GitLab

Suppress the following tests again:
  ContextLost_WebGLBlockedAfterJSNavigation
  ContextLost_WebGLUnblockedAfterUserInitiatedReload

They're still failing reliably on these two bots.

Also restart the browser after running these two tests; this was
forgotten, and appears necessary for any of these tests which tear
down the GPU process.

Tbr: ynovikov@chromium.org
Tbr: sugoi@chromium.org
No-Try: true
Bug: 832886
Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel
Change-Id: I5918e7b731023b8efb8e1202f8c9b57e1089e1bd
Reviewed-on: https://chromium-review.googlesource.com/1025112
Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Alexis Hétu <sugoi@chromium.org>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553330}

Bug: 512595
Change-Id: Ifdab5416ba41171f07f097910060cc99a08585a4
Reviewed-on: https://chromium-review.googlesource.com/1026642Reviewed-by: Helen Li <xunjieli@chromium.org>
Commit-Queue: Eric Roman <eroman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553329}

Migrate MessageLoop::current() to MessageLoopCurrent in /third_party/blink/renderer/platform/scheduler

These changes were scripted as part of the effort to restrict exposure of the
raw MessageLoop*.

Note that many static methods exposed on MessageLoopCurrent are also deprecated
so it is very well possible that this is migrating to an already deprecated
method.

The goal of this pass is to reduce usage of MessageLoop::current() (and
ultimately make it impossible to obtain the MessageLoop* statically).

As such I will not edit this CL unless the script did something logically wrong.
I defer to owners to fix highlighted usage of already deprecated APIs.

Possible script screw ups / things to look out for in this review:
 - Storing MessageLoopCurrent in a MessageLoop* variable or comparing it against
   one (I will go over that in a separate pass).

Includes should have been stripped if that was the last usage of message_loop.h
in that file.

Please CQ if LGTY

This CL was uploaded by git cl split.

R=rmcilroy@chromium.org

Bug: 825327
Change-Id: Ic6d3ecfe4aed463c921eee5c1d13f5be2210e38a
Reviewed-on: https://chromium-review.googlesource.com/1024435Reviewed-by: Alexander Timin <altimin@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553328}

Bug: none
Change-Id: If0e57d955483f2ba7e55cc1bae744cfbc4567868
Reviewed-on: https://chromium-review.googlesource.com/1026551Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553327}

This CL makes Ozone/Wayland able to exchange clipboard data with
other applications on the system. It accomplishes the functionality
by making use of wl interfaces like wl_data_source, wl_data_offer,
wl_data_device and wl_data_device_manager.

In order to bridge the communication between Ozone backends (eg wayland)
and the rest of Chromium, CL also introduces a 'delegate' class
class named ClipboardDelegate. In practice, clients of ClipboardDelegate
(eg Mus' ClipboardImpl) delegate clipboard operations to the active
Ozone backend.

FakeServer is also extended to support the various compositor side
functionalities needed so that tests in wayland_data_device_unittest.cc run.
For instance, MockDataDeviceManager, MockDataDevice, MockDataOffer,
MockDataSource classes are added, implementing the respective Wayland
compositor hooks to communicate with Wayland clients (eg Ozone/Wayland).

Last, the CL focuses on the Ozone/Wayland side of the implementation,
leaving plumbing with the rest of Chromium for follow up CLs (eg [1]).
However in order to test the functionality, wayland_data_device_unittest.cc
is introduced and mimics how a client (eg Mus) would use of the functionality.

[1] https://crrev.com/c/981372/

API footprint:

As mentioned, CL adds a new Ozone public API class: ClipboardDelegate. The
API is designed to support "promise scheme" clipboard implementations,
like to Wayland's and MacOS' NSPasteboard.

Flow:

1) Offer clipboard data (writing)
  When offering content to the system wide Clipboard (eg Chromium/Wayland
  user performs a copy or cut operation), WaylandConnection creates a
  WaylandDataSource instance, which installs listeners, including
  ::OnSend for actually exchange the data.
  WaylandDataSource then calls wl_data_source_offer to
  communicate the compositor the mimetypes available for clipboard
  operations, and then wl_data_device_set_selection to flag itself
  as the current active "clipboard owner".

