v8/src/spaces-inl.h

365 lines
11 KiB
C
Raw Normal View History

// Copyright 2011 the V8 project authors. 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.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef V8_SPACES_INL_H_
#define V8_SPACES_INL_H_
#include "isolate.h"
#include "spaces.h"
#include "v8memory.h"
namespace v8 {
namespace internal {
// -----------------------------------------------------------------------------
// Bitmap
void Bitmap::Clear(MemoryChunk* chunk) {
Bitmap* bitmap = chunk->markbits();
for (int i = 0; i < bitmap->CellsCount(); i++) bitmap->cells()[i] = 0;
chunk->ResetLiveBytes();
}
// -----------------------------------------------------------------------------
// PageIterator
PageIterator::PageIterator(PagedSpace* space)
: space_(space),
prev_page_(&space->anchor_),
next_page_(prev_page_->next_page()) { }
bool PageIterator::has_next() {
return next_page_ != &space_->anchor_;
}
Page* PageIterator::next() {
ASSERT(has_next());
prev_page_ = next_page_;
next_page_ = next_page_->next_page();
return prev_page_;
}
// -----------------------------------------------------------------------------
// NewSpacePageIterator
NewSpacePageIterator::NewSpacePageIterator(NewSpace* space)
: prev_page_(NewSpacePage::FromAddress(space->ToSpaceStart())->prev_page()),
next_page_(NewSpacePage::FromAddress(space->ToSpaceStart())),
last_page_(NewSpacePage::FromLimit(space->ToSpaceEnd())) { }
NewSpacePageIterator::NewSpacePageIterator(SemiSpace* space)
: prev_page_(space->anchor()),
next_page_(prev_page_->next_page()),
last_page_(prev_page_->prev_page()) { }
NewSpacePageIterator::NewSpacePageIterator(Address start, Address limit)
: prev_page_(NewSpacePage::FromAddress(start)->prev_page()),
next_page_(NewSpacePage::FromAddress(start)),
last_page_(NewSpacePage::FromLimit(limit)) {
SemiSpace::AssertValidRange(start, limit);
}
bool NewSpacePageIterator::has_next() {
return prev_page_ != last_page_;
}
NewSpacePage* NewSpacePageIterator::next() {
ASSERT(has_next());
prev_page_ = next_page_;
next_page_ = next_page_->next_page();
return prev_page_;
}
// -----------------------------------------------------------------------------
// HeapObjectIterator
HeapObject* HeapObjectIterator::FromCurrentPage() {
while (cur_addr_ != cur_end_) {
if (cur_addr_ == space_->top() && cur_addr_ != space_->limit()) {
cur_addr_ = space_->limit();
continue;
}
HeapObject* obj = HeapObject::FromAddress(cur_addr_);
int obj_size = (size_func_ == NULL) ? obj->Size() : size_func_(obj);
cur_addr_ += obj_size;
ASSERT(cur_addr_ <= cur_end_);
if (!obj->IsFiller()) {
ASSERT_OBJECT_SIZE(obj_size);
return obj;
}
}
return NULL;
}
// -----------------------------------------------------------------------------
// MemoryAllocator
#ifdef ENABLE_HEAP_PROTECTION
void MemoryAllocator::Protect(Address start, size_t size) {
OS::Protect(start, size);
}
void MemoryAllocator::Unprotect(Address start,
size_t size,
Executability executable) {
OS::Unprotect(start, size, executable);
}
void MemoryAllocator::ProtectChunkFromPage(Page* page) {
int id = GetChunkId(page);
OS::Protect(chunks_[id].address(), chunks_[id].size());
}
void MemoryAllocator::UnprotectChunkFromPage(Page* page) {
int id = GetChunkId(page);
OS::Unprotect(chunks_[id].address(), chunks_[id].size(),
