GitLab

Bug: 831648
Change-Id: I1154abdb22295cac7f396dd61c2c4f833b56dccf
Reviewed-on: https://chromium-review.googlesource.com/1093988
Commit-Queue: Sky Malice <skym@chromium.org>
Reviewed-by: Patrick Noland <pnoland@chromium.org>
Reviewed-by: Filip Gorski <fgorski@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567305}

This CL adds a DCHECK to make sure that PaintTiming::ReportSwapTime is
called on the main thread. This is the method that should be called
after a swap promise is concluded. This check is added in preparation of
the change to PerformanceEntry strings to AtomicString.

Bug: 852050
Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_slimming_paint_v2;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: Idc39c69b155c6e90fec32a3b3c0fae7fa987c378
Reviewed-on: https://chromium-review.googlesource.com/1100971Reviewed-by: Yoav Weiss <yoav@yoav.ws>
Commit-Queue: Nicolás Peña Moreno <npm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567304}

https://pdfium.googlesource.com/pdfium.git/+log/8e631772e342..e16ffd4fc3f2


git log 8e631772e342..e16ffd4fc3f2 --date=short --no-merges --format='%ad %ae %s'
2018-06-14 art-snake@yandex-team.ru Simplify the calculation of Pages Offsets within CPDF_HintsTable.


Created with:
  gclient setdep -r src/third_party/pdfium@e16ffd4fc3f2

The AutoRoll server is located here: https://pdfium-roll.skia.org

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

If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.


TBR=dsinclair@chromium.org

Change-Id: I8356eb7aa7cb1496ea97cdb801a874a891ff6f9e
Reviewed-on: https://chromium-review.googlesource.com/1101038Reviewed-by: pdfium-chromium-autoroll <pdfium-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: pdfium-chromium-autoroll <pdfium-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#567303}

PowerMonitorBroadcastSource suffers from raciness in that it can
procress incoming Mojo messages on one thread while both it and the
process-wide PowerMonitor instance are being destroyed on another
thread. A previous CL introduced a mitigation for this problem by
shutting down the processing of incoming Mojo messages in the
PowerMonitorBroadcastSource destructor before proceeding with
destruction (https://chromium-review.googlesource.com/1041215). However,
that turns out not to be sufficient to eliminate the race:
the PowerMonitorBroadcastSource instance can end up waiting for the
lock in its destructor while on another thread it is holding the lock
to process an incoming Mojo message. That second thread can then call
back into the process-wide PowerMonitor instance while it is being
torn down, resulting in crashes due to undefined state (specifically,
the PowerMonitor instance has already set its PowerMonitorSource
instance to nullptr, resulting in a crash when that instance is being
accessed -- ironically, the very instance that is currently blocked
in its destructor waiting for the processing of the incoming Mojo
message to finish :).

This CL removes the possibility of that race on shutdown by introducing
a two-phase shutdown process for PowerMonitorSource. In PowerMonitor's
destructor, it first invokes PowerMonitorSource::Shutdown() before
proceeding. The contract of PowerMonitorSource::Shutdown() is that
subclasses must take any necessary action to ensure that on return
from that method, they will no longer call into PowerMonitor.
PowerMonitorBroadcastSource fulfills this contract by stopping the
process of incoming Mojo messages in
PowerMonitorBroadcastSource::Shutdown(). In the above scenario,
the broadcast source will end up waiting for the processing of the
incoming Mojo message to finish *inside*
PowerMonitorBroadcastSource::Shutdown(), which is safe by design as it
is before PowerMonitor has done any teardown.

Note that it is possible that there are still other kinds of
raciness between PowerMonitorBroadcastSource and PowerMonitor; as I
expressed on
https://bugs.chromium.org/p/chromium/issues/detail?id=834312#c33, I am
not confident about the overall threading model here.

TBR=xjz@chromium.org, mmenke@chromium.org

Bug: 809031
Change-Id: I9d0f291995b8185e16dff1f88564b6e5344439b2
Reviewed-on: https://chromium-review.googlesource.com/1098932Reviewed-by: Matt Menke <mmenke@chromium.org>
Reviewed-by: Sunny Sachanandani <sunnyps@chromium.org>
Reviewed-by: Nico Weber <thakis@chromium.org>
Commit-Queue: Colin Blundell <blundell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567302}

Use the compositor frame ack instead of begin-frame callback to
trigger these callbacks. The frame ack is sufficient to prevent
frames from being dropped, and provides minimal latency
back-pressure for clients that implement more advanced
scheduling.

