GitLab

This patch changes |Adjust{Forward,Backward}PositionForUserSelectAll()| in
"SelectionModifier.cpp" to take |Const VisbilePosition&| as follow-up of
patch[1], since the parameter is "in"-parameter, we don't use it as "out"-
parameter for improving code health.

[1] http://crrev.com/c/562915 Refactor AdjustPositionForUserSelectAll()
in SelectionModifier.cpp

Change-Id: I618d48c6a05c015ece70ac1ffa103fbce2e64040
Reviewed-on: https://chromium-review.googlesource.com/564870Reviewed-by: Yoichi Osato <yoichio@chromium.org>
Commit-Queue: Yoshifumi Inoue <yosin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485205}

Unescaped custom validation messages could cause DCHECK failures.

Bug: 739091
Change-Id: Ia7993221cad5d7ff00767cfe676e88da2fba2114
Reviewed-on: https://chromium-review.googlesource.com/564194Reviewed-by: Keishi Hattori <keishi@chromium.org>
Commit-Queue: Kent Tamura <tkent@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485204}

The password_manager component now has a GetHumanReadableOrigin in its core
layer, so iOS settings should no longer use the outdated local replacement. As
work progresses on https://crbug.com/628988, this will also allow the iOS
settings to use more user-friendly names for credentials coming from Android
apps.

The difference is currently only subtle: the new function drops the path from
the web origin (URL). This seems reasonable, given that it highlights the
hostname and the full URL is available in the password details section (as soon
as viewing passwords is launched -- tracked at https://crbug.com/739404). Once
https://crbug.com/628988 is fixed, there will be a much bigger difference for
the credentials coming from Android apps, where the cryptic URL will be
replaced by a human readable app name.

Screenshots comparing the before and after state for this CL are at
https://crbug.com/669538#c2.

Bug: 669538
Change-Id: I4ae80be4326c48d44f684b18b73682da99abb8b7
Reviewed-on: https://chromium-review.googlesource.com/563664
Commit-Queue: Sylvain Defresne <sdefresne@chromium.org>
Reviewed-by: Sylvain Defresne <sdefresne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485203}

This patch only touches comments. No code change.

Change-Id: Ibc823deb7cfe35dfed400566767e1791ac78e48b
Reviewed-on: https://chromium-review.googlesource.com/564876Reviewed-by: Taiju Tsuiki <tzik@chromium.org>
Commit-Queue: Yuta Kitamura <yutak@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485202}

BUG=708487
TBR=primiano@chromium.org

Review-Url: https://codereview.chromium.org/2975743002
Cr-Commit-Position: refs/heads/master@{#485201}

This patch converts "empty-cell-right-click.html" to gTest to narrow scope of
verification for improving code health.

# Background
The bug[1] is happend only on MacOS due by MacOs specific behavior
|ShouldSelectOnContextualMenuClick()|. This means "empty-cell-right-click.html"
is only meaningful on MacOs. Also, The actual changes of fixing the bug[1] is in
|ViisibleSelection|. Hence, using unit test for |VisibleSeleciton| is more
suitable for this case.

[1] https://bugs.webkit.org/show_bug.cgi?id=32622 Right clicking in a table cell
creates a selection that extends into the next cell

Change-Id: I806907f6d13008f3b131514e3012fd9fb8bc0561
Reviewed-on: https://chromium-review.googlesource.com/564865Reviewed-by: Yoichi Osato <yoichio@chromium.org>
Commit-Queue: Yoshifumi Inoue <yosin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485200}

When there is an animation on the NTP screen with ContentSuggestions,
the appearing and disappearing icons are shown with an alpha of 1
during the whole animation.
This CL fixes this by reloading the data during the rotation.

Before: https://drive.google.com/open?id=0B21N-0pPGAuJMHVhWDRvdU55ODg
After: https://drive.google.com/open?id=0B21N-0pPGAuJYm13cW5IVDRiUjQ

BUG=700375

Change-Id: Ie2c928ff1d012439cf329948713f2ff46887b7d2
Reviewed-on: https://chromium-review.googlesource.com/561776
Commit-Queue: Gauthier Ambard <gambard@chromium.org>
Reviewed-by: Louis Romero <lpromero@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485199}

This applies when converting CSSUnitValues to other CSSUnitValues.

https://drafts.css-houdini.org/css-typed-om/#dom-cssnumericvalue-to

BUG=545318

Review-Url: https://codereview.chromium.org/2958233002
Cr-Commit-Position: refs/heads/master@{#485198}

Bug: 738794
Change-Id: I6926dd2db27f0bd824e4b4566278af27c9feb5cb
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/563304Reviewed-by: Hayato Ito <hayato@chromium.org>
Commit-Queue: Hayato Ito <hayato@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485197}

This CL introduces WebDocumentSubresourceFilter::Builder class.

