GitLab

Bug: none
Change-Id: I0a69f59192e400fba4ecbf563b4de64b6247f249
Reviewed-on: https://chromium-review.googlesource.com/570841Reviewed-by: Kurt Horimoto <kkhorimoto@chromium.org>
Commit-Queue: Kurt Horimoto <kkhorimoto@chromium.org>
Commit-Queue: Ed Chin <edchin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486881}

This patch creates GRC TabSignalObserver for TabManager to get tab scoped
signals from GRC, implements TabDoneLoading signal and eventually sends it to
TabManager::WebContentsData, and changes network idle signal to be persisted.

TBR=japhet@chromium.org
BUG=730098

Change-Id: I5ba9be4365cc79b0d1e6f19e44ec93eeaf0408ee
Reviewed-on: https://chromium-review.googlesource.com/563832
Commit-Queue: lpy <lpy@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Zhen Wang <zhenw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486880}

If shared memory returns bad data, then UserScriptSet might end up
allocating too much data.  This might distract crash reports to think
there was an OOM error. Set an arbitrary limit to number of scripts:
roughly 1000 extensions * 100 scripts = 100000u.

Also put |pickle_size| in crash stack to aid debugging crash reports.

Bug: 723381
Change-Id: Idf96137f5f7adec87adbacd5a276253dc5f24842
Reviewed-on: https://chromium-review.googlesource.com/570987
Commit-Queue: Istiaque Ahmed <lazyboy@chromium.org>
Reviewed-by: Devlin <rdevlin.cronin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486879}

This is a reland of 9d704fa8
Original change's description:
> [Zucchini] Introduce buffer_source and buffer_sink.
> 
> Cursors with convenient read and write methods to ease manipulation of
> raw data.
> 
> Bug: 729154
> Change-Id: I8a7303964b02b9df7ef033de8756b09716a91eae
> Reviewed-on: https://chromium-review.googlesource.com/560657
> Commit-Queue: Etienne Pierre-Doray <etiennep@chromium.org>
> Reviewed-by: Samuel Huang <huangs@chromium.org>
> Reviewed-by: Erik Chen <erikchen@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#486238}

Bug: 729154
Change-Id: I97552e41b452081bd8421d2a5be61a2f3b59e6b2
Reviewed-on: https://chromium-review.googlesource.com/571177Reviewed-by: Erik Chen <erikchen@chromium.org>
Reviewed-by: Samuel Huang <huangs@chromium.org>
Commit-Queue: Etienne Pierre-Doray <etiennep@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486878}

In https://codereview.chromium.org/2895323002, I slightly changed the 
way how verdict is cached, which makes previous cached entries no
longer valid. Therefore, CleanUpExpiredVerdicts should also remove
any invalid verdict entries together with the expired ones.

TBR=vakh@chromium.org

Bug: 742713
Change-Id: Id9e162d485e436e4992cb83c5a4378eea3df9771
Reviewed-on: https://chromium-review.googlesource.com/572086
Commit-Queue: Jialiu Lin <jialiul@chromium.org>
Reviewed-by: Luke Z <lpz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486877}

This doc was never added; if we want to add it in the future we can put
the link back at that time.

Change-Id: I2c23121f7844e6bae5d653259ef9578d70e98674
Reviewed-on: https://chromium-review.googlesource.com/572181Reviewed-by: Ben Henry <benhenry@chromium.org>
Commit-Queue: Annie Sullivan <sullivan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486876}