This makes it possible to improve scheduling of frames for
Linux apps significantly. Sommelier can have the next frame
prepared as a dmabuf and ready to commit when we receive the
callback. This minimizes the chance that we'll miss a frame.

BUG=845659
TEST=exo_unittests --gtest_filter=SurfaceTest.RequestFrameCallback
TEST=sommelier /usr/lib/weston/weston-simple-shm
TEST=sommelier -X glxgears

Change-Id: I20f06618d9d550ab501f3945b3d25d4a5c20c218
Reviewed-on: https://chromium-review.googlesource.com/1097776Reviewed-by: Daniele Castagna <dcastagna@chromium.org>
Commit-Queue: David Reveman <reveman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567301}

Handle all combinations of:
-reading cookies via JavaScript or by inspecting HTTP request headers
-setting cookies via JavaScript or by HTTP response headers

Bug: 789632
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_mojo
Change-Id: I9473c9b0b514ada40ac36e07e4df2dc7e44613e1
Reviewed-on: https://chromium-review.googlesource.com/1099482Reviewed-by: Jochen Eisinger <jochen@chromium.org>
Commit-Queue: John Abd-El-Malek <jam@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567300}

This change fixes a https://crbug.com/851937 where users are unable to
clear auto-opening downloads settings.

Bug: chromium:851937
Test: verified manually
Cq-Include-Trybots: master.tryserver.chromium.linux:closure_compilation
Change-Id: I263ef9497a8a9b298a3514404fc3f31a61393687
Reviewed-on: https://chromium-review.googlesource.com/1099693Reviewed-by: Steven Bennetts <stevenjb@chromium.org>
Reviewed-by: Demetrios Papadopoulos <dpapad@chromium.org>
Reviewed-by: Zentaro Kavanagh <zentaro@chromium.org>
Commit-Queue: Bailey Berro <baileyberro@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567299}

Adds a condition to history_tab_helper to not return if navigation_context->IsSameDocument YES
and navigation hasn't been commited.

Bug: 843529
Cq-Include-Trybots: luci.chromium.try:ios-simulator-full-configs;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: I19bd101b975c784270e7af61852ad77d9ce7f2a1
Reviewed-on: https://chromium-review.googlesource.com/1098137Reviewed-by: Eugene But <eugenebut@chromium.org>
Commit-Queue: Sergio Collazos <sczs@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567298}

This changes how snapshot-requests via Page.captureScreenshot in
devtools are handled. Instead of notifying the browser directly from the
display-compositor, the renderer requests a presentation-time for the
new force-redraw frame. Upon receipt of the presentation time, renderer
notifies the browser, which can then capture the screenshot (either via
chrome, or directly from the underlying platform window).

This allows removing the snapshots from LatencyInfo, which was an odd
fit. This also enables the gpu and display-compositor code to be unaware
of snapshot-requests.

BUG=851504

Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel
Change-Id: I16d8cb070ce883e95ac40ed5fd05ebaf3e96f0b1
Reviewed-on: https://chromium-review.googlesource.com/1095562Reviewed-by: Ken Buchanan <kenrb@chromium.org>
Reviewed-by: Antoine Labour <piman@chromium.org>
Reviewed-by: Timothy Dresser <tdresser@chromium.org>
Commit-Queue: Sadrul Chowdhury <sadrul@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567297}

https://chromium.googlesource.com/angle/angle.git/+log/d668be9ce66f..f6e160fa9ca5


git log d668be9ce66f..f6e160fa9ca5 --date=short --no-merges --format='%ad %ae %s'
2018-06-14 lucferron@chromium.org Vulkan: Don't use swizzle state when its not needed


Created with:
  gclient setdep -r src/third_party/angle@f6e160fa9ca5

The AutoRoll server is located here: https://angle-chromium-roll.skia.org

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

If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.

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
TBR=cwallez@chromium.org

Change-Id: Id3528b97656dcab53cbf3c4faff02677b9d34fb8
Reviewed-on: https://chromium-review.googlesource.com/1101037Reviewed-by: angle-chromium-autoroll <angle-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: angle-chromium-autoroll <angle-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#567296}

This CL coalesces functions PerformCollectionTask(), RecordSample(), and
ScheduleNextSample() of the SamplingThread class into a new class member
function named RecordSampleTask(). Redundant code is removed.

