GitLab

Previously, DialogOverlayCore didn't receive a release() call from
DSImpl, since DSImpl drops the reference to DSCore immediately
after notifying the client of surface destruction.  This caused
problems on some cast devices, where the overlay wasn't being
deleted properly.

Now, DSCore calls release() synchronously after waiting for
DSImpl to close the overlay.  This keeps the dismiss() call sync
with respect to onSurfaceDestroyed.  Even if DSImpl did call
DSCore.release(), it would be after onSurfaceDestroyed completed.

Bug: 782282
Change-Id: I25f98643063545e0a44ccb7ab04145e7c6728d3a
Reviewed-on: https://chromium-review.googlesource.com/756932
Commit-Queue: Frank Liberato <liberato@chromium.org>
Reviewed-by: Bo <boliu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515234}

This CL converts a few more tests to use embedded test server. Most of
them are not listed in the mojo filter but also not working correctly.


Change-Id: I02a9e0962a5b3efc61a3fbc05eb9df618c4fc1a7

Bug: 779121, 715630
Change-Id: I02a9e0962a5b3efc61a3fbc05eb9df618c4fc1a7
Reviewed-on: https://chromium-review.googlesource.com/755839
Commit-Queue: Xing Liu <xingliu@chromium.org>
Reviewed-by: Min Qin <qinmin@chromium.org>
Reviewed-by: David Trainor <dtrainor@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515233}

Design doc @ go/locked-fullscreen-mode

Bug: chromium:755696
Change-Id: I78b774a40993b5672f24a2e1ae2f8a8aaa4eed8a
Reviewed-on: https://chromium-review.googlesource.com/723422
Commit-Queue: Ivan Šandrk <isandrk@chromium.org>
Reviewed-by: James Cook <jamescook@chromium.org>
Reviewed-by: Istiaque Ahmed <lazyboy@chromium.org>
Reviewed-by: Maksim Ivanov <emaxx@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515232}

changes:
Crash happens when doing profile switching, we reset SearchController
but didn't reset AppListModel's search results. When switching back,
search controller is accessing stale search results' stale pointers,
which causes the crash.

Bug: 781499
Test: test that crash doesn't happen any more
Change-Id: Id020581d2c94fd8e14637ec25b8bd1ab266fecde
Reviewed-on: https://chromium-review.googlesource.com/758520Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Reviewed-by: Vadim Tryshev <vadimt@chromium.org>
Commit-Queue: Qiang(Joe) Xu <warx@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515231}

Change-Id: Ia56ff060ef0f478d5ce577290cd0322c2ae9aed8
Reviewed-on: https://chromium-review.googlesource.com/759376Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Vadim Tryshev <vadimt@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515230}

Currently, the non-persistent client hints may be sent on non-secure
URLs. This CL changes it so that client hints (persistent or
non-persistent) are sent only on secure or local origins.

Bug: 782381
Change-Id: I50a184d7aa19813eacc59e4d9fca5e74ec8855f2
Reviewed-on: https://chromium-review.googlesource.com/758464
Commit-Queue: Tarun Bansal <tbansal@chromium.org>
Reviewed-by: Ryan Sturm <ryansturm@chromium.org>
Reviewed-by: Kinuko Yasuda <kinuko@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515229}

These tests take a very long time to execute on emulated ARM cores,
but they do pass on real ARM hardware.

Bug: 780652
Change-Id: Iaa608ab980dd63616b9c44b4e5b386340af2664b
Reviewed-on: https://chromium-review.googlesource.com/750076
Commit-Queue: Kevin Marshall <kmarshall@chromium.org>
Reviewed-by: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515228}

Bug: 782864
Change-Id: I445d55f1a17b43b95850aaaf4a5e45237ae87d38
Reviewed-on: https://chromium-review.googlesource.com/759404
Commit-Queue: Doug Arnett <dougarnett@chromium.org>
Reviewed-by: Ryan Sturm <ryansturm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515227}

TBR=chrome-os-gardeners@google.com

Change-Id: I99afde99885e3fbefc0295a6172cec0df53aae49
Reviewed-on: https://chromium-review.googlesource.com/761076Reviewed-by: Steven Bennetts <stevenjb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515226}

Add some margin so the watch expression prompt does not go off screen.
Today, text can overflow outside of the prompt, so we should hide
overflow.

