GitLab

- Adding an horizontal Separator between the list of apps and the "Rememeber
my choice" checkbox.
- Adjusting the row size for app candidates from 40 to 32, reducing the
icon from 20 to 16 pt.
- Changing the color of the InkDrop from black to gray.

Bug: 678141
Test: Manually checked the UI. Build
Change-Id: I5697d0aa3e0aa2b2f2b2e9cb627061c43c74f587
Reviewed-on: https://chromium-review.googlesource.com/825963Reviewed-by: Yusuke Sato <yusukes@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: David Jacobo <djacobo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523962}

Change-Id: Ic856f6d5c20ab05ca566c18b8e0490f792f472ef
Reviewed-on: https://chromium-review.googlesource.com/825763
Commit-Queue: Joel Einbinder <einbinder@chromium.org>
Reviewed-by: Andrey Lushnikov <lushnikov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523961}

It supports transforming attributes into css values.

Bug: none
Change-Id: I67c2b67807ceed0ad3d5872aa019d1e59ef3b724
Reviewed-on: https://chromium-review.googlesource.com/826129
Commit-Queue: Dmitry Gozman <dgozman@chromium.org>
Reviewed-by: Andrey Lushnikov <lushnikov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523960}

Modify GetShmemTempDir to check for a runtime switch to always use the temp dir
Changes tests in shared_memory_unittest.cc to be parameterized and
test also for when /dev/shm is disabled.

Bug: 736452,792117
Change-Id: Id930ea525980e97c4ab70e37689301370776fc79
Reviewed-on: https://chromium-review.googlesource.com/810325Reviewed-by: Robert Sesek <rsesek@chromium.org>
Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: David Vallet <dvallet@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523959}

This adds some NOTREACHED() on ChromeOS to figure out why tests are
disabling gpu compositing.

R=zmo@chromium.org

Bug: 793303
Change-Id: I6ee99fc334c34c292bcb1e3568108e422d43da4c
Reviewed-on: https://chromium-review.googlesource.com/826382Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Commit-Queue: danakj <danakj@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523958}

As a preparation step to deprecate resource_messages.h, this change
migrates ResourceDispatcherHostTest to LoadingWithMojo. This change
also removes some tests that are no longer needed.

Bug: 788241
Change-Id: I04ca5649fd621404a177d9c2571384a00a0d7741
Reviewed-on: https://chromium-review.googlesource.com/818925
Commit-Queue: Yutaka Hirano <yhirano@chromium.org>
Reviewed-by: Kinuko Yasuda <kinuko@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523957}

As Blink-side maintenance of spellcheck enabled status will be removed,
toggling the Blink-side status no longer makes sense. Hence, any test
code toggling the Blink-side status should be removed.

This patch:

1. Removes internals.setSpellCheckingEnabled() since it is the tool for
   toggling the Blink-side status

2. Removes all the callers from layout tests, most of which are
   redundant. Test cases that really depend on the removed function are
   also removed.

3. An Android test is temporarily disabled since it depends on the
   removed function. It will be re-enabled when the cleanup is finished.

All-in-one patch: crrev.com/c/821496

Removal plan of Blink-side status: https://goo.gl/grtURZ

Bug: 710097
Change-Id: I0ace095c3d082e9168a342164479a69b44813abc
Reviewed-on: https://chromium-review.googlesource.com/825250Reviewed-by: Emil A Eklund <eae@chromium.org>
Reviewed-by: Ryan Landay <rlanday@chromium.org>
Reviewed-by: Changwan Ryu <changwan@chromium.org>
Commit-Queue: Xiaocheng Hu <xiaochengh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523956}

There are a few cases where a user might open a
SystemWebDialogDelegate while NetworkConfigView is open.

NetworkConfigView uses SystemTrayClient::CreateUnownedDialogWidget
which does not interact well with SystemWebDialogDelegate
(see issue).

Since NetworkConfigView will soon be deprecated, the simplest
short term fix is to not show any SystemWebDialogDelegate
instance while NetworkConfigView is shown. This will not seem
to strange to a user since NetworkConfigView is a modal dialog.

Bug: 791955
Change-Id: Ibe37ff737df69d5d02670c435124e2837ddb0bab
Reviewed-on: https://chromium-review.googlesource.com/825626Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Reviewed-by: Mitsuru Oshima <oshima@chromium.org>
Commit-Queue: Steven Bennetts <stevenjb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523955}

It was replaced by setting actual focus. To avoid a huge CL start with
hardcoding the flag as true and some small obvious cleanups.

