GitLab

Version 2 additions:

- set_parent request that can be used to provide an initial position
  relative to parent for shell surfaces.

- zaura_output extension to wl_output that provides detailed
  information about supported scale factors.

Test: compile
Bug: 
Change-Id: I00aebee4475a7004db84f6f0e2cb5d41d8a823da
Reviewed-on: https://chromium-review.googlesource.com/804354
Commit-Queue: David Reveman <reveman@chromium.org>
Reviewed-by: Daniele Castagna <dcastagna@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521077}

Component build was broken on Fuchsia because
base::RaiseProcessToHighPriority() was not defined.

Bug: 706592
Change-Id: Ie88ffa6c7ad09729249417c6fb094e2194026854
Reviewed-on: https://chromium-review.googlesource.com/804210
Commit-Queue: Sergey Ulanov <sergeyu@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Reviewed-by: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521076}

This can't be done by mucking with ShouldWindowContentsBeTransparent() as I
opriginally thought; update comment there to describe better why.

Instead, do this directly in TabStripImpl.

The net effect of this should be that on Win 10, the titlebar background color
will no longer show through the background tabs in the default theme; for white
titlebars, this should make active and inactive tabs a little easier to
distinguish.

BUG=none
TEST=none

Change-Id: I1b722ba7ce887a1a32470e794e04b436bc06c22b
Reviewed-on: https://chromium-review.googlesource.com/802522Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Peter Kasting <pkasting@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521075}

For both the initial check and the subsequent verification, we were
using the url to be loaded to handle DAL verification. But DAL is for
origins only, so we should instead split the path from the url before
sending it for verification

BUG=782905

Change-Id: I380076b3c1de5d10fdaba14238609fde9fdc3463
Reviewed-on: https://chromium-review.googlesource.com/801986
Commit-Queue: Yusuf Ozuysal <yusufo@chromium.org>
Reviewed-by: Yaron Friedman <yfriedman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521074}

We have added Chrome.ProcessSingleton.TerminateProcessErrorCode.Windows
histogram here: https://codereview.chromium.org/2871793003/

Sadly, it was added to process_singleton_win.cc with mispell as:
Chrome.ProcessSingleton.ProcessTerminateErrorCode.Windows

This CL fixes histogram name to proper one.

R=gab@chromium.org,holte@chromium.org

Change-Id: Ice0989f3646aec6942138788e3cce0ebeb4dcffa
Reviewed-on: https://chromium-review.googlesource.com/800630
Commit-Queue: Steven Holte <holte@chromium.org>
Reviewed-by: Jochen Eisinger <jochen@chromium.org>
Reviewed-by: Steven Holte <holte@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521073}

This adds the new TabIcon view to the TabExperimental class and adds the
associated plumbing to hook up the basic favicon and throb functionality.

Rename experimental observer variables to help differentiate from the
regular observers.

Bug: 778461
Change-Id: I56a67a6f14d4003ee49818568bb59298e15bf1b8
Reviewed-on: https://chromium-review.googlesource.com/804206
Commit-Queue: Brett Wilson <brettw@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521072}

This change:

* Ensures that the Observers are notified when an Instance's connection
  errors out.
* Relaxes the restrictions upon calling SetInstance() when there is
  already an InstancePtr bound. This matches the semantics prior to the
  ConnectionHolder refactoring.

Bug: None
Test: git cl try
Change-Id: I933792f2846dfe614b8d71a52c9f9ebbce3f7c09
Reviewed-on: https://chromium-review.googlesource.com/797970
Commit-Queue: Luis Hector Chavez <lhchavez@chromium.org>
Reviewed-by: Hidehiko Abe <hidehiko@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521071}

This CL disables tests failing on the Site Isolation Win FYI bot until
they can be investigated and fixed.

Bug: 788390
Change-Id: Ie51cd7dc3fe3559da7e2e3eec8ad9ca19169d794
Reviewed-on: https://chromium-review.googlesource.com/804146Reviewed-by: Alex Moshchuk <alexmos@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521070}

Perf bisects indicate that the hb_set changes may be the reason for
seing performance regressions in issues 781794 and 782220. This is
somewhat speculative as there were issues with reproducibility in the
perf_bisect runs. Results from the main perf bot graphs should help
figure out whether the hb_set changes were the culprit here.

