +#define GCEA_SDP_IO_PRIORITY__WR_GROUP0_PRIORITY_MASK 0x000F0000L +#define GCEA_SDP_IO_PRIORITY__WR_GROUP1_PRIORITY_MASK 0x00F00000L +#define GCEA_SDP_IO_PRIORITY__WR_GROUP2_PRIORITY_MASK 0x0F000000L +#define GCEA_SDP_IO_PRIORITY__WR_GROUP3_PRIORITY_MASK 0xF0000000L +//GCEA_SDP_CREDITS +#define GCEA_SDP_CREDITS__TAG_LIMIT__SHIFT 0x0 +#define GCEA_SDP_CREDITS__WR_RESP_CREDITS__SHIFT 0x8 +#define GCEA_SDP_CREDITS__RD_RESP_CREDITS__SHIFT 0x10 +#define GCEA_SDP_CREDITS__PRB_REQ_CREDITS__SHIFT 0x18 +#define GCEA_SDP_CREDITS__TAG_LIMIT_MASK 0x000000FFL +#define GCEA_SDP_CREDITS__WR_RESP_CREDITS_MASK 0x00007F00L +#define GCEA_SDP_CREDITS__RD_RESP_CREDITS_MASK 0x007F0000L +#define GCEA_SDP_CREDITS__PRB_REQ_CREDITS_MASK 0x3F000000L +//GCEA_SDP_TAG_RESERVE0 +#define GCEA_SDP_TAG_RESERVE0__VC0__SHIFT 0x0 +#define GCEA_SDP_TAG_RESERVE0__VC1__SHIFT 0x8 +#define GCEA_SDP_TAG_RESERVE0__VC2__SHIFT 0x10 +#define GCEA_SDP_TAG_RESERVE0__VC3__SHIFT 0x18 +#define GCEA_SDP_TAG_RESERVE0__VC0_MASK 0x000000FFL +#define GCEA_SDP_TAG_RESERVE0__VC1_MASK 0x0000FF00L +#define GCEA_SDP_TAG_RESERVE0__VC2_MASK 0x00FF0000L +#define GCEA_SDP_TAG_RESERVE0__VC3_MASK 0xFF000000L +//GCEA_SDP_TAG_RESERVE1 +#define GCEA_SDP_TAG_RESERVE1__VC4__SHIFT 0x0 +#define GCEA_SDP_TAG_RESERVE1__VC5__SHIFT 0x8 +#define GCEA_SDP_TAG_RESERVE1__VC6__SHIFT 0x10 +#define GCEA_SDP_TAG_RESERVE1__VC7__SHIFT 0x18 +#define GCEA_SDP_TAG_RESERVE1__VC4_MASK 0x000000FFL +#define GCEA_SDP_TAG_RESERVE1__VC5_MASK 0x0000FF00L +#define GCEA_SDP_TAG_RESERVE1__VC6_MASK 0x00FF0000L +#define GCEA_SDP_TAG_RESERVE1__VC7_MASK 0xFF000000L +//GCEA_SDP_VCC_RESERVE0 +#define GCEA_SDP_VCC_RESERVE0__VC0_CREDITS__SHIFT 0x0 +#define GCEA_SDP_VCC_RESERVE0__VC1_CREDITS__SHIFT 0x6 +#define GCEA_SDP_VCC_RESERVE0__VC2_CREDITS__SHIFT 0xc +#define GCEA_SDP_VCC_RESERVE0__VC3_CREDITS__SHIFT 0x12 +#define GCEA_SDP_VCC_RESERVE0__VC4_CREDITS__SHIFT 0x18 +#define GCEA_SDP_VCC_RESERVE0__VC0_CREDITS_MASK 0x0000003FL +#define GCEA_SDP_VCC_RESERVE0__VC1_CREDITS_MASK 0x00000FC0L +#define GCEA_SDP_VCC_RESERVE0__VC2_CREDITS_MASK 0x0003F000L +#define GCEA_SDP_VCC_RESERVE0__VC3_CREDITS_MASK 0x00FC0000L +#define GCEA_SDP_VCC_RESERVE0__VC4_CREDITS_MASK 0x3F000000L +//GCEA_SDP_VCC_RESERVE1 +#define GCEA_SDP_VCC_RESERVE1__VC5_CREDITS__SHIFT 0x0 +#define GCEA_SDP_VCC_RESERVE1__VC6_CREDITS__SHIFT 0x6 +#define GCEA_SDP_VCC_RESERVE1__VC7_CREDITS__SHIFT 0xc +#define GCEA_SDP_VCC_RESERVE1__DISTRIBUTE_POOL__SHIFT 0x1f +#define GCEA_SDP_VCC_RESERVE1__VC5_CREDITS_MASK 0x0000003FL +#define GCEA_SDP_VCC_RESERVE1__VC6_CREDITS_MASK 0x00000FC0L +#define GCEA_SDP_VCC_RESERVE1__VC7_CREDITS_MASK 0x0003F000L +#define GCEA_SDP_VCC_RESERVE1__DISTRIBUTE_POOL_MASK 0x80000000L +//GCEA_SDP_VCD_RESERVE0 +#define GCEA_SDP_VCD_RESERVE0__VC0_CREDITS__SHIFT 0x0 +#define GCEA_SDP_VCD_RESERVE0__VC1_CREDITS__SHIFT 0x6 +#define GCEA_SDP_VCD_RESERVE0__VC2_CREDITS__SHIFT 0xc +#define GCEA_SDP_VCD_RESERVE0__VC3_CREDITS__SHIFT 0x12 +#define GCEA_SDP_VCD_RESERVE0__VC4_CREDITS__SHIFT 0x18 +#define GCEA_SDP_VCD_RESERVE0__VC0_CREDITS_MASK 0x0000003FL +#define GCEA_SDP_VCD_RESERVE0__VC1_CREDITS_MASK 0x00000FC0L +#define GCEA_SDP_VCD_RESERVE0__VC2_CREDITS_MASK 0x0003F000L +#define GCEA_SDP_VCD_RESERVE0__VC3_CREDITS_MASK 0x00FC0000L +#define GCEA_SDP_VCD_RESERVE0__VC4_CREDITS_MASK 0x3F000000L +//GCEA_SDP_VCD_RESERVE1 +#define GCEA_SDP_VCD_RESERVE1__VC5_CREDITS__SHIFT 0x0 +#define GCEA_SDP_VCD_RESERVE1__VC6_CREDITS__SHIFT 0x6 +#define GCEA_SDP_VCD_RESERVE1__VC7_CREDITS__SHIFT 0xc +#define GCEA_SDP_VCD_RESERVE1__DISTRIBUTE_POOL__SHIFT 0x1f +#define GCEA_SDP_VCD_RESERVE1__VC5_CREDITS_MASK 0x0000003FL +#define GCEA_SDP_VCD_RESERVE1__VC6_CREDITS_MASK 0x00000FC0L +#define GCEA_SDP_VCD_RESERVE1__VC7_CREDITS_MASK 0x0003F000L +#define GCEA_SDP_VCD_RESERVE1__DISTRIBUTE_POOL_MASK 0x80000000L +//GCEA_SDP_REQ_CNTL +#define GCEA_SDP_REQ_CNTL__REQ_PASS_PW_OVERRIDE_READ__SHIFT 0x0 +#define GCEA_SDP_REQ_CNTL__REQ_PASS_PW_OVERRIDE_WRITE__SHIFT 0x1 +#define GCEA_SDP_REQ_CNTL__REQ_PASS_PW_OVERRIDE_ATOMIC__SHIFT 0x2 +#define GCEA_SDP_REQ_CNTL__REQ_CHAIN_OVERRIDE_DRAM__SHIFT 0x3 +#define GCEA_SDP_REQ_CNTL__INNER_DOMAIN_MODE__SHIFT 0x4 +#define GCEA_SDP_REQ_CNTL__REQ_PASS_PW_OVERRIDE_READ_MASK 0x00000001L +#define GCEA_SDP_REQ_CNTL__REQ_PASS_PW_OVERRIDE_WRITE_MASK 0x00000002L +#define GCEA_SDP_REQ_CNTL__REQ_PASS_PW_OVERRIDE_ATOMIC_MASK 0x00000004L +#define GCEA_SDP_REQ_CNTL__REQ_CHAIN_OVERRIDE_DRAM_MASK 0x00000008L +#define GCEA_SDP_REQ_CNTL__INNER_DOMAIN_MODE_MASK 0x00000010L +//GCEA_MISC +#define GCEA_MISC__RELATIVE_PRI_IN_DRAM_RD_ARB__SHIFT 0x0 +#define GCEA_MISC__RELATIVE_PRI_IN_DRAM_WR_ARB__SHIFT 0x1 +#define GCEA_MISC__RELATIVE_PRI_IN_GMI_RD_ARB__SHIFT 0x2 +#define GCEA_MISC__RELATIVE_PRI_IN_GMI_WR_ARB__SHIFT 0x3 +#define GCEA_MISC__RELATIVE_PRI_IN_IO_RD_ARB__SHIFT 0x4 +#define GCEA_MISC__RELATIVE_PRI_IN_IO_WR_ARB__SHIFT 0x5 +#define GCEA_MISC__EARLYWRRET_ENABLE_VC0__SHIFT 0x6 +#define GCEA_MISC__EARLYWRRET_ENABLE_VC1__SHIFT 0x7 +#define GCEA_MISC__EARLYWRRET_ENABLE_VC2__SHIFT 0x8 +#define GCEA_MISC__EARLYWRRET_ENABLE_VC3__SHIFT 0x9 +#define GCEA_MISC__EARLYWRRET_ENABLE_VC4__SHIFT 0xa +#define GCEA_MISC__EARLYWRRET_ENABLE_VC5__SHIFT 0xb +#define GCEA_MISC__EARLYWRRET_ENABLE_VC6__SHIFT 0xc +#define GCEA_MISC__EARLYWRRET_ENABLE_VC7__SHIFT 0xd +#define GCEA_MISC__EARLY_SDP_ORIGDATA__SHIFT 0xe +#define GCEA_MISC__LINKMGR_DYNAMIC_MODE__SHIFT 0xf +#define GCEA_MISC__LINKMGR_HALT_THRESHOLD__SHIFT 0x11 +#define GCEA_MISC__LINKMGR_RECONNECT_DELAY__SHIFT 0x13 +#define GCEA_MISC__LINKMGR_IDLE_THRESHOLD__SHIFT 0x15 +#define GCEA_MISC__FAVOUR_MIDCHAIN_CS_IN_DRAM_ARB__SHIFT 0x1a +#define GCEA_MISC__FAVOUR_MIDCHAIN_CS_IN_GMI_ARB__SHIFT 0x1b +#define GCEA_MISC__FAVOUR_LAST_CS_IN_DRAM_ARB__SHIFT 0x1c +#define GCEA_MISC__FAVOUR_LAST_CS_IN_GMI_ARB__SHIFT 0x1d +#define GCEA_MISC__SWITCH_CS_ON_W2R_IN_DRAM_ARB__SHIFT 0x1e +#define GCEA_MISC__SWITCH_CS_ON_W2R_IN_GMI_ARB__SHIFT 0x1f +#define GCEA_MISC__RELATIVE_PRI_IN_DRAM_RD_ARB_MASK 0x00000001L +#define GCEA_MISC__RELATIVE_PRI_IN_DRAM_WR_ARB_MASK 0x00000002L +#define GCEA_MISC__RELATIVE_PRI_IN_GMI_RD_ARB_MASK 0x00000004L +#define GCEA_MISC__RELATIVE_PRI_IN_GMI_WR_ARB_MASK 0x00000008L +#define GCEA_MISC__RELATIVE_PRI_IN_IO_RD_ARB_MASK 0x00000010L +#define GCEA_MISC__RELATIVE_PRI_IN_IO_WR_ARB_MASK 0x00000020L +#define GCEA_MISC__EARLYWRRET_ENABLE_VC0_MASK 0x00000040L +#define GCEA_MISC__EARLYWRRET_ENABLE_VC1_MASK 0x00000080L +#define GCEA_MISC__EARLYWRRET_ENABLE_VC2_MASK 0x00000100L +#define GCEA_MISC__EARLYWRRET_ENABLE_VC3_MASK 0x00000200L +#define GCEA_MISC__EARLYWRRET_ENABLE_VC4_MASK 0x00000400L +#define GCEA_MISC__EARLYWRRET_ENABLE_VC5_MASK 0x00000800L +#define GCEA_MISC__EARLYWRRET_ENABLE_VC6_MASK 0x00001000L +#define GCEA_MISC__EARLYWRRET_ENABLE_VC7_MASK 0x00002000L +#define GCEA_MISC__EARLY_SDP_ORIGDATA_MASK 0x00004000L +#define GCEA_MISC__LINKMGR_DYNAMIC_MODE_MASK 0x00018000L +#define GCEA_MISC__LINKMGR_HALT_THRESHOLD_MASK 0x00060000L +#define GCEA_MISC__LINKMGR_RECONNECT_DELAY_MASK 0x00180000L +#define GCEA_MISC__LINKMGR_IDLE_THRESHOLD_MASK 0x03E00000L +#define GCEA_MISC__FAVOUR_MIDCHAIN_CS_IN_DRAM_ARB_MASK 0x04000000L +#define GCEA_MISC__FAVOUR_MIDCHAIN_CS_IN_GMI_ARB_MASK 0x08000000L +#define GCEA_MISC__FAVOUR_LAST_CS_IN_DRAM_ARB_MASK 0x10000000L +#define GCEA_MISC__FAVOUR_LAST_CS_IN_GMI_ARB_MASK 0x20000000L +#define GCEA_MISC__SWITCH_CS_ON_W2R_IN_DRAM_ARB_MASK 0x40000000L +#define GCEA_MISC__SWITCH_CS_ON_W2R_IN_GMI_ARB_MASK 0x80000000L +//GCEA_LATENCY_SAMPLING +#define GCEA_LATENCY_SAMPLING__SAMPLER0_DRAM__SHIFT 0x0 +#define GCEA_LATENCY_SAMPLING__SAMPLER1_DRAM__SHIFT 0x1 +#define GCEA_LATENCY_SAMPLING__SAMPLER0_GMI__SHIFT 0x2 +#define GCEA_LATENCY_SAMPLING__SAMPLER1_GMI__SHIFT 0x3 +#define GCEA_LATENCY_SAMPLING__SAMPLER0_IO__SHIFT 0x4 +#define GCEA_LATENCY_SAMPLING__SAMPLER1_IO__SHIFT 0x5 +#define GCEA_LATENCY_SAMPLING__SAMPLER0_READ__SHIFT 0x6 +#define GCEA_LATENCY_SAMPLING__SAMPLER1_READ__SHIFT 0x7 +#define GCEA_LATENCY_SAMPLING__SAMPLER0_WRITE__SHIFT 0x8 +#define GCEA_LATENCY_SAMPLING__SAMPLER1_WRITE__SHIFT 0x9 +#define GCEA_LATENCY_SAMPLING__SAMPLER0_ATOMIC_RET__SHIFT 0xa +#define GCEA_LATENCY_SAMPLING__SAMPLER1_ATOMIC_RET__SHIFT 0xb +#define GCEA_LATENCY_SAMPLING__SAMPLER0_ATOMIC_NORET__SHIFT 0xc +#define GCEA_LATENCY_SAMPLING__SAMPLER1_ATOMIC_NORET__SHIFT 0xd +#define GCEA_LATENCY_SAMPLING__SAMPLER0_VC__SHIFT 0xe +#define GCEA_LATENCY_SAMPLING__SAMPLER1_VC__SHIFT 0x16 +#define GCEA_LATENCY_SAMPLING__SAMPLER0_DRAM_MASK 0x00000001L +#define GCEA_LATENCY_SAMPLING__SAMPLER1_DRAM_MASK 0x00000002L +#define GCEA_LATENCY_SAMPLING__SAMPLER0_GMI_MASK 0x00000004L +#define GCEA_LATENCY_SAMPLING__SAMPLER1_GMI_MASK 0x00000008L +#define GCEA_LATENCY_SAMPLING__SAMPLER0_IO_MASK 0x00000010L +#define GCEA_LATENCY_SAMPLING__SAMPLER1_IO_MASK 0x00000020L +#define GCEA_LATENCY_SAMPLING__SAMPLER0_READ_MASK 0x00000040L +#define GCEA_LATENCY_SAMPLING__SAMPLER1_READ_MASK 0x00000080L +#define GCEA_LATENCY_SAMPLING__SAMPLER0_WRITE_MASK 0x00000100L +#define GCEA_LATENCY_SAMPLING__SAMPLER1_WRITE_MASK 0x00000200L +#define GCEA_LATENCY_SAMPLING__SAMPLER0_ATOMIC_RET_MASK 0x00000400L +#define GCEA_LATENCY_SAMPLING__SAMPLER1_ATOMIC_RET_MASK 0x00000800L +#define GCEA_LATENCY_SAMPLING__SAMPLER0_ATOMIC_NORET_MASK 0x00001000L +#define GCEA_LATENCY_SAMPLING__SAMPLER1_ATOMIC_NORET_MASK 0x00002000L +#define GCEA_LATENCY_SAMPLING__SAMPLER0_VC_MASK 0x003FC000L +#define GCEA_LATENCY_SAMPLING__SAMPLER1_VC_MASK 0x3FC00000L +//GCEA_PERFCOUNTER_LO +#define GCEA_PERFCOUNTER_LO__COUNTER_LO__SHIFT 0x0 +#define GCEA_PERFCOUNTER_LO__COUNTER_LO_MASK 0xFFFFFFFFL +//GCEA_PERFCOUNTER_HI +#define GCEA_PERFCOUNTER_HI__COUNTER_HI__SHIFT 0x0 +#define GCEA_PERFCOUNTER_HI__COMPARE_VALUE__SHIFT 0x10 +#define GCEA_PERFCOUNTER_HI__COUNTER_HI_MASK 0x0000FFFFL +#define GCEA_PERFCOUNTER_HI__COMPARE_VALUE_MASK 0xFFFF0000L +//GCEA_PERFCOUNTER0_CFG +#define GCEA_PERFCOUNTER0_CFG__PERF_SEL__SHIFT 0x0 +#define GCEA_PERFCOUNTER0_CFG__PERF_SEL_END__SHIFT 0x8 +#define GCEA_PERFCOUNTER0_CFG__PERF_MODE__SHIFT 0x18 +#define GCEA_PERFCOUNTER0_CFG__ENABLE__SHIFT 0x1c +#define GCEA_PERFCOUNTER0_CFG__CLEAR__SHIFT 0x1d +#define GCEA_PERFCOUNTER0_CFG__PERF_SEL_MASK 0x000000FFL +#define GCEA_PERFCOUNTER0_CFG__PERF_SEL_END_MASK 0x0000FF00L +#define GCEA_PERFCOUNTER0_CFG__PERF_MODE_MASK 0x0F000000L +#define GCEA_PERFCOUNTER0_CFG__ENABLE_MASK 0x10000000L +#define GCEA_PERFCOUNTER0_CFG__CLEAR_MASK 0x20000000L +//GCEA_PERFCOUNTER1_CFG +#define GCEA_PERFCOUNTER1_CFG__PERF_SEL__SHIFT 0x0 +#define GCEA_PERFCOUNTER1_CFG__PERF_SEL_END__SHIFT 0x8 +#define GCEA_PERFCOUNTER1_CFG__PERF_MODE__SHIFT 0x18 +#define GCEA_PERFCOUNTER1_CFG__ENABLE__SHIFT 0x1c +#define GCEA_PERFCOUNTER1_CFG__CLEAR__SHIFT 0x1d +#define GCEA_PERFCOUNTER1_CFG__PERF_SEL_MASK 0x000000FFL +#define GCEA_PERFCOUNTER1_CFG__PERF_SEL_END_MASK 0x0000FF00L +#define GCEA_PERFCOUNTER1_CFG__PERF_MODE_MASK 0x0F000000L +#define GCEA_PERFCOUNTER1_CFG__ENABLE_MASK 0x10000000L +#define GCEA_PERFCOUNTER1_CFG__CLEAR_MASK 0x20000000L +//GCEA_PERFCOUNTER_RSLT_CNTL +#define GCEA_PERFCOUNTER_RSLT_CNTL__PERF_COUNTER_SELECT__SHIFT 0x0 +#define GCEA_PERFCOUNTER_RSLT_CNTL__START_TRIGGER__SHIFT 0x8 +#define GCEA_PERFCOUNTER_RSLT_CNTL__STOP_TRIGGER__SHIFT 0x10 +#define GCEA_PERFCOUNTER_RSLT_CNTL__ENABLE_ANY__SHIFT 0x18 +#define GCEA_PERFCOUNTER_RSLT_CNTL__CLEAR_ALL__SHIFT 0x19 +#define GCEA_PERFCOUNTER_RSLT_CNTL__STOP_ALL_ON_SATURATE__SHIFT 0x1a +#define GCEA_PERFCOUNTER_RSLT_CNTL__PERF_COUNTER_SELECT_MASK 0x0000000FL +#define GCEA_PERFCOUNTER_RSLT_CNTL__START_TRIGGER_MASK 0x0000FF00L +#define GCEA_PERFCOUNTER_RSLT_CNTL__STOP_TRIGGER_MASK 0x00FF0000L +#define GCEA_PERFCOUNTER_RSLT_CNTL__ENABLE_ANY_MASK 0x01000000L +#define GCEA_PERFCOUNTER_RSLT_CNTL__CLEAR_ALL_MASK 0x02000000L +#define GCEA_PERFCOUNTER_RSLT_CNTL__STOP_ALL_ON_SATURATE_MASK 0x04000000L + + +// addressBlock: gc_tcdec +//TCP_INVALIDATE +#define TCP_INVALIDATE__START__SHIFT 0x0 +#define TCP_INVALIDATE__START_MASK 0x00000001L +//TCP_STATUS +#define TCP_STATUS__TCP_BUSY__SHIFT 0x0 +#define TCP_STATUS__INPUT_BUSY__SHIFT 0x1 +#define TCP_STATUS__ADRS_BUSY__SHIFT 0x2 +#define TCP_STATUS__TAGRAMS_BUSY__SHIFT 0x3 +#define TCP_STATUS__CNTRL_BUSY__SHIFT 0x4 +#define TCP_STATUS__LFIFO_BUSY__SHIFT 0x5 +#define TCP_STATUS__READ_BUSY__SHIFT 0x6 +#define TCP_STATUS__FORMAT_BUSY__SHIFT 0x7 +#define TCP_STATUS__VM_BUSY__SHIFT 0x8 +#define TCP_STATUS__TCP_BUSY_MASK 0x00000001L +#define TCP_STATUS__INPUT_BUSY_MASK 0x00000002L +#define TCP_STATUS__ADRS_BUSY_MASK 0x00000004L +#define TCP_STATUS__TAGRAMS_BUSY_MASK 0x00000008L +#define TCP_STATUS__CNTRL_BUSY_MASK 0x00000010L +#define TCP_STATUS__LFIFO_BUSY_MASK 0x00000020L +#define TCP_STATUS__READ_BUSY_MASK 0x00000040L +#define TCP_STATUS__FORMAT_BUSY_MASK 0x00000080L +#define TCP_STATUS__VM_BUSY_MASK 0x00000100L +//TCP_CNTL +#define TCP_CNTL__FORCE_HIT__SHIFT 0x0 +#define TCP_CNTL__FORCE_MISS__SHIFT 0x1 +#define TCP_CNTL__L1_SIZE__SHIFT 0x2 +#define TCP_CNTL__FLAT_BUF_HASH_ENABLE__SHIFT 0x4 +#define TCP_CNTL__FLAT_BUF_CACHE_SWIZZLE__SHIFT 0x5 +#define TCP_CNTL__FORCE_EOW_TOTAL_CNT__SHIFT 0xf +#define TCP_CNTL__FORCE_EOW_TAGRAM_CNT__SHIFT 0x16 +#define TCP_CNTL__DISABLE_Z_MAP__SHIFT 0x1c +#define TCP_CNTL__INV_ALL_VMIDS__SHIFT 0x1d +#define TCP_CNTL__ASTC_VE_MSB_TOLERANT__SHIFT 0x1e +#define TCP_CNTL__FORCE_HIT_MASK 0x00000001L +#define TCP_CNTL__FORCE_MISS_MASK 0x00000002L +#define TCP_CNTL__L1_SIZE_MASK 0x0000000CL +#define TCP_CNTL__FLAT_BUF_HASH_ENABLE_MASK 0x00000010L +#define TCP_CNTL__FLAT_BUF_CACHE_SWIZZLE_MASK 0x00000020L +#define TCP_CNTL__FORCE_EOW_TOTAL_CNT_MASK 0x001F8000L +#define TCP_CNTL__FORCE_EOW_TAGRAM_CNT_MASK 0x0FC00000L +#define TCP_CNTL__DISABLE_Z_MAP_MASK 0x10000000L +#define TCP_CNTL__INV_ALL_VMIDS_MASK 0x20000000L +#define TCP_CNTL__ASTC_VE_MSB_TOLERANT_MASK 0x40000000L +//TCP_CHAN_STEER_LO +#define TCP_CHAN_STEER_LO__CHAN0__SHIFT 0x0 +#define TCP_CHAN_STEER_LO__CHAN1__SHIFT 0x4 +#define TCP_CHAN_STEER_LO__CHAN2__SHIFT 0x8 +#define TCP_CHAN_STEER_LO__CHAN3__SHIFT 0xc +#define TCP_CHAN_STEER_LO__CHAN4__SHIFT 0x10 +#define TCP_CHAN_STEER_LO__CHAN5__SHIFT 0x14 +#define TCP_CHAN_STEER_LO__CHAN6__SHIFT 0x18 +#define TCP_CHAN_STEER_LO__CHAN7__SHIFT 0x1c +#define TCP_CHAN_STEER_LO__CHAN0_MASK 0x0000000FL +#define TCP_CHAN_STEER_LO__CHAN1_MASK 0x000000F0L +#define TCP_CHAN_STEER_LO__CHAN2_MASK 0x00000F00L +#define TCP_CHAN_STEER_LO__CHAN3_MASK 0x0000F000L +#define TCP_CHAN_STEER_LO__CHAN4_MASK 0x000F0000L +#define TCP_CHAN_STEER_LO__CHAN5_MASK 0x00F00000L +#define TCP_CHAN_STEER_LO__CHAN6_MASK 0x0F000000L +#define TCP_CHAN_STEER_LO__CHAN7_MASK 0xF0000000L +//TCP_CHAN_STEER_HI +#define TCP_CHAN_STEER_HI__CHAN8__SHIFT 0x0 +#define TCP_CHAN_STEER_HI__CHAN9__SHIFT 0x4 +#define TCP_CHAN_STEER_HI__CHANA__SHIFT 0x8 +#define TCP_CHAN_STEER_HI__CHANB__SHIFT 0xc +#define TCP_CHAN_STEER_HI__CHANC__SHIFT 0x10 +#define TCP_CHAN_STEER_HI__CHAND__SHIFT 0x14 +#define TCP_CHAN_STEER_HI__CHANE__SHIFT 0x18 +#define TCP_CHAN_STEER_HI__CHANF__SHIFT 0x1c +#define TCP_CHAN_STEER_HI__CHAN8_MASK 0x0000000FL +#define TCP_CHAN_STEER_HI__CHAN9_MASK 0x000000F0L +#define TCP_CHAN_STEER_HI__CHANA_MASK 0x00000F00L +#define TCP_CHAN_STEER_HI__CHANB_MASK 0x0000F000L +#define TCP_CHAN_STEER_HI__CHANC_MASK 0x000F0000L +#define TCP_CHAN_STEER_HI__CHAND_MASK 0x00F00000L +#define TCP_CHAN_STEER_HI__CHANE_MASK 0x0F000000L +#define TCP_CHAN_STEER_HI__CHANF_MASK 0xF0000000L +//TCP_ADDR_CONFIG +#define TCP_ADDR_CONFIG__NUM_TCC_BANKS__SHIFT 0x0 +#define TCP_ADDR_CONFIG__NUM_BANKS__SHIFT 0x4 +#define TCP_ADDR_CONFIG__COLHI_WIDTH__SHIFT 0x6 +#define TCP_ADDR_CONFIG__RB_SPLIT_COLHI__SHIFT 0x9 +#define TCP_ADDR_CONFIG__NUM_TCC_BANKS_MASK 0x0000000FL +#define TCP_ADDR_CONFIG__NUM_BANKS_MASK 0x00000030L +#define TCP_ADDR_CONFIG__COLHI_WIDTH_MASK 0x000001C0L +#define TCP_ADDR_CONFIG__RB_SPLIT_COLHI_MASK 0x00000200L +//TCP_CREDIT +#define TCP_CREDIT__LFIFO_CREDIT__SHIFT 0x0 +#define TCP_CREDIT__REQ_FIFO_CREDIT__SHIFT 0x10 +#define TCP_CREDIT__TD_CREDIT__SHIFT 0x1d +#define TCP_CREDIT__LFIFO_CREDIT_MASK 0x000003FFL +#define TCP_CREDIT__REQ_FIFO_CREDIT_MASK 0x007F0000L +#define TCP_CREDIT__TD_CREDIT_MASK 0xE0000000L +//TCP_BUFFER_ADDR_HASH_CNTL +#define TCP_BUFFER_ADDR_HASH_CNTL__CHANNEL_BITS__SHIFT 0x0 +#define TCP_BUFFER_ADDR_HASH_CNTL__BANK_BITS__SHIFT 0x8 +#define TCP_BUFFER_ADDR_HASH_CNTL__CHANNEL_XOR_COUNT__SHIFT 0x10 +#define TCP_BUFFER_ADDR_HASH_CNTL__BANK_XOR_COUNT__SHIFT 0x18 +#define TCP_BUFFER_ADDR_HASH_CNTL__CHANNEL_BITS_MASK 0x00000007L +#define TCP_BUFFER_ADDR_HASH_CNTL__BANK_BITS_MASK 0x00000700L +#define TCP_BUFFER_ADDR_HASH_CNTL__CHANNEL_XOR_COUNT_MASK 0x00070000L +#define TCP_BUFFER_ADDR_HASH_CNTL__BANK_XOR_COUNT_MASK 0x07000000L +//TCP_EDC_CNT +#define TCP_EDC_CNT__SEC_COUNT__SHIFT 0x0 +#define TCP_EDC_CNT__LFIFO_SED_COUNT__SHIFT 0x8 +#define TCP_EDC_CNT__DED_COUNT__SHIFT 0x10 +#define TCP_EDC_CNT__SEC_COUNT_MASK 0x000000FFL +#define TCP_EDC_CNT__LFIFO_SED_COUNT_MASK 0x0000FF00L +#define TCP_EDC_CNT__DED_COUNT_MASK 0x00FF0000L +//TC_CFG_L1_LOAD_POLICY0 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_0__SHIFT 0x0 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_1__SHIFT 0x2 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_2__SHIFT 0x4 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_3__SHIFT 0x6 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_4__SHIFT 0x8 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_5__SHIFT 0xa +#define TC_CFG_L1_LOAD_POLICY0__POLICY_6__SHIFT 0xc +#define TC_CFG_L1_LOAD_POLICY0__POLICY_7__SHIFT 0xe +#define TC_CFG_L1_LOAD_POLICY0__POLICY_8__SHIFT 0x10 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_9__SHIFT 0x12 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_10__SHIFT 0x14 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_11__SHIFT 0x16 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_12__SHIFT 0x18 +#define TC_CFG_L1_LOAD_POLICY0__POLICY_13__SHIFT 0x1a +#define TC_CFG_L1_LOAD_POLICY0__POLICY_14__SHIFT 0x1c +#define TC_CFG_L1_LOAD_POLICY0__POLICY_15__SHIFT 0x1e +#define TC_CFG_L1_LOAD_POLICY0__POLICY_0_MASK 0x00000003L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_1_MASK 0x0000000CL +#define TC_CFG_L1_LOAD_POLICY0__POLICY_2_MASK 0x00000030L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_3_MASK 0x000000C0L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_4_MASK 0x00000300L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_5_MASK 0x00000C00L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_6_MASK 0x00003000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_7_MASK 0x0000C000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_8_MASK 0x00030000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_9_MASK 0x000C0000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_10_MASK 0x00300000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_11_MASK 0x00C00000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_12_MASK 0x03000000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_13_MASK 0x0C000000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_14_MASK 0x30000000L +#define TC_CFG_L1_LOAD_POLICY0__POLICY_15_MASK 0xC0000000L +//TC_CFG_L1_LOAD_POLICY1 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_16__SHIFT 0x0 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_17__SHIFT 0x2 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_18__SHIFT 0x4 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_19__SHIFT 0x6 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_20__SHIFT 0x8 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_21__SHIFT 0xa +#define TC_CFG_L1_LOAD_POLICY1__POLICY_22__SHIFT 0xc +#define TC_CFG_L1_LOAD_POLICY1__POLICY_23__SHIFT 0xe +#define TC_CFG_L1_LOAD_POLICY1__POLICY_24__SHIFT 0x10 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_25__SHIFT 0x12 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_26__SHIFT 0x14 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_27__SHIFT 0x16 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_28__SHIFT 0x18 +#define TC_CFG_L1_LOAD_POLICY1__POLICY_29__SHIFT 0x1a +#define TC_CFG_L1_LOAD_POLICY1__POLICY_30__SHIFT 0x1c +#define TC_CFG_L1_LOAD_POLICY1__POLICY_31__SHIFT 0x1e +#define TC_CFG_L1_LOAD_POLICY1__POLICY_16_MASK 0x00000003L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_17_MASK 0x0000000CL +#define TC_CFG_L1_LOAD_POLICY1__POLICY_18_MASK 0x00000030L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_19_MASK 0x000000C0L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_20_MASK 0x00000300L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_21_MASK 0x00000C00L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_22_MASK 0x00003000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_23_MASK 0x0000C000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_24_MASK 0x00030000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_25_MASK 0x000C0000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_26_MASK 0x00300000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_27_MASK 0x00C00000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_28_MASK 0x03000000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_29_MASK 0x0C000000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_30_MASK 0x30000000L +#define TC_CFG_L1_LOAD_POLICY1__POLICY_31_MASK 0xC0000000L +//TC_CFG_L1_STORE_POLICY +#define TC_CFG_L1_STORE_POLICY__POLICY_0__SHIFT 0x0 +#define TC_CFG_L1_STORE_POLICY__POLICY_1__SHIFT 0x1 +#define TC_CFG_L1_STORE_POLICY__POLICY_2__SHIFT 0x2 +#define TC_CFG_L1_STORE_POLICY__POLICY_3__SHIFT 0x3 +#define TC_CFG_L1_STORE_POLICY__POLICY_4__SHIFT 0x4 +#define TC_CFG_L1_STORE_POLICY__POLICY_5__SHIFT 0x5 +#define TC_CFG_L1_STORE_POLICY__POLICY_6__SHIFT 0x6 +#define TC_CFG_L1_STORE_POLICY__POLICY_7__SHIFT 0x7 +#define TC_CFG_L1_STORE_POLICY__POLICY_8__SHIFT 0x8 +#define TC_CFG_L1_STORE_POLICY__POLICY_9__SHIFT 0x9 +#define TC_CFG_L1_STORE_POLICY__POLICY_10__SHIFT 0xa +#define TC_CFG_L1_STORE_POLICY__POLICY_11__SHIFT 0xb +#define TC_CFG_L1_STORE_POLICY__POLICY_12__SHIFT 0xc +#define TC_CFG_L1_STORE_POLICY__POLICY_13__SHIFT 0xd +#define TC_CFG_L1_STORE_POLICY__POLICY_14__SHIFT 0xe +#define TC_CFG_L1_STORE_POLICY__POLICY_15__SHIFT 0xf +#define TC_CFG_L1_STORE_POLICY__POLICY_16__SHIFT 0x10 +#define TC_CFG_L1_STORE_POLICY__POLICY_17__SHIFT 0x11 +#define TC_CFG_L1_STORE_POLICY__POLICY_18__SHIFT 0x12 +#define TC_CFG_L1_STORE_POLICY__POLICY_19__SHIFT 0x13 +#define TC_CFG_L1_STORE_POLICY__POLICY_20__SHIFT 0x14 +#define TC_CFG_L1_STORE_POLICY__POLICY_21__SHIFT 0x15 +#define TC_CFG_L1_STORE_POLICY__POLICY_22__SHIFT 0x16 +#define TC_CFG_L1_STORE_POLICY__POLICY_23__SHIFT 0x17 +#define TC_CFG_L1_STORE_POLICY__POLICY_24__SHIFT 0x18 +#define TC_CFG_L1_STORE_POLICY__POLICY_25__SHIFT 0x19 +#define TC_CFG_L1_STORE_POLICY__POLICY_26__SHIFT 0x1a +#define TC_CFG_L1_STORE_POLICY__POLICY_27__SHIFT 0x1b +#define TC_CFG_L1_STORE_POLICY__POLICY_28__SHIFT 0x1c +#define TC_CFG_L1_STORE_POLICY__POLICY_29__SHIFT 0x1d +#define TC_CFG_L1_STORE_POLICY__POLICY_30__SHIFT 0x1e +#define TC_CFG_L1_STORE_POLICY__POLICY_31__SHIFT 0x1f +#define TC_CFG_L1_STORE_POLICY__POLICY_0_MASK 0x00000001L +#define TC_CFG_L1_STORE_POLICY__POLICY_1_MASK 0x00000002L +#define TC_CFG_L1_STORE_POLICY__POLICY_2_MASK 0x00000004L +#define TC_CFG_L1_STORE_POLICY__POLICY_3_MASK 0x00000008L +#define TC_CFG_L1_STORE_POLICY__POLICY_4_MASK 0x00000010L +#define TC_CFG_L1_STORE_POLICY__POLICY_5_MASK 0x00000020L +#define TC_CFG_L1_STORE_POLICY__POLICY_6_MASK 0x00000040L +#define TC_CFG_L1_STORE_POLICY__POLICY_7_MASK 0x00000080L +#define TC_CFG_L1_STORE_POLICY__POLICY_8_MASK 0x00000100L +#define TC_CFG_L1_STORE_POLICY__POLICY_9_MASK 0x00000200L +#define TC_CFG_L1_STORE_POLICY__POLICY_10_MASK 0x00000400L +#define TC_CFG_L1_STORE_POLICY__POLICY_11_MASK 0x00000800L +#define TC_CFG_L1_STORE_POLICY__POLICY_12_MASK 0x00001000L +#define TC_CFG_L1_STORE_POLICY__POLICY_13_MASK 0x00002000L +#define TC_CFG_L1_STORE_POLICY__POLICY_14_MASK 0x00004000L +#define TC_CFG_L1_STORE_POLICY__POLICY_15_MASK 0x00008000L +#define TC_CFG_L1_STORE_POLICY__POLICY_16_MASK 0x00010000L +#define TC_CFG_L1_STORE_POLICY__POLICY_17_MASK 0x00020000L +#define TC_CFG_L1_STORE_POLICY__POLICY_18_MASK 0x00040000L +#define TC_CFG_L1_STORE_POLICY__POLICY_19_MASK 0x00080000L +#define TC_CFG_L1_STORE_POLICY__POLICY_20_MASK 0x00100000L +#define TC_CFG_L1_STORE_POLICY__POLICY_21_MASK 0x00200000L +#define TC_CFG_L1_STORE_POLICY__POLICY_22_MASK 0x00400000L +#define TC_CFG_L1_STORE_POLICY__POLICY_23_MASK 0x00800000L +#define TC_CFG_L1_STORE_POLICY__POLICY_24_MASK 0x01000000L +#define TC_CFG_L1_STORE_POLICY__POLICY_25_MASK 0x02000000L +#define TC_CFG_L1_STORE_POLICY__POLICY_26_MASK 0x04000000L +#define TC_CFG_L1_STORE_POLICY__POLICY_27_MASK 0x08000000L +#define TC_CFG_L1_STORE_POLICY__POLICY_28_MASK 0x10000000L +#define TC_CFG_L1_STORE_POLICY__POLICY_29_MASK 0x20000000L +#define TC_CFG_L1_STORE_POLICY__POLICY_30_MASK 0x40000000L +#define TC_CFG_L1_STORE_POLICY__POLICY_31_MASK 0x80000000L +//TC_CFG_L2_LOAD_POLICY0 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_0__SHIFT 0x0 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_1__SHIFT 0x2 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_2__SHIFT 0x4 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_3__SHIFT 0x6 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_4__SHIFT 0x8 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_5__SHIFT 0xa +#define TC_CFG_L2_LOAD_POLICY0__POLICY_6__SHIFT 0xc +#define TC_CFG_L2_LOAD_POLICY0__POLICY_7__SHIFT 0xe +#define TC_CFG_L2_LOAD_POLICY0__POLICY_8__SHIFT 0x10 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_9__SHIFT 0x12 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_10__SHIFT 0x14 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_11__SHIFT 0x16 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_12__SHIFT 0x18 +#define TC_CFG_L2_LOAD_POLICY0__POLICY_13__SHIFT 0x1a +#define TC_CFG_L2_LOAD_POLICY0__POLICY_14__SHIFT 0x1c +#define TC_CFG_L2_LOAD_POLICY0__POLICY_15__SHIFT 0x1e +#define TC_CFG_L2_LOAD_POLICY0__POLICY_0_MASK 0x00000003L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_1_MASK 0x0000000CL +#define TC_CFG_L2_LOAD_POLICY0__POLICY_2_MASK 0x00000030L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_3_MASK 0x000000C0L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_4_MASK 0x00000300L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_5_MASK 0x00000C00L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_6_MASK 0x00003000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_7_MASK 0x0000C000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_8_MASK 0x00030000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_9_MASK 0x000C0000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_10_MASK 0x00300000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_11_MASK 0x00C00000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_12_MASK 0x03000000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_13_MASK 0x0C000000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_14_MASK 0x30000000L +#define TC_CFG_L2_LOAD_POLICY0__POLICY_15_MASK 0xC0000000L +//TC_CFG_L2_LOAD_POLICY1 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_16__SHIFT 0x0 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_17__SHIFT 0x2 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_18__SHIFT 0x4 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_19__SHIFT 0x6 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_20__SHIFT 0x8 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_21__SHIFT 0xa +#define TC_CFG_L2_LOAD_POLICY1__POLICY_22__SHIFT 0xc +#define TC_CFG_L2_LOAD_POLICY1__POLICY_23__SHIFT 0xe +#define TC_CFG_L2_LOAD_POLICY1__POLICY_24__SHIFT 0x10 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_25__SHIFT 0x12 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_26__SHIFT 0x14 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_27__SHIFT 0x16 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_28__SHIFT 0x18 +#define TC_CFG_L2_LOAD_POLICY1__POLICY_29__SHIFT 0x1a +#define TC_CFG_L2_LOAD_POLICY1__POLICY_30__SHIFT 0x1c +#define TC_CFG_L2_LOAD_POLICY1__POLICY_31__SHIFT 0x1e +#define TC_CFG_L2_LOAD_POLICY1__POLICY_16_MASK 0x00000003L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_17_MASK 0x0000000CL +#define TC_CFG_L2_LOAD_POLICY1__POLICY_18_MASK 0x00000030L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_19_MASK 0x000000C0L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_20_MASK 0x00000300L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_21_MASK 0x00000C00L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_22_MASK 0x00003000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_23_MASK 0x0000C000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_24_MASK 0x00030000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_25_MASK 0x000C0000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_26_MASK 0x00300000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_27_MASK 0x00C00000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_28_MASK 0x03000000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_29_MASK 0x0C000000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_30_MASK 0x30000000L +#define TC_CFG_L2_LOAD_POLICY1__POLICY_31_MASK 0xC0000000L +//TC_CFG_L2_STORE_POLICY0 +#define TC_CFG_L2_STORE_POLICY0__POLICY_0__SHIFT 0x0 +#define TC_CFG_L2_STORE_POLICY0__POLICY_1__SHIFT 0x2 +#define TC_CFG_L2_STORE_POLICY0__POLICY_2__SHIFT 0x4 +#define TC_CFG_L2_STORE_POLICY0__POLICY_3__SHIFT 0x6 +#define TC_CFG_L2_STORE_POLICY0__POLICY_4__SHIFT 0x8 +#define TC_CFG_L2_STORE_POLICY0__POLICY_5__SHIFT 0xa +#define TC_CFG_L2_STORE_POLICY0__POLICY_6__SHIFT 0xc +#define TC_CFG_L2_STORE_POLICY0__POLICY_7__SHIFT 0xe +#define TC_CFG_L2_STORE_POLICY0__POLICY_8__SHIFT 0x10 +#define TC_CFG_L2_STORE_POLICY0__POLICY_9__SHIFT 0x12 +#define TC_CFG_L2_STORE_POLICY0__POLICY_10__SHIFT 0x14 +#define TC_CFG_L2_STORE_POLICY0__POLICY_11__SHIFT 0x16 +#define TC_CFG_L2_STORE_POLICY0__POLICY_12__SHIFT 0x18 +#define TC_CFG_L2_STORE_POLICY0__POLICY_13__SHIFT 0x1a +#define TC_CFG_L2_STORE_POLICY0__POLICY_14__SHIFT 0x1c +#define TC_CFG_L2_STORE_POLICY0__POLICY_15__SHIFT 0x1e +#define TC_CFG_L2_STORE_POLICY0__POLICY_0_MASK 0x00000003L +#define TC_CFG_L2_STORE_POLICY0__POLICY_1_MASK 0x0000000CL +#define TC_CFG_L2_STORE_POLICY0__POLICY_2_MASK 0x00000030L +#define TC_CFG_L2_STORE_POLICY0__POLICY_3_MASK 0x000000C0L +#define TC_CFG_L2_STORE_POLICY0__POLICY_4_MASK 0x00000300L +#define TC_CFG_L2_STORE_POLICY0__POLICY_5_MASK 0x00000C00L +#define TC_CFG_L2_STORE_POLICY0__POLICY_6_MASK 0x00003000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_7_MASK 0x0000C000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_8_MASK 0x00030000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_9_MASK 0x000C0000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_10_MASK 0x00300000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_11_MASK 0x00C00000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_12_MASK 0x03000000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_13_MASK 0x0C000000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_14_MASK 0x30000000L +#define TC_CFG_L2_STORE_POLICY0__POLICY_15_MASK 0xC0000000L +//TC_CFG_L2_STORE_POLICY1 +#define TC_CFG_L2_STORE_POLICY1__POLICY_16__SHIFT 0x0 +#define TC_CFG_L2_STORE_POLICY1__POLICY_17__SHIFT 0x2 +#define TC_CFG_L2_STORE_POLICY1__POLICY_18__SHIFT 0x4 +#define TC_CFG_L2_STORE_POLICY1__POLICY_19__SHIFT 0x6 +#define TC_CFG_L2_STORE_POLICY1__POLICY_20__SHIFT 0x8 +#define TC_CFG_L2_STORE_POLICY1__POLICY_21__SHIFT 0xa +#define TC_CFG_L2_STORE_POLICY1__POLICY_22__SHIFT 0xc +#define TC_CFG_L2_STORE_POLICY1__POLICY_23__SHIFT 0xe +#define TC_CFG_L2_STORE_POLICY1__POLICY_24__SHIFT 0x10 +#define TC_CFG_L2_STORE_POLICY1__POLICY_25__SHIFT 0x12 +#define TC_CFG_L2_STORE_POLICY1__POLICY_26__SHIFT 0x14 +#define TC_CFG_L2_STORE_POLICY1__POLICY_27__SHIFT 0x16 +#define TC_CFG_L2_STORE_POLICY1__POLICY_28__SHIFT 0x18 +#define TC_CFG_L2_STORE_POLICY1__POLICY_29__SHIFT 0x1a +#define TC_CFG_L2_STORE_POLICY1__POLICY_30__SHIFT 0x1c +#define TC_CFG_L2_STORE_POLICY1__POLICY_31__SHIFT 0x1e +#define TC_CFG_L2_STORE_POLICY1__POLICY_16_MASK 0x00000003L +#define TC_CFG_L2_STORE_POLICY1__POLICY_17_MASK 0x0000000CL +#define TC_CFG_L2_STORE_POLICY1__POLICY_18_MASK 0x00000030L +#define TC_CFG_L2_STORE_POLICY1__POLICY_19_MASK 0x000000C0L +#define TC_CFG_L2_STORE_POLICY1__POLICY_20_MASK 0x00000300L +#define TC_CFG_L2_STORE_POLICY1__POLICY_21_MASK 0x00000C00L +#define TC_CFG_L2_STORE_POLICY1__POLICY_22_MASK 0x00003000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_23_MASK 0x0000C000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_24_MASK 0x00030000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_25_MASK 0x000C0000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_26_MASK 0x00300000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_27_MASK 0x00C00000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_28_MASK 0x03000000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_29_MASK 0x0C000000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_30_MASK 0x30000000L +#define TC_CFG_L2_STORE_POLICY1__POLICY_31_MASK 0xC0000000L +//TC_CFG_L2_ATOMIC_POLICY +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_0__SHIFT 0x0 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_1__SHIFT 0x2 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_2__SHIFT 0x4 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_3__SHIFT 0x6 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_4__SHIFT 0x8 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_5__SHIFT 0xa +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_6__SHIFT 0xc +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_7__SHIFT 0xe +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_8__SHIFT 0x10 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_9__SHIFT 0x12 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_10__SHIFT 0x14 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_11__SHIFT 0x16 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_12__SHIFT 0x18 +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_13__SHIFT 0x1a +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_14__SHIFT 0x1c +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_15__SHIFT 0x1e +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_0_MASK 0x00000003L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_1_MASK 0x0000000CL +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_2_MASK 0x00000030L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_3_MASK 0x000000C0L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_4_MASK 0x00000300L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_5_MASK 0x00000C00L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_6_MASK 0x00003000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_7_MASK 0x0000C000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_8_MASK 0x00030000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_9_MASK 0x000C0000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_10_MASK 0x00300000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_11_MASK 0x00C00000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_12_MASK 0x03000000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_13_MASK 0x0C000000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_14_MASK 0x30000000L +#define TC_CFG_L2_ATOMIC_POLICY__POLICY_15_MASK 0xC0000000L +//TC_CFG_L1_VOLATILE +#define TC_CFG_L1_VOLATILE__VOL__SHIFT 0x0 +#define TC_CFG_L1_VOLATILE__VOL_MASK 0x0000000FL +//TC_CFG_L2_VOLATILE +#define TC_CFG_L2_VOLATILE__VOL__SHIFT 0x0 +#define TC_CFG_L2_VOLATILE__VOL_MASK 0x0000000FL +//TCI_STATUS +#define TCI_STATUS__TCI_BUSY__SHIFT 0x0 +#define TCI_STATUS__TCI_BUSY_MASK 0x00000001L +//TCI_CNTL_1 +#define TCI_CNTL_1__WBINVL1_NUM_CYCLES__SHIFT 0x0 +#define TCI_CNTL_1__REQ_FIFO_DEPTH__SHIFT 0x10 +#define TCI_CNTL_1__WDATA_RAM_DEPTH__SHIFT 0x18 +#define TCI_CNTL_1__WBINVL1_NUM_CYCLES_MASK 0x0000FFFFL +#define TCI_CNTL_1__REQ_FIFO_DEPTH_MASK 0x00FF0000L +#define TCI_CNTL_1__WDATA_RAM_DEPTH_MASK 0xFF000000L +//TCI_CNTL_2 +#define TCI_CNTL_2__L1_INVAL_ON_WBINVL2__SHIFT 0x0 +#define TCI_CNTL_2__TCA_MAX_CREDIT__SHIFT 0x1 +#define TCI_CNTL_2__L1_INVAL_ON_WBINVL2_MASK 0x00000001L +#define TCI_CNTL_2__TCA_MAX_CREDIT_MASK 0x000001FEL +//TCC_CTRL +#define TCC_CTRL__CACHE_SIZE__SHIFT 0x0 +#define TCC_CTRL__RATE__SHIFT 0x2 +#define TCC_CTRL__WRITEBACK_MARGIN__SHIFT 0x4 +#define TCC_CTRL__METADATA_LATENCY_FIFO_SIZE__SHIFT 0x8 +#define TCC_CTRL__SRC_FIFO_SIZE__SHIFT 0xc +#define TCC_CTRL__LATENCY_FIFO_SIZE__SHIFT 0x10 +#define TCC_CTRL__LINEAR_SET_HASH__SHIFT 0x15 +#define TCC_CTRL__MDC_SIZE__SHIFT 0x18 +#define TCC_CTRL__MDC_SECTOR_SIZE__SHIFT 0x1a +#define TCC_CTRL__MDC_SIDEBAND_FIFO_SIZE__SHIFT 0x1c +#define TCC_CTRL__CACHE_SIZE_MASK 0x00000003L +#define TCC_CTRL__RATE_MASK 0x0000000CL +#define TCC_CTRL__WRITEBACK_MARGIN_MASK 0x000000F0L +#define TCC_CTRL__METADATA_LATENCY_FIFO_SIZE_MASK 0x00000F00L +#define TCC_CTRL__SRC_FIFO_SIZE_MASK 0x0000F000L +#define TCC_CTRL__LATENCY_FIFO_SIZE_MASK 0x000F0000L +#define TCC_CTRL__LINEAR_SET_HASH_MASK 0x00200000L +#define TCC_CTRL__MDC_SIZE_MASK 0x03000000L +#define TCC_CTRL__MDC_SECTOR_SIZE_MASK 0x0C000000L +#define TCC_CTRL__MDC_SIDEBAND_FIFO_SIZE_MASK 0xF0000000L +//TCC_CTRL2 +#define TCC_CTRL2__PROBE_FIFO_SIZE__SHIFT 0x0 +#define TCC_CTRL2__PROBE_FIFO_SIZE_MASK 0x0000000FL +//TCC_EDC_CNT +#define TCC_EDC_CNT__CACHE_DATA_SEC_COUNT__SHIFT 0x0 +#define TCC_EDC_CNT__CACHE_DATA_DED_COUNT__SHIFT 0x2 +#define TCC_EDC_CNT__CACHE_DIRTY_SEC_COUNT__SHIFT 0x4 +#define TCC_EDC_CNT__CACHE_DIRTY_DED_COUNT__SHIFT 0x6 +#define TCC_EDC_CNT__HIGH_RATE_TAG_SEC_COUNT__SHIFT 0x8 +#define TCC_EDC_CNT__HIGH_RATE_TAG_DED_COUNT__SHIFT 0xa +#define TCC_EDC_CNT__LOW_RATE_TAG_SEC_COUNT__SHIFT 0xc +#define TCC_EDC_CNT__LOW_RATE_TAG_DED_COUNT__SHIFT 0xe +#define TCC_EDC_CNT__SRC_FIFO_SEC_COUNT__SHIFT 0x10 +#define TCC_EDC_CNT__SRC_FIFO_DED_COUNT__SHIFT 0x12 +#define TCC_EDC_CNT__IN_USE_DEC_SED_COUNT__SHIFT 0x14 +#define TCC_EDC_CNT__IN_USE_TRANSFER_SED_COUNT__SHIFT 0x16 +#define TCC_EDC_CNT__LATENCY_FIFO_SED_COUNT__SHIFT 0x18 +#define TCC_EDC_CNT__RETURN_DATA_SED_COUNT__SHIFT 0x1a +#define TCC_EDC_CNT__RETURN_CONTROL_SED_COUNT__SHIFT 0x1c +#define TCC_EDC_CNT__UC_ATOMIC_FIFO_SED_COUNT__SHIFT 0x1e +#define TCC_EDC_CNT__CACHE_DATA_SEC_COUNT_MASK 0x00000003L +#define TCC_EDC_CNT__CACHE_DATA_DED_COUNT_MASK 0x0000000CL +#define TCC_EDC_CNT__CACHE_DIRTY_SEC_COUNT_MASK 0x00000030L +#define TCC_EDC_CNT__CACHE_DIRTY_DED_COUNT_MASK 0x000000C0L +#define TCC_EDC_CNT__HIGH_RATE_TAG_SEC_COUNT_MASK 0x00000300L +#define TCC_EDC_CNT__HIGH_RATE_TAG_DED_COUNT_MASK 0x00000C00L +#define TCC_EDC_CNT__LOW_RATE_TAG_SEC_COUNT_MASK 0x00003000L +#define TCC_EDC_CNT__LOW_RATE_TAG_DED_COUNT_MASK 0x0000C000L +#define TCC_EDC_CNT__SRC_FIFO_SEC_COUNT_MASK 0x00030000L +#define TCC_EDC_CNT__SRC_FIFO_DED_COUNT_MASK 0x000C0000L +#define TCC_EDC_CNT__IN_USE_DEC_SED_COUNT_MASK 0x00300000L +#define TCC_EDC_CNT__IN_USE_TRANSFER_SED_COUNT_MASK 0x00C00000L +#define TCC_EDC_CNT__LATENCY_FIFO_SED_COUNT_MASK 0x03000000L +#define TCC_EDC_CNT__RETURN_DATA_SED_COUNT_MASK 0x0C000000L +#define TCC_EDC_CNT__RETURN_CONTROL_SED_COUNT_MASK 0x30000000L +#define TCC_EDC_CNT__UC_ATOMIC_FIFO_SED_COUNT_MASK 0xC0000000L +//TCC_EDC_CNT2 +#define TCC_EDC_CNT2__WRITE_RETURN_SED_COUNT__SHIFT 0x0 +#define TCC_EDC_CNT2__WRITE_CACHE_READ_SED_COUNT__SHIFT 0x2 +#define TCC_EDC_CNT2__SRC_FIFO_NEXT_RAM_SED_COUNT__SHIFT 0x4 +#define TCC_EDC_CNT2__LATENCY_FIFO_NEXT_RAM_SED_COUNT__SHIFT 0x6 +#define TCC_EDC_CNT2__CACHE_TAG_PROBE_FIFO_SED_COUNT__SHIFT 0x8 +#define TCC_EDC_CNT2__WRITE_RETURN_SED_COUNT_MASK 0x00000003L +#define TCC_EDC_CNT2__WRITE_CACHE_READ_SED_COUNT_MASK 0x0000000CL +#define TCC_EDC_CNT2__SRC_FIFO_NEXT_RAM_SED_COUNT_MASK 0x00000030L +#define TCC_EDC_CNT2__LATENCY_FIFO_NEXT_RAM_SED_COUNT_MASK 0x000000C0L +#define TCC_EDC_CNT2__CACHE_TAG_PROBE_FIFO_SED_COUNT_MASK 0x00000300L +//TCC_REDUNDANCY +#define TCC_REDUNDANCY__MC_SEL0__SHIFT 0x0 +#define TCC_REDUNDANCY__MC_SEL1__SHIFT 0x1 +#define TCC_REDUNDANCY__MC_SEL0_MASK 0x00000001L +#define TCC_REDUNDANCY__MC_SEL1_MASK 0x00000002L +//TCC_EXE_DISABLE +#define TCC_EXE_DISABLE__EXE_DISABLE__SHIFT 0x1 +#define TCC_EXE_DISABLE__EXE_DISABLE_MASK 0x00000002L +//TCC_DSM_CNTL +#define TCC_DSM_CNTL__CACHE_DATA_IRRITATOR_DATA_SEL__SHIFT 0x0 +#define TCC_DSM_CNTL__CACHE_DATA_IRRITATOR_SINGLE_WRITE__SHIFT 0x2 +#define TCC_DSM_CNTL__CACHE_DATA_BANK_0_1_IRRITATOR_DATA_SEL__SHIFT 0x3 +#define TCC_DSM_CNTL__CACHE_DATA_BANK_0_1_IRRITATOR_SINGLE_WRITE__SHIFT 0x5 +#define TCC_DSM_CNTL__CACHE_DATA_BANK_1_0_IRRITATOR_DATA_SEL__SHIFT 0x6 +#define TCC_DSM_CNTL__CACHE_DATA_BANK_1_0_IRRITATOR_SINGLE_WRITE__SHIFT 0x8 +#define TCC_DSM_CNTL__CACHE_DATA_BANK_1_1_IRRITATOR_DATA_SEL__SHIFT 0x9 +#define TCC_DSM_CNTL__CACHE_DATA_BANK_1_1_IRRITATOR_SINGLE_WRITE__SHIFT 0xb +#define TCC_DSM_CNTL__CACHE_DIRTY_BANK_0_IRRITATOR_DATA_SEL__SHIFT 0xc +#define TCC_DSM_CNTL__CACHE_DIRTY_BANK_0_IRRITATOR_SINGLE_WRITE__SHIFT 0xe +#define TCC_DSM_CNTL__CACHE_DIRTY_BANK_1_IRRITATOR_DATA_SEL__SHIFT 0xf +#define TCC_DSM_CNTL__CACHE_DIRTY_BANK_1_IRRITATOR_SINGLE_WRITE__SHIFT 0x11 +#define TCC_DSM_CNTL__HIGH_RATE_TAG_IRRITATOR_DATA_SEL__SHIFT 0x12 +#define TCC_DSM_CNTL__HIGH_RATE_TAG_IRRITATOR_SINGLE_WRITE__SHIFT 0x14 +#define TCC_DSM_CNTL__LOW_RATE_TAG_IRRITATOR_DATA_SEL__SHIFT 0x15 +#define TCC_DSM_CNTL__LOW_RATE_TAG_IRRITATOR_SINGLE_WRITE__SHIFT 0x17 +#define TCC_DSM_CNTL__IN_USE_DEC_IRRITATOR_DATA_SEL__SHIFT 0x18 +#define TCC_DSM_CNTL__IN_USE_DEC_IRRITATOR_SINGLE_WRITE__SHIFT 0x1a +#define TCC_DSM_CNTL__IN_USE_TRANSFER_IRRITATOR_DATA_SEL__SHIFT 0x1b +#define TCC_DSM_CNTL__IN_USE_TRANSFER_IRRITATOR_SINGLE_WRITE__SHIFT 0x1d +#define TCC_DSM_CNTL__CACHE_DATA_IRRITATOR_DATA_SEL_MASK 0x00000003L +#define TCC_DSM_CNTL__CACHE_DATA_IRRITATOR_SINGLE_WRITE_MASK 0x00000004L +#define TCC_DSM_CNTL__CACHE_DATA_BANK_0_1_IRRITATOR_DATA_SEL_MASK 0x00000018L +#define TCC_DSM_CNTL__CACHE_DATA_BANK_0_1_IRRITATOR_SINGLE_WRITE_MASK 0x00000020L +#define TCC_DSM_CNTL__CACHE_DATA_BANK_1_0_IRRITATOR_DATA_SEL_MASK 0x000000C0L +#define TCC_DSM_CNTL__CACHE_DATA_BANK_1_0_IRRITATOR_SINGLE_WRITE_MASK 0x00000100L +#define TCC_DSM_CNTL__CACHE_DATA_BANK_1_1_IRRITATOR_DATA_SEL_MASK 0x00000600L +#define TCC_DSM_CNTL__CACHE_DATA_BANK_1_1_IRRITATOR_SINGLE_WRITE_MASK 0x00000800L +#define TCC_DSM_CNTL__CACHE_DIRTY_BANK_0_IRRITATOR_DATA_SEL_MASK 0x00003000L +#define TCC_DSM_CNTL__CACHE_DIRTY_BANK_0_IRRITATOR_SINGLE_WRITE_MASK 0x00004000L +#define TCC_DSM_CNTL__CACHE_DIRTY_BANK_1_IRRITATOR_DATA_SEL_MASK 0x00018000L +#define TCC_DSM_CNTL__CACHE_DIRTY_BANK_1_IRRITATOR_SINGLE_WRITE_MASK 0x00020000L +#define TCC_DSM_CNTL__HIGH_RATE_TAG_IRRITATOR_DATA_SEL_MASK 0x000C0000L +#define TCC_DSM_CNTL__HIGH_RATE_TAG_IRRITATOR_SINGLE_WRITE_MASK 0x00100000L +#define TCC_DSM_CNTL__LOW_RATE_TAG_IRRITATOR_DATA_SEL_MASK 0x00600000L +#define TCC_DSM_CNTL__LOW_RATE_TAG_IRRITATOR_SINGLE_WRITE_MASK 0x00800000L +#define TCC_DSM_CNTL__IN_USE_DEC_IRRITATOR_DATA_SEL_MASK 0x03000000L +#define TCC_DSM_CNTL__IN_USE_DEC_IRRITATOR_SINGLE_WRITE_MASK 0x04000000L +#define TCC_DSM_CNTL__IN_USE_TRANSFER_IRRITATOR_DATA_SEL_MASK 0x18000000L +#define TCC_DSM_CNTL__IN_USE_TRANSFER_IRRITATOR_SINGLE_WRITE_MASK 0x20000000L +//TCC_DSM_CNTLA +#define TCC_DSM_CNTLA__SRC_FIFO_IRRITATOR_DATA_SEL__SHIFT 0x0 +#define TCC_DSM_CNTLA__SRC_FIFO_IRRITATOR_SINGLE_WRITE__SHIFT 0x2 +#define TCC_DSM_CNTLA__UC_ATOMIC_FIFO_IRRITATOR_DATA_SEL__SHIFT 0x3 +#define TCC_DSM_CNTLA__UC_ATOMIC_FIFO_IRRITATOR_SINGLE_WRITE__SHIFT 0x5 +#define TCC_DSM_CNTLA__WRITE_RETURN_IRRITATOR_DATA_SEL__SHIFT 0x6 +#define TCC_DSM_CNTLA__WRITE_RETURN_IRRITATOR_SINGLE_WRITE__SHIFT 0x8 +#define TCC_DSM_CNTLA__WRITE_CACHE_READ_IRRITATOR_DATA_SEL__SHIFT 0x9 +#define TCC_DSM_CNTLA__WRITE_CACHE_READ_IRRITATOR_SINGLE_WRITE__SHIFT 0xb +#define TCC_DSM_CNTLA__SRC_FIFO_NEXT_RAM_IRRITATOR_DATA_SEL__SHIFT 0xc +#define TCC_DSM_CNTLA__SRC_FIFO_NEXT_RAM_IRRITATOR_SINGLE_WRITE__SHIFT 0xe +#define TCC_DSM_CNTLA__LATENCY_FIFO_NEXT_RAM_IRRITATOR_DATA_SEL__SHIFT 0xf +#define TCC_DSM_CNTLA__LATENCY_FIFO_NEXT_RAM_IRRITATOR_SINGLE_WRITE__SHIFT 0x11 +#define TCC_DSM_CNTLA__CACHE_TAG_PROBE_FIFO_IRRITATOR_DATA_SEL__SHIFT 0x12 +#define TCC_DSM_CNTLA__CACHE_TAG_PROBE_FIFO_IRRITATOR_SINGLE_WRITE__SHIFT 0x14 +#define TCC_DSM_CNTLA__LATENCY_FIFO_IRRITATOR_DATA_SEL__SHIFT 0x15 +#define TCC_DSM_CNTLA__LATENCY_FIFO_IRRITATOR_SINGLE_WRITE__SHIFT 0x17 +#define TCC_DSM_CNTLA__RETURN_DATA_IRRITATOR_DATA_SEL__SHIFT 0x18 +#define TCC_DSM_CNTLA__RETURN_DATA_IRRITATOR_SINGLE_WRITE__SHIFT 0x1a +#define TCC_DSM_CNTLA__RETURN_CONTROL_IRRITATOR_DATA_SEL__SHIFT 0x1b +#define TCC_DSM_CNTLA__RETURN_CONTROL_IRRITATOR_SINGLE_WRITE__SHIFT 0x1d +#define TCC_DSM_CNTLA__SRC_FIFO_IRRITATOR_DATA_SEL_MASK 0x00000003L +#define TCC_DSM_CNTLA__SRC_FIFO_IRRITATOR_SINGLE_WRITE_MASK 0x00000004L +#define TCC_DSM_CNTLA__UC_ATOMIC_FIFO_IRRITATOR_DATA_SEL_MASK 0x00000018L +#define TCC_DSM_CNTLA__UC_ATOMIC_FIFO_IRRITATOR_SINGLE_WRITE_MASK 0x00000020L +#define TCC_DSM_CNTLA__WRITE_RETURN_IRRITATOR_DATA_SEL_MASK 0x000000C0L +#define TCC_DSM_CNTLA__WRITE_RETURN_IRRITATOR_SINGLE_WRITE_MASK 0x00000100L +#define TCC_DSM_CNTLA__WRITE_CACHE_READ_IRRITATOR_DATA_SEL_MASK 0x00000600L +#define TCC_DSM_CNTLA__WRITE_CACHE_READ_IRRITATOR_SINGLE_WRITE_MASK 0x00000800L +#define TCC_DSM_CNTLA__SRC_FIFO_NEXT_RAM_IRRITATOR_DATA_SEL_MASK 0x00003000L +#define TCC_DSM_CNTLA__SRC_FIFO_NEXT_RAM_IRRITATOR_SINGLE_WRITE_MASK 0x00004000L +#define TCC_DSM_CNTLA__LATENCY_FIFO_NEXT_RAM_IRRITATOR_DATA_SEL_MASK 0x00018000L +#define TCC_DSM_CNTLA__LATENCY_FIFO_NEXT_RAM_IRRITATOR_SINGLE_WRITE_MASK 0x00020000L +#define TCC_DSM_CNTLA__CACHE_TAG_PROBE_FIFO_IRRITATOR_DATA_SEL_MASK 0x000C0000L +#define TCC_DSM_CNTLA__CACHE_TAG_PROBE_FIFO_IRRITATOR_SINGLE_WRITE_MASK 0x00100000L +#define TCC_DSM_CNTLA__LATENCY_FIFO_IRRITATOR_DATA_SEL_MASK 0x00600000L +#define TCC_DSM_CNTLA__LATENCY_FIFO_IRRITATOR_SINGLE_WRITE_MASK 0x00800000L +#define TCC_DSM_CNTLA__RETURN_DATA_IRRITATOR_DATA_SEL_MASK 0x03000000L +#define TCC_DSM_CNTLA__RETURN_DATA_IRRITATOR_SINGLE_WRITE_MASK 0x04000000L +#define TCC_DSM_CNTLA__RETURN_CONTROL_IRRITATOR_DATA_SEL_MASK 0x18000000L +#define TCC_DSM_CNTLA__RETURN_CONTROL_IRRITATOR_SINGLE_WRITE_MASK 0x20000000L +//TCC_DSM_CNTL2 +#define TCC_DSM_CNTL2__CACHE_DATA_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define TCC_DSM_CNTL2__CACHE_DATA_SELECT_INJECT_DELAY__SHIFT 0x2 +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_0_1_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_0_1_SELECT_INJECT_DELAY__SHIFT 0x5 +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_1_0_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_1_0_SELECT_INJECT_DELAY__SHIFT 0x8 +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_1_1_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_1_1_SELECT_INJECT_DELAY__SHIFT 0xb +#define TCC_DSM_CNTL2__CACHE_DIRTY_BANK_0_ENABLE_ERROR_INJECT__SHIFT 0xc +#define TCC_DSM_CNTL2__CACHE_DIRTY_BANK_0_SELECT_INJECT_DELAY__SHIFT 0xe +#define TCC_DSM_CNTL2__CACHE_DIRTY_BANK_1_ENABLE_ERROR_INJECT__SHIFT 0xf +#define TCC_DSM_CNTL2__CACHE_DIRTY_BANK_1_SELECT_INJECT_DELAY__SHIFT 0x11 +#define TCC_DSM_CNTL2__HIGH_RATE_TAG_ENABLE_ERROR_INJECT__SHIFT 0x12 +#define TCC_DSM_CNTL2__HIGH_RATE_TAG_SELECT_INJECT_DELAY__SHIFT 0x14 +#define TCC_DSM_CNTL2__LOW_RATE_TAG_ENABLE_ERROR_INJECT__SHIFT 0x15 +#define TCC_DSM_CNTL2__LOW_RATE_TAG_SELECT_INJECT_DELAY__SHIFT 0x17 +#define TCC_DSM_CNTL2__INJECT_DELAY__SHIFT 0x1a +#define TCC_DSM_CNTL2__CACHE_DATA_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define TCC_DSM_CNTL2__CACHE_DATA_SELECT_INJECT_DELAY_MASK 0x00000004L +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_0_1_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_0_1_SELECT_INJECT_DELAY_MASK 0x00000020L +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_1_0_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_1_0_SELECT_INJECT_DELAY_MASK 0x00000100L +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_1_1_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define TCC_DSM_CNTL2__CACHE_DATA_BANK_1_1_SELECT_INJECT_DELAY_MASK 0x00000800L +#define TCC_DSM_CNTL2__CACHE_DIRTY_BANK_0_ENABLE_ERROR_INJECT_MASK 0x00003000L +#define TCC_DSM_CNTL2__CACHE_DIRTY_BANK_0_SELECT_INJECT_DELAY_MASK 0x00004000L +#define TCC_DSM_CNTL2__CACHE_DIRTY_BANK_1_ENABLE_ERROR_INJECT_MASK 0x00018000L +#define TCC_DSM_CNTL2__CACHE_DIRTY_BANK_1_SELECT_INJECT_DELAY_MASK 0x00020000L +#define TCC_DSM_CNTL2__HIGH_RATE_TAG_ENABLE_ERROR_INJECT_MASK 0x000C0000L +#define TCC_DSM_CNTL2__HIGH_RATE_TAG_SELECT_INJECT_DELAY_MASK 0x00100000L +#define TCC_DSM_CNTL2__LOW_RATE_TAG_ENABLE_ERROR_INJECT_MASK 0x00600000L +#define TCC_DSM_CNTL2__LOW_RATE_TAG_SELECT_INJECT_DELAY_MASK 0x00800000L +#define TCC_DSM_CNTL2__INJECT_DELAY_MASK 0xFC000000L +//TCC_DSM_CNTL2A +#define TCC_DSM_CNTL2A__IN_USE_DEC_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define TCC_DSM_CNTL2A__IN_USE_DEC_SELECT_INJECT_DELAY__SHIFT 0x2 +#define TCC_DSM_CNTL2A__IN_USE_TRANSFER_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define TCC_DSM_CNTL2A__IN_USE_TRANSFER_SELECT_INJECT_DELAY__SHIFT 0x5 +#define TCC_DSM_CNTL2A__RETURN_DATA_ENABLE_ERROR_INJECT__SHIFT 0x6 +#define TCC_DSM_CNTL2A__RETURN_DATA_SELECT_INJECT_DELAY__SHIFT 0x8 +#define TCC_DSM_CNTL2A__RETURN_CONTROL_ENABLE_ERROR_INJECT__SHIFT 0x9 +#define TCC_DSM_CNTL2A__RETURN_CONTROL_SELECT_INJECT_DELAY__SHIFT 0xb +#define TCC_DSM_CNTL2A__UC_ATOMIC_FIFO_ENABLE_ERROR_INJECT__SHIFT 0xc +#define TCC_DSM_CNTL2A__UC_ATOMIC_FIFO_SELECT_INJECT_DELAY__SHIFT 0xe +#define TCC_DSM_CNTL2A__WRITE_RETURN_ENABLE_ERROR_INJECT__SHIFT 0xf +#define TCC_DSM_CNTL2A__WRITE_RETURN_SELECT_INJECT_DELAY__SHIFT 0x11 +#define TCC_DSM_CNTL2A__WRITE_CACHE_READ_ENABLE_ERROR_INJECT__SHIFT 0x12 +#define TCC_DSM_CNTL2A__WRITE_CACHE_READ_SELECT_INJECT_DELAY__SHIFT 0x14 +#define TCC_DSM_CNTL2A__SRC_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x15 +#define TCC_DSM_CNTL2A__SRC_FIFO_SELECT_INJECT_DELAY__SHIFT 0x17 +#define TCC_DSM_CNTL2A__SRC_FIFO_NEXT_RAM_ENABLE_ERROR_INJECT__SHIFT 0x18 +#define TCC_DSM_CNTL2A__SRC_FIFO_NEXT_RAM_SELECT_INJECT_DELAY__SHIFT 0x1a +#define TCC_DSM_CNTL2A__CACHE_TAG_PROBE_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x1b +#define TCC_DSM_CNTL2A__CACHE_TAG_PROBE_FIFO_SELECT_INJECT_DELAY__SHIFT 0x1d +#define TCC_DSM_CNTL2A__IN_USE_DEC_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define TCC_DSM_CNTL2A__IN_USE_DEC_SELECT_INJECT_DELAY_MASK 0x00000004L +#define TCC_DSM_CNTL2A__IN_USE_TRANSFER_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define TCC_DSM_CNTL2A__IN_USE_TRANSFER_SELECT_INJECT_DELAY_MASK 0x00000020L +#define TCC_DSM_CNTL2A__RETURN_DATA_ENABLE_ERROR_INJECT_MASK 0x000000C0L +#define TCC_DSM_CNTL2A__RETURN_DATA_SELECT_INJECT_DELAY_MASK 0x00000100L +#define TCC_DSM_CNTL2A__RETURN_CONTROL_ENABLE_ERROR_INJECT_MASK 0x00000600L +#define TCC_DSM_CNTL2A__RETURN_CONTROL_SELECT_INJECT_DELAY_MASK 0x00000800L +#define TCC_DSM_CNTL2A__UC_ATOMIC_FIFO_ENABLE_ERROR_INJECT_MASK 0x00003000L +#define TCC_DSM_CNTL2A__UC_ATOMIC_FIFO_SELECT_INJECT_DELAY_MASK 0x00004000L +#define TCC_DSM_CNTL2A__WRITE_RETURN_ENABLE_ERROR_INJECT_MASK 0x00018000L +#define TCC_DSM_CNTL2A__WRITE_RETURN_SELECT_INJECT_DELAY_MASK 0x00020000L +#define TCC_DSM_CNTL2A__WRITE_CACHE_READ_ENABLE_ERROR_INJECT_MASK 0x000C0000L +#define TCC_DSM_CNTL2A__WRITE_CACHE_READ_SELECT_INJECT_DELAY_MASK 0x00100000L +#define TCC_DSM_CNTL2A__SRC_FIFO_ENABLE_ERROR_INJECT_MASK 0x00600000L +#define TCC_DSM_CNTL2A__SRC_FIFO_SELECT_INJECT_DELAY_MASK 0x00800000L +#define TCC_DSM_CNTL2A__SRC_FIFO_NEXT_RAM_ENABLE_ERROR_INJECT_MASK 0x03000000L +#define TCC_DSM_CNTL2A__SRC_FIFO_NEXT_RAM_SELECT_INJECT_DELAY_MASK 0x04000000L +#define TCC_DSM_CNTL2A__CACHE_TAG_PROBE_FIFO_ENABLE_ERROR_INJECT_MASK 0x18000000L +#define TCC_DSM_CNTL2A__CACHE_TAG_PROBE_FIFO_SELECT_INJECT_DELAY_MASK 0x20000000L +//TCC_DSM_CNTL2B +#define TCC_DSM_CNTL2B__LATENCY_FIFO_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define TCC_DSM_CNTL2B__LATENCY_FIFO_SELECT_INJECT_DELAY__SHIFT 0x2 +#define TCC_DSM_CNTL2B__LATENCY_FIFO_NEXT_RAM_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define TCC_DSM_CNTL2B__LATENCY_FIFO_NEXT_RAM_SELECT_INJECT_DELAY__SHIFT 0x5 +#define TCC_DSM_CNTL2B__LATENCY_FIFO_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define TCC_DSM_CNTL2B__LATENCY_FIFO_SELECT_INJECT_DELAY_MASK 0x00000004L +#define TCC_DSM_CNTL2B__LATENCY_FIFO_NEXT_RAM_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define TCC_DSM_CNTL2B__LATENCY_FIFO_NEXT_RAM_SELECT_INJECT_DELAY_MASK 0x00000020L +//TCC_WBINVL2 +#define TCC_WBINVL2__DONE__SHIFT 0x4 +#define TCC_WBINVL2__DONE_MASK 0x00000010L +//TCC_SOFT_RESET +#define TCC_SOFT_RESET__HALT_FOR_RESET__SHIFT 0x0 +#define TCC_SOFT_RESET__HALT_FOR_RESET_MASK 0x00000001L +//TCA_CTRL +#define TCA_CTRL__HOLE_TIMEOUT__SHIFT 0x0 +#define TCA_CTRL__RB_STILL_4_PHASE__SHIFT 0x4 +#define TCA_CTRL__RB_AS_TCI__SHIFT 0x5 +#define TCA_CTRL__DISABLE_UTCL2_PRIORITY__SHIFT 0x6 +#define TCA_CTRL__DISABLE_RB_ONLY_TCA_ARBITER__SHIFT 0x7 +#define TCA_CTRL__HOLE_TIMEOUT_MASK 0x0000000FL +#define TCA_CTRL__RB_STILL_4_PHASE_MASK 0x00000010L +#define TCA_CTRL__RB_AS_TCI_MASK 0x00000020L +#define TCA_CTRL__DISABLE_UTCL2_PRIORITY_MASK 0x00000040L +#define TCA_CTRL__DISABLE_RB_ONLY_TCA_ARBITER_MASK 0x00000080L +//TCA_BURST_MASK +#define TCA_BURST_MASK__ADDR_MASK__SHIFT 0x0 +#define TCA_BURST_MASK__ADDR_MASK_MASK 0xFFFFFFFFL +//TCA_BURST_CTRL +#define TCA_BURST_CTRL__MAX_BURST__SHIFT 0x0 +#define TCA_BURST_CTRL__RB_DISABLE__SHIFT 0x3 +#define TCA_BURST_CTRL__TCP_DISABLE__SHIFT 0x4 +#define TCA_BURST_CTRL__SQC_DISABLE__SHIFT 0x5 +#define TCA_BURST_CTRL__CPF_DISABLE__SHIFT 0x6 +#define TCA_BURST_CTRL__CPG_DISABLE__SHIFT 0x7 +#define TCA_BURST_CTRL__IA_DISABLE__SHIFT 0x8 +#define TCA_BURST_CTRL__WD_DISABLE__SHIFT 0x9 +#define TCA_BURST_CTRL__SQG_DISABLE__SHIFT 0xa +#define TCA_BURST_CTRL__UTCL2_DISABLE__SHIFT 0xb +#define TCA_BURST_CTRL__TPI_DISABLE__SHIFT 0xc +#define TCA_BURST_CTRL__RLC_DISABLE__SHIFT 0xd +#define TCA_BURST_CTRL__PA_DISABLE__SHIFT 0xe +#define TCA_BURST_CTRL__MAX_BURST_MASK 0x00000007L +#define TCA_BURST_CTRL__RB_DISABLE_MASK 0x00000008L +#define TCA_BURST_CTRL__TCP_DISABLE_MASK 0x00000010L +#define TCA_BURST_CTRL__SQC_DISABLE_MASK 0x00000020L +#define TCA_BURST_CTRL__CPF_DISABLE_MASK 0x00000040L +#define TCA_BURST_CTRL__CPG_DISABLE_MASK 0x00000080L +#define TCA_BURST_CTRL__IA_DISABLE_MASK 0x00000100L +#define TCA_BURST_CTRL__WD_DISABLE_MASK 0x00000200L +#define TCA_BURST_CTRL__SQG_DISABLE_MASK 0x00000400L +#define TCA_BURST_CTRL__UTCL2_DISABLE_MASK 0x00000800L +#define TCA_BURST_CTRL__TPI_DISABLE_MASK 0x00001000L +#define TCA_BURST_CTRL__RLC_DISABLE_MASK 0x00002000L +#define TCA_BURST_CTRL__PA_DISABLE_MASK 0x00004000L +//TCA_DSM_CNTL +#define TCA_DSM_CNTL__HOLE_FIFO_SED_IRRITATOR_DATA_SEL__SHIFT 0x0 +#define TCA_DSM_CNTL__HOLE_FIFO_SED_IRRITATOR_SINGLE_WRITE__SHIFT 