blob: b0e233f4e2d24a442be2936fc08401c1fddc2866 [file] [log] [blame]
* Copyright (C) 2013 Google Inc. All rights reserved.
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are
* met:
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following disclaimer
* in the documentation and/or other materials provided with the
* distribution.
* * Neither the name of Google Inc. nor the names of its
* contributors may be used to endorse or promote products derived from
* this software without specific prior written permission.
#ifndef WTF_PartitionAlloc_h
#define WTF_PartitionAlloc_h
// partitionAlloc() / partitionAllocGeneric() and partitionFree() /
// partitionFreeGeneric() are approximately analagous to malloc() and free().
// The main difference is that a PartitionRoot / PartitionRootGeneric object
// must be supplied to these functions, representing a specific "heap partition"
// that will be used to satisfy the allocation. Different partitions are
// guaranteed to exist in separate address spaces, including being separate from
// the main system heap. If the contained objects are all freed, physical memory
// is returned to the system but the address space remains reserved.
// SizeSpecificPartitionAllocator / PartitionAllocatorGeneric classes. To
// minimize the instruction count to the fullest extent possible, the
// PartitionRoot is really just a header adjacent to other data areas provided
// by the allocator class.
// The partitionAlloc() variant of the API has the following caveats:
// - Allocations and frees against a single partition must be single threaded.
// - Allocations must not exceed a max size, chosen at compile-time via a
// templated parameter to PartitionAllocator.
// - Allocation sizes must be aligned to the system pointer size.
// - Allocations are bucketed exactly according to size.
// And for partitionAllocGeneric():
// - Multi-threaded use against a single partition is ok; locking is handled.
// - Allocations of any arbitrary size can be handled (subject to a limit of
// INT_MAX bytes for security reasons).
// - Bucketing is by approximate size, for example an allocation of 4000 bytes
// might be placed into a 4096-byte bucket. Bucket sizes are chosen to try and
// keep worst-case waste to ~10%.
// The allocators are designed to be extremely fast, thanks to the following
// properties and design:
// - Just a single (reasonably predicatable) branch in the hot / fast path for
// both allocating and (significantly) freeing.
// - A minimal number of operations in the hot / fast path, with the slow paths
// in separate functions, leading to the possibility of inlining.
// - Each partition page (which is usually multiple physical pages) has a
// metadata structure which allows fast mapping of free() address to an
// underlying bucket.
// - Supports a lock-free API for fast performance in single-threaded cases.
// - The freelist for a given bucket is split across a number of partition
// pages, enabling various simple tricks to try and minimize fragmentation.
// - Fine-grained bucket sizes leading to less waste and better packing.
// The following security properties are provided at this time:
// - Linear overflows cannot corrupt into the partition.
// - Linear overflows cannot corrupt out of the partition.
// - Freed pages will only be re-used within the partition.
// (exception: large allocations > ~1MB)
// - Freed pages will only hold same-sized objects when re-used.
// - Dereference of freelist pointer should fault.
// - Out-of-line main metadata: linear over or underflow cannot corrupt it.
// - Partial pointer overwrite of freelist pointer should fault.
// - Rudimentary double-free detection.
// - Large allocations (> ~1MB) are guard-paged at the beginning and end.
// The following security properties could be investigated in the future:
// - Per-object bucketing (instead of per-size) is mostly available at the API,
// but not used yet.
// - No randomness of freelist entries or bucket position.
// - Better checking for wild pointers in free().
// - Better freelist masking function to guarantee fault on 32-bit.
#include "wtf/Assertions.h"
#include "wtf/BitwiseOperations.h"
#include "wtf/ByteSwap.h"
#include "wtf/CPU.h"
#include "wtf/PageAllocator.h"
#include "wtf/SpinLock.h"
#include <limits.h>
#include <stdlib.h>
#include <string.h>
namespace WTF {
// Allocation granularity of sizeof(void*) bytes.
static const size_t kAllocationGranularity = sizeof(void*);
static const size_t kAllocationGranularityMask = kAllocationGranularity - 1;
static const size_t kBucketShift = (kAllocationGranularity == 8) ? 3 : 2;
// Underlying partition storage pages are a power-of-two size. It is typical
// for a partition page to be based on multiple system pages. Most references to
// "page" refer to partition pages.
// We also have the concept of "super pages" -- these are the underlying system
// allocations we make. Super pages contain multiple partition pages inside them
// and include space for a small amount of metadata per partition page.
