GitLab

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/4bf1ce2..e67164d

CQ_EXTRA_TRYBOTS=tryserver.blink:linux_blink_rel,linux_blink_dbg
TBR=stephana@google.com

Review URL: https://codereview.chromium.org/890893008

Cr-Commit-Position: refs/heads/master@{#314258}

BUG=448817

Review URL: https://codereview.chromium.org/850013004

Cr-Commit-Position: refs/heads/master@{#314257}

[DOMStorage] Offset the session storage namespace ids generated by different contexts to help identify when an id from one is mistakenly used in another.

Review URL: https://codereview.chromium.org/881253004

Cr-Commit-Position: refs/heads/master@{#314256}

This adds details about each individual DisplayItem to
DisplayItemList's trace output. Each DrawingDisplayItem's
SkPicture is also included in the trace output.

Review URL: https://codereview.chromium.org/884473005

Cr-Commit-Position: refs/heads/master@{#314255}

This is proving useful for test filters involving >1 test, which I'm
using heavily as I work on the net test server.

BUG=428729

Review URL: https://codereview.chromium.org/892273002

Cr-Commit-Position: refs/heads/master@{#314254}

It's flaky on Mac and Linux.
BUG=442752
TBR=horo@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/876593004

Cr-Commit-Position: refs/heads/master@{#314253}

This method is no longer used. All Android code assumes that new profile
management is enabled.

BUG=393335
R=aruslan@chromium.org

Review URL: https://codereview.chromium.org/890113002

Cr-Commit-Position: refs/heads/master@{#314252}

This adds a new third_party directory, android_data_chart, which
contains the data usage chart UI widget from Android's Settings app. The
widget shows a graph of how much data the user has used and saved over
time.

BUG=428885
R=bengr@chromium.org, cpu@chromium.org, dannyb@google.com, tedchoc@chromium.org

Review URL: https://codereview.chromium.org/866813004

Cr-Commit-Position: refs/heads/master@{#314251}

ChromeSwitchPreference uses setWidgetLayoutResource() to show a Material
switch, but on BLU Life Play devices, this has no effect and the default
switch is shown instead. As a result ChromeSwitchPreference.onBindView()
couldn't find the Material switch and was crashing with a NPE. This
fixes the crash by adding a null check, though these devices will still
show a non-material switch.

BUG=451447
R=dtrainor@chromium.org

Review URL: https://codereview.chromium.org/894873004

Cr-Commit-Position: refs/heads/master@{#314250}

BUG=454037

Review URL: https://codereview.chromium.org/878523005

Cr-Commit-Position: refs/heads/master@{#314249}

Otherwise, on Ash, it seems to steal the initial focus away from the
widget.

BUG=453284
TEST=Start Chrome with --enable-new-avatar menu, and in Metro mode.
Choose "Switch Person" from the avatar menu. In the user manager, start
tabbing (without clicking on the user manager). This should work, and
tabbing should change the selected user pod.

Review URL: https://codereview.chromium.org/894723003

Cr-Commit-Position: refs/heads/master@{#314248}

BUG=433539

Review URL: https://codereview.chromium.org/879993002

Cr-Commit-Position: refs/heads/master@{#314247}

Revert of [Cast] Software encoder support for varying video frame sizes. (patchset #3 id:60001 of https://codereview.chromium.org/877393003/)

Reason for revert:
Linux MSAN fails with uninit memory access.

Original issue's description:
> [Cast] Software encoder support for varying video frame sizes.
>
> Adds support for automatic on-line reconfiguration of the VP8 software
> encoder whenever the video frame size changes.  libvpx supports
> shrinking frame sizes without tearing down the encoder, but growing
> frames sizes still requires a full tear-down and re-creation.
>
> Most of this change involves adding extensive unit testing to confirm
> media/cast is now capable of handling variable frames sizes end-to-end.
> In addition, the cast_sender_app and cast_receiver_app diagnostic tools
> have been updated.
>
> An upcoming change will eliminate the width and height properties from
> media::cast::VideoSenderConfig, which are both being ignored now.
>
> BUG=451277
>
> Committed: https://crrev.com/dbba1332d2d3c6d03d0e86f5dd72164b38b11442
> Cr-Commit-Position: refs/heads/master@{#314196}

TBR=hubbe@chromium.org,miu@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=451277

Review URL: https://codereview.chromium.org/899583002

Cr-Commit-Position: refs/heads/master@{#314246}

Revert of Fixing inbox test by adding sha1 files and enabling inbox benchmark again. (patchset #1 id:1 of https://codereview.chromium.org/883693006/)

