GitLab

BUG=546411

Review URL: https://codereview.chromium.org/1412613004

Cr-Commit-Position: refs/heads/master@{#357068}

BUG=546596

Review URL: https://codereview.chromium.org/1423773003

Cr-Commit-Position: refs/heads/master@{#357067}

This CL finds the closest supported format in allocate() from the requested resolution, and uses that to open the camera. This is consistent with how VideoCaptureCamera and other camera capture devices work.

BUG=541232

Review URL: https://codereview.chromium.org/1416253005

Cr-Commit-Position: refs/heads/master@{#357066}

This CL renames DISALLOW_ALLOCATION to DISALLOW_NEW, and renames
ALLOW_ONLY_INLINE_ALLOCATION to DISALLOW_NEW_EXCEPT_PLACEMENT_NEW
per the plan in https://docs.google.com/spreadsheets/d/1AZtNeaUN_2YTbm4GhCGuy6V19PaDm4cqGElkfZyo2ug/edit?pli=1#gid=0

BUG=548217

Review URL: https://codereview.chromium.org/1406923009

Cr-Commit-Position: refs/heads/master@{#357065}

This CL includes:
- Replace custom progress bar with paper-slider.
- Pause track temporalily while dragging progress bar to be consistent with video player.
- Adjust layout of control panel.
- Show playing time as "{current} / {duration}".

BUG=488229
TEST=manually

Review URL: https://codereview.chromium.org/1403303011

Cr-Commit-Position: refs/heads/master@{#357064}

BUG=543522

Review URL: https://codereview.chromium.org/1418453004

Cr-Commit-Position: refs/heads/master@{#357063}

TBR=aiolos@chromium.org
BUG=549428
CQ_EXTRA_TRYBOTS=tryserver.chromium.perf:linux_perf_bisect;tryserver.chromium.perf:mac_10_10_perf_bisect;tryserver.chromium.perf:win_perf_bisect;tryserver.chromium.perf:android_nexus5_perf_bisect

Review URL: https://codereview.chromium.org/1428883003

Cr-Commit-Position: refs/heads/master@{#357062}

This CL paves the way for elimination of the FOREIGN_SESSION_UPDATED
notification as part of componentizing sync code. Concretely, it replaces the
notification via a SyncServiceObserver method and changes consumers to use the
observer method instead.

Currently, ProfileSyncService triggers the observer callback by itself
listening for the FOREIGN_SESSION_UPDATED method. Once SessionsSyncManager has
been changed to communicate directly with ProfileSyncService rather than
sending the notification, the notification can be eliminated entirely.

BUG=512825

Review URL: https://codereview.chromium.org/1406943004

Cr-Commit-Position: refs/heads/master@{#357061}

R=dpranke@chromium.org
BUG=549254

Review URL: https://codereview.chromium.org/1422333005

Cr-Commit-Position: refs/heads/master@{#357060}

It is used only in XSLTProcessorLibxslt.cpp.
This CL makes a static helper method for XSLTProcessorLibxslt and removes
Partitions::fastStrDup.

BUG=548582

Review URL: https://codereview.chromium.org/1409113010

Cr-Commit-Position: refs/heads/master@{#357059}

MessageLoopInterruptor and PendingGCRunner are always created and destroyed at the same time.
This CL puts these two classes into GCTaskRunner.
This CL also renames PendingGCRunner to GCTaskObserver.

BUG=

Review URL: https://codereview.chromium.org/1419513007

Cr-Commit-Position: refs/heads/master@{#357058}

We send other UMAs about sweeping only from the main thread.
BlinkGC.CompleteSweep should follow the rule.

BUG=548628

Review URL: https://codereview.chromium.org/1425643003

Cr-Commit-Position: refs/heads/master@{#357057}

On Android, on-screen keyboard will flicker when you move focus from
one input form to another. We have a delayed OSK dismissal hack, but
we found that this is not a good solution for three reasons:

1) Flickering still happens, even though it's less severe.
2) Hiding of OSK will be delayed when touching a non-editable web node.
3) It makes keyboard showing & hiding very difficult to test.