SubresourceFilterAgent::WillCreateWorkerFetchContext() creates the builder
object and set to the WebWorkerFetchContext on the main thread.

WebWorkerFetchContext::TakeSubresourceFilter() is called by the constructor of
WorkerFetchContext on the worker thread to creates a WebDocumentSubresourceFilter
using the builder object.

Note: This CL adds SubresourceFilter supports only for Dedicated Workers with
off-main-thread-fetch optimization. Currently SubresourceFilter is not supported
in Shared Workers and Service Workers regardless of the off-main-thread-fetch
optimization. See crbug.com/739597#c2

Bug: 443374,739597
Change-Id: I5719f3b9de33a3e552a3271304e0d530e9c48cc8
Reviewed-on: https://chromium-review.googlesource.com/561263Reviewed-by: Kinuko Yasuda <kinuko@chromium.org>
Reviewed-by: Charlie Harrison <csharrison@chromium.org>
Commit-Queue: Tsuyoshi Horo <horo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485196}

AudioRendererSinkCacheImpl::DeleteLaterIfUnused didn't retain a reference
to |sink_ptr| when it posts |sink_ptr| to |task_runner_| for deletion.
If multiple cache deletions are scheduled, and a new allocation of
AudioRendererSink reuses that address between the first deletion and the
second deletion, the second deletion may delete unexpected cached item.

This CL adds RetainedRef on the async cache deletion to keep
AudioRendererSink alive while it's in the task queue, so that the
task doesn't hold the dangling pointer.

Change-Id: I53a04a2a7fbc5f2d5c08013e98e3aff0d5ba613e
Reviewed-on: https://chromium-review.googlesource.com/562837
Commit-Queue: Taiju Tsuiki <tzik@chromium.org>
Reviewed-by: Olga Sharonova <olka@chromium.org>
Reviewed-by: Peter Kasting <pkasting@chromium.org>
Reviewed-by: Xiaohan Wang <xhwang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485195}

EarlGrey matchers should not wait, this has been replaced by
[ChromeEarlGrey waitForWebViewNotContainingText:]

Bug: 739073
Change-Id: I742a65025f2814bb29bb32f753623fd88ed179a2
Reviewed-on: https://chromium-review.googlesource.com/558988
Commit-Queue: Mike Baxley <baxley@chromium.org>
Reviewed-by: Eugene But <eugenebut@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485194}

(1) treat content_settings change as user gesture
   This is because we have seen website triggers redirection based on
  content setting change to obfuscate landing page.
(2) propagate EVENT_URL entry for html 5 download
  EVENT_URL will always show up in both case of attribution, such 
 that it is easier for backend to process data.

Bug: 739453, 735510
Change-Id: Ibff64bc474dcfc1883a7159d715d97e1ff173f09
Reviewed-on: https://chromium-review.googlesource.com/562318
Commit-Queue: Jialiu Lin <jialiul@chromium.org>
Reviewed-by: John Abd-El-Malek <jam@chromium.org>
Reviewed-by: Chris Palmer <palmer@chromium.org>
Reviewed-by: Varun Khaneja <vakh@chromium.org>
Reviewed-by: Luke Z <lpz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485193}

This reverts commit 2faacc5b27449d42f79505b8003c6309b57e6667.

BUG=b:32308760
TEST=tot Chrome OS (9726.0.0) has all fixes for the test failures
 (CL:560921, CL:563150, and ag/2501460)

Change-Id: Id20c584d280cd19cc5c33bd391bd9108335a9e25
Reviewed-on: https://chromium-review.googlesource.com/563868Reviewed-by: Hidehiko Abe <hidehiko@chromium.org>
Commit-Queue: Yusuke Sato <yusukes@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485192}

I saw this while I was exploring a code refactor. This is supposed to copy one
vector to another but really just accesses memory out-of-bounds if the passed-in
vector is not empty. Yikes!

It appears I introduced this bug in https://codereview.chromium.org/2568093003.

