GitLab

Invalidation, search engines, and omnibox preferences are removed from
the persistent storage list for incognito mode.
After this CL, if these preferences are changed from incognito mode,
they are only stored in memory and won't affect user profile.

Bug: 861722
Change-Id: I8c2fbfac8ff8e9002ab828c908e826419bff2abc
Reviewed-on: https://chromium-review.googlesource.com/1185599Reviewed-by: Peter Kasting <pkasting@chromium.org>
Reviewed-by: Nicolas Zea <zea@chromium.org>
Commit-Queue: Ramin Halavati <rhalavati@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585703}

Running Chrome normally and watching it for a few minutes, I don't
see this being delay loaded by anything, so this is probably
a load time improvement.

BUG=871847

Change-Id: I7c6647325f63d7d7dfa0e6b0807c15757cf8fda0
Reviewed-on: https://chromium-review.googlesource.com/1187711Reviewed-by: Bruce Dawson <brucedawson@chromium.org>
Commit-Queue: Will Harris <wfh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585702}

This patch refactors the function by extracing DirectionAt() out of it,
so that the logic of getting direction from VisiblePosition is
abstracted. It also reduces one client of InlineBox, making it easier to
transition to LayoutNG.

Bug: 877263
Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng
Change-Id: Ifba556e3fb03b7bfa6e4ed27df5e39d07a45eb7c
Reviewed-on: https://chromium-review.googlesource.com/1187630Reviewed-by: Yoichi Osato <yoichio@chromium.org>
Reviewed-by: Yoshifumi Inoue <yosin@chromium.org>
Commit-Queue: Xiaocheng Hu <xiaochengh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585701}

More variables in this file are said to be const but are declared var.
Make the directory tree selectors const.

ENTRY_SET is too generic (not really helpful for code searching): make
it FOLDER_ENTRY_SET, also make it const. Ahhh DIRECTORY: be const till
I think of a better name.

Bug: 876267
Change-Id: Ic0c1a25c4d019088cf73e45119081b4582c640c0
Reviewed-on: https://chromium-review.googlesource.com/1186521
Commit-Queue: Noel Gordon <noel@chromium.org>
Reviewed-by: Luciano Pacheco <lucmult@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585700}

Add new test file for tests related to Launcher.

Add test util to call the Files app event listener that listens to
chrome.launcherSearchProvider.OnOpenResult event, which emulates a
call from the Launcher app.

Bug: 869481
Change-Id: I178b20a8860348d613dd655291747a98e94cce95
Reviewed-on: https://chromium-review.googlesource.com/1186295Reviewed-by: Noel Gordon <noel@chromium.org>
Commit-Queue: Luciano Pacheco <lucmult@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585699}

The IF condition is inverted, it should return when it isn't a native
FileEntry/DirectoryEntry, only FilesAppEntry's types have "type_name"
attribute defined.

Test: New browser test being added on crrev.com/c/1186295
Bug: 869481
Change-Id: Icfdbdedc8289c65447f02c0604e6a36ac8176e6b
Reviewed-on: https://chromium-review.googlesource.com/1186006
Commit-Queue: Luciano Pacheco <lucmult@chromium.org>
Reviewed-by: Noel Gordon <noel@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585698}

- Make sure the base::SharedMemory passed in is valid.
- Add a helper function in PdfCompositorImplTest to create test data.
- Remove some unused headers.

Change-Id: I2e6efeec7ec6efb75bd973edc9d1a532e1919ff0
Reviewed-on: https://chromium-review.googlesource.com/1186271Reviewed-by: Wei Li <weili@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585697}

TREEITEM_DRIVE and TREEITEM_DOWNLOADS are shared between various tests
but are var-defined in multiple places.

Multiple definitions are error-prone: choose one place and define them
as const therein (CreateNewFolder for now). Update folder_shortcuts.js
test to use this new TREEITEM_DRIVE definition, keyboard_operations.js
already uses the TREEITEM_DRIVE/DOWNLOADS definitions.

