Optimize outline paint invalidation

Optimize outline paint invalidation

Shape of an outline can be affected by layout/style change of
descendants. Previously we assumed that an outline will change
when an object having outline is checked for paint invalidation
for any reason, which covered all cases of outline changes, but
also produced over invalidations, which is especially bad
when we move caret paint invalidation into paint invalidation
stage: a blinking caret will cause full paint invalidation of
the edit box having focus ring.

Now tighten the condition. If a box's outline is not affected
by descendants, don't do full paint invalidation.

TEST=paint/invalidation/outline-*.html
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2669003002
Cr-Commit-Position: refs/heads/master@{#448079}
Committed: https://chromium.googlesource.com/chromium/src/+/2b30e8534ceb11cd138fa1360e7ae6efc6b4c046

[Xianzhu, 2017-02-01 20:27:32]

Description was changed from

==========
Optimize outline paint invalidation

Shape of an outline can be affected by layout/style change of
descendants. Previously we assumed that an outline will change
when an object having outline is checked for paint invalidation
for any reason, which covered all cases of outline changes, but
also produced over invalidations, which is especially bad
when we move caret paint invalidation into paint invalidation
stage: a blinking caret will cause full paint invalidation of
the edit box having focus ring.

Now tighten the condition. If a box's outline is not affected
by descendants, don't do full paint invalidation.

TEST=paint/invalidation/outline-*.html
==========

to

==========
Optimize outline paint invalidation

Shape of an outline can be affected by layout/style change of
descendants. Previously we assumed that an outline will change
when an object having outline is checked for paint invalidation
for any reason, which covered all cases of outline changes, but
also produced over invalidations, which is especially bad
when we move caret paint invalidation into paint invalidation
stage: a blinking caret will cause full paint invalidation of
the edit box having focus ring.

Now tighten the condition. If a box's outline is not affected
by descendants, don't do full paint invalidation.

TEST=paint/invalidation/outline-*.html
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
==========

[Xianzhu, 2017-02-01 20:27:38]

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

[commit-bot: I haz the power, 2017-02-01 20:28:12]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Xianzhu, 2017-02-01 21:22:35]

wangxianzhu@chromium.org changed reviewers:
+ wkorman@chromium.org

[commit-bot: I haz the power, 2017-02-01 21:53:37]

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

[commit-bot: I haz the power, 2017-02-01 21:53:38]

Dry run: Try jobs failed on following builders:
  mac_chromium_rel_ng on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_rel_...)

[Xianzhu, 2017-02-02 17:49:33]

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

[commit-bot: I haz the power, 2017-02-02 17:50:04]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-02-02 20:17:56]

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

[commit-bot: I haz the power, 2017-02-02 20:17:57]

Dry run: Try jobs failed on following builders:
  linux_chromium_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[chrishtr, 2017-02-03 02:00:10]

chrishtr@chromium.org changed reviewers:
+ chrishtr@chromium.org

[chrishtr, 2017-02-03 02:00:11]

https://codereview.chromium.org/2669003002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/paint/ObjectPaintInvalidator.cpp (right):

https://codereview.chromium.org/2669003002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/paint/ObjectPaintInvalidator.cpp:492: //
simplicity, just force full paint invalidation if this object is marked for
Comment needs updating.

[Xianzhu, 2017-02-03 04:37:59]

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

[Xianzhu, 2017-02-03 04:38:35]

https://codereview.chromium.org/2669003002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/paint/ObjectPaintInvalidator.cpp (right):

https://codereview.chromium.org/2669003002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/paint/ObjectPaintInvalidator.cpp:492: //
simplicity, just force full paint invalidation if this object is marked for
On 2017/02/03 02:00:11, chrishtr wrote:
> Comment needs updating.

Done.

[commit-bot: I haz the power, 2017-02-03 04:39:39]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Xianzhu, 2017-02-03 04:40:46]

https://codereview.chromium.org/2669003002/diff/40001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/TestExpectations (right):

https://codereview.chromium.org/2669003002/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/TestExpectations:107: crbug.com/636271
virtual/stable/paint/invalidation/resize-iframe-text.html [ Pass Failure Crash ]
This CL seems to make the test flaky also on other platforms. I don't know the
reason, but the flakiness had been reproducible locally before this CL, and
there is a theory of the flakiness in crbug.com/636271.

[commit-bot: I haz the power, 2017-02-03 06:16:58]

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

[commit-bot: I haz the power, 2017-02-03 06:16:59]

Dry run: This issue passed the CQ dry run.

[wkorman, 2017-02-03 18:30:38]

lgtm

https://codereview.chromium.org/2669003002/diff/40001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/TestExpectations (right):

https://codereview.chromium.org/2669003002/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/TestExpectations:107: crbug.com/636271
virtual/stable/paint/invalidation/resize-iframe-text.html [ Pass Failure Crash ]
On 2017/02/03 at 04:40:46, Xianzhu wrote:
> This CL seems to make the test flaky also on other platforms. I don't know the
reason, but the flakiness had been reproducible locally before this CL, and
there is a theory of the flakiness in crbug.com/636271.

Sorry about the continuing hassle with this test. I think the method the test
uses internally was documented as old and potentially flaky. I will look at it
soon.

[Xianzhu, 2017-02-03 21:49:07]

The CQ bit was checked by wangxianzhu@chromium.org

[commit-bot: I haz the power, 2017-02-03 21:49:40]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-02-03 21:56:37]

CQ is committing da patch.

Bot data: {"patchset_id": 40001, "attempt_start_ts": 1486158547371340,
"parent_rev": "4d68b33ea8398044d86aa76353e6f7de6880bad6", "commit_rev":
"2b30e8534ceb11cd138fa1360e7ae6efc6b4c046"}

[commit-bot: I haz the power, 2017-02-03 21:57:38]

Description was changed from

==========
Optimize outline paint invalidation

Shape of an outline can be affected by layout/style change of
descendants. Previously we assumed that an outline will change
when an object having outline is checked for paint invalidation
for any reason, which covered all cases of outline changes, but
also produced over invalidations, which is especially bad
when we move caret paint invalidation into paint invalidation
stage: a blinking caret will cause full paint invalidation of
the edit box having focus ring.

Now tighten the condition. If a box's outline is not affected
by descendants, don't do full paint invalidation.

TEST=paint/invalidation/outline-*.html
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
==========

to

==========
Optimize outline paint invalidation

Shape of an outline can be affected by layout/style change of
descendants. Previously we assumed that an outline will change
when an object having outline is checked for paint invalidation
for any reason, which covered all cases of outline changes, but
also produced over invalidations, which is especially bad
when we move caret paint invalidation into paint invalidation
stage: a blinking caret will cause full paint invalidation of
the edit box having focus ring.

Now tighten the condition. If a box's outline is not affected
by descendants, don't do full paint invalidation.

TEST=paint/invalidation/outline-*.html
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2669003002
Cr-Commit-Position: refs/heads/master@{#448079}
Committed:
https://chromium.googlesource.com/chromium/src/+/2b30e8534ceb11cd138fa1360e7a...
==========

[commit-bot: I haz the power, 2017-02-03 21:57:40]

Committed patchset #3 (id:40001) as
https://chromium.googlesource.com/chromium/src/+/2b30e8534ceb11cd138fa1360e7a...