// Inside super pages, we store "slot spans". A slot span is a continguous range
// of one or more partition pages that stores allocations of the same size.
// Slot span sizes are adjusted depending on the allocation size, to make sure
// the packing does not lead to unused (wasted) space at the end of the last
// system page of the span. For our current max slot span size of 64k and other
// constant values, we pack _all_ partitionAllocGeneric() sizes perfectly up
// against the end of a system page.
static const size_t kPartitionPageShift = 14; // 16KB
static const size_t kPartitionPageSize = 1 << kPartitionPageShift;
static const size_t kPartitionPageOffsetMask = kPartitionPageSize - 1;
static const size_t kPartitionPageBaseMask = ~kPartitionPageOffsetMask;
static const size_t kMaxPartitionPagesPerSlotSpan = 4;
// To avoid fragmentation via never-used freelist entries, we hand out partition
// freelist sections gradually, in units of the dominant system page size.
// What we're actually doing is avoiding filling the full partition page (16 KB)
// with freelist pointers right away. Writing freelist pointers will fault and
// dirty a private page, which is very wasteful if we never actually store
// objects there.
static const size_t kNumSystemPagesPerPartitionPage = kPartitionPageSize / kSystemPageSize;
static const size_t kMaxSystemPagesPerSlotSpan = kNumSystemPagesPerPartitionPage * kMaxPartitionPagesPerSlotSpan;
// We reserve virtual address space in 2MB chunks (aligned to 2MB as well).
// These chunks are called "super pages". We do this so that we can store
// metadata in the first few pages of each 2MB aligned section. This leads to
// a very fast free(). We specifically choose 2MB because this virtual address
// block represents a full but single PTE allocation on ARM, ia32 and x64.
// The layout of the super page is as follows. The sizes below are the same
// for 32 bit and 64 bit.
// | Guard page (4KB) | Metadata page (4KB) | Guard pages (8KB) | Slot span | Slot span | ... | Slot span | Guard page (4KB) |
// - Each slot span is a contiguous range of one or more PartitionPages.
// - The metadata page has the following format. Note that the PartitionPage
// that is not at the head of a slot span is "unused". In other words,
// the metadata for the slot span is stored only in the first PartitionPage
// of the slot span. Metadata accesses to other PartitionPages are
// redirected to the first PartitionPage.
// | SuperPageExtentEntry (32B) | PartitionPage of slot span 1 (32B, used) | PartitionPage of slot span 1 (32B, unused) | PartitionPage of slot span 1 (32B, unused) | PartitionPage of slot span 2 (32B, used) | PartitionPage of slot span 3 (32B, used) | ... | PartitionPage of slot span N (32B, unused) |
// A direct mapped page has a similar layout to fake it looking like a super page:
// | Guard page (4KB) | Metadata page (4KB) | Guard pages (8KB) | Direct mapped object | Guard page (4KB) |
// - The metadata page has the following layout:
// | SuperPageExtentEntry (32B) | PartitionPage (32B) | PartitionBucket (32B) | PartitionDirectMapExtent (8B) |
static const size_t kSuperPageShift = 21; // 2MB
static const size_t kSuperPageSize = 1 << kSuperPageShift;
static const size_t kSuperPageOffsetMask = kSuperPageSize - 1;
static const size_t kSuperPageBaseMask = ~kSuperPageOffsetMask;
static const size_t kNumPartitionPagesPerSuperPage = kSuperPageSize / kPartitionPageSize;
static const size_t kPageMetadataShift = 5; // 32 bytes per partition page.
static const size_t kPageMetadataSize = 1 << kPageMetadataShift;
// The following kGeneric* constants apply to the generic variants of the API.
// The "order" of an allocation is closely related to the power-of-two size of
// the allocation. More precisely, the order is the bit index of the
// most-significant-bit in the allocation size, where the bit numbers starts
// at index 1 for the least-significant-bit.