Bug: 851163
Change-Id: I0ec209a7bebaf295e213ff7e321e52af60465bf8
Reviewed-on: https://chromium-review.googlesource.com/1100047Reviewed-by: Mike Wittman <wittman@chromium.org>
Commit-Queue: Xi Cheng <chengx@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567295}

This reverts commit f17419f4.

Reason for revert: Microsoft responded to twitter reports and fixed
their detection so this isn't needed anymore. I'm retaining the typo
fix though!

Original change's description:
> Modify markdown to satisfy Windows Defender
> 
> debugging_gpu_related_code.md explains how to use a Chrome command-line
> switch which has been used for arbitrary code execution. Windows
> Defender therefore treats it as malware. This means that some Windows
> Chromium developers will get anti-virus warnings (with mandatory
> quarantining of the file). If Chrome is in a directory that is
> excluded from scanning then they will avoid warnings but will hit
> problems if they try to use Windows backup.
> 
> This is all very silly since there is no way - short of manually
> following the steps, modify to be malicious - to activate this payload.
> Windows Defender even complains about the file created by this:
> 
>   > echo --no-sandbox --gpu-launcher="x" >foo.txt
> 
> But, sometimes it's not worth arguing.
> 
> This also fixes a typo.
> 
> R=​kainino@chromium.org
> BUG=851562
> 
> Change-Id: I85403a1cb1667f45784684179927119058608d40
> Reviewed-on: https://chromium-review.googlesource.com/1096311
> Commit-Queue: Bruce Dawson <brucedawson@chromium.org>
> Reviewed-by: Kai Ninomiya <kainino@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#566245}

TBR=brucedawson@chromium.org,kainino@chromium.org

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

Bug: 851562
Change-Id: I60b09d8e6a698c1646f5c4bb0ecab16f999bfb9c
Reviewed-on: https://chromium-review.googlesource.com/1100255
Commit-Queue: Bruce Dawson <brucedawson@chromium.org>
Reviewed-by: Bruce Dawson <brucedawson@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567294}

This CL is a reland of crrev.com/c/1079834 and crrev.com/c/1086131
that got reverted (the second is just adding a few forgotten DCHECKs).
This is because  GpuArcVideoDecodeAccelerator sends empty client and
texture id vectors [0]; this CL handles gracefully these cases.

[0] https://cs.chromium.org/chromium/src/components/arc/video_accelerator/gpu_arc_video_decode_accelerator.cc?l=468

original cl description ------------------------------------------------

This CL moves part of the logic of VaapiPicture creation from VaVDA to
VaapiPictureFactory, passing the PictureBuffer wholesale since we use
most of it anyway.  The net result is less lines of code.

VaapiPicture and derived classes hold on to a |texture_id_| and a
|client_texture_id_|, the latter only used with |bind_image_cb_|:
- |texture_id_| cannot be zero, either in the tests or in reality, because
  is a GL texture id; this CL enforces this via a DCHECK() in ctor. That
  simplifies the body a few methods.
- |client_texture_id_| can be zero, and is always zero in the v_d_a_unittests,
  but it doesn't matter because |bind_image_cb_| is dummy for tests ([1]),
  so there's no point of enforcing it to be non zero.  Moreover, since
  it's a client-side texture id, it can be any number, so let's skip
  the confusing checks.

Sprinkled a couple of consts and also, there's no need to do
!callback.is_null() because CallbackBase provides bool(), so
s/!callback.is_null()/callback/

[1] https://cs.chromium.org/chromium/src/media/gpu/video_decode_accelerator_unittest.cc?sq=package:chromium&dr&g=0&l=452

Bug: 822346
Test: compiled+run simplechrome, v_d_a_unittests on eve. media_unittests.
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: I8b211d97e6bcb3f86149224a1c130ca6a24984e8
Reviewed-on: https://chromium-review.googlesource.com/1100535Reviewed-by: Daniele Castagna <dcastagna@chromium.org>
Reviewed-by: Hirokazu Honda <hiroh@chromium.org>
Commit-Queue: Miguel Casas <mcasas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567293}

https://webrtc.googlesource.com/src.git/+log/d1f970dc435a..724a97d08d35


git log d1f970dc435a..724a97d08d35 --date=short --no-merges --format='%ad %ae %s'
2018-06-14 nisse@webrtc.org Drop tools/gyp from dependencies.
2018-06-14 saza@webrtc.org Add ivoc@ and saza@ to audio_processing OWNERS
2018-06-14 magjed@webrtc.org Android: Add tests for VideoFrame.Buffer.toI420() and cropAndScale()


Created with:
  gclient setdep -r src/third_party/webrtc@724a97d08d35

The AutoRoll server is located here: https://webrtc-chromium-roll.skia.org

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

If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.

CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_chromium_archive_rel_ng;master.tryserver.chromium.mac:mac_chromium_archive_rel_ng

BUG=chromium:None
TBR=webrtc-chromium-sheriffs-robots@google.com

Change-Id: I61e9cbca6c322134bf4b9b5df9bc5dd44da8055d
Reviewed-on: https://chromium-review.googlesource.com/1101039Reviewed-by: webrtc-chromium-autoroll <webrtc-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: webrtc-chromium-autoroll <webrtc-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#567292}

Setting the plugin before attaching print-preview-app to the document
will no longer work in Polymer2. Add an interface for the plugin,
similar to the approach for the NativeLayer, and use this to stub out
checks for plugin compatibility and other plugin functions.

Bug: 738611
Cq-Include-Trybots: master.tryserver.chromium.linux:closure_compilation
Change-Id: I8428fba35ff5783e18679f99cc02f8d085f82a2e
Reviewed-on: https://chromium-review.googlesource.com/1097514
Commit-Queue: Rebekah Potter <rbpotter@chromium.org>
Reviewed-by: Demetrios Papadopoulos <dpapad@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567291}

This is a reland of 1ff37758

Original change's description:
> Multi-global WebUSB WPT for window and worker
>
> This change updates the existing web platform tests for WebUSB to use
> multi-global tests to test the API in the supported contexts. The
> change updates the following tests:
> * idlharness
> * usb
> * usbConnectionEvent
>
> Bug: 841510
> Change-Id: I6b118aa4121109cd6a48054c2ebc69b831660afc
> Reviewed-on: https://chromium-review.googlesource.com/1087209
> Reviewed-by: Reilly Grant <reillyg@chromium.org>
> Commit-Queue: Ovidio Henriquez <odejesush@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#564981}

Bug: 841510
Change-Id: Ie354fdd54a4cd94a4801ee634c8d3263e0aa6e93
Reviewed-on: https://chromium-review.googlesource.com/1089647
Commit-Queue: Ovidio Henriquez <odejesush@chromium.org>
Reviewed-by: Chong Zhang <chongz@chromium.org>
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567290}

The CSSWG resolved about it in the last meeting:
https://github.com/w3c/csswg-drafts/issues/2746

BUG=843329
TEST=external/wpt/css/css-contain/contain-size-012.html

Change-Id: Ibb037f9ab1b95bed03e46fdbeb0e68520ff741b4
Reviewed-on: https://chromium-review.googlesource.com/1100824Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Commit-Queue: Manuel Rego Casasnovas <rego@igalia.com>
Cr-Commit-Position: refs/heads/master@{#567289}

The test tries to symbolize Chromium Framework independently, so the outer executable's
rpath isn't on the stack. The Framework does have an rpath in component builds, but
it is currently os.path.join(5 * ['..']), which relative to
out/Release/Chromium.app/Contents/Versions/69.0.3456.0/Chromium Framework.framework/Versions/A
gets us to out/Release/Chromium.app/Contents.  That doesn't seem useful for anything
(from what I can tell, it's always been this way since at least the gn switch),
so use 7 * ['..'] instead to get to the build dir.

This possibly fixes a regression from the ancient https://codereview.chromium.org/2700013002

Bug: 850055
Change-Id: I9e65a7fa76620539b2492fd6096b70533a523a74
Reviewed-on: https://chromium-review.googlesource.com/1101064
Commit-Queue: Nico Weber <thakis@chromium.org>
Reviewed-by: Robert Sesek <rsesek@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567288}

TimeTicks are less confusing than double (which can be variously seconds
and milliseconds).

Change-Id: I70badbc63c8f8a1c8226f50b5ccb7d70d8209314
Reviewed-on: https://chromium-review.googlesource.com/1099370Reviewed-by: Kentaro Hara <haraken@chromium.org>
Commit-Queue: Jeremy Roman <jbroman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567287}

LocalDataContainer is owned by a unique_ptr in CookiesTreeModel but
it is created from the outside and passed as a raw pointer.
Fix this by directly creating it as a unique_ptr.

