Call LayoutFullScreen::updateStyle() when exiting Fullscreen

Call LayoutFullScreen::updateStyle() when exiting Fullscreen

https://codereview.chromium.org/2550703002 changed the definition of
FullscreenController::isFullscreen. Previously, by using
m_fullscreenFrame, it would return true from having entered fullscreen
to having exited fullscreen, i.e. also while exiting. It was changed to
only return true while properly in fullscreen.

This difference turned out to matter, as a call to
LayoutFullScreen::updateStyle() in FullscreenController::updateSize() is
no longer made.

In hangouts.google.com, this broke the layout upon exiting fullscreen.

TEST=https://bugs.chromium.org/p/chromium/issues/detail?id=672804#c21

A regression test for this has proven difficult. With or without the
change, there is a LayoutObject::assertLaidOut() failure upon exiting
fullscreen, which is most likely related to LayoutFullScreen. Yet, it
appears that a non-trivial layout is required to reproduce, which would
require minimizing hangouts.google.com into a test case.

LayoutFullScreen and thus the updateStyle() call is slated for full
removal with top layer (issue 240576) so this does not seem like a good
use of time. Instead, an issue to re-check for failing assertLaidOut()
after top layer has landed was filed: https://crbug.com/685068

BUG=672804
Review-Url: https://codereview.chromium.org/2652763011
Cr-Commit-Position: refs/heads/master@{#446062}
Committed: https://chromium.googlesource.com/chromium/src/+/60d7317f407f575a17410c1a05f52f1ee4941d0c

[foolip, 2017-01-25 08:29:30]

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

[commit-bot: I haz the power, 2017-01-25 08:29:45]

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

[foolip, 2017-01-25 08:39:43]

Description was changed from

==========
Call LayoutFullScreen::updateStyle() when exiting Fullscreen

BUG=672804
==========

to

==========
Call LayoutFullScreen::updateStyle() when exiting Fullscreen

https://codereview.chromium.org/2550703002 changed the definition of
FullscreenController::isFullscreen. Previously, by using
m_fullscreenFrame, it would return true from having entered fullscreen
to having exited fullscreen, i.e. also while exiting. It was changed to
only return true while properly in fullscreen.

This difference turned out to matter, as a call to
LayoutFullScreen::updateStyle() in FullscreenController::updateSize() is
no longer made.

In hangouts.google.com, this broke the layout upon exiting fullscreen.

TEST=https://bugs.chromium.org/p/chromium/issues/detail?id=672804#c21

BUG=672804
==========

[foolip, 2017-01-25 08:53:25]

Description was changed from

==========
Call LayoutFullScreen::updateStyle() when exiting Fullscreen

https://codereview.chromium.org/2550703002 changed the definition of
FullscreenController::isFullscreen. Previously, by using
m_fullscreenFrame, it would return true from having entered fullscreen
to having exited fullscreen, i.e. also while exiting. It was changed to
only return true while properly in fullscreen.

This difference turned out to matter, as a call to
LayoutFullScreen::updateStyle() in FullscreenController::updateSize() is
no longer made.

In hangouts.google.com, this broke the layout upon exiting fullscreen.

TEST=https://bugs.chromium.org/p/chromium/issues/detail?id=672804#c21

BUG=672804
==========

to

==========
Call LayoutFullScreen::updateStyle() when exiting Fullscreen

https://codereview.chromium.org/2550703002 changed the definition of
FullscreenController::isFullscreen. Previously, by using
m_fullscreenFrame, it would return true from having entered fullscreen
to having exited fullscreen, i.e. also while exiting. It was changed to
only return true while properly in fullscreen.

This difference turned out to matter, as a call to
LayoutFullScreen::updateStyle() in FullscreenController::updateSize() is
no longer made.

In hangouts.google.com, this broke the layout upon exiting fullscreen.

TEST=https://bugs.chromium.org/p/chromium/issues/detail?id=672804#c21

A regression test for this has proven difficult. With or without the
change, there is a LayoutObject::assertLaidOut() failure upon exiting
fullscreen, which is most likely related to LayoutFullScreen. Yet, it
appears that a non-trivial layout is required to reproduce, which would
require minimizing hangouts.google.com into a teset case.

LayoutFullScreen and thus the updateStyle() call is slated for full
removal with top layer (issue 240576) so this does not seem like a good
use of time. Instead, an issue to re-check for failing assertLaidOut()
after top layer has landed was filed: https://crbug.com/685068

BUG=672804
==========

[foolip, 2017-01-25 08:53:36]

Description was changed from

==========
Call LayoutFullScreen::updateStyle() when exiting Fullscreen

https://codereview.chromium.org/2550703002 changed the definition of
FullscreenController::isFullscreen. Previously, by using
m_fullscreenFrame, it would return true from having entered fullscreen
to having exited fullscreen, i.e. also while exiting. It was changed to
only return true while properly in fullscreen.

This difference turned out to matter, as a call to
LayoutFullScreen::updateStyle() in FullscreenController::updateSize() is
no longer made.

In hangouts.google.com, this broke the layout upon exiting fullscreen.

TEST=https://bugs.chromium.org/p/chromium/issues/detail?id=672804#c21

A regression test for this has proven difficult. With or without the
change, there is a LayoutObject::assertLaidOut() failure upon exiting
fullscreen, which is most likely related to LayoutFullScreen. Yet, it
appears that a non-trivial layout is required to reproduce, which would
require minimizing hangouts.google.com into a teset case.

