GitLab

This is a reland of 088dc9df

GpuProcessHost::Get() is now called with force_create set to false.

BUG=877058
TBR=dgozman@chromium.org,kbr@chromium.org

Original change's description:
> DevTools: Restore GPU tasks recording in performance profiles.
>
> Add GPU process pid to the list of processes where tracing is enabled.
>
> Change-Id: Ic401ee4ea7d218cad36f687be226d092cd4e54a0
> Reviewed-on: https://chromium-review.googlesource.com/1182304
> Commit-Queue: Alexei Filippov <alph@chromium.org>
> Reviewed-by: Kenneth Russell <kbr@chromium.org>
> Reviewed-by: Dmitry Gozman <dgozman@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#585301}

Change-Id: I21d6bc8e28724ced518d37a128af6797addaa1c0
Reviewed-on: https://chromium-review.googlesource.com/1187085
Commit-Queue: Alexei Filippov <alph@chromium.org>
Reviewed-by: Alexei Filippov <alph@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586102}

This CL keeps OffscreenTabOwner in TabCapture API, but moves
OffscreenTab to chrome/browser/media to allow it being used by the
mirroring service to support OffscreenTab mirroring. Implementation is
not changed at all. No functional or behavior change either.

Bug: 734672
Change-Id: Ia8de3680853b86824809d70dc96c6180e9a0c957
Reviewed-on: https://chromium-review.googlesource.com/1180588
Commit-Queue: Xiangjun Zhang <xjz@chromium.org>
Reviewed-by: Lei Zhang <thestig@chromium.org>
Reviewed-by: Yuri Wiitala <miu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586101}

This reverts commit 4117b35a.

Reason for revert: The issue which was leading to the original crashes here has been addressed.

Original change's description:
> Sheriff: Mark animations/animation-shorthand-overriding.html etc with Crash on Android
> 
> Bug: 874737
> Change-Id: I11c23cf99defa75625cd14284b010e28bb297c72
> Tbr: nhiroki@chromium.org
> Notry: true
> Reviewed-on: https://chromium-review.googlesource.com/1177219
> Reviewed-by: Hiroki Nakagawa <nhiroki@chromium.org>
> Commit-Queue: Hiroki Nakagawa <nhiroki@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#583565}

TBR=nhiroki@chromium.org

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

Bug: 874737
Change-Id: I44fa371cdc4d2ee1b1d513484643372774c6943f
Reviewed-on: https://chromium-review.googlesource.com/1188495
Commit-Queue: Eric Karl <ericrk@chromium.org>
Reviewed-by: Eric Karl <ericrk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586100}

This is a reland of f933e705

Original change's description:
> Chrome OS: Add Discover window UI.
>
> This CL adds Chrome OS Discover UI in separate window.
>
> Window need further changes to match the specs, but this
> change mostly adds infrastructure.
>
> Bug: 864686
>
> Change-Id: I2834d8f9efb3660050a31768508992acb1f3f262
> Reviewed-on: https://chromium-review.googlesource.com/1182880
> Commit-Queue: Alexander Alekseev <alemate@chromium.org>
> Reviewed-by: Ilya Sherman <isherman@chromium.org>
> Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#586034}

TBR=isherman@chromium.org,xiyuan@chromium.org

Bug: 864686
Change-Id: I384e150da88aa60b09a5ed93b06403f2bab8e4eb
Reviewed-on: https://chromium-review.googlesource.com/1189166
Commit-Queue: Alexander Alekseev <alemate@chromium.org>
Reviewed-by: Alexander Alekseev <alemate@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586099}

This CL
  - Introduces url_pattern_index::kUrlPatternIndexFormatVersion to keep a track
    of the current format version of the UrlPatternIndex schema.
  - Adds checks to url_pattern_index clients (Declarative net request and
    subresource_filter) to ensure that their ruleset format versions are updated
    whenever kUrlPatternIndexFormatVersion is changed.
  - Adds presubmit checks to components/url_pattern_index to ensure that
    kUrlPatternIndexFormatVersion is kept updated.

BUG=877214, 755717

Change-Id: Ib82b775e3da2112d5f17a2040ec495a7ff8e18d9
Reviewed-on: https://chromium-review.googlesource.com/1187706
Commit-Queue: Karan Bhatia <karandeepb@chromium.org>
Reviewed-by: Charlie Harrison <csharrison@chromium.org>
Reviewed-by: Istiaque Ahmed <lazyboy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586098}

Observe the target WebContents for Tab/OffscreenTab mirroring.