The main cause of the flickering is that we call UpdateTextInputState()
in the middle of focus change:

- FocusController::setFocusedElement()
- FocusController::clearSelectionIfNecessary()
- FrameSelection::clear()
- Editor::notifyComponentsOnChangedSelection()
- ...
- RenderWidget::UpdateTextInputState()
- WebViewImpl::textInputInfo()

This can be avoided if we guard non-keyboard events in OnHandleInputType.

Also because focuschange (which should result in state update) can happen
through / not through OnHandleInputType, I made
StartHandlingImeEvent and FinishHandlingImeEvent nestable.

BUG=484139, 542619

Review URL: https://codereview.chromium.org/1388283002

Cr-Commit-Position: refs/heads/master@{#357056}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/dce364cc..82fcf88a

Please follow these instructions for assigning/CC'ing issues:
https://code.google.com/p/v8-wiki/wiki/TriagingIssues

Please close rolling in case of a roll revert:
https://v8-roll.appspot.com/

TBR=hablich@chromium.org,machenbach@chromium.org,yangguo@chromium.org,vogelheim@chromium.org

Review URL: https://codereview.chromium.org/1423363002

Cr-Commit-Position: refs/heads/master@{#357055}

Intent to Ship: https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/uajNnbH8UPU

BUG=521466

Review URL: https://codereview.chromium.org/1371393003

Cr-Commit-Position: refs/heads/master@{#357054}

Copied correct file from LayoutTests/resources/Ahem.ttf.
Also, updated http tests to use the fixed one.

BUG=548543

Review URL: https://codereview.chromium.org/1417943005

Cr-Commit-Position: refs/heads/master@{#357053}

BUG=529938

Review URL: https://codereview.chromium.org/1410573008

Cr-Commit-Position: refs/heads/master@{#357052}

This should get reverted once the reference build is updated, which should be
sometime next week.

BUG=547833

Review URL: https://codereview.chromium.org/1404423010

Cr-Commit-Position: refs/heads/master@{#357051}

Revert of Web Animations: Use a single animation clock (patchset #7 id:120001 of https://codereview.chromium.org/1410313004/ )

Reason for revert:
Still breaks the Android bot (at least): https://build.chromium.org/p/chromium.webkit/waterfall?show=WebKit%20Android%20(Nexus4)

Please do not reland without first ensuring you have run the webkit_unit_tests step a few times on an equivalent Android bot.

Original issue's description:
> Web Animations: Use a single animation clock
>
> All animation clocks for a page should tick at the same time, but since we
> had an animation clock per document there were some cases where clocks
> could tick independently.
>
> This patch moves the clock from Document to PageAnimator.
> animation-timeline-detached-no-crash.html is removed as the test was not valid
> due to the animation's document not being active.
>
> Committed: https://crrev.com/b45c39903df3727d21cbc217a0b14c2c85b0715a
> Cr-Commit-Position: refs/heads/master@{#356776}
>
> Committed: https://crrev.com/ddd06a6080e1e7d7d02b15595289940980b04973
> Cr-Commit-Position: refs/heads/master@{#357044}

TBR=samli@chromium.org,dstockwell@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1425143006

Cr-Commit-Position: refs/heads/master@{#357050}

BUG=532905

Review URL: https://codereview.chromium.org/1425143004

Cr-Commit-Position: refs/heads/master@{#357049}

Missing these header files cause unclean builds.

TBR=haraken
BUG=548915

Review URL: https://codereview.chromium.org/1409033006

Cr-Commit-Position: refs/heads/master@{#357048}

This patch address the following corner cases:

1. If OnConnectionError() is invoked while there are queued requests, this
happens when extension is woken up but MRPM did not register with MR
successfully.
2. Wake up extension failed.
3. After connection is established (RegisterMediaRouteProvider), we try to run
all pending requests, but it's possible that the extension is already suspended
at that point. It looks like an uncommon timing issue and does not appear to
indicate a connection error.