LayoutFullScreen and thus the updateStyle() call is slated for full
removal with top layer (issue 240576) so this does not seem like a good
use of time. Instead, an issue to re-check for failing assertLaidOut()
after top layer has landed was filed: https://crbug.com/685068

BUG=672804
==========

to

==========
Call LayoutFullScreen::updateStyle() when exiting Fullscreen

https://codereview.chromium.org/2550703002 changed the definition of
FullscreenController::isFullscreen. Previously, by using
m_fullscreenFrame, it would return true from having entered fullscreen
to having exited fullscreen, i.e. also while exiting. It was changed to
only return true while properly in fullscreen.

This difference turned out to matter, as a call to
LayoutFullScreen::updateStyle() in FullscreenController::updateSize() is
no longer made.

In hangouts.google.com, this broke the layout upon exiting fullscreen.

TEST=https://bugs.chromium.org/p/chromium/issues/detail?id=672804#c21

A regression test for this has proven difficult. With or without the
change, there is a LayoutObject::assertLaidOut() failure upon exiting
fullscreen, which is most likely related to LayoutFullScreen. Yet, it
appears that a non-trivial layout is required to reproduce, which would
require minimizing hangouts.google.com into a test case.

LayoutFullScreen and thus the updateStyle() call is slated for full
removal with top layer (issue 240576) so this does not seem like a good
use of time. Instead, an issue to re-check for failing assertLaidOut()
after top layer has landed was filed: https://crbug.com/685068

BUG=672804
==========

[foolip, 2017-01-25 08:57:34]

foolip@chromium.org changed reviewers:
+ eae@chromium.org, esprehn@chromium.org

[foolip, 2017-01-25 08:57:34]

PTAL. This is is M57 regression and I'd like to do a merge request. I've already
made a considerable effort to test this using as a unit test, but wasn't able to
find any layout object where needsLayout() changed with and without this change,
so I concluded that time-consuming minimization would be needed and propose the
follow-up check as described.

[commit-bot: I haz the power, 2017-01-25 09:56:39]

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

[commit-bot: I haz the power, 2017-01-25 09:56:40]

Dry run: This issue passed the CQ dry run.

[eae, 2017-01-25 17:20:26]

OK, LGTM

[foolip, 2017-01-25 18:25:52]

The CQ bit was checked by foolip@chromium.org

[commit-bot: I haz the power, 2017-01-25 18:26:16]

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, 2017-01-25 18:31:12]

CQ is committing da patch.

Bot data: {"patchset_id": 1, "attempt_start_ts": 1485368752942490, "parent_rev":
"c98415fefbb279f0f0cca9eae96bc6ee05dda7b3", "commit_rev":
"60d7317f407f575a17410c1a05f52f1ee4941d0c"}

[commit-bot: I haz the power, 2017-01-25 18:31:47]

Description was changed from

==========
Call LayoutFullScreen::updateStyle() when exiting Fullscreen

https://codereview.chromium.org/2550703002 changed the definition of
FullscreenController::isFullscreen. Previously, by using
m_fullscreenFrame, it would return true from having entered fullscreen
to having exited fullscreen, i.e. also while exiting. It was changed to
only return true while properly in fullscreen.

This difference turned out to matter, as a call to
LayoutFullScreen::updateStyle() in FullscreenController::updateSize() is
no longer made.

In hangouts.google.com, this broke the layout upon exiting fullscreen.

TEST=https://bugs.chromium.org/p/chromium/issues/detail?id=672804#c21

A regression test for this has proven difficult. With or without the
change, there is a LayoutObject::assertLaidOut() failure upon exiting
fullscreen, which is most likely related to LayoutFullScreen. Yet, it
appears that a non-trivial layout is required to reproduce, which would
require minimizing hangouts.google.com into a test case.

LayoutFullScreen and thus the updateStyle() call is slated for full
removal with top layer (issue 240576) so this does not seem like a good
use of time. Instead, an issue to re-check for failing assertLaidOut()
after top layer has landed was filed: https://crbug.com/685068

BUG=672804
==========

to

==========
Call LayoutFullScreen::updateStyle() when exiting Fullscreen

https://codereview.chromium.org/2550703002 changed the definition of
FullscreenController::isFullscreen. Previously, by using
m_fullscreenFrame, it would return true from having entered fullscreen
to having exited fullscreen, i.e. also while exiting. It was changed to
only return true while properly in fullscreen.

This difference turned out to matter, as a call to
LayoutFullScreen::updateStyle() in FullscreenController::updateSize() is
no longer made.

In hangouts.google.com, this broke the layout upon exiting fullscreen.

TEST=https://bugs.chromium.org/p/chromium/issues/detail?id=672804#c21

A regression test for this has proven difficult. With or without the
change, there is a LayoutObject::assertLaidOut() failure upon exiting
fullscreen, which is most likely related to LayoutFullScreen. Yet, it
appears that a non-trivial layout is required to reproduce, which would
require minimizing hangouts.google.com into a test case.

LayoutFullScreen and thus the updateStyle() call is slated for full
removal with top layer (issue 240576) so this does not seem like a good
use of time. Instead, an issue to re-check for failing assertLaidOut()
after top layer has landed was filed: https://crbug.com/685068

BUG=672804

Review-Url: https://codereview.chromium.org/2652763011
Cr-Commit-Position: refs/heads/master@{#446062}
Committed:
https://chromium.googlesource.com/chromium/src/+/60d7317f407f575a17410c1a05f5...
==========

[commit-bot: I haz the power, 2017-01-25 18:31:49]

Committed patchset #1 (id:1) as
https://chromium.googlesource.com/chromium/src/+/60d7317f407f575a17410c1a05f5...