Bug: 734672
Change-Id: I65194a345f81b38967396c80e5f614d4119782e4
Reviewed-on: https://chromium-review.googlesource.com/1176110Reviewed-by: Yuri Wiitala <miu@chromium.org>
Commit-Queue: Xiangjun Zhang <xjz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586097}

CAOS was refactored to simplify the logic for the service connector,
multithreading, and unit tests.

The CAOS::Backend object was renamed to CAOS::CmaWrapper, and much of
the implementation was moved out into CAOS to simplify the object.

The service connector is created in CastAudioManager and passed to
CAOS's audio thread. Instead of passing this then to the media thread
as done previously, we only utilize the connector in the audio thread.
The connector can be easily overriden for tests.

CAOS runs on two threads. It is constructed and controlled from
CastAudioManager in the audio thread. All CMA construction and control
happens in the media thread in the CAOS::CmaWrapper object.

The thread usage in the unit tests has been significantly simplified.
Instead of using RunLoops and multiple threads, we use a single thread
retrieved from a ScopedTaskEnvironment in CAM, CAOS:audio, and CAOS:media.
This means that everything in the unit tests are running on a single
thread, and the tests are far less flaky.

In order to allow everything to run on a single thread, the usages of
WaitableEvent were removed. WaitableEvent was previously used to allow
the synchronous behavior of the audio thread to work with the
asynchronous thread hopping. To compensate for this, certain task
postings between threads can be cached or cancelled, and errors are
returned late.

The following unit tests were added to verify the above behavior and
fill in untested implementation:
- CastAudioManagerTest
  .HasValidOutputStreamParameters
  .CanMakeStream
  .CanMakeMixerStream
- CastAudioOutputStreamTest
  .CloseWithoutStart
  .CloseWithoutStop
  .CloseCancelsOpen
  .CloseCancelsStart
  .CloseCancelsStop
  .StartImmediatelyAfterOpen
  .SetVolumeImmediatelyAfterOpen
  .StopCancelsStart
  .StopDoesNotCancelSetVolume
  .PushFrameAfterStop
  .PushFrameAfterClose
  .MultiroomInfo
  .SessionId

Bug: b/111993375
Test: cast_media_unittests, udon build
Change-Id: Ib7dce8df8a36a2311e13d6a0062b7e56d4928995
Reviewed-on: https://chromium-review.googlesource.com/1159824
Commit-Queue: Aidan Wolter <awolter@chromium.org>
Reviewed-by: Luke Halliwell <halliwell@chromium.org>
Reviewed-by: Kenneth MacKay <kmackay@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586096}

This CL adds functionality for two edge cases:
(1) If all sub-features are prohibited by policy, the Better Together
    suite should also be prohibited.
(2) If a feature would have been enabled except that the suite as a
    whole is disabled, the feature is now in the
    kUnavailableSuiteDisabled state.

Note: This CL also renames the kDisabledByPolicy enum value to be
kProhibitedByPolicy instead for clarity.

Bug: 870113, 824568
Change-Id: I11bdfcfbfd3313490599831606d31cb5ffa22983
Reviewed-on: https://chromium-review.googlesource.com/1187834
Commit-Queue: Kyle Horimoto <khorimoto@chromium.org>
Reviewed-by: Tommy Li <tommycli@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Ryan Hansberry <hansberry@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586095}

If services are launched without full browser process, they need to retrieve
the network service from the service manager connector they have.
This CL adds a helper method to allow that to work.

BUG=695115

Change-Id: I362b814b03d60d9645523664a7243555bac6f053
Reviewed-on: https://chromium-review.googlesource.com/1188776Reviewed-by: John Abd-El-Malek <jam@chromium.org>
Commit-Queue: Min Qin <qinmin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586094}

https://chromium.googlesource.com/angle/angle.git/+log/dc248f716822..8707af516c46


git log dc248f716822..8707af516c46 --date=short --no-merges --format='%ad %ae %s'
2018-08-24 syoussefi@google.com Add a test case for enabled-but-inactive vertex attributes
2018-08-24 syoussefi@google.com Disable vertex attributes if not active in program