Bug: 
Change-Id: Ie529e5030e32a370419b47fcb03bf09ece810114
Reviewed-on: https://chromium-review.googlesource.com/564282Reviewed-by: Istiaque Ahmed <lazyboy@chromium.org>
Commit-Queue: Ryan Landay <rlanday@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485191}

This patch fixes an error introduced in patch crrev.com/2882603002,
where the CSSPropertyBackgroundColor property was pulled out into a
single property API when it should have been grouped with the
CSSPropertyColor property. This error left CSSPropertyColor parsing
logic to fall through to border color parsing logic.

This patch
- moved CSSPropertyColor to be grouped with CSSPropertyBorderColor in
  existing API implementation
- renamed CSSPropertyAPIBackgroundColor to CSSPropertyAPIColor as a
  group API name for CSSPropertyColor and CSSPropertyBorderColor
- because CSSPropertyAPIColor was already taken as an API name,
  renamed the pre-existing CSSPropertyAPIColor to
  CSSPropertyAPIColorNoQuirks (the difference between the 2 APIs
  parsing logic is only whether quirks mode is permitted)

Bug: 668012
Change-Id: I119a2cafbd51b93bd1c9323745aae728cc74a2ca
Reviewed-on: https://chromium-review.googlesource.com/562944Reviewed-by: meade_UTC10 <meade@chromium.org>
Reviewed-by: Eric Willigers <ericwilligers@chromium.org>
Reviewed-by: Suzy Howlett <suzyh@chromium.org>
Commit-Queue: Bugs Nash <bugsnash@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485190}

Revert of [cr-action-menu] Use clientWidth for rtl flipping. (patchset #2 id:60001 of https://codereview.chromium.org/2966163004/ )

Reason for revert:
Breaks on https://uberchromegw.corp.google.com/i/chromium.mac/builders/Mac10.11%20Tests/builds/15007

Original issue's description:
> [cr-action-menu] Use clientWidth for rtl flipping.
>
> This CL fixes a bug where the action menu would use the entire body's
> length as the viewport when flipping coordinates in RTL.
>
> It also moves the scroll position saving into showAtPosition so that
> both show methods account for scroll correctly.
>
> BUG=734984
> CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation
>
> Review-Url: https://codereview.chromium.org/2966163004
> Cr-Commit-Position: refs/heads/master@{#485181}
> Committed: https://chromium.googlesource.com/chromium/src/+/87be8f4c1eaef26b7c6d7ddb2c56d84ef504a232

TBR=dpapad@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=734984

Review-Url: https://codereview.chromium.org/2980463002
Cr-Commit-Position: refs/heads/master@{#485189}

This CL proposes a few editorial changes to some VEA-related
files, namely to rtc_video_encoder.h:
- Makes RecordInitEncodeUMA() a file-static function 
- Uses __func__ pervasively in DVLOG() statements to save some
 lines.
In gpu_video_encode_accelerator_host.cc:
- prefers early return
- micro DVLOG cleanup

** No new code, except the early returns **

Bug: 736517
Change-Id: I2c6cd351a0716810338fa3a2e87e2ef3ca708cbf
Reviewed-on: https://chromium-review.googlesource.com/564743Reviewed-by: Wu-Cheng Li <wuchengli@chromium.org>
Commit-Queue: Miguel Casas <mcasas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485188}

This patch changes VisibleSelectionTemplate::Validate() to use early-return
for improving readability.

Change-Id: If1395322fd0fea262af37c9840c0976439f78cef
Reviewed-on: https://chromium-review.googlesource.com/563220Reviewed-by: Yoichi Osato <yoichio@chromium.org>
Commit-Queue: Yoshifumi Inoue <yosin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485187}

This simplifies many conditions around the code.

BUG=561800

Review-Url: https://codereview.chromium.org/2972333002
Cr-Commit-Position: refs/heads/master@{#485186}

When the availability callback is added from the native controls, the
low end device check is by-passed. Add an extra check and also return
availability as true on such devices so that the button is shown.

BUG=731462
TEST=manual+existing tests

Change-Id: I34c6815a82c1a75ea7e174341703c19fee33af7a
Reviewed-on: https://chromium-review.googlesource.com/545275Reviewed-by: Kentaro Hara <haraken@chromium.org>
Reviewed-by: Mounir Lamouri <mlamouri@chromium.org>
Commit-Queue: Anton Vayvod <avayvod@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485185}

There is a case that CreateVisiblePosition returns {start, null} range
 in LayoutSelection::CalcSelection.
VisibleSelection and EphemeralRange constructors prohibit such case
 but it is by DCHECK so we need cover on Release build failure.

