Reland "[heap] Fine-grained JSArrayBuffer tracking"

Track based on JSArrayBuffer addresses on pages instead of the attached
backing store.

Details of tracking:
- Scavenge: New space pages are processes in bulk on the main thread
- MC: Unswept pages are processed in bulk in parallel. All other pages
  are processed by the sweeper concurrently.

BUG=chromium:611688
LOG=N
TEST=cctest/test-array-buffer-tracker/*
CQ_EXTRA_TRYBOTS=tryserver.v8:v8_linux_arm64_gc_stress_dbg,v8_linux_gc_stress_dbg,v8_mac_gc_stress_dbg,v8_linux64_tsan_rel,v8_mac64_asan_rel

Committed: https://crrev.com/839f3fd406426a221d74eb7a33a72794c3c7a548
Cr-Commit-Position: refs/heads/master@{#36798}

[Michael Lippautz, 2016-06-02 16:47:27]

Description was changed from

==========
Reland "[heap] Fine-grained JSArrayBuffer tracking"

BUG=
==========

to

==========
Reland "[heap] Fine-grained JSArrayBuffer tracking"

BUG=
==========

[Michael Lippautz, 2016-06-02 16:47:27]

mlippautz@chromium.org changed reviewers:
+ hpayer@chromium.org, ulan@chromium.org

[Michael Lippautz, 2016-06-02 16:48:51]

ptal; +ulan

- I removed the atomic variable for the tracker. It's now all guarded by the
page lock.
- More DCHECKs
- Destroying the local tracker checks for an empty set. This means that for
TearDown we need to explicitly finalize the buffers.

[Michael Lippautz, 2016-06-02 20:30:48]

Patchset #3 (id:40001) has been deleted

[Hannes Payer (out of office), 2016-06-03 07:43:16]

https://codereview.chromium.org/2036643002/diff/60001/src/heap/spaces.cc
File src/heap/spaces.cc (right):

https://codereview.chromium.org/2036643002/diff/60001/src/heap/spaces.cc#newc...
src/heap/spaces.cc:1091: PageIterator iterator(this);
Factor that out into a function.

https://codereview.chromium.org/2036643002/diff/60001/src/heap/spaces.cc#newc...
src/heap/spaces.cc:1719: NewSpacePageIterator it(this);
Factor that out into a function. It should reuse the implementation from above.

[Michael Lippautz, 2016-06-03 08:07:36]

I refactored a bit more so that we can bring back the DCHECK for sweeping which
I think is a good sanity check. Also, the additional mode for Free() now means
that there's no need to reset the markbits during tear down.

https://codereview.chromium.org/2036643002/diff/60001/src/heap/spaces.cc
File src/heap/spaces.cc (right):

https://codereview.chromium.org/2036643002/diff/60001/src/heap/spaces.cc#newc...
src/heap/spaces.cc:1091: PageIterator iterator(this);
On 2016/06/03 07:43:15, Hannes Payer wrote:
> Factor that out into a function.

Done.

https://codereview.chromium.org/2036643002/diff/60001/src/heap/spaces.cc#newc...
src/heap/spaces.cc:1719: NewSpacePageIterator it(this);
On 2016/06/03 07:43:15, Hannes Payer wrote:
> Factor that out into a function. It should reuse the implementation from
above.

Done.

[ulan, 2016-06-03 11:31:32]

https://codereview.chromium.org/2036643002/diff/80001/src/heap/array-buffer-t...
File src/heap/array-buffer-tracker.cc (right):

https://codereview.chromium.org/2036643002/diff/80001/src/heap/array-buffer-t...
src/heap/array-buffer-tracker.cc:65: Page* target_page =
Page::FromAddress(new_buffer->address());
Don't we need to lock the target page here?

[Michael Lippautz, 2016-06-07 09:21:40]

Patchset #5 (id:100001) has been deleted

[Michael Lippautz, 2016-06-07 09:52:05]

Patchset #5 (id:120001) has been deleted

[Michael Lippautz, 2016-06-07 09:54:29]

PTAL

https://codereview.chromium.org/2036643002/diff/80001/src/heap/array-buffer-t...
File src/heap/array-buffer-tracker.cc (right):

https://codereview.chromium.org/2036643002/diff/80001/src/heap/array-buffer-t...
src/heap/array-buffer-tracker.cc:65: Page* target_page =
Page::FromAddress(new_buffer->address());
On 2016/06/03 11:31:32, ulan wrote:
> Don't we need to lock the target page here?

Thanks a lot!

For old space pages we have exclusive access but we also use this one during
parallel semispace copy in new space.

[Michael Lippautz, 2016-06-07 09:55:04]

Description was changed from

==========
Reland "[heap] Fine-grained JSArrayBuffer tracking"

BUG=
==========

to

==========
Track based on JSArrayBuffer addresses on pages instead of the attached
backing store.

Details of tracking:
- Scavenge: New space pages are processes in bulk on the main thread
- MC: Unswept pages are processed in bulk in parallel. All other pages
  are processed by the sweeper concurrently.