Created with:
  gclient setdep -r src/third_party/angle@8707af516c46

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

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: Ied6acdbd1d1776e21827cc16e315f32360be4ef3
Reviewed-on: https://chromium-review.googlesource.com/1188926Reviewed-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@{#586093}

We're seeing crashes in GetContextOwner(), which is used to return an
identifier for a v8::Context that is constant across multiple contexts
for that same owner (e.g., an extension id). The crash is happening
at a failed CHECK:

ScriptContext* script_context = GetScriptContextFromV8ContextChecked(context);
const std::string& extension_id = script_context->GetExtensionID();
bool id_is_valid = crx_file::id_util::IdIsValid(extension_id);
CHECK(id_is_valid || script_context->url().is_valid());

Looking at the crash dumps, it appears that this crash happens with a
ScriptContext that has neither an associated extension nor a valid URL
(the URL is empty). The context type is for a Feature::WEB_PAGE_CONTEXT.

In theory, this should never happen, because contexts should only be
classified as a WEB_PAGE_CONTEXT with a valid URL. However, there is a
chance that it could, as described in https://crbug.com/877658. My
suspicion is that this happens during page navigation, when we are
loading/initializing a new context.

Both native and JS-based bindings use the same check for validity, but
it appears this crash only happens in native bindings. The primary
difference is that native bindings check a context owner's identity
immediately on event construction, whereas JS bindings do it lazily on
listener addition/removal.

In order to try and mitigate the crashes, change native bindings to
mimic JS bindings behavior, where the context identity is determined
lazily. This will not fix the underlying context issue, but should
hopefully address the crashes. If it doesn't, we'll have to go back to
the drawing board.

Bug: 877401

Change-Id: I8b29ec872b1fdb33f74b96bfa85e721b5e26b870
Reviewed-on: https://chromium-review.googlesource.com/1188764
Commit-Queue: Devlin <rdevlin.cronin@chromium.org>
Reviewed-by: Istiaque Ahmed <lazyboy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586092}

https://chromium.googlesource.com/catapult.git/+log/72712b0d6078..bc2c0a930728


git log 72712b0d6078..bc2c0a930728 --date=short --no-merges --format='%ad %ae %s'
2018-08-24 zmo@chromium.org Exclude a few more folders with large data from telemetry data set


Created with:
  gclient setdep -r src/third_party/catapult@bc2c0a930728

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

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

BUG=chromium:872521
TBR=sullivan@chromium.org

Change-Id: I905c391cc4d42a8e3d7334203aa1af701a1bbc66
Reviewed-on: https://chromium-review.googlesource.com/1188931Reviewed-by: catapult-chromium-autoroll <catapult-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: catapult-chromium-autoroll <catapult-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#586091}

The caller expects the decoder to populate memory in the allocator if
set, but the MockImageDecoder was not doing this. This causes failures
in MSAN where the memory remains unitinitialized.

Fix MockImageDecoder to respect this expectation.

R=chrishtr@chromium.org

Bug: 877497
Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_slimming_paint_v2;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: I441095eb3fb89c90cf3b9bf592438cd20b126391
Reviewed-on: https://chromium-review.googlesource.com/1188877
Commit-Queue: Khushal <khushalsagar@chromium.org>
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586090}

https://chrome-internal.googlesource.com/chrome/src-internal.git/+log/2916d264903a..28763cdcff7e


Created with:
  gclient setdep -r src-internal@28763cdcff7e

The AutoRoll server is located here: https://autoroll-internal.skia.org/r/src-internal-chromium-autoroll

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=mmoss@chromium.org

Change-Id: I9ab941a0a6d3ec57a6f5c3ea687396e7f4b9a826
Reviewed-on: https://chromium-review.googlesource.com/1189123Reviewed-by: src-internal-chromium-autoroll <src-internal-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: src-internal-chromium-autoroll <src-internal-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#586089}

Since Bluetooth power preference became user-specific, we disallowed
non-primary users to change Bluetooth power setting for UX consistency
with system tray.

Bug: 853374
Change-Id: If61d2c0c048e7d8c764e24798da0ebbf2ab6c68f
Reviewed-on: https://chromium-review.googlesource.com/1187601
Commit-Queue: Sonny Sasaka <sonnysasaka@chromium.org>
Reviewed-by: Michael Giuffrida <michaelpg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586088}

This CL adds ref count to a ServiceWorker while dispatching an
extension event to it. The ref count is decreased when we receive an
ack for worker thread, similar to vanilla extension events.

