blob: 6725c3279e2d05291611565f37093ac5848aa9bd [file] [log] [blame]
// Copyright 2021 the V8 project authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef V8_CODEGEN_LOONG64_ASSEMBLER_LOONG64_H_
#define V8_CODEGEN_LOONG64_ASSEMBLER_LOONG64_H_
#include <stdio.h>
#include <memory>
#include <set>
#include "src/codegen/assembler.h"
#include "src/codegen/external-reference.h"
#include "src/codegen/label.h"
#include "src/codegen/loong64/constants-loong64.h"
#include "src/codegen/loong64/register-loong64.h"
#include "src/codegen/machine-type.h"
#include "src/objects/contexts.h"
#include "src/objects/smi.h"
namespace v8 {
namespace internal {
class SafepointTableBuilder;
// -----------------------------------------------------------------------------
// Machine instruction Operands.
constexpr int kSmiShift = kSmiTagSize + kSmiShiftSize;
constexpr uint64_t kSmiShiftMask = (1UL << kSmiShift) - 1;
// Class Operand represents a shifter operand in data processing instructions.
class Operand {
public:
// Immediate.
V8_INLINE explicit Operand(int64_t immediate,
RelocInfo::Mode rmode = RelocInfo::NO_INFO)
: rm_(no_reg), rmode_(rmode) {
value_.immediate = immediate;
}
V8_INLINE explicit Operand(const ExternalReference& f)
: rm_(no_reg), rmode_(RelocInfo::EXTERNAL_REFERENCE) {
value_.immediate = static_cast<int64_t>(f.address());
}
V8_INLINE explicit Operand(const char* s);
explicit Operand(Handle<HeapObject> handle);
V8_INLINE explicit Operand(Smi value)
: rm_(no_reg), rmode_(RelocInfo::NO_INFO) {
value_.immediate = static_cast<intptr_t>(value.ptr());
}
static Operand EmbeddedNumber(double number); // Smi or HeapNumber.
static Operand EmbeddedStringConstant(const StringConstantBase* str);
// Register.
V8_INLINE explicit Operand(Register rm) : rm_(rm) {}
// Return true if this is a register operand.
V8_INLINE bool is_reg() const;
inline int64_t immediate() const;
bool IsImmediate() const { return !rm_.is_valid(); }
HeapObjectRequest heap_object_request() const {
DCHECK(IsHeapObjectRequest());
return value_.heap_object_request;
}
bool IsHeapObjectRequest() const {
DCHECK_IMPLIES(is_heap_object_request_, IsImmediate());
DCHECK_IMPLIES(is_heap_object_request_,
rmode_ == RelocInfo::FULL_EMBEDDED_OBJECT ||
rmode_ == RelocInfo::CODE_TARGET);
return is_heap_object_request_;
}
Register rm() const { return rm_; }
RelocInfo::Mode rmode() const { return rmode_; }
private:
Register rm_;
union Value {
Value() {}
HeapObjectRequest heap_object_request; // if is_heap_object_request_
int64_t immediate; // otherwise
} value_; // valid if rm_ == no_reg
bool is_heap_object_request_ = false;
RelocInfo::Mode rmode_;
friend class Assembler;
friend class MacroAssembler;
};
// Class MemOperand represents a memory operand in load and store instructions.
// 1: base_reg + off_imm( si12 | si14<<2)
// 2: base_reg + offset_reg
class V8_EXPORT_PRIVATE MemOperand {
public:
explicit MemOperand(Register rj, int32_t offset = 0);
explicit MemOperand(Register rj, Register offset = no_reg);
Register base() const { return base_; }
Register index() const { return index_; }
int32_t offset() const { return offset_; }
bool hasIndexReg() const { return index_ != no_reg; }
private:
Register base_; // base
Register index_; // index
int32_t offset_; // offset
friend class Assembler;
};
class V8_EXPORT_PRIVATE Assembler : public AssemblerBase {
public:
// Create an assembler. Instructions and relocation information are emitted
// into a buffer, with the instructions starting from the beginning and the
// relocation information starting from the end of the buffer. See CodeDesc
// for a detailed comment on the layout (globals.h).
//
// If the provided buffer is nullptr, the assembler allocates and grows its
// own buffer. Otherwise it takes ownership of the provided buffer.
explicit Assembler(const AssemblerOptions&,
std::unique_ptr<AssemblerBuffer> = {});
virtual ~Assembler() {}
// GetCode emits any pending (non-emitted) code and fills the descriptor desc.
static constexpr int kNoHandlerTable = 0;
static constexpr SafepointTableBuilder* kNoSafepointTable = nullptr;
void GetCode(Isolate* isolate, CodeDesc* desc,
SafepointTableBuilder* safepoint_table_builder,
int handler_table_offset);
// Convenience wrapper for code without safepoint or handler tables.
void GetCode(Isolate* isolate, CodeDesc* desc) {
GetCode(isolate, desc, kNoSafepointTable, kNoHandlerTable);
}
// Unused on this architecture.
void MaybeEmitOutOfLineConstantPool() {}
// Loong64 uses BlockTrampolinePool to prevent generating trampoline inside a
// continuous instruction block. In the destructor of BlockTrampolinePool, it
// must check if it needs to generate trampoline immediately, if it does not
// do this, the branch range will go beyond the max branch offset, that means
// the pc_offset after call CheckTrampolinePool may have changed. So we use
// pc_for_safepoint_ here for safepoint record.
int pc_offset_for_safepoint() {
return static_cast<int>(pc_for_safepoint_ - buffer_start_);
}
// TODO(LOONG_dev): LOONG64 Check this comment
// Label operations & relative jumps (PPUM Appendix D).