Screenshot: https://imgur.com/a/pQfSV

Bug: 639284
Change-Id: Ib23fc2561ce6895164850474788279e9889fe53d
Reviewed-on: https://chromium-review.googlesource.com/759426Reviewed-by: Dmitry Gozman <dgozman@chromium.org>
Commit-Queue: Erik Luo <luoe@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515225}

Change-Id: I5ac1891d1ceb8b3809b607e6f81b016ed72e912e
Reviewed-on: https://chromium-review.googlesource.com/760639Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Cr-Commit-Position: refs/heads/master@{#515224}

This tests most of the inputs to the oop raster extension so
that we can verify that the GLES2Implementation behaves the
same as RasterSource.

Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel
Change-Id: I4592627423ec543232263401bb9d86e1d4fcf934
Reviewed-on: https://chromium-review.googlesource.com/758489Reviewed-by: Antoine Labour <piman@chromium.org>
Reviewed-by: vmpstr <vmpstr@chromium.org>
Commit-Queue: enne <enne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515223}

Replace home-baked ScopedRTL and other Scoped RTL objects with
the more commonly used version.

Modify ctor of ScopedRestoreICUDefaultLocale to optionally accept a
locale code to set.

Bug: 777997
Change-Id: I23bebd574b008b1f7e8d16f2be5dda780d46ba63
Reviewed-on: https://chromium-review.googlesource.com/742442Reviewed-by: danakj <danakj@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Commit-Queue: Alex Newcomer <newcomer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515222}

In OSX 10.10+, some combination of the titlebar and
toolbar always show in fullscreen mode. We do not want
either to show so we hide the window that contains these
views. This way, only our custom toolbar will show on
the screen.

However, this cause an issue where the titlebar
disappears for windows without our custom toolbar
(Ex/ Chrome app windows and popups). This CL adds a check
so that that we only hide the window if it has a toolbar.

Bug: 601038
Change-Id: Ia125273a534d69b4ad406dc2d5329d31f2d97602
Reviewed-on: https://chromium-review.googlesource.com/758500
Commit-Queue: Sarah Chan <spqchan@chromium.org>
Reviewed-by: Elly Fong-Jones <ellyjones@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515221}

The android targets are not compiled on windows, therefore
the "is_win" related bits can be removed.

(as per original comment on
https://chromium-review.googlesource.com/c/chromium/src/+/738240/1)

BUG=668515

Change-Id: Ibf16b209dc6fac1f83790c2abf40b3add43f5db4
Reviewed-on: https://chromium-review.googlesource.com/753344
Commit-Queue: Tim Volodine <timvolodine@chromium.org>
Reviewed-by: Tim Volodine <timvolodine@chromium.org>
Reviewed-by: Nathan Parker <nparker@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515220}

Bug: 782399
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: Icafc8d57aaa229acb3d66628378aa49146d9b8cd
Reviewed-on: https://chromium-review.googlesource.com/758083Reviewed-by: Eugene But <eugenebut@chromium.org>
Reviewed-by: Kurt Horimoto <kkhorimoto@chromium.org>
Commit-Queue: Kurt Horimoto <kkhorimoto@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515219}

This validates the hints protobuf file and sends a notification when the hints have been validated and processed.

Bug: 777892
Change-Id: I72ca225a0d327f11849f0cd3f3198fe6c2abb06b
Reviewed-on: https://chromium-review.googlesource.com/753290
Commit-Queue: Sophie Chang <sophiechang@chromium.org>
Reviewed-by: Colin Blundell <blundell@chromium.org>
Reviewed-by: Sorin Jianu <sorin@chromium.org>
Reviewed-by: Joshua Pawlicki <waffles@chromium.org>
Reviewed-by: Doug Arnett <dougarnett@chromium.org>
Reviewed-by: Ryan Sturm <ryansturm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515218}

Bug: 720383
Change-Id: Ibfa33fcf247473361e8f7b2496eff480163ffd80
Reviewed-on: https://chromium-review.googlesource.com/758582
Commit-Queue: Michael van Ouwerkerk <mvanouwerkerk@chromium.org>
Reviewed-by: Alexei Svitkine <asvitkine@chromium.org>
Reviewed-by: Bernhard Bauer <bauerb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515217}

This tests is making sure the tests are able to run with
ContentSuggestions enabled.