CompositorFrames and SharedBitmaps' memory handles come from two different
mojo pipes and currently there is no guarantee that for all SharedBitmaps 
referenced by a CompositorFrame are available when it arrives. In this CL
ClientSharedBitmapManager assigns a sequence number to each SharedBitmap
it allocates. These sequence numbers are sent to the browser process along
with the CompositorFrame that references those SharedBitmaps. When a
CompositorFrame arrives, RenderWidgetHostImpl asks from
SharedBitmapAllocationNotifierImpl the latest sequence number from that
renderer known to the browser and if the CompositorFrame references higher
sequence numbers it won't be processed until
SharedBitmapAllocationNotifierImpl notifies RenderWidgetHostImpl through the
SharedBitmapAllocationObserver interface that the necessary SharedBitmaps are
ready.
A few notes:
- This CL is meant to be a quick fix for crbug.com/734058. Long term we
would like this logic to be in viz and not content, but that's
substantially more work.
- This CL only fixes the issue for RenderWidgets, which got broken after
crrev.com/463845. Everything else including OffscreenCanvas has always
been broken and remains broken.
- SharedBitmapAllocationNotifier will no longer be associated with
RenderMessageFilter. It used to be associated to preserve ordering with
respect to CompositorFrames, but after crrev.com/463845 CompositorFrames
no longer share a pipe with RenderMessageFilter so there is no point.
As a result SharedBitmapAllocationNotifierImpl now has its own separate
pipe and it lives on the UI thread which is nicer because that's where
the observers (RenderWidgetHostImpls) live and we can avoid doing PostTasks.
The separate pipe also prepares us for an out of process display compositor
because RenderMessageFilter will always be a browser<->renderer pipe and 
being associated with it doesn't make much sense.
- With SharedBitmapAllocationNotifierImpl living on UI thread, I thought
we should be able to get rid of the mutex in ServerSharedBitmapManager,
but it seems like OnMemoryDump is still being called from another thread.
We should investigate whether something can be done about OnMemoryDump
so we can get rid of the lock because it feels very unnecessary now.

BUG=734058

Change-Id: I3ba32f1e1cb4fce1589445b2f17e9f9855aa0c40
Reviewed-on: https://chromium-review.googlesource.com/565645Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Antoine Labour <piman@chromium.org>
Commit-Queue: Saman Sami <samans@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486875}

This CL adds TRACE_EVENTs to several heap profiler-related functions
to surface their timings.

BUG=739378

Review-Url: https://codereview.chromium.org/2980083002
Cr-Commit-Position: refs/heads/master@{#486874}

This is the first step of removing another use of Native App Whitelist
Manager. At this time, the Native App is only used for checking whether
certain Google native iOS apps are installed. This CL creates a new
null-implementation API in ChromeBrowserProvider for doing just that.

The next several steps will be
- implementing this API downstream
- switch upstream to use this new API
- remove GetNativeAppWhitelistManager() API

Bug: 742617
Change-Id: I6553a85743f1328ff8d6f5fca2a466af53d9a393
Reviewed-on: https://chromium-review.googlesource.com/571313
Commit-Queue: Peter Lee <pkl@chromium.org>
Reviewed-by: Rohit Rao <rohitrao@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486873}

CastAudioManager should only be used on platforms that use ALSA.

BUG=internal 35197382
TEST=internal CQ

Change-Id: I259dc5a5798c5a7144641a09ae6aede18125c0a6
Reviewed-on: https://chromium-review.googlesource.com/565625
Commit-Queue: Jenny Wong <jyw@chromium.org>
Reviewed-by: Alok Priyadarshi <alokp@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486872}

Passing ownership by raw pointer is discouraged, and std::unique_ptr is now
permissible in the Blink public API.

Change-Id: Icd65d137468feb27a1fe9c217347968ab9d5451d
Reviewed-on: https://chromium-review.googlesource.com/568684
Commit-Queue: Jeremy Roman <jbroman@chromium.org>
Reviewed-by: John Rummell <jrummell@chromium.org>
Reviewed-by: Xiaohan Wang <xhwang@chromium.org>
Reviewed-by: Kentaro Hara <haraken@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486871}

For synchronous requests, Resource::WillFollowRedirect() is not
called and thus redirect chain is not set in Resources and
redirects are not reported to Resource.
Therefore, this breaks assumptions around redirects and
revalidations:
- The URLs of the original response and revalidation response
  can be different without WillFollowRedirect() calls,
  and this causes CHECK failure in Resource::RevalidationSucceeded().
- We forbid successful revalidation if either the old or new request
  is redirected, but this restriction relies on WillFollowRedirect()
  and thus is bypassed in sync cases.