For 1 and 3, we will retry the wakeup (up to a max number of attempts) since
there is no way to tell if the error was transient. For 2, it is usually due to
a nonretriable error, so we will just drain the queue.

The retry / drain logic also ensures pending requests do not sit in the queue
indefinitely.

Also, introduced a max limit queue size. When the queue size exceeds that limit,
the oldest pending request will be dropped.

A couple of things considered but not implemented:
- Try to fail the pending request in some way instead of dropping them. First,
not all types of requests can be failed, i.e. majority of them do not contain
callbacks. Also because it is a corner case, the added complexity probably does
not justify it.
- Adding a permanent_failure bit to MediaRouterMojoImpl and automatically drop
requests if set to true. Locking down the media router due to connection error
feels too restrictive and I am not sure if it will help.

BUG=490787

Review URL: https://codereview.chromium.org/1419853003

Cr-Commit-Position: refs/heads/master@{#357047}

Controlled by the AutofillFieldMetadata trial, we send the form name as described in the bug.

BUG=548366

Review URL: https://codereview.chromium.org/1422853003

Cr-Commit-Position: refs/heads/master@{#357046}

Previously we were filtering out PropertyHandle types while iterating over
an AnimationStack.
This change pushes the filter into the construction of the AnimationStack
to avoid having undesirable entries in the first place.

Review URL: https://codereview.chromium.org/1412423004

Cr-Commit-Position: refs/heads/master@{#357045}

All animation clocks for a page should tick at the same time, but since we
had an animation clock per document there were some cases where clocks
could tick independently.

This patch moves the clock from Document to PageAnimator.
animation-timeline-detached-no-crash.html is removed as the test was not valid
due to the animation's document not being active.

Committed: https://crrev.com/b45c39903df3727d21cbc217a0b14c2c85b0715a
Cr-Commit-Position: refs/heads/master@{#356776}

Review URL: https://codereview.chromium.org/1410313004

Cr-Commit-Position: refs/heads/master@{#357044}

The first reland had two mistakes.
  1) It tried to turn on Mach-based shared memory on 10.6, which isn't supported.
  2) It hit a race condition in AttachmentBrokering code in GPU conformance tests.

I submitted two speculative fixes for (2), but haven't been able to reproduce the problem
locally, so can't verify that the problem is definitively fixed. Please revert this CL if
problems are seen in GPU conformance tests.

> The CL changes the default behavior of resource buffer to use a Mach-backed
> SharedMemory region. This CL adds a field trial to measure the effect of Mach vs
> POSIX backed SharedMemory regions.
>
> BUG=547261, 466437
> Committed: https://crrev.com/e0e2fd398f3d07c8eebbe662d71a0f9286862476
> Cr-Commit-Position: refs/heads/master@{#356513}

BUG=547261, 466437
TBR=mark@chromium.org, isherman@chromium.org, asvitkine@chromium.org, avi@chromium.org

Review URL: https://codereview.chromium.org/1406123010

Cr-Commit-Position: refs/heads/master@{#357043}

Also updating interface due to thread safety concerns. Unit tests
have been modified to ensure that Observers and Validators can call
DeviceCapabilities methods safely without hitting deadlocks or
unexpected behavior.

TEST=cast_base_unittests
BUG=

Review URL: https://codereview.chromium.org/1409173006

Cr-Commit-Position: refs/heads/master@{#357042}

This change adds a method in StyleResolver which checks the animation-name
property of an element and returns all CSSOM keyframes rule representations
for that element. This method can then be used in InspectorCSSAgent to propogate
to the front-end and show the keyframes there.

BUG=266276

Review URL: https://codereview.chromium.org/1371723002

Cr-Commit-Position: refs/heads/master@{#357041}

Add to shelf on ChromeOS installs a bookmark app, which uses an
extension under the hood. Extension blacklist policies which prevent
extensions from being installed also disable the add to shelf feature
silently. This CL skips the policy checks for bookmark app installation,
allowing add to shelf to function when extensions are blacklisted.

BUG=545541
R=benwells

Review URL: https://codereview.chromium.org/1409393009

Cr-Commit-Position: refs/heads/master@{#357040}

These functions were introduced by [1] and all callsites were removed by [2].