// In terms of allocation sizes, order 0 covers 0, order 1 covers 1, order 2
// covers 2->3, order 3 covers 4->7, order 4 covers 8->15.
static const size_t kGenericMinBucketedOrder = 4; // 8 bytes.
static const size_t kGenericMaxBucketedOrder = 20; // Largest bucketed order is 1<<(20-1) (storing 512KB -> almost 1MB)
static const size_t kGenericNumBucketedOrders = (kGenericMaxBucketedOrder - kGenericMinBucketedOrder) + 1;
static const size_t kGenericNumBucketsPerOrderBits = 3; // Eight buckets per order (for the higher orders), e.g. order 8 is 128, 144, 160, ..., 240
static const size_t kGenericNumBucketsPerOrder = 1 << kGenericNumBucketsPerOrderBits;
static const size_t kGenericNumBuckets = kGenericNumBucketedOrders * kGenericNumBucketsPerOrder;
static const size_t kGenericSmallestBucket = 1 << (kGenericMinBucketedOrder - 1);
static const size_t kGenericMaxBucketSpacing = 1 << ((kGenericMaxBucketedOrder - 1) - kGenericNumBucketsPerOrderBits);
static const size_t kGenericMaxBucketed = (1 << (kGenericMaxBucketedOrder - 1)) + ((kGenericNumBucketsPerOrder - 1) * kGenericMaxBucketSpacing);
static const size_t kGenericMinDirectMappedDownsize = kGenericMaxBucketed + 1; // Limit when downsizing a direct mapping using realloc().
static const size_t kGenericMaxDirectMapped = INT_MAX - kSystemPageSize;
static const size_t kBitsPerSizet = sizeof(void*) * CHAR_BIT;
// Constants for the memory reclaim logic.
static const size_t kMaxFreeableSpans = 16;
// If the total size in bytes of allocated but not committed pages exceeds this
// value (probably it is a "out of virtual address space" crash),
// a special crash stack trace is generated at |partitionOutOfMemory|.
// This is to distinguish "out of virtual address space" from
// "out of physical memory" in crash reports.
static const size_t kReasonableSizeOfUnusedPages = 1024 * 1024 * 1024; // 1GiB
// These two byte values match tcmalloc.
static const unsigned char kUninitializedByte = 0xAB;
static const unsigned char kFreedByte = 0xCD;
static const size_t kCookieSize = 16; // Handles alignment up to XMM instructions on Intel.
static const unsigned char kCookieValue[kCookieSize] = { 0xDE, 0xAD, 0xBE, 0xEF, 0xCA, 0xFE, 0xD0, 0x0D, 0x13, 0x37, 0xF0, 0x05, 0xBA, 0x11, 0xAB, 0x1E };
struct PartitionBucket;
struct PartitionRootBase;
struct PartitionFreelistEntry {
PartitionFreelistEntry* next;
// Some notes on page states. A page can be in one of four major states:
// 1) Active.
// 2) Full.
// 3) Empty.
// 4) Decommitted.
// An active page has available free slots. A full page has no free slots. An
// empty page has no free slots, and a decommitted page is an empty page that
// had its backing memory released back to the system.
// There are two linked lists tracking the pages. The "active page" list is an
// approximation of a list of active pages. It is an approximation because
// full, empty and decommitted pages may briefly be present in the list until
// we next do a scan over it.
// The "empty page" list is an accurate list of pages which are either empty
// or decommitted.
// The significant page transitions are:
// - free() will detect when a full page has a slot free()'d and immediately
// return the page to the head of the active list.
// - free() will detect when a page is fully emptied. It _may_ add it to the
// empty list or it _may_ leave it on the active list until a future list scan.
// - malloc() _may_ scan the active page list in order to fulfil the request.
// If it does this, full, empty and decommitted pages encountered will be
// booted out of the active list. If there are no suitable active pages found,
// an empty or decommitted page (if one exists) will be pulled from the empty
// list on to the active list.
struct PartitionPage {
PartitionFreelistEntry* freelistHead;
PartitionPage* nextPage;
PartitionBucket* bucket;
int16_t numAllocatedSlots; // Deliberately signed, 0 for empty or decommitted page, -n for full pages.
uint16_t numUnprovisionedSlots;
uint16_t pageOffset;
int16_t emptyCacheIndex; // -1 if not in the empty cache.
struct PartitionBucket {
PartitionPage* activePagesHead; // Accessed most in hot path => goes first.