This CL adds the event ack IPC from worker thread to browser process.

The ref counting methods in //content (Starting/FinishedExternalRequest)
requires service worker version id within a SWContext, this CL adds
support for this:
  - in browser process: EventListeners now also include service
    worker version id.
  - in browser process: ServiceWorkerContext::StartWorkerForPattern
    returns version id along with existing information (render process
    id and worker thread id). ServiceWorkerTaskQueue uses it to
    dispatch event through EventRouter.
  - in worker thread: ScriptContext of a worker also stores the
    version id, used while sending event ack from WorkerThreadDispatcher.

Bug: 876207
Change-Id: Ia0d2df39091ac07504444523020fef9579e42696
Reviewed-on: https://chromium-review.googlesource.com/1175420
Commit-Queue: Istiaque Ahmed <lazyboy@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Matt Falkenhagen <falken@chromium.org>
Reviewed-by: Kinuko Yasuda <kinuko@chromium.org>
Reviewed-by: Devlin <rdevlin.cronin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586087}

When we seek a video being casted with the FlingingRenderer, time never
gets updated on the phone after the seek completes. This is because
Blink doesn't get the appropriate seek completion signals, since WMPI's
buffering state never goes back to BUFFERING_HAVE_ENOUGH.

This CL fixes issue by always returning BUFFERING_HAVE_ENOUGH in
FlingingRenderer::StartPlayingFrom(). Because there is no need to buffer
when using the FR, we can always consider the media pipeline to be fully
buffered. Sending a buffering state change in StartPlayingFrom() might
be suprising, but we do the same with the MediaPlayerRenderer, and it
has not caused any (known) problems so far.

Bug: 875517
Change-Id: If6069d501af12ec20f844d0adc1ea980935fb99f
Reviewed-on: https://chromium-review.googlesource.com/1184060
Commit-Queue: Thomas Guilbert <tguilbert@chromium.org>
Reviewed-by: Dale Curtis <dalecurtis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586086}

This should guard against a crash.
It's not clear exactly what the cause of this null is, but it's
probably some kind of failure to allocate the SkImage (e.g.
out-of-memory, GPU process loss, or other unpredictable failure case).

Bug: 852303
Change-Id: I10a817f2dccd801f5834be343ad687af09e06721
Reviewed-on: https://chromium-review.googlesource.com/1188515Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Kai Ninomiya <kainino@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586085}

This simplifies many conditions around the code in
chromeos/components/tether.

Bug: 561800
Change-Id: I69c25666ae1b1362df6df67889a9a45226961387
Reviewed-on: https://chromium-review.googlesource.com/1188060Reviewed-by: Kyle Horimoto <khorimoto@chromium.org>
Commit-Queue: DongJun Kim <djmix.kim@samsung.com>
Cr-Commit-Position: refs/heads/master@{#586084}

Avoids linker errors from LLD about -Wl, not being a recognized command
line flag.  This will fix the Windows CFI and ThinLTO FYI bots.

Factor out a variable for the optimization level so we need one less
conditional.

TBR=agreive@chromium.org, agrieve@chromium.org, gbiv@chromium.org, pcc@chromium.org

Bug: 807147
Change-Id: I8a928eaab93ec181ba057cdda4cb635389f30926
Reviewed-on: https://chromium-review.googlesource.com/1188879Reviewed-by: Reid Kleckner <rnk@chromium.org>
Reviewed-by: George Burgess <gbiv@chromium.org>
Reviewed-by: Peter Collingbourne <pcc@chromium.org>
Reviewed-by: agrieve <agrieve@chromium.org>
Commit-Queue: Reid Kleckner <rnk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586083}

Replace the GPU.DirectComposition.SwapBuffersFailed UMA which was being
emitted on failure only with SwapBuffersResult which is emitted on both
success and failure.

Bug: 869677
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: If950beb70d22f64997e7533751abcf9448d608db
Reviewed-on: https://chromium-review.googlesource.com/1187128
Commit-Queue: Sunny Sachanandani <sunnyps@chromium.org>
Reviewed-by: Ilya Sherman <isherman@chromium.org>
Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586082}

This reverts commit bfb341f2.