This CL forbids reusing and revalidating synchronously-loaded
Resources for sync requests, to avoid these issues.

Bug: 618967
Change-Id: Ic8c668b751a69f594f3d08b7d63b9832b6bcc56e
Reviewed-on: https://chromium-review.googlesource.com/567679
Commit-Queue: Hiroshige Hayashizaki <hiroshige@chromium.org>
Reviewed-by: Takeshi Yoshino <tyoshino@chromium.org>
Reviewed-by: Yutaka Hirano <yhirano@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486870}

I was not able to reproduce the crash reported but here is a speculative fix.

This is a scenario where a partial transaction needs to create a new entry and doom
the existing one since it cannot validate itself against the new entry, due to,
say, there are no strong validators.

The crash requires a specific race with another transaction for the same entry
with LOAD_BYPASS_CACHE set due to which the second transaction straight away goes
and dooms the entry (DoDoomEntry*) while the first one had opened the entry but
not yet reached DoomPartialEntry.

This race is much more likely after the parallel validation CL
(https://codereview.chromium.org/2721933002) because add to entry is always an
async operation now.

This CL adds a check that the entry is not already doomed before calling
cache_->DoomEntry in DoomPartialEntry and in one other calling location in
DoSuccessfulSendRequest as it may happen there as well.

BUG=736993

Review-Url: https://codereview.chromium.org/2970133002
Cr-Commit-Position: refs/heads/master@{#486869}

'importScripts()' is not a parser-inserted script-loading mechanism; it
ought to continue loading script in the presence of 'strict-dynamic'.

Bug: 742354
Change-Id: Ice327c6f69183e1b27912c808646d16d0030b934
Reviewed-on: https://chromium-review.googlesource.com/571723Reviewed-by: Jochen Eisinger <jochen@chromium.org>
Commit-Queue: Mike West <mkwst@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486868}

This is the safest way to avoid potential dead lock issues with
dispatch_once and ios_web_view::InitializeGlobalState().

BUG=

Review-Url: https://codereview.chromium.org/2980983002
Cr-Commit-Position: refs/heads/master@{#486867}

Many fonts tests use local helper classes TestRun, ExpectedRun and
helper macros DECLARE_RUNSVECTOR and CHECK_RUNS. With jumbo builds
those names start colliding so this patch gives the classes and
macros a prefix related to the actual test.

Change-Id: I494cfc182070ae37a87b68f08303fc8cc3ef9cfe
Reviewed-on: https://chromium-review.googlesource.com/570424
Commit-Queue: Emil A Eklund <eae@chromium.org>
Reviewed-by: Emil A Eklund <eae@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486866}

This CL adds a SensorReadingSharedBufferReader class at
//services/device/public/cpp/generic_sensor so that it can be
used by both //content/renderer/device_sensors and Blink code.

Bug: 727788
Change-Id: I9d03b64e5cf191738e87bbe29fd7611654f3fd5b
Reviewed-on: https://chromium-review.googlesource.com/541685Reviewed-by: Tim Volodine <timvolodine@chromium.org>
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Reviewed-by: Alexander Shalamov <alexander.shalamov@intel.com>
Commit-Queue: Jun Cai <juncai@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486865}

When re-aligning a line after removing its ellipsis we need to take
into account any floats on the line as well as its alignment. So
reposition the line when the floats are still present, rather than
later on when they've already been deleted.

Bug: 275883
Change-Id: I48913da2be5d8e60d10376a911e31bc7a87655ce
Reviewed-on: https://chromium-review.googlesource.com/570432Reviewed-by: Emil A Eklund <eae@chromium.org>
Commit-Queue: Emil A Eklund <eae@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486864}

https://chromium.googlesource.com/external/github.com/catapult-project/catapult.git/+log/ebf00ecf2e83..1b26db0a1723

$ git log ebf00ecf2..1b26db0a1 --date=short --no-merges --format='%ad %ae %s'
2017-07-14 etienneb Remove useless debug variable

Created with:
  roll-dep src/third_party/catapult
BUG=739783


Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls


CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel
TBR=sullivan@chromium.org

Change-Id: I032b7a73221c674df6d5a93bed297a672d6048c8
Reviewed-on: https://chromium-review.googlesource.com/572120
Reviewed-by: <catapult-deps-roller@chromium.org>
Commit-Queue: <catapult-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486863}

https://pdfium.googlesource.com/pdfium.git/+log/21dd189fe50d..3a4ebcc7c490

$ git log 21dd189fe..3a4ebcc7c --date=short --no-merges --format='%ad %ae %s'
2017-07-14 hnakashima Create ParseDataType unit tests based on specs.

