GitLab

This will allow custom media renderers to instantiate
default renderer factory, which expects media log.

BUG=457959

Review URL: https://codereview.chromium.org/943903002

Cr-Commit-Position: refs/heads/master@{#317359}

This patch fixes an exception thrown while running ChromiumLinkerTest
which was caused by null Context. The solution is to make
ChromeLinkerTestApplication be a subclass of BaseChromiumApplication
because it automatically calls ApplicationStatus.initialize() in
onCreate().

BUG=459669

Review URL: https://codereview.chromium.org/941953003

Cr-Commit-Position: refs/heads/master@{#317358}

patch from issue 937933002 at patchset 1 (http://crrev.com/937933002#ps1)

BUG=

Review URL: https://codereview.chromium.org/942843002

Cr-Commit-Position: refs/heads/master@{#317357}

Revert of Revert of Revert of Now that the cq runs tests on rel, the compile-only bug should be dbg. (patchset #1 id:1 of https://codereview.chromium.org/941363002/)

Reason for revert:
`findbugs` is running and failing. Undo this again while I figure that out.

Original issue's description:
> Revert of Revert of Now that the cq runs tests on rel, the compile-only bug should be dbg. (patchset #1 id:1 of https://codereview.chromium.org/914023007/)
>
> Reason for revert:
> The regular android dbg bots also do static buils, so trying this again. I'm watching the bot.
>
> Original issue's description:
> > Revert of Now that the cq runs tests on rel, the compile-only bug should be dbg. (patchset #1 id:1 of https://codereview.chromium.org/937853002/)
> >
> > Reason for revert:
> > Too many red builds on the dbg not, need to investigate tomorrow. Also, it does a static build for some reason.
> >
> > Original issue's description:
> > > Now that the cq runs tests on rel, the compile-only bug should be dbg.
> > >
> > > Depends on https://codereview.chromium.org/913223007
> > >
> > > BUG=456396
> > > TBR=sergiyb@chromium.org
> > >
> > > Committed: https://crrev.com/3d0efdf9afa56e55b8d8c62fd883a1992f35cf45
> > > Cr-Commit-Position: refs/heads/master@{#316998}
> >
> > TBR=jam@chromium.org,sergiyb@chromium.org
> > NOPRESUBMIT=true
> > NOTREECHECKS=true
> > NOTRY=true
> > BUG=456396
> >
> > Committed: https://crrev.com/f3a85349d2da16aa306f1fdc43ea3b5d21a9f956
> > Cr-Commit-Position: refs/heads/master@{#316999}
>
> TBR=jam@chromium.org,sergiyb@chromium.org
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=456396
>
> Committed: https://crrev.com/89f83d5a826c0bfdee8a8cafea307e84d8958df3
> Cr-Commit-Position: refs/heads/master@{#317341}

TBR=jam@chromium.org,sergiyb@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=456396

Review URL: https://codereview.chromium.org/947513002

Cr-Commit-Position: refs/heads/master@{#317356}

BUG=423481

Review URL: https://codereview.chromium.org/937203002

Cr-Commit-Position: refs/heads/master@{#317355}

This pulls in the following Native Client changes:

33b6a18: (yusukes@chromium.org) Implement faster NaClSysFutex functions for NACL_LINUX
0eec380: (yusukes@chromium.org) Revert "Implement faster NaClSysFutex functions for NACL_LINUX"
8dfb438: (bradnelson@google.com) Adding android opt bot to the default trybot list.
c5b278d: (mseaborn@chromium.org) irt.h: Add a comment to document that futexes are process-private
62b6f76: (bradnelson@google.com) Pull in new llvm rev, drop old build flag.
43b8d01: (shyamsundarr@chromium.org) use objdump instead of validator's disassembler in proofs.
10c2124: (jfb@chromium.org) PNaCl: Update LLVM revision in pnacl/COMPONENT_REVISIONS
456d6e4: (yusukes@chromium.org) Implement faster NaClSysFutex functions for NACL_LINUX

BUG=none
TEST=browser_tests and nacl_integration
CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_chromium_dbg_32_ng,linux_arm_compile,linux_nacl_sdk_build

Review URL: https://codereview.chromium.org/942033003

Cr-Commit-Position: refs/heads/master@{#317354}

Currently the automatic option of "Enable touch events" flag cannot
send the touch events to the pages, because the touch screen for
Linux is considered as an attach-slave device, the render thread sets
the "touchEnable" to false in this case. We now take consider both
floating-salve device and attach-slave device in order to send the
touch events to render.

