GitLab

It is actually a server misbehavior, but currently it is unknown,
how widespread it is.

Added UMA to track such events.

Drive-by: fix reported encoding in case of unadvertised encoding
to be "REJECTED".

BUG=714514

Review-Url: https://codereview.chromium.org/2835403003
Cr-Commit-Position: refs/heads/master@{#468032}

The interface taking a raw pointer is deprecated and being removed.

BUG=690937

Review-Url: https://codereview.chromium.org/2845263002
Cr-Commit-Position: refs/heads/master@{#468031}

Split out the parts that are specific to the committed Document, and
move those to a new Document::CheckCompleted(). Most current callers of
FrameLoader::CheckCompleted() will now call Document::CheckCompleted()
instead. Rename the remainder of FrameLoader::CheckCompleted() to
DidFinishNavigation(), and have it handle the logic that is tied to
setting Frame::loading_ to false and firing DidStopLoading() callbacks.

Originally reviewed in https://codereview.chromium.org/2809733003/

BUG=

Review-Url: https://codereview.chromium.org/2837763003
Cr-Commit-Position: refs/heads/master@{#468030}

No one is using it, so we've decided to delete the code. The removals
from api.txt are safe because that code is known to be unused, partly
because it wouldn't have worked anyway.

BUG=716108
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_cronet_tester

Review-Url: https://codereview.chromium.org/2839413003
Cr-Commit-Position: refs/heads/master@{#468029}

The interface taking a raw pointer is deprecated and being removed.

BUG=690937

Review-Url: https://codereview.chromium.org/2851673002
Cr-Commit-Position: refs/heads/master@{#468028}

When the zygote starts it sends a "boot message" to the browser process
across a socket. If the browser has already exited or crashed the
sendmsg() will fail. We've seen this happen in the Chrome OS autotest
lab.

Instead of crashing the zygote, just log an error and cleanly exit.

BUG=692227
TEST=add sleep to zygote startup, kill browser before zygote boot
message is sent, verify zygote shuts down cleanly

Review-Url: https://codereview.chromium.org/2851473005
Cr-Commit-Position: refs/heads/master@{#468027}

BUG=716149

Review-Url: https://codereview.chromium.org/2844373003
Cr-Commit-Position: refs/heads/master@{#468026}

BUG=714827

Review-Url: https://codereview.chromium.org/2843463004
Cr-Commit-Position: refs/heads/master@{#468025}

PatchSet #1 is the state as previously reviewed and reverted.
PatchSet #2 fixes the reason for the revert.

Description of Fix:
The test "AccessIsRevokedOnSecondAccess" was flaky because had an incorrect
expectation on the order in which events from the service arrive. This is fixed
by removing this expectation.

Original CL description:

This CL is part of the Mojo Video Capture work. For the bigger picture,
see [1] CL22b.

Note: The video capture service implementation is currently in an incomplete and
outdated state and its tests are disabled. With the refactoring of the legacy
video capture stack now being complete, the next goal is to update the service
implementation and fit it into the refactored stack.

Changes in this CL:
* Update Mojo interfaces to more closely resemble their native counterparts.
  - video_capture.mojom.Device offers functionality similar to
    VideoCaptureDeviceLauncher + LaunchedVideoCaptureDevice.
  - video_capture.mojom.DeviceFactory offers functionality similar to
    media::VideoCaptureProvider.
  - video_capture.mojom.Receiver offers functionality similar to
    media::VideoFrameReceiver.
* In the service implementation, use a VideoCaptureSystem instead of a
  VideoCaptureDeviceFactory directly.
* Add new adapter classes for plumbing between Mojo service and its usage in
  the native code.
* Re-enable existing video_capture_unittests and add a few more.

BUG=584797
TEST=
  service_unittests --gtest_filter="*Video*"
  content_unittests --gtest_filter="*Video*"
  content_browsertests --gtest_filter="VideoCaptureBrowserTest.*"
TBR=emircan@chromium.org,ochang@chromium.org

[1] https://docs.google.com/a/chromium.org/document/d/1Qw7rw1AJy0QHXjha36jZNiEuxsxWslJ_X-zpOhijvI8/edit?usp=sharing

Review-Url: https://codereview.chromium.org/2843333006
Cr-Commit-Position: refs/heads/master@{#468024}

Basic dispatcher to help with ios/clean migration.  Handlers UrlLoader,
OmniboxFocuser and the WebToolbarDelegate, all used by the
GoogleLandingController.

BUG=694750

Review-Url: https://codereview.chromium.org/2829003002
Cr-Commit-Position: refs/heads/master@{#468023}

It's no longer part of the spec
(https://github.com/whatwg/fetch/commit/42464c)

intent-to-deprecate-and-remove
https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/x3YXciXQWM0

BUG=645492

Review-Url: https://codereview.chromium.org/2591723003
Cr-Commit-Position: refs/heads/master@{#468022}

The throttling fix landed in https://codereview.chromium.org/2817603004/
has severely impacted rendering throughput due to the delay in
compositor round trips.  To avoid blocking the renderer script's
animation loop, we need to allow the rendering proceed when there
is still a pending frame.