Bug: 764720
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: Ia7ea99614f6020841273733ae4ea126e3d84155a
Reviewed-on: https://chromium-review.googlesource.com/758648
Commit-Queue: Gauthier Ambard <gambard@chromium.org>
Reviewed-by: Sylvain Defresne <sdefresne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515216}

Bug: 781283
Change-Id: I273fc466d0af9115193972f3d862fb1bcbb6e539
Reviewed-on: https://chromium-review.googlesource.com/757090
Commit-Queue: Ganggui Tang <gogerald@chromium.org>
Reviewed-by: Rouslan Solomakhin <rouslan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515215}

Add EarlGrey tests tryjobs to ios changes.  We want to have EG
tests running for at least ios changes to reduce tree breakages.

Since the ios-simulator-full-configs has only 8 builders, iOS CLs might
experience extra landing time due to the tryjob size and limited capacity.

Bug: 782735
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: I8c00dbe1963893af19c37d7410e6d57c8d37ab5d
Reviewed-on: https://chromium-review.googlesource.com/759166Reviewed-by: Rohit Rao (ping after 24h) <rohitrao@chromium.org>
Commit-Queue: Menglu Huang <huangml@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515214}

v1 packaged apps use the app.launch.local_path key. This was incorrectly
changed when the doc was expanded.

Change-Id: Iebcb59acf3e878364b2ed46f9554a7f0424e6f22
Reviewed-on: https://chromium-review.googlesource.com/728410Reviewed-by: Devlin <rdevlin.cronin@chromium.org>
Commit-Queue: Michael Giuffrida <michaelpg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515213}

Currently, we don't check for the 3 hours to pass before opening the
bottom sheet to half if opened on the tab switcher. This adds that
check.

Bug: 781052
Change-Id: If680a1ed06d5e18161fdc8df60892592a6a55ddb
Reviewed-on: https://chromium-review.googlesource.com/758825Reviewed-by: Ted Choc <tedchoc@chromium.org>
Reviewed-by: Theresa <twellington@chromium.org>
Commit-Queue: Troy Hildebrandt <thildebr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515212}

This is the revert of crrev.com/b77864ad
Turns out that fix causes crbug.com/775526.
In the mean time, revert won't re-trigger the crbug.com/759007 as it's
fixed by crrev.com/9990e87e.

Bug: 759007
Bug: 775526
Change-Id: I2cd5acefb3d8588613629a99dde592f648650a93
Reviewed-on: https://chromium-review.googlesource.com/757718Reviewed-by: anthonyvd <anthonyvd@chromium.org>
Commit-Queue: Owen Min <zmin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515211}

Also fixes the bugs that were uncovered from adding these checks.

Cq-Include-Trybots: master.tryserver.chromium.android:android_cronet_tester;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: Iecac97f4b91c24bc3aeae4b2fc336e39c25df781
Reviewed-on: https://chromium-review.googlesource.com/759690Reviewed-by: Ryan Hamilton <rch@chromium.org>
Commit-Queue: Yixin Wang <wangyix@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515210}

Those tests calls UIView -frame on a background thread which
is an error on iOS 11.0+ causing the tests to fail.

Bug: 782760
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-eg
Change-Id: Iadb1e5d917724dbc6d1341d4de3dcbe2fcf7f3cd
Reviewed-on: https://chromium-review.googlesource.com/758867Reviewed-by: Elodie Banel <lod@chromium.org>
Commit-Queue: Sylvain Defresne <sdefresne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515209}

Those tests calls UIView -frame on a background thread which
is an error on iOS 11.0+ causing the tests to fail.

Bug: 782760
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-eg
Change-Id: I83d4338263bde003000be22ff38ef7199a35da64
Reviewed-on: https://chromium-review.googlesource.com/760361
Commit-Queue: Sylvain Defresne <sdefresne@chromium.org>
Reviewed-by: Elodie Banel <lod@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515208}

AVDA used to receive a "frame hidden" signal from WMPI.  This would
let it give up the overlay in case some other frame needs it.

Now that DialogOverlayImpl destroys the overlay when the WebContents
is hidden / etc., this isn't needed.  It will be handled normally.

