GitLab

Fixes a minor copy/paste artifact in the string descriptions.

Bug: b:156774894
Change-Id: Ie1c42dd29ffbd917f3f11b575c4526e1cb56911b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2228140
Commit-Queue: Bailey Berro <baileyberro@chromium.org>
Commit-Queue: Kyle Horimoto <khorimoto@chromium.org>
Reviewed-by: Kyle Horimoto <khorimoto@chromium.org>
Auto-Submit: Bailey Berro <baileyberro@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774731}

Because of crbug.com/1080691, WPT import constantly adds redundant
virtual expectations, and the warnings (though not errors) are
annoying to most developers.

Actually the virtual fallback feature was to relieve burden of
sheriffs and developers for adding duplicated virtual and base
expectations. Preventing redundant expectations is not the main goal.
Existing redundant expectations are not very harmful.

Bug: 1080691
Change-Id: Id0a4ac931d417a223d337cd260aba47ddd74395f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2227454
Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
Auto-Submit: Xianzhu Wang <wangxianzhu@chromium.org>
Reviewed-by: Robert Ma <robertma@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774730}

- Copies the clang asan library into iOS app bundles. This library is
  pulled from WebKitTestRunnerApp.app/Frameworks.

- Updates the iOS test runner to set DYLD_LIBRARY_PATH to pick up the
  clang asan library from its copied location.

- Fixes the macOS build to stop rewriting WK_API_AVAILABLE() calls, as
  this was causing compiler errors.

BUG=None

Change-Id: Ie30b40b01d3724b0d922f0edc1c39544bc251f2d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2228116
Commit-Queue: Rohit Rao <rohitrao@chromium.org>
Reviewed-by: Ali Juma <ajuma@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774729}

TriChrome consists of {Library, Chrome, WebView}. Each of these units
can be an input to SuperSize. To get SuperSize to aggregate these into
one .size file with a common pool of symbols, we'd like to to abstract
these units as "containers", represented by the Container class, which
relates to other data structure by the following:
* A SizeInfo has 1 or more Containers.
* A Container consists of {metadata, section_sizes} (formerly in
  SizeInfo), and has a distinct name.
* Each Symbol belongs to a Container.

This CL adds the Container class to SuperSize, and rewire former usage
of {metadata, section_sizes} to go through it. Only the *single*
Container case (with name='') is handled, to simplify review and
commits.

The format of .size files is kept invariant. And as seen in .golden
file, very few changes to SuperSize-console output. A number of TODOs
are added to indicate upcoming code changes for multi-Container support.
Other changes:
* describe.py: Use itertools.chain.from_iterable() so that chained
  generators can be appended into array, instead of needing distinct
  local variables.
* diff.py: Add _DiffObj() to generalize diffing.

Bug: 1040645
Change-Id: I799c960d0510157f55876016726aa2eda25d55a9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2228169
Commit-Queue: Samuel Huang <huangs@chromium.org>
Reviewed-by: Andrew Grieve <agrieve@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774728}

This CL deletes FakeTileProvider from java.

Bug: 1090140
Change-Id: I47c1e5a50ad8984aaf85d2f59605cef1756898d6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2219389
Commit-Queue: Shakti Sahu <shaktisahu@chromium.org>
Reviewed-by: David Trainor <dtrainor@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774727}

Work is being done on shelf notification indicators as a separate
feature from context menu notifications, so create a new flag to
separate the two features.

Bug: 1080827
Change-Id: I9e6a2a20a30b692a5e9c185b904b0ff337eaad39
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2228189
Commit-Queue: Matthew Mourgos <mmourgos@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774726}

Using wpt-import in Chromium 66a4d901.
With Chromium commits locally applied on WPT:
f5fcca96 "Revert "[COOP access reporting] Preliminary WPT tests.""