BUG=716138

Review-Url: https://codereview.chromium.org/2848553003
Cr-Commit-Position: refs/heads/master@{#468021}

Followup for http://crrev.com/2844553003

R=dbeam@chromium.org
BUG=706310

Review-Url: https://codereview.chromium.org/2850663002
Cr-Commit-Position: refs/heads/master@{#468020}

Returns a specific error saying the keys are not available in the
signin profile.

BUG=715121
TEST=unit tests

Review-Url: https://codereview.chromium.org/2838423003
Cr-Commit-Position: refs/heads/master@{#468019}

ScopedTaskEnvironment allows usage of ThreadTaskRunnerHandle and
base/task_scheduler/post_task.h within its scope. It should be
instantiated in everytest that uses either of these APIs
(i.e. no test should instantiate a MessageLoop directly).

Motivation for ScopedTaskEnvironment can be found in:
https://docs.google.com/document/d/1QabRo8c7D9LsYY3cEcaPQbOCLo8Tu-6VLykYXyl3Pkk/edit

BUG=708584
R=gab@chromium.org
TBR=garykac@chromium.org

Review-Url: https://codereview.chromium.org/2846943002
Cr-Commit-Position: refs/heads/master@{#468018}

Since bug 630103 was fixed, these crash keys are no longer necessary.

BUG=630103

Review-Url: https://codereview.chromium.org/2841203003
Cr-Commit-Position: refs/heads/master@{#468017}

This CL fixes confusing field name. Field has AccountManagerDelegate type
and should be named accordinly.

BUG=698258

Review-Url: https://codereview.chromium.org/2850813002
Cr-Commit-Position: refs/heads/master@{#468016}

It looks like this function got cargo-culted around during refactors.

* Remove unused copies for NaCl, utility process and PPAPI
* Inline the one used copy from RenderProcessHostImpl into BrowserMainLoop

No behavior changes.

BUG=none
TEST=bots

Review-Url: https://codereview.chromium.org/2848603003
Cr-Commit-Position: refs/heads/master@{#468015}

The interface taking a raw pointer is deprecated and being removed.

BUG=690937

Review-Url: https://codereview.chromium.org/2845253002
Cr-Commit-Position: refs/heads/master@{#468014}

Chrome also detects when the connection to mus is lost and further
chrome needs to control when the shell it shutdown, so adds the option
to set a callback that controls what should happen when connection to
mus is lost.

I think there is more that needs to happen, but this is a good start.

BUG=709593
TEST=covered by tests
R=jonross@chromium.org

Review-Url: https://codereview.chromium.org/2846893004
Cr-Commit-Position: refs/heads/master@{#468013}

BUG=716091

Review-Url: https://codereview.chromium.org/2851573002
Cr-Commit-Position: refs/heads/master@{#468012}

Recently, fake-bidi was introduced into the list of supported locales:
  https://codereview.chromium.org/2773453003

This introduced a bug where fake-bidi appeared on our list of supported
locales, yet could not be selected as a locale. Since we are missing
dedeicated resources for fake-bidi, and not of our tests explicitly
support it, remove support everywhere for consistency. Also unify the
locale list into a single place that can be referenced from anywhere in
GN.

BUG= internal b/37215118
Test: All .pak files generate properly, setup test works.

Change-Id: I8bc7e7335716d14288654d9912479d31f6438555
Reviewed-on: https://chromium-review.googlesource.com/488166Reviewed-by: Luke Halliwell <halliwell@chromium.org>
Commit-Queue: Stephen Lanham <slan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#468011}

BUG=637415

Review-Url: https://codereview.chromium.org/2843153002
Cr-Commit-Position: refs/heads/master@{#468010}

ConsolePrompt's moveCaretToEndOfPrompt() works by setting the CodeMirror
editor's selection to the end. After a recent CodeMirror roll, doing so now
scrolls the editor into view.

This change in behavior led to an error where expanding an object in console
would jump the viewport to bottom. This CL only focuses the prompt without
moving the cursor, in some cases.

BUG=706128

Review-Url: https://codereview.chromium.org/2840663002
Cr-Commit-Position: refs/heads/master@{#468009}

The interface taking a raw pointer is deprecated and being removed.

BUG=690937

Review-Url: https://codereview.chromium.org/2853443002
Cr-Commit-Position: refs/heads/master@{#468008}

Screenshots at
https://drive.google.com/drive/folders/0B7eRCSITD4qmelN0aEFMQXNwNTA?usp=sharing

BUG=716022

Review-Url: https://codereview.chromium.org/2849523004
Cr-Commit-Position: refs/heads/master@{#468007}

The interface taking a raw pointer is deprecated and being removed.

