GitLab

Changes in this patch as below:
1. Replace references to constants in ui/views/layout/layout_constants.h with their
   equivalents using ChromeLayoutProvider.

2. Changed the RequestFileSystemDialogView class to use a simpler FillLayout instead of
   a BoxLayout. This class now subclasses DialogDelegateView instead of DialogDelegate for
   simplicity. Removed the dependency on Extension and Volume datatypes from this class, which
   enables us to easily test the dialog. The extension name, volume label and id are passed in.
   We maintain the contract of invoking the cancel callback if the volume label is empty. That is
   done in the caller ConsentProviderDelegate::ShowDialog().

3. Added browser tests for displaying the TryChromeDialogView and RequestFileSystemDialogView dialogs.

4. Made the TryChromeDialogTest class a friend of the TryChromeDialogView class. This allows it to
   instantiate an instance of this class. The TryChromeDialogView class no longer puts up a modal
   dialog by default. This is controlled by the DialogType enum, which for chrome is always MODAL.
   For the browser tests we pass in MODELESS here, as the loop is run by the test. The ShowModal() method
   is renamed to ShowDialog(). It takes an additional enum UsageType which controls whether we are in testing
   mode. In this case we always want the dialog to be displayed.

BUG=691897
TEST=Covered by tests RequestFileSystemDialogTest.InvokeDialog_default and TryChromeDialogTest.InvokeDialog_default
TBR=grt

Review-Url: https://codereview.chromium.org/2816293002
Cr-Commit-Position: refs/heads/master@{#468195}

When the tab model switches, the bottom sheet content changes to
suit the normal or incognito tab. This can happen in the tab switcher
while the toolbar is transparent causing the toolbar to flash grey.
This change skips the transition animation if in tab switcher mode.

BUG=702717

Review-Url: https://codereview.chromium.org/2849993002
Cr-Commit-Position: refs/heads/master@{#468194}

Previously if you wanted to reset the starting point of the interval
(for measuring an idle timeout, for example) you had to construct a new
instance on the heap.

Now it's possible to keep a stack/member variable (potentially
base::Optional) and assign fresh values to it to reset it.

Review-Url: https://codereview.chromium.org/2851493004
Cr-Commit-Position: refs/heads/master@{#468193}

I ran check_gn_headers.py on linux and android full builds, merged the
missing header file lists, fed to fix_gn_headers.py (without
AddHeadersToSources), chose which matches to apply interactively, and
manually fixed the results.

Read more about the manual fixes in the Review-Url.

The number of missing header files on linux and android decreased by 9,
from 840 to 831.

None of the fixed files are included in the following spreadsheet:
https://docs.google.com/spreadsheets/d/15az3FMl-jAS0mx4E9XVSBVHVpmEzo-9EAGY0ywe7bZs/edit#gid=0

fix_gn_headers.py is from https://codereview.chromium.org/2790563003

BUG=661774

Review-Url: https://codereview.chromium.org/2846463002
Cr-Commit-Position: refs/heads/master@{#468192}

It's currently flaking on the NVIDIA Shield TV.

BUG=716564
TBR=kbr@chromium.org
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

Review-Url: https://codereview.chromium.org/2850973002
Cr-Commit-Position: refs/heads/master@{#468191}

After more testing on top-of-tree, it seems that the fix for Issue
713127 should have addressed all of these. Remove the suppressions in
order to show any remaining issues on the waterfall.

BUG=694359,715696
NOTRY=true
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

Review-Url: https://codereview.chromium.org/2852623003
Cr-Commit-Position: refs/heads/master@{#468190}

BUG=715972
TBR=dpranke@chromium.org

Review-Url: https://codereview.chromium.org/2847783002
Cr-Commit-Position: refs/heads/master@{#468189}

Atomics.wait is a blocking call, and is only allowed on certain workers. See
the new web platform tests here:
https://github.com/w3c/web-platform-tests/pull/5569

BUG=chromium:711809

Review-Url: https://codereview.chromium.org/2841003002
Cr-Commit-Position: refs/heads/master@{#468188}

https://pdfium.googlesource.com/pdfium.git/+log/60cd033adf6c..3b91290ba31e

$ git log 60cd033ad..3b91290ba --date=short --no-merges --format='%ad %ae %s'
2017-04-28 tsepez Remove more |new|s, part 3
2017-04-28 tsepez CXFA_FMStringExpression: int16_t is not a good type for a string index.