BUG=443667

Review URL: https://codereview.chromium.org/914393002

Cr-Commit-Position: refs/heads/master@{#317353}

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/87fa631..792c80f

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

Review URL: https://codereview.chromium.org/945913002

Cr-Commit-Position: refs/heads/master@{#317352}

BUG=452193

Review URL: https://codereview.chromium.org/941033002

Cr-Commit-Position: refs/heads/master@{#317351}

There was a typo that caused the class to confuse megabytes and kilobytes.

BUG=442546

Review URL: https://codereview.chromium.org/939283002

Cr-Commit-Position: refs/heads/master@{#317350}

Previously WebURLLoaderImpl would keep track of cancelling the request by reseting bridge_. Now that we use request_id_ there instead, we need to do the same and reset its value.

BUG=460335,338338

Review URL: https://codereview.chromium.org/941223002

Cr-Commit-Position: refs/heads/master@{#317349}

The bubble isn't instantiated yet and the new code is probably completely removed by linker.
See the screenshots in the bug tracker.

BUG=400674

Review URL: https://codereview.chromium.org/943933004

Cr-Commit-Position: refs/heads/master@{#317348}

BUG=

Review URL: https://codereview.chromium.org/936553002

Cr-Commit-Position: refs/heads/master@{#317347}

Use GetPreferredDataTypes rather than GetActiveDataTypes when performing sync check for chrome suggestions. This ensure we do not clear Chrome Suggestions when the sync service is starting up and not yet active, unless the user explicitely indicated he didn't want to sync history.

BUG=459729

Review URL: https://codereview.chromium.org/940043002

Cr-Commit-Position: refs/heads/master@{#317346}

BUG=395700

Review URL: https://codereview.chromium.org/942893002

Cr-Commit-Position: refs/heads/master@{#317345}

When frame size changes VideoDecoderShim is supposed to dismiss all
textures with the old size, but it wasn't doing it properly.

BUG=460280

Review URL: https://codereview.chromium.org/937413002

Cr-Commit-Position: refs/heads/master@{#317344}

This CL updates the documentation for vpnProvider interface to
reflect the behavior of the API.

BUG=459609
TBR=kalman@chromium.org

Review URL: https://codereview.chromium.org/941343002

Cr-Commit-Position: refs/heads/master@{#317343}

It's not clear what MPArch.RWH_TabSwitchPaintDuration histogram measures on Android.

BUG=460084

Review URL: https://codereview.chromium.org/944013002

Cr-Commit-Position: refs/heads/master@{#317342}

Revert of Revert of Now that the cq runs tests on rel, the compile-only bug should be dbg. (patchset #1 id:1 of https://codereview.chromium.org/914023007/)

Reason for revert:
The regular android dbg bots also do static buils, so trying this again. I'm watching the bot.

Original issue's description:
> Revert of Now that the cq runs tests on rel, the compile-only bug should be dbg. (patchset #1 id:1 of https://codereview.chromium.org/937853002/)
>
> Reason for revert:
> Too many red builds on the dbg not, need to investigate tomorrow. Also, it does a static build for some reason.
>
> Original issue's description:
> > Now that the cq runs tests on rel, the compile-only bug should be dbg.
> >
> > Depends on https://codereview.chromium.org/913223007
> >
> > BUG=456396
> > TBR=sergiyb@chromium.org
> >
> > Committed: https://crrev.com/3d0efdf9afa56e55b8d8c62fd883a1992f35cf45
> > Cr-Commit-Position: refs/heads/master@{#316998}
>
> TBR=jam@chromium.org,sergiyb@chromium.org
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=456396
>
> Committed: https://crrev.com/f3a85349d2da16aa306f1fdc43ea3b5d21a9f956
> Cr-Commit-Position: refs/heads/master@{#316999}

TBR=jam@chromium.org,sergiyb@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=456396

Review URL: https://codereview.chromium.org/941363002

Cr-Commit-Position: refs/heads/master@{#317341}

Disable flaking keyboard tests. Seems these tests have been flaking for a while, disable them until the root cause can be determined.

BUG=459420

Review URL: https://codereview.chromium.org/942563002

Cr-Commit-Position: refs/heads/master@{#317340}

(It rarely, if ever, makes sense to call isLocked() unless you've already
locked the thing.  We're keeping it only as a concession for asserts.)