BUG=739996

Review-Url: https://codereview.chromium.org/2968363002
Cr-Commit-Position: refs/heads/master@{#485184}

If you minimise Gmail, it disappears from the task manager, so allow the
LAST_ACTIVITY process state to be displayed for ARC processes.

Additionally allow FOREGROUND_SERVICE since foreground services are user
visible.

BUG=738651

Change-Id: I403b9946fbbb29b115fe1877f63bb64e3fe3b2c7
Reviewed-on: https://chromium-review.googlesource.com/562906Reviewed-by: Nick Carter <nick@chromium.org>
Reviewed-by: Yusuke Sato <yusukes@chromium.org>
Commit-Queue: Eliot Courtney <edcourtney@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485183}

It looks the purpose of the SHMBuffer is to store the size of
the SharedMemory. However, the info is also available in
SharedMemory::mapped_size().

BUG=None
TEST=None (No behavior change)

Change-Id: I71d4ae4ea8c27597c260e10a3cda9edd3530f268
Reviewed-on: https://chromium-review.googlesource.com/562907Reviewed-by: Dale Curtis <dalecurtis@chromium.org>
Commit-Queue: Owen Lin <owenlin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485182}

This CL fixes a bug where the action menu would use the entire body's
length as the viewport when flipping coordinates in RTL.

It also moves the scroll position saving into showAtPosition so that
both show methods account for scroll correctly.

BUG=734984
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2966163004
Cr-Commit-Position: refs/heads/master@{#485181}

Bug: 
Change-Id: I7adc36e09d570d3034c2882817489bacdb84b493
Reviewed-on: https://chromium-review.googlesource.com/562911
Commit-Queue: David Vallet <dvallet@chromium.org>
Reviewed-by: Sami Kyöstilä <skyostil@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485180}

Bug: 667961
Change-Id: Ifed211edbc99290178daf60c39c7ff0e013213b7
Reviewed-on: https://chromium-review.googlesource.com/558308
Commit-Queue: Stuart Langley <slangley@chromium.org>
Reviewed-by: meade_UTC10 <meade@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485179}

TEST=None
BUG=b:32692654

Review-Url: https://codereview.chromium.org/2892863002
Cr-Commit-Position: refs/heads/master@{#485178}

TBR=sky@chromium.org

Change-Id: If31023bb2734c5e9fd3df8575f472dfcec826047
Reviewed-on: https://chromium-review.googlesource.com/563723Reviewed-by: chrome-release-bot@chromium.org <chrome-release-bot@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485177}

PermissionRequestManager's destructor currently clears its request
vectors, but as it is a WebContentsUserData these should already have
been cleared in WebContentsDestroyed. This patch replaces this with
assertions the vectors are empty.

Change-Id: I71d56e91356435698c9dbcb4fcc347b997d05e6f
Reviewed-on: https://chromium-review.googlesource.com/562845Reviewed-by: Raymes Khoury <raymes@chromium.org>
Commit-Queue: Timothy Loh <timloh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485176}

The new function ConsumeShorthandVia4LonghandsAPI is a modified version
of Consume4Values of CSSPropertyParser. They both parse shorthands 
consisting of 4 longhands (top, right, bottom, left). The difference
between this new version and the old/existing Consume4Values is that
the new version expects component longhands to have API impl already 
as it will forward parsing logic to component longhands API via 
ParseLonghandViaAPI.

This cl contains one change only, i.e. the impl of 
ConsumeShorthandVia4LonghandsAPI. This cl is also a use case of
ParseLonghandViaAPI. The next cl will have a shorthand API impl
that will use ConsumeShorthandVia4LonghandsAPI.

Bug: 668012
Change-Id: I9a638d9c654a1bb734b60dccadaf2c52ddd761f0
Reviewed-on: https://chromium-review.googlesource.com/560920
Commit-Queue: Jia Meng <jiameng@chromium.org>
Reviewed-by: Alan Cutter <alancutter@chromium.org>
Reviewed-by: Darren Shen <shend@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485175}

BUG=616950

Review-Url: https://codereview.chromium.org/2962673003
Cr-Commit-Position: refs/heads/master@{#485174}

BUG=554289

Review-Url: https://codereview.chromium.org/2972053003
Cr-Commit-Position: refs/heads/master@{#485173}

Currently, FrameSinkManager::UnregisterFrameSinkManagerClient removes
the entry associated with |frame_sink_id| from |frame_sink_source_map_|
if the |frame_sink_id| does not have any children. This would result in
a client not connected with a BeginFrameSource when it is unregistered
and re-registered with the same FrameSinkId.

