Avoid issuing write barriers for unboxed double fields in Heap::CopyJSObject().

Avoid issuing write barriers for unboxed double fields in Heap::CopyJSObject().

Committed: https://crrev.com/ba8409d2f66ee56fd18198019591f49c9c675244
Cr-Commit-Position: refs/heads/master@{#26394}

[Igor Sheludko, 2015-02-02 18:34:09]

Patchset #1 (id:1) has been deleted

[Igor Sheludko, 2015-02-02 18:34:44]

ishell@chromium.org changed reviewers:
+ hpayer@chromium.org

[Igor Sheludko, 2015-02-02 18:34:45]

PTAL

[Hannes Payer (out of office), 2015-02-02 21:00:54]

LGTM, please simplify the test.

https://codereview.chromium.org/880043003/diff/20001/test/cctest/test-unboxed...
File test/cctest/test-unboxed-doubles.cc (right):

https://codereview.chromium.org/880043003/diff/20001/test/cctest/test-unboxed...
test/cctest/test-unboxed-doubles.cc:1237: // Increase the chance of
'bump-the-pointer' allocation in old space.
I do not get this line. Why does that increase the chance of bump-pointer
allocation?

https://codereview.chromium.org/880043003/diff/20001/test/cctest/test-unboxed...
test/cctest/test-unboxed-doubles.cc:1270: while ((*limit_addr - *top_addr) >
allocation_amount) {
Just create a filler here, that's easier. Moreover, we should make sure that new
space is not growing. Just set the max_semi_space_size=1.

[Igor Sheludko, 2015-02-03 09:14:21]

https://codereview.chromium.org/880043003/diff/20001/test/cctest/test-unboxed...
File test/cctest/test-unboxed-doubles.cc (right):

https://codereview.chromium.org/880043003/diff/20001/test/cctest/test-unboxed...
test/cctest/test-unboxed-doubles.cc:1237: // Increase the chance of
'bump-the-pointer' allocation in old space.
On 2015/02/02 21:00:54, Hannes Payer wrote:
> I do not get this line. Why does that increase the chance of bump-pointer
> allocation?

Sorry, that was a copy/paste. Done.

https://codereview.chromium.org/880043003/diff/20001/test/cctest/test-unboxed...
test/cctest/test-unboxed-doubles.cc:1270: while ((*limit_addr - *top_addr) >
allocation_amount) {
On 2015/02/02 21:00:54, Hannes Payer wrote:
> Just create a filler here, that's easier. Moreover, we should make sure that
new
> space is not growing. Just set the max_semi_space_size=1.

Thanks for the max_semi_space_size suggestion! 
Unfortunately Heap::AllocateFillerObject() is a private method and I'm not sure
if we should change that, so I'll keep allocating arrays here.

[Igor Sheludko, 2015-02-03 09:58:45]

The CQ bit was checked by ishell@chromium.org

[commit-bot: I haz the power, 2015-02-03 09:59:44]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/880043003/40001

[commit-bot: I haz the power, 2015-02-03 10:28:26]

Committed patchset #2 (id:40001)

[commit-bot: I haz the power, 2015-02-03 10:28:45]

Patchset 2 (id:??) landed as
https://crrev.com/ba8409d2f66ee56fd18198019591f49c9c675244
Cr-Commit-Position: refs/heads/master@{#26394}