Note to sheriffs: This CL imports external tests and adds
expectations for those tests; if this CL is large and causes
a few new failures, please fix the failures by adding new
lines to TestExpectations rather than reverting. See:
https://chromium.googlesource.com/chromium/src/+/master/docs/testing/web_platform_tests.md

NOAUTOREVERT=true
TBR=robertma@google.com

No-Export: true
Change-Id: I42faef3f10430cf6f443b0c2e8f0c2bbdffca4e8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2228941Reviewed-by: WPT Autoroller <wpt-autoroller@chops-service-accounts.iam.gserviceaccount.com>
Commit-Queue: WPT Autoroller <wpt-autoroller@chops-service-accounts.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#774725}

It can be replaced with TestUkmRecorder::SetOnAddEntryCallback

Bug: 1057770
Change-Id: If49262570bfbfeb90b5d090c42ece84b804c1c5e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2216800Reviewed-by: Mugdha Lakhani <nator@chromium.org>
Commit-Queue: Evan Stade <estade@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774724}

After https://crrev.com/c/1861280 legacy supervised users are
deprecated, but some code about them was left in an unreachable clause.
Now this clause is removed.

Bug: 1090844

Change-Id: I3dc00348d659fc9231c6697d109dc7679da630d9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2219948Reviewed-by: Devlin <rdevlin.cronin@chromium.org>
Commit-Queue: Oleg Davydov <burunduk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774723}

The CL is fixing the issue where the IO-Thread was not receiving the
power manager notifications.

The observer list doesn't register observers without a sequence
being set.

  void AddObserver(ObserverType* observer) {
    // TODO(fdoray): Change this to a DCHECK once all call sites have a
    // SequencedTaskRunnerHandle.
    if (!SequencedTaskRunnerHandle::IsSet())
      return;

Previously, these observers were not registered and notifications
were not received.

Currently, the thread controller bind to the current thread but has
conditional code depending if the task runner is set.

  see: ThreadControllerWithMessagePumpImpl::BindToCurrentThread

  if (task_runner_)
      InitializeThreadTaskRunnerHandle();


The function InitializeThreadTaskRunnerHandle(...) is the place to make
the link between the PowerMonitor and the Controller since a
task_runner is required for power notifications.


Bug: 1074332
Change-Id: If3e652e6516fdc3df4e3ea8ace909389d668e690
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2225627Reviewed-by: François Doray <fdoray@chromium.org>
Reviewed-by: Gabriel Charette <gab@chromium.org>
Commit-Queue: Etienne Bergeron <etienneb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774722}

This CL fixes :
1 - A NPE in query tiles omnibox suggestions.
2 - Removed caching of TileProvider from java layer.

Bug: 1087750
Change-Id: I6977db9f862c62150d51d8a57d59a84602718df3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2225515Reviewed-by: Shakti Sahu <shaktisahu@chromium.org>
Reviewed-by: David Trainor <dtrainor@chromium.org>
Commit-Queue: Shakti Sahu <shaktisahu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774721}

This completes a portion of a long-coming migration.

(TBR for mechanical changes outside of //base)
TBR=fdoray@chromium.org

Bug: 781352
Change-Id: Ia3a43684f0a5a496bd0eed30ecb49d6db1d7f96a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2225921
Commit-Queue: Gabriel Charette <gab@chromium.org>
Reviewed-by: François Doray <fdoray@chromium.org>
Auto-Submit: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774720}

This CL removes the assumption that overlays are visible elements
that extend past 64px (the fixed bar height) because it does not account
for the push-down promo which can increase the height of the bar below
the 64px threshold.

This CL also removes the code that attempts to reduce the number of
overlays that are tracked by trying to find a common parent overlay that
houses child overlay elements.

Bug: b/157846704, b/157136561
Change-Id: Ibb872a4ee1c4d7f146676b124b578fe0ba80a358
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2226634
Commit-Queue: Esmael Elmoslimany <aee@chromium.org>
Reviewed-by: Tibor Goldschwendt <tiborg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774719}

This feature flag improves support for WebXR on computers with more
than one GPU, particularly on Windows. This feature will be landed in a
series of changes.

The CL that will use this flag in Blink: https://crrev.com/c/2219780

Other relevant active CLs:
https://crrev.com/c/2216166
https://crrev.com/c/2225366
+ 1 more that will be uploaded and linked shortly

This change was also initially reviewed as one larger CL here:
https://crrev.com/c/2096778

Bug: 792657
Change-Id: Icc5b40e4d0b7aad702199ce2514573864d5e57f9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2226071Reviewed-by: Rafael Cintron <rafael.cintron@microsoft.com>
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Patrick To <patrto@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#774718}

Roll Chrome Mac PGO profile from chrome-mac-master-1591184865-5fd7e217cbf2739ffd4335727dd5978110d7b63d.profdata to chrome-mac-master-1591199990-53a8f554cee1144b8cf9ac7f99b06fbe1ec5521a.profdata

If this roll has caused a breakage, revert this CL and stop the roller
using the controls here:
https://autoroll.skia.org/r/pgo-mac-chromium
Please CC jeffyoon@google.com,liaoyuke@google.com on the revert to ensure that a human
is aware of the problem.