Bug: 782220, 781794
Change-Id: I5462b37503ca76a299c99cf41388d0d195722358
Tbr: behdad@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/804374Reviewed-by: Dominik Röttsches <drott@chromium.org>
Commit-Queue: Dominik Röttsches <drott@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521069}

This CL launches xvfb with +extension RANDR to suppress an error message of the
form:

Xlib:  extension "RANDR" missing on display ":99".

that gets printed by Xlib on every test run on the bots.  This message cannot be
suppressed AFAIK since it's part of libx11.

R=dpranke@chromium.org

Change-Id: I50e249102a082074dbbc5f501921a2a8d0d99c0b
Reviewed-on: https://chromium-review.googlesource.com/802624
Commit-Queue: Thomas Anderson <thomasanderson@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521068}

Previously, MCVD ran on a thread that was owned by the mojo service.
If the service was destroyed, then the thread would be also.  This
made it difficult for things to outlive MCVD (e.g., overlays).

Now, it is sufficient to hold the task runner instead.  This CL
removes the ServiceContextRef, in favor of the task runner.

Bug: 789447
Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_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
Change-Id: I1f07382a3977685d6a92f6f13a92d5455602d6ec
Reviewed-on: https://chromium-review.googlesource.com/801334Reviewed-by: Sergey Volk <servolk@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Chris Watkins <watk@chromium.org>
Reviewed-by: Xiaohan Wang <xhwang@chromium.org>
Reviewed-by: Dan Sanders <sandersd@chromium.org>
Commit-Queue: Frank Liberato <liberato@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521067}

Adds both a positive check and negative check.

Bug: 787185
Change-Id: I4a5c9c450d4500604bbfc4690f7ef6fcc643c87a
Reviewed-on: https://chromium-review.googlesource.com/804198Reviewed-by: Ryan Sturm <ryansturm@chromium.org>
Commit-Queue: Doug Arnett <dougarnett@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521066}

flow is complete.

When showing the second run promo, Chrome checks if the
first run promo has completed. This CL adds a switch to
disable that check if a command line switch is set.
This switch would be used for UI Automator tests.

Bug: 752972,752973
Change-Id: I345fbc89dd7c51cdca4ddf1203a9e1e29045cab4
Reviewed-on: https://chromium-review.googlesource.com/780264
Commit-Queue: Tarun Bansal <tbansal@chromium.org>
Reviewed-by: Tommy Nyquist <nyquist@chromium.org>
Reviewed-by: Tarun Bansal <tbansal@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521065}

Update the GetPrintersURL call to utilize metadata_v2. This allows for certain
restrictions to be placed on the model to determine whether or not the printer should be displayed

Bug: chromium:773766
Change-Id: Iaade00489a066340a095b3b07bbd853f13f8e35a
Reviewed-on: https://chromium-review.googlesource.com/797476
Commit-Queue: Luum Habtemariam <luum@chromium.org>
Reviewed-by: Sean Kau <skau@chromium.org>
Reviewed-by: Steven Bennetts <stevenjb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521064}

changes:
When window is added to a workspace from another workspace, its current
window bounds possibly is not the original-not-changed-by-user bounds,
for example a resized bounds truncated by available workarea.

Using a member in ash::wm::WindowState to store the windows bounds. It
is used in window added to workspace event from another workspace. It is
set when window is removing from a workspace and it is nullptr (using
pre_auto_manage_window_bounds() if it exists otherwise current window bounds).
It gets reset when window bounds is changed by user.

DisplayMoveWindowUtilTest.KeepWindowBoundsIfNotChangedByUser, also
tested on emulator. Not-user-resized window bounds will not affect the
original bounds when moving between displays.

Bug: 785017, 778438
Test: added the test coverage in
Change-Id: I37d2c79238ea7200b32b5e3a9d5064addc575976
Reviewed-on: https://chromium-review.googlesource.com/770891
Commit-Queue: Qiang(Joe) Xu <warx@chromium.org>
Reviewed-by: Mitsuru Oshima <oshima@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521063}

Bug: 789509
Cq-Include-Trybots: master.tryserver.chromium.linux:closure_compilation
Change-Id: I0f25e5f282508975ba2b969c1c1f7cf74fc51c30
Reviewed-on: https://chromium-review.googlesource.com/801118
Commit-Queue: Wenzhao (Colin) Zang <wzang@chromium.org>
Reviewed-by: Alexander Alekseev <alemate@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521062}

Query the --enable-viz flag and plumb it to the DelegatedFrameHost.