//
// Takes a branch opcode (cc) and a label (L) and generates
// either a backward branch or a forward branch and links it
// to the label fixup chain. Usage:
//
// Label L; // unbound label
// j(cc, &L); // forward branch to unbound label
// bind(&L); // bind label to the current pc
// j(cc, &L); // backward branch to bound label
// bind(&L); // illegal: a label may be bound only once
//
// Note: The same Label can be used for forward and backward branches
// but it may be bound only once.
void bind(Label* L); // Binds an unbound label L to current code position.
enum OffsetSize : int { kOffset26 = 26, kOffset21 = 21, kOffset16 = 16 };
// Determines if Label is bound and near enough so that branch instruction
// can be used to reach it, instead of jump instruction.
// c means conditinal branch, a means always branch.
bool is_near_c(Label* L);
bool is_near(Label* L, OffsetSize bits);
bool is_near_a(Label* L);
int BranchOffset(Instr instr);
// Returns the branch offset to the given label from the current code
// position. Links the label to the current position if it is still unbound.
// Manages the jump elimination optimization if the second parameter is true.
int32_t branch_offset_helper(Label* L, OffsetSize bits);
inline int32_t branch_offset(Label* L) {
return branch_offset_helper(L, OffsetSize::kOffset16);
}
inline int32_t branch_offset21(Label* L) {
return branch_offset_helper(L, OffsetSize::kOffset21);
}
inline int32_t branch_offset26(Label* L) {
return branch_offset_helper(L, OffsetSize::kOffset26);
}
inline int32_t shifted_branch_offset(Label* L) {
return branch_offset(L) >> 2;
}
inline int32_t shifted_branch_offset21(Label* L) {
return branch_offset21(L) >> 2;
}
inline int32_t shifted_branch_offset26(Label* L) {
return branch_offset26(L) >> 2;
}
uint64_t jump_address(Label* L);
uint64_t jump_offset(Label* L);
uint64_t branch_long_offset(Label* L);
// Puts a labels target address at the given position.
// The high 8 bits are set to zero.
void label_at_put(Label* L, int at_offset);
// Read/Modify the code target address in the branch/call instruction at pc.
// The isolate argument is unused (and may be nullptr) when skipping flushing.
static Address target_address_at(Address pc);
V8_INLINE static void set_target_address_at(
Address pc, Address target,
ICacheFlushMode icache_flush_mode = FLUSH_ICACHE_IF_NEEDED) {
set_target_value_at(pc, target, icache_flush_mode);
}
// On LOONG64 there is no Constant Pool so we skip that parameter.
V8_INLINE static Address target_address_at(Address pc,
Address constant_pool) {
return target_address_at(pc);
}
V8_INLINE static void set_target_address_at(
Address pc, Address constant_pool, Address target,
ICacheFlushMode icache_flush_mode = FLUSH_ICACHE_IF_NEEDED) {
set_target_address_at(pc, target, icache_flush_mode);
}
static void set_target_value_at(
Address pc, uint64_t target,
ICacheFlushMode icache_flush_mode = FLUSH_ICACHE_IF_NEEDED);
static void JumpLabelToJumpRegister(Address pc);
// This sets the branch destination (which gets loaded at the call address).
// This is for calls and branches within generated code. The serializer
// has already deserialized the lui/ori instructions etc.
inline static void deserialization_set_special_target_at(
Address instruction_payload, Code code, Address target);
// Get the size of the special target encoded at 'instruction_payload'.
inline static int deserialization_special_target_size(
Address instruction_payload);
// This sets the internal reference at the pc.
inline static void deserialization_set_target_internal_reference_at(
Address pc, Address target,
RelocInfo::Mode mode = RelocInfo::INTERNAL_REFERENCE);
// Here we are patching the address in the LUI/ORI instruction pair.
// These values are used in the serialization process and must be zero for
// LOONG platform, as Code, Embedded Object or External-reference pointers
// are split across two consecutive instructions and don't exist separately
// in the code, so the serializer should not step forwards in memory after
// a target is resolved and written.
static constexpr int kSpecialTargetSize = 0;
// Number of consecutive instructions used to store 32bit/64bit constant.
// This constant was used in RelocInfo::target_address_address() function
// to tell serializer address of the instruction that follows
// LUI/ORI instruction pair.
// TODO(LOONG_dev): check this
static constexpr int kInstructionsFor64BitConstant = 4;
// Max offset for instructions with 16-bit offset field
static constexpr int kMax16BranchOffset = (1 << (18 - 1)) - 1;
// Max offset for instructions with 21-bit offset field
static constexpr int kMax21BranchOffset = (1 << (23 - 1)) - 1;
// Max offset for compact branch instructions with 26-bit offset field
static constexpr int kMax26BranchOffset = (1 << (28 - 1)) - 1;
static constexpr int kTrampolineSlotsSize = 2 * kInstrSize;
RegList* GetScratchRegisterList() { return &scratch_register_list_; }
// ---------------------------------------------------------------------------
// Code generation.
// Insert the smallest number of nop instructions
// possible to align the pc offset to a multiple
// of m. m must be a power of 2 (>= 4).
void Align(int m);
// Insert the smallest number of zero bytes possible to align the pc offset
// to a mulitple of m. m must be a power of 2 (>= 2).
void DataAlign(int m);
// Aligns code to something that's optimal for a jump target for the platform.
void CodeTargetAlign();
void LoopHeaderAlign() { CodeTargetAlign(); }
// Different nop operations are used by the code generator to detect certain
// states of the generated code.
enum NopMarkerTypes {
NON_MARKING_NOP = 0,
DEBUG_BREAK_NOP,
// IC markers.