To report a problem with the AutoRoller itself, please file a bug:
https://bugs.chromium.org/p/skia/issues/entry?template=Autoroller+Bug

Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+doc/master/autoroll/README.md

Cq-Include-Trybots: chrome/try:mac-chrome
Tbr: jeffyoon@google.com,liaoyuke@google.com
Change-Id: I82f875a302543eecc40189fd9169f912826ae051
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2229147Reviewed-by: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com>
Commit-Queue: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#774717}

Previously only "PIN" was supported.

Fixed: 1090167
Change-Id: I7b1879769cebb3d3a8ab56ac67642d7ab3e799e7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2227510
Auto-Submit: Kyle Horimoto <khorimoto@chromium.org>
Reviewed-by: Regan Hsu <hsuregan@chromium.org>
Commit-Queue: Kyle Horimoto <khorimoto@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774716}

This reverts commit f11b606d.

Reason for revert: Did not fix the problem completely. Will check in a universal fix.

Original change's description:
> Fix spacing issue on hover cards.
> 
> Issue was caused by a previous fix for CrOS not taking hover card images
> into account. Now there is a uniform margin around preview images, as
> there was before.
> 
> Also removed a redundant SetProperty call that was overridden later in
> the same function.
> 
> Bug: 1087759
> Change-Id: I5d1ab7081f9e1195961ead27cfc365bb6e25ede8
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2225239
> Reviewed-by: Connie Wan <connily@chromium.org>
> Commit-Queue: Connie Wan <connily@chromium.org>
> Commit-Queue: Dana Fried <dfried@chromium.org>
> Auto-Submit: Dana Fried <dfried@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#774326}

TBR=dfried@chromium.org,connily@chromium.org

Change-Id: Ia3fa918d758687df80be358c79803413ecc9869d
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 1087759
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2229320Reviewed-by: Dana Fried <dfried@chromium.org>
Commit-Queue: Dana Fried <dfried@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774715}

.... instead of WebMediaStreamTracker.

This is part of the effort to reduce the needless use of
public Blink APIs (wrappers) within renderer/modules.

BUG=704136
R=guidou@chromium.org

Change-Id: I8b89b7b715c98a3a0cf3e26707848de091732c64
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2228942
Commit-Queue: Antonio Gomes <tonikitoo@igalia.com>
Reviewed-by: Guido Urdaneta <guidou@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774714}

Fixing the expectation so it covers all platforms where it is flaky.

Bug: 984467
Change-Id: I5daf2f5f02e6140163d659a75b74a39d24f7c3a4
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2228137
Auto-Submit: Juanmi Huertas <juanmihd@chromium.org>
Reviewed-by: Yi Xu <yiyix@chromium.org>
Commit-Queue: Juanmi Huertas <juanmihd@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774713}

This CL moves AccountManagerTestRule to the package of SigninTestUtil
so we can use SigninTestUtil.signIn/signOut methods in the rule later.

Bug: 1004418
Change-Id: I82204f8661ecda00ebbf1e741ce7ae5bec793a95
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2209214Reviewed-by: Boris Sazonov <bsazonov@chromium.org>
Reviewed-by: Alex Ilin <alexilin@chromium.org>
Reviewed-by: Tanmoy Mollik <triploblastic@chromium.org>
Commit-Queue: Alice Wang <aliceywang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774712}

Generate UkmEventNameHash histogram enum automatically from ukm.xml if
and only if ukm.xml path is provided. Also, add some unit tests for the
merge_xml.py.

Bug: 1087533
Change-Id: Ibef6dc02644d49dcb3c9b132e93fd72c759f505d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2224597Reviewed-by: Robert Kaplow <rkaplow@chromium.org>
Reviewed-by: Alexei Svitkine <asvitkine@chromium.org>
Commit-Queue: Weilun Shi <sweilun@chromium.org>
Auto-Submit: Weilun Shi <sweilun@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774711}

As a step towards improving spellcheck initialization performance,
with the ultimate goal of not loading the Windows spellcheck platform
DLLs until needed (e.g. when actually performing a spellcheck), this
CL adds a new feature flag allowing the initialization of the
spellcheck service on browser startup to be prevented. This should
allow addressing the rest of the performance issue in stages.