BUG=chromium:611688
LOG=N
TEST=cctest/test-array-buffer-tracker/*
CQ_EXTRA_TRYBOTS=tryserver.v8:v8_linux_arm64_gc_stress_dbg,v8_linux_gc_stress_dbg,v8_mac_gc_stress_dbg,v8_linux64_tsan_rel,v8_mac64_asan_rel
==========

[Hannes Payer (out of office), 2016-06-07 12:17:05]

LGTM if performance is good

[Michael Lippautz, 2016-06-07 13:35:58]

The CQ bit was checked by mlippautz@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-07 13:36:06]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2036643002/180001

[commit-bot: I haz the power, 2016-06-07 13:38:56]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-07 13:38:57]

Dry run: Try jobs failed on following builders:
  v8_linux64_avx2_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_avx2_rel_ng/buil...)
  v8_mac_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/2760)

[Michael Lippautz, 2016-06-07 13:41:35]

Patchset #7 (id:180001) has been deleted

[Michael Lippautz, 2016-06-07 13:41:52]

The CQ bit was checked by mlippautz@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-07 13:41:57]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2036643002/200001

[commit-bot: I haz the power, 2016-06-07 14:06:07]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-07 14:06:08]

Dry run: Try jobs failed on following builders:
  v8_linux_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_rel_ng/builds/6842)
  v8_linux_rel_ng_triggered on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_rel_ng_triggered/b...)

[Michael Lippautz, 2016-06-07 14:07:50]

The CQ bit was checked by mlippautz@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-07 14:07:57]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2036643002/220001

[commit-bot: I haz the power, 2016-06-07 14:36:18]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-07 14:36:19]

Dry run: Try jobs failed on following builders:
  v8_linux64_tsan_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_tsan_rel/builds/262)

[Michael Lippautz, 2016-06-07 16:22:25]

The CQ bit was checked by mlippautz@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-07 16:33:09]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2036643002/240001

[commit-bot: I haz the power, 2016-06-07 17:15:33]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-07 17:15:35]

Dry run: This issue passed the CQ dry run.

[ulan, 2016-06-07 17:20:23]

lgtm

[Michael Lippautz, 2016-06-07 17:20:41]

The CQ bit was checked by mlippautz@chromium.org

[Michael Lippautz, 2016-06-07 17:20:41]

The patchset sent to the CQ was uploaded after l-g-t-m from hpayer@chromium.org
Link to the patchset: https://codereview.chromium.org/2036643002/#ps240001
(title: "Remove unneeded locking to avoid lock-inversion-order errors in TSAN")

[commit-bot: I haz the power, 2016-06-07 17:20:56]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2036643002/240001

[commit-bot: I haz the power, 2016-06-07 17:28:39]

Description was changed from

==========
Track based on JSArrayBuffer addresses on pages instead of the attached
backing store.

Details of tracking:
- Scavenge: New space pages are processes in bulk on the main thread
- MC: Unswept pages are processed in bulk in parallel. All other pages
  are processed by the sweeper concurrently.

BUG=chromium:611688
LOG=N
TEST=cctest/test-array-buffer-tracker/*
CQ_EXTRA_TRYBOTS=tryserver.v8:v8_linux_arm64_gc_stress_dbg,v8_linux_gc_stress_dbg,v8_mac_gc_stress_dbg,v8_linux64_tsan_rel,v8_mac64_asan_rel
==========

to

==========
Track based on JSArrayBuffer addresses on pages instead of the attached
backing store.

Details of tracking:
- Scavenge: New space pages are processes in bulk on the main thread
- MC: Unswept pages are processed in bulk in parallel. All other pages
  are processed by the sweeper concurrently.

BUG=chromium:611688
LOG=N
TEST=cctest/test-array-buffer-tracker/*
CQ_EXTRA_TRYBOTS=tryserver.v8:v8_linux_arm64_gc_stress_dbg,v8_linux_gc_stress_dbg,v8_mac_gc_stress_dbg,v8_linux64_tsan_rel,v8_mac64_asan_rel
==========

[commit-bot: I haz the power, 2016-06-07 17:28:40]

Committed patchset #9 (id:240001)

[commit-bot: I haz the power, 2016-06-07 17:29:50]

Description was changed from

==========
Track based on JSArrayBuffer addresses on pages instead of the attached
backing store.

Details of tracking:
- Scavenge: New space pages are processes in bulk on the main thread
- MC: Unswept pages are processed in bulk in parallel. All other pages
  are processed by the sweeper concurrently.

BUG=chromium:611688
LOG=N
TEST=cctest/test-array-buffer-tracker/*
CQ_EXTRA_TRYBOTS=tryserver.v8:v8_linux_arm64_gc_stress_dbg,v8_linux_gc_stress_dbg,v8_mac_gc_stress_dbg,v8_linux64_tsan_rel,v8_mac64_asan_rel
==========

to

==========
Track based on JSArrayBuffer addresses on pages instead of the attached
backing store.

Details of tracking:
- Scavenge: New space pages are processes in bulk on the main thread
- MC: Unswept pages are processed in bulk in parallel. All other pages
  are processed by the sweeper concurrently.

BUG=chromium:611688
LOG=N
TEST=cctest/test-array-buffer-tracker/*
CQ_EXTRA_TRYBOTS=tryserver.v8:v8_linux_arm64_gc_stress_dbg,v8_linux_gc_stress_dbg,v8_mac_gc_stress_dbg,v8_linux64_tsan_rel,v8_mac64_asan_rel

Committed: https://crrev.com/839f3fd406426a221d74eb7a33a72794c3c7a548
Cr-Commit-Position: refs/heads/master@{#36798}
==========

[commit-bot: I haz the power, 2016-06-07 17:29:51]

Patchset 9 (id:??) landed as
https://crrev.com/839f3fd406426a221d74eb7a33a72794c3c7a548
Cr-Commit-Position: refs/heads/master@{#36798}