BUG=

Review URL: https://codereview.chromium.org/924943008

Cr-Commit-Position: refs/heads/master@{#317339}

This patch stops the linker test runner from printing each character in
the logcat output on a separate line.

BUG=460468

Review URL: https://codereview.chromium.org/948433002

Cr-Commit-Position: refs/heads/master@{#317338}

This removes the BrowserCompositorOutputSurfaceProxy class. It adds
a UpdateVSyncParameters callback to CommandBufferProxyImpl similar to
its sibling SwapBuffersCompleted callback.

R=sievers
BUG=329552

Review URL: https://codereview.chromium.org/939373002

Cr-Commit-Position: refs/heads/master@{#317337}

This CL introduces a dump provider which is able to capture
the process Resident Set Size and the corresponding unittest.

BUG=442731,458295

Review URL: https://codereview.chromium.org/934323002

Cr-Commit-Position: refs/heads/master@{#317336}

With this change the platformKeys API is functional.
The only restriction it is effectively disabled if the user is policy managed.
This restriction will be removed once the according policy is in place.

Depends on
https://codereview.chromium.org/932553002/

BUG=450167

Review URL: https://codereview.chromium.org/927293002

Cr-Commit-Position: refs/heads/master@{#317335}

The semantics of VideoCaptureDevice::Client::Buffer do not belong
where they are now [1]. Make it pure virtual and move the implementation
to VCController. Rename the derived class to show its purpose.

This also forces adaptation in StubClient of
web_contents_video_capture...unittest.cc.

[1] https://code.google.com/p/chromium/codesearch#chromium/src/media/video/capture/video_capture_device.h&sq=package:chromium&type=cs&l=176

BUG=440843

Review URL: https://codereview.chromium.org/941833004

Cr-Commit-Position: refs/heads/master@{#317334}

The frame size we receive from AVFoundation is sometimes excessive, e.g. the size of the uncompressed frame. For example, for a JPEG YUV422, the reported size might be 1280*720*2 = 1.8MB, while the actual compressed data is only 260kB.

In these cases, libyuv::MJPGToI420 spends a lot of time in libyuv::ValidateJpeg, as much as 25%. The reason for this is that ValidateJpeg starts at the end of the buffer and loops backwards looking for the End Of Image (EOI) marker.

This CL tries to optimize these cases by finding the EOI by looping forwards instead of backwards, and reporting the actual frame size to libyuv::MJPGToI420.

BUG=346634

Review URL: https://codereview.chromium.org/912293005

Cr-Commit-Position: refs/heads/master@{#317333}

Initial patch in https://codereview.chromium.org/848613003 used
DCHECK in too many methods that should function even when
the adapter is not present.

BUG=452705

Review URL: https://codereview.chromium.org/933823003

Cr-Commit-Position: refs/heads/master@{#317332}

In order to remove the dependency of HistoryBackend on AndroidProviderBackend,
remove the adapter method that just forward their implementation to it. Also
remove methods that advance or delete a AndroidStatement as they do depend on
the HistoryBackend.

Convert the adapter method in adapter free function wrapped in HistoryDBTask
so that the code only use the public API of HistoryService and HistoryBackend.

Return the base::CancelableTaskTracker::TaskId from the ScheduleDBTask method
on HistoryService so that AndroidHistoryProviderService can return it and the
task be properly cancellable.

BUG=453790

Review URL: https://codereview.chromium.org/930363002

Cr-Commit-Position: refs/heads/master@{#317331}

BUG=371859

Review URL: https://codereview.chromium.org/935053003

Cr-Commit-Position: refs/heads/master@{#317330}

Once NPAPI is pushed behind a flag in M42, enterprise polices that are set need to re-enable NPAPI plugins correctly.

This CL covers all the policy cases and is intended to be a wide net to catch any/all policies.  We don't particularly mind if this has a few false positives, because NPAPI will be totally gone soon.