PROPERTY_ACCESS_INLINED,
PROPERTY_ACCESS_INLINED_CONTEXT,
PROPERTY_ACCESS_INLINED_CONTEXT_DONT_DELETE,
// Helper values.
LAST_CODE_MARKER,
FIRST_IC_MARKER = PROPERTY_ACCESS_INLINED,
};
// Type == 0 is the default non-marking nop. For LoongArch this is a
// andi(zero_reg, zero_reg, 0).
void nop(unsigned int type = 0) {
DCHECK_LT(type, 32);
andi(zero_reg, zero_reg, type);
}
// --------Branch-and-jump-instructions----------
// We don't use likely variant of instructions.
void b(int32_t offset);
inline void b(Label* L) { b(shifted_branch_offset26(L)); }
void bl(int32_t offset);
inline void bl(Label* L) { bl(shifted_branch_offset26(L)); }
void beq(Register rj, Register rd, int32_t offset);
inline void beq(Register rj, Register rd, Label* L) {
beq(rj, rd, shifted_branch_offset(L));
}
void bne(Register rj, Register rd, int32_t offset);
inline void bne(Register rj, Register rd, Label* L) {
bne(rj, rd, shifted_branch_offset(L));
}
void blt(Register rj, Register rd, int32_t offset);
inline void blt(Register rj, Register rd, Label* L) {
blt(rj, rd, shifted_branch_offset(L));
}
void bge(Register rj, Register rd, int32_t offset);
inline void bge(Register rj, Register rd, Label* L) {
bge(rj, rd, shifted_branch_offset(L));
}
void bltu(Register rj, Register rd, int32_t offset);
inline void bltu(Register rj, Register rd, Label* L) {
bltu(rj, rd, shifted_branch_offset(L));
}
void bgeu(Register rj, Register rd, int32_t offset);
inline void bgeu(Register rj, Register rd, Label* L) {
bgeu(rj, rd, shifted_branch_offset(L));
}
void beqz(Register rj, int32_t offset);
inline void beqz(Register rj, Label* L) {
beqz(rj, shifted_branch_offset21(L));
}
void bnez(Register rj, int32_t offset);
inline void bnez(Register rj, Label* L) {
bnez(rj, shifted_branch_offset21(L));
}
void jirl(Register rd, Register rj, int32_t offset);
void bceqz(CFRegister cj, int32_t si21);
inline void bceqz(CFRegister cj, Label* L) {
bceqz(cj, shifted_branch_offset21(L));
}
void bcnez(CFRegister cj, int32_t si21);
inline void bcnez(CFRegister cj, Label* L) {
bcnez(cj, shifted_branch_offset21(L));
}
// -------Data-processing-instructions---------
// Arithmetic.
void add_w(Register rd, Register rj, Register rk);
void add_d(Register rd, Register rj, Register rk);
void sub_w(Register rd, Register rj, Register rk);
void sub_d(Register rd, Register rj, Register rk);
void addi_w(Register rd, Register rj, int32_t si12);
void addi_d(Register rd, Register rj, int32_t si12);
void addu16i_d(Register rd, Register rj, int32_t si16);
void alsl_w(Register rd, Register rj, Register rk, int32_t sa2);
void alsl_wu(Register rd, Register rj, Register rk, int32_t sa2);
void alsl_d(Register rd, Register rj, Register rk, int32_t sa2);
void lu12i_w(Register rd, int32_t si20);
void lu32i_d(Register rd, int32_t si20);
void lu52i_d(Register rd, Register rj, int32_t si12);
void slt(Register rd, Register rj, Register rk);
void sltu(Register rd, Register rj, Register rk);
void slti(Register rd, Register rj, int32_t si12);
void sltui(Register rd, Register rj, int32_t si12);
void pcaddi(Register rd, int32_t si20);
void pcaddu12i(Register rd, int32_t si20);
void pcaddu18i(Register rd, int32_t si20);
void pcalau12i(Register rd, int32_t si20);
void and_(Register rd, Register rj, Register rk);
void or_(Register rd, Register rj, Register rk);
void xor_(Register rd, Register rj, Register rk);
void nor(Register rd, Register rj, Register rk);
void andn(Register rd, Register rj, Register rk);
void orn(Register rd, Register rj, Register rk);
void andi(Register rd, Register rj, int32_t ui12);
void ori(Register rd, Register rj, int32_t ui12);
void xori(Register rd, Register rj, int32_t ui12);
void mul_w(Register rd, Register rj, Register rk);
void mulh_w(Register rd, Register rj, Register rk);
void mulh_wu(Register rd, Register rj, Register rk);
void mul_d(Register rd, Register rj, Register rk);
void mulh_d(Register rd, Register rj, Register rk);
void mulh_du(Register rd, Register rj, Register rk);
void mulw_d_w(Register rd, Register rj, Register rk);
void mulw_d_wu(Register rd, Register rj, Register rk);
void div_w(Register rd, Register rj, Register rk);
void mod_w(Register rd, Register rj, Register rk);
void div_wu(Register rd, Register rj, Register rk);
void mod_wu(Register rd, Register rj, Register rk);
void div_d(Register rd, Register rj, Register rk);
void mod_d(Register rd, Register rj, Register rk);
void div_du(Register rd, Register rj, Register rk);
void mod_du(Register rd, Register rj, Register rk);
// Shifts.