Change-Id: I11dadcd91359f2b07f8d021fa3daa62dde80711f
Reviewed-on: https://chromium-review.googlesource.com/1100832
Commit-Queue: Christian Dullweber <dullweber@chromium.org>
Reviewed-by: Bernhard Bauer <bauerb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567286}

TBR=rdevlin.cronin@chromium.org

Bug: 851655
Change-Id: Ie13d383c007d206aa120e2993fe4f93e4ee6e407
Reviewed-on: https://chromium-review.googlesource.com/1100901
Commit-Queue: François Doray <fdoray@chromium.org>
Reviewed-by: François Doray <fdoray@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567285}

Adds --manual-symbol-offsets, --manual-libname and --manual-objdir to
the orderfile generator, allowing custom profiling. For example, with
this one can easily collect profile information manual, process it with
phased_orderfile.py, and then use the resulting symbol lists to produce
a properly patched orderfile that is suitable for performance comparison
against the production orderfile.

Bug: 758566
Change-Id: I0e41319dfcd537cd7f076e0236dc96b853dd7a4f
Reviewed-on: https://chromium-review.googlesource.com/1095274Reviewed-by: Benoit L <lizeb@chromium.org>
Commit-Queue: Matthew Cary <mattcary@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567284}

Tests that can be expressed as LayoutTests (preferreably Web Platform
Tests) rather than browser tests should. Browser tests are slow to run
and a maintenance burden.

The WebRtcRtpBrowserTests are mainly testing behaviors of adding and
removing tracks/streams and asserting that the expected outcomes.

- The essentials of these tests are already covered in
  external/wpt/webrtc/ which we are actively maintaining and updating in
  preparation for Unified Plan and RTCRtpTransceiver support. Almost all
  of it is overlapping.
- The WebRtcRtpBrowserTests were not written to test behaviors, but
  functions. Each test is asserting a lot of unrelated behaviors.
- When running with Unified Plan/RTCRtpTransceiver (WIP CL:
  https://chromium-review.googlesource.com/c/chromium/src/+/1025771)
  these tests start failing. I started updating these tests but the
  failures were due to Plan B assumptions (problems in the tests, not in
  the implementation).
- Debugging this part of the code is slow and cumbersome.
- I won't miss any of these tests.

Conclusion: Nuke it.

Bug: 773472, 777617
Change-Id: I099992f1783914d4fb1d5c2d952ff977ec4cc513
Reviewed-on: https://chromium-review.googlesource.com/1100891Reviewed-by: Patrik Höglund <phoglund@chromium.org>
Commit-Queue: Henrik Boström <hbos@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567283}

https://skia.googlesource.com/skia.git/+log/f018b7d607a0..8760e2f42857


git log f018b7d607a0..8760e2f42857 --date=short --no-merges --format='%ad %ae %s'
2018-06-14 jvanverth@google.com Add outline concave shadow support


Created with:
  gclient setdep -r src/third_party/skia@8760e2f42857

The AutoRoll server is located here: https://autoroll.skia.org

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

If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;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
TBR=csmartdalton@chromium.org

Change-Id: I71a1f925f875fe77e05837d9b96eb0c6f6b00d79
Reviewed-on: https://chromium-review.googlesource.com/1100917Reviewed-by: skia-chromium-autoroll <skia-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: skia-chromium-autoroll <skia-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#567282}

Added LoadingErrorPage flag to NavigationContext to mark
LoadHTMLString: navigations which feed Error page to
WKWebView. These navigations will be ignored in
WKWebViewNavigationDelegate callbacks.

GoBackFromErrorPage and testGoBackFromErrorPage tests were
updated to cover this test case.

Bug: 725241
Cq-Include-Trybots: luci.chromium.try:ios-simulator-full-configs;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: I1a38685d96997328c4be2a250dcedb5837e76d77
Reviewed-on: https://chromium-review.googlesource.com/1093515Reviewed-by: Danyao Wang <danyao@chromium.org>
Commit-Queue: Eugene But <eugenebut@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567281}

A result looks crowded with a tab switch button. The spec asks that
we use the two line format if the tab switch button is present.

This CL also expands the button to 32px, per spec.