  By the time data is requested from this client (eg another application on
  the system "paste" to an editor), the WaylandDataSource::OnSend callback
  is called with a mime type and a valid file descriptor passed in as
  parameters. Ozone/Wayland then writes it clipboard data into it.
  At this point, the application requesting the data is responsible for
  reading the content from the other end of pipe, and proceed with the
  editing operation on its side, once data is exchanged.

  This is exercised by WaylandDataDeviceManagerTest::WriteToClipboard
  end-to-end, ie client_1 writes data -> compositor -> client_2 reads data.

2) Requesting clibpard data (reading)
  When requesting data from the system wide Clipboard (eg Chromium/Wayland
  user performs a paste operation), WaylandConnection calls wl_data_offer_receive.
  with a valid "WriteIn" file descriptor for the current "clipboard owner"
  write data into. Once written, the data is read by Ozone/Wayland, from
  the other end of the same pipe.

  This is exercised by WaylandDataDeviceManagerTest::ReadFromClibpard
  end-to-end, ie client_2 request data -> compositor -> client_1 writes data.

BUG=578890

Change-Id: Ibfb8c26fe817e7639f32ac4bdc2f106d4b8bb513
Reviewed-on: https://chromium-review.googlesource.com/976461
Commit-Queue: Antonio Gomes <tonikitoo@igalia.com>
Reviewed-by: Robert Kroeger <rjkroege@chromium.org>
Reviewed-by: Maksim Sisov <msisov@igalia.com>
Cr-Commit-Position: refs/heads/master@{#553326}

GetErrorPage function will be used in ChromeWebClient::PrepareErrorPage.

Bug: 725241
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: Id2f9f3e86d9685ad0ac448bc767f7daf4a80e3be
Reviewed-on: https://chromium-review.googlesource.com/1024277Reviewed-by: Danyao Wang <danyao@chromium.org>
Commit-Queue: Eugene But <eugenebut@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553325}

Most notably we'd like to include
https://github.com/w3c/web-platform-tests/pull/10599,
which added support for test variants in .any.js tests.

Other changes in this roll:
* Fixed paths in chromium.patch after the Blink mv.
* Added the new wptserve/config.py to the whitelist.

This roll also exposed an upstream issue. Filed issue
836377 to track it (fix is under way in the upstream).

Bug: 836276, 829697, 836377
Change-Id: I5fc0af49c0149871222eaacd1bf5473b187c38a5
Reviewed-on: https://chromium-review.googlesource.com/1026165
Commit-Queue: Robert Ma <robertma@chromium.org>
Reviewed-by: Quinten Yearsley <qyearsley@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553324}

This patch uses OffTheRecord profiles to implement
|Target.createBrowserContext| and |Target.diposeBrowserContext| methods
for non-headless mode.

R=dgozman
BUG=836272

Change-Id: Ibd8b33bd9c804b69f17ef9cf1283b8c4bfc405ac
Reviewed-on: https://chromium-review.googlesource.com/1023591
Commit-Queue: Andrey Lushnikov <lushnikov@chromium.org>
Reviewed-by: Stefan Kuhne <skuhne@chromium.org>
Reviewed-by: Dmitry Gozman <dgozman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553323}

BUG=833380
TBR=mcnee@chromium.org
NOTRY=true

Change-Id: Ia38901113736af2fa923c301769282e834d531c3
Reviewed-on: https://chromium-review.googlesource.com/1026830Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553322}

Migrate ui/events/ozone/evdev to the new callback types as appropriate.

Bug: 714018
Test: CQ & run chrome on eve
Change-Id: Ic518057debb61cd6dbe297bebb13fec0538063c6
Reviewed-on: https://chromium-review.googlesource.com/1026523
Commit-Queue: Michael Spang <spang@chromium.org>
Reviewed-by: Kevin Schoedel <kpschoedel@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553321}

onTextEditorContentSet is a legacy method from when SourceFrame had
many subclasses. Overriding setContent instead is simpler and makes
for a cleaner plugin story around dispose/ensure plugins.