void sll_w(Register rd, Register rj, Register rk);
void srl_w(Register rd, Register rj, Register rk);
void sra_w(Register rd, Register rj, Register rk);
void rotr_w(Register rd, Register rj, Register rk);
void slli_w(Register rd, Register rj, int32_t ui5);
void srli_w(Register rd, Register rj, int32_t ui5);
void srai_w(Register rd, Register rj, int32_t ui5);
void rotri_w(Register rd, Register rj, int32_t ui5);
void sll_d(Register rd, Register rj, Register rk);
void srl_d(Register rd, Register rj, Register rk);
void sra_d(Register rd, Register rj, Register rk);
void rotr_d(Register rd, Register rj, Register rk);
void slli_d(Register rd, Register rj, int32_t ui6);
void srli_d(Register rd, Register rj, int32_t ui6);
void srai_d(Register rd, Register rj, int32_t ui6);
void rotri_d(Register rd, Register rj, int32_t ui6);
// Bit twiddling.
void ext_w_b(Register rd, Register rj);
void ext_w_h(Register rd, Register rj);
void clo_w(Register rd, Register rj);
void clz_w(Register rd, Register rj);
void cto_w(Register rd, Register rj);
void ctz_w(Register rd, Register rj);
void clo_d(Register rd, Register rj);
void clz_d(Register rd, Register rj);
void cto_d(Register rd, Register rj);
void ctz_d(Register rd, Register rj);
void bytepick_w(Register rd, Register rj, Register rk, int32_t sa2);
void bytepick_d(Register rd, Register rj, Register rk, int32_t sa3);
void revb_2h(Register rd, Register rj);
void revb_4h(Register rd, Register rj);
void revb_2w(Register rd, Register rj);
void revb_d(Register rd, Register rj);
void revh_2w(Register rd, Register rj);
void revh_d(Register rd, Register rj);
void bitrev_4b(Register rd, Register rj);
void bitrev_8b(Register rd, Register rj);
void bitrev_w(Register rd, Register rj);
void bitrev_d(Register rd, Register rj);
void bstrins_w(Register rd, Register rj, int32_t msbw, int32_t lsbw);
void bstrins_d(Register rd, Register rj, int32_t msbd, int32_t lsbd);
void bstrpick_w(Register rd, Register rj, int32_t msbw, int32_t lsbw);
void bstrpick_d(Register rd, Register rj, int32_t msbd, int32_t lsbd);
void maskeqz(Register rd, Register rj, Register rk);
void masknez(Register rd, Register rj, Register rk);
// Memory-instructions
void ld_b(Register rd, Register rj, int32_t si12);
void ld_h(Register rd, Register rj, int32_t si12);
void ld_w(Register rd, Register rj, int32_t si12);
void ld_d(Register rd, Register rj, int32_t si12);
void ld_bu(Register rd, Register rj, int32_t si12);
void ld_hu(Register rd, Register rj, int32_t si12);
void ld_wu(Register rd, Register rj, int32_t si12);
void st_b(Register rd, Register rj, int32_t si12);
void st_h(Register rd, Register rj, int32_t si12);
void st_w(Register rd, Register rj, int32_t si12);
void st_d(Register rd, Register rj, int32_t si12);
void ldx_b(Register rd, Register rj, Register rk);
void ldx_h(Register rd, Register rj, Register rk);
void ldx_w(Register rd, Register rj, Register rk);
void ldx_d(Register rd, Register rj, Register rk);
void ldx_bu(Register rd, Register rj, Register rk);
void ldx_hu(Register rd, Register rj, Register rk);
void ldx_wu(Register rd, Register rj, Register rk);
void stx_b(Register rd, Register rj, Register rk);
void stx_h(Register rd, Register rj, Register rk);
void stx_w(Register rd, Register rj, Register rk);
void stx_d(Register rd, Register rj, Register rk);
void ldptr_w(Register rd, Register rj, int32_t si14);
void ldptr_d(Register rd, Register rj, int32_t si14);
void stptr_w(Register rd, Register rj, int32_t si14);
void stptr_d(Register rd, Register rj, int32_t si14);
void amswap_w(Register rd, Register rk, Register rj);
void amswap_d(Register rd, Register rk, Register rj);
void amadd_w(Register rd, Register rk, Register rj);
void amadd_d(Register rd, Register rk, Register rj);
void amand_w(Register rd, Register rk, Register rj);
void amand_d(Register rd, Register rk, Register rj);
void amor_w(Register rd, Register rk, Register rj);
void amor_d(Register rd, Register rk, Register rj);
void amxor_w(Register rd, Register rk, Register rj);
void amxor_d(Register rd, Register rk, Register rj);
void ammax_w(Register rd, Register rk, Register rj);
void ammax_d(Register rd, Register rk, Register rj);
void ammin_w(Register rd, Register rk, Register rj);
void ammin_d(Register rd, Register rk, Register rj);
void ammax_wu(Register rd, Register rk, Register rj);
void ammax_du(Register rd, Register rk, Register rj);
void ammin_wu(Register rd, Register rk, Register rj);
void ammin_du(Register rd, Register rk, Register rj);
void amswap_db_w(Register rd, Register rk, Register rj);
void amswap_db_d(Register rd, Register rk, Register rj);
void amadd_db_w(Register rd, Register rk, Register rj);
void amadd_db_d(Register rd, Register rk, Register rj);
void amand_db_w(Register rd, Register rk, Register rj);
void amand_db_d(Register rd, Register rk, Register rj);
void amor_db_w(Register rd, Register rk, Register rj);
void amor_db_d(Register rd, Register rk, Register rj);
void amxor_db_w(Register rd, Register rk, Register rj);
void amxor_db_d(Register rd, Register rk, Register rj);
void ammax_db_w(Register rd, Register rk, Register rj);
void ammax_db_d(Register rd, Register rk, Register rj);
void ammin_db_w(Register rd, Register rk, Register rj);
void ammin_db_d(Register rd, Register rk, Register rj);
void ammax_db_wu(Register rd, Register rk, Register rj);
void ammax_db_du(Register rd, Register rk, Register rj);
void ammin_db_wu(Register rd, Register rk, Register rj);
void ammin_db_du(Register rd, Register rk, Register rj);
void ll_w(Register rd, Register rj, int32_t si14);
void ll_d(Register rd, Register rj, int32_t si14);
void sc_w(Register rd, Register rj, int32_t si14);
void sc_d(Register rd, Register rj, int32_t si14);
void dbar(int32_t hint);
void ibar(int32_t hint);
// Break instruction
void break_(uint32_t code, bool break_as_stop = false);
void stop(uint32_t code = kMaxStopCode);
// Arithmetic.