Test: browser_tests --gtest_filter="*/FilesApp*"
No-presubmit: true
Bug: 876267
Change-Id: I7b86719a29a25e9d649c33edbe8af6cddae529bc
Reviewed-on: https://chromium-review.googlesource.com/1186390
Commit-Queue: Noel Gordon <noel@chromium.org>
Reviewed-by: Luciano Pacheco <lucmult@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585696}

-keepname -> -keepnames

Change-Id: Ie8febb9c4239e5344b0e3e6b333b4576cccb84b3
Reviewed-on: https://chromium-review.googlesource.com/1185692Reviewed-by: John Budorick <jbudorick@chromium.org>
Commit-Queue: agrieve <agrieve@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585695}

Convert list:
snav-display-contents-crash
snav-div-in-anchor
snav-fully-aligned-horizontally
snav-fully-aligned-vertically
snav-hidden-focusable-element
snav-imagemap-area-not-focusable
snav-imagemap-area-without-image
snav-imagemap-overlapped-areas
snav-imagemap-simple
snav-media-elements

Bug: 803086

Change-Id: Iebe20bba5868fec1e1c38f2a188346432c42d568
Reviewed-on: https://chromium-review.googlesource.com/1184945Reviewed-by: Hugo Holgersson <hugoh@vewd.com>
Commit-Queue: JunHo Seo <junho0924.seo@lge.com>
Cr-Commit-Position: refs/heads/master@{#585694}

A follow-up commit will use the IDs as a map key, to de-duplicate
installation tasks and avoid re-installing a web app that's already
installed. That requires that different web apps have different IDs.

Change-Id: I8a86e6cd20ed59a2f5df0683353935004a2c39a5
Reviewed-on: https://chromium-review.googlesource.com/1186927Reviewed-by: Giovanni Ortuño Urquidi <ortuno@chromium.org>
Commit-Queue: Nigel Tao <nigeltao@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585693}

Bug: 875087
Change-Id: I983ea386fa6ea80d46ea0ed732f620003c1c4099
Reviewed-on: https://chromium-review.googlesource.com/1185855Reviewed-by: Filip Gorski <fgorski@chromium.org>
Commit-Queue: Brandon Wylie <wylieb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585692}

Currently, the new print preview always sends fit to page = false for
HTML documents where this setting is not available. Instead, send true
unless the destination is Save as PDF, to match the old UI behavior.

Bug: None
Change-Id: I6f303aa176dd87fe90f6bde3a603800e9617b831
Reviewed-on: https://chromium-review.googlesource.com/1180590Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: Rebekah Potter <rbpotter@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585691}

This is a follow-up CL for http://crrev.com/585412

We can't find a specific description in spec, but it is natural
to put optional arguments after all non-optional arguments.

This changes the order of arguments in a code generator test case
from: method(arg1, optional arg2, arg3)
to  : method(arg1, arg2, optional arg3)


Bug: 876608
Change-Id: Ief33bbdeef7fefc832a03d78b5ca28ab68aa63f2
Reviewed-on: https://chromium-review.googlesource.com/1187847Reviewed-by: Yuki Shiino <yukishiino@chromium.org>
Reviewed-by: Kenichi Ishibashi <bashi@chromium.org>
Reviewed-by: Kentaro Hara <haraken@chromium.org>
Commit-Queue: Hitoshi Yoshida <peria@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585690}

This lets users turn ThinLTO optimizations on with a single gn arg,
and adds some special optimization settings for Android.

No functionality change is intended by this patch.

Bug: 807147
Test: `ninja`
Change-Id: I7c2da21c1d5c2e779abce598ed775109453c8fa0
Reviewed-on: https://chromium-review.googlesource.com/1148997Reviewed-by: Peter Collingbourne <pcc@chromium.org>
Reviewed-by: agrieve <agrieve@chromium.org>
Commit-Queue: George Burgess <gbiv@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585689}

The animation on network feature pod button was drawin in a wrong color
while a wireless network is connecting, when in OOBE and Unified System
Tray is enabled.

Bug: 874829
Change-Id: I4506008e5584e6f3830198d7236a7d75f2ca0aa5
Reviewed-on: https://chromium-review.googlesource.com/1183008Reviewed-by: Tetsui Ohkubo <tetsui@chromium.org>
Reviewed-by: Yoshiki Iguchi <yoshiki@chromium.org>
Commit-Queue: Tatsuhisa Yamaguchi <yamaguchi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585688}

This CL adds SignInPreference to SyncAndServicesPreferences that is only
visible when the user is not signed in. SignInPreference in
SyncAndServicesPreferences shouldn't show personalized sign-in promo, so
setPersonalizedPromoEnabled method is added to SignInPreference. This CL
also hides 'Use Sync and all services' toggle and collapses 'Sync and
personalization' section in SyncAndServicesPreferences if the user is
not signed in.

Bug: 814728
Change-Id: If09fa20144bc4669fe03e95cb5719347aa652afc
Reviewed-on: https://chromium-review.googlesource.com/1187154Reviewed-by: Theresa <twellington@chromium.org>
Commit-Queue: Boris Sazonov <bsazonov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585687}

Remove the Delegate interface from
ProxyResolutionServiceProvider. This code lives in
//chrome/browser/chromeos/dbus now, so it's able to access
profiles directly.