0x2 +#define TCA_DSM_CNTL__REQ_FIFO_SED_IRRITATOR_DATA_SEL__SHIFT 0x3 +#define TCA_DSM_CNTL__REQ_FIFO_SED_IRRITATOR_SINGLE_WRITE__SHIFT 0x5 +#define TCA_DSM_CNTL__HOLE_FIFO_SED_IRRITATOR_DATA_SEL_MASK 0x00000003L +#define TCA_DSM_CNTL__HOLE_FIFO_SED_IRRITATOR_SINGLE_WRITE_MASK 0x00000004L +#define TCA_DSM_CNTL__REQ_FIFO_SED_IRRITATOR_DATA_SEL_MASK 0x00000018L +#define TCA_DSM_CNTL__REQ_FIFO_SED_IRRITATOR_SINGLE_WRITE_MASK 0x00000020L +//TCA_DSM_CNTL2 +#define TCA_DSM_CNTL2__HOLE_FIFO_SED_ENABLE_ERROR_INJECT__SHIFT 0x0 +#define TCA_DSM_CNTL2__HOLE_FIFO_SED_SELECT_INJECT_DELAY__SHIFT 0x2 +#define TCA_DSM_CNTL2__REQ_FIFO_SED_ENABLE_ERROR_INJECT__SHIFT 0x3 +#define TCA_DSM_CNTL2__REQ_FIFO_SED_SELECT_INJECT_DELAY__SHIFT 0x5 +#define TCA_DSM_CNTL2__INJECT_DELAY__SHIFT 0x1a +#define TCA_DSM_CNTL2__HOLE_FIFO_SED_ENABLE_ERROR_INJECT_MASK 0x00000003L +#define TCA_DSM_CNTL2__HOLE_FIFO_SED_SELECT_INJECT_DELAY_MASK 0x00000004L +#define TCA_DSM_CNTL2__REQ_FIFO_SED_ENABLE_ERROR_INJECT_MASK 0x00000018L +#define TCA_DSM_CNTL2__REQ_FIFO_SED_SELECT_INJECT_DELAY_MASK 0x00000020L +#define TCA_DSM_CNTL2__INJECT_DELAY_MASK 0xFC000000L +//TCA_EDC_CNT +#define TCA_EDC_CNT__HOLE_FIFO_SED_COUNT__SHIFT 0x0 +#define TCA_EDC_CNT__REQ_FIFO_SED_COUNT__SHIFT 0x2 +#define TCA_EDC_CNT__HOLE_FIFO_SED_COUNT_MASK 0x00000003L +#define TCA_EDC_CNT__REQ_FIFO_SED_COUNT_MASK 0x0000000CL + + +// addressBlock: gc_shdec +//SPI_SHADER_PGM_RSRC3_PS +#define SPI_SHADER_PGM_RSRC3_PS__CU_EN__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC3_PS__WAVE_LIMIT__SHIFT 0x10 +#define SPI_SHADER_PGM_RSRC3_PS__LOCK_LOW_THRESHOLD__SHIFT 0x16 +#define SPI_SHADER_PGM_RSRC3_PS__SIMD_DISABLE__SHIFT 0x1a +#define SPI_SHADER_PGM_RSRC3_PS__CU_EN_MASK 0x0000FFFFL +#define SPI_SHADER_PGM_RSRC3_PS__WAVE_LIMIT_MASK 0x003F0000L +#define SPI_SHADER_PGM_RSRC3_PS__LOCK_LOW_THRESHOLD_MASK 0x03C00000L +#define SPI_SHADER_PGM_RSRC3_PS__SIMD_DISABLE_MASK 0x3C000000L +//SPI_SHADER_PGM_LO_PS +#define SPI_SHADER_PGM_LO_PS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_LO_PS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_HI_PS +#define SPI_SHADER_PGM_HI_PS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_HI_PS__MEM_BASE_MASK 0xFFL +//SPI_SHADER_PGM_RSRC1_PS +#define SPI_SHADER_PGM_RSRC1_PS__VGPRS__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC1_PS__SGPRS__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC1_PS__PRIORITY__SHIFT 0xa +#define SPI_SHADER_PGM_RSRC1_PS__FLOAT_MODE__SHIFT 0xc +#define SPI_SHADER_PGM_RSRC1_PS__PRIV__SHIFT 0x14 +#define SPI_SHADER_PGM_RSRC1_PS__DX10_CLAMP__SHIFT 0x15 +#define SPI_SHADER_PGM_RSRC1_PS__IEEE_MODE__SHIFT 0x17 +#define SPI_SHADER_PGM_RSRC1_PS__CU_GROUP_DISABLE__SHIFT 0x18 +#define SPI_SHADER_PGM_RSRC1_PS__FP16_OVFL__SHIFT 0x1d +#define SPI_SHADER_PGM_RSRC1_PS__VGPRS_MASK 0x0000003FL +#define SPI_SHADER_PGM_RSRC1_PS__SGPRS_MASK 0x000003C0L +#define SPI_SHADER_PGM_RSRC1_PS__PRIORITY_MASK 0x00000C00L +#define SPI_SHADER_PGM_RSRC1_PS__FLOAT_MODE_MASK 0x000FF000L +#define SPI_SHADER_PGM_RSRC1_PS__PRIV_MASK 0x00100000L +#define SPI_SHADER_PGM_RSRC1_PS__DX10_CLAMP_MASK 0x00200000L +#define SPI_SHADER_PGM_RSRC1_PS__IEEE_MODE_MASK 0x00800000L +#define SPI_SHADER_PGM_RSRC1_PS__CU_GROUP_DISABLE_MASK 0x01000000L +#define SPI_SHADER_PGM_RSRC1_PS__FP16_OVFL_MASK 0x20000000L +//SPI_SHADER_PGM_RSRC2_PS +#define SPI_SHADER_PGM_RSRC2_PS__SCRATCH_EN__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC2_PS__USER_SGPR__SHIFT 0x1 +#define SPI_SHADER_PGM_RSRC2_PS__TRAP_PRESENT__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC2_PS__WAVE_CNT_EN__SHIFT 0x7 +#define SPI_SHADER_PGM_RSRC2_PS__EXTRA_LDS_SIZE__SHIFT 0x8 +#define SPI_SHADER_PGM_RSRC2_PS__EXCP_EN__SHIFT 0x10 +#define SPI_SHADER_PGM_RSRC2_PS__LOAD_COLLISION_WAVEID__SHIFT 0x19 +#define SPI_SHADER_PGM_RSRC2_PS__LOAD_INTRAWAVE_COLLISION__SHIFT 0x1a +#define SPI_SHADER_PGM_RSRC2_PS__SKIP_USGPR0__SHIFT 0x1b +#define SPI_SHADER_PGM_RSRC2_PS__USER_SGPR_MSB__SHIFT 0x1c +#define SPI_SHADER_PGM_RSRC2_PS__SCRATCH_EN_MASK 0x00000001L +#define SPI_SHADER_PGM_RSRC2_PS__USER_SGPR_MASK 0x0000003EL +#define SPI_SHADER_PGM_RSRC2_PS__TRAP_PRESENT_MASK 0x00000040L +#define SPI_SHADER_PGM_RSRC2_PS__WAVE_CNT_EN_MASK 0x00000080L +#define SPI_SHADER_PGM_RSRC2_PS__EXTRA_LDS_SIZE_MASK 0x0000FF00L +#define SPI_SHADER_PGM_RSRC2_PS__EXCP_EN_MASK 0x01FF0000L +#define SPI_SHADER_PGM_RSRC2_PS__LOAD_COLLISION_WAVEID_MASK 0x02000000L +#define SPI_SHADER_PGM_RSRC2_PS__LOAD_INTRAWAVE_COLLISION_MASK 0x04000000L +#define SPI_SHADER_PGM_RSRC2_PS__SKIP_USGPR0_MASK 0x08000000L +#define SPI_SHADER_PGM_RSRC2_PS__USER_SGPR_MSB_MASK 0x10000000L +//SPI_SHADER_USER_DATA_PS_0 +#define SPI_SHADER_USER_DATA_PS_0__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_0__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_1 +#define SPI_SHADER_USER_DATA_PS_1__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_1__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_2 +#define SPI_SHADER_USER_DATA_PS_2__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_2__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_3 +#define SPI_SHADER_USER_DATA_PS_3__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_3__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_4 +#define SPI_SHADER_USER_DATA_PS_4__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_4__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_5 +#define SPI_SHADER_USER_DATA_PS_5__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_5__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_6 +#define SPI_SHADER_USER_DATA_PS_6__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_6__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_7 +#define SPI_SHADER_USER_DATA_PS_7__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_7__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_8 +#define SPI_SHADER_USER_DATA_PS_8__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_8__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_9 +#define SPI_SHADER_USER_DATA_PS_9__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_9__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_10 +#define SPI_SHADER_USER_DATA_PS_10__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_10__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_11 +#define SPI_SHADER_USER_DATA_PS_11__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_11__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_12 +#define SPI_SHADER_USER_DATA_PS_12__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_12__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_13 +#define SPI_SHADER_USER_DATA_PS_13__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_13__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_14 +#define SPI_SHADER_USER_DATA_PS_14__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_14__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_15 +#define SPI_SHADER_USER_DATA_PS_15__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_15__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_16 +#define SPI_SHADER_USER_DATA_PS_16__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_16__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_17 +#define SPI_SHADER_USER_DATA_PS_17__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_17__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_18 +#define SPI_SHADER_USER_DATA_PS_18__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_18__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_19 +#define SPI_SHADER_USER_DATA_PS_19__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_19__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_20 +#define SPI_SHADER_USER_DATA_PS_20__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_20__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_21 +#define SPI_SHADER_USER_DATA_PS_21__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_21__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_22 +#define SPI_SHADER_USER_DATA_PS_22__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_22__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_23 +#define SPI_SHADER_USER_DATA_PS_23__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_23__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_24 +#define SPI_SHADER_USER_DATA_PS_24__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_24__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_25 +#define SPI_SHADER_USER_DATA_PS_25__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_25__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_26 +#define SPI_SHADER_USER_DATA_PS_26__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_26__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_27 +#define SPI_SHADER_USER_DATA_PS_27__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_27__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_28 +#define SPI_SHADER_USER_DATA_PS_28__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_28__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_29 +#define SPI_SHADER_USER_DATA_PS_29__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_29__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_30 +#define SPI_SHADER_USER_DATA_PS_30__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_30__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_PS_31 +#define SPI_SHADER_USER_DATA_PS_31__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_PS_31__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_RSRC3_VS +#define SPI_SHADER_PGM_RSRC3_VS__CU_EN__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC3_VS__WAVE_LIMIT__SHIFT 0x10 +#define SPI_SHADER_PGM_RSRC3_VS__LOCK_LOW_THRESHOLD__SHIFT 0x16 +#define SPI_SHADER_PGM_RSRC3_VS__SIMD_DISABLE__SHIFT 0x1a +#define SPI_SHADER_PGM_RSRC3_VS__CU_EN_MASK 0x0000FFFFL +#define SPI_SHADER_PGM_RSRC3_VS__WAVE_LIMIT_MASK 0x003F0000L +#define SPI_SHADER_PGM_RSRC3_VS__LOCK_LOW_THRESHOLD_MASK 0x03C00000L +#define SPI_SHADER_PGM_RSRC3_VS__SIMD_DISABLE_MASK 0x3C000000L +//SPI_SHADER_LATE_ALLOC_VS +#define SPI_SHADER_LATE_ALLOC_VS__LIMIT__SHIFT 0x0 +#define SPI_SHADER_LATE_ALLOC_VS__LIMIT_MASK 0x0000003FL +//SPI_SHADER_PGM_LO_VS +#define SPI_SHADER_PGM_LO_VS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_LO_VS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_HI_VS +#define SPI_SHADER_PGM_HI_VS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_HI_VS__MEM_BASE_MASK 0xFFL +//SPI_SHADER_PGM_RSRC1_VS +#define SPI_SHADER_PGM_RSRC1_VS__VGPRS__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC1_VS__SGPRS__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC1_VS__PRIORITY__SHIFT 0xa +#define SPI_SHADER_PGM_RSRC1_VS__FLOAT_MODE__SHIFT 0xc +#define SPI_SHADER_PGM_RSRC1_VS__PRIV__SHIFT 0x14 +#define SPI_SHADER_PGM_RSRC1_VS__DX10_CLAMP__SHIFT 0x15 +#define SPI_SHADER_PGM_RSRC1_VS__IEEE_MODE__SHIFT 0x17 +#define SPI_SHADER_PGM_RSRC1_VS__VGPR_COMP_CNT__SHIFT 0x18 +#define SPI_SHADER_PGM_RSRC1_VS__CU_GROUP_ENABLE__SHIFT 0x1a +#define SPI_SHADER_PGM_RSRC1_VS__FP16_OVFL__SHIFT 0x1f +#define SPI_SHADER_PGM_RSRC1_VS__VGPRS_MASK 0x0000003FL +#define SPI_SHADER_PGM_RSRC1_VS__SGPRS_MASK 0x000003C0L +#define SPI_SHADER_PGM_RSRC1_VS__PRIORITY_MASK 0x00000C00L +#define SPI_SHADER_PGM_RSRC1_VS__FLOAT_MODE_MASK 0x000FF000L +#define SPI_SHADER_PGM_RSRC1_VS__PRIV_MASK 0x00100000L +#define SPI_SHADER_PGM_RSRC1_VS__DX10_CLAMP_MASK 0x00200000L +#define SPI_SHADER_PGM_RSRC1_VS__IEEE_MODE_MASK 0x00800000L +#define SPI_SHADER_PGM_RSRC1_VS__VGPR_COMP_CNT_MASK 0x03000000L +#define SPI_SHADER_PGM_RSRC1_VS__CU_GROUP_ENABLE_MASK 0x04000000L +#define SPI_SHADER_PGM_RSRC1_VS__FP16_OVFL_MASK 0x80000000L +//SPI_SHADER_PGM_RSRC2_VS +#define SPI_SHADER_PGM_RSRC2_VS__SCRATCH_EN__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC2_VS__USER_SGPR__SHIFT 0x1 +#define SPI_SHADER_PGM_RSRC2_VS__TRAP_PRESENT__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC2_VS__OC_LDS_EN__SHIFT 0x7 +#define SPI_SHADER_PGM_RSRC2_VS__SO_BASE0_EN__SHIFT 0x8 +#define SPI_SHADER_PGM_RSRC2_VS__SO_BASE1_EN__SHIFT 0x9 +#define SPI_SHADER_PGM_RSRC2_VS__SO_BASE2_EN__SHIFT 0xa +#define SPI_SHADER_PGM_RSRC2_VS__SO_BASE3_EN__SHIFT 0xb +#define SPI_SHADER_PGM_RSRC2_VS__SO_EN__SHIFT 0xc +#define SPI_SHADER_PGM_RSRC2_VS__EXCP_EN__SHIFT 0xd +#define SPI_SHADER_PGM_RSRC2_VS__PC_BASE_EN__SHIFT 0x16 +#define SPI_SHADER_PGM_RSRC2_VS__DISPATCH_DRAW_EN__SHIFT 0x18 +#define SPI_SHADER_PGM_RSRC2_VS__SKIP_USGPR0__SHIFT 0x1b +#define SPI_SHADER_PGM_RSRC2_VS__USER_SGPR_MSB__SHIFT 0x1c +#define SPI_SHADER_PGM_RSRC2_VS__SCRATCH_EN_MASK 0x00000001L +#define SPI_SHADER_PGM_RSRC2_VS__USER_SGPR_MASK 0x0000003EL +#define SPI_SHADER_PGM_RSRC2_VS__TRAP_PRESENT_MASK 0x00000040L +#define SPI_SHADER_PGM_RSRC2_VS__OC_LDS_EN_MASK 0x00000080L +#define SPI_SHADER_PGM_RSRC2_VS__SO_BASE0_EN_MASK 0x00000100L +#define SPI_SHADER_PGM_RSRC2_VS__SO_BASE1_EN_MASK 0x00000200L +#define SPI_SHADER_PGM_RSRC2_VS__SO_BASE2_EN_MASK 0x00000400L +#define SPI_SHADER_PGM_RSRC2_VS__SO_BASE3_EN_MASK 0x00000800L +#define SPI_SHADER_PGM_RSRC2_VS__SO_EN_MASK 0x00001000L +#define SPI_SHADER_PGM_RSRC2_VS__EXCP_EN_MASK 0x003FE000L +#define SPI_SHADER_PGM_RSRC2_VS__PC_BASE_EN_MASK 0x00400000L +#define SPI_SHADER_PGM_RSRC2_VS__DISPATCH_DRAW_EN_MASK 0x01000000L +#define SPI_SHADER_PGM_RSRC2_VS__SKIP_USGPR0_MASK 0x08000000L +#define SPI_SHADER_PGM_RSRC2_VS__USER_SGPR_MSB_MASK 0x10000000L +//SPI_SHADER_USER_DATA_VS_0 +#define SPI_SHADER_USER_DATA_VS_0__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_0__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_1 +#define SPI_SHADER_USER_DATA_VS_1__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_1__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_2 +#define SPI_SHADER_USER_DATA_VS_2__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_2__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_3 +#define SPI_SHADER_USER_DATA_VS_3__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_3__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_4 +#define SPI_SHADER_USER_DATA_VS_4__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_4__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_5 +#define SPI_SHADER_USER_DATA_VS_5__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_5__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_6 +#define SPI_SHADER_USER_DATA_VS_6__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_6__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_7 +#define SPI_SHADER_USER_DATA_VS_7__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_7__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_8 +#define SPI_SHADER_USER_DATA_VS_8__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_8__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_9 +#define SPI_SHADER_USER_DATA_VS_9__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_9__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_10 +#define SPI_SHADER_USER_DATA_VS_10__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_10__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_11 +#define SPI_SHADER_USER_DATA_VS_11__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_11__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_12 +#define SPI_SHADER_USER_DATA_VS_12__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_12__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_13 +#define SPI_SHADER_USER_DATA_VS_13__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_13__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_14 +#define SPI_SHADER_USER_DATA_VS_14__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_14__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_15 +#define SPI_SHADER_USER_DATA_VS_15__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_15__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_16 +#define SPI_SHADER_USER_DATA_VS_16__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_16__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_17 +#define SPI_SHADER_USER_DATA_VS_17__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_17__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_18 +#define SPI_SHADER_USER_DATA_VS_18__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_18__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_19 +#define SPI_SHADER_USER_DATA_VS_19__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_19__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_20 +#define SPI_SHADER_USER_DATA_VS_20__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_20__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_21 +#define SPI_SHADER_USER_DATA_VS_21__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_21__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_22 +#define SPI_SHADER_USER_DATA_VS_22__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_22__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_23 +#define SPI_SHADER_USER_DATA_VS_23__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_23__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_24 +#define SPI_SHADER_USER_DATA_VS_24__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_24__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_25 +#define SPI_SHADER_USER_DATA_VS_25__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_25__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_26 +#define SPI_SHADER_USER_DATA_VS_26__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_26__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_27 +#define SPI_SHADER_USER_DATA_VS_27__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_27__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_28 +#define SPI_SHADER_USER_DATA_VS_28__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_28__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_29 +#define SPI_SHADER_USER_DATA_VS_29__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_29__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_30 +#define SPI_SHADER_USER_DATA_VS_30__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_30__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_VS_31 +#define SPI_SHADER_USER_DATA_VS_31__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_VS_31__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_RSRC2_GS_VS +#define SPI_SHADER_PGM_RSRC2_GS_VS__SCRATCH_EN__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC2_GS_VS__USER_SGPR__SHIFT 0x1 +#define SPI_SHADER_PGM_RSRC2_GS_VS__TRAP_PRESENT__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC2_GS_VS__EXCP_EN__SHIFT 0x7 +#define SPI_SHADER_PGM_RSRC2_GS_VS__VGPR_COMP_CNT__SHIFT 0x10 +#define SPI_SHADER_PGM_RSRC2_GS_VS__OC_LDS_EN__SHIFT 0x12 +#define SPI_SHADER_PGM_RSRC2_GS_VS__LDS_SIZE__SHIFT 0x13 +#define SPI_SHADER_PGM_RSRC2_GS_VS__SKIP_USGPR0__SHIFT 0x1b +#define SPI_SHADER_PGM_RSRC2_GS_VS__USER_SGPR_MSB__SHIFT 0x1c +#define SPI_SHADER_PGM_RSRC2_GS_VS__SCRATCH_EN_MASK 0x00000001L +#define SPI_SHADER_PGM_RSRC2_GS_VS__USER_SGPR_MASK 0x0000003EL +#define SPI_SHADER_PGM_RSRC2_GS_VS__TRAP_PRESENT_MASK 0x00000040L +#define SPI_SHADER_PGM_RSRC2_GS_VS__EXCP_EN_MASK 0x0000FF80L +#define SPI_SHADER_PGM_RSRC2_GS_VS__VGPR_COMP_CNT_MASK 0x00030000L +#define SPI_SHADER_PGM_RSRC2_GS_VS__OC_LDS_EN_MASK 0x00040000L +#define SPI_SHADER_PGM_RSRC2_GS_VS__LDS_SIZE_MASK 0x07F80000L +#define SPI_SHADER_PGM_RSRC2_GS_VS__SKIP_USGPR0_MASK 0x08000000L +#define SPI_SHADER_PGM_RSRC2_GS_VS__USER_SGPR_MSB_MASK 0x10000000L +//SPI_SHADER_PGM_RSRC4_GS +#define SPI_SHADER_PGM_RSRC4_GS__GROUP_FIFO_DEPTH__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC4_GS__SPI_SHADER_LATE_ALLOC_GS__SHIFT 0x7 +#define SPI_SHADER_PGM_RSRC4_GS__GROUP_FIFO_DEPTH_MASK 0x0000007FL +#define SPI_SHADER_PGM_RSRC4_GS__SPI_SHADER_LATE_ALLOC_GS_MASK 0x00003F80L +//SPI_SHADER_USER_DATA_ADDR_LO_GS +#define SPI_SHADER_USER_DATA_ADDR_LO_GS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ADDR_LO_GS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ADDR_HI_GS +#define SPI_SHADER_USER_DATA_ADDR_HI_GS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ADDR_HI_GS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_LO_ES +#define SPI_SHADER_PGM_LO_ES__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_LO_ES__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_HI_ES +#define SPI_SHADER_PGM_HI_ES__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_HI_ES__MEM_BASE_MASK 0xFFL +//SPI_SHADER_PGM_RSRC3_GS +#define SPI_SHADER_PGM_RSRC3_GS__CU_EN__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC3_GS__WAVE_LIMIT__SHIFT 0x10 +#define SPI_SHADER_PGM_RSRC3_GS__LOCK_LOW_THRESHOLD__SHIFT 0x16 +#define SPI_SHADER_PGM_RSRC3_GS__SIMD_DISABLE__SHIFT 0x1a +#define SPI_SHADER_PGM_RSRC3_GS__CU_EN_MASK 0x0000FFFFL +#define SPI_SHADER_PGM_RSRC3_GS__WAVE_LIMIT_MASK 0x003F0000L +#define SPI_SHADER_PGM_RSRC3_GS__LOCK_LOW_THRESHOLD_MASK 0x03C00000L +#define SPI_SHADER_PGM_RSRC3_GS__SIMD_DISABLE_MASK 0x3C000000L +//SPI_SHADER_PGM_LO_GS +#define SPI_SHADER_PGM_LO_GS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_LO_GS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_HI_GS +#define SPI_SHADER_PGM_HI_GS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_HI_GS__MEM_BASE_MASK 0xFFL +//SPI_SHADER_PGM_RSRC1_GS +#define SPI_SHADER_PGM_RSRC1_GS__VGPRS__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC1_GS__SGPRS__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC1_GS__PRIORITY__SHIFT 0xa +#define SPI_SHADER_PGM_RSRC1_GS__FLOAT_MODE__SHIFT 0xc +#define SPI_SHADER_PGM_RSRC1_GS__PRIV__SHIFT 0x14 +#define SPI_SHADER_PGM_RSRC1_GS__DX10_CLAMP__SHIFT 0x15 +#define SPI_SHADER_PGM_RSRC1_GS__IEEE_MODE__SHIFT 0x17 +#define SPI_SHADER_PGM_RSRC1_GS__CU_GROUP_ENABLE__SHIFT 0x18 +#define SPI_SHADER_PGM_RSRC1_GS__GS_VGPR_COMP_CNT__SHIFT 0x1d +#define SPI_SHADER_PGM_RSRC1_GS__FP16_OVFL__SHIFT 0x1f +#define SPI_SHADER_PGM_RSRC1_GS__VGPRS_MASK 0x0000003FL +#define SPI_SHADER_PGM_RSRC1_GS__SGPRS_MASK 0x000003C0L +#define SPI_SHADER_PGM_RSRC1_GS__PRIORITY_MASK 0x00000C00L +#define SPI_SHADER_PGM_RSRC1_GS__FLOAT_MODE_MASK 0x000FF000L +#define SPI_SHADER_PGM_RSRC1_GS__PRIV_MASK 0x00100000L +#define SPI_SHADER_PGM_RSRC1_GS__DX10_CLAMP_MASK 0x00200000L +#define SPI_SHADER_PGM_RSRC1_GS__IEEE_MODE_MASK 0x00800000L +#define SPI_SHADER_PGM_RSRC1_GS__CU_GROUP_ENABLE_MASK 0x01000000L +#define SPI_SHADER_PGM_RSRC1_GS__GS_VGPR_COMP_CNT_MASK 0x60000000L +#define SPI_SHADER_PGM_RSRC1_GS__FP16_OVFL_MASK 0x80000000L +//SPI_SHADER_PGM_RSRC2_GS +#define SPI_SHADER_PGM_RSRC2_GS__SCRATCH_EN__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC2_GS__USER_SGPR__SHIFT 0x1 +#define SPI_SHADER_PGM_RSRC2_GS__TRAP_PRESENT__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC2_GS__EXCP_EN__SHIFT 0x7 +#define SPI_SHADER_PGM_RSRC2_GS__ES_VGPR_COMP_CNT__SHIFT 0x10 +#define SPI_SHADER_PGM_RSRC2_GS__OC_LDS_EN__SHIFT 0x12 +#define SPI_SHADER_PGM_RSRC2_GS__LDS_SIZE__SHIFT 0x13 +#define SPI_SHADER_PGM_RSRC2_GS__SKIP_USGPR0__SHIFT 0x1b +#define SPI_SHADER_PGM_RSRC2_GS__USER_SGPR_MSB__SHIFT 0x1c +#define SPI_SHADER_PGM_RSRC2_GS__SCRATCH_EN_MASK 0x00000001L +#define SPI_SHADER_PGM_RSRC2_GS__USER_SGPR_MASK 0x0000003EL +#define SPI_SHADER_PGM_RSRC2_GS__TRAP_PRESENT_MASK 0x00000040L +#define SPI_SHADER_PGM_RSRC2_GS__EXCP_EN_MASK 0x0000FF80L +#define SPI_SHADER_PGM_RSRC2_GS__ES_VGPR_COMP_CNT_MASK 0x00030000L +#define SPI_SHADER_PGM_RSRC2_GS__OC_LDS_EN_MASK 0x00040000L +#define SPI_SHADER_PGM_RSRC2_GS__LDS_SIZE_MASK 0x07F80000L +#define SPI_SHADER_PGM_RSRC2_GS__SKIP_USGPR0_MASK 0x08000000L +#define SPI_SHADER_PGM_RSRC2_GS__USER_SGPR_MSB_MASK 0x10000000L +//SPI_SHADER_USER_DATA_ES_0 +#define SPI_SHADER_USER_DATA_ES_0__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_0__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_1 +#define SPI_SHADER_USER_DATA_ES_1__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_1__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_2 +#define SPI_SHADER_USER_DATA_ES_2__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_2__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_3 +#define SPI_SHADER_USER_DATA_ES_3__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_3__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_4 +#define SPI_SHADER_USER_DATA_ES_4__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_4__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_5 +#define SPI_SHADER_USER_DATA_ES_5__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_5__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_6 +#define SPI_SHADER_USER_DATA_ES_6__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_6__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_7 +#define SPI_SHADER_USER_DATA_ES_7__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_7__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_8 +#define SPI_SHADER_USER_DATA_ES_8__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_8__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_9 +#define SPI_SHADER_USER_DATA_ES_9__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_9__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_10 +#define SPI_SHADER_USER_DATA_ES_10__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_10__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_11 +#define SPI_SHADER_USER_DATA_ES_11__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_11__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_12 +#define SPI_SHADER_USER_DATA_ES_12__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_12__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_13 +#define SPI_SHADER_USER_DATA_ES_13__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_13__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_14 +#define SPI_SHADER_USER_DATA_ES_14__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_14__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_15 +#define SPI_SHADER_USER_DATA_ES_15__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_15__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_16 +#define SPI_SHADER_USER_DATA_ES_16__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_16__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_17 +#define SPI_SHADER_USER_DATA_ES_17__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_17__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_18 +#define SPI_SHADER_USER_DATA_ES_18__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_18__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_19 +#define SPI_SHADER_USER_DATA_ES_19__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_19__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_20 +#define SPI_SHADER_USER_DATA_ES_20__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_20__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_21 +#define SPI_SHADER_USER_DATA_ES_21__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_21__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_22 +#define SPI_SHADER_USER_DATA_ES_22__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_22__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_23 +#define SPI_SHADER_USER_DATA_ES_23__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_23__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_24 +#define SPI_SHADER_USER_DATA_ES_24__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_24__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_25 +#define SPI_SHADER_USER_DATA_ES_25__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_25__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_26 +#define