Resize the ui::Compositor along with the NSView when under viz, because
otherwise it will never be resized (in non-viz, we resize it when we
get compositor frames, which we no longer get).

Consolidate the places that we query NSView/display properties (size,
scale factor, and color space) in BrowserCompositorMac.

Bug: 794839
Change-Id: I3d62f340a8c318c963f24f3732cdde4bb9328126
Reviewed-on: https://chromium-review.googlesource.com/794840
Commit-Queue: ccameron <ccameron@chromium.org>
Reviewed-by: danakj <danakj@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521061}

https://boringssl.googlesource.com/boringssl/+log/fc9c67599d9bdeb2e0467085133b81a8e28f77a4..48eaa28a124e702edf373e212011cb283f0f0161

BUG=none

Change-Id: I4e722641068bb53d07ad06c634657d79960faab8
Reviewed-on: https://chromium-review.googlesource.com/803840
Commit-Queue: David Benjamin <davidben@chromium.org>
Commit-Queue: Steven Valdez <svaldez@chromium.org>
Reviewed-by: Steven Valdez <svaldez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521060}

Bug: chromedriver:2105
Change-Id: I7a4274ae6f670d39071809a34e66fbac09001fba
Reviewed-on: https://chromium-review.googlesource.com/803631Reviewed-by: John Chen <johnchen@chromium.org>
Commit-Queue: John Chen <johnchen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521059}

This debugging/presentation feature was tied to Ctrl-Alt-P. Usage is
very low per UMA metrics. We suspect most usage was from users
mistakenly hitting it when pressing Ctrl-P to print. Since it was tied
to a pref the user could get stuck in this state. It also didn't work
at the login screen and would turn on and off when switching
multiprofile users.

This feature is useful for debugging, recording videos of touch bugs,
and for projecting demos. UI review suggested moving it to about:flags
alongside the other touch debug flags.

* Remove the pref and add cleanup code to wipe the existing value.
* Remove the accelerator key.
* Add --show-taps. The name is inspired by the Android OS developer
  option that does the same thing.

TBR=rockot@chromium.org

Bug: 786115
Test: existing ash_unittests
Change-Id: I07579bff6ab3dd16ec5c14347df1a047b1aa935c
Reviewed-on: https://chromium-review.googlesource.com/802075
Commit-Queue: James Cook <jamescook@chromium.org>
Reviewed-by: Dominic Battré <battre@chromium.org>
Reviewed-by: James Cook <jamescook@chromium.org>
Reviewed-by: Ilya Sherman <isherman@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521058}

Consume fixed storage disk events in StorageMonitorChromeos.
Only observe stateful partition, because it is relevant for users.

Bug: 420633
Change-Id: Ia235b8f60b94c396c90da856bc68e3a61b2292dd
Reviewed-on: https://chromium-review.googlesource.com/786649
Commit-Queue: Aga Wronska <agawronska@chromium.org>
Reviewed-by: Steven Bennetts <stevenjb@chromium.org>
Reviewed-by: Lei Zhang <thestig@chromium.org>
Reviewed-by: Toni Barzic <tbarzic@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521057}

The [SecureContext] extended attribute now applies correctly to
interfaces, so make sure that the StorageManager interface is not
exposed in non-secure contexts, not just the `navigator.storage` entry
point.

Bug: 726510
Change-Id: Ie141aa4e29e261f1107af292f74f8063370c3748
Reviewed-on: https://chromium-review.googlesource.com/802676Reviewed-by: Victor Costan <pwnall@chromium.org>
Commit-Queue: Joshua Bell <jsbell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521056}

This is an update to https://chromium-review.googlesource.com/783707 that
ensures that blocking GetCollectStatsConsent is used only on a blocking thread.

Bug: 782641
Change-Id: I5629e00616f1ab24b4a63f1c3433f683370c3239
Reviewed-on: https://chromium-review.googlesource.com/802764
Commit-Queue: Alexander Alekseev <alemate@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521055}

BUG=none
TEST=none

Change-Id: If31621b94082b463ec2d90f7f8d59f11e6052bed
Reviewed-on: https://chromium-review.googlesource.com/802975Reviewed-by: Kevin Bailey <krb@chromium.org>
Commit-Queue: Peter Kasting <pkasting@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521054}

PAGE_DONE is not useful on non-Windows platforms. Stop dispatching the
event in these cases.