Reason for revert: No longer necessary to attempt to expand
SSHPortForwarder's functionality to support bidirectional port mapping.
Original change's description:
> [Fuchsia] TestServer rework for new SSHPortForwarder functionality.
>
> Made a minor change to make sure that SetupTestServer interacts
> properly with the new _ConnectPortForwardingTask and ForwardingScheme
> versions.
>
> Change-Id: I16add737f7595fac416973c1296882f9f40d0dc1
> Reviewed-on: https://chromium-review.googlesource.com/1178381
> Reviewed-by: Kevin Marshall <kmarshall@chromium.org>
> Commit-Queue: Kevin Marshall <kmarshall@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#583900}

TBR=kmarshall@chromium.org,sergeyu@chromium.org,stephanstross@google.com

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

Change-Id: I1fd3a0f48cc21fba0d599a9fba26a62aa04e5dc1
Reviewed-on: https://chromium-review.googlesource.com/1188493
Commit-Queue: Stephan Stross <stephanstross@google.com>
Reviewed-by: Sergey Ulanov <sergeyu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586081}

autofill_jni_headers used to run for
components/autofill/browser/android/java/src/org/chromium/components/browser/autofill/PersonalAutofillPopulator.java
in the old gyp build, but that file has been removed in crbug.com/403872.

Bug: 403872
Change-Id: I7551a517087e1da2d3ebede156107ca358712aa7
Reviewed-on: https://chromium-review.googlesource.com/1187764Reviewed-by: Evan Stade <estade@chromium.org>
Commit-Queue: Nico Weber <thakis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586080}

This policy allows administrators to prohibit users from utilizing the
upcoming feature which allows users to sync their SMS messages between
their laptops and phones.

Bug: 870113, 824568
Change-Id: I12515b75c1f62d5c99bfb1a0adacddad80de0196
Reviewed-on: https://chromium-review.googlesource.com/1187730Reviewed-by: Maksim Ivanov <emaxx@chromium.org>
Reviewed-by: Gayane Petrosyan <gayane@chromium.org>
Reviewed-by: Ryan Hansberry <hansberry@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586079}

In GTK4, the GDestroyNotify that SelectFileDialogImplGtk used does not get fired
when the GtkFileChooser is closed (perhaps GTK is now doing some lazy
deletion?).  But the "destroy" signal does get fired, so move the event
re-enabling code there.

Also clean up some messy code.

BUG=876586
R=sky

Change-Id: I2d78ff292d167fe526bd5fb35aaf6296d32da1af
Reviewed-on: https://chromium-review.googlesource.com/1188585Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Thomas Anderson <thomasanderson@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586078}

This behaves similar to DelegatedFrameHost. As discussed in
https://chromium-review.googlesource.com/c/chromium/src/+/1173663,
this implements composition instead of inheritance.

Bug: 859977
Change-Id: Icf5e0c83c4b3789a2173933a26067c6275f62765
Reviewed-on: https://chromium-review.googlesource.com/1186199Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Reviewed-by: Fady Samuel <fsamuel@chromium.org>
Commit-Queue: Xing Xu <xing.xu@intel.com>
Cr-Commit-Position: refs/heads/master@{#586077}

This CL updates the What's New texts to reflect these changes:
* Between Android client M61 and M70
* Between iOS client M67 and M70

Bug: 877704
Change-Id: Ifb3aff593c1bfc3a8b1f02f018dd458670c3100e
Reviewed-on: https://chromium-review.googlesource.com/1189055Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Commit-Queue: Yuwei Huang <yuweih@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586076}

This CL includes concrete implementations for the abstract Executor
classes defined in the Nearby library. A ScheduledExecutor allows
scheduled tasks that asynchronously execute in the future. A
MultiThreadExecutor allows tasks with or without return values to be
asynchronously executed in parallel, and a SingleThreadExecutor allows
tasks with or without return values to be executed asynchronously, but
in sequence relative to each other.

Because the actual Nearby library has yet to be merged into the CrOS
directory, this CL includes stand-in Nearby abstract classes under
the temporary directory //chromeos/components/nearby/library.
This directory will be removed after the Nearby library gets imported
to //third_party.

Bug: 861813
Change-Id: Idc561f2cedea7360eaf6d74940e1be8e56efe7cb
Reviewed-on: https://chromium-review.googlesource.com/1154134
Commit-Queue: Kyle Qian <kyleqian@google.com>
Reviewed-by: Kyle Horimoto <khorimoto@chromium.org>
Reviewed-by: Ryan Hansberry <hansberry@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586075}

This reverts commit acf1c986, which
speculatively reverted the change to establish whether it might be
responsible for a spike in BlobReader crashes. Re-landing now, since the
revert had no impact on crash rate of that signature in Canary (see
https://crbug.com/864351).