SPI_SHADER_USER_DATA_ES_26__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_26__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_27 +#define SPI_SHADER_USER_DATA_ES_27__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_27__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_28 +#define SPI_SHADER_USER_DATA_ES_28__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_28__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_29 +#define SPI_SHADER_USER_DATA_ES_29__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_29__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_30 +#define SPI_SHADER_USER_DATA_ES_30__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_30__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ES_31 +#define SPI_SHADER_USER_DATA_ES_31__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ES_31__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_RSRC4_HS +#define SPI_SHADER_PGM_RSRC4_HS__GROUP_FIFO_DEPTH__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC4_HS__GROUP_FIFO_DEPTH_MASK 0x0000007FL +//SPI_SHADER_USER_DATA_ADDR_LO_HS +#define SPI_SHADER_USER_DATA_ADDR_LO_HS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ADDR_LO_HS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_ADDR_HI_HS +#define SPI_SHADER_USER_DATA_ADDR_HI_HS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_ADDR_HI_HS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_LO_LS +#define SPI_SHADER_PGM_LO_LS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_LO_LS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_HI_LS +#define SPI_SHADER_PGM_HI_LS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_HI_LS__MEM_BASE_MASK 0xFFL +//SPI_SHADER_PGM_RSRC3_HS +#define SPI_SHADER_PGM_RSRC3_HS__WAVE_LIMIT__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC3_HS__LOCK_LOW_THRESHOLD__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC3_HS__SIMD_DISABLE__SHIFT 0xa +#define SPI_SHADER_PGM_RSRC3_HS__CU_EN__SHIFT 0x10 +#define SPI_SHADER_PGM_RSRC3_HS__WAVE_LIMIT_MASK 0x0000003FL +#define SPI_SHADER_PGM_RSRC3_HS__LOCK_LOW_THRESHOLD_MASK 0x000003C0L +#define SPI_SHADER_PGM_RSRC3_HS__SIMD_DISABLE_MASK 0x00003C00L +#define SPI_SHADER_PGM_RSRC3_HS__CU_EN_MASK 0xFFFF0000L +//SPI_SHADER_PGM_LO_HS +#define SPI_SHADER_PGM_LO_HS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_LO_HS__MEM_BASE_MASK 0xFFFFFFFFL +//SPI_SHADER_PGM_HI_HS +#define SPI_SHADER_PGM_HI_HS__MEM_BASE__SHIFT 0x0 +#define SPI_SHADER_PGM_HI_HS__MEM_BASE_MASK 0xFFL +//SPI_SHADER_PGM_RSRC1_HS +#define SPI_SHADER_PGM_RSRC1_HS__VGPRS__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC1_HS__SGPRS__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC1_HS__PRIORITY__SHIFT 0xa +#define SPI_SHADER_PGM_RSRC1_HS__FLOAT_MODE__SHIFT 0xc +#define SPI_SHADER_PGM_RSRC1_HS__PRIV__SHIFT 0x14 +#define SPI_SHADER_PGM_RSRC1_HS__DX10_CLAMP__SHIFT 0x15 +#define SPI_SHADER_PGM_RSRC1_HS__IEEE_MODE__SHIFT 0x17 +#define SPI_SHADER_PGM_RSRC1_HS__LS_VGPR_COMP_CNT__SHIFT 0x1c +#define SPI_SHADER_PGM_RSRC1_HS__FP16_OVFL__SHIFT 0x1e +#define SPI_SHADER_PGM_RSRC1_HS__VGPRS_MASK 0x0000003FL +#define SPI_SHADER_PGM_RSRC1_HS__SGPRS_MASK 0x000003C0L +#define SPI_SHADER_PGM_RSRC1_HS__PRIORITY_MASK 0x00000C00L +#define SPI_SHADER_PGM_RSRC1_HS__FLOAT_MODE_MASK 0x000FF000L +#define SPI_SHADER_PGM_RSRC1_HS__PRIV_MASK 0x00100000L +#define SPI_SHADER_PGM_RSRC1_HS__DX10_CLAMP_MASK 0x00200000L +#define SPI_SHADER_PGM_RSRC1_HS__IEEE_MODE_MASK 0x00800000L +#define SPI_SHADER_PGM_RSRC1_HS__LS_VGPR_COMP_CNT_MASK 0x30000000L +#define SPI_SHADER_PGM_RSRC1_HS__FP16_OVFL_MASK 0x40000000L +//SPI_SHADER_PGM_RSRC2_HS +#define SPI_SHADER_PGM_RSRC2_HS__SCRATCH_EN__SHIFT 0x0 +#define SPI_SHADER_PGM_RSRC2_HS__USER_SGPR__SHIFT 0x1 +#define SPI_SHADER_PGM_RSRC2_HS__TRAP_PRESENT__SHIFT 0x6 +#define SPI_SHADER_PGM_RSRC2_HS__EXCP_EN__SHIFT 0x7 +#define SPI_SHADER_PGM_RSRC2_HS__LDS_SIZE__SHIFT 0x10 +#define SPI_SHADER_PGM_RSRC2_HS__SKIP_USGPR0__SHIFT 0x1b +#define SPI_SHADER_PGM_RSRC2_HS__USER_SGPR_MSB__SHIFT 0x1c +#define SPI_SHADER_PGM_RSRC2_HS__SCRATCH_EN_MASK 0x00000001L +#define SPI_SHADER_PGM_RSRC2_HS__USER_SGPR_MASK 0x0000003EL +#define SPI_SHADER_PGM_RSRC2_HS__TRAP_PRESENT_MASK 0x00000040L +#define SPI_SHADER_PGM_RSRC2_HS__EXCP_EN_MASK 0x0000FF80L +#define SPI_SHADER_PGM_RSRC2_HS__LDS_SIZE_MASK 0x01FF0000L +#define SPI_SHADER_PGM_RSRC2_HS__SKIP_USGPR0_MASK 0x08000000L +#define SPI_SHADER_PGM_RSRC2_HS__USER_SGPR_MSB_MASK 0x10000000L +//SPI_SHADER_USER_DATA_LS_0 +#define SPI_SHADER_USER_DATA_LS_0__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_0__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_1 +#define SPI_SHADER_USER_DATA_LS_1__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_1__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_2 +#define SPI_SHADER_USER_DATA_LS_2__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_2__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_3 +#define SPI_SHADER_USER_DATA_LS_3__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_3__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_4 +#define SPI_SHADER_USER_DATA_LS_4__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_4__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_5 +#define SPI_SHADER_USER_DATA_LS_5__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_5__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_6 +#define SPI_SHADER_USER_DATA_LS_6__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_6__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_7 +#define SPI_SHADER_USER_DATA_LS_7__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_7__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_8 +#define SPI_SHADER_USER_DATA_LS_8__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_8__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_9 +#define SPI_SHADER_USER_DATA_LS_9__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_9__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_10 +#define SPI_SHADER_USER_DATA_LS_10__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_10__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_11 +#define SPI_SHADER_USER_DATA_LS_11__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_11__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_12 +#define SPI_SHADER_USER_DATA_LS_12__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_12__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_13 +#define SPI_SHADER_USER_DATA_LS_13__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_13__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_14 +#define SPI_SHADER_USER_DATA_LS_14__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_14__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_15 +#define SPI_SHADER_USER_DATA_LS_15__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_15__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_16 +#define SPI_SHADER_USER_DATA_LS_16__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_16__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_17 +#define SPI_SHADER_USER_DATA_LS_17__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_17__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_18 +#define SPI_SHADER_USER_DATA_LS_18__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_18__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_19 +#define SPI_SHADER_USER_DATA_LS_19__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_19__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_20 +#define SPI_SHADER_USER_DATA_LS_20__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_20__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_21 +#define SPI_SHADER_USER_DATA_LS_21__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_21__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_22 +#define SPI_SHADER_USER_DATA_LS_22__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_22__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_23 +#define SPI_SHADER_USER_DATA_LS_23__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_23__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_24 +#define SPI_SHADER_USER_DATA_LS_24__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_24__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_25 +#define SPI_SHADER_USER_DATA_LS_25__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_25__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_26 +#define SPI_SHADER_USER_DATA_LS_26__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_26__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_27 +#define SPI_SHADER_USER_DATA_LS_27__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_27__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_28 +#define SPI_SHADER_USER_DATA_LS_28__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_28__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_29 +#define SPI_SHADER_USER_DATA_LS_29__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_29__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_30 +#define SPI_SHADER_USER_DATA_LS_30__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_30__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_LS_31 +#define SPI_SHADER_USER_DATA_LS_31__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_LS_31__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_0 +#define SPI_SHADER_USER_DATA_COMMON_0__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_0__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_1 +#define SPI_SHADER_USER_DATA_COMMON_1__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_1__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_2 +#define SPI_SHADER_USER_DATA_COMMON_2__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_2__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_3 +#define SPI_SHADER_USER_DATA_COMMON_3__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_3__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_4 +#define SPI_SHADER_USER_DATA_COMMON_4__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_4__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_5 +#define SPI_SHADER_USER_DATA_COMMON_5__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_5__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_6 +#define SPI_SHADER_USER_DATA_COMMON_6__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_6__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_7 +#define SPI_SHADER_USER_DATA_COMMON_7__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_7__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_8 +#define SPI_SHADER_USER_DATA_COMMON_8__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_8__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_9 +#define SPI_SHADER_USER_DATA_COMMON_9__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_9__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_10 +#define SPI_SHADER_USER_DATA_COMMON_10__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_10__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_11 +#define SPI_SHADER_USER_DATA_COMMON_11__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_11__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_12 +#define SPI_SHADER_USER_DATA_COMMON_12__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_12__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_13 +#define SPI_SHADER_USER_DATA_COMMON_13__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_13__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_14 +#define SPI_SHADER_USER_DATA_COMMON_14__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_14__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_15 +#define SPI_SHADER_USER_DATA_COMMON_15__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_15__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_16 +#define SPI_SHADER_USER_DATA_COMMON_16__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_16__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_17 +#define SPI_SHADER_USER_DATA_COMMON_17__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_17__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_18 +#define SPI_SHADER_USER_DATA_COMMON_18__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_18__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_19 +#define SPI_SHADER_USER_DATA_COMMON_19__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_19__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_20 +#define SPI_SHADER_USER_DATA_COMMON_20__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_20__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_21 +#define SPI_SHADER_USER_DATA_COMMON_21__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_21__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_22 +#define SPI_SHADER_USER_DATA_COMMON_22__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_22__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_23 +#define SPI_SHADER_USER_DATA_COMMON_23__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_23__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_24 +#define SPI_SHADER_USER_DATA_COMMON_24__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_24__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_25 +#define SPI_SHADER_USER_DATA_COMMON_25__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_25__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_26 +#define SPI_SHADER_USER_DATA_COMMON_26__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_26__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_27 +#define SPI_SHADER_USER_DATA_COMMON_27__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_27__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_28 +#define SPI_SHADER_USER_DATA_COMMON_28__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_28__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_29 +#define SPI_SHADER_USER_DATA_COMMON_29__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_29__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_30 +#define SPI_SHADER_USER_DATA_COMMON_30__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_30__DATA_MASK 0xFFFFFFFFL +//SPI_SHADER_USER_DATA_COMMON_31 +#define SPI_SHADER_USER_DATA_COMMON_31__DATA__SHIFT 0x0 +#define SPI_SHADER_USER_DATA_COMMON_31__DATA_MASK 0xFFFFFFFFL +//COMPUTE_DISPATCH_INITIATOR +#define COMPUTE_DISPATCH_INITIATOR__COMPUTE_SHADER_EN__SHIFT 0x0 +#define COMPUTE_DISPATCH_INITIATOR__PARTIAL_TG_EN__SHIFT 0x1 +#define COMPUTE_DISPATCH_INITIATOR__FORCE_START_AT_000__SHIFT 0x2 +#define COMPUTE_DISPATCH_INITIATOR__ORDERED_APPEND_ENBL__SHIFT 0x3 +#define COMPUTE_DISPATCH_INITIATOR__ORDERED_APPEND_MODE__SHIFT 0x4 +#define COMPUTE_DISPATCH_INITIATOR__USE_THREAD_DIMENSIONS__SHIFT 0x5 +#define COMPUTE_DISPATCH_INITIATOR__ORDER_MODE__SHIFT 0x6 +#define COMPUTE_DISPATCH_INITIATOR__SCALAR_L1_INV_VOL__SHIFT 0xa +#define COMPUTE_DISPATCH_INITIATOR__VECTOR_L1_INV_VOL__SHIFT 0xb +#define COMPUTE_DISPATCH_INITIATOR__RESERVED__SHIFT 0xc +#define COMPUTE_DISPATCH_INITIATOR__RESTORE__SHIFT 0xe +#define COMPUTE_DISPATCH_INITIATOR__COMPUTE_SHADER_EN_MASK 0x00000001L +#define COMPUTE_DISPATCH_INITIATOR__PARTIAL_TG_EN_MASK 0x00000002L +#define COMPUTE_DISPATCH_INITIATOR__FORCE_START_AT_000_MASK 0x00000004L +#define COMPUTE_DISPATCH_INITIATOR__ORDERED_APPEND_ENBL_MASK 0x00000008L +#define COMPUTE_DISPATCH_INITIATOR__ORDERED_APPEND_MODE_MASK 0x00000010L +#define COMPUTE_DISPATCH_INITIATOR__USE_THREAD_DIMENSIONS_MASK 0x00000020L +#define COMPUTE_DISPATCH_INITIATOR__ORDER_MODE_MASK 0x00000040L +#define COMPUTE_DISPATCH_INITIATOR__SCALAR_L1_INV_VOL_MASK 0x00000400L +#define COMPUTE_DISPATCH_INITIATOR__VECTOR_L1_INV_VOL_MASK 0x00000800L +#define COMPUTE_DISPATCH_INITIATOR__RESERVED_MASK 0x00001000L +#define COMPUTE_DISPATCH_INITIATOR__RESTORE_MASK 0x00004000L +//COMPUTE_DIM_X +#define COMPUTE_DIM_X__SIZE__SHIFT 0x0 +#define COMPUTE_DIM_X__SIZE_MASK 0xFFFFFFFFL +//COMPUTE_DIM_Y +#define COMPUTE_DIM_Y__SIZE__SHIFT 0x0 +#define COMPUTE_DIM_Y__SIZE_MASK 0xFFFFFFFFL +//COMPUTE_DIM_Z +#define COMPUTE_DIM_Z__SIZE__SHIFT 0x0 +#define COMPUTE_DIM_Z__SIZE_MASK 0xFFFFFFFFL +//COMPUTE_START_X +#define COMPUTE_START_X__START__SHIFT 0x0 +#define COMPUTE_START_X__START_MASK 0xFFFFFFFFL +//COMPUTE_START_Y +#define COMPUTE_START_Y__START__SHIFT 0x0 +#define COMPUTE_START_Y__START_MASK 0xFFFFFFFFL +//COMPUTE_START_Z +#define COMPUTE_START_Z__START__SHIFT 0x0 +#define COMPUTE_START_Z__START_MASK 0xFFFFFFFFL +//COMPUTE_NUM_THREAD_X +#define COMPUTE_NUM_THREAD_X__NUM_THREAD_FULL__SHIFT 0x0 +#define COMPUTE_NUM_THREAD_X__NUM_THREAD_PARTIAL__SHIFT 0x10 +#define COMPUTE_NUM_THREAD_X__NUM_THREAD_FULL_MASK 0x0000FFFFL +#define COMPUTE_NUM_THREAD_X__NUM_THREAD_PARTIAL_MASK 0xFFFF0000L +//COMPUTE_NUM_THREAD_Y +#define COMPUTE_NUM_THREAD_Y__NUM_THREAD_FULL__SHIFT 0x0 +#define COMPUTE_NUM_THREAD_Y__NUM_THREAD_PARTIAL__SHIFT 0x10 +#define COMPUTE_NUM_THREAD_Y__NUM_THREAD_FULL_MASK 0x0000FFFFL +#define COMPUTE_NUM_THREAD_Y__NUM_THREAD_PARTIAL_MASK 0xFFFF0000L +//COMPUTE_NUM_THREAD_Z +#define COMPUTE_NUM_THREAD_Z__NUM_THREAD_FULL__SHIFT 0x0 +#define COMPUTE_NUM_THREAD_Z__NUM_THREAD_PARTIAL__SHIFT 0x10 +#define COMPUTE_NUM_THREAD_Z__NUM_THREAD_FULL_MASK 0x0000FFFFL +#define COMPUTE_NUM_THREAD_Z__NUM_THREAD_PARTIAL_MASK 0xFFFF0000L +//COMPUTE_PIPELINESTAT_ENABLE +#define COMPUTE_PIPELINESTAT_ENABLE__PIPELINESTAT_ENABLE__SHIFT 0x0 +#define COMPUTE_PIPELINESTAT_ENABLE__PIPELINESTAT_ENABLE_MASK 0x00000001L +//COMPUTE_PERFCOUNT_ENABLE +#define COMPUTE_PERFCOUNT_ENABLE__PERFCOUNT_ENABLE__SHIFT 0x0 +#define COMPUTE_PERFCOUNT_ENABLE__PERFCOUNT_ENABLE_MASK 0x00000001L +//COMPUTE_PGM_LO +#define COMPUTE_PGM_LO__DATA__SHIFT 0x0 +#define COMPUTE_PGM_LO__DATA_MASK 0xFFFFFFFFL +//COMPUTE_PGM_HI +#define COMPUTE_PGM_HI__DATA__SHIFT 0x0 +#define COMPUTE_PGM_HI__DATA_MASK 0x000000FFL +//COMPUTE_DISPATCH_PKT_ADDR_LO +#define COMPUTE_DISPATCH_PKT_ADDR_LO__DATA__SHIFT 0x0 +#define COMPUTE_DISPATCH_PKT_ADDR_LO__DATA_MASK 0xFFFFFFFFL +//COMPUTE_DISPATCH_PKT_ADDR_HI +#define COMPUTE_DISPATCH_PKT_ADDR_HI__DATA__SHIFT 0x0 +#define COMPUTE_DISPATCH_PKT_ADDR_HI__DATA_MASK 0x000000FFL +//COMPUTE_DISPATCH_SCRATCH_BASE_LO +#define COMPUTE_DISPATCH_SCRATCH_BASE_LO__DATA__SHIFT 0x0 +#define COMPUTE_DISPATCH_SCRATCH_BASE_LO__DATA_MASK 0xFFFFFFFFL +//COMPUTE_DISPATCH_SCRATCH_BASE_HI +#define COMPUTE_DISPATCH_SCRATCH_BASE_HI__DATA__SHIFT 0x0 +#define COMPUTE_DISPATCH_SCRATCH_BASE_HI__DATA_MASK 0x000000FFL +//COMPUTE_PGM_RSRC1 +#define COMPUTE_PGM_RSRC1__VGPRS__SHIFT 0x0 +#define COMPUTE_PGM_RSRC1__SGPRS__SHIFT 0x6 +#define COMPUTE_PGM_RSRC1__PRIORITY__SHIFT 0xa +#define COMPUTE_PGM_RSRC1__FLOAT_MODE__SHIFT 0xc +#define COMPUTE_PGM_RSRC1__PRIV__SHIFT 0x14 +#define COMPUTE_PGM_RSRC1__DX10_CLAMP__SHIFT 0x15 +#define COMPUTE_PGM_RSRC1__IEEE_MODE__SHIFT 0x17 +#define COMPUTE_PGM_RSRC1__BULKY__SHIFT 0x18 +#define COMPUTE_PGM_RSRC1__FP16_OVFL__SHIFT 0x1a +#define COMPUTE_PGM_RSRC1__VGPRS_MASK 0x0000003FL +#define COMPUTE_PGM_RSRC1__SGPRS_MASK 0x000003C0L +#define COMPUTE_PGM_RSRC1__PRIORITY_MASK 0x00000C00L +#define COMPUTE_PGM_RSRC1__FLOAT_MODE_MASK 0x000FF000L +#define COMPUTE_PGM_RSRC1__PRIV_MASK 0x00100000L +#define COMPUTE_PGM_RSRC1__DX10_CLAMP_MASK 0x00200000L +#define COMPUTE_PGM_RSRC1__IEEE_MODE_MASK 0x00800000L +#define COMPUTE_PGM_RSRC1__BULKY_MASK 0x01000000L +#define COMPUTE_PGM_RSRC1__FP16_OVFL_MASK 0x04000000L +//COMPUTE_PGM_RSRC2 +#define COMPUTE_PGM_RSRC2__SCRATCH_EN__SHIFT 0x0 +#define COMPUTE_PGM_RSRC2__USER_SGPR__SHIFT 0x1 +#define COMPUTE_PGM_RSRC2__TRAP_PRESENT__SHIFT 0x6 +#define COMPUTE_PGM_RSRC2__TGID_X_EN__SHIFT 0x7 +#define COMPUTE_PGM_RSRC2__TGID_Y_EN__SHIFT 0x8 +#define COMPUTE_PGM_RSRC2__TGID_Z_EN__SHIFT 0x9 +#define COMPUTE_PGM_RSRC2__TG_SIZE_EN__SHIFT 0xa +#define COMPUTE_PGM_RSRC2__TIDIG_COMP_CNT__SHIFT 0xb +#define COMPUTE_PGM_RSRC2__EXCP_EN_MSB__SHIFT 0xd +#define COMPUTE_PGM_RSRC2__LDS_SIZE__SHIFT 0xf +#define COMPUTE_PGM_RSRC2__EXCP_EN__SHIFT 0x18 +#define COMPUTE_PGM_RSRC2__SKIP_USGPR0__SHIFT 0x1f +#define COMPUTE_PGM_RSRC2__SCRATCH_EN_MASK 0x00000001L +#define COMPUTE_PGM_RSRC2__USER_SGPR_MASK 0x0000003EL +#define COMPUTE_PGM_RSRC2__TRAP_PRESENT_MASK 0x00000040L +#define COMPUTE_PGM_RSRC2__TGID_X_EN_MASK 0x00000080L +#define COMPUTE_PGM_RSRC2__TGID_Y_EN_MASK 0x00000100L +#define COMPUTE_PGM_RSRC2__TGID_Z_EN_MASK 0x00000200L +#define COMPUTE_PGM_RSRC2__TG_SIZE_EN_MASK 0x00000400L +#define COMPUTE_PGM_RSRC2__TIDIG_COMP_CNT_MASK 0x00001800L +#define COMPUTE_PGM_RSRC2__EXCP_EN_MSB_MASK 0x00006000L +#define COMPUTE_PGM_RSRC2__LDS_SIZE_MASK 0x00FF8000L +#define COMPUTE_PGM_RSRC2__EXCP_EN_MASK 0x7F000000L +#define COMPUTE_PGM_RSRC2__SKIP_USGPR0_MASK 0x80000000L +//COMPUTE_VMID +#define COMPUTE_VMID__DATA__SHIFT 0x0 +#define COMPUTE_VMID__DATA_MASK 0x0000000FL +//COMPUTE_RESOURCE_LIMITS +#define COMPUTE_RESOURCE_LIMITS__WAVES_PER_SH__SHIFT 0x0 +#define COMPUTE_RESOURCE_LIMITS__TG_PER_CU__SHIFT 0xc +#define COMPUTE_RESOURCE_LIMITS__LOCK_THRESHOLD__SHIFT 0x10 +#define COMPUTE_RESOURCE_LIMITS__SIMD_DEST_CNTL__SHIFT 0x16 +#define COMPUTE_RESOURCE_LIMITS__FORCE_SIMD_DIST__SHIFT 0x17 +#define COMPUTE_RESOURCE_LIMITS__CU_GROUP_COUNT__SHIFT 0x18 +#define COMPUTE_RESOURCE_LIMITS__SIMD_DISABLE__SHIFT 0x1b +#define COMPUTE_RESOURCE_LIMITS__WAVES_PER_SH_MASK 0x000003FFL +#define COMPUTE_RESOURCE_LIMITS__TG_PER_CU_MASK 0x0000F000L +#define COMPUTE_RESOURCE_LIMITS__LOCK_THRESHOLD_MASK 0x003F0000L +#define COMPUTE_RESOURCE_LIMITS__SIMD_DEST_CNTL_MASK 0x00400000L +#define COMPUTE_RESOURCE_LIMITS__FORCE_SIMD_DIST_MASK 0x00800000L +#define COMPUTE_RESOURCE_LIMITS__CU_GROUP_COUNT_MASK 0x07000000L +#define COMPUTE_RESOURCE_LIMITS__SIMD_DISABLE_MASK 0x78000000L +//COMPUTE_STATIC_THREAD_MGMT_SE0 +#define COMPUTE_STATIC_THREAD_MGMT_SE0__SH0_CU_EN__SHIFT 0x0 +#define COMPUTE_STATIC_THREAD_MGMT_SE0__SH1_CU_EN__SHIFT 0x10 +#define COMPUTE_STATIC_THREAD_MGMT_SE0__SH0_CU_EN_MASK 0x0000FFFFL +#define COMPUTE_STATIC_THREAD_MGMT_SE0__SH1_CU_EN_MASK 0xFFFF0000L +//COMPUTE_STATIC_THREAD_MGMT_SE1 +#define COMPUTE_STATIC_THREAD_MGMT_SE1__SH0_CU_EN__SHIFT 0x0 +#define COMPUTE_STATIC_THREAD_MGMT_SE1__SH1_CU_EN__SHIFT 0x10 +#define COMPUTE_STATIC_THREAD_MGMT_SE1__SH0_CU_EN_MASK 0x0000FFFFL +#define COMPUTE_STATIC_THREAD_MGMT_SE1__SH1_CU_EN_MASK 0xFFFF0000L +//COMPUTE_TMPRING_SIZE +#define COMPUTE_TMPRING_SIZE__WAVES__SHIFT 0x0 +#define COMPUTE_TMPRING_SIZE__WAVESIZE__SHIFT 0xc +#define COMPUTE_TMPRING_SIZE__WAVES_MASK 0x00000FFFL +#define COMPUTE_TMPRING_SIZE__WAVESIZE_MASK 0x01FFF000L +//COMPUTE_STATIC_THREAD_MGMT_SE2 +#define COMPUTE_STATIC_THREAD_MGMT_SE2__SH0_CU_EN__SHIFT 0x0 +#define COMPUTE_STATIC_THREAD_MGMT_SE2__SH1_CU_EN__SHIFT 0x10 +#define COMPUTE_STATIC_THREAD_MGMT_SE2__SH0_CU_EN_MASK 0x0000FFFFL +#define COMPUTE_STATIC_THREAD_MGMT_SE2__SH1_CU_EN_MASK 0xFFFF0000L +//COMPUTE_STATIC_THREAD_MGMT_SE3 +#define COMPUTE_STATIC_THREAD_MGMT_SE3__SH0_CU_EN__SHIFT 0x0 +#define COMPUTE_STATIC_THREAD_MGMT_SE3__SH1_CU_EN__SHIFT 0x10 +#define COMPUTE_STATIC_THREAD_MGMT_SE3__SH0_CU_EN_MASK 0x0000FFFFL +#define COMPUTE_STATIC_THREAD_MGMT_SE3__SH1_CU_EN_MASK 0xFFFF0000L +//COMPUTE_RESTART_X +#define COMPUTE_RESTART_X__RESTART__SHIFT 0x0 +#define COMPUTE_RESTART_X__RESTART_MASK 0xFFFFFFFFL +//COMPUTE_RESTART_Y +#define COMPUTE_RESTART_Y__RESTART__SHIFT 0x0 +#define COMPUTE_RESTART_Y__RESTART_MASK 0xFFFFFFFFL +//COMPUTE_RESTART_Z +#define COMPUTE_RESTART_Z__RESTART__SHIFT 0x0 +#define COMPUTE_RESTART_Z__RESTART_MASK 0xFFFFFFFFL +//COMPUTE_THREAD_TRACE_ENABLE +#define COMPUTE_THREAD_TRACE_ENABLE__THREAD_TRACE_ENABLE__SHIFT 0x0 +#define COMPUTE_THREAD_TRACE_ENABLE__THREAD_TRACE_ENABLE_MASK 0x00000001L +//COMPUTE_MISC_RESERVED +#define COMPUTE_MISC_RESERVED__SEND_SEID__SHIFT 0x0 +#define COMPUTE_MISC_RESERVED__RESERVED2__SHIFT 0x2 +#define COMPUTE_MISC_RESERVED__RESERVED3__SHIFT 0x3 +#define COMPUTE_MISC_RESERVED__RESERVED4__SHIFT 0x4 +#define COMPUTE_MISC_RESERVED__WAVE_ID_BASE__SHIFT 0x5 +#define COMPUTE_MISC_RESERVED__SEND_SEID_MASK 0x00000003L +#define COMPUTE_MISC_RESERVED__RESERVED2_MASK 0x00000004L +#define COMPUTE_MISC_RESERVED__RESERVED3_MASK 0x00000008L +#define COMPUTE_MISC_RESERVED__RESERVED4_MASK 0x00000010L +#define COMPUTE_MISC_RESERVED__WAVE_ID_BASE_MASK 0x0001FFE0L +//COMPUTE_DISPATCH_ID +#define COMPUTE_DISPATCH_ID__DISPATCH_ID__SHIFT 0x0 +#define COMPUTE_DISPATCH_ID__DISPATCH_ID_MASK 0xFFFFFFFFL +//COMPUTE_THREADGROUP_ID +#define COMPUTE_THREADGROUP_ID__THREADGROUP_ID__SHIFT 0x0 +#define COMPUTE_THREADGROUP_ID__THREADGROUP_ID_MASK 0xFFFFFFFFL +//COMPUTE_RELAUNCH +#define COMPUTE_RELAUNCH__PAYLOAD__SHIFT 0x0 +#define COMPUTE_RELAUNCH__IS_EVENT__SHIFT 0x1e +#define COMPUTE_RELAUNCH__IS_STATE__SHIFT 0x1f +#define COMPUTE_RELAUNCH__PAYLOAD_MASK 0x3FFFFFFFL +#define COMPUTE_RELAUNCH__IS_EVENT_MASK 0x40000000L +#define COMPUTE_RELAUNCH__IS_STATE_MASK 0x80000000L +//COMPUTE_WAVE_RESTORE_ADDR_LO +#define COMPUTE_WAVE_RESTORE_ADDR_LO__ADDR__SHIFT 0x0 +#define COMPUTE_WAVE_RESTORE_ADDR_LO__ADDR_MASK 0xFFFFFFFFL +//COMPUTE_WAVE_RESTORE_ADDR_HI +#define COMPUTE_WAVE_RESTORE_ADDR_HI__ADDR__SHIFT 0x0 +#define COMPUTE_WAVE_RESTORE_ADDR_HI__ADDR_MASK 0xFFFFL +//COMPUTE_USER_DATA_0 +#define COMPUTE_USER_DATA_0__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_0__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_1 +#define COMPUTE_USER_DATA_1__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_1__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_2 +#define COMPUTE_USER_DATA_2__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_2__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_3 +#define COMPUTE_USER_DATA_3__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_3__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_4 +#define COMPUTE_USER_DATA_4__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_4__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_5 +#define COMPUTE_USER_DATA_5__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_5__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_6 +#define COMPUTE_USER_DATA_6__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_6__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_7 +#define COMPUTE_USER_DATA_7__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_7__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_8 +#define COMPUTE_USER_DATA_8__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_8__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_9 +#define COMPUTE_USER_DATA_9__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_9__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_10 +#define COMPUTE_USER_DATA_10__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_10__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_11 +#define COMPUTE_USER_DATA_11__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_11__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_12 +#define COMPUTE_USER_DATA_12__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_12__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_13 +#define COMPUTE_USER_DATA_13__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_13__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_14 +#define COMPUTE_USER_DATA_14__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_14__DATA_MASK 0xFFFFFFFFL +//COMPUTE_USER_DATA_15 +#define COMPUTE_USER_DATA_15__DATA__SHIFT 0x0 +#define COMPUTE_USER_DATA_15__DATA_MASK 0xFFFFFFFFL +//COMPUTE_NOWHERE +#define COMPUTE_NOWHERE__DATA__SHIFT 0x0 +#define COMPUTE_NOWHERE__DATA_MASK 0xFFFFFFFFL + + +// addressBlock: gc_cppdec +//CP_DFY_CNTL +#define