Bug: 782827
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: I9b5857f3b37545da2c44462de0424b00de8cb3ae
Reviewed-on: https://chromium-review.googlesource.com/759143Reviewed-by: Thomas Guilbert <tguilbert@chromium.org>
Reviewed-by: Chris Palmer <palmer@chromium.org>
Commit-Queue: Frank Liberato <liberato@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515207}

For net::NetworkQualitiesPrefsManager.

Cq-Include-Trybots: master.tryserver.chromium.android:android_cronet_tester;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: Ie4fddd3363d078dc468d3df61aac1d6336c8ac49
Reviewed-on: https://chromium-review.googlesource.com/760216Reviewed-by: Tarun Bansal <tbansal@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515206}

"Distribution" is likely a word which implies "an old mechanism without IncrementalShadowDOM".
SupportsAssignment would be better here.

Bug: 776656
Change-Id: Id1cfbbb9df054b7f0a47352c15aaf0bd57203792
Reviewed-on: https://chromium-review.googlesource.com/760117
Commit-Queue: Hayato Ito <hayato@chromium.org>
Reviewed-by: Takayoshi Kochi <kochi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515205}

This CL adds some extra info re. support for encrypted streams to
the chrome://gpu video acceleration entry, to avoid the duplication
of formats described in the bug.

Bug: 783100
Change-Id: I06bf1bd450e71cdf2db88ec27f4ffff98814accb
Reviewed-on: https://chromium-review.googlesource.com/760478Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Commit-Queue: Miguel Casas <mcasas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515204}

Promotes ResizeObserver to stable

Intent to ship: https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/z6ienONUb5A

LGTM1 from dglazgov, LGTM2 from foolip, LGTM3 from chrishtr on Oct 17

Bug: 612962
Change-Id: I3dc59fd0ab725acc6f40bf5a3bbd70de5e889f0f
Reviewed-on: https://chromium-review.googlesource.com/754562
Commit-Queue: Aleks Totic <atotic@chromium.org>
Reviewed-by: Yoav Weiss <yoav@yoav.ws>
Cr-Commit-Position: refs/heads/master@{#515203}

We want UKM scroll latency reporting to work for --enable-viz (i.e. for the
case when display compositor is in the GPU process). This means that we need to
report these metrics in LatencyTracker::ReportUkmScrollLatency, not in
RWHLatencyTrcker::ReportUkmScrollLatency.

This CL adds ukm::SourceId to LatencyInfo. The source id is set on the
LatencyInfo object in RWHLatencyTracker::OnInputEvent. This way LatencyTracker
(which does not have access to RWHDelegate) can use the ukm::SourceId from
LatencyInfo when reporting UKM scroll latency.

TBR=sadrul@chromium.org

Bug: 717629
Change-Id: Ib449bcaaf6153b67d2ffdf2f28bf4749f78c80fa
Reviewed-on: https://chromium-review.googlesource.com/744702
Commit-Queue: Mikhail Fomitchev <mfomitchev@chromium.org>
Reviewed-by: Steven Holte <holte@chromium.org>
Reviewed-by: Khushal <khushalsagar@chromium.org>
Reviewed-by: Timothy Dresser <tdresser@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515202}

Service and Manager classes used to depend on each other, making
their initialization timing complicated.

This CL moves more logic into Manager, such that the dependency cycle
is broken, Service no longer depends on Manager. As a result of this
simplification, there is now no need to keep track of both the aggregate
list (previously Service#extensions_) as well as the two sublists
(Manager#apps and Manager#extensions). Manger now only keeps track of
the sublists, and some unnecessary linear searching has been eliminated.

Bug: None
Cq-Include-Trybots: master.tryserver.chromium.linux:closure_compilation
Change-Id: I3bdc0e9b630cc1adf98034ee4fda862b17dbfbfa
Reviewed-on: https://chromium-review.googlesource.com/745903
Commit-Queue: Demetrios Papadopoulos <dpapad@chromium.org>
Reviewed-by: Devlin <rdevlin.cronin@chromium.org>
Reviewed-by: Scott Chen <scottchen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515201}

Bug: 720786
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: I80e08d02f49e452596a21e4a91f0d0d7f3810236
Reviewed-on: https://chromium-review.googlesource.com/758484
Commit-Queue: Danyao Wang <danyao@chromium.org>
Reviewed-by: Eugene But <eugenebut@chromium.org>
Reviewed-by: Sylvain Defresne <sdefresne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515200}

|description_class| array can be empty. Add conditional to handle if so.