Created with:
  roll-dep src/third_party/pdfium


Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls


TBR=dsinclair@chromium.org

Change-Id: Iefe43530b57fbad468c200bba61eaaa81898ab2b
Reviewed-on: https://chromium-review.googlesource.com/572006
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486862}

The file was removed in https://chromium-review.googlesource.com/c/569444/

But tryserver.chromium.perf is still using its custom code path, and not
the standard trybots.py that other tryservers use, so it still requires
tryserver.chromium.perf.json.

I'm in the process of converting it to trybots.py for Pinpoint, and we can
remove it later.

Bug: 742894
Change-Id: I062de1e561a36b186c20abb071bd5562578a31d1
Reviewed-on: https://chromium-review.googlesource.com/571858Reviewed-by: Annie Sullivan <sullivan@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Annie Sullivan <sullivan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486861}

This CL updates various functions in the ExtensionScopedPrefs and ExtensionPrefs
API to use StringPiece while accepting preference keys. Currently most of the
clients to these functions pass the preference keys as a const char pointer. The
functions accept the keys as std::string. During most function calls, these are
further converted to a StringPiece leading to a "const char* -> std::string ->
StringPiece" conversion chain. This CL avoids these redundant conversions by
accepting preference keys as StringPiece.

BUG=None.

Change-Id: Ib0727d45ac3701d08ab0118a84df97ac1123058f
Reviewed-on: https://chromium-review.googlesource.com/570892
Commit-Queue: Karan Bhatia <karandeepb@chromium.org>
Reviewed-by: Devlin <rdevlin.cronin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486860}

Convert fixed array of pointers to a flat map. Previously, we
were requiring a direct-mapped array of 128 pointers per byte of
input string.

Add a small unit test for SuffixTree.

Bug: 740873
Change-Id: I03836f5f9f893861bec3076fafaffd7c4e3e04eb
Reviewed-on: https://chromium-review.googlesource.com/571193Reviewed-by: Kentaro Hara <haraken@chromium.org>
Reviewed-by: Mike West <mkwst@chromium.org>
Commit-Queue: Tom Sepez <tsepez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486859}

Doing an implicit template instantiation before doing an explicit
template instantiation in the same translation unit is a C++
violation. That happens if you jumbo compile blink platform but is
easy to avoid with a forward declaration such as the one this patch
adds.

Change-Id: I7b22e28b2bf8f2583598e97db69d68fa1613550e
Reviewed-on: https://chromium-review.googlesource.com/570444
Commit-Queue: Emil A Eklund <eae@chromium.org>
Reviewed-by: Emil A Eklund <eae@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486858}

There is a use after free in recent commit. Fix it.

Bug: 742374
Change-Id: Iab561ccdc7f863a77a1c70460a87b6714894e954
Reviewed-on: https://chromium-review.googlesource.com/571323Reviewed-by: Fady Samuel <fsamuel@chromium.org>
Commit-Queue: kylechar <kylechar@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486857}

There were two implementation of SkiaScalarToHarfBuzzPosition and
in jumbo builds they collided. This patch merges them.

Bug: 
Change-Id: Ibea7aeffbfdf77eb3fd088949f912eb41e56c462
Reviewed-on: https://chromium-review.googlesource.com/570428
Commit-Queue: Emil A Eklund <eae@chromium.org>
Reviewed-by: Emil A Eklund <eae@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486856}

Bug: 736249
Change-Id: I2de2dc831228167bd24473940077ee36844993cf
Reviewed-on: https://chromium-review.googlesource.com/571035Reviewed-by: Shivani Sharma <shivanisha@chromium.org>
Commit-Queue: Charlie Harrison <csharrison@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486855}

apply to loads in the main frame.