CP_DFY_CNTL__POLICY__SHIFT 0x0 +#define CP_DFY_CNTL__MTYPE__SHIFT 0x2 +#define CP_DFY_CNTL__TPI_SDP_SEL__SHIFT 0x1a +#define CP_DFY_CNTL__LFSR_RESET__SHIFT 0x1c +#define CP_DFY_CNTL__MODE__SHIFT 0x1d +#define CP_DFY_CNTL__ENABLE__SHIFT 0x1f +#define CP_DFY_CNTL__POLICY_MASK 0x00000001L +#define CP_DFY_CNTL__MTYPE_MASK 0x0000000CL +#define CP_DFY_CNTL__TPI_SDP_SEL_MASK 0x04000000L +#define CP_DFY_CNTL__LFSR_RESET_MASK 0x10000000L +#define CP_DFY_CNTL__MODE_MASK 0x60000000L +#define CP_DFY_CNTL__ENABLE_MASK 0x80000000L +//CP_DFY_STAT +#define CP_DFY_STAT__BURST_COUNT__SHIFT 0x0 +#define CP_DFY_STAT__TAGS_PENDING__SHIFT 0x10 +#define CP_DFY_STAT__BUSY__SHIFT 0x1f +#define CP_DFY_STAT__BURST_COUNT_MASK 0x0000FFFFL +#define CP_DFY_STAT__TAGS_PENDING_MASK 0x07FF0000L +#define CP_DFY_STAT__BUSY_MASK 0x80000000L +//CP_DFY_ADDR_HI +#define CP_DFY_ADDR_HI__ADDR_HI__SHIFT 0x0 +#define CP_DFY_ADDR_HI__ADDR_HI_MASK 0xFFFFFFFFL +//CP_DFY_ADDR_LO +#define CP_DFY_ADDR_LO__ADDR_LO__SHIFT 0x5 +#define CP_DFY_ADDR_LO__ADDR_LO_MASK 0xFFFFFFE0L +//CP_DFY_DATA_0 +#define CP_DFY_DATA_0__DATA__SHIFT 0x0 +#define CP_DFY_DATA_0__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_1 +#define CP_DFY_DATA_1__DATA__SHIFT 0x0 +#define CP_DFY_DATA_1__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_2 +#define CP_DFY_DATA_2__DATA__SHIFT 0x0 +#define CP_DFY_DATA_2__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_3 +#define CP_DFY_DATA_3__DATA__SHIFT 0x0 +#define CP_DFY_DATA_3__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_4 +#define CP_DFY_DATA_4__DATA__SHIFT 0x0 +#define CP_DFY_DATA_4__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_5 +#define CP_DFY_DATA_5__DATA__SHIFT 0x0 +#define CP_DFY_DATA_5__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_6 +#define CP_DFY_DATA_6__DATA__SHIFT 0x0 +#define CP_DFY_DATA_6__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_7 +#define CP_DFY_DATA_7__DATA__SHIFT 0x0 +#define CP_DFY_DATA_7__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_8 +#define CP_DFY_DATA_8__DATA__SHIFT 0x0 +#define CP_DFY_DATA_8__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_9 +#define CP_DFY_DATA_9__DATA__SHIFT 0x0 +#define CP_DFY_DATA_9__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_10 +#define CP_DFY_DATA_10__DATA__SHIFT 0x0 +#define CP_DFY_DATA_10__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_11 +#define CP_DFY_DATA_11__DATA__SHIFT 0x0 +#define CP_DFY_DATA_11__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_12 +#define CP_DFY_DATA_12__DATA__SHIFT 0x0 +#define CP_DFY_DATA_12__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_13 +#define CP_DFY_DATA_13__DATA__SHIFT 0x0 +#define CP_DFY_DATA_13__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_14 +#define CP_DFY_DATA_14__DATA__SHIFT 0x0 +#define CP_DFY_DATA_14__DATA_MASK 0xFFFFFFFFL +//CP_DFY_DATA_15 +#define CP_DFY_DATA_15__DATA__SHIFT 0x0 +#define CP_DFY_DATA_15__DATA_MASK 0xFFFFFFFFL +//CP_DFY_CMD +#define CP_DFY_CMD__OFFSET__SHIFT 0x0 +#define CP_DFY_CMD__SIZE__SHIFT 0x10 +#define CP_DFY_CMD__OFFSET_MASK 0x000001FFL +#define CP_DFY_CMD__SIZE_MASK 0xFFFF0000L +//CP_EOPQ_WAIT_TIME +#define CP_EOPQ_WAIT_TIME__WAIT_TIME__SHIFT 0x0 +#define CP_EOPQ_WAIT_TIME__SCALE_COUNT__SHIFT 0xa +#define CP_EOPQ_WAIT_TIME__WAIT_TIME_MASK 0x000003FFL +#define CP_EOPQ_WAIT_TIME__SCALE_COUNT_MASK 0x0003FC00L +//CP_CPC_MGCG_SYNC_CNTL +#define CP_CPC_MGCG_SYNC_CNTL__COOLDOWN_PERIOD__SHIFT 0x0 +#define CP_CPC_MGCG_SYNC_CNTL__WARMUP_PERIOD__SHIFT 0x8 +#define CP_CPC_MGCG_SYNC_CNTL__COOLDOWN_PERIOD_MASK 0x000000FFL +#define CP_CPC_MGCG_SYNC_CNTL__WARMUP_PERIOD_MASK 0x0000FF00L +//CPC_INT_INFO +#define CPC_INT_INFO__ADDR_HI__SHIFT 0x0 +#define CPC_INT_INFO__TYPE__SHIFT 0x10 +#define CPC_INT_INFO__VMID__SHIFT 0x14 +#define CPC_INT_INFO__QUEUE_ID__SHIFT 0x1c +#define CPC_INT_INFO__ADDR_HI_MASK 0x0000FFFFL +#define CPC_INT_INFO__TYPE_MASK 0x00010000L +#define CPC_INT_INFO__VMID_MASK 0x00F00000L +#define CPC_INT_INFO__QUEUE_ID_MASK 0x70000000L +//CP_VIRT_STATUS +#define CP_VIRT_STATUS__VIRT_STATUS__SHIFT 0x0 +#define CP_VIRT_STATUS__VIRT_STATUS_MASK 0xFFFFFFFFL +//CPC_INT_ADDR +#define CPC_INT_ADDR__ADDR__SHIFT 0x0 +#define CPC_INT_ADDR__ADDR_MASK 0xFFFFFFFFL +//CPC_INT_PASID +#define CPC_INT_PASID__PASID__SHIFT 0x0 +#define CPC_INT_PASID__PASID_MASK 0x0000FFFFL +//CP_GFX_ERROR +#define CP_GFX_ERROR__EDC_ERROR_ID__SHIFT 0x0 +#define CP_GFX_ERROR__SUA_ERROR__SHIFT 0x4 +#define CP_GFX_ERROR__RSVD1_ERROR__SHIFT 0x5 +#define CP_GFX_ERROR__RSVD2_ERROR__SHIFT 0x6 +#define CP_GFX_ERROR__SEM_UTCL1_ERROR__SHIFT 0x7 +#define CP_GFX_ERROR__QU_STRM_UTCL1_ERROR__SHIFT 0x8 +#define CP_GFX_ERROR__QU_EOP_UTCL1_ERROR__SHIFT 0x9 +#define CP_GFX_ERROR__QU_PIPE_UTCL1_ERROR__SHIFT 0xa +#define CP_GFX_ERROR__QU_READ_UTCL1_ERROR__SHIFT 0xb +#define CP_GFX_ERROR__SYNC_MEMRD_UTCL1_ERROR__SHIFT 0xc +#define CP_GFX_ERROR__SYNC_MEMWR_UTCL1_ERROR__SHIFT 0xd +#define CP_GFX_ERROR__SHADOW_UTCL1_ERROR__SHIFT 0xe +#define CP_GFX_ERROR__APPEND_UTCL1_ERROR__SHIFT 0xf +#define CP_GFX_ERROR__CE_DMA_UTCL1_ERROR__SHIFT 0x10 +#define CP_GFX_ERROR__PFP_VGTDMA_UTCL1_ERROR__SHIFT 0x11 +#define CP_GFX_ERROR__DMA_SRC_UTCL1_ERROR__SHIFT 0x12 +#define CP_GFX_ERROR__DMA_DST_UTCL1_ERROR__SHIFT 0x13 +#define CP_GFX_ERROR__PFP_TC_UTCL1_ERROR__SHIFT 0x14 +#define CP_GFX_ERROR__ME_TC_UTCL1_ERROR__SHIFT 0x15 +#define CP_GFX_ERROR__CE_TC_UTCL1_ERROR__SHIFT 0x16 +#define CP_GFX_ERROR__PRT_LOD_UTCL1_ERROR__SHIFT 0x17 +#define CP_GFX_ERROR__RDPTR_RPT_UTCL1_ERROR__SHIFT 0x18 +#define CP_GFX_ERROR__RB_FETCHER_UTCL1_ERROR__SHIFT 0x19 +#define CP_GFX_ERROR__I1_FETCHER_UTCL1_ERROR__SHIFT 0x1a +#define CP_GFX_ERROR__I2_FETCHER_UTCL1_ERROR__SHIFT 0x1b +#define CP_GFX_ERROR__C1_FETCHER_UTCL1_ERROR__SHIFT 0x1c +#define CP_GFX_ERROR__C2_FETCHER_UTCL1_ERROR__SHIFT 0x1d +#define CP_GFX_ERROR__ST_FETCHER_UTCL1_ERROR__SHIFT 0x1e +#define CP_GFX_ERROR__CE_INIT_UTCL1_ERROR__SHIFT 0x1f +#define CP_GFX_ERROR__EDC_ERROR_ID_MASK 0x0000000FL +#define CP_GFX_ERROR__SUA_ERROR_MASK 0x00000010L +#define CP_GFX_ERROR__RSVD1_ERROR_MASK 0x00000020L +#define CP_GFX_ERROR__RSVD2_ERROR_MASK 0x00000040L +#define CP_GFX_ERROR__SEM_UTCL1_ERROR_MASK 0x00000080L +#define CP_GFX_ERROR__QU_STRM_UTCL1_ERROR_MASK 0x00000100L +#define CP_GFX_ERROR__QU_EOP_UTCL1_ERROR_MASK 0x00000200L +#define CP_GFX_ERROR__QU_PIPE_UTCL1_ERROR_MASK 0x00000400L +#define CP_GFX_ERROR__QU_READ_UTCL1_ERROR_MASK 0x00000800L +#define CP_GFX_ERROR__SYNC_MEMRD_UTCL1_ERROR_MASK 0x00001000L +#define CP_GFX_ERROR__SYNC_MEMWR_UTCL1_ERROR_MASK 0x00002000L +#define CP_GFX_ERROR__SHADOW_UTCL1_ERROR_MASK 0x00004000L +#define CP_GFX_ERROR__APPEND_UTCL1_ERROR_MASK 0x00008000L +#define CP_GFX_ERROR__CE_DMA_UTCL1_ERROR_MASK 0x00010000L +#define CP_GFX_ERROR__PFP_VGTDMA_UTCL1_ERROR_MASK 0x00020000L +#define CP_GFX_ERROR__DMA_SRC_UTCL1_ERROR_MASK 0x00040000L +#define CP_GFX_ERROR__DMA_DST_UTCL1_ERROR_MASK 0x00080000L +#define CP_GFX_ERROR__PFP_TC_UTCL1_ERROR_MASK 0x00100000L +#define CP_GFX_ERROR__ME_TC_UTCL1_ERROR_MASK 0x00200000L +#define CP_GFX_ERROR__CE_TC_UTCL1_ERROR_MASK 0x00400000L +#define CP_GFX_ERROR__PRT_LOD_UTCL1_ERROR_MASK 0x00800000L +#define CP_GFX_ERROR__RDPTR_RPT_UTCL1_ERROR_MASK 0x01000000L +#define CP_GFX_ERROR__RB_FETCHER_UTCL1_ERROR_MASK 0x02000000L +#define CP_GFX_ERROR__I1_FETCHER_UTCL1_ERROR_MASK 0x04000000L +#define CP_GFX_ERROR__I2_FETCHER_UTCL1_ERROR_MASK 0x08000000L +#define CP_GFX_ERROR__C1_FETCHER_UTCL1_ERROR_MASK 0x10000000L +#define CP_GFX_ERROR__C2_FETCHER_UTCL1_ERROR_MASK 0x20000000L +#define CP_GFX_ERROR__ST_FETCHER_UTCL1_ERROR_MASK 0x40000000L +#define CP_GFX_ERROR__CE_INIT_UTCL1_ERROR_MASK 0x80000000L +//CPG_UTCL1_CNTL +#define CPG_UTCL1_CNTL__XNACK_REDO_TIMER_CNT__SHIFT 0x0 +#define CPG_UTCL1_CNTL__VMID_RESET_MODE__SHIFT 0x17 +#define CPG_UTCL1_CNTL__DROP_MODE__SHIFT 0x18 +#define CPG_UTCL1_CNTL__BYPASS__SHIFT 0x19 +#define CPG_UTCL1_CNTL__INVALIDATE__SHIFT 0x1a +#define CPG_UTCL1_CNTL__FRAG_LIMIT_MODE__SHIFT 0x1b +#define CPG_UTCL1_CNTL__FORCE_SNOOP__SHIFT 0x1c +#define CPG_UTCL1_CNTL__FORCE_SD_VMID_DIRTY__SHIFT 0x1d +#define CPG_UTCL1_CNTL__MTYPE_NO_PTE_MODE__SHIFT 0x1e +#define CPG_UTCL1_CNTL__XNACK_REDO_TIMER_CNT_MASK 0x000FFFFFL +#define CPG_UTCL1_CNTL__VMID_RESET_MODE_MASK 0x00800000L +#define CPG_UTCL1_CNTL__DROP_MODE_MASK 0x01000000L +#define CPG_UTCL1_CNTL__BYPASS_MASK 0x02000000L +#define CPG_UTCL1_CNTL__INVALIDATE_MASK 0x04000000L +#define CPG_UTCL1_CNTL__FRAG_LIMIT_MODE_MASK 0x08000000L +#define CPG_UTCL1_CNTL__FORCE_SNOOP_MASK 0x10000000L +#define CPG_UTCL1_CNTL__FORCE_SD_VMID_DIRTY_MASK 0x20000000L +#define CPG_UTCL1_CNTL__MTYPE_NO_PTE_MODE_MASK 0x40000000L +//CPC_UTCL1_CNTL +#define CPC_UTCL1_CNTL__XNACK_REDO_TIMER_CNT__SHIFT 0x0 +#define CPC_UTCL1_CNTL__DROP_MODE__SHIFT 0x18 +#define CPC_UTCL1_CNTL__BYPASS__SHIFT 0x19 +#define CPC_UTCL1_CNTL__INVALIDATE__SHIFT 0x1a +#define CPC_UTCL1_CNTL__FRAG_LIMIT_MODE__SHIFT 0x1b +#define CPC_UTCL1_CNTL__FORCE_SNOOP__SHIFT 0x1c +#define CPC_UTCL1_CNTL__FORCE_SD_VMID_DIRTY__SHIFT 0x1d +#define CPC_UTCL1_CNTL__MTYPE_NO_PTE_MODE__SHIFT 0x1e +#define CPC_UTCL1_CNTL__XNACK_REDO_TIMER_CNT_MASK 0x000FFFFFL +#define CPC_UTCL1_CNTL__DROP_MODE_MASK 0x01000000L +#define CPC_UTCL1_CNTL__BYPASS_MASK 0x02000000L +#define CPC_UTCL1_CNTL__INVALIDATE_MASK 0x04000000L +#define CPC_UTCL1_CNTL__FRAG_LIMIT_MODE_MASK 0x08000000L +#define CPC_UTCL1_CNTL__FORCE_SNOOP_MASK 0x10000000L +#define CPC_UTCL1_CNTL__FORCE_SD_VMID_DIRTY_MASK 0x20000000L +#define CPC_UTCL1_CNTL__MTYPE_NO_PTE_MODE_MASK 0x40000000L +//CPF_UTCL1_CNTL +#define CPF_UTCL1_CNTL__XNACK_REDO_TIMER_CNT__SHIFT 0x0 +#define CPF_UTCL1_CNTL__VMID_RESET_MODE__SHIFT 0x17 +#define CPF_UTCL1_CNTL__DROP_MODE__SHIFT 0x18 +#define CPF_UTCL1_CNTL__BYPASS__SHIFT 0x19 +#define CPF_UTCL1_CNTL__INVALIDATE__SHIFT 0x1a +#define CPF_UTCL1_CNTL__FRAG_LIMIT_MODE__SHIFT 0x1b +#define CPF_UTCL1_CNTL__FORCE_SNOOP__SHIFT 0x1c +#define CPF_UTCL1_CNTL__FORCE_SD_VMID_DIRTY__SHIFT 0x1d +#define CPF_UTCL1_CNTL__MTYPE_NO_PTE_MODE__SHIFT 0x1e +#define CPF_UTCL1_CNTL__FORCE_NO_EXE__SHIFT 0x1f +#define CPF_UTCL1_CNTL__XNACK_REDO_TIMER_CNT_MASK 0x000FFFFFL +#define CPF_UTCL1_CNTL__VMID_RESET_MODE_MASK 0x00800000L +#define CPF_UTCL1_CNTL__DROP_MODE_MASK 0x01000000L +#define CPF_UTCL1_CNTL__BYPASS_MASK 0x02000000L +#define CPF_UTCL1_CNTL__INVALIDATE_MASK 0x04000000L +#define CPF_UTCL1_CNTL__FRAG_LIMIT_MODE_MASK 0x08000000L +#define CPF_UTCL1_CNTL__FORCE_SNOOP_MASK 0x10000000L +#define CPF_UTCL1_CNTL__FORCE_SD_VMID_DIRTY_MASK 0x20000000L +#define CPF_UTCL1_CNTL__MTYPE_NO_PTE_MODE_MASK 0x40000000L +#define CPF_UTCL1_CNTL__FORCE_NO_EXE_MASK 0x80000000L +//CP_AQL_SMM_STATUS +#define CP_AQL_SMM_STATUS__AQL_QUEUE_SMM__SHIFT 0x0 +#define CP_AQL_SMM_STATUS__AQL_QUEUE_SMM_MASK 0xFFFFFFFFL +//CP_RB0_BASE +#define CP_RB0_BASE__RB_BASE__SHIFT 0x0 +#define CP_RB0_BASE__RB_BASE_MASK 0xFFFFFFFFL +//CP_RB_BASE +#define CP_RB_BASE__RB_BASE__SHIFT 0x0 +#define CP_RB_BASE__RB_BASE_MASK 0xFFFFFFFFL +//CP_RB0_CNTL +#define CP_RB0_CNTL__RB_BUFSZ__SHIFT 0x0 +#define CP_RB0_CNTL__RB_BLKSZ__SHIFT 0x8 +#define CP_RB0_CNTL__BUF_SWAP__SHIFT 0x11 +#define CP_RB0_CNTL__MIN_AVAILSZ__SHIFT 0x14 +#define CP_RB0_CNTL__MIN_IB_AVAILSZ__SHIFT 0x16 +#define CP_RB0_CNTL__CACHE_POLICY__SHIFT 0x18 +#define CP_RB0_CNTL__RB_NO_UPDATE__SHIFT 0x1b +#define CP_RB0_CNTL__RB_RPTR_WR_ENA__SHIFT 0x1f +#define CP_RB0_CNTL__RB_BUFSZ_MASK 0x0000003FL +#define CP_RB0_CNTL__RB_BLKSZ_MASK 0x00003F00L +#define CP_RB0_CNTL__BUF_SWAP_MASK 0x00060000L +#define CP_RB0_CNTL__MIN_AVAILSZ_MASK 0x00300000L +#define CP_RB0_CNTL__MIN_IB_AVAILSZ_MASK 0x00C00000L +#define CP_RB0_CNTL__CACHE_POLICY_MASK 0x01000000L +#define CP_RB0_CNTL__RB_NO_UPDATE_MASK 0x08000000L +#define CP_RB0_CNTL__RB_RPTR_WR_ENA_MASK 0x80000000L +//CP_RB_CNTL +#define CP_RB_CNTL__RB_BUFSZ__SHIFT 0x0 +#define CP_RB_CNTL__RB_BLKSZ__SHIFT 0x8 +#define CP_RB_CNTL__MIN_AVAILSZ__SHIFT 0x14 +#define CP_RB_CNTL__MIN_IB_AVAILSZ__SHIFT 0x16 +#define CP_RB_CNTL__CACHE_POLICY__SHIFT 0x18 +#define CP_RB_CNTL__RB_NO_UPDATE__SHIFT 0x1b +#define CP_RB_CNTL__RB_RPTR_WR_ENA__SHIFT 0x1f +#define CP_RB_CNTL__RB_BUFSZ_MASK 0x0000003FL +#define CP_RB_CNTL__RB_BLKSZ_MASK 0x00003F00L +#define CP_RB_CNTL__MIN_AVAILSZ_MASK 0x00300000L +#define CP_RB_CNTL__MIN_IB_AVAILSZ_MASK 0x00C00000L +#define CP_RB_CNTL__CACHE_POLICY_MASK 0x01000000L +#define CP_RB_CNTL__RB_NO_UPDATE_MASK 0x08000000L +#define CP_RB_CNTL__RB_RPTR_WR_ENA_MASK 0x80000000L +//CP_RB_RPTR_WR +#define CP_RB_RPTR_WR__RB_RPTR_WR__SHIFT 0x0 +#define CP_RB_RPTR_WR__RB_RPTR_WR_MASK 0x000FFFFFL +//CP_RB0_RPTR_ADDR +#define CP_RB0_RPTR_ADDR__RB_RPTR_ADDR__SHIFT 0x2 +#define CP_RB0_RPTR_ADDR__RB_RPTR_ADDR_MASK 0xFFFFFFFCL +//CP_RB_RPTR_ADDR +#define CP_RB_RPTR_ADDR__RB_RPTR_ADDR__SHIFT 0x2 +#define CP_RB_RPTR_ADDR__RB_RPTR_ADDR_MASK 0xFFFFFFFCL +//CP_RB0_RPTR_ADDR_HI +#define CP_RB0_RPTR_ADDR_HI__RB_RPTR_ADDR_HI__SHIFT 0x0 +#define CP_RB0_RPTR_ADDR_HI__RB_RPTR_ADDR_HI_MASK 0x0000FFFFL +//CP_RB_RPTR_ADDR_HI +#define CP_RB_RPTR_ADDR_HI__RB_RPTR_ADDR_HI__SHIFT 0x0 +#define CP_RB_RPTR_ADDR_HI__RB_RPTR_ADDR_HI_MASK 0x0000FFFFL +//CP_RB0_BUFSZ_MASK +#define CP_RB0_BUFSZ_MASK__DATA__SHIFT 0x0 +#define CP_RB0_BUFSZ_MASK__DATA_MASK 0x000FFFFFL +//CP_RB_BUFSZ_MASK +#define CP_RB_BUFSZ_MASK__DATA__SHIFT 0x0 +#define CP_RB_BUFSZ_MASK__DATA_MASK 0x000FFFFFL +//CP_RB_WPTR_POLL_ADDR_LO +#define CP_RB_WPTR_POLL_ADDR_LO__RB_WPTR_POLL_ADDR_LO__SHIFT 0x2 +#define CP_RB_WPTR_POLL_ADDR_LO__RB_WPTR_POLL_ADDR_LO_MASK 0xFFFFFFFCL +//CP_RB_WPTR_POLL_ADDR_HI +#define CP_RB_WPTR_POLL_ADDR_HI__RB_WPTR_POLL_ADDR_HI__SHIFT 0x0 +#define CP_RB_WPTR_POLL_ADDR_HI__RB_WPTR_POLL_ADDR_HI_MASK 0x0000FFFFL +//GC_PRIV_MODE +//CP_INT_CNTL +#define CP_INT_CNTL__CP_VM_DOORBELL_WR_INT_ENABLE__SHIFT 0xb +#define CP_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_INT_CNTL__CMP_BUSY_INT_ENABLE__SHIFT 0x12 +#define CP_INT_CNTL__CNTX_BUSY_INT_ENABLE__SHIFT 0x13 +#define CP_INT_CNTL__CNTX_EMPTY_INT_ENABLE__SHIFT 0x14 +#define CP_INT_CNTL__GFX_IDLE_INT_ENABLE__SHIFT 0x15 +#define CP_INT_CNTL__PRIV_INSTR_INT_ENABLE__SHIFT 0x16 +#define CP_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_INT_CNTL__CP_VM_DOORBELL_WR_INT_ENABLE_MASK 0x00000800L +#define CP_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_INT_CNTL__CMP_BUSY_INT_ENABLE_MASK 0x00040000L +#define CP_INT_CNTL__CNTX_BUSY_INT_ENABLE_MASK 0x00080000L +#define CP_INT_CNTL__CNTX_EMPTY_INT_ENABLE_MASK 0x00100000L +#define CP_INT_CNTL__GFX_IDLE_INT_ENABLE_MASK 0x00200000L +#define CP_INT_CNTL__PRIV_INSTR_INT_ENABLE_MASK 0x00400000L +#define CP_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_INT_STATUS +#define CP_INT_STATUS__CP_VM_DOORBELL_WR_INT_STAT__SHIFT 0xb +#define CP_INT_STATUS__CP_ECC_ERROR_INT_STAT__SHIFT 0xe +#define CP_INT_STATUS__GPF_INT_STAT__SHIFT 0x10 +#define CP_INT_STATUS__WRM_POLL_TIMEOUT_INT_STAT__SHIFT 0x11 +#define CP_INT_STATUS__CMP_BUSY_INT_STAT__SHIFT 0x12 +#define CP_INT_STATUS__CNTX_BUSY_INT_STAT__SHIFT 0x13 +#define CP_INT_STATUS__CNTX_EMPTY_INT_STAT__SHIFT 0x14 +#define CP_INT_STATUS__GFX_IDLE_INT_STAT__SHIFT 0x15 +#define CP_INT_STATUS__PRIV_INSTR_INT_STAT__SHIFT 0x16 +#define CP_INT_STATUS__PRIV_REG_INT_STAT__SHIFT 0x17 +#define CP_INT_STATUS__OPCODE_ERROR_INT_STAT__SHIFT 0x18 +#define CP_INT_STATUS__TIME_STAMP_INT_STAT__SHIFT 0x1a +#define CP_INT_STATUS__RESERVED_BIT_ERROR_INT_STAT__SHIFT 0x1b +#define CP_INT_STATUS__GENERIC2_INT_STAT__SHIFT 0x1d +#define CP_INT_STATUS__GENERIC1_INT_STAT__SHIFT 0x1e +#define CP_INT_STATUS__GENERIC0_INT_STAT__SHIFT 0x1f +#define CP_INT_STATUS__CP_VM_DOORBELL_WR_INT_STAT_MASK 0x00000800L +#define CP_INT_STATUS__CP_ECC_ERROR_INT_STAT_MASK 0x00004000L +#define CP_INT_STATUS__GPF_INT_STAT_MASK 0x00010000L +#define CP_INT_STATUS__WRM_POLL_TIMEOUT_INT_STAT_MASK 0x00020000L +#define CP_INT_STATUS__CMP_BUSY_INT_STAT_MASK 0x00040000L +#define CP_INT_STATUS__CNTX_BUSY_INT_STAT_MASK 0x00080000L +#define CP_INT_STATUS__CNTX_EMPTY_INT_STAT_MASK 0x00100000L +#define CP_INT_STATUS__GFX_IDLE_INT_STAT_MASK 0x00200000L +#define CP_INT_STATUS__PRIV_INSTR_INT_STAT_MASK 0x00400000L +#define CP_INT_STATUS__PRIV_REG_INT_STAT_MASK 0x00800000L +#define CP_INT_STATUS__OPCODE_ERROR_INT_STAT_MASK 0x01000000L +#define CP_INT_STATUS__TIME_STAMP_INT_STAT_MASK 0x04000000L +#define CP_INT_STATUS__RESERVED_BIT_ERROR_INT_STAT_MASK 0x08000000L +#define CP_INT_STATUS__GENERIC2_INT_STAT_MASK 0x20000000L +#define CP_INT_STATUS__GENERIC1_INT_STAT_MASK 0x40000000L +#define CP_INT_STATUS__GENERIC0_INT_STAT_MASK 0x80000000L +//CP_DEVICE_ID +#define CP_DEVICE_ID__DEVICE_ID__SHIFT 0x0 +#define CP_DEVICE_ID__DEVICE_ID_MASK 0x000000FFL +//CP_ME0_PIPE_PRIORITY_CNTS +#define CP_ME0_PIPE_PRIORITY_CNTS__PRIORITY1_CNT__SHIFT 0x0 +#define CP_ME0_PIPE_PRIORITY_CNTS__PRIORITY2A_CNT__SHIFT 0x8 +#define CP_ME0_PIPE_PRIORITY_CNTS__PRIORITY2B_CNT__SHIFT 0x10 +#define CP_ME0_PIPE_PRIORITY_CNTS__PRIORITY3_CNT__SHIFT 0x18 +#define CP_ME0_PIPE_PRIORITY_CNTS__PRIORITY1_CNT_MASK 0x000000FFL +#define CP_ME0_PIPE_PRIORITY_CNTS__PRIORITY2A_CNT_MASK 0x0000FF00L +#define CP_ME0_PIPE_PRIORITY_CNTS__PRIORITY2B_CNT_MASK 0x00FF0000L +#define CP_ME0_PIPE_PRIORITY_CNTS__PRIORITY3_CNT_MASK 0xFF000000L +//CP_RING_PRIORITY_CNTS +#define CP_RING_PRIORITY_CNTS__PRIORITY1_CNT__SHIFT 0x0 +#define CP_RING_PRIORITY_CNTS__PRIORITY2A_CNT__SHIFT 0x8 +#define CP_RING_PRIORITY_CNTS__PRIORITY2B_CNT__SHIFT 0x10 +#define CP_RING_PRIORITY_CNTS__PRIORITY3_CNT__SHIFT 0x18 +#define CP_RING_PRIORITY_CNTS__PRIORITY1_CNT_MASK 0x000000FFL +#define CP_RING_PRIORITY_CNTS__PRIORITY2A_CNT_MASK 0x0000FF00L +#define CP_RING_PRIORITY_CNTS__PRIORITY2B_CNT_MASK 0x00FF0000L +#define CP_RING_PRIORITY_CNTS__PRIORITY3_CNT_MASK 0xFF000000L +//CP_ME0_PIPE0_PRIORITY +#define CP_ME0_PIPE0_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME0_PIPE0_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_RING0_PRIORITY +#define CP_RING0_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_RING0_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME0_PIPE1_PRIORITY +#define CP_ME0_PIPE1_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME0_PIPE1_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_RING1_PRIORITY +#define CP_RING1_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_RING1_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME0_PIPE2_PRIORITY +#define CP_ME0_PIPE2_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME0_PIPE2_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_RING2_PRIORITY +#define CP_RING2_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_RING2_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_FATAL_ERROR +#define CP_FATAL_ERROR__CPF_FATAL_ERROR__SHIFT 0x0 +#define CP_FATAL_ERROR__CPG_FATAL_ERROR__SHIFT 0x1 +#define CP_FATAL_ERROR__GFX_HALT_PROC__SHIFT 0x2 +#define CP_FATAL_ERROR__DIS_CPG_FATAL_ERROR__SHIFT 0x3 +#define CP_FATAL_ERROR__CPG_TAG_FATAL_ERROR_EN__SHIFT 0x4 +#define CP_FATAL_ERROR__CPF_FATAL_ERROR_MASK 0x00000001L +#define CP_FATAL_ERROR__CPG_FATAL_ERROR_MASK 0x00000002L +#define CP_FATAL_ERROR__GFX_HALT_PROC_MASK 0x00000004L +#define CP_FATAL_ERROR__DIS_CPG_FATAL_ERROR_MASK 0x00000008L +#define CP_FATAL_ERROR__CPG_TAG_FATAL_ERROR_EN_MASK 0x00000010L +//CP_RB_VMID +#define CP_RB_VMID__RB0_VMID__SHIFT 0x0 +#define CP_RB_VMID__RB1_VMID__SHIFT 0x8 +#define CP_RB_VMID__RB2_VMID__SHIFT 0x10 +#define CP_RB_VMID__RB0_VMID_MASK 0x0000000FL +#define CP_RB_VMID__RB1_VMID_MASK 0x00000F00L +#define CP_RB_VMID__RB2_VMID_MASK 0x000F0000L +//CP_ME0_PIPE0_VMID +#define CP_ME0_PIPE0_VMID__VMID__SHIFT 0x0 +#define CP_ME0_PIPE0_VMID__VMID_MASK 0x0000000FL +//CP_ME0_PIPE1_VMID +#define CP_ME0_PIPE1_VMID__VMID__SHIFT 0x0 +#define CP_ME0_PIPE1_VMID__VMID_MASK 0x0000000FL +//CP_RB0_WPTR +#define CP_RB0_WPTR__RB_WPTR__SHIFT 0x0 +#define CP_RB0_WPTR__RB_WPTR_MASK 0xFFFFFFFFL +//CP_RB_WPTR +#define CP_RB_WPTR__RB_WPTR__SHIFT 0x0 +#define CP_RB_WPTR__RB_WPTR_MASK 0xFFFFFFFFL +//CP_RB0_WPTR_HI +#define CP_RB0_WPTR_HI__RB_WPTR__SHIFT 0x0 +#define CP_RB0_WPTR_HI__RB_WPTR_MASK 0xFFFFFFFFL +//CP_RB_WPTR_HI +#define CP_RB_WPTR_HI__RB_WPTR__SHIFT 0x0 +#define CP_RB_WPTR_HI__RB_WPTR_MASK 0xFFFFFFFFL +//CP_RB1_WPTR +#define CP_RB1_WPTR__RB_WPTR__SHIFT 0x0 +#define CP_RB1_WPTR__RB_WPTR_MASK 0xFFFFFFFFL +//CP_RB1_WPTR_HI +#define CP_RB1_WPTR_HI__RB_WPTR__SHIFT 0x0 +#define CP_RB1_WPTR_HI__RB_WPTR_MASK 0xFFFFFFFFL +//CP_RB2_WPTR +#define CP_RB2_WPTR__RB_WPTR__SHIFT 0x0 +#define CP_RB2_WPTR__RB_WPTR_MASK 0x000FFFFFL +//CP_RB_DOORBELL_CONTROL +#define CP_RB_DOORBELL_CONTROL__DOORBELL_BIF_DROP__SHIFT 0x1 +#define CP_RB_DOORBELL_CONTROL__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL__DOORBELL_BIF_DROP_MASK 0x00000002L +#define CP_RB_DOORBELL_CONTROL__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_RANGE_LOWER +#define CP_RB_DOORBELL_RANGE_LOWER__DOORBELL_RANGE_LOWER__SHIFT 0x2 +#define CP_RB_DOORBELL_RANGE_LOWER__DOORBELL_RANGE_LOWER_MASK 0x0FFFFFFCL +//CP_RB_DOORBELL_RANGE_UPPER +#define CP_RB_DOORBELL_RANGE_UPPER__DOORBELL_RANGE_UPPER__SHIFT 0x2 +#define CP_RB_DOORBELL_RANGE_UPPER__DOORBELL_RANGE_UPPER_MASK 0x0FFFFFFCL +//CP_MEC_DOORBELL_RANGE_LOWER +#define CP_MEC_DOORBELL_RANGE_LOWER__DOORBELL_RANGE_LOWER__SHIFT 0x2 +#define CP_MEC_DOORBELL_RANGE_LOWER__DOORBELL_RANGE_LOWER_MASK 0x0FFFFFFCL +//CP_MEC_DOORBELL_RANGE_UPPER +#define CP_MEC_DOORBELL_RANGE_UPPER__DOORBELL_RANGE_UPPER__SHIFT 0x2 +#define CP_MEC_DOORBELL_RANGE_UPPER__DOORBELL_RANGE_UPPER_MASK 0x0FFFFFFCL +//CPG_UTCL1_ERROR +#define CPG_UTCL1_ERROR__ERROR_DETECTED_HALT__SHIFT 0x0 +#define CPG_UTCL1_ERROR__ERROR_DETECTED_HALT_MASK 0x00000001L +//CPC_UTCL1_ERROR +#define CPC_UTCL1_ERROR__ERROR_DETECTED_HALT__SHIFT 0x0 +#define CPC_UTCL1_ERROR__ERROR_DETECTED_HALT_MASK 0x00000001L +//CP_RB1_BASE +#define CP_RB1_BASE__RB_BASE__SHIFT 0x0 +#define CP_RB1_BASE__RB_BASE_MASK 0xFFFFFFFFL +//CP_RB1_CNTL +#define CP_RB1_CNTL__RB_BUFSZ__SHIFT 0x0 +#define CP_RB1_CNTL__RB_BLKSZ__SHIFT 0x8 +#define CP_RB1_CNTL__MIN_AVAILSZ__SHIFT 0x14 +#define CP_RB1_CNTL__MIN_IB_AVAILSZ__SHIFT 0x16 +#define CP_RB1_CNTL__CACHE_POLICY__SHIFT 0x18 +#define CP_RB1_CNTL__RB_NO_UPDATE__SHIFT 0x1b +#define CP_RB1_CNTL__RB_RPTR_WR_ENA__SHIFT 0x1f +#define CP_RB1_CNTL__RB_BUFSZ_MASK 0x0000003FL +#define CP_RB1_CNTL__RB_BLKSZ_MASK 0x00003F00L +#define CP_RB1_CNTL__MIN_AVAILSZ_MASK 0x00300000L +#define CP_RB1_CNTL__MIN_IB_AVAILSZ_MASK 0x00C00000L +#define CP_RB1_CNTL__CACHE_POLICY_MASK 0x01000000L +#define CP_RB1_CNTL__RB_NO_UPDATE_MASK 0x08000000L +#define CP_RB1_CNTL__RB_RPTR_WR_ENA_MASK 0x80000000L +//CP_RB1_RPTR_ADDR +#define CP_RB1_RPTR_ADDR__RB_RPTR_ADDR__SHIFT 0x2 +#define CP_RB1_RPTR_ADDR__RB_RPTR_ADDR_MASK 0xFFFFFFFCL +//CP_RB1_RPTR_ADDR_HI +#define CP_RB1_RPTR_ADDR_HI__RB_RPTR_ADDR_HI__SHIFT 0x0 +#define CP_RB1_RPTR_ADDR_HI__RB_RPTR_ADDR_HI_MASK 0x0000FFFFL +//CP_RB2_BASE +#define CP_RB2_BASE__RB_BASE__SHIFT 0x0 +#define CP_RB2_BASE__RB_BASE_MASK 0xFFFFFFFFL +//CP_RB2_CNTL +#define CP_RB2_CNTL__RB_BUFSZ__SHIFT 0x0 +#define CP_RB2_CNTL__RB_BLKSZ__SHIFT 0x8 +#define CP_RB2_CNTL__MIN_AVAILSZ__SHIFT 0x14 +#define CP_RB2_CNTL__MIN_IB_AVAILSZ__SHIFT 0x16 +#define CP_RB2_CNTL__CACHE_POLICY__SHIFT 0x18 +#define CP_RB2_CNTL__RB_NO_UPDATE__SHIFT 0x1b +#define CP_RB2_CNTL__RB_RPTR_WR_ENA__SHIFT 0x1f +#define CP_RB2_CNTL__RB_BUFSZ_MASK 0x0000003FL +#define CP_RB2_CNTL__RB_BLKSZ_MASK 0x00003F00L +#define CP_RB2_CNTL__MIN_AVAILSZ_MASK 0x00300000L +#define CP_RB2_CNTL__MIN_IB_AVAILSZ_MASK 0x00C00000L +#define CP_RB2_CNTL__CACHE_POLICY_MASK 0x01000000L +#define CP_RB2_CNTL__RB_NO_UPDATE_MASK 0x08000000L +#define CP_RB2_CNTL__RB_RPTR_WR_ENA_MASK 0x80000000L +//CP_RB2_RPTR_ADDR +#define CP_RB2_RPTR_ADDR__RB_RPTR_ADDR__SHIFT 0x2 +#define CP_RB2_RPTR_ADDR__RB_RPTR_ADDR_MASK 0xFFFFFFFCL +//CP_RB2_RPTR_ADDR_HI +#define CP_RB2_RPTR_ADDR_HI__RB_RPTR_ADDR_HI__SHIFT 0x0 +#define CP_RB2_RPTR_ADDR_HI__RB_RPTR_ADDR_HI_MASK 0x0000FFFFL +//CP_RB0_ACTIVE +#define CP_RB0_ACTIVE__ACTIVE__SHIFT 0x0 +#define CP_RB0_ACTIVE__ACTIVE_MASK 0x00000001L +//CP_RB_ACTIVE +#define CP_RB_ACTIVE__ACTIVE__SHIFT 0x0 +#define CP_RB_ACTIVE__ACTIVE_MASK 0x00000001L +//CP_INT_CNTL_RING0 +#define CP_INT_CNTL_RING0__CP_VM_DOORBELL_WR_INT_ENABLE__SHIFT 0xb +#define CP_INT_CNTL_RING0__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_INT_CNTL_RING0__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_INT_CNTL_RING0__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_INT_CNTL_RING0__CMP_BUSY_INT_ENABLE__SHIFT 0x12 +#define CP_INT_CNTL_RING0__CNTX_BUSY_INT_ENABLE__SHIFT 0x13 +#define CP_INT_CNTL_RING0__CNTX_EMPTY_INT_ENABLE__SHIFT 0x14 +#define CP_INT_CNTL_RING0__GFX_IDLE_INT_ENABLE__SHIFT 0x15 +#define CP_INT_CNTL_RING0__PRIV_INSTR_INT_ENABLE__SHIFT 0x16 +#define CP_INT_CNTL_RING0__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_INT_CNTL_RING0__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_INT_CNTL_RING0__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_INT_CNTL_RING0__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_INT_CNTL_RING0__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_INT_CNTL_RING0__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_INT_CNTL_RING0__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_INT_CNTL_RING0__CP_VM_DOORBELL_WR_INT_ENABLE_MASK 0x00000800L +#define CP_INT_CNTL_RING0__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_INT_CNTL_RING0__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_INT_CNTL_RING0__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_INT_CNTL_RING0__CMP_BUSY_INT_ENABLE_MASK 0x00040000L +#define CP_INT_CNTL_RING0__CNTX_BUSY_INT_ENABLE_MASK 0x00080000L +#define CP_INT_CNTL_RING0__CNTX_EMPTY_INT_ENABLE_MASK 0x00100000L +#define CP_INT_CNTL_RING0__GFX_IDLE_INT_ENABLE_MASK 0x00200000L +#define CP_INT_CNTL_RING0__PRIV_INSTR_INT_ENABLE_MASK 0x00400000L +#define CP_INT_CNTL_RING0__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_INT_CNTL_RING0__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_INT_CNTL_RING0__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_INT_CNTL_RING0__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_INT_CNTL_RING0__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_INT_CNTL_RING0__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_INT_CNTL_RING0__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_INT_CNTL_RING1 +#define CP_INT_CNTL_RING1__CP_VM_DOORBELL_WR_INT_ENABLE__SHIFT 0xb +#define CP_INT_CNTL_RING1__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_INT_CNTL_RING1__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_INT_CNTL_RING1__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_INT_CNTL_RING1__CMP_BUSY_INT_ENABLE__SHIFT 0x12 +#define CP_INT_CNTL_RING1__CNTX_BUSY_INT_ENABLE__SHIFT 0x13 +#define CP_INT_CNTL_RING1__CNTX_EMPTY_INT_ENABLE__SHIFT 0x14 +#define CP_INT_CNTL_RING1__GFX_IDLE_INT_ENABLE__SHIFT 0x15 +#define CP_INT_CNTL_RING1__PRIV_INSTR_INT_ENABLE__SHIFT 0x16 +#define CP_INT_CNTL_RING1__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_INT_CNTL_RING1__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_INT_CNTL_RING1__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_INT_CNTL_RING1__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_INT_CNTL_RING1__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_INT_CNTL_RING1__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_INT_CNTL_RING1__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_INT_CNTL_RING1__CP_VM_DOORBELL_WR_INT_ENABLE_MASK 0x00000800L +#define CP_INT_CNTL_RING1__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_INT_CNTL_RING1__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_INT_CNTL_RING1__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_INT_CNTL_RING1__CMP_BUSY_INT_ENABLE_MASK 