Bug: None
Change-Id: I317b8d7ecd265c8386d474316685bcfb7113061c
Reviewed-on: https://chromium-review.googlesource.com/802822
Commit-Queue: Rebekah Potter <rbpotter@chromium.org>
Reviewed-by: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521053}

FieldTrials should be recorded through sync when an event logger client
registers a dependency, their dependent event is recorded, and the
trial is finalized. All trials are recorded whenever the FieldTrial
event is emitted, which should be on a change or period, but no more.
Feature [and params] exist but may never actually change.

Because FieldTrials (and other imminent types) are not tied to
navigations, reworked the ShouldRecordEvent() check to be more
permissive, but also enforce presence (or lack) of navigation id.

Bug: 719038
Change-Id: Ib62a18949dde41d194935347768b3fef0d2623d6
Reviewed-on: https://chromium-review.googlesource.com/764408
Commit-Queue: Sky Malice <skym@chromium.org>
Reviewed-by: Alexei Svitkine <asvitkine@chromium.org>
Reviewed-by: Pavel Yatsuk <pavely@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521052}

Distinguish between the permission errors the broker process
fabricates vs. organically ocurring EPERM from a syscall.

Use block-scope rather than unique_ptrs to force broker destruction.

Fix a dubious use of ScopedFD(). ScopeFD's only legal bad FD value
is -1, which co-incidentally happens to be -EPERM.  So don't apply
this where we know the file is invalid, since other negative values
cause a CHECK().

Convert some ASSERTS to EXPECTS, to avoid crashes.

Change-Id: I822e15e1ff386687627ba27c8f9fff158c10f316
Reviewed-on: https://chromium-review.googlesource.com/797919
Commit-Queue: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Robert Sesek <rsesek@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521051}

It is possible to set fixed sample rate in cast_shell on Linux with
--alsa-fixed-output-sample-rate, but it wasn't possible on Fuchsia.
Added --audio-output-sample-rate, which allows to set fixed sample rate
on both platforms. For backward compatibility kept the old ALSA-specific
flag, but it's marked as deprecated now.

Bug: 772488
Change-Id: I0f980f09b432768ff9bb51191a9b1ec31e89274f
Reviewed-on: https://chromium-review.googlesource.com/798066Reviewed-by: Luke Halliwell <halliwell@chromium.org>
Reviewed-by: Kenneth MacKay <kmackay@chromium.org>
Commit-Queue: Sergey Ulanov <sergeyu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521050}

This change migrates the omnibox sizing, layout, text and colors to
match UX spec. Things like rounded corners are still missing.

Also, fix the UI testapp.

BUG=
R=bshe

Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_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
Change-Id: I96bc91c8c80d070dd3e09275d35eb61774468559
Reviewed-on: https://chromium-review.googlesource.com/801954Reviewed-by: Biao She <bshe@chromium.org>
Commit-Queue: Christopher Grant <cjgrant@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521049}

Adding a bound accessor to service_manager::Connector so clients that
require a fresh (unbound) connector can check that.

Bug: None
Change-Id: I555e32b71d7259713544f6283c8868827409382c
Reviewed-on: https://chromium-review.googlesource.com/803907Reviewed-by: Ken Rockot <rockot@chromium.org>
Commit-Queue: Jay Civelli <jcivelli@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521048}

The current logic uses ResourceDispatcherHostDelegate.
This doesn't work when network service is enabled.
This CL rewrites the logic using ContentBrowserClient,
so test can work with and without network service enabled

BUG=715630

Cq-Include-Trybots: master.tryserver.chromium.linux:linux_mojo
Change-Id: I109a7f3fafe6a7d81a98ef39f2b828e7e0551cf0
Reviewed-on: https://chromium-review.googlesource.com/801370Reviewed-by: John Abd-El-Malek <jam@chromium.org>
Reviewed-by: Matt Menke <mmenke@chromium.org>
Commit-Queue: Min Qin <qinmin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521047}