Bug: 783194
Change-Id: I3f023e7b279f2eceaa6ea4432f68a2c09de19171
Reviewed-on: https://chromium-review.googlesource.com/760660Reviewed-by: Justin Donnelly <jdonnelly@chromium.org>
Commit-Queue: Kevin Bailey <krb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515199}

https://chromium.googlesource.com/angle/angle.git/+log/d80f2944cf13..70b715c9b8bc

$ git log d80f2944c..70b715c9b --date=short --no-merges --format='%ad %ae %s'
2017-11-07 yunchao.he ES31: Add test for DispatchCompute with rendering program
2017-11-03 jiawei.shao Skip hasMappedBuffer check in draw validations in WebGL contexts
2017-10-26 jiawei.shao ES31: Support EXT_geometry_shader in GLSL compiler
2017-10-31 jgilbert KHR_create_context with GLES forbids _KHR reset notification enum.
2017-11-07 jmadill Move LazyResource::resolveImpl to the cpp.
2017-11-08 jmadill Link interface blocks in ProgramImpl::link.
2017-10-31 jgilbert attachmentsHaveSameDimensions should not compare depths.
2017-10-30 cwallez Split pixelBuffer from pack/unpack state
2017-11-06 cwallez Add back dirty bits for pack / unpack buffer binding
2017-11-06 jmadill Rename UniformLinker.cpp/h to ProgramLinkedResources.
2017-11-07 oetuaho Size const unsized arrays without an initializer
2017-10-27 oetuaho Only store innermost array offset in VariableLocation
2017-10-24 geofflang Disallow null pixel data for TexSubImage that have non-zero size.
2017-11-07 jmadill Pass InterfaceBlockLinker to ProgramImpl::link.

Created with:
  roll-dep src/third_party/angle
BUG=781729


The AutoRoll server is located here: https://angle-chromium-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.


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
TBR=geofflang@chromium.org

Change-Id: I895af7e8c73982e88413866985c1ca9649d585ae
Reviewed-on: https://chromium-review.googlesource.com/760656Reviewed-by: angle-deps-roller . <angle-deps-roller@chromium.org>
Commit-Queue: angle-deps-roller . <angle-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515198}

The fetchForms... uses an empty form name to fetch all the forms.
This conflicts with the empty form name for filling formless checkouts.

Add a specific parameter and fix the javascript method to fill these fields.


Bug: 780050
Change-Id: Ia9fc514db6118a79af40dcd12b2312387f7ae726
Reviewed-on: https://chromium-review.googlesource.com/747101
Commit-Queue: Olivier Robin <olivierrobin@chromium.org>
Reviewed-by: Moe Ahmadi <mahmadi@chromium.org>
Reviewed-by: Peter Lee <pkl@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515197}

Now that crbug.com/755119 has been fixed, remove TODO and add a
NOTREACHED for BACKOFF behaviour which we don't currently support.

In reality the NOTREACHED is very unlikely to trigger as you'd need to
launch 15+ background fetches simultaneously, which is very unlikely for
a flagged feature.

Bug: 755119
Change-Id: Ic70d056486e695e7953a8c58515eca31c1d3f935
Reviewed-on: https://chromium-review.googlesource.com/758781Reviewed-by: Peter Beverloo <peter@chromium.org>
Commit-Queue: Peter Beverloo <peter@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515196}

SetCursor has no effect today unless a commit takes place
afterwards. This is fixed by making it possible to commit
contents before a surface has a delegate and there's a
surface tree host (STH). That allows us to add a surface to
a STH and then submit a valid frame that can be captured.

Bug: 
Test: exo_unittests --gtest_filter=PointerTest.SetCursor
Change-Id: I4e6688b2a83b971cd935a70f91642e075b079cb4
Reviewed-on: https://chromium-review.googlesource.com/760541Reviewed-by: Peng Huang <penghuang@chromium.org>
Commit-Queue: David Reveman <reveman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515195}