0x00040000L +#define CP_INT_CNTL_RING1__CNTX_BUSY_INT_ENABLE_MASK 0x00080000L +#define CP_INT_CNTL_RING1__CNTX_EMPTY_INT_ENABLE_MASK 0x00100000L +#define CP_INT_CNTL_RING1__GFX_IDLE_INT_ENABLE_MASK 0x00200000L +#define CP_INT_CNTL_RING1__PRIV_INSTR_INT_ENABLE_MASK 0x00400000L +#define CP_INT_CNTL_RING1__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_INT_CNTL_RING1__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_INT_CNTL_RING1__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_INT_CNTL_RING1__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_INT_CNTL_RING1__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_INT_CNTL_RING1__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_INT_CNTL_RING1__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_INT_CNTL_RING2 +#define CP_INT_CNTL_RING2__CP_VM_DOORBELL_WR_INT_ENABLE__SHIFT 0xb +#define CP_INT_CNTL_RING2__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_INT_CNTL_RING2__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_INT_CNTL_RING2__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_INT_CNTL_RING2__CMP_BUSY_INT_ENABLE__SHIFT 0x12 +#define CP_INT_CNTL_RING2__CNTX_BUSY_INT_ENABLE__SHIFT 0x13 +#define CP_INT_CNTL_RING2__CNTX_EMPTY_INT_ENABLE__SHIFT 0x14 +#define CP_INT_CNTL_RING2__GFX_IDLE_INT_ENABLE__SHIFT 0x15 +#define CP_INT_CNTL_RING2__PRIV_INSTR_INT_ENABLE__SHIFT 0x16 +#define CP_INT_CNTL_RING2__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_INT_CNTL_RING2__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_INT_CNTL_RING2__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_INT_CNTL_RING2__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_INT_CNTL_RING2__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_INT_CNTL_RING2__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_INT_CNTL_RING2__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_INT_CNTL_RING2__CP_VM_DOORBELL_WR_INT_ENABLE_MASK 0x00000800L +#define CP_INT_CNTL_RING2__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_INT_CNTL_RING2__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_INT_CNTL_RING2__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_INT_CNTL_RING2__CMP_BUSY_INT_ENABLE_MASK 0x00040000L +#define CP_INT_CNTL_RING2__CNTX_BUSY_INT_ENABLE_MASK 0x00080000L +#define CP_INT_CNTL_RING2__CNTX_EMPTY_INT_ENABLE_MASK 0x00100000L +#define CP_INT_CNTL_RING2__GFX_IDLE_INT_ENABLE_MASK 0x00200000L +#define CP_INT_CNTL_RING2__PRIV_INSTR_INT_ENABLE_MASK 0x00400000L +#define CP_INT_CNTL_RING2__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_INT_CNTL_RING2__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_INT_CNTL_RING2__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_INT_CNTL_RING2__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_INT_CNTL_RING2__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_INT_CNTL_RING2__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_INT_CNTL_RING2__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_INT_STATUS_RING0 +#define CP_INT_STATUS_RING0__CP_VM_DOORBELL_WR_INT_STAT__SHIFT 0xb +#define CP_INT_STATUS_RING0__CP_ECC_ERROR_INT_STAT__SHIFT 0xe +#define CP_INT_STATUS_RING0__GPF_INT_STAT__SHIFT 0x10 +#define CP_INT_STATUS_RING0__WRM_POLL_TIMEOUT_INT_STAT__SHIFT 0x11 +#define CP_INT_STATUS_RING0__CMP_BUSY_INT_STAT__SHIFT 0x12 +#define CP_INT_STATUS_RING0__GCNTX_BUSY_INT_STAT__SHIFT 0x13 +#define CP_INT_STATUS_RING0__CNTX_EMPTY_INT_STAT__SHIFT 0x14 +#define CP_INT_STATUS_RING0__GFX_IDLE_INT_STAT__SHIFT 0x15 +#define CP_INT_STATUS_RING0__PRIV_INSTR_INT_STAT__SHIFT 0x16 +#define CP_INT_STATUS_RING0__PRIV_REG_INT_STAT__SHIFT 0x17 +#define CP_INT_STATUS_RING0__OPCODE_ERROR_INT_STAT__SHIFT 0x18 +#define CP_INT_STATUS_RING0__TIME_STAMP_INT_STAT__SHIFT 0x1a +#define CP_INT_STATUS_RING0__RESERVED_BIT_ERROR_INT_STAT__SHIFT 0x1b +#define CP_INT_STATUS_RING0__GENERIC2_INT_STAT__SHIFT 0x1d +#define CP_INT_STATUS_RING0__GENERIC1_INT_STAT__SHIFT 0x1e +#define CP_INT_STATUS_RING0__GENERIC0_INT_STAT__SHIFT 0x1f +#define CP_INT_STATUS_RING0__CP_VM_DOORBELL_WR_INT_STAT_MASK 0x00000800L +#define CP_INT_STATUS_RING0__CP_ECC_ERROR_INT_STAT_MASK 0x00004000L +#define CP_INT_STATUS_RING0__GPF_INT_STAT_MASK 0x00010000L +#define CP_INT_STATUS_RING0__WRM_POLL_TIMEOUT_INT_STAT_MASK 0x00020000L +#define CP_INT_STATUS_RING0__CMP_BUSY_INT_STAT_MASK 0x00040000L +#define CP_INT_STATUS_RING0__GCNTX_BUSY_INT_STAT_MASK 0x00080000L +#define CP_INT_STATUS_RING0__CNTX_EMPTY_INT_STAT_MASK 0x00100000L +#define CP_INT_STATUS_RING0__GFX_IDLE_INT_STAT_MASK 0x00200000L +#define CP_INT_STATUS_RING0__PRIV_INSTR_INT_STAT_MASK 0x00400000L +#define CP_INT_STATUS_RING0__PRIV_REG_INT_STAT_MASK 0x00800000L +#define CP_INT_STATUS_RING0__OPCODE_ERROR_INT_STAT_MASK 0x01000000L +#define CP_INT_STATUS_RING0__TIME_STAMP_INT_STAT_MASK 0x04000000L +#define CP_INT_STATUS_RING0__RESERVED_BIT_ERROR_INT_STAT_MASK 0x08000000L +#define CP_INT_STATUS_RING0__GENERIC2_INT_STAT_MASK 0x20000000L +#define CP_INT_STATUS_RING0__GENERIC1_INT_STAT_MASK 0x40000000L +#define CP_INT_STATUS_RING0__GENERIC0_INT_STAT_MASK 0x80000000L +//CP_INT_STATUS_RING1 +#define CP_INT_STATUS_RING1__CP_VM_DOORBELL_WR_INT_STAT__SHIFT 0xb +#define CP_INT_STATUS_RING1__CP_ECC_ERROR_INT_STAT__SHIFT 0xe +#define CP_INT_STATUS_RING1__GPF_INT_STAT__SHIFT 0x10 +#define CP_INT_STATUS_RING1__WRM_POLL_TIMEOUT_INT_STAT__SHIFT 0x11 +#define CP_INT_STATUS_RING1__CMP_BUSY_INT_STAT__SHIFT 0x12 +#define CP_INT_STATUS_RING1__CNTX_BUSY_INT_STAT__SHIFT 0x13 +#define CP_INT_STATUS_RING1__CNTX_EMPTY_INT_STAT__SHIFT 0x14 +#define CP_INT_STATUS_RING1__GFX_IDLE_INT_STAT__SHIFT 0x15 +#define CP_INT_STATUS_RING1__PRIV_INSTR_INT_STAT__SHIFT 0x16 +#define CP_INT_STATUS_RING1__PRIV_REG_INT_STAT__SHIFT 0x17 +#define CP_INT_STATUS_RING1__OPCODE_ERROR_INT_STAT__SHIFT 0x18 +#define CP_INT_STATUS_RING1__TIME_STAMP_INT_STAT__SHIFT 0x1a +#define CP_INT_STATUS_RING1__RESERVED_BIT_ERROR_INT_STAT__SHIFT 0x1b +#define CP_INT_STATUS_RING1__GENERIC2_INT_STAT__SHIFT 0x1d +#define CP_INT_STATUS_RING1__GENERIC1_INT_STAT__SHIFT 0x1e +#define CP_INT_STATUS_RING1__GENERIC0_INT_STAT__SHIFT 0x1f +#define CP_INT_STATUS_RING1__CP_VM_DOORBELL_WR_INT_STAT_MASK 0x00000800L +#define CP_INT_STATUS_RING1__CP_ECC_ERROR_INT_STAT_MASK 0x00004000L +#define CP_INT_STATUS_RING1__GPF_INT_STAT_MASK 0x00010000L +#define CP_INT_STATUS_RING1__WRM_POLL_TIMEOUT_INT_STAT_MASK 0x00020000L +#define CP_INT_STATUS_RING1__CMP_BUSY_INT_STAT_MASK 0x00040000L +#define CP_INT_STATUS_RING1__CNTX_BUSY_INT_STAT_MASK 0x00080000L +#define CP_INT_STATUS_RING1__CNTX_EMPTY_INT_STAT_MASK 0x00100000L +#define CP_INT_STATUS_RING1__GFX_IDLE_INT_STAT_MASK 0x00200000L +#define CP_INT_STATUS_RING1__PRIV_INSTR_INT_STAT_MASK 0x00400000L +#define CP_INT_STATUS_RING1__PRIV_REG_INT_STAT_MASK 0x00800000L +#define CP_INT_STATUS_RING1__OPCODE_ERROR_INT_STAT_MASK 0x01000000L +#define CP_INT_STATUS_RING1__TIME_STAMP_INT_STAT_MASK 0x04000000L +#define CP_INT_STATUS_RING1__RESERVED_BIT_ERROR_INT_STAT_MASK 0x08000000L +#define CP_INT_STATUS_RING1__GENERIC2_INT_STAT_MASK 0x20000000L +#define CP_INT_STATUS_RING1__GENERIC1_INT_STAT_MASK 0x40000000L +#define CP_INT_STATUS_RING1__GENERIC0_INT_STAT_MASK 0x80000000L +//CP_INT_STATUS_RING2 +#define CP_INT_STATUS_RING2__CP_VM_DOORBELL_WR_INT_STAT__SHIFT 0xb +#define CP_INT_STATUS_RING2__CP_ECC_ERROR_INT_STAT__SHIFT 0xe +#define CP_INT_STATUS_RING2__GPF_INT_STAT__SHIFT 0x10 +#define CP_INT_STATUS_RING2__WRM_POLL_TIMEOUT_INT_STAT__SHIFT 0x11 +#define CP_INT_STATUS_RING2__CMP_BUSY_INT_STAT__SHIFT 0x12 +#define CP_INT_STATUS_RING2__CNTX_BUSY_INT_STAT__SHIFT 0x13 +#define CP_INT_STATUS_RING2__CNTX_EMPTY_INT_STAT__SHIFT 0x14 +#define CP_INT_STATUS_RING2__GFX_IDLE_INT_STAT__SHIFT 0x15 +#define CP_INT_STATUS_RING2__PRIV_INSTR_INT_STAT__SHIFT 0x16 +#define CP_INT_STATUS_RING2__PRIV_REG_INT_STAT__SHIFT 0x17 +#define CP_INT_STATUS_RING2__OPCODE_ERROR_INT_STAT__SHIFT 0x18 +#define CP_INT_STATUS_RING2__TIME_STAMP_INT_STAT__SHIFT 0x1a +#define CP_INT_STATUS_RING2__RESERVED_BIT_ERROR_INT_STAT__SHIFT 0x1b +#define CP_INT_STATUS_RING2__GENERIC2_INT_STAT__SHIFT 0x1d +#define CP_INT_STATUS_RING2__GENERIC1_INT_STAT__SHIFT 0x1e +#define CP_INT_STATUS_RING2__GENERIC0_INT_STAT__SHIFT 0x1f +#define CP_INT_STATUS_RING2__CP_VM_DOORBELL_WR_INT_STAT_MASK 0x00000800L +#define CP_INT_STATUS_RING2__CP_ECC_ERROR_INT_STAT_MASK 0x00004000L +#define CP_INT_STATUS_RING2__GPF_INT_STAT_MASK 0x00010000L +#define CP_INT_STATUS_RING2__WRM_POLL_TIMEOUT_INT_STAT_MASK 0x00020000L +#define CP_INT_STATUS_RING2__CMP_BUSY_INT_STAT_MASK 0x00040000L +#define CP_INT_STATUS_RING2__CNTX_BUSY_INT_STAT_MASK 0x00080000L +#define CP_INT_STATUS_RING2__CNTX_EMPTY_INT_STAT_MASK 0x00100000L +#define CP_INT_STATUS_RING2__GFX_IDLE_INT_STAT_MASK 0x00200000L +#define CP_INT_STATUS_RING2__PRIV_INSTR_INT_STAT_MASK 0x00400000L +#define CP_INT_STATUS_RING2__PRIV_REG_INT_STAT_MASK 0x00800000L +#define CP_INT_STATUS_RING2__OPCODE_ERROR_INT_STAT_MASK 0x01000000L +#define CP_INT_STATUS_RING2__TIME_STAMP_INT_STAT_MASK 0x04000000L +#define CP_INT_STATUS_RING2__RESERVED_BIT_ERROR_INT_STAT_MASK 0x08000000L +#define CP_INT_STATUS_RING2__GENERIC2_INT_STAT_MASK 0x20000000L +#define CP_INT_STATUS_RING2__GENERIC1_INT_STAT_MASK 0x40000000L +#define CP_INT_STATUS_RING2__GENERIC0_INT_STAT_MASK 0x80000000L +#define CP_PFP_F32_INTERRUPT__PRIV_REG_INT__SHIFT 0x1 +#define CP_PFP_F32_INTERRUPT__PRIV_REG_INT_MASK 0x00000002L +#define CP_MEC1_F32_INTERRUPT__PRIV_REG_INT__SHIFT 0x1 +#define CP_MEC1_F32_INTERRUPT__PRIV_REG_INT_MASK 0x00000002L +#define CP_MEC2_F32_INTERRUPT__PRIV_REG_INT__SHIFT 0x1 +#define CP_MEC2_F32_INTERRUPT__PRIV_REG_INT_MASK 0x00000002L +//CP_PWR_CNTL +#define CP_PWR_CNTL__GFX_CLK_HALT_ME0_PIPE0__SHIFT 0x0 +#define CP_PWR_CNTL__GFX_CLK_HALT_ME0_PIPE1__SHIFT 0x1 +#define CP_PWR_CNTL__CMP_CLK_HALT_ME1_PIPE0__SHIFT 0x8 +#define CP_PWR_CNTL__CMP_CLK_HALT_ME1_PIPE1__SHIFT 0x9 +#define CP_PWR_CNTL__CMP_CLK_HALT_ME1_PIPE2__SHIFT 0xa +#define CP_PWR_CNTL__CMP_CLK_HALT_ME1_PIPE3__SHIFT 0xb +#define CP_PWR_CNTL__CMP_CLK_HALT_ME2_PIPE0__SHIFT 0x10 +#define CP_PWR_CNTL__CMP_CLK_HALT_ME2_PIPE1__SHIFT 0x11 +#define CP_PWR_CNTL__CMP_CLK_HALT_ME2_PIPE2__SHIFT 0x12 +#define CP_PWR_CNTL__CMP_CLK_HALT_ME2_PIPE3__SHIFT 0x13 +#define CP_PWR_CNTL__GFX_CLK_HALT_ME0_PIPE0_MASK 0x00000001L +#define CP_PWR_CNTL__GFX_CLK_HALT_ME0_PIPE1_MASK 0x00000002L +#define CP_PWR_CNTL__CMP_CLK_HALT_ME1_PIPE0_MASK 0x00000100L +#define CP_PWR_CNTL__CMP_CLK_HALT_ME1_PIPE1_MASK 0x00000200L +#define CP_PWR_CNTL__CMP_CLK_HALT_ME1_PIPE2_MASK 0x00000400L +#define CP_PWR_CNTL__CMP_CLK_HALT_ME1_PIPE3_MASK 0x00000800L +#define CP_PWR_CNTL__CMP_CLK_HALT_ME2_PIPE0_MASK 0x00010000L +#define CP_PWR_CNTL__CMP_CLK_HALT_ME2_PIPE1_MASK 0x00020000L +#define CP_PWR_CNTL__CMP_CLK_HALT_ME2_PIPE2_MASK 0x00040000L +#define CP_PWR_CNTL__CMP_CLK_HALT_ME2_PIPE3_MASK 0x00080000L +//CP_MEM_SLP_CNTL +#define CP_MEM_SLP_CNTL__CP_MEM_LS_EN__SHIFT 0x0 +#define CP_MEM_SLP_CNTL__CP_MEM_DS_EN__SHIFT 0x1 +#define CP_MEM_SLP_CNTL__RESERVED__SHIFT 0x2 +#define CP_MEM_SLP_CNTL__CP_LS_DS_BUSY_OVERRIDE__SHIFT 0x7 +#define CP_MEM_SLP_CNTL__CP_MEM_LS_ON_DELAY__SHIFT 0x8 +#define CP_MEM_SLP_CNTL__CP_MEM_LS_OFF_DELAY__SHIFT 0x10 +#define CP_MEM_SLP_CNTL__RESERVED1__SHIFT 0x18 +#define CP_MEM_SLP_CNTL__CP_MEM_LS_EN_MASK 0x00000001L +#define CP_MEM_SLP_CNTL__CP_MEM_DS_EN_MASK 0x00000002L +#define CP_MEM_SLP_CNTL__RESERVED_MASK 0x0000007CL +#define CP_MEM_SLP_CNTL__CP_LS_DS_BUSY_OVERRIDE_MASK 0x00000080L +#define CP_MEM_SLP_CNTL__CP_MEM_LS_ON_DELAY_MASK 0x0000FF00L +#define CP_MEM_SLP_CNTL__CP_MEM_LS_OFF_DELAY_MASK 0x00FF0000L +#define CP_MEM_SLP_CNTL__RESERVED1_MASK 0xFF000000L +//CP_ECC_FIRSTOCCURRENCE +#define CP_ECC_FIRSTOCCURRENCE__INTERFACE__SHIFT 0x0 +#define CP_ECC_FIRSTOCCURRENCE__CLIENT__SHIFT 0x4 +#define CP_ECC_FIRSTOCCURRENCE__ME__SHIFT 0x8 +#define CP_ECC_FIRSTOCCURRENCE__PIPE__SHIFT 0xa +#define CP_ECC_FIRSTOCCURRENCE__QUEUE__SHIFT 0xc +#define CP_ECC_FIRSTOCCURRENCE__VMID__SHIFT 0x10 +#define CP_ECC_FIRSTOCCURRENCE__INTERFACE_MASK 0x00000003L +#define CP_ECC_FIRSTOCCURRENCE__CLIENT_MASK 0x000000F0L +#define CP_ECC_FIRSTOCCURRENCE__ME_MASK 0x00000300L +#define CP_ECC_FIRSTOCCURRENCE__PIPE_MASK 0x00000C00L +#define CP_ECC_FIRSTOCCURRENCE__QUEUE_MASK 0x00007000L +#define CP_ECC_FIRSTOCCURRENCE__VMID_MASK 0x000F0000L +//CP_ECC_FIRSTOCCURRENCE_RING0 +#define CP_ECC_FIRSTOCCURRENCE_RING0__OBSOLETE__SHIFT 0x0 +#define CP_ECC_FIRSTOCCURRENCE_RING0__OBSOLETE_MASK 0xFFFFFFFFL +//CP_ECC_FIRSTOCCURRENCE_RING1 +#define CP_ECC_FIRSTOCCURRENCE_RING1__OBSOLETE__SHIFT 0x0 +#define CP_ECC_FIRSTOCCURRENCE_RING1__OBSOLETE_MASK 0xFFFFFFFFL +//CP_ECC_FIRSTOCCURRENCE_RING2 +#define CP_ECC_FIRSTOCCURRENCE_RING2__OBSOLETE__SHIFT 0x0 +#define CP_ECC_FIRSTOCCURRENCE_RING2__OBSOLETE_MASK 0xFFFFFFFFL +//GB_EDC_MODE +#define GB_EDC_MODE__FORCE_SEC_ON_DED__SHIFT 0xf +#define GB_EDC_MODE__COUNT_FED_OUT__SHIFT 0x10 +#define GB_EDC_MODE__GATE_FUE__SHIFT 0x11 +#define GB_EDC_MODE__DED_MODE__SHIFT 0x14 +#define GB_EDC_MODE__PROP_FED__SHIFT 0x1d +#define GB_EDC_MODE__BYPASS__SHIFT 0x1f +#define GB_EDC_MODE__FORCE_SEC_ON_DED_MASK 0x00008000L +#define GB_EDC_MODE__COUNT_FED_OUT_MASK 0x00010000L +#define GB_EDC_MODE__GATE_FUE_MASK 0x00020000L +#define GB_EDC_MODE__DED_MODE_MASK 0x00300000L +#define GB_EDC_MODE__PROP_FED_MASK 0x20000000L +#define GB_EDC_MODE__BYPASS_MASK 0x80000000L +//CP_PQ_WPTR_POLL_CNTL +#define CP_PQ_WPTR_POLL_CNTL__PERIOD__SHIFT 0x0 +#define CP_PQ_WPTR_POLL_CNTL__DISABLE_PEND_REQ_ONE_SHOT__SHIFT 0x1d +#define CP_PQ_WPTR_POLL_CNTL__POLL_ACTIVE__SHIFT 0x1e +#define CP_PQ_WPTR_POLL_CNTL__EN__SHIFT 0x1f +#define CP_PQ_WPTR_POLL_CNTL__PERIOD_MASK 0x000000FFL +#define CP_PQ_WPTR_POLL_CNTL__DISABLE_PEND_REQ_ONE_SHOT_MASK 0x20000000L +#define CP_PQ_WPTR_POLL_CNTL__POLL_ACTIVE_MASK 0x40000000L +#define CP_PQ_WPTR_POLL_CNTL__EN_MASK 0x80000000L +//CP_PQ_WPTR_POLL_CNTL1 +#define CP_PQ_WPTR_POLL_CNTL1__QUEUE_MASK__SHIFT 0x0 +#define CP_PQ_WPTR_POLL_CNTL1__QUEUE_MASK_MASK 0xFFFFFFFFL +//CP_ME1_PIPE0_INT_CNTL +#define CP_ME1_PIPE0_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CP_ME1_PIPE0_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CP_ME1_PIPE0_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_ME1_PIPE0_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CP_ME1_PIPE0_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_ME1_PIPE0_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_ME1_PIPE0_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_ME1_PIPE0_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_ME1_PIPE0_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_ME1_PIPE0_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_ME1_PIPE0_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_ME1_PIPE0_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_ME1_PIPE0_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_ME1_PIPE0_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CP_ME1_PIPE0_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CP_ME1_PIPE0_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_ME1_PIPE0_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CP_ME1_PIPE0_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_ME1_PIPE0_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_ME1_PIPE0_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_ME1_PIPE0_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_ME1_PIPE0_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_ME1_PIPE0_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_ME1_PIPE0_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_ME1_PIPE0_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_ME1_PIPE0_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_ME1_PIPE1_INT_CNTL +#define CP_ME1_PIPE1_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CP_ME1_PIPE1_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CP_ME1_PIPE1_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_ME1_PIPE1_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CP_ME1_PIPE1_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_ME1_PIPE1_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_ME1_PIPE1_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_ME1_PIPE1_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_ME1_PIPE1_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_ME1_PIPE1_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_ME1_PIPE1_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_ME1_PIPE1_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_ME1_PIPE1_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_ME1_PIPE1_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CP_ME1_PIPE1_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CP_ME1_PIPE1_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_ME1_PIPE1_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CP_ME1_PIPE1_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_ME1_PIPE1_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_ME1_PIPE1_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_ME1_PIPE1_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_ME1_PIPE1_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_ME1_PIPE1_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_ME1_PIPE1_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_ME1_PIPE1_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_ME1_PIPE1_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_ME1_PIPE2_INT_CNTL +#define CP_ME1_PIPE2_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CP_ME1_PIPE2_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CP_ME1_PIPE2_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_ME1_PIPE2_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CP_ME1_PIPE2_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_ME1_PIPE2_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_ME1_PIPE2_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_ME1_PIPE2_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_ME1_PIPE2_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_ME1_PIPE2_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_ME1_PIPE2_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_ME1_PIPE2_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_ME1_PIPE2_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_ME1_PIPE2_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CP_ME1_PIPE2_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CP_ME1_PIPE2_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_ME1_PIPE2_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CP_ME1_PIPE2_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_ME1_PIPE2_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_ME1_PIPE2_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_ME1_PIPE2_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_ME1_PIPE2_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_ME1_PIPE2_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_ME1_PIPE2_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_ME1_PIPE2_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_ME1_PIPE2_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_ME1_PIPE3_INT_CNTL +#define CP_ME1_PIPE3_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CP_ME1_PIPE3_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CP_ME1_PIPE3_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_ME1_PIPE3_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CP_ME1_PIPE3_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_ME1_PIPE3_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_ME1_PIPE3_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_ME1_PIPE3_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_ME1_PIPE3_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_ME1_PIPE3_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_ME1_PIPE3_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_ME1_PIPE3_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_ME1_PIPE3_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_ME1_PIPE3_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CP_ME1_PIPE3_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CP_ME1_PIPE3_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_ME1_PIPE3_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CP_ME1_PIPE3_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_ME1_PIPE3_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_ME1_PIPE3_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_ME1_PIPE3_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_ME1_PIPE3_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_ME1_PIPE3_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_ME1_PIPE3_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_ME1_PIPE3_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_ME1_PIPE3_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_ME2_PIPE0_INT_CNTL +#define CP_ME2_PIPE0_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CP_ME2_PIPE0_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CP_ME2_PIPE0_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_ME2_PIPE0_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CP_ME2_PIPE0_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_ME2_PIPE0_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_ME2_PIPE0_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_ME2_PIPE0_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_ME2_PIPE0_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_ME2_PIPE0_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_ME2_PIPE0_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_ME2_PIPE0_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_ME2_PIPE0_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_ME2_PIPE0_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CP_ME2_PIPE0_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CP_ME2_PIPE0_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_ME2_PIPE0_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CP_ME2_PIPE0_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_ME2_PIPE0_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_ME2_PIPE0_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_ME2_PIPE0_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_ME2_PIPE0_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_ME2_PIPE0_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_ME2_PIPE0_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_ME2_PIPE0_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_ME2_PIPE0_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_ME2_PIPE1_INT_CNTL +#define CP_ME2_PIPE1_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CP_ME2_PIPE1_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CP_ME2_PIPE1_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_ME2_PIPE1_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CP_ME2_PIPE1_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_ME2_PIPE1_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_ME2_PIPE1_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_ME2_PIPE1_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_ME2_PIPE1_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_ME2_PIPE1_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_ME2_PIPE1_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_ME2_PIPE1_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_ME2_PIPE1_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_ME2_PIPE1_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CP_ME2_PIPE1_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CP_ME2_PIPE1_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_ME2_PIPE1_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CP_ME2_PIPE1_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_ME2_PIPE1_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_ME2_PIPE1_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_ME2_PIPE1_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_ME2_PIPE1_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_ME2_PIPE1_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_ME2_PIPE1_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_ME2_PIPE1_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_ME2_PIPE1_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_ME2_PIPE2_INT_CNTL +#define CP_ME2_PIPE2_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CP_ME2_PIPE2_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CP_ME2_PIPE2_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_ME2_PIPE2_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CP_ME2_PIPE2_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_ME2_PIPE2_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_ME2_PIPE2_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_ME2_PIPE2_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_ME2_PIPE2_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_ME2_PIPE2_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_ME2_PIPE2_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_ME2_PIPE2_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_ME2_PIPE2_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_ME2_PIPE2_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CP_ME2_PIPE2_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CP_ME2_PIPE2_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_ME2_PIPE2_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CP_ME2_PIPE2_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_ME2_PIPE2_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_ME2_PIPE2_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_ME2_PIPE2_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_ME2_PIPE2_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_ME2_PIPE2_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_ME2_PIPE2_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_ME2_PIPE2_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_ME2_PIPE2_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_ME2_PIPE3_INT_CNTL +#define CP_ME2_PIPE3_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CP_ME2_PIPE3_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CP_ME2_PIPE3_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CP_ME2_PIPE3_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CP_ME2_PIPE3_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CP_ME2_PIPE3_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CP_ME2_PIPE3_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CP_ME2_PIPE3_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CP_ME2_PIPE3_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CP_ME2_PIPE3_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CP_ME2_PIPE3_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CP_ME2_PIPE3_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CP_ME2_PIPE3_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CP_ME2_PIPE3_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CP_ME2_PIPE3_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CP_ME2_PIPE3_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CP_ME2_PIPE3_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CP_ME2_PIPE3_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CP_ME2_PIPE3_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CP_ME2_PIPE3_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CP_ME2_PIPE3_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CP_ME2_PIPE3_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CP_ME2_PIPE3_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CP_ME2_PIPE3_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CP_ME2_PIPE3_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CP_ME2_PIPE3_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CP_ME1_PIPE0_INT_STATUS +#define CP_ME1_PIPE0_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CP_ME1_PIPE0_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CP_ME1_PIPE0_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CP_ME1_PIPE0_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CP_ME1_PIPE0_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CP_ME1_PIPE0_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CP_ME1_PIPE0_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CP_ME1_PIPE0_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CP_ME1_PIPE0_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CP_ME1_PIPE0_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CP_ME1_PIPE0_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CP_ME1_PIPE0_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CP_ME1_PIPE0_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CP_ME1_PIPE0_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CP_ME1_PIPE0_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CP_ME1_PIPE0_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CP_ME1_PIPE0_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CP_ME1_PIPE0_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CP_ME1_PIPE0_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CP_ME1_PIPE0_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CP_ME1_PIPE0_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CP_ME1_PIPE0_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CP_ME1_PIPE0_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CP_ME1_PIPE0_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CP_ME1_PIPE0_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CP_ME1_PIPE0_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +//CP_ME1_PIPE1_INT_STATUS +#define CP_ME1_PIPE1_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CP_ME1_PIPE1_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CP_ME1_PIPE1_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CP_ME1_PIPE1_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CP_ME1_PIPE1_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CP_ME1_PIPE1_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CP_ME1_PIPE1_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CP_ME1_PIPE1_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CP_ME1_PIPE1_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CP_ME1_PIPE1_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CP_ME1_PIPE1_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CP_ME1_PIPE1_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CP_ME1_PIPE1_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CP_ME1_PIPE1_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CP_ME1_PIPE1_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CP_ME1_PIPE1_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CP_ME1_PIPE1_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CP_ME1_PIPE1_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CP_ME1_PIPE1_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CP_ME1_PIPE1_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CP_ME1_PIPE1_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CP_ME1_PIPE1_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CP_ME1_PIPE1_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CP_ME1_PIPE1_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CP_ME1_PIPE1_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CP_ME1_PIPE1_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +//CP_ME1_PIPE2_INT_STATUS +#define CP_ME1_PIPE2_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CP_ME1_PIPE2_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CP_ME1_PIPE2_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CP_ME1_PIPE2_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CP_ME1_PIPE2_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CP_ME1_PIPE2_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CP_ME1_PIPE2_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CP_ME1_PIPE2_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CP_ME1_PIPE2_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CP_ME1_PIPE2_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CP_ME1_PIPE2_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CP_ME1_PIPE2_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CP_ME1_PIPE2_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CP_ME1_PIPE2_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CP_ME1_PIPE2_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CP_ME1_PIPE2_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CP_ME1_PIPE2_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CP_ME1_PIPE2_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CP_ME1_PIPE2_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CP_ME1_PIPE2_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CP_ME1_PIPE2_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CP_ME1_PIPE2_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CP_ME1_PIPE2_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CP_ME1_PIPE2_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CP_ME1_PIPE2_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CP_ME1_PIPE2_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +//CP_ME1_PIPE3_INT_STATUS +#define CP_ME1_PIPE3_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CP_ME1_PIPE3_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CP_ME1_PIPE3_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CP_ME1_PIPE3_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CP_ME1_PIPE3_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CP_ME1_PIPE3_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CP_ME1_PIPE3_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CP_ME1_PIPE3_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CP_ME1_PIPE3_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CP_ME1_PIPE3_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CP_ME1_PIPE3_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CP_ME1_PIPE3_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CP_ME1_PIPE3_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CP_ME1_PIPE3_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CP_ME1_PIPE3_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CP_ME1_PIPE3_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CP_ME1_PIPE3_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CP_ME1_PIPE3_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CP_ME1_PIPE3_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CP_ME1_PIPE3_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CP_ME1_PIPE3_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CP_ME1_PIPE3_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CP_ME1_PIPE3_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CP_ME1_PIPE3_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CP_ME1_PIPE3_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CP_ME1_PIPE3_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +//CP_ME2_PIPE0_INT_STATUS +#define CP_ME2_PIPE0_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CP_ME2_PIPE0_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CP_ME2_PIPE0_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CP_ME2_PIPE0_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CP_ME2_PIPE0_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CP_ME2_PIPE0_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CP_ME2_PIPE0_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CP_ME2_PIPE0_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CP_ME2_PIPE0_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CP_ME2_PIPE0_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CP_ME2_PIPE0_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CP_ME2_PIPE0_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CP_ME2_PIPE0_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CP_ME2_PIPE0_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CP_ME2_PIPE0_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CP_ME2_PIPE0_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CP_ME2_PIPE0_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CP_ME2_PIPE0_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CP_ME2_PIPE0_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CP_ME2_PIPE0_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CP_ME2_PIPE0_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CP_ME2_PIPE0_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CP_ME2_PIPE0_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CP_ME2_PIPE0_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CP_ME2_PIPE0_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CP_ME2_PIPE0_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +//CP_ME2_PIPE1_INT_STATUS +#define CP_ME2_PIPE1_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CP_ME2_PIPE1_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CP_ME2_PIPE1_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CP_ME2_PIPE1_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CP_ME2_PIPE1_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CP_ME2_PIPE1_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CP_ME2_PIPE1_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CP_ME2_PIPE1_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CP_ME2_PIPE1_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CP_ME2_PIPE1_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CP_ME2_PIPE1_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CP_ME2_PIPE1_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CP_ME2_PIPE1_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CP_ME2_PIPE1_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CP_ME2_PIPE1_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CP_ME2_PIPE1_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CP_ME2_PIPE1_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CP_ME2_PIPE1_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CP_ME2_PIPE1_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CP_ME2_PIPE1_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CP_ME2_PIPE1_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CP_ME2_PIPE1_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CP_ME2_PIPE1_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CP_ME2_PIPE1_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CP_ME2_PIPE1_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CP_ME2_PIPE1_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +//CP_ME2_PIPE2_INT_STATUS +#define CP_ME2_PIPE2_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CP_ME2_PIPE2_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CP_ME2_PIPE2_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CP_ME2_PIPE2_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CP_ME2_PIPE2_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CP_ME2_PIPE2_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CP_ME2_PIPE2_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CP_ME2_PIPE2_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CP_ME2_PIPE2_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CP_ME2_PIPE2_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CP_ME2_PIPE2_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CP_ME2_PIPE2_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CP_ME2_PIPE2_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CP_ME2_PIPE2_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CP_ME2_PIPE2_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CP_ME2_PIPE2_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CP_ME2_PIPE2_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CP_ME2_PIPE2_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CP_ME2_PIPE2_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CP_ME2_PIPE2_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CP_ME2_PIPE2_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CP_ME2_PIPE2_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CP_ME2_PIPE2_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CP_ME2_PIPE2_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CP_ME2_PIPE2_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CP_ME2_PIPE2_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +//CP_ME2_PIPE3_INT_STATUS +#define CP_ME2_PIPE3_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CP_ME2_PIPE3_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CP_ME2_PIPE3_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CP_ME2_PIPE3_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CP_ME2_PIPE3_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CP_ME2_PIPE3_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CP_ME2_PIPE3_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CP_ME2_PIPE3_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CP_ME2_PIPE3_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CP_ME2_PIPE3_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CP_ME2_PIPE3_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CP_ME2_PIPE3_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CP_ME2_PIPE3_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CP_ME2_PIPE3_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CP_ME2_PIPE3_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CP_ME2_PIPE3_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CP_ME2_PIPE3_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CP_ME2_PIPE3_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CP_ME2_PIPE3_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CP_ME2_PIPE3_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CP_ME2_PIPE3_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CP_ME2_PIPE3_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CP_ME2_PIPE3_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CP_ME2_PIPE3_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CP_ME2_PIPE3_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CP_ME2_PIPE3_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +#define CP_ME1_INT_STAT_DEBUG__PRIV_REG_INT_ASSERTED__SHIFT 0x17 +#define CP_ME1_INT_STAT_DEBUG__PRIV_REG_INT_ASSERTED_MASK 0x00800000L +#define CP_ME2_INT_STAT_DEBUG__PRIV_REG_INT_ASSERTED__SHIFT 0x17 +#define CP_ME2_INT_STAT_DEBUG__PRIV_REG_INT_ASSERTED_MASK 0x00800000L +//CC_GC_EDC_CONFIG +#define CC_GC_EDC_CONFIG__DIS_EDC__SHIFT 0x1 +#define CC_GC_EDC_CONFIG__DIS_EDC_MASK 0x00000002L +//CP_ME1_PIPE_PRIORITY_CNTS +#define CP_ME1_PIPE_PRIORITY_CNTS__PRIORITY1_CNT__SHIFT 0x0 +#define CP_ME1_PIPE_PRIORITY_CNTS__PRIORITY2A_CNT__SHIFT 0x8 +#define CP_ME1_PIPE_PRIORITY_CNTS__PRIORITY2B_CNT__SHIFT 0x10 +#define CP_ME1_PIPE_PRIORITY_CNTS__PRIORITY3_CNT__SHIFT 0x18 +#define CP_ME1_PIPE_PRIORITY_CNTS__PRIORITY1_CNT_MASK 0x000000FFL +#define CP_ME1_PIPE_PRIORITY_CNTS__PRIORITY2A_CNT_MASK 0x0000FF00L +#define CP_ME1_PIPE_PRIORITY_CNTS__PRIORITY2B_CNT_MASK 0x00FF0000L +#define CP_ME1_PIPE_PRIORITY_CNTS__PRIORITY3_CNT_MASK 0xFF000000L +//CP_ME1_PIPE0_PRIORITY +#define CP_ME1_PIPE0_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME1_PIPE0_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME1_PIPE1_PRIORITY +#define CP_ME1_PIPE1_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME1_PIPE1_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME1_PIPE2_PRIORITY +#define CP_ME1_PIPE2_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME1_PIPE2_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME1_PIPE3_PRIORITY +#define CP_ME1_PIPE3_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME1_PIPE3_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME2_PIPE_PRIORITY_CNTS +#define CP_ME2_PIPE_PRIORITY_CNTS__PRIORITY1_CNT__SHIFT 0x0 +#define CP_ME2_PIPE_PRIORITY_CNTS__PRIORITY2A_CNT__SHIFT 0x8 +#define CP_ME2_PIPE_PRIORITY_CNTS__PRIORITY2B_CNT__SHIFT 0x10 +#define CP_ME2_PIPE_PRIORITY_CNTS__PRIORITY3_CNT__SHIFT 0x18 +#define CP_ME2_PIPE_PRIORITY_CNTS__PRIORITY1_CNT_MASK 0x000000FFL +#define CP_ME2_PIPE_PRIORITY_CNTS__PRIORITY2A_CNT_MASK 0x0000FF00L +#define CP_ME2_PIPE_PRIORITY_CNTS__PRIORITY2B_CNT_MASK 0x00FF0000L +#define CP_ME2_PIPE_PRIORITY_CNTS__PRIORITY3_CNT_MASK 0xFF000000L +//CP_ME2_PIPE0_PRIORITY +#define CP_ME2_PIPE0_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME2_PIPE0_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME2_PIPE1_PRIORITY +#define CP_ME2_PIPE1_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME2_PIPE1_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME2_PIPE2_PRIORITY +#define CP_ME2_PIPE2_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME2_PIPE2_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_ME2_PIPE3_PRIORITY +#define CP_ME2_PIPE3_PRIORITY__PRIORITY__SHIFT 0x0 +#define CP_ME2_PIPE3_PRIORITY__PRIORITY_MASK 0x00000003L +//CP_CE_PRGRM_CNTR_START +#define CP_CE_PRGRM_CNTR_START__IP_START__SHIFT 0x0 +#define CP_CE_PRGRM_CNTR_START__IP_START_MASK 0x000007FFL +//CP_PFP_PRGRM_CNTR_START +#define CP_PFP_PRGRM_CNTR_START__IP_START__SHIFT 0x0 +#define CP_PFP_PRGRM_CNTR_START__IP_START_MASK 0x00001FFFL +//CP_ME_PRGRM_CNTR_START +#define CP_ME_PRGRM_CNTR_START__IP_START__SHIFT 0x0 +#define CP_ME_PRGRM_CNTR_START__IP_START_MASK 0x00000FFFL +//CP_MEC1_PRGRM_CNTR_START +#define CP_MEC1_PRGRM_CNTR_START__IP_START__SHIFT 0x0 +#define CP_MEC1_PRGRM_CNTR_START__IP_START_MASK 0x0000FFFFL +//CP_MEC2_PRGRM_CNTR_START +#define CP_MEC2_PRGRM_CNTR_START__IP_START__SHIFT 0x0 +#define CP_MEC2_PRGRM_CNTR_START__IP_START_MASK 0x0000FFFFL +//CP_CE_INTR_ROUTINE_START +#define CP_CE_INTR_ROUTINE_START__IR_START__SHIFT 0x0 +#define CP_CE_INTR_ROUTINE_START__IR_START_MASK 0x000007FFL +//CP_PFP_INTR_ROUTINE_START +#define CP_PFP_INTR_ROUTINE_START__IR_START__SHIFT 0x0 +#define CP_PFP_INTR_ROUTINE_START__IR_START_MASK 0x00001FFFL +//CP_ME_INTR_ROUTINE_START +#define CP_ME_INTR_ROUTINE_START__IR_START__SHIFT 0x0 +#define CP_ME_INTR_ROUTINE_START__IR_START_MASK 0x00000FFFL +//CP_MEC1_INTR_ROUTINE_START +#define CP_MEC1_INTR_ROUTINE_START__IR_START__SHIFT 0x0 +#define CP_MEC1_INTR_ROUTINE_START__IR_START_MASK 0x0000FFFFL +//CP_MEC2_INTR_ROUTINE_START +#define CP_MEC2_INTR_ROUTINE_START__IR_START__SHIFT 0x0 +#define CP_MEC2_INTR_ROUTINE_START__IR_START_MASK 0x0000FFFFL +//CP_CONTEXT_CNTL +#define CP_CONTEXT_CNTL__ME0PIPE0_MAX_WD_CNTX__SHIFT 0x0 +#define CP_CONTEXT_CNTL__ME0PIPE0_MAX_PIPE_CNTX__SHIFT 0x4 +#define CP_CONTEXT_CNTL__ME0PIPE1_MAX_WD_CNTX__SHIFT 0x10 +#define CP_CONTEXT_CNTL__ME0PIPE1_MAX_PIPE_CNTX__SHIFT 0x14 +#define CP_CONTEXT_CNTL__ME0PIPE0_MAX_WD_CNTX_MASK 0x00000007L +#define CP_CONTEXT_CNTL__ME0PIPE0_MAX_PIPE_CNTX_MASK 0x00000070L +#define CP_CONTEXT_CNTL__ME0PIPE1_MAX_WD_CNTX_MASK 0x00070000L +#define CP_CONTEXT_CNTL__ME0PIPE1_MAX_PIPE_CNTX_MASK 0x00700000L +//CP_MAX_CONTEXT +#define CP_MAX_CONTEXT__MAX_CONTEXT__SHIFT 0x0 +#define CP_MAX_CONTEXT__MAX_CONTEXT_MASK 0x00000007L +//CP_IQ_WAIT_TIME1 +#define CP_IQ_WAIT_TIME1__IB_OFFLOAD__SHIFT 0x0 +#define CP_IQ_WAIT_TIME1__ATOMIC_OFFLOAD__SHIFT 0x8 +#define CP_IQ_WAIT_TIME1__WRM_OFFLOAD__SHIFT 0x10 +#define CP_IQ_WAIT_TIME1__GWS__SHIFT 0x18 +#define CP_IQ_WAIT_TIME1__IB_OFFLOAD_MASK 0x000000FFL +#define CP_IQ_WAIT_TIME1__ATOMIC_OFFLOAD_MASK 0x0000FF00L +#define CP_IQ_WAIT_TIME1__WRM_OFFLOAD_MASK 0x00FF0000L +#define CP_IQ_WAIT_TIME1__GWS_MASK 0xFF000000L +//CP_IQ_WAIT_TIME2 +#define CP_IQ_WAIT_TIME2__QUE_SLEEP__SHIFT 0x0 +#define CP_IQ_WAIT_TIME2__SCH_WAVE__SHIFT 0x8 +#define CP_IQ_WAIT_TIME2__SEM_REARM__SHIFT 0x10 +#define CP_IQ_WAIT_TIME2__DEQ_RETRY__SHIFT 0x18 +#define CP_IQ_WAIT_TIME2__QUE_SLEEP_MASK 0x000000FFL +#define CP_IQ_WAIT_TIME2__SCH_WAVE_MASK 0x0000FF00L +#define CP_IQ_WAIT_TIME2__SEM_REARM_MASK 0x00FF0000L +#define CP_IQ_WAIT_TIME2__DEQ_RETRY_MASK 0xFF000000L +//CP_RB0_BASE_HI +#define CP_RB0_BASE_HI__RB_BASE_HI__SHIFT 0x0 +#define CP_RB0_BASE_HI__RB_BASE_HI_MASK 0x000000FFL +//CP_RB1_BASE_HI +#define CP_RB1_BASE_HI__RB_BASE_HI__SHIFT 0x0 +#define CP_RB1_BASE_HI__RB_BASE_HI_MASK 0x000000FFL +//CP_VMID_RESET +#define CP_VMID_RESET__RESET_REQUEST__SHIFT 0x0 +#define CP_VMID_RESET__RESET_REQUEST_MASK 0x0000FFFFL +//CPC_INT_CNTL +#define CPC_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE__SHIFT 0xc +#define CPC_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE__SHIFT 0xd +#define CPC_INT_CNTL__CP_ECC_ERROR_INT_ENABLE__SHIFT 0xe +#define CPC_INT_CNTL__SUA_VIOLATION_INT_ENABLE__SHIFT 0xf +#define CPC_INT_CNTL__GPF_INT_ENABLE__SHIFT 0x10 +#define CPC_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE__SHIFT 0x11 +#define CPC_INT_CNTL__PRIV_REG_INT_ENABLE__SHIFT 0x17 +#define CPC_INT_CNTL__OPCODE_ERROR_INT_ENABLE__SHIFT 0x18 +#define CPC_INT_CNTL__TIME_STAMP_INT_ENABLE__SHIFT 0x1a +#define CPC_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE__SHIFT 0x1b +#define CPC_INT_CNTL__GENERIC2_INT_ENABLE__SHIFT 0x1d +#define CPC_INT_CNTL__GENERIC1_INT_ENABLE__SHIFT 0x1e +#define CPC_INT_CNTL__GENERIC0_INT_ENABLE__SHIFT 0x1f +#define CPC_INT_CNTL__CMP_QUERY_STATUS_INT_ENABLE_MASK 0x00001000L +#define CPC_INT_CNTL__DEQUEUE_REQUEST_INT_ENABLE_MASK 0x00002000L +#define CPC_INT_CNTL__CP_ECC_ERROR_INT_ENABLE_MASK 0x00004000L +#define CPC_INT_CNTL__SUA_VIOLATION_INT_ENABLE_MASK 0x00008000L +#define CPC_INT_CNTL__GPF_INT_ENABLE_MASK 0x00010000L +#define CPC_INT_CNTL__WRM_POLL_TIMEOUT_INT_ENABLE_MASK 0x00020000L +#define CPC_INT_CNTL__PRIV_REG_INT_ENABLE_MASK 0x00800000L +#define CPC_INT_CNTL__OPCODE_ERROR_INT_ENABLE_MASK 0x01000000L +#define CPC_INT_CNTL__TIME_STAMP_INT_ENABLE_MASK 0x04000000L +#define CPC_INT_CNTL__RESERVED_BIT_ERROR_INT_ENABLE_MASK 0x08000000L +#define CPC_INT_CNTL__GENERIC2_INT_ENABLE_MASK 0x20000000L +#define CPC_INT_CNTL__GENERIC1_INT_ENABLE_MASK 0x40000000L +#define CPC_INT_CNTL__GENERIC0_INT_ENABLE_MASK 0x80000000L +//CPC_INT_STATUS +#define CPC_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS__SHIFT 0xc +#define CPC_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS__SHIFT 0xd +#define CPC_INT_STATUS__CP_ECC_ERROR_INT_STATUS__SHIFT 0xe +#define CPC_INT_STATUS__SUA_VIOLATION_INT_STATUS__SHIFT 0xf +#define CPC_INT_STATUS__GPF_INT_STATUS__SHIFT 0x10 +#define CPC_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS__SHIFT 0x11 +#define CPC_INT_STATUS__PRIV_REG_INT_STATUS__SHIFT 0x17 +#define CPC_INT_STATUS__OPCODE_ERROR_INT_STATUS__SHIFT 0x18 +#define CPC_INT_STATUS__TIME_STAMP_INT_STATUS__SHIFT 0x1a +#define CPC_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS__SHIFT 0x1b +#define CPC_INT_STATUS__GENERIC2_INT_STATUS__SHIFT 0x1d +#define CPC_INT_STATUS__GENERIC1_INT_STATUS__SHIFT 0x1e +#define CPC_INT_STATUS__GENERIC0_INT_STATUS__SHIFT 0x1f +#define CPC_INT_STATUS__CMP_QUERY_STATUS_INT_STATUS_MASK 0x00001000L +#define CPC_INT_STATUS__DEQUEUE_REQUEST_INT_STATUS_MASK 0x00002000L +#define CPC_INT_STATUS__CP_ECC_ERROR_INT_STATUS_MASK 0x00004000L +#define CPC_INT_STATUS__SUA_VIOLATION_INT_STATUS_MASK 0x00008000L +#define CPC_INT_STATUS__GPF_INT_STATUS_MASK 0x00010000L +#define CPC_INT_STATUS__WRM_POLL_TIMEOUT_INT_STATUS_MASK 0x00020000L +#define CPC_INT_STATUS__PRIV_REG_INT_STATUS_MASK 0x00800000L +#define CPC_INT_STATUS__OPCODE_ERROR_INT_STATUS_MASK 0x01000000L +#define CPC_INT_STATUS__TIME_STAMP_INT_STATUS_MASK 0x04000000L +#define CPC_INT_STATUS__RESERVED_BIT_ERROR_INT_STATUS_MASK 0x08000000L +#define CPC_INT_STATUS__GENERIC2_INT_STATUS_MASK 0x20000000L +#define CPC_INT_STATUS__GENERIC1_INT_STATUS_MASK 0x40000000L +#define CPC_INT_STATUS__GENERIC0_INT_STATUS_MASK 0x80000000L +//CP_VMID_PREEMPT +#define CP_VMID_PREEMPT__PREEMPT_REQUEST__SHIFT 0x0 +#define CP_VMID_PREEMPT__VIRT_COMMAND__SHIFT 0x10 +#define CP_VMID_PREEMPT__PREEMPT_REQUEST_MASK 0x0000FFFFL +#define CP_VMID_PREEMPT__VIRT_COMMAND_MASK 0x000F0000L +//CPC_INT_CNTX_ID +#define CPC_INT_CNTX_ID__CNTX_ID__SHIFT 0x0 +#define CPC_INT_CNTX_ID__CNTX_ID_MASK 0xFFFFFFFFL +//CP_PQ_STATUS +#define CP_PQ_STATUS__DOORBELL_UPDATED__SHIFT 0x0 +#define CP_PQ_STATUS__DOORBELL_ENABLE__SHIFT 0x1 +#define CP_PQ_STATUS__DOORBELL_UPDATED_MASK 0x00000001L +#define CP_PQ_STATUS__DOORBELL_ENABLE_MASK 0x00000002L +//CP_CPC_IC_BASE_LO +#define CP_CPC_IC_BASE_LO__IC_BASE_LO__SHIFT 0xc +#define CP_CPC_IC_BASE_LO__IC_BASE_LO_MASK 0xFFFFF000L +//CP_CPC_IC_BASE_HI +#define CP_CPC_IC_BASE_HI__IC_BASE_HI__SHIFT 0x0 +#define CP_CPC_IC_BASE_HI__IC_BASE_HI_MASK 0x0000FFFFL +//CP_CPC_IC_BASE_CNTL +#define CP_CPC_IC_BASE_CNTL__VMID__SHIFT 0x0 +#define CP_CPC_IC_BASE_CNTL__CACHE_POLICY__SHIFT 0x18 +#define CP_CPC_IC_BASE_CNTL__VMID_MASK 0x0000000FL +#define CP_CPC_IC_BASE_CNTL__CACHE_POLICY_MASK 0x01000000L +//CP_CPC_IC_OP_CNTL +#define CP_CPC_IC_OP_CNTL__INVALIDATE_CACHE__SHIFT 0x0 +#define CP_CPC_IC_OP_CNTL__PRIME_ICACHE__SHIFT 0x4 +#define CP_CPC_IC_OP_CNTL__ICACHE_PRIMED__SHIFT 0x5 +#define CP_CPC_IC_OP_CNTL__INVALIDATE_CACHE_MASK 0x00000001L +#define CP_CPC_IC_OP_CNTL__PRIME_ICACHE_MASK 0x00000010L +#define CP_CPC_IC_OP_CNTL__ICACHE_PRIMED_MASK 0x00000020L +//CP_MEC1_F32_INT_DIS +#define CP_MEC1_F32_INT_DIS__EDC_ROQ_FED_INT__SHIFT 0x0 +#define CP_MEC1_F32_INT_DIS__PRIV_REG_INT__SHIFT 0x1 +#define CP_MEC1_F32_INT_DIS__RESERVED_BIT_ERR_INT__SHIFT 0x2 +#define CP_MEC1_F32_INT_DIS__EDC_TC_FED_INT__SHIFT 0x3 +#define CP_MEC1_F32_INT_DIS__EDC_GDS_FED_INT__SHIFT 0x4 +#define CP_MEC1_F32_INT_DIS__EDC_SCRATCH_FED_INT__SHIFT 0x5 +#define CP_MEC1_F32_INT_DIS__WAVE_RESTORE_INT__SHIFT 0x6 +#define CP_MEC1_F32_INT_DIS__SUA_VIOLATION_INT__SHIFT 0x7 +#define CP_MEC1_F32_INT_DIS__EDC_DMA_FED_INT__SHIFT 0x8 +#define CP_MEC1_F32_INT_DIS__IQ_TIMER_INT__SHIFT 0x9 +#define CP_MEC1_F32_INT_DIS__GPF_INT_CPF__SHIFT 0xa +#define CP_MEC1_F32_INT_DIS__GPF_INT_DMA__SHIFT 0xb +#define CP_MEC1_F32_INT_DIS__GPF_INT_CPC__SHIFT 0xc +#define CP_MEC1_F32_INT_DIS__EDC_SR_MEM_FED_INT__SHIFT 0xd +#define CP_MEC1_F32_INT_DIS__QUEUE_MESSAGE_INT__SHIFT 0xe +#define CP_MEC1_F32_INT_DIS__FATAL_EDC_ERROR_INT__SHIFT 0xf +#define CP_MEC1_F32_INT_DIS__EDC_ROQ_FED_INT_MASK 0x00000001L +#define CP_MEC1_F32_INT_DIS__PRIV_REG_INT_MASK 0x00000002L +#define CP_MEC1_F32_INT_DIS__RESERVED_BIT_ERR_INT_MASK 0x00000004L +#define CP_MEC1_F32_INT_DIS__EDC_TC_FED_INT_MASK 0x00000008L +#define CP_MEC1_F32_INT_DIS__EDC_GDS_FED_INT_MASK 0x00000010L +#define CP_MEC1_F32_INT_DIS__EDC_SCRATCH_FED_INT_MASK 0x00000020L +#define CP_MEC1_F32_INT_DIS__WAVE_RESTORE_INT_MASK 0x00000040L +#define CP_MEC1_F32_INT_DIS__SUA_VIOLATION_INT_MASK 0x00000080L +#define CP_MEC1_F32_INT_DIS__EDC_DMA_FED_INT_MASK 0x00000100L +#define CP_MEC1_F32_INT_DIS__IQ_TIMER_INT_MASK 0x00000200L +#define CP_MEC1_F32_INT_DIS__GPF_INT_CPF_MASK 0x00000400L +#define CP_MEC1_F32_INT_DIS__GPF_INT_DMA_MASK 0x00000800L +#define CP_MEC1_F32_INT_DIS__GPF_INT_CPC_MASK 0x00001000L +#define CP_MEC1_F32_INT_DIS__EDC_SR_MEM_FED_INT_MASK 0x00002000L +#define CP_MEC1_F32_INT_DIS__QUEUE_MESSAGE_INT_MASK 0x00004000L +#define CP_MEC1_F32_INT_DIS__FATAL_EDC_ERROR_INT_MASK 0x00008000L +//CP_MEC2_F32_INT_DIS +#define CP_MEC2_F32_INT_DIS__EDC_ROQ_FED_INT__SHIFT 0x0 +#define CP_MEC2_F32_INT_DIS__PRIV_REG_INT__SHIFT 0x1 +#define CP_MEC2_F32_INT_DIS__RESERVED_BIT_ERR_INT__SHIFT 0x2 +#define CP_MEC2_F32_INT_DIS__EDC_TC_FED_INT__SHIFT 0x3 +#define CP_MEC2_F32_INT_DIS__EDC_GDS_FED_INT__SHIFT 0x4 +#define CP_MEC2_F32_INT_DIS__EDC_SCRATCH_FED_INT__SHIFT 0x5 +#define CP_MEC2_F32_INT_DIS__WAVE_RESTORE_INT__SHIFT 0x6 +#define