PartitionPage* emptyPagesHead;
PartitionPage* decommittedPagesHead;
uint32_t slotSize;
uint16_t numSystemPagesPerSlotSpan;
uint16_t numFullPages;
// An "extent" is a span of consecutive superpages. We link to the partition's
// next extent (if there is one) at the very start of a superpage's metadata
// area.
struct PartitionSuperPageExtentEntry {
PartitionRootBase* root;
char* superPageBase;
char* superPagesEnd;
PartitionSuperPageExtentEntry* next;
struct PartitionDirectMapExtent {
PartitionDirectMapExtent* nextExtent;
PartitionDirectMapExtent* prevExtent;
PartitionBucket* bucket;
size_t mapSize; // Mapped size, not including guard pages and meta-data.
struct WTF_EXPORT PartitionRootBase {
size_t totalSizeOfCommittedPages;
size_t totalSizeOfSuperPages;
size_t totalSizeOfDirectMappedPages;
// Invariant: totalSizeOfCommittedPages <= totalSizeOfSuperPages + totalSizeOfDirectMappedPages.
unsigned numBuckets;
unsigned maxAllocation;
bool initialized;
char* nextSuperPage;
char* nextPartitionPage;
char* nextPartitionPageEnd;
PartitionSuperPageExtentEntry* currentExtent;
PartitionSuperPageExtentEntry* firstExtent;
PartitionDirectMapExtent* directMapList;
PartitionPage* globalEmptyPageRing[kMaxFreeableSpans];
int16_t globalEmptyPageRingIndex;
uintptr_t invertedSelf;
static int gInitializedLock;
static bool gInitialized;
// gSeedPage is used as a sentinel to indicate that there is no page
// in the active page list. We can use nullptr, but in that case we need
// to add a null-check branch to the hot allocation path. We want to avoid
// that.
static PartitionPage gSeedPage;
static PartitionBucket gPagedBucket;
// gOomHandlingFunction is invoked when ParitionAlloc hits OutOfMemory.
static void (*gOomHandlingFunction)();
// Never instantiate a PartitionRoot directly, instead use PartitionAlloc.
struct PartitionRoot : public PartitionRootBase {
// The PartitionAlloc templated class ensures the following is correct.
ALWAYS_INLINE PartitionBucket* buckets() { return reinterpret_cast<PartitionBucket*>(this + 1); }
ALWAYS_INLINE const PartitionBucket* buckets() const { return reinterpret_cast<const PartitionBucket*>(this + 1); }
// Never instantiate a PartitionRootGeneric directly, instead use PartitionAllocatorGeneric.
struct PartitionRootGeneric : public PartitionRootBase {
int lock;
// Some pre-computed constants.
size_t orderIndexShifts[kBitsPerSizet + 1];
size_t orderSubIndexMasks[kBitsPerSizet + 1];
// The bucket lookup table lets us map a size_t to a bucket quickly.
// The trailing +1 caters for the overflow case for very large allocation sizes.
// It is one flat array instead of a 2D array because in the 2D world, we'd
// need to index array[blah][max+1] which risks undefined behavior.
PartitionBucket* bucketLookups[((kBitsPerSizet + 1) * kGenericNumBucketsPerOrder) + 1];
PartitionBucket buckets[kGenericNumBuckets];
// Flags for partitionAllocGenericFlags.
enum PartitionAllocFlags {
PartitionAllocReturnNull = 1 << 0,
// Struct used to retrieve total memory usage of a partition. Used by
// PartitionStatsDumper implementation.
struct PartitionMemoryStats {
size_t totalMmappedBytes; // Total bytes mmaped from the system.
size_t totalCommittedBytes; // Total size of commmitted pages.
size_t totalResidentBytes; // Total bytes provisioned by the partition.
size_t totalActiveBytes; // Total active bytes in the partition.
size_t totalDecommittableBytes; // Total bytes that could be decommitted.
size_t totalDiscardableBytes; // Total bytes that could be discarded.
// Struct used to retrieve memory statistics about a partition bucket. Used by
// PartitionStatsDumper implementation.
struct PartitionBucketMemoryStats {
bool isValid; // Used to check if the stats is valid.
bool isDirectMap; // True if this is a direct mapping; size will not be unique.
uint32_t bucketSlotSize; // The size of the slot in bytes.
uint32_t allocatedPageSize; // Total size the partition page allocated from the system.
uint32_t activeBytes; // Total active bytes used in the bucket.
uint32_t residentBytes; // Total bytes provisioned in the bucket.
uint32_t decommittableBytes; // Total bytes that could be decommitted.
uint32_t discardableBytes; // Total bytes that could be discarded.
uint32_t numFullPages; // Number of pages with all slots allocated.
uint32_t numActivePages; // Number of pages that have at least one provisioned slot.
uint32_t numEmptyPages; // Number of pages that are empty but not decommitted.
uint32_t numDecommittedPages; // Number of pages that are empty and decommitted.