Bug: 766807
Change-Id: I3d16523e276ad4bddc772f516fc3ff1d5d5d8e5f
Reviewed-on: https://chromium-review.googlesource.com/804685
Commit-Queue: Vadim Tryshev <vadimt@chromium.org>
Reviewed-by: Yury Khmel <khmel@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523954}

Some of these were from the old enable_mash_packaged_app flag (which
no longer exists) and from when the test_chrome telemetry binary
had separate dependencies (it doesn't anymore).

Bug: none
Test: compiles
Change-Id: I15dc5fb447302d4ff86c0819595092067f3c71ba
Reviewed-on: https://chromium-review.googlesource.com/823078Reviewed-by: Ken Rockot <rockot@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523953}

Instead of calling back into the ProxyDelegate interface this patch adds
the alternate proxy server information to the ProxyInfo class so that it
can be populated by DataReductionProxyDelegate during the initial call
to OnResolveProxy.

This reduces the number of round-trips between //net and the Data
Reduction Proxy implementation which should simplify the Network Service
implementation.

Bug: 721403
Change-Id: I8cc4280901325c39346a595fbaa41d03995102be
Reviewed-on: https://chromium-review.googlesource.com/821416Reviewed-by: Randy Smith <rdsmith@chromium.org>
Reviewed-by: Tarun Bansal <tbansal@chromium.org>
Commit-Queue: Reilly Grant <reillyg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523952}

DataReductionProxy would like to know about dogfood.

Bug: 771727
Change-Id: I7e69a9447abf49bec78dff8b36e927e942bd7448
Reviewed-on: https://chromium-review.googlesource.com/823254
Commit-Queue: Ryan Sturm <ryansturm@chromium.org>
Reviewed-by: Tarun Bansal <tbansal@chromium.org>
Reviewed-by: Steven Holte <holte@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523951}

In preparation for more printing related servicification (PDF to EMF
converter, printer capabilities...), creating a new Chrome printing
service to host these interfaces and moving PdfToPwgRasterConverter
service there.

There is no change in logic.

Bug: 792225
Change-Id: I2c71a9e035dafc1ade84e71da6d71a9ca38277cd
Reviewed-on: https://chromium-review.googlesource.com/809619Reviewed-by: Ken Rockot <rockot@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: Jay Civelli <jcivelli@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523950}

Previously, we didn't reset |shared_tab_loader_| every time when a test cast
got torn down, resulting in DCHECK failure when we tried to restore tabs in
another test case. This patch resets |shared_tab_loader_| every time when a
test cases is being torn down.

BUG=793625

Change-Id: Ic21faedd1748f6c9d26f1d493d97f7679c8e2dca
Reviewed-on: https://chromium-review.googlesource.com/823250Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: lpy <lpy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523949}

This new class avoids having separate mock definitions in
render_widget_host_view_(aura|mac)_unittest.(cc|mm).

Having a single mock definition will simplify writing
common tests for occlusion tracking in
RenderWidgetHostView(Aura|Mac).

Bug: 668690
Change-Id: I7d7c0c39bf266014374df0b307eecc2a9047820d
Reviewed-on: https://chromium-review.googlesource.com/797851Reviewed-by: Ken Buchanan <kenrb@chromium.org>
Reviewed-by: Alex Moshchuk <alexmos@chromium.org>
Commit-Queue: François Doray <fdoray@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523948}

I modified this file manually in the previous patch, and made a
typo.

BUG=711468
R=skobes@chromium.org

Change-Id: I63cc6c285ea47a10dd41aec4de4049413311a0ed
Reviewed-on: https://chromium-review.googlesource.com/824946Reviewed-by: Steve Kobes <skobes@chromium.org>
Commit-Queue: Stefan Zager <szager@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523947}

https://pdfium.googlesource.com/pdfium.git/+log/2614250d4f11..604f5b9e0422

$ git log 2614250d4..604f5b9e0 --date=short --no-merges --format='%ad %ae %s'
2017-12-13 rharrison Refactor CFXA_*List constructors

Created with:
  roll-dep src/third_party/pdfium


The AutoRoll server is located here: https://pdfium-roll.skia.org

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=dsinclair@chromium.org

Change-Id: I706ee144900976b61077f4f466dbb4709d5e9abb
Reviewed-on: https://chromium-review.googlesource.com/825983
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523946}

Otherwise it results in a situation that keydown is delivered to the
client but keyup is not. The consequence is never-ending key repeat
in the client side.