Bug: 843392
Change-Id: I1fe04762b260fdd4f4cc792f8921c085f1ded7b8
Reviewed-on: https://chromium-review.googlesource.com/1187727Reviewed-by: James Cook <jamescook@chromium.org>
Commit-Queue: Dan Erat <derat@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585686}

Motion implementation has stabilized. Deleting the flag and removing
dead code paths.

Bug: b:113107934
Change-Id: I3dac44473d176fbe241a4b344527e8e37d69b0c9
Reviewed-on: https://chromium-review.googlesource.com/1187266
Commit-Queue: David Black <dmblack@google.com>
Reviewed-by: Xiaohui Chen <xiaohuic@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585685}

This reverts commit ecdb6a3b.

Reason for revert:
This is causing a weird behavior where if you mouse-down on the button and
then hold and drag way, select-to-speak is activated anyway. Select-to-Speak
can also be activated from behind the on-screen keyboard.

TBR=jamescook@chromium.org,dtseng@chromium.org

Bug: 870922
Change-Id: Ib9b7395e3a66c491a6c3274dd93898b4e064dd4d
Reviewed-on: https://chromium-review.googlesource.com/1187728
Commit-Queue: Katie Dektar <katie@chromium.org>
Reviewed-by: Katie Dektar <katie@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585684}

Turns out the new version of iBus (starting from 1.5.17) will multiply the coordinates with the scale factor specified by GDK_SCALE.
Chrome on linux now send the system based coordinates to the system's gtk-base IME, despite the scale factor specified by the app.
The correct coordinates should be divided by the GDK_SCALE of the app window.

Bug: 774974
Change-Id: I654bacaa1a4eaba13680a33f2516e41b35d31a13
Reviewed-on: https://chromium-review.googlesource.com/1186291Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Shu Chen <shuchen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585683}

Bug: 648475

Change-Id: Ibbafd96a020850881c0c515e0e4dd71ee530c099
Reviewed-on: https://chromium-review.googlesource.com/1187922Reviewed-by: Peter Collingbourne <pcc@chromium.org>
Commit-Queue: agrieve <agrieve@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585682}

Hosted app windows using GlassBrowserFrameView (Windows 10 only)
had their controls and title misplaced when maximized.
This CL adds missing padding values to the titlebar layout.

Before: https://bugs.chromium.org/p/chromium/issues/attachment?aid=354512&signed_aid=Tmrgizah9A_QhgrbRHXHFQ==&inline=1
After: https://bugs.chromium.org/p/chromium/issues/attachment?aid=354513&signed_aid=h_aEa4bnb7fIu3UutTuNrA==&inline=1

Bug: 876168, 872203
Change-Id: I9e4d3e1ff0776cfba09981cad30a0cfa39469e7c
Reviewed-on: https://chromium-review.googlesource.com/1183019
Commit-Queue: Alan Cutter <alancutter@chromium.org>
Reviewed-by: Bret Sepulveda <bsep@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585681}

This reverts commit a2a4e75d.

Reason for revert: [sheriff] a culprit for cronet_tests failing on multiple builders
See https://crbug.com/877355 for details.

Original change's description:
> Check that TaskScheduler is not leaked between tests, or test-cases.
> 
> - Move some APIs off of base::TestSuite, that we unnecessarily public.
> - Rename the various internal TestEventListeners to express purpose.
> - Add CheckForLeakedGlobals check, and DisableCheckForLeakedGlobals API.
>   The latter is called by the content::ContentTestSuiteBase, since the
>   various browser- and ui-tests running under that suite base run more
>   like the browser itself, and so are expected to let global singletons
>   leak.
> 
> Bug: 875486
> Change-Id: If749dfe8530abc4a5c13a840bbf8258d48c377c5
> Reviewed-on: https://chromium-review.googlesource.com/1182612
> Reviewed-by: Scott Violet <sky@chromium.org>
> Reviewed-by: Gabriel Charette <gab@chromium.org>
> Commit-Queue: Wez <wez@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#585652}

TBR=sky@chromium.org,wez@chromium.org,gab@chromium.org

Change-Id: I906458f276100d9f924a0a78c7cca3a55bcb28b6
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 875486
Reviewed-on: https://chromium-review.googlesource.com/1187982Reviewed-by: Hayato Ito <hayato@chromium.org>
Commit-Queue: Hayato Ito <hayato@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585680}

Neither the mustash team nor the accessibility team has any concrete
plans to ship this. We can always restore it if we change our minds.

This eliminates a couple IsAshInBrowserProcess() checks in //ash.