Since Chromium already by default initializes the spellcheck service
on startup for custom dictionary synchronization, the command line
for launching the browser with Windows hybrid spellchecking enabled
but no initialization of the spellcheck service is:

chrome --enable-features=
WinUseBrowserSpellChecker,WinDelaySpellcheckServiceInit
--disable-sync-types="Dictionary"

Downstream Edge has Windows spellcheck enabled by default and does
not perform a custom dictionary sync, so the corresponding command
line for Edge is:

msEdge --enable-features=WinDelaySpellcheckServiceInit

Bug: 1064351
Change-Id: I91559193918cce6895133a9e9c440513b55fdd65
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2227450Reviewed-by: Rouslan Solomakhin <rouslan@chromium.org>
Reviewed-by: Avi Drissman <avi@chromium.org>
Reviewed-by: Guillaume Jenkins <gujen@google.com>
Commit-Queue: Bruce Long <brlong@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#774710}

TBR=fdoray@chromium.org

No-Try: true
Bug: 1083293
Change-Id: Idfc19377184872f9712ec42c33c7dc7a12896d4f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2229298Reviewed-by: François Doray <fdoray@chromium.org>
Commit-Queue: François Doray <fdoray@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774709}

AverageLagTracker::AddLatencyInFrame is already logically splited
(most part of its body is inside an if/else body) into 2 cases:
ScrollBegin and ScrollUpdate. This CL aims to create a method
for each case so AddLatencyInFrame is leaner and simpler to read.

Also, as this class in being refactored at
https://chromium-review.googlesource.com/c/chromium/src/+/2189633
this simple refactor will reduce the amount of noise created by
CL #2189633.


Bug: 1079024, 1079028
Change-Id: I7c9d2567ac3a875ca4c03b33d77a7a169bbb6375
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2216707Reviewed-by: David Bokan <bokan@chromium.org>
Reviewed-by: Navid Zolghadr <nzolghadr@chromium.org>
Reviewed-by: Daniel Libby <dlibby@microsoft.com>
Commit-Queue: João Victor Almeida de Aguiar <joalmei@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#774708}

... instead of WebMediaStreamTrack.

Note that two occurrences of WebMediaStreamTrack are left off,
since they are part of signature of an overridden method.

This is part of the effort to reduce the needless use of
public Blink APIs (wrappers) within renderer/modules.

BUG=704136
R=guidou@chromium.org

Change-Id: Id3baa879075dbc030fec03cdeb35cc6e2026a57d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2225543
Commit-Queue: Antonio Gomes <tonikitoo@igalia.com>
Commit-Queue: Guido Urdaneta <guidou@chromium.org>
Auto-Submit: Antonio Gomes <tonikitoo@igalia.com>
Reviewed-by: Guido Urdaneta <guidou@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774707}

base::PostTask with no traits implicitly posted to base::ThreadPool so
this CL is a no-op.

This change is blocking https://chromium-review.googlesource.com/c/chromium/src/+/2225541

Bug: 968047, 1026641
Change-Id: I0656db21fbe55f6c4d79643ff5605380047d46b7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2229004Reviewed-by: Eugene But <eugenebut@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774706}

Previously, we had one settings enum value for idle power, but I renamed
it and split it into two: one while charging and one while on battery,
which matches the settings page.

Bug: 1083518
Change-Id: Ib5f51ed5094252878dfe1f287a031dc644112eab
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2227970
Commit-Queue: Kyle Horimoto <khorimoto@chromium.org>
Auto-Submit: Kyle Horimoto <khorimoto@chromium.org>
Reviewed-by: Regan Hsu <hsuregan@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774705}

I discovered that AXObject::Language() was not always returning the
correct thing for the root element. The rest of the tree had the
right value, but not the root.

I updated an existing test to cover the case where the document's
language is something other than English.

AX-Relnotes: Fixes the computed language for some web pages.
Bug: none
Change-Id: I3e3bec73e9bded0df3c37bb3c8ba634d3d5018b6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2226338
Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
Reviewed-by: Chris Hall <chrishall@chromium.org>
Auto-Submit: Dominic Mazzoni <dmazzoni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774704}

* Remove the separate OS sync and browser sync toggles
* Add "Yes, I'm in" and "No, thanks" buttons to the footer
* "Yes, I'm in" enables both OS and browser sync
* "No, thanks" disables both OS and browser sync

Screenshot: http://screen/Nm9c79LW4Qs

I will add metrics for the user's choice in a follow-up CL.