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: I74ec80e85a9540d9dcdf2b845b3739ddb8cb8f10
Reviewed-on: https://chromium-review.googlesource.com/490634
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#468187}

This patch ensures that we don't even try to checker images that are
too big to fit into our cache.

R=khushalsagar@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2844273004
Cr-Commit-Position: refs/heads/master@{#468186}

Fixed bug where if a websocket frame was received it would sometimes
move the selected node in network panel to a new request.

Offending patch: https://codereview.chromium.org/2694103007

R=dgozman,caseq
BUG=714229

Review-Url: https://codereview.chromium.org/2838673003
Cr-Commit-Position: refs/heads/master@{#468185}

BUG=671152

Review-Url: https://codereview.chromium.org/2844243003
Cr-Commit-Position: refs/heads/master@{#468184}

BUG=716416
TBR=ben@chromium.org

Change-Id: I765d24506682167248dbc7d51c0d399cda41a741
Reviewed-on: https://chromium-review.googlesource.com/490613Reviewed-by: Ken Rockot <rockot@chromium.org>
Reviewed-by: Ben Goodger <ben@chromium.org>
Commit-Queue: Ken Rockot <rockot@chromium.org>
Cr-Commit-Position: refs/heads/master@{#468183}

The user-visible change here is that Select-to-speak on Chrome OS now
works on the virtual keyboard. It also makes it visible to ChromeVox
so we can make that work better too.

Three fixes were required:

* We need to always set the window property on the window for a
  RenderWidgetHostViewAura, whether that WebContents already has
  accessibility enabled or not. That's solved by moving the
  trigger for the window property to RenderViewReady and
  WebContentsImpl::NotifySwappedFromRenderManager.

* AXAuraObjCache needs to listen for newly-created windows,
  changes to window bounds, and window property changes.
  Otherwise the virtual keyboard window is created but no
  accessibility events fire allowing automation clients to
  find it.

* The bounds for a AXWindowObjWrapper should be global bounds and not
  local. This wasn't caught before because widget bounds were correct,
  but the virtual keyboard is directly in an aura Window, with no Widget
  and no Views.

BUG=699617
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation;master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2803823002
Cr-Commit-Position: refs/heads/master@{#468182}

We recently made a change to treat the MHTML page already being saved to
success, and we added a unit test for the background loader offliner,
but didn't add one for the prerendering offliner.  This adds the test
for the prerendering offliner.

BUG=715152

Review-Url: https://codereview.chromium.org/2850953002
Cr-Commit-Position: refs/heads/master@{#468181}

Calling MaybeReleaseAdapter directly from
BluetoothEventRouter::AdapterDiscoveringChanged could be a problem since
this can happen:
BluetoothAdapterBlueZ::AdapterPropertyChanged ->
BluetoothAdapterBlueZ::DiscoveringChanged ->
BluetoothEventRouter::AdapterDiscoveringChanged -> MaybeReleaseAdapter

If BluetoothEventRouter is the only pointer holder left then the caller
of BluetoothEventRouter::AdapterDiscoveringChanged could be gone even
before that function returns.

BUG=710216

Review-Url: https://codereview.chromium.org/2852773002
Cr-Commit-Position: refs/heads/master@{#468180}

RenderWidgetHostViewChildFrame expects its parent to have a valid
FrameSinkId. Make sure RenderWidgetHostViewAura has a FrameSinkId even
if DelegatedFrameHost is not used (in mus+ash).

BUG=706553
TBR=jam@chromium.org

Review-Url: https://codereview.chromium.org/2847253003
Cr-Commit-Position: refs/heads/master@{#468179}

BUG=705865,715366,716652
TBR=zmo@chromium.org
NOTRY=true
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

Review-Url: https://codereview.chromium.org/2848123002
Cr-Commit-Position: refs/heads/master@{#468178}

Move auxclick manual test under uievent
in wpt as the spec got merged into UIEvent
spec. Also add the test automation.
Currently it seems that the double click
is not fully functional in GPU benchmarking.

BUG=716164

Review-Url: https://codereview.chromium.org/2851633002
Cr-Commit-Position: refs/heads/master@{#468177}

chrome://choose-mobile-network can be launched on a Chromebook without
any cellular modem. This CL changes the UI to handle such case more
gracefully:

- Without overhauling the UI too much, the UI is changed to simply show
  'No networks found'.
- When the Cancel button is pressed, the UI performs nothing.