// Interface that is passed to partitionDumpStats and
// partitionDumpStatsGeneric for using the memory statistics.
class WTF_EXPORT PartitionStatsDumper {
// Called to dump total memory used by partition, once per partition.
virtual void partitionDumpTotals(const char* partitionName, const PartitionMemoryStats*) = 0;
// Called to dump stats about buckets, for each bucket.
virtual void partitionsDumpBucketStats(const char* partitionName, const PartitionBucketMemoryStats*) = 0;
WTF_EXPORT void partitionAllocGlobalInit(void (*oomHandlingFunction)());
WTF_EXPORT void partitionAllocInit(PartitionRoot*, size_t numBuckets, size_t maxAllocation);
WTF_EXPORT bool partitionAllocShutdown(PartitionRoot*);
WTF_EXPORT void partitionAllocGenericInit(PartitionRootGeneric*);
WTF_EXPORT bool partitionAllocGenericShutdown(PartitionRootGeneric*);
enum PartitionPurgeFlags {
// Decommitting the ring list of empty pages is reasonably fast.
PartitionPurgeDecommitEmptyPages = 1 << 0,
// Discarding unused system pages is slower, because it involves walking all
// freelists in all active partition pages of all buckets >= system page
// size. It often frees a similar amount of memory to decommitting the empty
// pages, though.
PartitionPurgeDiscardUnusedSystemPages = 1 << 1,
WTF_EXPORT void partitionPurgeMemory(PartitionRoot*, int);
WTF_EXPORT void partitionPurgeMemoryGeneric(PartitionRootGeneric*, int);
WTF_EXPORT NEVER_INLINE void* partitionAllocSlowPath(PartitionRootBase*, int, size_t, PartitionBucket*);
WTF_EXPORT NEVER_INLINE void partitionFreeSlowPath(PartitionPage*);
WTF_EXPORT NEVER_INLINE void* partitionReallocGeneric(PartitionRootGeneric*, void*, size_t);
WTF_EXPORT void partitionDumpStats(PartitionRoot*, const char* partitionName, bool isLightDump, PartitionStatsDumper*);
WTF_EXPORT void partitionDumpStatsGeneric(PartitionRootGeneric*, const char* partitionName, bool isLightDump, PartitionStatsDumper*);
ALWAYS_INLINE PartitionFreelistEntry* partitionFreelistMask(PartitionFreelistEntry* ptr)
// We use bswap on little endian as a fast mask for two reasons:
// 1) If an object is freed and its vtable used where the attacker doesn't
// get the chance to run allocations between the free and use, the vtable
// dereference is likely to fault.
// 2) If the attacker has a linear buffer overflow and elects to try and
// corrupt a freelist pointer, partial pointer overwrite attacks are
// thwarted.
// For big endian, similar guarantees are arrived at with a negation.
uintptr_t masked = ~reinterpret_cast<uintptr_t>(ptr);
uintptr_t masked = bswapuintptrt(reinterpret_cast<uintptr_t>(ptr));
return reinterpret_cast<PartitionFreelistEntry*>(masked);
ALWAYS_INLINE size_t partitionCookieSizeAdjustAdd(size_t size)
// Add space for cookies, checking for integer overflow.
ASSERT(size + (2 * kCookieSize) > size);
size += 2 * kCookieSize;
return size;
ALWAYS_INLINE size_t partitionCookieSizeAdjustSubtract(size_t size)
// Remove space for cookies.
ASSERT(size >= 2 * kCookieSize);
size -= 2 * kCookieSize;
return size;
ALWAYS_INLINE void* partitionCookieFreePointerAdjust(void* ptr)
// The value given to the application is actually just after the cookie.
ptr = static_cast<char*>(ptr) - kCookieSize;
return ptr;
ALWAYS_INLINE void partitionCookieWriteValue(void* ptr)
unsigned char* cookiePtr = reinterpret_cast<unsigned char*>(ptr);
for (size_t i = 0; i < kCookieSize; ++i, ++cookiePtr)
*cookiePtr = kCookieValue[i];
ALWAYS_INLINE void partitionCookieCheckValue(void* ptr)
unsigned char* cookiePtr = reinterpret_cast<unsigned char*>(ptr);
for (size_t i = 0; i < kCookieSize; ++i, ++cookiePtr)
ASSERT(*cookiePtr == kCookieValue[i]);
ALWAYS_INLINE char* partitionSuperPageToMetadataArea(char* ptr)
uintptr_t pointerAsUint = reinterpret_cast<uintptr_t>(ptr);
ASSERT(!(pointerAsUint & kSuperPageOffsetMask));
// The metadata area is exactly one system page (the guard page) into the
// super page.