BUG=690937

Review-Url: https://codereview.chromium.org/2847013003
Cr-Commit-Position: refs/heads/master@{#468006}

The interface taking a raw pointer is deprecated and being removed.

BUG=690937

Review-Url: https://codereview.chromium.org/2844413004
Cr-Commit-Position: refs/heads/master@{#468005}

The interface taking a raw pointer is deprecated and being removed.

BUG=690937

Review-Url: https://codereview.chromium.org/2847763003
Cr-Commit-Position: refs/heads/master@{#468004}

This CL fixes presubmit_scheme_histograms.py and
presubmit_bad_message_reasons.py which after r462591 would always report
presubmit errors (because it started returning a tuple from HistogramNeedsUpdate
which always coerces to True in the broken presubmit scripts).

After r462591 HistogramNeedsUpdate effectively checks all
histogram-enum-related presubmits (both for updating histograms.xml and
for detecting duplicate values).  Therefore the current CL renames this
function to CheckPresubmitErrors and makes it return a presubmit error
(rather than a tuple of (needs_updating, duplicates)).  This refactoring
has a nice side benefit - after this CL, all presubmits going through
CheckPresubmitErrors will check for duplicate values (r462591 only added
this check for UseCounter histograms).

This CL also tweaks content/browser/PRESUBMIT.py so that it applies
*both* to upload-time and commit-time checks (without this change
|git cl presubmit| would not hit the checks related to
content/browser/bad_message.h - they would only be hit when running
|git cl presubmit --upload|).

Manual testing done:
- Make changes in content/browser/bad_message.h,
  chrome/browser/extensions/chrome_content_browser_client_extensions_part.cc,
  third_party/WebKit/Source/code/frame
- Run git cl presubmit and verify that the presubmit
  checks detect 1) duplicates and 2) the need to update histograms.xml

BUG=577772
TEST=See "Manual testing done" above.
TBR=sky@chromium.org, jochen@chromium.org, rdevlin.cronin@chromium.org

Review-Url: https://codereview.chromium.org/2841823007
Cr-Commit-Position: refs/heads/master@{#468003}

This adds the basic internals of the GRC; CoordinationUnits which can form a DAG, receive events, and send back updated resource usage policies.

A full prototype of GRC usage can be seen here: https://codereview.chromium.org/2710823003

Service architecture doc: https://docs.google.com/document/d/1qec4DNDM2pLLIFfCBtnNQTxlNXQzjml69yC8SGU9bzI/edit#

GRC parent doc (internal only): https://docs.google.com/document/d/1dx4KDbDFvP-GWwwrSPg8Gxx4kboIoPi8kDKTSXoTbC4/edit#heading=h.td4yhfm12fe3

R=primiano@chromium.org,skyostil@chromium.org,fmeawad@chromium.org,rockot@chromium.org
BUG=691886

Review-Url: https://codereview.chromium.org/2798713002
Cr-Commit-Position: refs/heads/master@{#468002}

When forcible termination is requested, WorkerThread::WillProcessTask() calls
PrepareForShutdownOnWorkerThread() that disposes of the global scope scheduler,
which in turn disposes of the task to be run. This results in crashes.

This CL moves PrepareForShutdownOnWorkerThread() call from WillProcessTask() to
DidProcessTask() to fix the crashes. This could still cause a crash if the
termination request is issued between DidProcessTask() and WillProcessTask()
because in the case a task runs and may access an empty handle returned by V8
API. However, such a case would be rare because generally forcible termination
is scheduled as a delayed task(*) and DidProcessTask() can have a chance to
check IsForciblyTerminated() before it's fired.

(*) Forcible termination could synchronously be called in some path, but it'll
be removed (see issue 641846).

BUG=713914

Review-Url: https://codereview.chromium.org/2849583004
Cr-Commit-Position: refs/heads/master@{#468001}

If it happens that user policy is retrieved before session is started,
then there's no policy to apply to the user which could be used as
policy enforcement escape. This is a case when some flow went wrong,
and we abandon the user sign in.

BUG=689206

Review-Url: https://codereview.chromium.org/2801993002
Cr-Commit-Position: refs/heads/master@{#468000}

Spec: https://drafts.css-houdini.org/worklets/#dom-worklet-addmodule

BUG=627945

Review-Url: https://codereview.chromium.org/2851693002
Cr-Commit-Position: refs/heads/master@{#467999}

This feature was reverted due to the new browsertest being flaky on
internal Cast infrastructure: crrev.com/2838813003

=== Original Commit Message ===
In Chromium, Finch-enabled features are controlled through base::FeatureList,
a class which abstracts the experiment framework and developer overrides
from client code. Though Chromecast's experiment framework is fundamentally
different (in that it is server-driven) Cast builds can still make use of
this class. Introduce some utilities to help.