Reason for revert:
Still flaky on Mac 10.9 Perf (5)

Original issue's description:
> Fixing inbox test by adding sha1 files and enabling inbox benchmark again.
>
> BUG=452257
>
> Committed: https://crrev.com/fd0b94cd15337a91d167a3b931ac4252e1174042
> Cr-Commit-Position: refs/heads/master@{#314164}

TBR=sullivan@google.com,tonyg@google.com,aiolos@google.com,dtu@chromium.org,tonyg@chromium.org,manisharora@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=452257

Review URL: https://codereview.chromium.org/893073005

Cr-Commit-Position: refs/heads/master@{#314245}

This CL implements Register function for BluetoothAudioSinkChromeOS
and comes up with the corresponding tests. For the tests,
FakeBluetoothMediaClient is changed to simulate
the behavior of the
real D-Bus APIs.

BUG=441581
TEST=device_unittests --gtest_filter=*AudioSink*

Review URL: https://codereview.chromium.org/876153002

Cr-Commit-Position: refs/heads/master@{#314244}

ui_controls functions are typically only used in interactive_ui_tests
since it's not supported in sharded test suites.

We want to use its SynthesizeKeyEvent() to generate events in
event_generator_delegate_mac.mm (which is designed for non-interactive
tests and can be sharded).

This CL moves it to cocoa_test_event_utils.h

BUG=378134

Review URL: https://codereview.chromium.org/882053002

Cr-Commit-Position: refs/heads/master@{#314243}

Windows, Linux, and Chrome OS use the cross-platform
RenderTextHarfBuzz class as of M40. Remove the
platform-specific RenderText implementations.

Also update chrome://flags so that only Mac users can
control whether RenderTextHarfBuzz is used. RenderTextMac is
still used there by default.

BUG=321868,398885

Review URL: https://codereview.chromium.org/891013003

Cr-Commit-Position: refs/heads/master@{#314242}

For MediaPlayerListener, the callbacks can happen on a non-UI thread.
Weak_ptr is not thread safe, so we shouldn't bind the weak_ptr in those callbacks.
This change creates a threadsafe proxy class to bridge all the callbacks.

BUG=453581, 447368, 447367

Review URL: https://codereview.chromium.org/899473002

Cr-Commit-Position: refs/heads/master@{#314241}

Revert of Componentize WebHistoryService (patchset #4 id:60001 of https://codereview.chromium.org/882753002/)

Reason for revert:
BUG=454524

Original issue's description:
> Componentize WebHistoryService
>
> Instead of using the factories to retrieve the dependended upon keyed service,
> pass them explicitly to WebHistoryService from the WebHistoryServiceFactory.
>
> Fix the dependency of WebHistoryService to add the service used and to remove
> the service unused.
>
> Move file into //components/history/core/browser, remove unnecessary #include
> and fix files that did not #include all the files they used.
>
> BUG=371835
>
> Committed: https://crrev.com/12dabec1af75b80952821fee2f37a28637fa15b7
> Cr-Commit-Position: refs/heads/master@{#313926}

TBR=droger@chromium.org,jhawkins@chromium.org,rlp@chromium.org,rogerta@chromium.org,sdefresne@chromium.org
NOPRESUBMIT=true
BUG=371835

Review URL: https://codereview.chromium.org/889553003

Cr-Commit-Position: refs/heads/master@{#314240}

Revert of Unify GPU raster and accelerated canvas blacklists on Android. (patchset #3 id:40001 of https://codereview.chromium.org/877343008/)

Reason for revert:
Causes blank screen on Daisy boot (ChromeOS T604)

BUG=454547

Original issue's description:
> Unify GPU raster and accelerated canvas blacklists on Android.
>
> Accelerated canvas and GPU raster both use the same Skia codepaths and
> have generally displayed the same bugs.  It doesn't make sense to
> blacklist one but not the other.  Since GPU raster sees more usage than
> canvas nowadays, and I researched GPU raster compatibility as part of
> https://codereview.chromium.org/760053003, it makes sense to fold canvas
> into the same blacklist.
>
> This immediately fixes an Adreno 225 canvas bug which slipped through a
> driver-reported-version based blacklist, which is apparently unreliable
> and which I deleted in this patch.  More generally it also should avoid
> this kind of issue in the future.
>
> BUG=453283
> NOTRY=true
>
> Committed: https://crrev.com/044dfb80b1f612f3c84b35182db0f9b4d93e086c
> Cr-Commit-Position: refs/heads/master@{#313813}