return reinterpret_cast<char*>(pointerAsUint + kSystemPageSize);
ALWAYS_INLINE PartitionPage* partitionPointerToPageNoAlignmentCheck(void* ptr)
uintptr_t pointerAsUint = reinterpret_cast<uintptr_t>(ptr);
char* superPagePtr = reinterpret_cast<char*>(pointerAsUint & kSuperPageBaseMask);
uintptr_t partitionPageIndex = (pointerAsUint & kSuperPageOffsetMask) >> kPartitionPageShift;
// Index 0 is invalid because it is the metadata and guard area and
// the last index is invalid because it is a guard page.
ASSERT(partitionPageIndex < kNumPartitionPagesPerSuperPage - 1);
PartitionPage* page = reinterpret_cast<PartitionPage*>(partitionSuperPageToMetadataArea(superPagePtr) + (partitionPageIndex << kPageMetadataShift));
// Partition pages in the same slot span can share the same page object. Adjust for that.
size_t delta = page->pageOffset << kPageMetadataShift;
page = reinterpret_cast<PartitionPage*>(reinterpret_cast<char*>(page) - delta);
return page;
ALWAYS_INLINE void* partitionPageToPointer(const PartitionPage* page)
uintptr_t pointerAsUint = reinterpret_cast<uintptr_t>(page);
uintptr_t superPageOffset = (pointerAsUint & kSuperPageOffsetMask);
ASSERT(superPageOffset > kSystemPageSize);
ASSERT(superPageOffset < kSystemPageSize + (kNumPartitionPagesPerSuperPage * kPageMetadataSize));
uintptr_t partitionPageIndex = (superPageOffset - kSystemPageSize) >> kPageMetadataShift;
// Index 0 is invalid because it is the metadata area and the last index is invalid because it is a guard page.
ASSERT(partitionPageIndex < kNumPartitionPagesPerSuperPage - 1);
uintptr_t superPageBase = (pointerAsUint & kSuperPageBaseMask);
void* ret = reinterpret_cast<void*>(superPageBase + (partitionPageIndex << kPartitionPageShift));
return ret;
ALWAYS_INLINE PartitionPage* partitionPointerToPage(void* ptr)
PartitionPage* page = partitionPointerToPageNoAlignmentCheck(ptr);
// Checks that the pointer is a multiple of bucket size.
ASSERT(!((reinterpret_cast<uintptr_t>(ptr) - reinterpret_cast<uintptr_t>(partitionPageToPointer(page))) % page->bucket->slotSize));
return page;
ALWAYS_INLINE bool partitionBucketIsDirectMapped(const PartitionBucket* bucket)
return !bucket->numSystemPagesPerSlotSpan;
ALWAYS_INLINE size_t partitionBucketBytes(const PartitionBucket* bucket)
return bucket->numSystemPagesPerSlotSpan * kSystemPageSize;
ALWAYS_INLINE uint16_t partitionBucketSlots(const PartitionBucket* bucket)
return static_cast<uint16_t>(partitionBucketBytes(bucket) / bucket->slotSize);
ALWAYS_INLINE size_t* partitionPageGetRawSizePtr(PartitionPage* page)
// For single-slot buckets which span more than one partition page, we
// have some spare metadata space to store the raw allocation size. We
// can use this to report better statistics.
PartitionBucket* bucket = page->bucket;
if (bucket->slotSize <= kMaxSystemPagesPerSlotSpan * kSystemPageSize)
return nullptr;
ASSERT((bucket->slotSize % kSystemPageSize) == 0);
ASSERT(partitionBucketIsDirectMapped(bucket) || partitionBucketSlots(bucket) == 1);
return reinterpret_cast<size_t*>(&page->freelistHead);
ALWAYS_INLINE size_t partitionPageGetRawSize(PartitionPage* page)
size_t* rawSizePtr = partitionPageGetRawSizePtr(page);
if (UNLIKELY(rawSizePtr != nullptr))
return *rawSizePtr;
return 0;
ALWAYS_INLINE PartitionRootBase* partitionPageToRoot(PartitionPage* page)
PartitionSuperPageExtentEntry* extentEntry = reinterpret_cast<PartitionSuperPageExtentEntry*>(reinterpret_cast<uintptr_t>(page) & kSystemPageBaseMask);
return extentEntry->root;
ALWAYS_INLINE bool partitionPointerIsValid(void* ptr)
PartitionPage* page = partitionPointerToPage(ptr);
PartitionRootBase* root = partitionPageToRoot(page);
return root->invertedSelf == ~reinterpret_cast<uintptr_t>(root);
ALWAYS_INLINE void* partitionBucketAlloc(PartitionRootBase* root, int flags, size_t size, PartitionBucket* bucket)
PartitionPage* page = bucket->activePagesHead;
// Check that this page is neither full nor freed.