Bug: 876115
Test: ash_unittests
Change-Id: I60928f52facc4a4d382dff2c4c373d967f203f0a
Reviewed-on: https://chromium-review.googlesource.com/1187563Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585679}

This CL makes HostViewController request to autohide the home indicator
(the bar acting like a virtual home button), so that user can see
things on the bottom after the home indicator times out.

This CL is dependent on internal CL:
https://chrome-internal-review.googlesource.com/c/chrome/ios_internal/+/666568

Bug: 876014
Change-Id: I4d5e3f983100dd43f84f5d422bba31207707c651
Reviewed-on: https://chromium-review.googlesource.com/1187655Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Commit-Queue: Yuwei Huang <yuweih@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585678}

This is the last Gallery unittest not closure compiled.

Bug: 867700
Change-Id: I6e4c2b2cf23b80967e060d838dd3872b55306daf
Reviewed-on: https://chromium-review.googlesource.com/1184644Reviewed-by: Noel Gordon <noel@chromium.org>
Commit-Queue: Trent Apted <tapted@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585677}

We're low on capacity. Will re-enable the tests when we've refreshed
the pool.

NOTRY=true

Bug: 877280
Change-Id: Ibfe66c0e0a6cdbf7c7c954b2f636fee1f970d0c4
Reviewed-on: https://chromium-review.googlesource.com/1187678
Commit-Queue: Ben Pastene <bpastene@chromium.org>
Reviewed-by: John Budorick <jbudorick@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585676}

Some informational messages in preferences use text appearance that is
normally used for preference summaries (for example, the link to 'Sync
and other Google services' in PrivacyPreferences that is shown if
UNIFIED_CONSENT is enabled). This CL improves TextMessagePreference to
support messages in summaries and updates PrivacyPreferences to use
TextMessagePreference, thus fixing the alignment issue there. No longer
used TextMessageWithLinkAndIconPreference is removed.

Bug: 864806
Change-Id: Ia4375a767a34eadbdc82a7ebd8fa0e90a3e254a5
Reviewed-on: https://chromium-review.googlesource.com/1187152Reviewed-by: Theresa <twellington@chromium.org>
Commit-Queue: Boris Sazonov <bsazonov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585675}

Add a line to the tooltip of the extension action indicating if the
extension either wants access to the current page or has access to the
current page (don't indicate if it neither has nor wants access).

Add tests for the same.

Bug: 876930

Change-Id: Id3da1e77df0f1f1a3109d06be8da1dfa23f543fb
Reviewed-on: https://chromium-review.googlesource.com/1185401
Commit-Queue: Devlin <rdevlin.cronin@chromium.org>
Reviewed-by: Karan Bhatia <karandeepb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585674}

This reverts commit 1501aabf.

Reason for revert: crbug.com/877063 speculative fix.
Original change's description:
> Clang: stop building with GCC/libstdc++ from Ubuntu Precise (roll 338452-1:338452-2)
> 
> Previously we used a GCC/libstdc++ package from Ubuntu Precise to build a Clang
> that would work on Precise or later. These days, all bots and developers are on
> Trusty or later, so this should no longer be necessary.
> 
> Bug: none
> Change-Id: I518ee55d07c04dc89163fe262c81d935e05fed78
> Reviewed-on: https://chromium-review.googlesource.com/1183915
> Reviewed-by: Nico Weber <thakis@chromium.org>
> Reviewed-by: Thomas Anderson <thomasanderson@chromium.org>
> Commit-Queue: Hans Wennborg <hans@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#585067}

TBR=thakis@chromium.org,hans@chromium.org,thomasanderson@chromium.org

# Not skipping CQ checks because original CL landed > 1 day ago.

Bug: none
Change-Id: I632e2e47d856d2c7df7e629f12068903e149d166
Reviewed-on: https://chromium-review.googlesource.com/1187344Reviewed-by: Justin DeWitt <dewittj@chromium.org>
Reviewed-by: Hans Wennborg <hans@chromium.org>
Commit-Queue: Hans Wennborg <hans@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585673}

This CL defines the live tab count buckets, adds a namespace and helper
methods for defining metrics bucketed by live tab count, and adds unit tests
that test example metrics. Actual metrics will be implemented in follow-up CLs.

Design doc: https://docs.google.com/document/d/1FB6Pv5Pu0jbYJwSfnqSeufHVOGEt9-d1agpzA4_CC-k/edit?usp=sharing

Bug: 861796
Change-Id: Idd510da5f3c40f9c4d5d1dc292fb6b4139217af2
Reviewed-on: https://chromium-review.googlesource.com/1166215
Commit-Queue: Scott Haseley <shaseley@google.com>
Reviewed-by: Chris Hamilton <chrisha@chromium.org>
Reviewed-by: Mark Pearson <mpearson@chromium.org>
Reviewed-by: Jochen Eisinger <jochen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585672}

This patch ensures that when the fragmentainer iterator returns a
fragment that has an empty clip, we skip generating a fragment for it.