chunks_[id].owner()->executable() == EXECUTABLE);
}
#endif
// --------------------------------------------------------------------------
// PagedSpace
Page* Page::Initialize(Heap* heap,
MemoryChunk* chunk,
Executability executable,
PagedSpace* owner) {
Page* page = reinterpret_cast<Page*>(chunk);
ASSERT(page->area_size() <= kNonCodeObjectAreaSize);
ASSERT(chunk->owner() == owner);
owner->IncreaseCapacity(page->area_size());
owner->Free(page->area_start(), page->area_size());
heap->incremental_marking()->SetOldSpacePageFlags(chunk);
return page;
}
bool PagedSpace::Contains(Address addr) {
Page* p = Page::FromAddress(addr);
if (!p->is_valid()) return false;
return p->owner() == this;
}
void MemoryChunk::set_scan_on_scavenge(bool scan) {
if (scan) {
if (!scan_on_scavenge()) heap_->increment_scan_on_scavenge_pages();
SetFlag(SCAN_ON_SCAVENGE);
} else {
if (scan_on_scavenge()) heap_->decrement_scan_on_scavenge_pages();
ClearFlag(SCAN_ON_SCAVENGE);
}
heap_->incremental_marking()->SetOldSpacePageFlags(this);
}
MemoryChunk* MemoryChunk::FromAnyPointerAddress(Address addr) {
MemoryChunk* maybe = reinterpret_cast<MemoryChunk*>(
OffsetFrom(addr) & ~Page::kPageAlignmentMask);
if (maybe->owner() != NULL) return maybe;
LargeObjectIterator iterator(HEAP->lo_space());
for (HeapObject* o = iterator.Next(); o != NULL; o = iterator.Next()) {
// Fixed arrays are the only pointer-containing objects in large object
// space.
if (o->IsFixedArray()) {
MemoryChunk* chunk = MemoryChunk::FromAddress(o->address());
if (chunk->Contains(addr)) {
return chunk;
}
}
}
UNREACHABLE();
return NULL;
}
void MemoryChunk::UpdateHighWaterMark(Address mark) {
if (mark == NULL) return;
// Need to subtract one from the mark because when a chunk is full the
// top points to the next address after the chunk, which effectively belongs
// to another chunk. See the comment to Page::FromAllocationTop.
MemoryChunk* chunk = MemoryChunk::FromAddress(mark - 1);
int new_mark = static_cast<int>(mark - chunk->address());
if (new_mark > chunk->high_water_mark_) {
chunk->high_water_mark_ = new_mark;
}
}
PointerChunkIterator::PointerChunkIterator(Heap* heap)
: state_(kOldPointerState),
old_pointer_iterator_(heap->old_pointer_space()),
map_iterator_(heap->map_space()),
lo_iterator_(heap->lo_space()) { }
Page* Page::next_page() {
ASSERT(next_chunk()->owner() == owner());
return static_cast<Page*>(next_chunk());
}
Page* Page::prev_page() {
ASSERT(prev_chunk()->owner() == owner());
return static_cast<Page*>(prev_chunk());
}
void Page::set_next_page(Page* page) {
ASSERT(page->owner() == owner());
set_next_chunk(page);
}
void Page::set_prev_page(Page* page) {
ASSERT(page->owner() == owner());
set_prev_chunk(page);
}
// Try linear allocation in the page of alloc_info's allocation top. Does
// not contain slow case logic (e.g. move to the next page or try free list
// allocation) so it can be used by all the allocation functions and for all
// the paged spaces.
HeapObject* PagedSpace::AllocateLinearly(int size_in_bytes) {
Address current_top = allocation_info_.top;
Address new_top = current_top + size_in_bytes;
if (new_top > allocation_info_.limit) return NULL;
allocation_info_.top = new_top;
return HeapObject::FromAddress(current_top);
}
// Raw allocation.