TBR=zmo@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=453283

Review URL: https://codereview.chromium.org/895673004

Cr-Commit-Position: refs/heads/master@{#314239}

This deletes stubs and renames all uses in tests for the following
methods that were renamed in https://codereview.chromium.org/800613009:

- LayerImpl::TotalScrollOffset was renamed to CurrentScrollOffset.
  TotalScrollOffset is now a LayerTreeImpl-specific concept meaning the
  sum of the inner and outer viewport current scroll offsets.
- LayerImpl::scroll_offset() was renamed to BaseScrollOffset.  This is
  the latest scroll offset pushed from the main thread.
- LayerImpl::SetScrollOffset() was renamed to
  PushScrollOffsetFromMainThread.  There are two ways to change the
  scroll offset, either pushing from the main thread or directly doing
  SetCurrentScrollOffset, corresponding to main-thread or impl-thread
  scrolling.

BUG=

Review URL: https://codereview.chromium.org/878413005

Cr-Commit-Position: refs/heads/master@{#314238}

Impl sents truncated ScrollDelta to main. Recent change to use
SyncedProperty changes the truncation behavior in the case when
scroll_offset(active_base) is non-integer. This CL restores the
behavior.

The whole flooring will be gone soon when we are able to send
non-integer scroll delta to main. This CL is only meant to make
sure that SyncedProperty switch has no functional change.

BUG=

Review URL: https://codereview.chromium.org/867823007

Cr-Commit-Position: refs/heads/master@{#314237}

Otherwise, minimising it sends it into the void, from where it can
never be recovered (unless you manually re-open it again)

BUG=450959
TEST=Start Chrome with --enable-new-avatar menu, and in Metro mode.
Choose "Switch Person" from the avatar menu. The window should have a
shelf item.

Review URL: https://codereview.chromium.org/891173005

Cr-Commit-Position: refs/heads/master@{#314236}

test_opus_api, test_opus_encode, test_opus_decode, and test_opus_padding
can be run directly, with no command-line options, and they print out
"OK" during the test and an overall status message such as
"All API tests passed." or "Tests completed successfully." at the end.

The celt component has several unit tests. I didn't add build targets
for them because some of them don't print any output, so you'd need a
test harness script to examine their exit status.

R=henrika@chromium.org,sergeyu@chromium.org
BUG=452984

Review URL: https://codereview.chromium.org/894473002

Cr-Commit-Position: refs/heads/master@{#314235}

To reduce the amount of repetition, this patch adds some helper methods
to ServiceWorkerJobTest that run an asynchronous task and verifie the
result.

TEST=content_unittests --gtest_filter='ServiceWorkerJobTest.*'
BUG=

Review URL: https://codereview.chromium.org/893983002

Cr-Commit-Position: refs/heads/master@{#314234}

BUG=444869

Review URL: https://codereview.chromium.org/892923002

Cr-Commit-Position: refs/heads/master@{#314233}

This time it should work properly. I checked all the images manually.

BUG=407781
TBR=newt@chromium.org,tedchoc@chromium.org

Review URL: https://codereview.chromium.org/896633002

Cr-Commit-Position: refs/heads/master@{#314232}

BUG=451321

Review URL: https://codereview.chromium.org/889283002

Cr-Commit-Position: refs/heads/master@{#314231}

1% is a lot less noisy than 15%, and still gives up some visibility into new
regressions.

BUG=384552

Review URL: https://codereview.chromium.org/893083002

Cr-Commit-Position: refs/heads/master@{#314230}

This is done to ensure consistency when naming Polymer elements in the PDF viewer.

BUG=110020

Review URL: https://codereview.chromium.org/867483006

Cr-Commit-Position: refs/heads/master@{#314229}

The function is no longer used by anyone.

