Simplified large object allocation strategy.

Simplified large object allocation strategy.

BUG=
R=mstarzinger@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=15840

[Michael Starzinger, 2013-07-22 19:27:44]

LGTM (with comments). Mostly style nits. Just one real comment about the
assertion in MacroAssembler::Allocate.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc
File src/heap.cc (right):

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5356
src/heap.cc:5356: 
nit: Let's drop this empty new-line for consistency.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5359
src/heap.cc:5359: 
nit: Let's drop this empty new-line for consistency.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5542
src/heap.cc:5542: 
nit: Let's drop this empty new-line for consistency.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5670
src/heap.cc:5670: 
nit: Let's drop this empty new-line for consistency.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5672
src/heap.cc:5672: (pretenure == TENURED) ? OLD_DATA_SPACE : NEW_SPACE;
nit: Should fit into one line.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5675
src/heap.cc:5675: int size = FixedDoubleArray::SizeFor(length);
nit: Let's move this line (i.e. the size calculation) to before the assignment
of "space" for consistency.

https://codereview.chromium.org/19934006/diff/1/src/ia32/macro-assembler-ia32.cc
File src/ia32/macro-assembler-ia32.cc (right):

https://codereview.chromium.org/19934006/diff/1/src/ia32/macro-assembler-ia32...
src/ia32/macro-assembler-ia32.cc:1251: ASSERT(object_size <
Page::kMaxNonCodeHeapObjectSize);
This should actually be less-or-equal (i.e. "<=") here I guess? Applies to other
architectures as well.

https://codereview.chromium.org/19934006/diff/1/test/cctest/test-mark-compact.cc
File test/cctest/test-mark-compact.cc (right):

https://codereview.chromium.org/19934006/diff/1/test/cctest/test-mark-compact...
test/cctest/test-mark-compact.cc:123: Min(Page::kMaxNonCodeHeapObjectSize,
Page::kMaxNonCodeHeapObjectSize);
nit: This Min() looks kind of redundant now. I guess the left constant is
slightly smaller than the right one on Mondays and Wednesdays, but I think it's
safe to ignore that. :)

[Hannes Payer (out of office), 2013-07-23 20:00:51]

https://codereview.chromium.org/19934006/diff/1/src/heap.cc
File src/heap.cc (right):

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5356
src/heap.cc:5356: 
On 2013/07/22 19:27:50, Michael Starzinger wrote:
> nit: Let's drop this empty new-line for consistency.

Done.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5359
src/heap.cc:5359: 
On 2013/07/22 19:27:50, Michael Starzinger wrote:
> nit: Let's drop this empty new-line for consistency.

Done.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5542
src/heap.cc:5542: 
On 2013/07/22 19:27:50, Michael Starzinger wrote:
> nit: Let's drop this empty new-line for consistency.

Done.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5670
src/heap.cc:5670: 
On 2013/07/22 19:27:50, Michael Starzinger wrote:
> nit: Let's drop this empty new-line for consistency.

Done.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5672
src/heap.cc:5672: (pretenure == TENURED) ? OLD_DATA_SPACE : NEW_SPACE;
On 2013/07/22 19:27:50, Michael Starzinger wrote:
> nit: Should fit into one line.

Done.

https://codereview.chromium.org/19934006/diff/1/src/heap.cc#newcode5675
src/heap.cc:5675: int size = FixedDoubleArray::SizeFor(length);
On 2013/07/22 19:27:50, Michael Starzinger wrote:
> nit: Let's move this line (i.e. the size calculation) to before the assignment
> of "space" for consistency.

Done.

https://codereview.chromium.org/19934006/diff/1/src/ia32/macro-assembler-ia32.cc
File src/ia32/macro-assembler-ia32.cc (right):

https://codereview.chromium.org/19934006/diff/1/src/ia32/macro-assembler-ia32...
src/ia32/macro-assembler-ia32.cc:1251: ASSERT(object_size <
Page::kMaxNonCodeHeapObjectSize);
On 2013/07/22 19:27:50, Michael Starzinger wrote:
> This should actually be less-or-equal (i.e. "<=") here I guess? Applies to
other
> architectures as well.

Done.

https://codereview.chromium.org/19934006/diff/1/test/cctest/test-mark-compact.cc
File test/cctest/test-mark-compact.cc (right):

https://codereview.chromium.org/19934006/diff/1/test/cctest/test-mark-compact...
test/cctest/test-mark-compact.cc:123: Min(Page::kMaxNonCodeHeapObjectSize,
Page::kMaxNonCodeHeapObjectSize);
On 2013/07/22 19:27:50, Michael Starzinger wrote:
> nit: This Min() looks kind of redundant now. I guess the left constant is
> slightly smaller than the right one on Mondays and Wednesdays, but I think
it's
> safe to ignore that. :)

Done.

[Hannes Payer (out of office), 2013-07-23 20:01:46]

Committed patchset #2 manually as r15840 (presubmit successful).