At boot-up, the pref store will be queried for experiment configs, which
were cached to disk on the most recent config fetch from the last boot
cycle. If a developer overrides these features from the command line,
that value takes precedence. These features will be used to initialize
base::FeatureList, which can then be statically queried from any client
code that depends on //base.

This patch does not actually introduce or convert any existing features
to use this framework.

BUG=714291
BUG= internal b/35424335

Review-Url: https://codereview.chromium.org/2836263003
Cr-Commit-Position: refs/heads/master@{#467998}

When a pen enters and leaves a tablet's proximity and enters again, the
PointerEvents' isPrimary should always be true.

Testing page: https://patrickhlauke.github.io/touch/tracker/multi-touch-tracker-pointer-hud.html

Webplatform test pointerevent_pointerleave_pen-manual.html does not work
right now because of no support of PointerLeave events when pen leaves
the range of the tablet in PointerActionSequence API. I will add it soon.

BUG=713745

Review-Url: https://codereview.chromium.org/2831933002
Cr-Commit-Position: refs/heads/master@{#467997}

https://skia.googlesource.com/skia.git/+log/56de05fec3fa..0122af08f6af

$ git log 56de05fec..0122af08f --date=short --no-merges --format='%ad %ae %s'
2017-04-27 msarett Revert "Delete copyTo(Allocator), hide copyTo() behind flag"
2017-04-27 herb Add instrumentation into SkArenaAlloc.
2017-04-27 mtklein jumper, remove C(int)
2017-04-27 egdaniel Revert "Plumb the use of GrBackendRenderTarget throughout Skia"
2017-04-26 brianosman Fold SkImageCacherator into SkImage_Lazy
2017-04-27 mtklein jumper, replace _i with normal constants
2017-04-26 egdaniel Plumb the use of GrBackendRenderTarget throughout Skia
2017-04-27 egdaniel Add additional checks when creating vulkan context
2017-04-27 kjlubick Don't reboot Pixel C or Nexus Player before task
2017-04-27 msarett Delete copyTo(Allocator), hide copyTo() behind flag
2017-04-27 mtklein In ok thread mode, use wait_util instead of wait_for.
2017-04-27 mtklein clear out C(), _i, and _f constants from SkJumper_vectors.h
2017-04-24 reed remove legacy code
2017-04-26 egdaniel Change vulkan image memory allocation back to using sub allocation
2017-04-25 brianosman Finish removing GrContext param from onRefEncodedData
2017-04-26 brianosman Fix color space handling in SkImage_Gpu::getROPixels

Created with:
  roll-dep src/third_party/skia


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

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls


CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel
TBR=benjaminwagner@chromium.org

Change-Id: I29643b478fde4d32304518327ea933b4e9176491
Reviewed-on: https://chromium-review.googlesource.com/489483Reviewed-by: Skia Deps Roller <skia-deps-roller@chromium.org>
Commit-Queue: Ravi Mistry <rmistry@chromium.org>
Cr-Commit-Position: refs/heads/master@{#467996}

Every LayoutBox with a PaintLayer will do some layer updating at the end of its layout; this consists of:

- updating the layer's transform
- updating the layer's scrollable area, if necessary
- in the near future, updating the layer's size

This change consolidates all of those into LayoutBox::UpdateAfterLayout.

BUG=701575
R=skobes@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2839183002
Cr-Commit-Position: refs/heads/master@{#467995}

Changes the confusing ACTION_funnel_ names in scheduler state machine to
did_ACTION_. Replaces prepare_tiles_funnel_ (the only real "funnel")
with a bool that does the same job. Changes the state machine to use
default member initialization.

One SchedulerTest which explicitly tested that we reset the funnel after
visibility change had to be modified. Since we don't reset the prepare
tiles state on begin frame any more that old test doesn't work as is.
Following along the discussion in crbug.com/497473#c14 the test's exact
behavior isn't needed as we never end up in a situation with an ever
increasing prepare tiles funnel after this change.

R=danakj
BUG=NONE

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Change-Id: I1541b358d05ad1b824955ae63e3bcb173c3ffda3
Reviewed-on: https://chromium-review.googlesource.com/487322
Commit-Queue: danakj <danakj@chromium.org>
Reviewed-by: danakj <danakj@chromium.org>
Cr-Commit-Position: refs/heads/master@{#467994}

This will run the unit test suites for the try bots and on the main
chromium waterfall. Once this is done, we can remove them from the
separate webkit waterfall.

R=dpranke@chromium.org
BUG=713180, 703894

Review-Url: https://codereview.chromium.org/2849673002
Cr-Commit-Position: refs/heads/master@{#467993}