void fadd_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fadd_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fsub_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fsub_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fmul_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fmul_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fdiv_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fdiv_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fmadd_s(FPURegister fd, FPURegister fj, FPURegister fk, FPURegister fa);
void fmadd_d(FPURegister fd, FPURegister fj, FPURegister fk, FPURegister fa);
void fmsub_s(FPURegister fd, FPURegister fj, FPURegister fk, FPURegister fa);
void fmsub_d(FPURegister fd, FPURegister fj, FPURegister fk, FPURegister fa);
void fnmadd_s(FPURegister fd, FPURegister fj, FPURegister fk, FPURegister fa);
void fnmadd_d(FPURegister fd, FPURegister fj, FPURegister fk, FPURegister fa);
void fnmsub_s(FPURegister fd, FPURegister fj, FPURegister fk, FPURegister fa);
void fnmsub_d(FPURegister fd, FPURegister fj, FPURegister fk, FPURegister fa);
void fmax_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fmax_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fmin_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fmin_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fmaxa_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fmaxa_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fmina_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fmina_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fabs_s(FPURegister fd, FPURegister fj);
void fabs_d(FPURegister fd, FPURegister fj);
void fneg_s(FPURegister fd, FPURegister fj);
void fneg_d(FPURegister fd, FPURegister fj);
void fsqrt_s(FPURegister fd, FPURegister fj);
void fsqrt_d(FPURegister fd, FPURegister fj);
void frecip_s(FPURegister fd, FPURegister fj);
void frecip_d(FPURegister fd, FPURegister fj);
void frsqrt_s(FPURegister fd, FPURegister fj);
void frsqrt_d(FPURegister fd, FPURegister fj);
void fscaleb_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fscaleb_d(FPURegister fd, FPURegister fj, FPURegister fk);
void flogb_s(FPURegister fd, FPURegister fj);
void flogb_d(FPURegister fd, FPURegister fj);
void fcopysign_s(FPURegister fd, FPURegister fj, FPURegister fk);
void fcopysign_d(FPURegister fd, FPURegister fj, FPURegister fk);
void fclass_s(FPURegister fd, FPURegister fj);
void fclass_d(FPURegister fd, FPURegister fj);
void fcmp_cond_s(FPUCondition cc, FPURegister fj, FPURegister fk,
CFRegister cd);
void fcmp_cond_d(FPUCondition cc, FPURegister fj, FPURegister fk,
CFRegister cd);
void fcvt_s_d(FPURegister fd, FPURegister fj);
void fcvt_d_s(FPURegister fd, FPURegister fj);
void ffint_s_w(FPURegister fd, FPURegister fj);
void ffint_s_l(FPURegister fd, FPURegister fj);
void ffint_d_w(FPURegister fd, FPURegister fj);
void ffint_d_l(FPURegister fd, FPURegister fj);
void ftint_w_s(FPURegister fd, FPURegister fj);
void ftint_w_d(FPURegister fd, FPURegister fj);
void ftint_l_s(FPURegister fd, FPURegister fj);
void ftint_l_d(FPURegister fd, FPURegister fj);
void ftintrm_w_s(FPURegister fd, FPURegister fj);
void ftintrm_w_d(FPURegister fd, FPURegister fj);
void ftintrm_l_s(FPURegister fd, FPURegister fj);
void ftintrm_l_d(FPURegister fd, FPURegister fj);
void ftintrp_w_s(FPURegister fd, FPURegister fj);
void ftintrp_w_d(FPURegister fd, FPURegister fj);
void ftintrp_l_s(FPURegister fd, FPURegister fj);
void ftintrp_l_d(FPURegister fd, FPURegister fj);
void ftintrz_w_s(FPURegister fd, FPURegister fj);
void ftintrz_w_d(FPURegister fd, FPURegister fj);
void ftintrz_l_s(FPURegister fd, FPURegister fj);
void ftintrz_l_d(FPURegister fd, FPURegister fj);
void ftintrne_w_s(FPURegister fd, FPURegister fj);
void ftintrne_w_d(FPURegister fd, FPURegister fj);
void ftintrne_l_s(FPURegister fd, FPURegister fj);
void ftintrne_l_d(FPURegister fd, FPURegister fj);
void frint_s(FPURegister fd, FPURegister fj);
void frint_d(FPURegister fd, FPURegister fj);
void fmov_s(FPURegister fd, FPURegister fj);
void fmov_d(FPURegister fd, FPURegister fj);
void fsel(CFRegister ca, FPURegister fd, FPURegister fj, FPURegister fk);
void movgr2fr_w(FPURegister fd, Register rj);
void movgr2fr_d(FPURegister fd, Register rj);
void movgr2frh_w(FPURegister fd, Register rj);
void movfr2gr_s(Register rd, FPURegister fj);
void movfr2gr_d(Register rd, FPURegister fj);
void movfrh2gr_s(Register rd, FPURegister fj);
void movgr2fcsr(Register rj, FPUControlRegister fcsr = FCSR0);
void movfcsr2gr(Register rd, FPUControlRegister fcsr = FCSR0);
void movfr2cf(CFRegister cd, FPURegister fj);
void movcf2fr(FPURegister fd, CFRegister cj);
void movgr2cf(CFRegister cd, Register rj);
void movcf2gr(Register rd, CFRegister cj);
void fld_s(FPURegister fd, Register rj, int32_t si12);
void fld_d(FPURegister fd, Register rj, int32_t si12);
void fst_s(FPURegister fd, Register rj, int32_t si12);
void fst_d(FPURegister fd, Register rj, int32_t si12);
void fldx_s(FPURegister fd, Register rj, Register rk);
void fldx_d(FPURegister fd, Register rj, Register rk);
void fstx_s(FPURegister fd, Register rj, Register rk);
void fstx_d(FPURegister fd, Register rj, Register rk);
// Check the code size generated from label to here.