As a drive-by, the currentSearchResultIndex getter was removed.

Change-Id: I7f467da9297288b9f654eaf1cdd73b8be281cbaa
Reviewed-on: https://chromium-review.googlesource.com/1020812
Commit-Queue: Andrey Lushnikov <lushnikov@chromium.org>
Reviewed-by: Andrey Lushnikov <lushnikov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553320}

Bug: 778043
Change-Id: I7f80eb1553eb48456441d5ccab6be903875bbcff
Reviewed-on: https://chromium-review.googlesource.com/1023033
Commit-Queue: Andrey Lushnikov <lushnikov@chromium.org>
Reviewed-by: Andrey Lushnikov <lushnikov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553319}

This allows more easy testing of the feature under development/
experimentation.

Bug: 681921
Change-Id: Ibdf822b04312e794ec2a20051d02f1847275a9f9
Reviewed-on: https://chromium-review.googlesource.com/1024831
Commit-Queue: Mikel Astiz <mastiz@chromium.org>
Reviewed-by: Marc Treib <treib@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553318}

This CL is a refactor with no intended behavior change.

It updates the semantics to explicitly take ownership of the WebContents.

Bug: 832879
Change-Id: I62a6945aada916ff9538679934b2b767ea3adeb3
Reviewed-on: https://chromium-review.googlesource.com/1026311
Commit-Queue: Erik Chen <erikchen@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553317}

On a slow connection, if user taps Request Mobile/Desktop Site in rapid
succession, NavigationManagerImpl::ReloadWithUserAgentType() can be
called before the reload redirect completes, and |reload_url| will be
the restore_session.html redirect URL in this case. Added code to catch
this case and reload the original target URL.

Tested on GIN-3g network.

Bug: 833958
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: I958dde59f8bfc3469948aedfa83f5d5110f5ca40
Reviewed-on: https://chromium-review.googlesource.com/1026693Reviewed-by: Eugene But <eugenebut@chromium.org>
Commit-Queue: Danyao Wang <danyao@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553316}

When a conenction has path degrading timer fired, the connection will remember
that it's path degrading until forward progress is made, i.e., new ack has
been received on the degrading path. It will then spin timer to detect future
path degrading.

Flag protected by a new flag --quic_path_degrading_alarm2.

Manual merge internal change: 193990670

Bug: 835444
Change-Id: Idf487fafb10ccd61551f4a9905714ff83bd91cb9
Reviewed-on: https://chromium-review.googlesource.com/1025143Reviewed-by: Ryan Hamilton <rch@chromium.org>
Commit-Queue: Zhongyi Shi <zhongyi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553315}

This CL whitelists indexed and named property getters of common
collections with [Affects] extended attribute. This enables:
- div.classList[0]
- div.attributes.specific_attribute

to evaluate with throwOnSideEffect: true.

Bug: 829571
Change-Id: I64fb6ad154c14e0119d3b09870f4c9dd10c9de41
Reviewed-on: https://chromium-review.googlesource.com/1025256
Commit-Queue: Erik Luo <luoe@chromium.org>
Reviewed-by: Kentaro Hara <haraken@chromium.org>
Reviewed-by: Yuki Shiino <yukishiino@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553314}

PrefHashBrowserTestBase tests have been updated to manifest v2.
Remove the allowance for manifest v1 extensions.

Bug: 816679

Change-Id: Idfa6c27efac318842ad5169370e2073d02edb615
Reviewed-on: https://chromium-review.googlesource.com/1014561Reviewed-by: Gabriel Charette <gab@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553313}

Fixes this warning log:
[0411/153122.161864:WARNING:resource_bundle.cc(565)] locale resources are not loaded

Also shows "This plugin is not supported" when visiting websites with flash content.

Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: I33e7e65dcb8193a93f5208f847a39a20e8472987
Bug: 811603
Reviewed-on: https://chromium-review.googlesource.com/1024488Reviewed-by: Jochen Eisinger <jochen@chromium.org>
Reviewed-by: Mike Dougherty <michaeldo@chromium.org>
Commit-Queue: John Wu <jzw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553312}

Involved:
- Creating UI element classes.
- Combining Card/Text containers into single UiElementContainer.
- Adding support for multiple cards.
- Implementing queue to pend UI elements to ensure ordering. This is
  necessary as cards are rendered asynchronously before being added to
  the view hierarchy.

Also adds a background to text elements.

See bug for further details.

Bug: b:78130077
Change-Id: Ic10517de487de2e8aafe649062e8da56234557b5
Reviewed-on: https://chromium-review.googlesource.com/1019974Reviewed-by: Xiaohui Chen <xiaohuic@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Commit-Queue: David Black <dmblack@google.com>
Cr-Commit-Position: refs/heads/master@{#553311}

Use NOTIMPLEMENTED_LOG_ONCE().

Change-Id: I671c321331d1ef10e07d22d31e8bfd5312783129
Reviewed-on: https://chromium-review.googlesource.com/1026431Reviewed-by: Ken Buchanan <kenrb@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553310}

These changes were scripted as part of the effort to restrict exposure of the
raw MessageLoop*.

Note that many static methods exposed on MessageLoopCurrent are also deprecated
so it is very well possible that this is migrating to an already deprecated
method.

The goal of this pass is to reduce usage of MessageLoop::current() (and
ultimately make it impossible to obtain the MessageLoop* statically).

As such I will not edit this CL unless the script did something logically wrong.
I defer to owners to fix highlighted usage of already deprecated APIs.

Possible script screw ups / things to look out for in this review:
 - Storing MessageLoopCurrent in a MessageLoop* variable or comparing it against
   one (I will go over that in a separate pass).

Includes should have been stripped if that was the last usage of message_loop.h
in that file.

Please CQ if LGTY

This CL was uploaded by git cl split.

R=dtrainor@chromium.org

Bug: 825327
Change-Id: I6dcd9290b2c6021b914375e8f7fb03065f6299ce
Reviewed-on: https://chromium-review.googlesource.com/1024871Reviewed-by: David Trainor <dtrainor@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553309}

These changes were scripted as part of the effort to restrict exposure of the
raw MessageLoop*.

Note that many static methods exposed on MessageLoopCurrent are also deprecated
so it is very well possible that this is migrating to an already deprecated
method.

The goal of this pass is to reduce usage of MessageLoop::current() (and
ultimately make it impossible to obtain the MessageLoop* statically).

As such I will not edit this CL unless the script did something logically wrong.
I defer to owners to fix highlighted usage of already deprecated APIs.

Possible script screw ups / things to look out for in this review:
 - Storing MessageLoopCurrent in a MessageLoop* variable or comparing it against
   one (I will go over that in a separate pass).

Includes should have been stripped if that was the last usage of message_loop.h
in that file.

Please CQ if LGTY

This CL was uploaded by git cl split.

R=eroman@chromium.org

Bug: 825327
Change-Id: I981a208163dbff6560debcbdcd2dfb70f64ef717
Reviewed-on: https://chromium-review.googlesource.com/1024406Reviewed-by: Eric Roman <eroman@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553308}

These changes were scripted as part of the effort to restrict exposure of the
raw MessageLoop*.

Note that many static methods exposed on MessageLoopCurrent are also deprecated
so it is very well possible that this is migrating to an already deprecated
method.

The goal of this pass is to reduce usage of MessageLoop::current() (and
ultimately make it impossible to obtain the MessageLoop* statically).

As such I will not edit this CL unless the script did something logically wrong.
I defer to owners to fix highlighted usage of already deprecated APIs.

Possible script screw ups / things to look out for in this review:
 - Storing MessageLoopCurrent in a MessageLoop* variable or comparing it against
   one (I will go over that in a separate pass).

Includes should have been stripped if that was the last usage of message_loop.h
in that file.

Please CQ if LGTY

This CL was uploaded by git cl split.

R=sky@chromium.org

Bug: 825327
Change-Id: I26056af155629353a7a8e0fccf38fc3d2308ecd9
Reviewed-on: https://chromium-review.googlesource.com/1024792Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553307}

These changes were scripted as part of the effort to restrict exposure of the
raw MessageLoop*.