This CL removes the lines that removes the entry from |frame_sink_source_map_|
if the |frame_sink_id| has no child in
FrameSinkManager::UnregisterFrameSinkManagerClient. It also adds a unit
test to verify that the client gets reconnected to the BeginFrameSource
after unregistering and re-registering.

Bug: 735805
Change-Id: I93ae860b1192e23aa116749d0bbcdb553ddc2246
Reviewed-on: https://chromium-review.googlesource.com/563705
Commit-Queue: Xingyu Zhang <staraz@chromium.org>
Reviewed-by: Fady Samuel <fsamuel@chromium.org>
Reviewed-by: enne <enne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485172}

This implements the ARIA tree pattern: https://www.w3.org/TR/wai-aria-practices-1.1/#TreeView

BUG=560525

Review-Url: https://codereview.chromium.org/2967443002
Cr-Commit-Position: refs/heads/master@{#485171}

The ShouldBeginRequest override in TestResourceDispatcherHostDelegate
is not necessary as it does not change the default behavior of

content: :ResourceDispatcherHostDelegate.
Change-Id: I4678bd494a5aa94698fffe2e1c8bafcabf1a8b71
Reviewed-on: https://chromium-review.googlesource.com/563524Reviewed-by: Peter Kotwicz <pkotwicz@chromium.org>
Commit-Queue: Reilly Grant <reillyg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485170}

Bug: N/A
Test: Build passes.
Change-Id: I1f0b7788b52a0c15b1fda57ec7a153406ef2676f
Reviewed-on: https://chromium-review.googlesource.com/562847Reviewed-by: Keigo Oka <oka@chromium.org>
Reviewed-by: Jacob Dufault <jdufault@chromium.org>
Commit-Queue: Yuichiro Hanada <yhanada@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485169}

R=wez@chromium.org

Change-Id: Ibc345f53a4bbe14bb22ff03bfd871cd17443afaf
Reviewed-on: https://chromium-review.googlesource.com/559951Reviewed-by: Wez <wez@chromium.org>
Commit-Queue: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485168}

This cl contains the following changes
* Implemented ParseLonghandViaAPI in CSSPropertyParserHelpers.
- This function will be used by Consume2Values, Consume4Values and
ConsumeShorthandGreedily. These shorthand parsing functions all need to
forward parsing ops to component longhands.
- At the moment, these functions call CSSPropertyParser's ParseSingleValue.
However, as we will later move these shorthand parsing functions to CSSPropertyParserHelpers,
we should not continue to call CSSPropertyParser.
- Therefore, this cl adds ParseLonghandViaAPI, which contains similar logic to
CSSPropertyParser's ParseSingleValue, with the exception that this new method only works
on longhands that have API impl.

* Moved CountKeywordOnlyPropertyUsage from CSSPropertyParser to CSSPropertyParserHelps.
- This function is used by ParseLonghandViaAPI.

Note, the next cl will have an impl of ConsumeShorthandVia4LonghandsAPI that will use
ParseLonghandViaAPI.

Bug: 668012
Change-Id: Ia9211f9eb5d70124536f5ef0e7d44eeafefa69e2
Reviewed-on: https://chromium-review.googlesource.com/557619
Commit-Queue: Jia Meng <jiameng@chromium.org>
Reviewed-by: Darren Shen <shend@chromium.org>
Reviewed-by: Bugs Nash <bugsnash@chromium.org>
Reviewed-by: Alan Cutter <alancutter@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485167}

In these tests we want to sample animation values at a certain point in
time. We are however sampling the animations at the next frame that is
painted - at which point the timeline may have advanced, and hence we
sample at t+<framedelay> (or thereabout) instead. Currently this works
"fine" because the first animation frame follows special scheduling
rules, making sure the timeline won't advance within this time window.
Rather than relying on this, just pause the timeline instead.

BUG=726362

Change-Id: I2314b435861050fd0163670928318563bd09f271
Reviewed-on: https://chromium-review.googlesource.com/563624
Commit-Queue: Fredrik Söderquist <fs@opera.com>
Reviewed-by: Philip Rogers <pdr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485166}