int SizeOfCodeGeneratedSince(Label* label) {
return pc_offset() - label->pos();
}
// Check the number of instructions generated from label to here.
int InstructionsGeneratedSince(Label* label) {
return SizeOfCodeGeneratedSince(label) / kInstrSize;
}
// Class for scoping postponing the trampoline pool generation.
class V8_NODISCARD BlockTrampolinePoolScope {
public:
explicit BlockTrampolinePoolScope(Assembler* assem) : assem_(assem) {
assem_->StartBlockTrampolinePool();
}
~BlockTrampolinePoolScope() { assem_->EndBlockTrampolinePool(); }
private:
Assembler* assem_;
DISALLOW_IMPLICIT_CONSTRUCTORS(BlockTrampolinePoolScope);
};
// Class for postponing the assembly buffer growth. Typically used for
// sequences of instructions that must be emitted as a unit, before
// buffer growth (and relocation) can occur.
// This blocking scope is not nestable.
class V8_NODISCARD BlockGrowBufferScope {
public:
explicit BlockGrowBufferScope(Assembler* assem) : assem_(assem) {
assem_->StartBlockGrowBuffer();
}
~BlockGrowBufferScope() { assem_->EndBlockGrowBuffer(); }
private:
Assembler* assem_;
DISALLOW_IMPLICIT_CONSTRUCTORS(BlockGrowBufferScope);
};
// Record a deoptimization reason that can be used by a log or cpu profiler.
// Use --trace-deopt to enable.
void RecordDeoptReason(DeoptimizeReason reason, uint32_t node_id,
SourcePosition position, int id);
static int RelocateInternalReference(RelocInfo::Mode rmode, Address pc,
intptr_t pc_delta);
static void RelocateRelativeReference(RelocInfo::Mode rmode, Address pc,
intptr_t pc_delta);
// Writes a single byte or word of data in the code stream. Used for
// inline tables, e.g., jump-tables.
void db(uint8_t data);
void dd(uint32_t data, RelocInfo::Mode rmode = RelocInfo::NO_INFO);
void dq(uint64_t data, RelocInfo::Mode rmode = RelocInfo::NO_INFO);
void dp(uintptr_t data, RelocInfo::Mode rmode = RelocInfo::NO_INFO) {
dq(data, rmode);
}
void dd(Label* label);
// Postpone the generation of the trampoline pool for the specified number of
// instructions.
void BlockTrampolinePoolFor(int instructions);
// Check if there is less than kGap bytes available in the buffer.
// If this is the case, we need to grow the buffer before emitting
// an instruction or relocation information.
inline bool overflow() const { return pc_ >= reloc_info_writer.pos() - kGap; }
// Get the number of bytes available in the buffer.
inline intptr_t available_space() const {
return reloc_info_writer.pos() - pc_;
}
// Read/patch instructions.
static Instr instr_at(Address pc) { return *reinterpret_cast<Instr*>(pc); }
static void instr_at_put(Address pc, Instr instr) {
*reinterpret_cast<Instr*>(pc) = instr;
}
Instr instr_at(int pos) {
return *reinterpret_cast<Instr*>(buffer_start_ + pos);
}
void instr_at_put(int pos, Instr instr) {
*reinterpret_cast<Instr*>(buffer_start_ + pos) = instr;
}
// Check if an instruction is a branch of some kind.