[1] https://codereview.chromium.org/810583002
[2] https://codereview.chromium.org/1237353002

BUG=548977
TBR=rdevlin.cronin@chromium.org

Review URL: https://codereview.chromium.org/1429753002

Cr-Commit-Position: refs/heads/master@{#357039}

Revert of Change shortcut install location to non-subdir. (patchset #23 id:440001 of https://codereview.chromium.org/1289333005/ )

Reason for revert:
Cause of http://crbug.com/548965 and http://crbug.com/548964, need to figure out what's going on there before this goes to a larger population.

Only Canary was affected for now and users can recover through uninstall/reinstall (or even potentially just reboot if it's a Windows shell caching issue).

Original issue's description:
> Change shortcut install location to non-subdir.
>
> BUG=169669
>
> Committed: https://crrev.com/c5bab94fdde60b2bc9ed93822f35c5ce50202d26
> Cr-Commit-Position: refs/heads/master@{#356371}

TBR=grt@chromium.org,jhawkins@chromium.org,benwells@chromium.org,bcwhite@chromium.org
BUG=169669,548965,548964

Review URL: https://codereview.chromium.org/1405993006

Cr-Commit-Position: refs/heads/master@{#357038}

The registration was lost during refactoring how OOP apps are registered in
MojoShellContext.

BUG=403462
TEST=GN build with 'enable_mojo_media = "utility"' works with HTML5 audio.

Review URL: https://codereview.chromium.org/1425113003

Cr-Commit-Position: refs/heads/master@{#357037}

Cr-Commit-Position: refs/heads/master@{#357036}

In the GN build, chrome_android_core has a dependency on chrome/common
but it wasn't specified; this was causing a config to not get
correctly propagated up and a header file to get miscompiled as a
result.

R=pkotwicz@chromium.org, jam@chromium.org
BUG=547635

Review URL: https://codereview.chromium.org/1407353007

Cr-Commit-Position: refs/heads/master@{#357035}

- Introduces a new scheme "exe" which loads an executable alongside mojo_runner. TODO is some more security checking around this to prevent arbitrary apps from launching executables but since this is implemented in mojo_runner it's not in any production code right now.
- Includes an apptest of a .mojo app that connects to an executable and binds a service provided by it.
- Creates a new child library that makes it easy to bring up a runner controller thread in an arbitrary executable.

R=jam@chromium.org
http://crbug.com/548694

Review URL: https://codereview.chromium.org/1419293003

Cr-Commit-Position: refs/heads/master@{#357034}

SVG <use> can be used to gain access to a UA ShadowRoot, from there you could
then appendChild an input and querySelector across the boundary, or alter the
internals of the ShadowRoot during the SVG <use> cloning process.

This patch introduces a check in the callback queue that makes us drop all
callbacks for elements in UA shadows.

Review URL: https://codereview.chromium.org/1427783004

Cr-Commit-Position: refs/heads/master@{#357033}

This adds PageLoad.Timing2.NavigationToFirstContentfulPaint histogram,
which is the minimum of first text paint and first image paint.

BUG=542943

Review URL: https://codereview.chromium.org/1423263003

Cr-Commit-Position: refs/heads/master@{#357032}

BUG=n/a

Review URL: https://codereview.chromium.org/1420723009

Cr-Commit-Position: refs/heads/master@{#357031}

This change defers application of SVG Web Animations until after we
have sampled all current Web Animations. This allows us to group
InvalidatableInterpolations into PropertyHandle specific animation
stacks and composite them appropriately.

There are no SVGInterpolationTypes implemented yet that support
non-replace composite modes to test the additive pipeline.

BUG=530436

Review URL: https://codereview.chromium.org/1420553004

Cr-Commit-Position: refs/heads/master@{#357030}

A special resize callback is unnecessary because the decoder has access to the GLSurface and can call Resize on that.

Review URL: https://codereview.chromium.org/1420503011

Cr-Commit-Position: refs/heads/master@{#357029}