- Use a base::AutoReset instance to manage a single integer tracking
  whether the completion callback for an operation is invoked before
  we have actually returned to the caller.
- Remove the now-unused OperationHandle and BeginOperationScoper.
- Replace SupportsWeakPtr with an internal WeakPtrFactory.
- Replace potentially unsafe use of GetWeakPtr() with a |weak_ptr_|
  member.
- Temporarily disables the recently-added FileSystemURLLoaderFactoryTest
  tests, which have incorrect threading, causing WeakPtr checks to
  fire.

TBR=wez@chromium.org,rockot@chromium.org,alemate@chromium.org,noel@chromium.org,mek@chromium.org,fukino@chromium.org,cmumford@chromium.org,govind@chromium.org,tzik@chromium.org,abdulsyed@chromium.org

Bug: 846985, 860547, 864351
Change-Id: Ia843811a2e85be9347f410ec9f97dc3dc0f1713a
Reviewed-on: https://chromium-review.googlesource.com/1189048
Commit-Queue: Wez <wez@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586074}

Bug: 731383
Change-Id: Id563bd36de50c7685a524336906fc8542d1c6f51
Reviewed-on: https://chromium-review.googlesource.com/1188451Reviewed-by: Nodir Turakulov <nodir@chromium.org>
Commit-Queue: Eric Foo <efoo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586073}

Specifically the mojoms in services/ws now are all in the ws namespace.

BUG=876924
TEST=covered by tests
TBR=msw@chromium.org,tsepez@chromium.org

Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng
Change-Id: Ib92b167f8b528b6c0423ec4be99e25558d1d43da
Reviewed-on: https://chromium-review.googlesource.com/1189030Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586072}

App ranking system uses install time in case launch time is not
available to rank apps. We don't want apps, installed by policy or by
default appear in recent app. This CL extends app ranking system by
handling such apps. Previously we had ARC apps folled by this approach
by not-setting install time installed internally apps. However this was
not good once install time should not depend on ranking system. So this
CL makes ranking centralized and switch ARC and Extension apps to it.

TEST=Manually + unit tests added.
BUG=757639

Change-Id: If567e7e72908c0186e3096b1e83fa917df47d082
Reviewed-on: https://chromium-review.googlesource.com/1188591
Commit-Queue: Yury Khmel <khmel@google.com>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586071}

This CL includes a concrete implementation for the abstract classes
Lock (which must be recursive) and ConditionVariable defined in the
Nearby library.

A related CL was previously abandoned
(https://chromium-review.googlesource.com/c/chromium/src/+/1138871)
because the Nearby C++ library already includes a DefaultLock and
DefaultConditionVariable. However, that CL is being revived as this CL
because a Chromium-specific Lock/CV implementation is required. For
instance, Chromium requires a ScopedBlockingCall to be instantiated in
scopes where blocking calls are made. DefaultConditionVariable does not
include an instantiation of ScopedBlockingCall. This implementation of
ConditionVariable uses a base::WaitableEvent, which instantiates a
ScopedBlockingCall. This CL also includes a Lock implementation that is
recursive, which is a requirement of Nearby, but is not originally
supported by base::Lock.

Because the actual Nearby library has yet to be merged into the CrOS
directory, this CL includes stand-in Nearby abstract classes under
the temporary directory //chromeos/components/nearby/library.
This directory will be removed after the Nearby library gets imported
to //third_party.

Bug: 861813
Change-Id: I72660ed302e239d8f38a21728fc715b000057c13
Reviewed-on: https://chromium-review.googlesource.com/1145730
Commit-Queue: Kyle Qian <kyleqian@google.com>
Reviewed-by: François Doray <fdoray@chromium.org>
Reviewed-by: Kyle Horimoto <khorimoto@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586070}

remove the manual source list and use the auto-generated one.

BUG=872521
TEST=bots
R=nednguyen@chromium.org,kbr@chromium.org,nednguyen@google.com

Change-Id: Ic5d00299edc39af23802db71c43ad0716669837d
Reviewed-on: https://chromium-review.googlesource.com/1188901Reviewed-by: Ned Nguyen <nednguyen@google.com>
Commit-Queue: Zhenyao Mo <zmo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586069}

Create a utility method to unify the way we create image buttons. This
turns on default ink drop click and focus styles and make them focusable.