static bool IsBranch(Instr instr);
static bool IsB(Instr instr);
static bool IsBz(Instr instr);
static bool IsNal(Instr instr);
static bool IsBeq(Instr instr);
static bool IsBne(Instr instr);
static bool IsJump(Instr instr);
static bool IsMov(Instr instr, Register rd, Register rs);
static bool IsPcAddi(Instr instr, Register rd, int32_t si20);
static bool IsJ(Instr instr);
static bool IsLu12i_w(Instr instr);
static bool IsOri(Instr instr);
static bool IsLu32i_d(Instr instr);
static bool IsLu52i_d(Instr instr);
static bool IsNop(Instr instr, unsigned int type);
static Register GetRjReg(Instr instr);
static Register GetRkReg(Instr instr);
static Register GetRdReg(Instr instr);
static uint32_t GetRj(Instr instr);
static uint32_t GetRjField(Instr instr);
static uint32_t GetRk(Instr instr);
static uint32_t GetRkField(Instr instr);
static uint32_t GetRd(Instr instr);
static uint32_t GetRdField(Instr instr);
static uint32_t GetSa2(Instr instr);
static uint32_t GetSa3(Instr instr);
static uint32_t GetSa2Field(Instr instr);
static uint32_t GetSa3Field(Instr instr);
static uint32_t GetOpcodeField(Instr instr);
static uint32_t GetFunction(Instr instr);
static uint32_t GetFunctionField(Instr instr);
static uint32_t GetImmediate16(Instr instr);
static uint32_t GetLabelConst(Instr instr);
static bool IsAddImmediate(Instr instr);
static Instr SetAddImmediateOffset(Instr instr, int16_t offset);
static bool IsAndImmediate(Instr instr);
static bool IsEmittedConstant(Instr instr);
void CheckTrampolinePool();
// Get the code target object for a pc-relative call or jump.
V8_INLINE Handle<Code> relative_code_target_object_handle_at(
Address pc_) const;
inline int UnboundLabelsCount() { return unbound_labels_count_; }
protected:
// Helper function for memory load/store.
void AdjustBaseAndOffset(MemOperand* src);
inline static void set_target_internal_reference_encoded_at(Address pc,
Address target);
int64_t buffer_space() const { return reloc_info_writer.pos() - pc_; }
// Decode branch instruction at pos and return branch target pos.
int target_at(int pos, bool is_internal);
// Patch branch instruction at pos to branch to given branch target pos.
void target_at_put(int pos, int target_pos, bool is_internal);
// Say if we need to relocate with this mode.
bool MustUseReg(RelocInfo::Mode rmode);
// Record reloc info for current pc_.
void RecordRelocInfo(RelocInfo::Mode rmode, intptr_t data = 0);
// Block the emission of the trampoline pool before pc_offset.
void BlockTrampolinePoolBefore(int pc_offset) {
if (no_trampoline_pool_before_ < pc_offset)
no_trampoline_pool_before_ = pc_offset;
}
void StartBlockTrampolinePool() { trampoline_pool_blocked_nesting_++; }
void EndBlockTrampolinePool() {
trampoline_pool_blocked_nesting_--;
if (trampoline_pool_blocked_nesting_ == 0) {
CheckTrampolinePoolQuick(1);
}
}
bool is_trampoline_pool_blocked() const {
return trampoline_pool_blocked_nesting_ > 0;
}
bool has_exception() const { return internal_trampoline_exception_; }
bool is_trampoline_emitted() const { return trampoline_emitted_; }
// Temporarily block automatic assembly buffer growth.
void StartBlockGrowBuffer() {
DCHECK(!block_buffer_growth_);
block_buffer_growth_ = true;
}
void EndBlockGrowBuffer() {
DCHECK(block_buffer_growth_);
block_buffer_growth_ = false;
}
bool is_buffer_growth_blocked() const { return block_buffer_growth_; }
void CheckTrampolinePoolQuick(int extra_instructions = 0) {
if (pc_offset() >= next_buffer_check_ - extra_instructions * kInstrSize) {
CheckTrampolinePool();
}
}
void set_pc_for_safepoint() { pc_for_safepoint_ = pc_; }
private:
// Avoid overflows for displacements etc.
static const int kMaximalBufferSize = 512 * MB;
// Buffer size and constant pool distance are checked together at regular
// intervals of kBufferCheckInterval emitted bytes.
static constexpr int kBufferCheckInterval = 1 * KB / 2;
// Code generation.
// The relocation writer's position is at least kGap bytes below the end of
// the generated instructions. This is so that multi-instruction sequences do
// not have to check for overflow. The same is true for writes of large
// relocation info entries.
static constexpr int kGap = 64;
STATIC_ASSERT(AssemblerBase::kMinimalBufferSize >= 2 * kGap);
// Repeated checking whether the trampoline pool should be emitted is rather
// expensive. By default we only check again once a number of instructions
// has been generated.
static constexpr int kCheckConstIntervalInst = 32;
static constexpr int kCheckConstInterval =
kCheckConstIntervalInst * kInstrSize;
int next_buffer_check_; // pc offset of next buffer check.
// Emission of the trampoline pool may be blocked in some code sequences.
int trampoline_pool_blocked_nesting_; // Block emission if this is not zero.
int no_trampoline_pool_before_; // Block emission before this pc offset.
// Keep track of the last emitted pool to guarantee a maximal distance.
int last_trampoline_pool_end_; // pc offset of the end of the last pool.
// Automatic growth of the assembly buffer may be blocked for some sequences.
bool block_buffer_growth_; // Block growth when true.
// Relocation information generation.
// Each relocation is encoded as a variable size value.
static constexpr int kMaxRelocSize = RelocInfoWriter::kMaxSize;
RelocInfoWriter reloc_info_writer;
// The bound position, before this we cannot do instruction elimination.
int last_bound_pos_;
// Code emission.