Bug: 1074436
Test: updated browser_tests
Change-Id: I4cb81b0e48869323c251f184dcce8aae4ee87913
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2226072Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774703}

This CL follows up on the session history changes in
https://chromium-review.googlesource.com/c/chromium/src/+/2181973.
Prior to that CL, inactive frames or frames with no
FrameNavigationEntry returned early from FindFramesToNavigate(),
without recursing down into subframes.  After that CL, we will now
recurse into subframes and keep looking for navigations in those
cases.  This is harmless, since an inactive frame's children should
all be inactive as well, and subframes of a frame with no FNE also
won't have FNEs, but it's also redundant.  This CL modifies the new
flow through DetermineActionForHistoryNavigation() to match the old
behavior and stop looking for navigations in subframes in those cases.

Bug: 705550
Change-Id: Ida707fc9e108f7102b26cf4e44f4d9761e7dd7a7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2225353
Commit-Queue: Alex Moshchuk <alexmos@chromium.org>
Auto-Submit: Alex Moshchuk <alexmos@chromium.org>
Reviewed-by: Charlie Reis <creis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774702}

Ack a pending surface early the client's scheduler to schedule frames
or future surface changes This case happens when screen rotation takes
place while there is an active frame when this takes place it has a
unique signature of 2 consecutive submit frame requests with
clients_needs_begin_frame_ is set (high) and 1st frame is not a new
surface and 2nd frame is a new surface.

Earlier review: https://chromium-review.googlesource.com/c/chromium/src/+/2222918

Bug: https://bugs.chromium.org/p/chromium/issues/detail?id=1060058
Bug: https://bugs.chromium.org/p/chromium/issues/detail?id=1087865
Change-Id: I6707d264064d16357a0ffe5bdfe5521a7e903bec
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2222803Reviewed-by: Jonathan Ross <jonross@chromium.org>
Commit-Queue: Nazih Almalki <nalmalki@google.com>
Cr-Commit-Position: refs/heads/master@{#774701}

Presently, BubbleDialogDelegate is a WidgetObserver, and observes both
its anchor widget and its own widget for various events. There is a
problem with this: subclasses are unintentionally abusing the fact that
BubbleDialogDelegate is a WidgetObserver to observe changes to either
the anchor widget or the bubble widget. Very few of these subclasses:

* Are aware that BubbleDialogDelegate actually observes multiple
  widgets
* Are aware of the intricacies of when BubbleDialogDelegate starts &
  stops observing widgets
* Bother to call the BubbleDialogDelegate implementations of the
  WidgetObserver methods from their overrides

Change-Id: I9de0d1429431f33e80b564ad3e545c976d141a7e
Bug: 1011446
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2227054
Commit-Queue: Robert Liao <robliao@chromium.org>
Reviewed-by: Robert Liao <robliao@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774700}

Bug: 973801
Change-Id: Ibc35f2e986fd9f9ed13a720a4c275aff132a4b6a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2226773Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Chris Fredrickson <cfredric@google.com>
Cr-Commit-Position: refs/heads/master@{#774699}

We need to pick the action upload endpoint based on
the Chrome channel. While I was here, I refactored some code
to hopefully make it more clear. There was one actual improvement:
we were needlessly copying the request body before gzipping it.

See b/156953207 for some additional information about the api endpoint.

Bug: 1044139
Change-Id: I7ddac54abca3471d68dfffb080ede929ad178848
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2222801Reviewed-by: Ian Wells <iwells@chromium.org>
Commit-Queue: Dan H <harringtond@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774698}

Previously, in order to do a hit test for accessibility we'd
send an action to the renderer with the x,y coordinates, and
the renderer would fire an accessibility event on the matching
node.

It'd be more flexible if we had an API that could do a hit
test and call a callback function with the result. That
eliminates the need for keeping track of requests and matching
them up with events.

This patch refactors the way hit testing works so that a
hit testing request may contain a request for an accessibility
event to fire, or a callback function to call, or both.

Existing tests are extended to cover both mechanisms.