MaybeObject* PagedSpace::AllocateRaw(int size_in_bytes) {
HeapObject* object = AllocateLinearly(size_in_bytes);
if (object != NULL) {
if (identity() == CODE_SPACE) {
SkipList::Update(object->address(), size_in_bytes);
}
return object;
}
Refactoring of snapshots. This simplifies and improves the speed of deserializing code. The current startup time improvement for V8 is around 6%, but code deserialization is speeded up disproportionately, and we will soon have more code in the snapshot. * Removed support for deserializing into large object space. The regular pages are 1Mbyte now and that is plenty. This is a big simplification. * Instead of reserving space for the snapshot we actually allocate it now. This removes some special casing from the memory management and simplifies deserialization since we are just bumping a pointer rather than calling the normal allocation routines during deserialization. * Record in the snapshot how much we need to boot up and allocate it instead of just assuming that allocations in a new VM will always be linear. * In the snapshot we always address an object as a negative offset from the current allocation point. We used to sometimes address from the start of the deserialized data, but this is less useful now that we have good support for roots and repetitions in the deserialization data. * Code objects were previously deserialized (like other objects) by alternating raw data (deserialized with memcpy) and pointers (to external references, other objects, etc.). Now we deserialize code objects with a single memcpy, followed by a series of skips and pointers that partially overwrite the code we memcopied out of the snapshot. The skips are sometimes merged into the following instruction in the deserialization data to reduce dispatch time. * Integers in the snapshot were stored in a variable length format that gives a compact representation for small positive integers. This is still the case, but the new encoding can be decoded without branches or conditional instructions, which is faster on a modern CPU. Review URL: https://chromiumcodereview.appspot.com/10918067 git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@12505 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2012-09-14 11:16:56 +00:00
ASSERT(!heap()->linear_allocation() ||
(anchor_.next_chunk() == &anchor_ &&
anchor_.prev_chunk() == &anchor_));
object = free_list_.Allocate(size_in_bytes);
if (object != NULL) {
if (identity() == CODE_SPACE) {
SkipList::Update(object->address(), size_in_bytes);
}
return object;
}
object = SlowAllocateRaw(size_in_bytes);
if (object != NULL) {
if (identity() == CODE_SPACE) {
SkipList::Update(object->address(), size_in_bytes);
}
return object;
}
return Failure::RetryAfterGC(identity());
}
// -----------------------------------------------------------------------------
// NewSpace
MaybeObject* NewSpace::AllocateRaw(int size_in_bytes) {
Address old_top = allocation_info_.top;
#ifdef DEBUG
// If we are stressing compaction we waste some memory in new space
// in order to get more frequent GCs.
if (FLAG_stress_compaction && !HEAP->linear_allocation()) {
if (allocation_info_.limit - old_top >= size_in_bytes * 4) {
int filler_size = size_in_bytes * 4;
for (int i = 0; i < filler_size; i += kPointerSize) {
*(reinterpret_cast<Object**>(old_top + i)) =
HEAP->one_pointer_filler_map();
}
old_top += filler_size;
allocation_info_.top += filler_size;
}
}
#endif
if (allocation_info_.limit - old_top < size_in_bytes) {
return SlowAllocateRaw(size_in_bytes);
}
Object* obj = HeapObject::FromAddress(old_top);
allocation_info_.top += size_in_bytes;
ASSERT_SEMISPACE_ALLOCATION_INFO(allocation_info_, to_space_);
return obj;
}
LargePage* LargePage::Initialize(Heap* heap, MemoryChunk* chunk) {
heap->incremental_marking()->SetOldSpacePageFlags(chunk);
return static_cast<LargePage*>(chunk);
}
intptr_t LargeObjectSpace::Available() {
return ObjectSizeFor(heap()->isolate()->memory_allocator()->Available());
}
bool FreeListNode::IsFreeListNode(HeapObject* object) {
Map* map = object->map();
Heap* heap = object->GetHeap();
return map == heap->raw_unchecked_free_space_map()
|| map == heap->raw_unchecked_one_pointer_filler_map()
|| map == heap->raw_unchecked_two_pointer_filler_map();
}
} } // namespace v8::internal
#endif // V8_SPACES_INL_H_