BUG=714547
TEST=Tested the following:
- Launch chrome://choose-mobile-network on a Chromebook without a
  cellular modem. Verify that the UI shows 'No networks found'.
  Click the Cancel button and verify that the UI doesn't crash.
- Launch chrome://choose-mobile-network on a Chromebook with a cellular
  modem. Verify that the UI can populate a list of networks after
  successfully scanning for networks, and the Connect and Cancel button
  function correctly.

Review-Url: https://codereview.chromium.org/2851863003
Cr-Commit-Position: refs/heads/master@{#468176}

* Make TrustStore implement CertIssuerSource
* Add a method for getting trust/distrust of a certificate
* Remove the TrustAnchor abstraction

All the integrations (including CertVerifyProc) required full certificates anyway, so TrustAnchor ended up being more of a hindrance than benefit.

BUG=649017
TBR=dougsteed@chromium.org

Review-Url: https://codereview.chromium.org/2832703002
Cr-Commit-Position: refs/heads/master@{#468175}

Previously we didn't cache results of collapsed borders computation
during layout or overflow recalculation, so we had to recompute all
collapsed borders before paint invalidation if the table's collapsed
borders are marked invalid.

Now when a table's collapsed borders need to be invalidated, mark all
cells' collapsed borders need to be invalidated [1]. When a cell needs
its collapsed border values (regardless of lifecycle phase), we check
the invalidation flag and update and cache the result when necessary.

As LayoutTableCell::CollapsedBorderValues will be used for both layout
and paint, the include_color paremeter of ComputeCollapsedXXXBorder()
is removed.

As adjacent cells with the same span share collapsed borders, we can
also get the collapsed border from the adjacent cell if its collapsed
borders are valid.

[1] In the next step, we'll only invalidate collapsed borders for
affected cells instead of all cells.

BUG=626748,663208

Review-Url: https://codereview.chromium.org/2846563002
Cr-Commit-Position: refs/heads/master@{#468174}

Everything calling getBindingManager is already on launcher thread.
This removes the last lock in android's ChildProcessLauncher.

BUG=689758

Review-Url: https://codereview.chromium.org/2849943002
Cr-Commit-Position: refs/heads/master@{#468173}

BUG=715371
TEST=tests updated

Review-Url: https://codereview.chromium.org/2841223002
Cr-Commit-Position: refs/heads/master@{#468172}

R=hcarmona@chromium.org
BUG=715866

Review-Url: https://codereview.chromium.org/2848983002
Cr-Commit-Position: refs/heads/master@{#468171}

Make use of the existing PurgeEntriesWithTypeIn to purge the directory
of entries of the type we just migrated. Add a test that ensures this
happens.

R=skym@chromium.org

BUG=716156

Review-Url: https://codereview.chromium.org/2842373004
Cr-Commit-Position: refs/heads/master@{#468170}

BUG=690020

patch from issue 2847643002 at patchset 20001 (http://crrev.com/2847643002#ps20001)

Review-Url: https://codereview.chromium.org/2848683002
Cr-Commit-Position: refs/heads/master@{#468169}

The render surface layer list structure consisted of a top level
list of layers representing render surfaces, along with a list of
layers at each surface representing contributing layers or surfaces.

This CL replaces the render surface layer list with a top level
list of surfaces. Individual surfaces no longer store a list of
contributors.

BUG=611883
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2834123002
Cr-Commit-Position: refs/heads/master@{#468168}

Applications sometimes use window.rAF while not presenting, then switch to
vrDisplay.rAF after presentation starts. Depending on the animation loop's
timing, this can cause a race condition where presentation has been started
but there's no vrDisplay.rAF pending yet. Ensure there's at least vsync
being processed after presentation starts so that a queued window.rAF
can run and schedule a vrDisplay.rAF.

BUG=711789

Review-Url: https://codereview.chromium.org/2848483003
Cr-Commit-Position: refs/heads/master@{#468167}

Replacing a TODO with actual store state.

BUG=714256
R=petewil@chromium.org

Review-Url: https://codereview.chromium.org/2841113003
Cr-Commit-Position: refs/heads/master@{#468166}

Revert of Enable HTTP/2 to use a LIFO scheduler to schedule write. (patchset #1 id:1 of https://codereview.chromium.org/2844963005/ )

Reason for revert:
Reverting in hopes of fixing net_unittests. Specifically NSS/ClientCertStoreTest/0.CertAuthorityFiltering . See 716630.