BUG=295137
TEST=Set e.g. EnabledPlugins policy.  Check that NPAPI plugins are listed in chrome://plugins

Review URL: https://codereview.chromium.org/930243008

Cr-Commit-Position: refs/heads/master@{#317329}

These are the new binary files from
https://codereview.chromium.org/935053003/

TBR=bshe
BUG=371859

Review URL: https://codereview.chromium.org/933043007

Cr-Commit-Position: refs/heads/master@{#317328}

BUG=None
TEST=None

Review URL: https://codereview.chromium.org/933053002

Cr-Commit-Position: refs/heads/master@{#317327}

BUG=

Review URL: https://codereview.chromium.org/944673002

Cr-Commit-Position: refs/heads/master@{#317326}

Revert of Roll libyuv d204db6...194f740 (r1172..1285) (patchset #1 id:1 of https://codereview.chromium.org/943513002/)

Reason for revert:
Failing Windows DrMemory memory test: content
Error report:
07:13:11 drmemory_analyze.py [INFO] Found 1 error reports
07:13:11 drmemory_analyze.py [INFO] Report #1
### BEGIN MEMORY TOOL REPORT (error hash=#FBECECDC73272255#)
UNINITIALIZED READ: reading 0x001ff960-0x001ff980 32 byte(s) within 0x001ff960-0x001ff980
# 0 content.dll!CopyRow_AVX                                                    [third_party\libyuv\source\row_win.cc:2687]
# 1 content.dll!CopyRow_Any_AVX                                                [third_party\libyuv\source\row_any.cc:672]
# 2 content.dll!CopyPlane                                                      [third_party\libyuv\source\planar_functions.cc:72]
# 3 content.dll!I420Copy                                                       [third_party\libyuv\source\convert.cc:90]
# 4 content.dll!I420Rotate                                                     [third_party\libyuv\source\rotate.cc:1193]
# 5 content.dll!ConvertToI420                                                  [third_party\libyuv\source\convert_to_i420.cc:241]
# 6 content.dll!content::VideoCaptureController::VideoCaptureDeviceClient::OnIncomingCapturedData [content\browser\renderer_host\media\video_capture_controller.cc:514]
# 7 content::VideoCaptureControllerTest_DataCaptureInEachVideoFormatInSequence_Test::TestBody [content\browser\renderer_host\media\video_capture_controller_unittest.cc:661]
# 8 testing::internal::HandleExceptionsInMethodIfSupported<>                   [testing\gtest\src\gtest.cc:2420]
Note: @0:05:20.487 in thread 3640
Note: instruction: vmovdqu (%eax) -> %ymm0

Original issue's description:
> Roll libyuv d204db6...194f740 (r1172..1285)
>
> This roll is primarily done to pick up ValidateJpeg performance improvement.
>
> TBR=fbarchard
>
> Committed: https://crrev.com/50bf61b785adf9366446bc26c09e691a1c0db15f
> Cr-Commit-Position: refs/heads/master@{#317304}

TBR=fbarchard@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/941573004

Cr-Commit-Position: refs/heads/master@{#317325}

This CL makes inclusion list and exclusion list to be passed to
shill during setparameters API call in third party VPN.

BUG=459609
TBR=stevenjb@chromium.org

Review URL: https://codereview.chromium.org/940423002

Cr-Commit-Position: refs/heads/master@{#317324}

When used in manual mode, SetZoom() should early-out if the new zoom
level is the same as the old one in order to avoid sending unnecessary
ZoomChangedEvents.

BUG=459995

Review URL: https://codereview.chromium.org/942023002

Cr-Commit-Position: refs/heads/master@{#317323}

The problem was that if the wallpaper of the non active user got changed on a different system, the remote system would replace the wallpaper of the active desktop with the change, no matter which user currently was active.

JS is supplying the browser-context to the api call which can be used to get the applications owner - and that is used to identify the proper user.

Creating a test case for this scenario is a lot of work (setting up multi profile, running the wallpaper manager for both users, triggering a change in the app of one user to force a change, waiting until the change has trickled through the system, ..).
Since the work required for an automated test feels rather complex and time consuming for this setup and the change itself is fairly straight forward and simple, so I did not create a test.

BUG=459599
TEST=visual

Review URL: https://codereview.chromium.org/939233004

Cr-Commit-Position: refs/heads/master@{#317322}

BatteryStatusObserver::Stop can be called twice from
BatteryStatusService::Shutdown, so power_handle_ and batter_change_handle_
should be set to NULL on the first call, otherwise the second call
will pass invalid handles to UnregisterNotification.

Review URL: https://codereview.chromium.org/924193006

Cr-Commit-Position: refs/heads/master@{#317321}

WebRTC 8439:8443
Changes: https://chromium.googlesource.com/external/webrtc/trunk/webrtc.git/+log/00316a4..de41287

Libjingle 8437:8444
Changes: https://chromium.googlesource.com/external/webrtc/trunk/talk.git/+log/92478da..0366d0b

TBR=niklase@chromium.org

Review URL: https://codereview.chromium.org/940413002

Cr-Commit-Position: refs/heads/master@{#317320}