Note that many static methods exposed on MessageLoopCurrent are also deprecated
so it is very well possible that this is migrating to an already deprecated
method.

The goal of this pass is to reduce usage of MessageLoop::current() (and
ultimately make it impossible to obtain the MessageLoop* statically).

As such I will not edit this CL unless the script did something logically wrong.
I defer to owners to fix highlighted usage of already deprecated APIs.

Possible script screw ups / things to look out for in this review:
 - Storing MessageLoopCurrent in a MessageLoop* variable or comparing it against
   one (I will go over that in a separate pass).

Includes should have been stripped if that was the last usage of message_loop.h
in that file.

Please CQ if LGTY

This CL was uploaded by git cl split.

R=haraken@chromium.org

Bug: 825327
Change-Id: Ifc815fd5b8c80c62144137d5a1bdf4c6d8444f85
Reviewed-on: https://chromium-review.googlesource.com/1024438Reviewed-by: Kentaro Hara <haraken@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553306}

This reverts commit 54c93b85.

Reason for revert: Multiple nontrivial failures in FYI https://ci.chromium.org/buildbot/chromium.fyi/Fuchsia/16409

Original change's description:
> Reland "Fuchsia: remove "shell" capability and add dynamic libraries to packages."
> 
> Removing the "shell" capability allows packaged apps to load dynamic libraries
> from the package. In addition, it prevents the app from using any
> system-provided libraries, in favor of the package supplying all of its
> data and dynamic library dependencies.
> 
> This CL includes the system libraries from the SDK's "dist" directory in
> packages built using the package() GN template.
> 
> 
> Bug: 823927
> Change-Id: Ibc22913658f73d992e563e829a377be66f61f805
> Reviewed-on: https://chromium-review.googlesource.com/1025046
> Reviewed-by: Scott Graham <scottmg@chromium.org>
> Commit-Queue: Kevin Marshall <kmarshall@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#552950}

TBR=kmarshall@chromium.org,scottmg@chromium.org

Change-Id: I1b0b9517c878d1c4483efbba3e46502d875b194f
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 823927
Reviewed-on: https://chromium-review.googlesource.com/1026811Reviewed-by: Kevin Marshall <kmarshall@chromium.org>
Commit-Queue: Kevin Marshall <kmarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553305}

Bug: none
Change-Id: If1cf7b3d30bf86a408d9b7efb3c5a3e16d570e34
Reviewed-on: https://chromium-review.googlesource.com/1026518
Commit-Queue: James Cook <jamescook@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553304}

When computing the scroll-chain for a gesture, only Elements can be
added to the chain. However, the document element (i.e. <html>) doesn't
have a ScrollableArea itself. Its scrolling is delegated to the
document *node* (i.e. the LayoutView) which cannot be added to the
chain. Thus, we use the documentElement to represent scrolling the
LayoutView. See ScrollManager::RecomputeScrollChain to see this
explicitly in code.

When actually performing the scroll we need to reverse the above
substitution but we currently have a bug: in Element::NativeApplyScroll
we apply this substitution if we're scrolling the effective root
scroller rather than the document element. If the root scroller API is
enabled and the root scroller is changed to a different element, the
document can no longer be scrolled since it will fail to reverse the
documentElement<->LayoutView substitution and try to use the document
element's ScrollableArea which doesn't exist.

A similar fix is needed in LogicalScroll for keyboard scrolls. This
complicated logic is encoded in the scroll-chain calculation so I've
rewritten that method to use the scroll chain for consistency.

An additional change is needed since LayoutView::Scroll uses
LocalFrameView::GetScrollableArea to scroll. In the main frame, this
will be the RootFrameViewport which uses the global root scroller as
the layout viewport. Thus, if the root scroller is set to another
element, scrolling the LayoutView would actually scroll that element.
The global root scroller will always be scrolled using the correct
ScrollableArea via the ViewportScrollCallback so LayoutView shouldn't
be treated specially in this case. In fact, I've removed
LayoutBox::Scroll and its overrides altogether and replaced its
call sites with ScrollableArea directly.