Original issue's description:
> Enable HTTP/2 to use a LIFO scheduler to schedule write.
>
> This CL lands server change 152524616 by yasong.
>
> BUG=488484
>
> Review-Url: https://codereview.chromium.org/2844963005
> Cr-Commit-Position: refs/heads/master@{#467971}
> Committed: https://chromium.googlesource.com/chromium/src/+/c443c00c0c9e10b8ced1064d7da18c3a9668cac2

TBR=bnc@chromium.org,yasong@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=488484

Review-Url: https://codereview.chromium.org/2849983002
Cr-Commit-Position: refs/heads/master@{#468165}

This CL:
* Fixes button spacing in the network details page.
* Fixes the on/off toggle button font and color for network and
  bluetooth
* Replaces incorrect primary/secondary button classes in dialogs
  with action/cancel-button
* Corrects the blue color for action buttons as per UX
* Converts the "add network" section to list-frame / list-item
  to correct borders and spacing.

BUG=710258,710259,710261
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2846763003
Cr-Commit-Position: refs/heads/master@{#468164}

Since there is no equivalent of ScopedComPtr::ReceiveVoid in ComPtr,
we have to mimic ComPtrRef and add ScopedComPtrRef to get the same
convenient void** access. This does introduce some complications for
uses of ScopedComPtr<T>* in templates and in concrete types, but there
aren't very many of those.

BUG=638327
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=dalecurtis@chromium.org, thakis@chromium.org, wez@chromium.org

Review-Url: https://codereview.chromium.org/2850683002
Cr-Commit-Position: refs/heads/master@{#468163}

Requested by localization folks. Existing strings are difficult to
understand.

BUG=716574
TEST=none
TBR=cpu

Review-Url: https://codereview.chromium.org/2840323002
Cr-Commit-Position: refs/heads/master@{#468162}

The action will be used in a future changelist, landing soon.

BUG=687231

Review-Url: https://codereview.chromium.org/2847023002
Cr-Commit-Position: refs/heads/master@{#468161}

BUG=715466
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2839413004
Cr-Commit-Position: refs/heads/master@{#468160}

BUG=696009

Review-Url: https://codereview.chromium.org/2714193004
Cr-Commit-Position: refs/heads/master@{#468159}

This change forwards navigator.requestKeyLock() and navigator.cancelKeyLock()
functions to the RenderFrameHostImpl.

System-Keyboard-Lock is a feature to detect the key presses which usually cannot
be received by the browser, and send them to the web page. It contains various
components to achieve the goal. This change is one of them, which receives
JavaScript (or Web Platform API in the design doc) function calls and forwards
them into RenderFrameHost. For detail, please refer to the design doc at,
https://docs.google.com/document/d/1T9gJHYdA1VGZ6QHQeOu0hOacWWWJ7yNEt1VDbLju4bs/edit#heading=h.cgwemqs2j4ta

W3C Working Draft: https://garykac.github.io/system-keyboard-lock/
Intent to implement: https://groups.google.com/a/chromium.org/forum/#!msg/blink-dev/9pauQUAvrcw/lfbG7eunCAAJ

BUG=680809
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2805763004
Cr-Commit-Position: refs/heads/master@{#468158}

View's native theme.

This fixes a few bugs where we weren't using the right native theme,
whether by not updating after the NativeTheme changes or by trying to
access the NativeTheme before the View is added to a hierarchy (which
yields the default NativeTheme --- for most platforms, this didn't
effectively create a bug as there's only one NativeTheme).

Best example is that now the sad tab respects the GTK native theme.

Get rid of Ash's tray_constants::kBackgroundColor in favor of using the
bubble background color from the NativeTheme.

This is a reland of 097f9cde with fix
for asan failure.

BUG=711183, 693282
TBR=stevenjb@chromium.org,tdanderson@chromium.org

Review-Url: https://codereview.chromium.org/2838273002
Cr-Commit-Position: refs/heads/master@{#468157}

When there are no windows opened, do not change the auto_hide status when
swiping down (or swiping up and then releasing) on the shelf.

BUG=310322

Review-Url: https://codereview.chromium.org/2824233003
Cr-Commit-Position: refs/heads/master@{#468156}