Bug: 784510
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: I3ca7d45d6c5cf04aa97a787b106a2529daa8e421
Reviewed-on: https://chromium-review.googlesource.com/788450
Commit-Queue: Justin Cohen <justincohen@chromium.org>
Reviewed-by: Rohit Rao <rohitrao@chromium.org>
Reviewed-by: Kurt Horimoto <kkhorimoto@chromium.org>
Reviewed-by: Gauthier Ambard <gambard@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521046}

Personally, I no longer feel confident to review code in ash/wm

BUG=None

Change-Id: I5258309ba7171eb34ad3a3b9b78ca2bb6cdc71f2
Reviewed-on: https://chromium-review.googlesource.com/783800Reviewed-by: Valery Arkhangorodsky <varkha@chromium.org>
Reviewed-by: Mitsuru Oshima <oshima@chromium.org>
Commit-Queue: Peter Kotwicz <pkotwicz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521045}

This change adds arc::WaitForInstanceReady(), so that tests that
manually invoke SetInstance() can wait until the instance has been fully
connected in a safe, race-free way.

Bug: 782895
Test: git cl try
Change-Id: I4480ab78401bb145ba99fa60a1417b2c8c6af94d
Reviewed-on: https://chromium-review.googlesource.com/802191
Commit-Queue: Luis Hector Chavez <lhchavez@chromium.org>
Reviewed-by: Hidehiko Abe <hidehiko@chromium.org>
Reviewed-by: Steven Bennetts <stevenjb@chromium.org>
Reviewed-by: Yusuke Sato <yusukes@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521044}

There was a bug that it was not possible to cancel a notification after
the download was paused. This is because in API < 21, calling
stopForeground(false) would make the notification continue to be
classified as on-going and therefore undismissable, which is what was
happening in the pause case. This CL fixes this problem by inducing a
flicker that dissociates the notification from the foreground service
in the case that the Android version is API < 21.

Bug: 788080
Change-Id: I9c8776d245425b0fe78aaccf7e75fcfb6e57f039
Reviewed-on: https://chromium-review.googlesource.com/802078Reviewed-by: David Trainor <dtrainor@chromium.org>
Commit-Queue: Joy Ming <jming@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521043}

This test now matches the non-RLS expectations except for extra
scrolling layers.

Cq-Include-Trybots: master.tryserver.chromium.linux:linux_layout_tests_root_layer_scrolls
Change-Id: I7eef36da8f1d7bbf04d1c9a3c0e9f6010579b814
Reviewed-on: https://chromium-review.googlesource.com/804204
Commit-Queue: Philip Rogers <pdr@chromium.org>
Commit-Queue: Stefan Zager <szager@chromium.org>
Reviewed-by: Stefan Zager <szager@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521042}

Bug: 624061
Change-Id: I743121dc676720fd92bae8e27a9c1d7e305e54f9
Reviewed-on: https://chromium-review.googlesource.com/755523
Commit-Queue: Patrick Monette <pmonette@chromium.org>
Reviewed-by: Trent Apted <tapted@chromium.org>
Reviewed-by: Evan Stade <estade@chromium.org>
Reviewed-by: Peter Kasting <pkasting@chromium.org>
Reviewed-by: Martin Šrámek <msramek@chromium.org>
Reviewed-by: Charlie Harrison <csharrison@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521041}

BUG=773403

Change-Id: Ie6eb72cb6b83321c3ca2094645bab8340879da9d
Reviewed-on: https://chromium-review.googlesource.com/798217Reviewed-by: Tommy Nyquist <nyquist@chromium.org>
Commit-Queue: Tommy Nyquist <nyquist@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521040}

Should have no change in behavior.

Explicitly generating the IDs on the root node makes it clear
at what point they are generated, and from what tree state.
It also makes it clear that IDs are assigned only once.

Before this change, there was a cache to ensure IDs were generated
only once, but it's not great to rely on a cache for semantics.

Change-Id: I24898c740709231e14e28dbd39e32d45945ddcec
Reviewed-on: https://chromium-review.googlesource.com/803754
Commit-Queue: agrieve <agrieve@chromium.org>
Reviewed-by: Robert Flack <flackr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521039}

Bug: 789711
Change-Id: Iae6bc67ba67f8f9e492bd065a8d75903b7cf92a9
Reviewed-on: https://chromium-review.googlesource.com/801457
Commit-Queue: Mathieu Perreault <mathp@chromium.org>
Reviewed-by: Roger McFarlane <rogerm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#521038}