The distinction between "IsLoading" and "IsLoadingToDifferentDocument"
appears to only be relevant in UI contexts (e.g. TabRendererData, which
controls whether to show the throbber in the tabstrip on desktop
Chrome). I think the distinction only makes sense in the context of the
main frame and a load in a subframe should always be lumped in with
non-top-level document loads. For example, on Android it's sometimes
used to conditionally call ToolbarLayout.onNavigatedToDifferentPage(),
which from other callsites definitely appears to only care about main
frame navigations.

Making this change means the throbber won't flicker when content is
loaded in an iframe right after the main frame finishes loading, which
happens a lot (see bug).

With this change, we can probably remove the timer added in
89603210 which attempts to smooth out favicon/throbber
flickering.

Bug: 734104
Change-Id: Ie14b1c54956479021fd9aa6c4c8aa690be52f193
Reviewed-on: https://chromium-review.googlesource.com/569060Reviewed-by: Pavel Feldman <pfeldman@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Evan Stade <estade@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486854}

Use BindOnce in RecognitionAllowedCallback for MediaRequestPermissionCallback

Bug: 714018
Change-Id: I01e4eed7b75cd8828bd0dba80ca1fc07015a517f
Reviewed-on: https://chromium-review.googlesource.com/570510
Commit-Queue: Mark Pilgrim <pilgrim@chromium.org>
Reviewed-by: John Abd-El-Malek <jam@chromium.org>
Reviewed-by: Tommi <tommi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486853}

Implements a HitTestAggregator component to facilitate cross process
hit testing with minimal process hops.

The HitTestAggregator maintains a list of HitTestData objects that will
be created and sent with each CompositorFrame.  These are aggregated
into a single DisplayHitTest data structure that can be made available
in shared memory to facilitate hit testing with minimal process hops
for any Display.

HitTestAggregator listens as surfaces are added to the current
CompositorFrame to ensure that the list only includes information that
matches what has been rendered in the current frame.

Aggregation occurs on the BeginFrame notification.

This change includes the implementaiton of the HitTestAggregator and
unit test cases to verify the intended behaviour.

Next steps will include completion of the creation and transport of
HitTestData objects, implementation of shared memory and using this
component to target incoming events.

BUG=732399

Review-Url: https://codereview.chromium.org/2938953002
Cr-Commit-Position: refs/heads/master@{#486852}

Bug: 722935
Change-Id: I6f6a26ac5405308efb3793ba0dbc0dd718e9a309
Reviewed-on: https://chromium-review.googlesource.com/563777Reviewed-by: Antoine Labour <piman@chromium.org>
Reviewed-by: Ken Rockot <rockot@chromium.org>
Commit-Queue: Fady Samuel <fsamuel@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486851}

https://skia.googlesource.com/skia.git/+log/fc4ee229a653..8abb9f453798

$ git log fc4ee229a..8abb9f453 --date=short --no-merges --format='%ad %ae %s'
2017-07-14 bsalomon Remove GrTestMeshDrawOp

Created with:
  roll-dep src/third_party/skia


Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls


CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_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;master.tryserver.chromium.android:android_optional_gpu_tests_rel
TBR=ethannicholas@chromium.org

Change-Id: I079d39f4c96134ba622dd721bd5add88132ffd15
Reviewed-on: https://chromium-review.googlesource.com/571849Reviewed-by: Skia Deps Roller <skia-deps-roller@chromium.org>
Commit-Queue: Skia Deps Roller <skia-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486850}

Running tools\win\ShowGlobals on Canary PDBs shows that the variable
mojo::edk::ports::kInvalidSequenceNum shows up 21 times in both
chrome.dll and chrome_child.dll. This appears to be due to a VC++
quirk in how it handles integer constants that are initialized by
some expression types. Tagging the definition as constexpr should
ensure that this doesn't happen.