inline void CheckBuffer();
void GrowBuffer();
inline void emit(Instr x);
inline void emit(uint64_t x);
template <typename T>
inline void EmitHelper(T x);
inline void EmitHelper(Instr x);
void GenB(Opcode opcode, Register rj, int32_t si21); // opcode:6
void GenB(Opcode opcode, CFRegister cj, int32_t si21, bool isEq);
void GenB(Opcode opcode, int32_t si26);
void GenBJ(Opcode opcode, Register rj, Register rd, int32_t si16);
void GenCmp(Opcode opcode, FPUCondition cond, FPURegister fk, FPURegister fj,
CFRegister cd);
void GenSel(Opcode opcode, CFRegister ca, FPURegister fk, FPURegister fj,
FPURegister rd);
void GenRegister(Opcode opcode, Register rj, Register rd, bool rjrd = true);
void GenRegister(Opcode opcode, FPURegister fj, FPURegister fd);
void GenRegister(Opcode opcode, Register rj, FPURegister fd);
void GenRegister(Opcode opcode, FPURegister fj, Register rd);
void GenRegister(Opcode opcode, Register rj, FPUControlRegister fd);
void GenRegister(Opcode opcode, FPUControlRegister fj, Register rd);
void GenRegister(Opcode opcode, FPURegister fj, CFRegister cd);
void GenRegister(Opcode opcode, CFRegister cj, FPURegister fd);
void GenRegister(Opcode opcode, Register rj, CFRegister cd);
void GenRegister(Opcode opcode, CFRegister cj, Register rd);
void GenRegister(Opcode opcode, Register rk, Register rj, Register rd);
void GenRegister(Opcode opcode, FPURegister fk, FPURegister fj,
FPURegister fd);
void GenRegister(Opcode opcode, FPURegister fa, FPURegister fk,
FPURegister fj, FPURegister fd);
void GenRegister(Opcode opcode, Register rk, Register rj, FPURegister fd);
void GenImm(Opcode opcode, int32_t bit3, Register rk, Register rj,
Register rd);
void GenImm(Opcode opcode, int32_t bit6m, int32_t bit6l, Register rj,
Register rd);
void GenImm(Opcode opcode, int32_t bit20, Register rd);
void GenImm(Opcode opcode, int32_t bit15);
void GenImm(Opcode opcode, int32_t value, Register rj, Register rd,
int32_t value_bits); // 6 | 12 | 14 | 16
void GenImm(Opcode opcode, int32_t bit12, Register rj, FPURegister fd);
// Labels.
void print(const Label* L);
void bind_to(Label* L, int pos);
void next(Label* L, bool is_internal);
// One trampoline consists of:
// - space for trampoline slots,
// - space for labels.
//
// Space for trampoline slots is equal to slot_count * 2 * kInstrSize.
// Space for trampoline slots precedes space for labels. Each label is of one
// instruction size, so total amount for labels is equal to
// label_count * kInstrSize.
class Trampoline {
public:
Trampoline() {
start_ = 0;
next_slot_ = 0;
free_slot_count_ = 0;
end_ = 0;
}
Trampoline(int start, int slot_count) {
start_ = start;
next_slot_ = start;
free_slot_count_ = slot_count;
end_ = start + slot_count * kTrampolineSlotsSize;
}
int start() { return start_; }
int end() { return end_; }
int take_slot() {
int trampoline_slot = kInvalidSlotPos;
if (free_slot_count_ <= 0) {
// We have run out of space on trampolines.
// Make sure we fail in debug mode, so we become aware of each case
// when this happens.
DCHECK(0);
// Internal exception will be caught.
} else {
trampoline_slot = next_slot_;
free_slot_count_--;
next_slot_ += kTrampolineSlotsSize;
}
return trampoline_slot;
}
private:
int start_;
int end_;
int next_slot_;
int free_slot_count_;
};
int32_t get_trampoline_entry(int32_t pos);
int unbound_labels_count_;
// After trampoline is emitted, long branches are used in generated code for
// the forward branches whose target offsets could be beyond reach of branch
// instruction. We use this information to trigger different mode of
// branch instruction generation, where we use jump instructions rather
// than regular branch instructions.
bool trampoline_emitted_;
static constexpr int kInvalidSlotPos = -1;
// Internal reference positions, required for unbounded internal reference
// labels.
std::set<int64_t> internal_reference_positions_;
bool is_internal_reference(Label* L) {
return internal_reference_positions_.find(L->pos()) !=
internal_reference_positions_.end();
}
void EmittedCompactBranchInstruction() { prev_instr_compact_branch_ = true; }
void ClearCompactBranchState() { prev_instr_compact_branch_ = false; }
bool prev_instr_compact_branch_ = false;
Trampoline trampoline_;
bool internal_trampoline_exception_;
// Keep track of the last Call's position to ensure that safepoint can get the
// correct information even if there is a trampoline immediately after the
// Call.
byte* pc_for_safepoint_;
RegList scratch_register_list_;
private:
void AllocateAndInstallRequestedHeapObjects(Isolate* isolate);
int WriteCodeComments();
friend class RegExpMacroAssemblerLOONG64;
friend class RelocInfo;
friend class BlockTrampolinePoolScope;
friend class EnsureSpace;
};
class EnsureSpace {
public:
explicit inline EnsureSpace(Assembler* assembler);
};
class V8_EXPORT_PRIVATE V8_NODISCARD UseScratchRegisterScope {
public:
explicit UseScratchRegisterScope(Assembler* assembler);
~UseScratchRegisterScope();
Register Acquire();
bool hasAvailable() const;
void Include(const RegList& list) { *available_ |= list; }
void Exclude(const RegList& list) { *available_ &= ~list; }
void Include(const Register& reg1, const Register& reg2 = no_reg) {
RegList list({reg1, reg2});
Include(list);
}
void Exclude(const Register& reg1, const Register& reg2 = no_reg) {
RegList list({reg1, reg2});
Exclude(list);
}
private:
RegList* available_;
RegList old_available_;
};
} // namespace internal
} // namespace v8
#endif // V8_CODEGEN_LOONG64_ASSEMBLER_LOONG64_H_