BUG=794442
TEST=manually on Google Sheets app.

Change-Id: I573b75ad62c37603dc0074de6ceffa7a773b5f8b
Reviewed-on: https://chromium-review.googlesource.com/823372Reviewed-by: Yuichiro Hanada <yhanada@chromium.org>
Reviewed-by: David Reveman <reveman@chromium.org>
Commit-Queue: Kazuhiro Inaba <kinaba@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523945}

As described.

Test: Manually + browser test extended
Bug: b/70525239
Bug: b/70525068
Bug: 794698

Change-Id: I3f3280e25ccaaea8cfc9636fcd5a700f3eb7dbaf
Reviewed-on: https://chromium-review.googlesource.com/822768
Commit-Queue: Yury Khmel <khmel@google.com>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523944}

The DefaultProtocolHandlers TabMetrics field is limited to schemes
the user has accepted the site as the default protocol handler for.

Bug: 784639
Change-Id: I21f5c924a668c1b983744fcec4656c6b04245452
Reviewed-on: https://chromium-review.googlesource.com/797355
Commit-Queue: Michael Giuffrida <michaelpg@chromium.org>
Reviewed-by: Martin Šrámek <msramek@chromium.org>
Reviewed-by: Bryan McQuade <bmcquade@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: Matt Giuca <mgiuca@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523943}

mDNS discovery can trigger a firewall prompt on a fresh user-level
Chrome install on Windows or if there are no existing firewall rules
for mDNS traffic(?). This patch applies the same gating logic to
the mdns logic in in-browser cast discovery.

TODO: merge this patch back to M64.

Bug: 794317
Change-Id: I5df1a2f60eb3a8411b642fc8996f2e0279912401
Reviewed-on: https://chromium-review.googlesource.com/823244
Commit-Queue: Derek Cheng <imcheng@chromium.org>
Reviewed-by: mark a. foltz <mfoltz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523942}

Bug: 786173
Change-Id: Idca7bbe2c209c8cf75930550e4462ca2b7ccdda3
Reviewed-on: https://chromium-review.googlesource.com/824702
Commit-Queue: Ilya Sherman <isherman@chromium.org>
Reviewed-by: Nathan Parker <nparker@chromium.org>
Reviewed-by: Ilya Sherman <isherman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523941}

Replaced elements (except SVGRoot which uses the normal overflow clip/
border-radius clip rule) with border-radius will apply border radius clip
on their foreground, so we need to output InnerBorderRadiusClip for them,
and set proper chunk properties when painting their foregrounds.

Bug: 794295
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Change-Id: I206db7fcf55a99187cf1f8961f5f286d482f1bf2
Reviewed-on: https://chromium-review.googlesource.com/823177
Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523940}

Bug: 
Change-Id: Ia72a555a03f5eb6a4b9ec9f5639f137c1ac60450
Reviewed-on: https://chromium-review.googlesource.com/826162Reviewed-by: David Trainor <dtrainor@chromium.org>
Commit-Queue: Michael Thiessen <mthiesse@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523939}

On Windows, the folder upload prompt previously used "OK" as the default
button text, leading to the possibility that the user would not
understand that their selected folder would be uploaded to the server.

This CL matches other platforms and uses a localized string ("Upload")
as the text of the prompt's commit button.

Bug: 780881
Change-Id: I256d42fc8ba3fdd72d5af360585509b6573b7900
Reviewed-on: https://chromium-review.googlesource.com/777681Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Eric Lawrence <elawrence@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523938}

kMaxBufferCount is defined in the anonymous namespace in two places in
//services/video_capture, which conflict in jumbo builds.  But since
each is only used in one location, they can be converted into function
statics instead.

BUG=794692

Change-Id: Ibf930a73d426ff7ef53fb921ec146952b84bc3de
Reviewed-on: https://chromium-review.googlesource.com/825823Reviewed-by: Christian Fremerey <chfremer@chromium.org>
Commit-Queue: Mostyn Bramley-Moore <mostynb@vewd.com>
Cr-Commit-Position: refs/heads/master@{#523937}

https://crrev.com/c/795479 does the resizing at
|CompositorViewHolder.onLayout| but it is not the ideal place to
get the content view size/browser control info in. The view size can
still remain zero when a tab is being initialized from links on CCT
as reported int the bug.
The better place where we can be sure |ContentView.onSizeChanged|
was invoked and the size is set as expected is
|LayoutChangeListener.onLayoutChanged|.
This CL moves resizing logic accordingly.