Bug: 813747
Change-Id: I9e9a27ff684abf969c27ac152c860778a7e00e39
Reviewed-on: https://chromium-review.googlesource.com/1018880Reviewed-by: Steve Kobes <skobes@chromium.org>
Commit-Queue: David Bokan <bokan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553303}

These changes were scripted as part of the effort to restrict exposure of the
raw MessageLoop*.

Note that many static methods exposed on MessageLoopCurrent are also deprecated
so it is very well possible that this is migrating to an already deprecated
method.

The goal of this pass is to reduce usage of MessageLoop::current() (and
ultimately make it impossible to obtain the MessageLoop* statically).

As such I will not edit this CL unless the script did something logically wrong.
I defer to owners to fix highlighted usage of already deprecated APIs.

Possible script screw ups / things to look out for in this review:
 - Storing MessageLoopCurrent in a MessageLoop* variable or comparing it against
   one (I will go over that in a separate pass).

Includes should have been stripped if that was the last usage of message_loop.h
in that file.

Please CQ if LGTY

This CL was uploaded by git cl split.

R=alexst@chromium.org

Bug: 825327
Change-Id: I4a356b3e6ba95ac0e6e500225ca6576519416623
Reviewed-on: https://chromium-review.googlesource.com/1024761Reviewed-by: Alex Sakhartchouk <alexst@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Commit-Queue: Michael Spang <spang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553302}

Change-Id: Ib33e2722e02d84e043f4d1b962490a64d6eee464
Reviewed-on: https://chromium-review.googlesource.com/1026210
Commit-Queue: Steven Valdez <svaldez@chromium.org>
Reviewed-by: David Benjamin <davidben@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553301}

This cl creates the windows platform for ozone.

These changes are not sufficient for a successful Windows/Ozone build.
See: https://chromium-review.googlesource.com/c/chromium/src/+/974285
and https://swiftshader-review.googlesource.com/c/SwiftShader/+/17968

Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel
Change-Id: If27cd7b56d5503db6881cf1f2c563826f72f0be0
Reviewed-on: https://chromium-review.googlesource.com/975902Reviewed-by: Robert Kroeger <rjkroege@chromium.org>
Commit-Queue: Mustafa Çamurcu <camurcu@google.com>
Cr-Commit-Position: refs/heads/master@{#553300}

BUG=836426
TBR=skym@chromium.org
NOTRY=true

Change-Id: Ie9c07c1417ab693d517e1ac4ee3dc8082465a58a
Reviewed-on: https://chromium-review.googlesource.com/1026790Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553299}

Bug: 776877
Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel
Change-Id: Ic3263a6a8701cb9f99bcae657f041e624dc8ca01
Reviewed-on: https://chromium-review.googlesource.com/1011674
Commit-Queue: Peng Huang <penghuang@chromium.org>
Reviewed-by: Antoine Labour <piman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553298}

Saw this when investigating 835741, and I think this makes the code
confusing that it appears like we need to run these ops when in
reality they don't do anything.

Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: Ia361cacc4c3760bc5caf2653b0a55b84ce4fdee9
Reviewed-on: https://chromium-review.googlesource.com/1024671Reviewed-by: Khushal <khushalsagar@chromium.org>
Commit-Queue: enne <enne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553297}

In RLS mode, the position of an iframe LayoutView's main GraphicsLayer
must account for any offset from the border or outline styling of the
<iframe> element.

Previously there were two different codepaths for applying this offset:
(1) LayoutView's CLM in the child frame's compositing update
(2) LayoutIFrame's CLM in the parent frame's compositing update

If both frames are invalidated at the same time, path (1) breaks,
because the parent document's lifecycle state disallows the child
frame's compositing queries.  Having two paths is also error-prone.
This patch removes (1) in favor of using (2) exclusively.