Bug: 780835
Change-Id: I372cb24646bdd448497f0bcf456d383a04a8a55b
Reviewed-on: https://chromium-review.googlesource.com/1087754
Commit-Queue: Justin Donnelly <jdonnelly@chromium.org>
Reviewed-by: Justin Donnelly <jdonnelly@chromium.org>
Reviewed-by: Dave Schuyler <dschuyler@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567280}

Bug: 810833
Change-Id: Ia001570d879a714c0002c2270ff8273bf046f508
Reviewed-on: https://chromium-review.googlesource.com/1097228Reviewed-by: Jeremy Roman <jbroman@chromium.org>
Reviewed-by: Ian Clelland <iclelland@chromium.org>
Commit-Queue: Luna Lu <loonybear@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567279}

This CL fixes an issue where if GPU process initialization fails and
then GPU process crashes we end up incrementing the wrong UMA histogram
and recent crash count. This happens because when the GPU process
initialization fails we do a GPU mode fallback, eg. disable hardware
aceleration or SwiftShader, which modifies values in
GpuDataManagerImplPrivate. When the GPU process crashes, it checks the
values in GpuDataManagerImplPrivate that were already changed.

For example, if hardware acceleration was tried and GPU process
initialization fails then hardware acceleration will be disabled from
GpuProcessHost::DidFailInitialize(). When the GPU process crashes,
GpuProcessHost::RecordProcessCrash() will increment the SwiftShader UMA
and recent crash count instead of hardware accelerated UMA and crash
count.

Add GpuMode which tracks this information and have GpuProcessHost store
this on initialization. This way the same UMA histograms are incremented
on launch and crash. Also make sure after DidFailInitialize() that
RecordProcessCrash() doesn't also trigger GPU mode fallback.  Move the
fallback logic into GpuDataManagerImplPrivate::GpuModeFallback() to
ensure it's the same for all cases.

Change Chrome OS to also LOG(FATAL) if GPU process crashes too many
times or fails to initialize. Chrome OS always crashed in this case, it
would just crash later in GpuProcessTransportFactory. Also increase max
crashes for Chrome OS from 3 to 6 to match Android.

Bug: 852063
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: I6870d0d877cdc2766d54428f90517c34cbd54343
Reviewed-on: https://chromium-review.googlesource.com/1099721
Commit-Queue: kylechar <kylechar@chromium.org>
Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Antoine Labour <piman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567278}

Move ShowSigninScreenForTest from SigninScreenHandler to GaiaView, which
is the gaia/login-specific interface.

Bug: 851012
Change-Id: If01d1052341b670ca30ffece6bbbc00e981bc61a
Reviewed-on: https://chromium-review.googlesource.com/1093376Reviewed-by: Achuith Bhandarkar <achuith@chromium.org>
Reviewed-by: Maksim Ivanov <emaxx@chromium.org>
Commit-Queue: Jacob Dufault <jdufault@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567277}

Bug: 852016
Cq-Include-Trybots: luci.chromium.try:ios-simulator-full-configs;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: I99082e8eee6f4bb001bc2e7c396088475a772217
Reviewed-on: https://chromium-review.googlesource.com/1100339
Commit-Queue: Lindsay Pasricha <lindsayw@chromium.org>
Reviewed-by: Justin Cohen <justincohen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567276}

Previously, there was a race condition. The ondemand update could be
requested before the component was registered.

Bug: 851666
Change-Id: I5d2162a87748f93bdf902f78c8cbe63292d65201
Reviewed-on: https://chromium-review.googlesource.com/1097705
Commit-Queue: Tibor Goldschwendt <tiborg@chromium.org>
Reviewed-by: Michael Thiessen <mthiesse@chromium.org>
Reviewed-by: Joshua Pawlicki <waffles@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567275}

The current logic in library_loader_hooks.cc for when to fork and
prefetch the native library has gotten complicated. This CL simplifies
it to current best practice: The prefetching is always done, restricted
to the ordered code if --orderfile-memory-optimization is set.

Bug: 758566
Change-Id: I86ecfff37b18eb7bda1edadfa5399db97b8be53e
Reviewed-on: https://chromium-review.googlesource.com/1095256Reviewed-by: Gabriel Charette <gab@chromium.org>
Reviewed-by: Benoit L <lizeb@chromium.org>
Commit-Queue: Matthew Cary <mattcary@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567274}

Before this CL the test was constructing URLLoaderInterceptor too late -
after the RenderFrameHostImpl under test has already been created.  This
CL moves the creation of URLLoaderInterceptor (wrapped in
RequestInterceptor) slightly earlier.  After this CL the test passes
with and without NetworkService.

