[heap] New Dijkstra marking write barrier.

[heap] New Dijkstra marking write barrier.

A.x = B
Change from mark grey A to mark grey B.

BUG=

Committed: https://crrev.com/1c5df4fb62e11d293f96c8573e417a29f6d3a39e
Cr-Commit-Position: refs/heads/master@{#32765}

[Hannes Payer (out of office), 2015-11-05 18:58:32]

Description was changed from

==========
[heap] New write barrier.

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

to

==========
[heap] New Dijkstra write barrier.

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

[Hannes Payer (out of office), 2015-11-05 18:58:32]

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

[Hannes Payer (out of office), 2015-11-05 23:10:52]

Description was changed from

==========
[heap] New Dijkstra write barrier.

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

to

==========
[heap] New Dijkstra write barrier.

A.x = B
Change from mark grey A to mark grey B.


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

[Hannes Payer (out of office), 2015-11-05 23:11:03]

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

[ulan, 2015-11-06 10:54:46]

https://codereview.chromium.org/1409813007/diff/60001/src/heap/incremental-ma...
File src/heap/incremental-marking.cc (right):

https://codereview.chromium.org/1409813007/diff/60001/src/heap/incremental-ma...
src/heap/incremental-marking.cc:66: // Object may not be rescanned we need to
record the slot.
With the new barrier a black object will not be rescanned for sure.

https://codereview.chromium.org/1409813007/diff/60001/src/heap/incremental-ma...
src/heap/incremental-marking.cc:141: BlackToGreyAndUnshift(obj, obj_bit);
Are you leaving this unchanged intentionally?

[Hannes Payer (out of office), 2015-11-10 20:23:11]

https://codereview.chromium.org/1409813007/diff/60001/src/heap/incremental-ma...
File src/heap/incremental-marking.cc (right):

https://codereview.chromium.org/1409813007/diff/60001/src/heap/incremental-ma...
src/heap/incremental-marking.cc:66: // Object may not be rescanned we need to
record the slot.
On 2015/11/06 10:54:45, ulan wrote:
> With the new barrier a black object will not be rescanned for sure.

Yes! Done!

https://codereview.chromium.org/1409813007/diff/60001/src/heap/incremental-ma...
src/heap/incremental-marking.cc:141: BlackToGreyAndUnshift(obj, obj_bit);
On 2015/11/06 10:54:45, ulan wrote:
> Are you leaving this unchanged intentionally?

I did not change this one because it is used in a way where we explicitly want
to revisit the object. Overall, I do not like this mechanism. Maybe I should go
to all the call sites of that function and re-evaluate its use.

[Hannes Payer (out of office), 2015-11-10 20:35:15]

Description was changed from

==========
[heap] New Dijkstra write barrier.

A.x = B
Change from mark grey A to mark grey B.


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

to

==========
[heap] New Dijkstra marking write barrier.

A.x = B
Change from mark grey A to mark grey B.


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

[Michael Lippautz, 2015-11-10 20:50:28]

lgtm from my side

https://codereview.chromium.org/1409813007/diff/80001/src/heap/incremental-ma...
File src/heap/incremental-marking.cc (right):

https://codereview.chromium.org/1409813007/diff/80001/src/heap/incremental-ma...
src/heap/incremental-marking.cc:65: if (BaseRecordWrite(obj, value) && slot !=
NULL) {
nit: add paren =>
  (slot != nullptr)

[ulan, 2015-11-11 10:24:10]

lgtm

> I did not change this one because it is used in a way where we explicitly want
to revisit the object.
I checked the call-site that relies on it: it looks like a hack :) Let's remove
it in future.

[Hannes Payer (out of office), 2015-12-04 10:50:13]

rebased and running try perf bots again

https://codereview.chromium.org/1409813007/diff/80001/src/heap/incremental-ma...
File src/heap/incremental-marking.cc (right):

https://codereview.chromium.org/1409813007/diff/80001/src/heap/incremental-ma...
src/heap/incremental-marking.cc:65: if (BaseRecordWrite(obj, value) && slot !=
NULL) {
On 2015/11/10 20:50:28, Michael Lippautz wrote:
> nit: add paren =>
>   (slot != nullptr)

not required by style guide

[Hannes Payer (out of office), 2015-12-10 11:33:14]

The CQ bit was checked by hpayer@chromium.org

[Hannes Payer (out of office), 2015-12-10 11:33:15]

The patchset sent to the CQ was uploaded after l-g-t-m from
mlippautz@chromium.org, ulan@chromium.org
Link to the patchset: https://codereview.chromium.org/1409813007/#ps100001
(title: " ")

[commit-bot: I haz the power, 2015-12-10 11:33:23]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1409813007/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1409813007/100001

[commit-bot: I haz the power, 2015-12-10 11:55:41]

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

[commit-bot: I haz the power, 2015-12-10 11:55:41]

Try jobs failed on following builders:
  v8_linux64_avx2_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_avx2_rel/builds/...)

[Hannes Payer (out of office), 2015-12-10 17:08:59]

The CQ bit was checked by hpayer@chromium.org

[commit-bot: I haz the power, 2015-12-10 17:09:06]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1409813007/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1409813007/100001

[commit-bot: I haz the power, 2015-12-10 17:28:56]

Description was changed from

==========
[heap] New Dijkstra marking write barrier.

A.x = B
Change from mark grey A to mark grey B.


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

to

==========
[heap] New Dijkstra marking write barrier.

A.x = B
Change from mark grey A to mark grey B.


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

[commit-bot: I haz the power, 2015-12-10 17:28:57]

Committed patchset #6 (id:100001)

[commit-bot: I haz the power, 2015-12-10 17:29:16]

Description was changed from

==========
[heap] New Dijkstra marking write barrier.

A.x = B
Change from mark grey A to mark grey B.


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

to

==========
[heap] New Dijkstra marking write barrier.

A.x = B
Change from mark grey A to mark grey B.

BUG=

Committed: https://crrev.com/1c5df4fb62e11d293f96c8573e417a29f6d3a39e
Cr-Commit-Position: refs/heads/master@{#32765}
==========

[commit-bot: I haz the power, 2015-12-10 17:29:17]

Patchset 6 (id:??) landed as
https://crrev.com/1c5df4fb62e11d293f96c8573e417a29f6d3a39e
Cr-Commit-Position: refs/heads/master@{#32765}