Test: manually build and test
Bug: b:112925587
Change-Id: I70c72e033fae2dbad39034d716483c77ea578161
Reviewed-on: https://chromium-review.googlesource.com/1186137
Commit-Queue: Xiaohui Chen <xiaohuic@chromium.org>
Reviewed-by: Muyuan Li <muyuanli@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586068}

It seems that Event's root_location is in display's coordinate
in WindowTreeClient (i.e. browser). This could complicate the
coordinate conversions like aura::ScreenCoordinateClient and
cause problems like crbug.com/876509.

Since root_location is relative to the root window in desktop
aura (like desktop Linux), I believe we want to do so too on
Mash.

This is actually contradicting with crrev.com/583488 but
I believe this is better for us.

Bug: 876509
Test: unit tests pass, manually checked crbug.com/873763 behavior
Change-Id: I51f03772e2872783cf5fb9ce78cdcfe3dd7b0e67
Reviewed-on: https://chromium-review.googlesource.com/1187134
Commit-Queue: Jun Mukai <mukai@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586067}

These tests were disabled by https://crrev.com/c/1176206 because
they were reported as flaky on debug builds. Based on the
flakiness dashboard, it appears the failures were timeouts on the
Win10 Tests x64 (dbg) bot. interactive_ui_tests is marked as
experimental on this bot due to crbug.com/840369. The
webui_polymer2_interactive_ui_tests suite is expected to have the same
behavior, so it should also be marked as experimental on this bot.

Remove the tests from the blacklist, as they should pass on other bots.

Bug: 875443
Change-Id: Ib5b3c99cf589b1ab9994346460e95ed746533ba6
Reviewed-on: https://chromium-review.googlesource.com/1187729Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Rebekah Potter <rbpotter@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586066}

This is a reland of a2a4e75d with a
work-around for cronet_tests' sharing libbase.so with libcronet.so in
component builds.

- Move some APIs off of base::TestSuite, that we unnecessarily public.
- Rename the various internal TestEventListeners to express purpose.
- Add CheckForLeakedGlobals check, and DisableCheckForLeakedGlobals API.
  The latter is called by the content::ContentTestSuiteBase, since the
  various browser- and ui-tests running under that suite base run more
  like the browser itself, and so are expected to let global singletons
  leak.

If tests fail or flake due to this patch, then this indicates that the
test is actually leaking global state that may break other tests if they
happen to run in the same process, leading to hard-to-diagnose flakes.

There are two main failure cases:
If the test directly or indirectly uses TaskScheduler, but does not
create a ScopedTaskEnvironment, then please add one to the test, or to
the test base-class, as appropriate, to fix it.
If the test suite cannot be fixed in this way then please add a call to
DisableCheckForLeakedGlobals(), with a comment linking to a bug for the
issue.

TBR: gab, sky, mef
Bug: 875486, 877355, 744567
Change-Id: Iaea38d24ede271c248a3abb0b3f7ee931c2538f5
Reviewed-on: https://chromium-review.googlesource.com/1187783Reviewed-by: Wez <wez@chromium.org>
Commit-Queue: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586065}

This field was dropped by a revert, but we have data for it, so re-add the description.
https://crrev.com/d38f4e3c1225c36574911f6fdeecdb1b236dd054

Change-Id: Ie68bc76ced75a1d6269b252801cda76baa5de9ff
Reviewed-on: https://chromium-review.googlesource.com/1188765Reviewed-by: Ilya Sherman <isherman@chromium.org>
Commit-Queue: Steven Holte <holte@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586064}

Fixes network_service_content_browsertests RenderThreadImplBrowserTest.NonResourceDispatchIPCTasksDontGoThroughScheduler
failing to find the utility service process exe on Windows, like:
[FATAL:utility_process_host.cc(263)] Check failed: false. Unable to get utility process binary name.

Bug: 876947
Cq-Include-Trybots: luci.chromium.try:linux_mojo
Change-Id: Ia7f51366e4b4db27ec8c3bbee3d57866c1bd67b9
Reviewed-on: https://chromium-review.googlesource.com/1185628
Commit-Queue: Matt Mueller <mattm@chromium.org>
Reviewed-by: John Abd-El-Malek <jam@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586063}