Bug: 846341
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_mojo
Change-Id: I47fe27b55c341cdf87292a8aede8462921a9cbab
Reviewed-on: https://chromium-review.googlesource.com/1099971
Commit-Queue: Łukasz Anforowicz <lukasza@chromium.org>
Reviewed-by: Nasko Oskov <nasko@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567273}

Remove PageSchedulerImpl::CreateFrameSchedulerImpl in favour of
FrameSchedulerImpl::Create and
PageSchedulerImpl::RegisterFrameSchedulerImpl.
This will allow to create subclasses of FrameSchedulerImpl in tests.

R=alexclarke@chromium.org

Change-Id: I401596bcae945a0039a8737cf795c13a19dd35e2
Reviewed-on: https://chromium-review.googlesource.com/1099071
Commit-Queue: Alexander Timin <altimin@chromium.org>
Reviewed-by: Alex Clarke <alexclarke@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567272}

Bug: 852427
Change-Id: I6a397b8fa5e3d3013e69d7e55e50679dc27c875e
Reviewed-on: https://chromium-review.googlesource.com/1099178Reviewed-by: Jared Saul <jsaul@google.com>
Reviewed-by: Mathieu Perreault <mathp@chromium.org>
Commit-Queue: Moe Ahmadi <mahmadi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567271}

Adding some missing manifest and permission features that
ChromeVox needs on chromecast.

Bug: b/73383411
Test: None
Change-Id: I2d85e18fcec6dceb02dccec551838ebca418aaab
Reviewed-on: https://chromium-review.googlesource.com/1093272Reviewed-by: Alex Sakhartchouk <alexst@chromium.org>
Commit-Queue: Randy Rossi <rmrossi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567270}

This is a follow-up from
https://chromium-review.googlesource.com/c/chromium/src/+/1089633

This CL adds a helper function which wraps UCKeyTranslate along with
some param conversions needed to call into it in a new helper function.
I've also updated all of the callsites which used UCKeyTranslate
directly to use the new helper function.

During the original CR, there was a suggestion to put the helper
method in ui/base/cocoa.  I've kept it in ui/events as referring
to ui/base from within ui/events was problematic.  Also ui/base
seemed like the wrong place given that this helper does work with
low level key event data and not higher level constructs.

BUG=832811

Change-Id: Iba6c78c846637afd74eb0256ac7e65dc3508d97a
Reviewed-on: https://chromium-review.googlesource.com/1097976
Commit-Queue: Joe Downing <joedow@chromium.org>
Reviewed-by: Shuotao Gao <stgao@chromium.org>
Reviewed-by: ccameron <ccameron@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567269}

In libaom CMake configuration:
-Wenum-conversion and -Wparentheses-equality are enabled by default
in clang and not explicitly disabled.
-Wunused-function is from CMake defaults and no longer disabled.
-Wsign-compare is explicitly enabled.

The symbol "_vpx_scaled_2d" is no longer present, so the bug that
required "-fno-common" should be resolved.

Bug: 783519
Change-Id: I099b89c3b418fb2cd50bc19af167deaf71f12a6c
Reviewed-on: https://chromium-review.googlesource.com/1100370Reviewed-by: Johann Koenig <johannkoenig@google.com>
Commit-Queue: Johann Koenig <johannkoenig@google.com>
Cr-Commit-Position: refs/heads/master@{#567268}

TBR=dsinclair@chromium.org

Bug: 808152
Change-Id: I006300209602701c98a5ce35e9d504c1347d8915
Reviewed-on: https://chromium-review.googlesource.com/1101059Reviewed-by: François Doray <fdoray@chromium.org>
Commit-Queue: François Doray <fdoray@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567267}

- Add explanatory text to each one
- Make the manual test path accessible to color blind testers
- Bring headers up to date
- Create a consistent style for :focus-visible and :focus:not(:focus-visible) for easier understanding of test results

Bug: 817199
Change-Id: I97191d5c8395576a42c923a025d05d2d61e79552
Reviewed-on: https://chromium-review.googlesource.com/1093697
Commit-Queue: Alice Boxhall <aboxhall@chromium.org>
Reviewed-by: Jonathon Kereliuk <kereliuk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#567266}