Bug: 794288
Change-Id: I12fc771ba5850b79a418a003618b7cbc94678207
Reviewed-on: https://chromium-review.googlesource.com/822316Reviewed-by: Matthew Jones <mdjones@chromium.org>
Commit-Queue: Jinsuk Kim <jinsukkim@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523936}

Bug: 598854
Cq-Include-Trybots: master.tryserver.chromium.android:android_arm64_dbg_recipe;master.tryserver.chromium.android:android_compile_x64_dbg;master.tryserver.chromium.android:android_compile_x86_dbg
Change-Id: I9aae58daee5ba3eedbfc4601e1fabdab18013912
Reviewed-on: https://chromium-review.googlesource.com/822953
Commit-Queue: Robert Sesek <rsesek@chromium.org>
Reviewed-by: Jorge Lucangeli Obes <jorgelo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523935}

Document that chromeos::PowerManagerClient's
SetBacklightsForcedOff method doesn't support multiple
callers, and that ash::BacklightsForcedOffSetter should be
used instead.

Bug: 760730
Change-Id: Id083f9e1b3d1b252fa259ae519cf12c07135fa5f
Reviewed-on: https://chromium-review.googlesource.com/825948Reviewed-by: Toni Barzic <tbarzic@chromium.org>
Commit-Queue: Dan Erat <derat@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523934}

Bug introduced with: crrev.com/c/652867
The sync cell didn't include an image when having an error. This was the cause of an UI glitch.

Updating to EarlGrey tests to make sure an image is set in the sync cell.

Bug: 793822
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: I189becd8e30ddaf5ee8d2d20f9cbd20258a20014
Reviewed-on: https://chromium-review.googlesource.com/821230
Commit-Queue: Jérôme Lebel <jlebel@chromium.org>
Reviewed-by: Eric Noyau <noyau@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523933}

Bug: 794688
Change-Id: I2e57bc570e8a8dc249c5fb53f83b561997caab5a
Reviewed-on: https://chromium-review.googlesource.com/825628Reviewed-by: Ryan Hamilton <rch@chromium.org>
Commit-Queue: Ryan Hamilton <rch@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523932}

This CL implements exising metrics reported by CRWPassKitDownloader.
UMA constants are duplicated, which is fine because old constants will
be removed along with CRWPassKitDownloader class.

Bug: 789735
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: I455e9247ed4493dbc7f3d59a74e1f5806e167956
Reviewed-on: https://chromium-review.googlesource.com/822528Reviewed-by: Gregory Chatzinoff <gchatz@chromium.org>
Commit-Queue: Eugene But <eugenebut@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523931}

This change would allow us to propagate extension install source to
the component updater protocol builder.

Bug: 
Change-Id: I8daa94f9a07c15b9301b1ca2a3f8cacdfd7fa894
Reviewed-on: https://chromium-review.googlesource.com/823289
Commit-Queue: Minh Nguyen <mxnguyen@chromium.org>
Reviewed-by: Joshua Pawlicki <waffles@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523930}

This reverts commit b592ae80.

Reason for revert:
Caused LLD assert on Google Chrome Linux x64.

Original change's description:
> Roll clang 318667:320471.
> 
> Bug: 787920
> 
> Cq-Include-Trybots: master.tryserver.chromium.android:android_arm64_dbg_recipe;master.tryserver.chromium.android:android_compile_x64_dbg;master.tryserver.chromium.android:android_compile_x86_dbg
> Change-Id: Ifcb76b0d95acc65eafbf58da624f9fbc920f7924
> Reviewed-on: https://chromium-review.googlesource.com/823219
> Reviewed-by: Robert Sesek <rsesek@chromium.org>
> Reviewed-by: Jorge Lucangeli Obes <jorgelo@chromium.org>
> Reviewed-by: Peter Collingbourne <pcc@chromium.org>
> Commit-Queue: Hans Wennborg <hans@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#523910}

TBR=jorgelo@chromium.org,hans@chromium.org,rnk@chromium.org,pcc@chromium.org,rsesek@chromium.org

Change-Id: Icbbd0da2640660ba0529236d5b800d1cfde357f5
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 787920
Cq-Include-Trybots: master.tryserver.chromium.android:android_arm64_dbg_recipe;master.tryserver.chromium.android:android_compile_x64_dbg;master.tryserver.chromium.android:android_compile_x86_dbg
Reviewed-on: https://chromium-review.googlesource.com/826322Reviewed-by: Hans Wennborg <hans@chromium.org>
Commit-Queue: Hans Wennborg <hans@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523929}

net/data/verify_certificate_chain_unittest/.