Bug: 825775
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Change-Id: I00cf19be083081315e7918efc1ebd2af00cae906
Reviewed-on: https://chromium-review.googlesource.com/1020146
Commit-Queue: Steve Kobes <skobes@chromium.org>
Reviewed-by: David Bokan <bokan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553296}

When a view fails an accessibility check, it can be very helpful to find
out what line of code constructed it. This provides information on how
to patch in code that does that.

We may want to build this in at some point:
See https://crrev.com/c/1025637 or https://crrev.com/c/963284.

Bug: 811293
Change-Id: I14046d7d40a5475e28691d371acdcb6538b66663
Reviewed-on: https://chromium-review.googlesource.com/1026055Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
Commit-Queue: Aaron Leventhal <aleventhal@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553295}

Launching a process is racy with browser shutdown, and blocks it. So we
see shutdown hangs when process launch takes a long time. This CL makes
the process launcher use a CONTINUE_ON_SHUTDOWN instead of
BLOCK_SHUTDOWN task on Windows where it's presumably safe to do so given
no non-leaky globals are used, and based on local testing. This is not
safe on Linux since the process launcher there uses SandboxHostLinux
which is a non-leaky base::Singleton.

R=brucedawson,gab
BUG=830954

Change-Id: Ia8c5295924288819dcfdcc087d282a05e6ee7e55
Reviewed-on: https://chromium-review.googlesource.com/1020720Reviewed-by: Antoine Labour <piman@chromium.org>
Reviewed-by: Bruce Dawson <brucedawson@chromium.org>
Reviewed-by: Gabriel Charette <gab@chromium.org>
Commit-Queue: Sunny Sachanandani <sunnyps@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553294}

These changes were scripted as part of the effort to restrict exposure of the
raw MessageLoop*.

Note that many static methods exposed on MessageLoopCurrent are also deprecated
so it is very well possible that this is migrating to an already deprecated
method.

The goal of this pass is to reduce usage of MessageLoop::current() (and
ultimately make it impossible to obtain the MessageLoop* statically).

As such I will not edit this CL unless the script did something logically wrong.
I defer to owners to fix highlighted usage of already deprecated APIs.

Possible script screw ups / things to look out for in this review:
 - Storing MessageLoopCurrent in a MessageLoop* variable or comparing it against
   one (I will go over that in a separate pass).

Includes should have been stripped if that was the last usage of message_loop.h
in that file.

Please CQ if LGTY

This CL was uploaded by git cl split.

R=lambroslambrou@chromium.org

Bug: 825327
Change-Id: Ic3627d36d9e562b112766c1d597351caae7c0718
Reviewed-on: https://chromium-review.googlesource.com/1024284Reviewed-by: Joe Downing <joedow@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553293}

These changes were scripted as part of the effort to restrict exposure of the
raw MessageLoop*.

Note that many static methods exposed on MessageLoopCurrent are also deprecated
so it is very well possible that this is migrating to an already deprecated
method.

The goal of this pass is to reduce usage of MessageLoop::current() (and
ultimately make it impossible to obtain the MessageLoop* statically).

As such I will not edit this CL unless the script did something logically wrong.
I defer to owners to fix highlighted usage of already deprecated APIs.

Possible script screw ups / things to look out for in this review:
 - Storing MessageLoopCurrent in a MessageLoop* variable or comparing it against
   one (I will go over that in a separate pass).

Includes should have been stripped if that was the last usage of message_loop.h
in that file.

Please CQ if LGTY

This CL was uploaded by git cl split.

R=rockot@chromium.org

Bug: 825327
Change-Id: If17acb98d53b8a6a352a0c09699ddb60915393d0
Reviewed-on: https://chromium-review.googlesource.com/1024808Reviewed-by: Ken Rockot <rockot@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553292}

BUG=chromium:829085
TEST=Tested manually on device

Cq-Include-Trybots: master.tryserver.chromium.linux:closure_compilation
Change-Id: I9353962f50ba3ea78abe8251e4ec3c2f140cb432
Reviewed-on: https://chromium-review.googlesource.com/1008967
Commit-Queue: David Valleau <valleau@chromium.org>
Reviewed-by: Sean Kau <skau@chromium.org>
Reviewed-by: Michael Giuffrida <michaelpg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#553291}