This is important since a fragment that has an empty clip can share
the logical top with the subsequent fragment, meaning that we cannot
use the logical top to uniquely identify a fragment. This patch addresses
this issue.

Furthermore, we typically skip painting fragments that have empty clips
anyway (for obvious reasons), so this should not impact painted contents.

R=chrishtr@chromium.org, wangxianzhu@chromium.org

Bug: 870521
Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_slimming_paint_v2;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: I9f2f6311e0c3c75e4a8d114f13476ce8774e8218
Reviewed-on: https://chromium-review.googlesource.com/1187234
Commit-Queue: Chris Harrelson <chrishtr@chromium.org>
Reviewed-by: Xianzhu Wang <wangxianzhu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585671}

Specifically, those *.json files live in
"/usr/share/chromium/extensions/web_apps", not one level above that.

This matches what arc_default_app_list.cc does for external ARC apps,
after substituting s/arc/web_apps/.

In the future, we may still merge all of those directories into one, but
we will start with segregrated extensions, ARC apps and web apps.

Change-Id: I79176f00b7c3c19b87468465c2640fde89498da6
Reviewed-on: https://chromium-review.googlesource.com/1166272Reviewed-by: Dominick Ng <dominickn@chromium.org>
Commit-Queue: Nigel Tao <nigeltao@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585670}

Change-Id: I5df92149a180050e7c1583f891dc8d1b181b32f3
Reviewed-on: https://chromium-review.googlesource.com/1185339Reviewed-by: Dominick Ng <dominickn@chromium.org>
Commit-Queue: Nigel Tao <nigeltao@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585669}

This test is flaking on the Fuchsia bot, e.g.
https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Fuchsia%20x64/16873

TBR=kmarshall@chromium.org

Bug: 876894
Change-Id: I455c56835af050aa1f9ff6715a6243500cc06e1d
Reviewed-on: https://chromium-review.googlesource.com/1187552Reviewed-by: Scott Little <sclittle@chromium.org>
Commit-Queue: Scott Little <sclittle@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585668}

Previously, ChildProcessLauncherHelper would bind a ServiceDirectory
connection on the launcher thread. The binding would attach to the
asynchronous dispatcher of the launcher task runner's thread,
rather than the client thread. This generated some test flakiness
during test teardown.

This CL moves the ServiceDirectory binding call so that it runs
in the client thread instead.

Bug: 876894
Change-Id: I53e05d79a24ab8347e3df697f65f84601f618f40
Reviewed-on: https://chromium-review.googlesource.com/1187125
Commit-Queue: Wez <wez@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585667}

This CL allows the iOS client to display desktop that is partially
obstructed by the notch while allowing user to pan the desktop out of
the obstructed area, similar to what we have done on the Android
client.

This CL also changes the previous logic that handles keyboard height.
We are now using DesktopViewport's SetSafeInsets() methods to adjust
for the keyboard rather than just changing the surface size.

Bug: 876014
Change-Id: I428cf2552acb3bc33cf932890a9395428d969a40
Reviewed-on: https://chromium-review.googlesource.com/1186027Reviewed-by: Joe Downing <joedow@chromium.org>
Commit-Queue: Yuwei Huang <yuweih@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585666}

Moves plain Mash tests to FYI waterfall, while retaining a login test on the
main waterfall as a sanity check.

Bug: 874090
Change-Id: I86c60344610c2a4f99b35f9116c5366771a16c64
Reviewed-on: https://chromium-review.googlesource.com/1175222
Commit-Queue: Ryan Cui <rcui@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585665}

When Android goes invisible in OOP-D, it wasn't tearing down the
display, which can lead to GL issues as we continue to use GL
after the window (used to create the GL surface) is destroyed.

In order to tear down the display for Viz, we need to invalidate
our root frame sink ID. This change refactors things so that we
always invalidate the root frame sink ID on going invisible, and
re-register it on becoming visible. This allows both viz/non-viz
to share the same logic.

As registering/unregistering isn't doing much in non-viz case,
this doesn't add significant overhead there.

Bug: 863049
Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel
Change-Id: I1589e402185fd9e2cdb007d3d8cd739f303ad48a
Reviewed-on: https://chromium-review.googlesource.com/1184376Reviewed-by: Khushal <khushalsagar@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Commit-Queue: Eric Karl <ericrk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585664}