Moves the utility code to net/data/gencerts for easier re-use by other
tests.

Change-Id: Ifa6172def50f185e1a5b7d42e4073e2da003fbd9
Reviewed-on: https://chromium-review.googlesource.com/823288Reviewed-by: Matt Mueller <mattm@chromium.org>
Commit-Queue: Eric Roman <eroman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523928}

In ARIA 1.1, aria-posinset and aria-setsize was added
to "article" role.

This patch will add above properties to article,
and include some changes for "aria-posinset"
in content_browsertests.

Bug: none
Change-Id: I98647cc4d938a0dcf054290f55bbddbdc7eab087
Reviewed-on: https://chromium-review.googlesource.com/810328
Commit-Queue: Nektarios Paisios <nektar@chromium.org>
Reviewed-by: Nektarios Paisios <nektar@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523927}

This CL updates unit tests as a preparation for
CachedMetadataHandler changes in [1].

This CL
- Calls Resource::SetResponse() in ScriptStreamerTest.cpp and
  V8ScriptRunnerTest.cpp, because [1] will depend on
  SetResponse().
- Sets ResourceResponse's URL, because CachedMetadataHandler
  depends on response URLs.
- Uses http: URLs instead of data: URLs in ResourceTest, because
  CachedMetadataHandler is not set for data: URLs.

These are also natural, as
- SetResponse() is called and response URLs are set in non-test code.
- revalidation never occurs on data: URLs.

This CL also adds ResourceTest.RevalidationFailed, and
adds unit test assertions related to CachedMetadataHandler.

[1] https://chromium-review.googlesource.com/786410

Bug: 784875
Change-Id: Ia58d3c6c9b2ad20c502d8ed9c69d505aa73be03b
Reviewed-on: https://chromium-review.googlesource.com/812569
Commit-Queue: Hiroshige Hayashizaki <hiroshige@chromium.org>
Reviewed-by: Kouhei Ueno <kouhei@chromium.org>
Reviewed-by: Nate Chapin <japhet@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523926}

This is workaround for b/70566216 to unblock tests and functionality
until it is properly adrressed. This manually corrects right call order
for ArcAppListPrefs.

Test: Unit tests passed, cheets_AuthPerf passed without perf regression.
Bug: b/70566216
Bug: b/70530828
Change-Id: I18e8455c3e46a7160b7719d1ca400c41cba6f5f9
Reviewed-on: https://chromium-review.googlesource.com/821562Reviewed-by: Stefan Kuhne <skuhne@chromium.org>
Reviewed-by: Yusuke Sato <yusukes@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Yury Khmel <khmel@google.com>
Cr-Commit-Position: refs/heads/master@{#523925}

Work around a High Sierra fullscreen bug by getting rid of fractional pixel differences in video aspect ratio.

When a 720x1280 video is made fullscreen on a 1440x900 DIP screen, this
happens to its dimensions:

    720x1280 -[scale]-> 506.25x900 -[integer]-> 506x900

The AVSampleBufferDisplayLayer used for video content layers has its
videoGravity set to AVLayerVideoGravityResize, which doesn't preserve
aspect ratio. Under certain circumstances, the small difference in
aspect ratio triggers a macOS bug with low power detachment where the
video's x position changes rapidly each frame as long as the window
stays in detached mode. I've been able to make it happen if:

- The machine is a MacBook Pro.
- The OS is at least 10.13.
- The display is set to native 2x resolution ("Looks like 1440x900").
- The machine is running on integrated graphics if it has dual graphics.

A recording of the issue is attached to the crbug.

This workaround is sort of terrible and maybe Apple can suggest a better
one: when a video layer's aspect ratio can be made to exactly match the
source aspect ratio by adding a fractional pixel to either dimension, do
so (making sure to keep the video centered).

Bug: 792632
Change-Id: I8b149606cbccb6033c36d3ad400f7a63a4d462ce
Reviewed-on: https://chromium-review.googlesource.com/812114
Commit-Queue: Sidney San Martín <sdy@chromium.org>
Reviewed-by: ccameron <ccameron@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523924}

Bug: None
Change-Id: I8d447df58d9231f2b112ef261ee6b522fafe0fc1
Reviewed-on: https://chromium-review.googlesource.com/826047Reviewed-by: Robert Sesek <rsesek@chromium.org>
Reviewed-by: Greg Kerr <kerrnel@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523923}