Review URL: https://codereview.chromium.org/858173004

Cr-Commit-Position: refs/heads/master@{#314228}

BUG=447462

Review URL: https://codereview.chromium.org/849293003

Cr-Commit-Position: refs/heads/master@{#314227}

Revert of Report viewport memory and timing correctly for Slimming Paint. (patchset #3 id:60001 of https://codereview.chromium.org/867803003/)

Reason for revert:
Reverting due to https://crbug.com/453131

Original issue's description:
> Report viewport memory and timing correctly for Slimming Paint.
>
> The rasterize_and_record benchmark was not using data from the
> DisplayItemList. Now it does.
>
> R=ajuma@chromium.org,vmpstr@chromium.org
> BUG=451448
>
> Committed: https://crrev.com/c76f528131949d7fe60b7facf1e7b40362aacfd5
> Cr-Commit-Position: refs/heads/master@{#313420}

TBR=ajuma@chromium.org,vmpstr@chromium.org,schenney@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=451448

Review URL: https://codereview.chromium.org/885753004

Cr-Commit-Position: refs/heads/master@{#314226}

This is a embarassing followup of http://crrev.com/886153003 which
introduced --disable-mojo-channel flag.

The change didn't work as RenderProcessHostImpl refers the function
that should've removed. This CL removes the useless function whose
name masked the one of the newly added.

R=nasko@chromium.org
BUG=377980

Review URL: https://codereview.chromium.org/891353004

Cr-Commit-Position: refs/heads/master@{#314225}

Revert of Revert of Standardize "gpu_times" metric names and add GPU rasterization versions. (patchset #1 id:1 of https://codereview.chromium.org/896593002/)

Reason for revert:
Reverting the revert; looks like this wasn't the culprit.

Original issue's description:
> Revert of Standardize "gpu_times" metric names and add GPU rasterization versions. (patchset #2 id:20001 of https://codereview.chromium.org/875473003/)
>
> Reason for revert:
> Speculative revert for crbug.com/453131
>
> Original issue's description:
> > Standardize "gpu_times" metric names and add GPU rasterization versions.
> >
> > The GPU Times metric event names have been standardized to match the
> > names for other metrics. GPU Rasterization versions have also been
> > added for both GPU times tests. Here is the list of test names:
> >   gpu_times.gpu_rasterization.key_mobile_sites_smooth
> >   gpu_times.gpu_rasterization.top_25_smooth
> >   gpu_times.key_mobile_sites_smooth
> >   gpu_times.top_25_smooth
> >
> > R=vmiura@chromium.org
> > BUG=none
> > test=trybots
> >
> > Committed: https://crrev.com/9fbe73b4158beb6beb278af819e075cc3d744210
> > Cr-Commit-Position: refs/heads/master@{#313825}
>
> TBR=vmiura@chromium.org,epenner@chromium.org,dyen@chromium.org
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=none
>
> Committed: https://crrev.com/92a5662c9732fb41f2afc2026ed0e6b4e9c1fcd3
> Cr-Commit-Position: refs/heads/master@{#314192}

TBR=vmiura@chromium.org,epenner@chromium.org,dyen@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=none

Review URL: https://codereview.chromium.org/898593004

Cr-Commit-Position: refs/heads/master@{#314224}

TCHAR may be a WCHAR, so sizeof would allow GetModuleFileName to write
twice the size of the buffer.

Thanks to Ilya Berdichevsky for the bug and fix.

BUG=446464

Review URL: https://codereview.chromium.org/898493003

Cr-Commit-Position: refs/heads/master@{#314223}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/7e00d71..fc49b4d

TBR=vsevik@chromium.org,japhet@chromium.org

Review URL: https://codereview.chromium.org/876713006

Cr-Commit-Position: refs/heads/master@{#314222}

Previously LibjingleTransport was emitting RouteChange notification
even before connection is established. As result the reported results
were incorrect and confusing. Specifically it was always reporting
connection as direct first even if it ends up going through the relay.

Review URL: https://codereview.chromium.org/889043002

Cr-Commit-Position: refs/heads/master@{#314221}

Summary of changes available at:
https://chromium.googlesource.com/chromium/third_party/ffmpeg/+log/43b80eb..4cdb373

R=dalecurtis@chromium.org
BUG=454387

Review URL: https://codereview.chromium.org/892153003

Cr-Commit-Position: refs/heads/master@{#314220}

Phase 2 of lucid sleep involves allowing GCM-enabled apps and extensions
to wake up the system to process push messages.  This brings up the
problem of knowing when those apps/extensions are done processing the
event so that the system can go back to sleep.

This CL introduces a new class whose responsibility is to monitor
GCM-enabled apps and extensions and delay the system suspend while they
are processing push messages.  Additionally, network requests that are
started while the app/extension is processing a push message are
considered related to that message and can further delay the system
suspend even after the push message itself has been acked by the
extension.

BUG=397328

Review URL: https://codereview.chromium.org/823703004

Cr-Commit-Position: refs/heads/master@{#314219}