Bug: 
Change-Id: I5d72acc494f14880cb7cd5a0107e7d6eefbf7c7f
Reviewed-on: https://chromium-review.googlesource.com/571241Reviewed-by: Ken Rockot <rockot@chromium.org>
Commit-Queue: Bruce Dawson <brucedawson@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486849}

This will make it much easier for //chrome code to test the subresource
filter logic.

Bug: None
Change-Id: Id34d9128059ef87463a9644b903d4d64061e2180
Reviewed-on: https://chromium-review.googlesource.com/571949Reviewed-by: Josh Karlin <jkarlin@chromium.org>
Commit-Queue: Charlie Harrison <csharrison@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486848}

Change-Id: I9161dc11fab6b709ae21170c9e62e5de8676de28

TBR=dpranke@chromium.org

Change-Id: I9161dc11fab6b709ae21170c9e62e5de8676de28
Reviewed-on: https://chromium-review.googlesource.com/572030
Commit-Queue: Yipeng Wang <yipengw@chromium.org>
Reviewed-by: Richard Coles <torne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486847}

This is being renamed in https://codereview.chromium.org/2965383002/
This CL should land after that CL rolls into chromium.

Change-Id: Id36c421709d40de22771a241d431aa9c19be355b
Reviewed-on: https://chromium-review.googlesource.com/570678Reviewed-by: Ned Nguyen <nednguyen@google.com>
Reviewed-by: Derek Cheng <imcheng@chromium.org>
Commit-Queue: Annie Sullivan <sullivan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486846}

WindowTreeHostManager may reuse a WindowTreeHost for a different
display. This happens if all displays are removed, and then a new
display is added. Ash now handles this by way of calling through to
SetDisplayRoot() and SetDisplayRoot() on mus has been updated to allow
this to happen.

In addition this changes ash's DisplaySynchronizer to only notify mus
of the current display layout after all changes have been
processed. This is necessary as DisplayManager notifies observers while
internally the list of displays contains both added and removed
entries. This means when FocusSynchronizer could notify of state
display state, which mus didn't like and caused ash to crash (hit a
CHECK).

BUG=740589
TEST=covered by tests

Change-Id: Ie0a2a05a1b592645669355f39997f9e3a5e77c3f
Reviewed-on: https://chromium-review.googlesource.com/571399Reviewed-by: kylechar <kylechar@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486845}

Fix an IWYU bug and isnan() call site.

TBR=jamescook@chromium.org
R=jamescook@chromium.org

Change-Id: I7b586f9d093150c9419fdcd5995803d4fa2be7b3
Reviewed-on: https://chromium-review.googlesource.com/572035Reviewed-by: Justin Schuh <jschuh@chromium.org>
Commit-Queue: Justin Schuh <jschuh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486844}

|dalvik_mb| no longer follows the expected heuristic pattern, causing us
to over-estimate memory on low-end devices.

This is a temporary fix that removes the Dalvik VM computation when on
a low-end device. Eventually we should clean up this code for all
Android devices.

Bug: 742534
Change-Id: I9325530d6dc8cc9369d6beb0dd072424e5b60be9
Reviewed-on: https://chromium-review.googlesource.com/571179
Commit-Queue: Eric Karl <ericrk@chromium.org>
Reviewed-by: Alexandre Elias <aelias@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486843}

Developers sometimes do not notice that they are using a partially
supported platform when trying Web Bluetooth. Especially when using
the generic --enable-experimental-web-platform-features.

E.g. https://plus.google.com/117707543167867525197/posts/4nKECjvLD9D

This patch adds a warning to the developer console.

The definition of supported platforms comes from 
https://cs.chromium.org/chromium/src/content/child/runtime_features.cc?type=cs&q=WebRuntimeFeatures::EnableWebBluetooth%5C(true%5C)&l=69-71

Change-Id: I82f7e6fd800607372520d8474e339e5a7659259c
Reviewed-on: https://chromium-review.googlesource.com/557372
Commit-Queue: Vincent Scheib <scheib@chromium.org>
Reviewed-by: Conley Owens <cco3@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486842}