ASSERT(page->numAllocatedSlots >= 0);
void* ret = page->freelistHead;
if (LIKELY(ret != 0)) {
// If these asserts fire, you probably corrupted memory.
// All large allocations must go through the slow path to correctly
// update the size metadata.
ASSERT(partitionPageGetRawSize(page) == 0);
PartitionFreelistEntry* newHead = partitionFreelistMask(static_cast<PartitionFreelistEntry*>(ret)->next);
page->freelistHead = newHead;
} else {
ret = partitionAllocSlowPath(root, flags, size, bucket);
ASSERT(!ret || partitionPointerIsValid(ret));
if (!ret)
return 0;
// Fill the uninitialized pattern, and write the cookies.
page = partitionPointerToPage(ret);
size_t slotSize = page->bucket->slotSize;
size_t rawSize = partitionPageGetRawSize(page);
if (rawSize) {
ASSERT(rawSize == size);
slotSize = rawSize;
size_t noCookieSize = partitionCookieSizeAdjustSubtract(slotSize);
char* charRet = static_cast<char*>(ret);
// The value given to the application is actually just after the cookie.
ret = charRet + kCookieSize;
memset(ret, kUninitializedByte, noCookieSize);
partitionCookieWriteValue(charRet + kCookieSize + noCookieSize);
return ret;
ALWAYS_INLINE void* partitionAlloc(PartitionRoot* root, size_t size)
void* result = malloc(size);
return result;
size = partitionCookieSizeAdjustAdd(size);
size_t index = size >> kBucketShift;
ASSERT(index < root->numBuckets);
ASSERT(size == index << kBucketShift);
PartitionBucket* bucket = &root->buckets()[index];
return partitionBucketAlloc(root, 0, size, bucket);
ALWAYS_INLINE void partitionFreeWithPage(void* ptr, PartitionPage* page)
// If these asserts fire, you probably corrupted memory.
size_t slotSize = page->bucket->slotSize;
size_t rawSize = partitionPageGetRawSize(page);
if (rawSize)
slotSize = rawSize;
partitionCookieCheckValue(reinterpret_cast<char*>(ptr) + slotSize - kCookieSize);
memset(ptr, kFreedByte, slotSize);
PartitionFreelistEntry* freelistHead = page->freelistHead;
ASSERT(!freelistHead || partitionPointerIsValid(freelistHead));
RELEASE_ASSERT_WITH_SECURITY_IMPLICATION(ptr != freelistHead); // Catches an immediate double free.
ASSERT_WITH_SECURITY_IMPLICATION(!freelistHead || ptr != partitionFreelistMask(freelistHead->next)); // Look for double free one level deeper in debug.
PartitionFreelistEntry* entry = static_cast<PartitionFreelistEntry*>(ptr);
entry->next = partitionFreelistMask(freelistHead);
page->freelistHead = entry;
if (UNLIKELY(page->numAllocatedSlots <= 0)) {
} else {
// All single-slot allocations must go through the slow path to
// correctly update the size metadata.
ASSERT(partitionPageGetRawSize(page) == 0);
ALWAYS_INLINE void partitionFree(void* ptr)
ptr = partitionCookieFreePointerAdjust(ptr);
PartitionPage* page = partitionPointerToPage(ptr);
partitionFreeWithPage(ptr, page);
ALWAYS_INLINE PartitionBucket* partitionGenericSizeToBucket(PartitionRootGeneric* root, size_t size)
size_t order = kBitsPerSizet - countLeadingZerosSizet(size);
// The order index is simply the next few bits after the most significant bit.
size_t orderIndex = (size >> root->orderIndexShifts[order]) & (kGenericNumBucketsPerOrder - 1);
// And if the remaining bits are non-zero we must bump the bucket up.