CP_MEC2_F32_INT_DIS__SUA_VIOLATION_INT__SHIFT 0x7 +#define CP_MEC2_F32_INT_DIS__EDC_DMA_FED_INT__SHIFT 0x8 +#define CP_MEC2_F32_INT_DIS__IQ_TIMER_INT__SHIFT 0x9 +#define CP_MEC2_F32_INT_DIS__GPF_INT_CPF__SHIFT 0xa +#define CP_MEC2_F32_INT_DIS__GPF_INT_DMA__SHIFT 0xb +#define CP_MEC2_F32_INT_DIS__GPF_INT_CPC__SHIFT 0xc +#define CP_MEC2_F32_INT_DIS__EDC_SR_MEM_FED_INT__SHIFT 0xd +#define CP_MEC2_F32_INT_DIS__QUEUE_MESSAGE_INT__SHIFT 0xe +#define CP_MEC2_F32_INT_DIS__FATAL_EDC_ERROR_INT__SHIFT 0xf +#define CP_MEC2_F32_INT_DIS__EDC_ROQ_FED_INT_MASK 0x00000001L +#define CP_MEC2_F32_INT_DIS__PRIV_REG_INT_MASK 0x00000002L +#define CP_MEC2_F32_INT_DIS__RESERVED_BIT_ERR_INT_MASK 0x00000004L +#define CP_MEC2_F32_INT_DIS__EDC_TC_FED_INT_MASK 0x00000008L +#define CP_MEC2_F32_INT_DIS__EDC_GDS_FED_INT_MASK 0x00000010L +#define CP_MEC2_F32_INT_DIS__EDC_SCRATCH_FED_INT_MASK 0x00000020L +#define CP_MEC2_F32_INT_DIS__WAVE_RESTORE_INT_MASK 0x00000040L +#define CP_MEC2_F32_INT_DIS__SUA_VIOLATION_INT_MASK 0x00000080L +#define CP_MEC2_F32_INT_DIS__EDC_DMA_FED_INT_MASK 0x00000100L +#define CP_MEC2_F32_INT_DIS__IQ_TIMER_INT_MASK 0x00000200L +#define CP_MEC2_F32_INT_DIS__GPF_INT_CPF_MASK 0x00000400L +#define CP_MEC2_F32_INT_DIS__GPF_INT_DMA_MASK 0x00000800L +#define CP_MEC2_F32_INT_DIS__GPF_INT_CPC_MASK 0x00001000L +#define CP_MEC2_F32_INT_DIS__EDC_SR_MEM_FED_INT_MASK 0x00002000L +#define CP_MEC2_F32_INT_DIS__QUEUE_MESSAGE_INT_MASK 0x00004000L +#define CP_MEC2_F32_INT_DIS__FATAL_EDC_ERROR_INT_MASK 0x00008000L +//CP_VMID_STATUS +#define CP_VMID_STATUS__PREEMPT_DE_STATUS__SHIFT 0x0 +#define CP_VMID_STATUS__PREEMPT_CE_STATUS__SHIFT 0x10 +#define CP_VMID_STATUS__PREEMPT_DE_STATUS_MASK 0x0000FFFFL +#define CP_VMID_STATUS__PREEMPT_CE_STATUS_MASK 0xFFFF0000L + + +// addressBlock: gc_cppdec2 +//CP_RB_DOORBELL_CONTROL_SCH_0 +#define CP_RB_DOORBELL_CONTROL_SCH_0__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL_SCH_0__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL_SCH_0__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL_SCH_0__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL_SCH_0__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL_SCH_0__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_CONTROL_SCH_1 +#define CP_RB_DOORBELL_CONTROL_SCH_1__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL_SCH_1__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL_SCH_1__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL_SCH_1__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL_SCH_1__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL_SCH_1__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_CONTROL_SCH_2 +#define CP_RB_DOORBELL_CONTROL_SCH_2__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL_SCH_2__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL_SCH_2__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL_SCH_2__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL_SCH_2__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL_SCH_2__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_CONTROL_SCH_3 +#define CP_RB_DOORBELL_CONTROL_SCH_3__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL_SCH_3__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL_SCH_3__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL_SCH_3__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL_SCH_3__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL_SCH_3__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_CONTROL_SCH_4 +#define CP_RB_DOORBELL_CONTROL_SCH_4__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL_SCH_4__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL_SCH_4__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL_SCH_4__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL_SCH_4__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL_SCH_4__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_CONTROL_SCH_5 +#define CP_RB_DOORBELL_CONTROL_SCH_5__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL_SCH_5__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL_SCH_5__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL_SCH_5__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL_SCH_5__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL_SCH_5__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_CONTROL_SCH_6 +#define CP_RB_DOORBELL_CONTROL_SCH_6__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL_SCH_6__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL_SCH_6__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL_SCH_6__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL_SCH_6__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL_SCH_6__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_CONTROL_SCH_7 +#define CP_RB_DOORBELL_CONTROL_SCH_7__DOORBELL_OFFSET__SHIFT 0x2 +#define CP_RB_DOORBELL_CONTROL_SCH_7__DOORBELL_EN__SHIFT 0x1e +#define CP_RB_DOORBELL_CONTROL_SCH_7__DOORBELL_HIT__SHIFT 0x1f +#define CP_RB_DOORBELL_CONTROL_SCH_7__DOORBELL_OFFSET_MASK 0x0FFFFFFCL +#define CP_RB_DOORBELL_CONTROL_SCH_7__DOORBELL_EN_MASK 0x40000000L +#define CP_RB_DOORBELL_CONTROL_SCH_7__DOORBELL_HIT_MASK 0x80000000L +//CP_RB_DOORBELL_CLEAR +#define CP_RB_DOORBELL_CLEAR__MAPPED_QUEUE__SHIFT 0x0 +#define CP_RB_DOORBELL_CLEAR__MAPPED_QUE_DOORBELL_EN_CLEAR__SHIFT 0x8 +#define CP_RB_DOORBELL_CLEAR__MAPPED_QUE_DOORBELL_HIT_CLEAR__SHIFT 0x9 +#define CP_RB_DOORBELL_CLEAR__MASTER_DOORBELL_EN_CLEAR__SHIFT 0xa +#define CP_RB_DOORBELL_CLEAR__MASTER_DOORBELL_HIT_CLEAR__SHIFT 0xb +#define CP_RB_DOORBELL_CLEAR__QUEUES_DOORBELL_EN_CLEAR__SHIFT 0xc +#define CP_RB_DOORBELL_CLEAR__QUEUES_DOORBELL_HIT_CLEAR__SHIFT 0xd +#define CP_RB_DOORBELL_CLEAR__MAPPED_QUEUE_MASK 0x00000007L +#define CP_RB_DOORBELL_CLEAR__MAPPED_QUE_DOORBELL_EN_CLEAR_MASK 0x00000100L +#define CP_RB_DOORBELL_CLEAR__MAPPED_QUE_DOORBELL_HIT_CLEAR_MASK 0x00000200L +#define CP_RB_DOORBELL_CLEAR__MASTER_DOORBELL_EN_CLEAR_MASK 0x00000400L +#define CP_RB_DOORBELL_CLEAR__MASTER_DOORBELL_HIT_CLEAR_MASK 0x00000800L +#define CP_RB_DOORBELL_CLEAR__QUEUES_DOORBELL_EN_CLEAR_MASK 0x00001000L +#define CP_RB_DOORBELL_CLEAR__QUEUES_DOORBELL_HIT_CLEAR_MASK 0x00002000L +//CP_GFX_MQD_CONTROL +#define CP_GFX_MQD_CONTROL__VMID__SHIFT 0x0 +#define CP_GFX_MQD_CONTROL__EXE_DISABLE__SHIFT 0x17 +#define CP_GFX_MQD_CONTROL__CACHE_POLICY__SHIFT 0x18 +#define CP_GFX_MQD_CONTROL__VMID_MASK 0x0000000FL +#define CP_GFX_MQD_CONTROL__EXE_DISABLE_MASK 0x00800000L +#define CP_GFX_MQD_CONTROL__CACHE_POLICY_MASK 0x01000000L +//CP_GFX_MQD_BASE_ADDR +#define CP_GFX_MQD_BASE_ADDR__BASE_ADDR__SHIFT 0x2 +#define CP_GFX_MQD_BASE_ADDR__BASE_ADDR_MASK 0xFFFFFFFCL +//CP_GFX_MQD_BASE_ADDR_HI +#define CP_GFX_MQD_BASE_ADDR_HI__BASE_ADDR_HI__SHIFT 0x0 +#define CP_GFX_MQD_BASE_ADDR_HI__BASE_ADDR_HI_MASK 0x0000FFFFL +//CP_RB_STATUS +#define CP_RB_STATUS__DOORBELL_UPDATED__SHIFT 0x0 +#define CP_RB_STATUS__DOORBELL_ENABLE__SHIFT 0x1 +#define CP_RB_STATUS__DOORBELL_UPDATED_MASK 0x00000001L +#define CP_RB_STATUS__DOORBELL_ENABLE_MASK 0x00000002L +//CPG_UTCL1_STATUS +#define CPG_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define CPG_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define CPG_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define CPG_UTCL1_STATUS__FAULT_UTCL1ID__SHIFT 0x8 +#define CPG_UTCL1_STATUS__RETRY_UTCL1ID__SHIFT 0x10 +#define CPG_UTCL1_STATUS__PRT_UTCL1ID__SHIFT 0x18 +#define CPG_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define CPG_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define CPG_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L +#define CPG_UTCL1_STATUS__FAULT_UTCL1ID_MASK 0x00003F00L +#define CPG_UTCL1_STATUS__RETRY_UTCL1ID_MASK 0x003F0000L +#define CPG_UTCL1_STATUS__PRT_UTCL1ID_MASK 0x3F000000L +//CPC_UTCL1_STATUS +#define CPC_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define CPC_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define CPC_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define CPC_UTCL1_STATUS__FAULT_UTCL1ID__SHIFT 0x8 +#define CPC_UTCL1_STATUS__RETRY_UTCL1ID__SHIFT 0x10 +#define CPC_UTCL1_STATUS__PRT_UTCL1ID__SHIFT 0x18 +#define CPC_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define CPC_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define CPC_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L +#define CPC_UTCL1_STATUS__FAULT_UTCL1ID_MASK 0x00003F00L +#define CPC_UTCL1_STATUS__RETRY_UTCL1ID_MASK 0x003F0000L +#define CPC_UTCL1_STATUS__PRT_UTCL1ID_MASK 0x3F000000L +//CPF_UTCL1_STATUS +#define CPF_UTCL1_STATUS__FAULT_DETECTED__SHIFT 0x0 +#define CPF_UTCL1_STATUS__RETRY_DETECTED__SHIFT 0x1 +#define CPF_UTCL1_STATUS__PRT_DETECTED__SHIFT 0x2 +#define CPF_UTCL1_STATUS__FAULT_UTCL1ID__SHIFT 0x8 +#define CPF_UTCL1_STATUS__RETRY_UTCL1ID__SHIFT 0x10 +#define CPF_UTCL1_STATUS__PRT_UTCL1ID__SHIFT 0x18 +#define CPF_UTCL1_STATUS__FAULT_DETECTED_MASK 0x00000001L +#define CPF_UTCL1_STATUS__RETRY_DETECTED_MASK 0x00000002L +#define CPF_UTCL1_STATUS__PRT_DETECTED_MASK 0x00000004L +#define CPF_UTCL1_STATUS__FAULT_UTCL1ID_MASK 0x00003F00L +#define CPF_UTCL1_STATUS__RETRY_UTCL1ID_MASK 0x003F0000L +#define CPF_UTCL1_STATUS__PRT_UTCL1ID_MASK 0x3F000000L +//CP_SD_CNTL +#define CP_SD_CNTL__CPF_EN__SHIFT 0x0 +#define CP_SD_CNTL__CPG_EN__SHIFT 0x1 +#define CP_SD_CNTL__CPC_EN__SHIFT 0x2 +#define CP_SD_CNTL__RLC_EN__SHIFT 0x3 +#define CP_SD_CNTL__SPI_EN__SHIFT 0x4 +#define CP_SD_CNTL__WD_EN__SHIFT 0x5 +#define CP_SD_CNTL__IA_EN__SHIFT 0x6 +#define CP_SD_CNTL__PA_EN__SHIFT 0x7 +#define CP_SD_CNTL__RMI_EN__SHIFT 0x8 +#define CP_SD_CNTL__EA_EN__SHIFT 0x9 +#define CP_SD_CNTL__CPF_EN_MASK 0x00000001L +#define CP_SD_CNTL__CPG_EN_MASK 0x00000002L +#define CP_SD_CNTL__CPC_EN_MASK 0x00000004L +#define CP_SD_CNTL__RLC_EN_MASK 0x00000008L +#define CP_SD_CNTL__SPI_EN_MASK 0x00000010L +#define CP_SD_CNTL__WD_EN_MASK 0x00000020L +#define CP_SD_CNTL__IA_EN_MASK 0x00000040L +#define CP_SD_CNTL__PA_EN_MASK 0x00000080L +#define CP_SD_CNTL__RMI_EN_MASK 0x00000100L +#define CP_SD_CNTL__EA_EN_MASK 0x00000200L +//CP_SOFT_RESET_CNTL +#define CP_SOFT_RESET_CNTL__CMP_ONLY_SOFT_RESET__SHIFT 0x0 +#define CP_SOFT_RESET_CNTL__GFX_ONLY_SOFT_RESET__SHIFT 0x1 +#define CP_SOFT_RESET_CNTL__CMP_HQD_REG_RESET__SHIFT 0x2 +#define CP_SOFT_RESET_CNTL__CMP_INTR_REG_RESET__SHIFT 0x3 +#define CP_SOFT_RESET_CNTL__CMP_HQD_QUEUE_DOORBELL_RESET__SHIFT 0x4 +#define CP_SOFT_RESET_CNTL__GFX_RB_DOORBELL_RESET__SHIFT 0x5 +#define CP_SOFT_RESET_CNTL__GFX_INTR_REG_RESET__SHIFT 0x6 +#define CP_SOFT_RESET_CNTL__CMP_ONLY_SOFT_RESET_MASK 0x00000001L +#define CP_SOFT_RESET_CNTL__GFX_ONLY_SOFT_RESET_MASK 0x00000002L +#define CP_SOFT_RESET_CNTL__CMP_HQD_REG_RESET_MASK 0x00000004L +#define CP_SOFT_RESET_CNTL__CMP_INTR_REG_RESET_MASK 0x00000008L +#define CP_SOFT_RESET_CNTL__CMP_HQD_QUEUE_DOORBELL_RESET_MASK 0x00000010L +#define CP_SOFT_RESET_CNTL__GFX_RB_DOORBELL_RESET_MASK 0x00000020L +#define CP_SOFT_RESET_CNTL__GFX_INTR_REG_RESET_MASK 0x00000040L +//CP_CPC_GFX_CNTL +#define CP_CPC_GFX_CNTL__QUEUEID__SHIFT 0x0 +#define CP_CPC_GFX_CNTL__PIPEID__SHIFT 0x3 +#define CP_CPC_GFX_CNTL__MEID__SHIFT 0x5 +#define CP_CPC_GFX_CNTL__VALID__SHIFT 0x7 +#define CP_CPC_GFX_CNTL__QUEUEID_MASK 0x00000007L +#define CP_CPC_GFX_CNTL__PIPEID_MASK 0x00000018L +#define CP_CPC_GFX_CNTL__MEID_MASK 0x00000060L +#define CP_CPC_GFX_CNTL__VALID_MASK 0x00000080L + + +// addressBlock: gc_spipdec +//SPI_ARB_PRIORITY +#define SPI_ARB_PRIORITY__PIPE_ORDER_TS0__SHIFT 0x0 +#define SPI_ARB_PRIORITY__PIPE_ORDER_TS1__SHIFT 0x3 +#define SPI_ARB_PRIORITY__PIPE_ORDER_TS2__SHIFT 0x6 +#define SPI_ARB_PRIORITY__PIPE_ORDER_TS3__SHIFT 0x9 +#define SPI_ARB_PRIORITY__TS0_DUR_MULT__SHIFT 0xc +#define SPI_ARB_PRIORITY__TS1_DUR_MULT__SHIFT 0xe +#define SPI_ARB_PRIORITY__TS2_DUR_MULT__SHIFT 0x10 +#define SPI_ARB_PRIORITY__TS3_DUR_MULT__SHIFT 0x12 +#define SPI_ARB_PRIORITY__PIPE_ORDER_TS0_MASK 0x00000007L +#define SPI_ARB_PRIORITY__PIPE_ORDER_TS1_MASK 0x00000038L +#define SPI_ARB_PRIORITY__PIPE_ORDER_TS2_MASK 0x000001C0L +#define SPI_ARB_PRIORITY__PIPE_ORDER_TS3_MASK 0x00000E00L +#define SPI_ARB_PRIORITY__TS0_DUR_MULT_MASK 0x00003000L +#define SPI_ARB_PRIORITY__TS1_DUR_MULT_MASK 0x0000C000L +#define SPI_ARB_PRIORITY__TS2_DUR_MULT_MASK 0x00030000L +#define SPI_ARB_PRIORITY__TS3_DUR_MULT_MASK 0x000C0000L +//SPI_ARB_CYCLES_0 +#define SPI_ARB_CYCLES_0__TS0_DURATION__SHIFT 0x0 +#define SPI_ARB_CYCLES_0__TS1_DURATION__SHIFT 0x10 +#define SPI_ARB_CYCLES_0__TS0_DURATION_MASK 0x0000FFFFL +#define SPI_ARB_CYCLES_0__TS1_DURATION_MASK 0xFFFF0000L +//SPI_ARB_CYCLES_1 +#define SPI_ARB_CYCLES_1__TS2_DURATION__SHIFT 0x0 +#define SPI_ARB_CYCLES_1__TS3_DURATION__SHIFT 0x10 +#define SPI_ARB_CYCLES_1__TS2_DURATION_MASK 0x0000FFFFL +#define SPI_ARB_CYCLES_1__TS3_DURATION_MASK 0xFFFF0000L +//SPI_WCL_PIPE_PERCENT_GFX +#define SPI_WCL_PIPE_PERCENT_GFX__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_GFX__LS_GRP_VALUE__SHIFT 0x7 +#define SPI_WCL_PIPE_PERCENT_GFX__HS_GRP_VALUE__SHIFT 0xc +#define SPI_WCL_PIPE_PERCENT_GFX__ES_GRP_VALUE__SHIFT 0x11 +#define SPI_WCL_PIPE_PERCENT_GFX__GS_GRP_VALUE__SHIFT 0x16 +#define SPI_WCL_PIPE_PERCENT_GFX__VALUE_MASK 0x0000007FL +#define SPI_WCL_PIPE_PERCENT_GFX__LS_GRP_VALUE_MASK 0x00000F80L +#define SPI_WCL_PIPE_PERCENT_GFX__HS_GRP_VALUE_MASK 0x0001F000L +#define SPI_WCL_PIPE_PERCENT_GFX__ES_GRP_VALUE_MASK 0x003E0000L +#define SPI_WCL_PIPE_PERCENT_GFX__GS_GRP_VALUE_MASK 0x07C00000L +//SPI_WCL_PIPE_PERCENT_HP3D +#define SPI_WCL_PIPE_PERCENT_HP3D__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_HP3D__HS_GRP_VALUE__SHIFT 0xc +#define SPI_WCL_PIPE_PERCENT_HP3D__GS_GRP_VALUE__SHIFT 0x16 +#define SPI_WCL_PIPE_PERCENT_HP3D__VALUE_MASK 0x0000007FL +#define SPI_WCL_PIPE_PERCENT_HP3D__HS_GRP_VALUE_MASK 0x0001F000L +#define SPI_WCL_PIPE_PERCENT_HP3D__GS_GRP_VALUE_MASK 0x07C00000L +//SPI_WCL_PIPE_PERCENT_CS0 +#define SPI_WCL_PIPE_PERCENT_CS0__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_CS0__VALUE_MASK 0x7FL +//SPI_WCL_PIPE_PERCENT_CS1 +#define SPI_WCL_PIPE_PERCENT_CS1__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_CS1__VALUE_MASK 0x7FL +//SPI_WCL_PIPE_PERCENT_CS2 +#define SPI_WCL_PIPE_PERCENT_CS2__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_CS2__VALUE_MASK 0x7FL +//SPI_WCL_PIPE_PERCENT_CS3 +#define SPI_WCL_PIPE_PERCENT_CS3__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_CS3__VALUE_MASK 0x7FL +//SPI_WCL_PIPE_PERCENT_CS4 +#define SPI_WCL_PIPE_PERCENT_CS4__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_CS4__VALUE_MASK 0x7FL +//SPI_WCL_PIPE_PERCENT_CS5 +#define SPI_WCL_PIPE_PERCENT_CS5__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_CS5__VALUE_MASK 0x7FL +//SPI_WCL_PIPE_PERCENT_CS6 +#define SPI_WCL_PIPE_PERCENT_CS6__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_CS6__VALUE_MASK 0x7FL +//SPI_WCL_PIPE_PERCENT_CS7 +#define SPI_WCL_PIPE_PERCENT_CS7__VALUE__SHIFT 0x0 +#define SPI_WCL_PIPE_PERCENT_CS7__VALUE_MASK 0x7FL +//SPI_COMPUTE_QUEUE_RESET +#define SPI_COMPUTE_QUEUE_RESET__RESET__SHIFT 0x0 +#define SPI_COMPUTE_QUEUE_RESET__RESET_MASK 0x01L +//SPI_RESOURCE_RESERVE_CU_0 +#define SPI_RESOURCE_RESERVE_CU_0__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_0__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_0__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_0__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_0__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_0__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_0__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_0__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_0__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_0__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_1 +#define SPI_RESOURCE_RESERVE_CU_1__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_1__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_1__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_1__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_1__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_1__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_1__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_1__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_1__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_1__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_2 +#define SPI_RESOURCE_RESERVE_CU_2__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_2__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_2__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_2__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_2__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_2__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_2__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_2__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_2__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_2__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_3 +#define SPI_RESOURCE_RESERVE_CU_3__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_3__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_3__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_3__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_3__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_3__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_3__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_3__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_3__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_3__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_4 +#define SPI_RESOURCE_RESERVE_CU_4__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_4__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_4__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_4__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_4__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_4__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_4__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_4__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_4__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_4__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_5 +#define SPI_RESOURCE_RESERVE_CU_5__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_5__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_5__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_5__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_5__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_5__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_5__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_5__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_5__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_5__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_6 +#define SPI_RESOURCE_RESERVE_CU_6__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_6__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_6__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_6__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_6__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_6__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_6__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_6__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_6__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_6__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_7 +#define SPI_RESOURCE_RESERVE_CU_7__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_7__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_7__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_7__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_7__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_7__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_7__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_7__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_7__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_7__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_8 +#define SPI_RESOURCE_RESERVE_CU_8__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_8__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_8__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_8__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_8__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_8__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_8__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_8__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_8__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_8__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_9 +#define SPI_RESOURCE_RESERVE_CU_9__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_9__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_9__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_9__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_9__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_9__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_9__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_9__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_9__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_9__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_EN_CU_0 +#define SPI_RESOURCE_RESERVE_EN_CU_0__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_0__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_0__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_0__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_0__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_0__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_0__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_0__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_1 +#define SPI_RESOURCE_RESERVE_EN_CU_1__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_1__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_1__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_1__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_1__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_1__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_1__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_1__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_2 +#define SPI_RESOURCE_RESERVE_EN_CU_2__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_2__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_2__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_2__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_2__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_2__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_2__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_2__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_3 +#define SPI_RESOURCE_RESERVE_EN_CU_3__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_3__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_3__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_3__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_3__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_3__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_3__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_3__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_4 +#define SPI_RESOURCE_RESERVE_EN_CU_4__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_4__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_4__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_4__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_4__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_4__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_4__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_4__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_5 +#define SPI_RESOURCE_RESERVE_EN_CU_5__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_5__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_5__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_5__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_5__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_5__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_5__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_5__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_6 +#define SPI_RESOURCE_RESERVE_EN_CU_6__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_6__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_6__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_6__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_6__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_6__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_6__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_6__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_7 +#define SPI_RESOURCE_RESERVE_EN_CU_7__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_7__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_7__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_7__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_7__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_7__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_7__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_7__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_8 +#define SPI_RESOURCE_RESERVE_EN_CU_8__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_8__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_8__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_8__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_8__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_8__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_8__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_8__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_EN_CU_9 +#define SPI_RESOURCE_RESERVE_EN_CU_9__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_9__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_9__QUEUE_MASK__SHIFT 0x10 +#define SPI_RESOURCE_RESERVE_EN_CU_9__RESERVE_SPACE_ONLY__SHIFT 0x18 +#define SPI_RESOURCE_RESERVE_EN_CU_9__EN_MASK 0x00000001L +#define SPI_RESOURCE_RESERVE_EN_CU_9__TYPE_MASK_MASK 0x0000FFFEL +#define SPI_RESOURCE_RESERVE_EN_CU_9__QUEUE_MASK_MASK 0x00FF0000L +#define SPI_RESOURCE_RESERVE_EN_CU_9__RESERVE_SPACE_ONLY_MASK 0x01000000L +//SPI_RESOURCE_RESERVE_CU_10 +#define SPI_RESOURCE_RESERVE_CU_10__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_10__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_10__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_10__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_10__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_10__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_10__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_10__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_10__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_10__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_CU_11 +#define SPI_RESOURCE_RESERVE_CU_11__VGPR__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_CU_11__SGPR__SHIFT 0x4 +#define SPI_RESOURCE_RESERVE_CU_11__LDS__SHIFT 0x8 +#define SPI_RESOURCE_RESERVE_CU_11__WAVES__SHIFT 0xc +#define SPI_RESOURCE_RESERVE_CU_11__BARRIERS__SHIFT 0xf +#define SPI_RESOURCE_RESERVE_CU_11__VGPR_MASK 0x0000000FL +#define SPI_RESOURCE_RESERVE_CU_11__SGPR_MASK 0x000000F0L +#define SPI_RESOURCE_RESERVE_CU_11__LDS_MASK 0x00000F00L +#define SPI_RESOURCE_RESERVE_CU_11__WAVES_MASK 0x00007000L +#define SPI_RESOURCE_RESERVE_CU_11__BARRIERS_MASK 0x00078000L +//SPI_RESOURCE_RESERVE_EN_CU_10 +#define SPI_RESOURCE_RESERVE_EN_CU_10__EN__SHIFT 0x0 +#define SPI_RESOURCE_RESERVE_EN_CU_10__TYPE_MASK__SHIFT 0x1 +#define SPI_RESOURCE_RESERVE_EN_CU_10__QUEUE_MASK__SHIFT 0x10