Bug: 1083834


Change-Id: Idb4c1f302361a0eb6d6344bcee9da7f086caf3c8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2209630Reviewed-by: Arthur Sonzogni <arthursonzogni@chromium.org>
Reviewed-by: Emily Stark <estark@chromium.org>
Reviewed-by: Aaron Leventhal <aleventhal@chromium.org>
Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
Auto-Submit: Dominic Mazzoni <dmazzoni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774697}

Directly accessing iteration_duration is unsafe since it is an optional
timing property.

Instead, this patch uses a getter function, which has the necessary
safeguards to prevent nullpot dereferencing.

Bug: 1084547
Change-Id: I617dfa8af9aee02253386ea9ee26934760a02cad
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2225511
Commit-Queue: Hao Sheng <haozhes@chromium.org>
Reviewed-by: Kevin Ellis <kevers@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774696}

The current suppression for tex-2d-rgb-rgb-unsigned_byte was
only for non-angle. However the test also flakes on angle.

This updates the suppression to reflect that.

TBR=kbr@chromium.org
TEST=conformance/textures/image_bitmap_from_video/tex-2d-rgb-rgb-unsigned_byte.html
webgl_conformance_tests
Bug:834933

Change-Id: I170c6c4cae5c0ba039b7505d4b4f5461f9774591
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2229238Reviewed-by: Jonathan Ross <jonross@chromium.org>
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774695}

This reverts commit e1f9cdbe.

Reason for revert: https://crbug.com/943668#c14

Original change's description:
> Third reland of Add LayoutInstabilityTest.OOPIFSubframeWeighting.
> 
> Refactor test to focus on subframe; print window width/height.
> 
> Bug: 943668
> Change-Id: I28f9c66350f709fa792b242334d117255c36dc1e
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2227184
> Reviewed-by: Steve Kobes <skobes@chromium.org>
> Commit-Queue: Annie Sullivan <sullivan@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#774591}

TBR=sullivan@chromium.org,skobes@chromium.org

Change-Id: I2c3747481e371bd918eb69a0983bae35939df50c
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 943668
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2227864Reviewed-by: François Doray <fdoray@chromium.org>
Commit-Queue: François Doray <fdoray@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774694}

The motivation behind this change is that for content like below,
Windows Narrator cannot trigger the links in the tabs.
  <div role="tablist">
    <a href="https://twitter.com" role="tab">tab1</a>
    <a href="https://wikipedia.org" role="tab">tab2</a>
    <a href="https://w3.org" role="tab">tab3</a>
  </div>
This is because for single-selection container we have the behavior
that selection follows focus.
https://www.w3.org/TR/wai-aria-practices-1.1/#kbd_selection_follows_focus
In Item Navigation, Narrator sets focus first to the tab, but upon
focus the a11y selected state is set to true. When Narrator attempts
to perform ISelectionItemProvider::Select (Enter) on the tab to
trigger the link, no action would occur because the selected state is
already set to true.

This CL fixes UIA::ISelectionItemProvider::Select() to still
perform action::kDoDefault if the element is already selected due to
focus. Previously, if an element is already selected, for any reason,
Select() would not perform any a11y action.
This CL also introduces ax::mojom::BoolAttribute::kSelectedFromFocus
and uses it to communicate between renderer process and browser
process that an element is selected due to it being focused.
This CL should not affect ISelectionItemProvider::Select()
behavior on multiselectable containers.

AX-Relnotes: Windows Narrator can now trigger default action on an
element in a single-selection container such as a tabitem.

Bug: 847971
Change-Id: I12026b37ed25b867384895eb520eba658aad6f20
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2219155Reviewed-by: Kevin Babbitt <kbabbitt@microsoft.com>
Reviewed-by: Kinuko Yasuda <kinuko@chromium.org>
Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
Commit-Queue: Victor Fei <vicfei@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#774693}

Currently, when the activity is destroyed and re-created due to
dark/light mode switching, the TabModelSelectorTabObserver in
TabGroupUiMediator is not properly destroyed. This CL fixes this issue
by destroying it when tab is detached from activity.

Bug: 1087826
Change-Id: Ia11138923cb9d4e16e2ce35db18ca74afa50412f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2227890Reviewed-by: Wei-Yin Chen (陳威尹) <wychen@chromium.org>
Commit-Queue: Yue Zhang <yuezhanggg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#774692}