size_t subOrderIndex = size & root->orderSubIndexMasks[order];
PartitionBucket* bucket = root->bucketLookups[(order << kGenericNumBucketsPerOrderBits) + orderIndex + !!subOrderIndex];
ASSERT(!bucket->slotSize || bucket->slotSize >= size);
ASSERT(!(bucket->slotSize % kGenericSmallestBucket));
return bucket;
ALWAYS_INLINE void* partitionAllocGenericFlags(PartitionRootGeneric* root, int flags, size_t size)
void* result = malloc(size);
return result;
size = partitionCookieSizeAdjustAdd(size);
PartitionBucket* bucket = partitionGenericSizeToBucket(root, size);
void* ret = partitionBucketAlloc(root, flags, size, bucket);
return ret;
ALWAYS_INLINE void* partitionAllocGeneric(PartitionRootGeneric* root, size_t size)
return partitionAllocGenericFlags(root, 0, size);
ALWAYS_INLINE void partitionFreeGeneric(PartitionRootGeneric* root, void* ptr)
if (UNLIKELY(!ptr))
ptr = partitionCookieFreePointerAdjust(ptr);
PartitionPage* page = partitionPointerToPage(ptr);
partitionFreeWithPage(ptr, page);
ALWAYS_INLINE size_t partitionDirectMapSize(size_t size)
// Caller must check that the size is not above the kGenericMaxDirectMapped
// limit before calling. This also guards against integer overflow in the
// calculation here.
ASSERT(size <= kGenericMaxDirectMapped);
return (size + kSystemPageOffsetMask) & kSystemPageBaseMask;
ALWAYS_INLINE size_t partitionAllocActualSize(PartitionRootGeneric* root, size_t size)
return size;
size = partitionCookieSizeAdjustAdd(size);
PartitionBucket* bucket = partitionGenericSizeToBucket(root, size);
if (LIKELY(!partitionBucketIsDirectMapped(bucket))) {
size = bucket->slotSize;
} else if (size > kGenericMaxDirectMapped) {
// Too large to allocate => return the size unchanged.
} else {
ASSERT(bucket == &PartitionRootBase::gPagedBucket);
size = partitionDirectMapSize(size);
return partitionCookieSizeAdjustSubtract(size);
ALWAYS_INLINE bool partitionAllocSupportsGetSize()
return false;
return true;
ALWAYS_INLINE size_t partitionAllocGetSize(void* ptr)
// No need to lock here. Only 'ptr' being freed by another thread could
// cause trouble, and the caller is responsible for that not happening.
ptr = partitionCookieFreePointerAdjust(ptr);
PartitionPage* page = partitionPointerToPage(ptr);
size_t size = page->bucket->slotSize;
return partitionCookieSizeAdjustSubtract(size);
// N (or more accurately, N - sizeof(void*)) represents the largest size in
// bytes that will be handled by a SizeSpecificPartitionAllocator.
// Attempts to partitionAlloc() more than this amount will fail.
template <size_t N>
class SizeSpecificPartitionAllocator {
static const size_t kMaxAllocation = N - kAllocationGranularity;
static const size_t kNumBuckets = N / kAllocationGranularity;
void init() { partitionAllocInit(&m_partitionRoot, kNumBuckets, kMaxAllocation); }
bool shutdown() { return partitionAllocShutdown(&m_partitionRoot); }
ALWAYS_INLINE PartitionRoot* root() { return &m_partitionRoot; }
PartitionRoot m_partitionRoot;
PartitionBucket m_actualBuckets[kNumBuckets];
class PartitionAllocatorGeneric {
void init() { partitionAllocGenericInit(&m_partitionRoot); }
bool shutdown() { return partitionAllocGenericShutdown(&m_partitionRoot); }
ALWAYS_INLINE PartitionRootGeneric* root() { return &m_partitionRoot; }
PartitionRootGeneric m_partitionRoot;
} // namespace WTF
using WTF::SizeSpecificPartitionAllocator;
using WTF::PartitionAllocatorGeneric;
using WTF::PartitionRoot;
using WTF::partitionAllocInit;
using WTF::partitionAllocShutdown;
using WTF::partitionAlloc;
using WTF::partitionFree;
using WTF::partitionAllocGeneric;
using WTF::partitionFreeGeneric;
using WTF::partitionReallocGeneric;
using WTF::partitionAllocActualSize;
using WTF::partitionAllocSupportsGetSize;
using WTF::partitionAllocGetSize;
#endif // WTF_PartitionAlloc_h