Bump MaxRegularSpaceAllocationSize to InitialSemiSpaceSize() * 4/5

Bump MaxRegularSpaceAllocationSize to InitialSemiSpaceSize() * 4/5

to allow allocation of large packed arrays in paged spaces.

BUG=v8:2790
R=mstarzinger@chromium.org

Committed: http://code.google.com/p/v8/source/detail?r=16475

[Jakob Kummerow, 2013-09-02 13:06:28]

As discussed. PTAL.

[Michael Starzinger, 2013-09-02 13:12:43]

LGTM with comments.

https://codereview.chromium.org/23604023/diff/1/src/heap.cc
File src/heap.cc (right):

https://codereview.chromium.org/23604023/diff/1/src/heap.cc#newcode195
src/heap.cc:195: ASSERT(MaxRegularSpaceAllocationSize() >=
The initial_semispace_size might change at runtime, so this assert should go at
the end of Heap::ConfigureHeap I think.

https://codereview.chromium.org/23604023/diff/1/src/heap.cc#newcode198
src/heap.cc:198: JSObject::kInitialMaxFastElementArray * kPointerSize +
nit: Lets use FixedArray::SizeFor for (part of) this calculation.

[Jakob Kummerow, 2013-09-02 13:35:10]

Thanks for the review, landing.

https://codereview.chromium.org/23604023/diff/1/src/heap.cc
File src/heap.cc (right):

https://codereview.chromium.org/23604023/diff/1/src/heap.cc#newcode195
src/heap.cc:195: ASSERT(MaxRegularSpaceAllocationSize() >=
On 2013/09/02 13:12:43, Michael Starzinger wrote:
> The initial_semispace_size might change at runtime, so this assert should go
at
> the end of Heap::ConfigureHeap I think.

Good point. Done.

https://codereview.chromium.org/23604023/diff/1/src/heap.cc#newcode198
src/heap.cc:198: JSObject::kInitialMaxFastElementArray * kPointerSize +
On 2013/09/02 13:12:43, Michael Starzinger wrote:
> nit: Lets use FixedArray::SizeFor for (part of) this calculation.

Done.

[Jakob Kummerow, 2013-09-02 13:37:08]

Committed patchset #2 manually as r16475 (presubmit successful).