GitLab

Review URL: https://codereview.chromium.org/896463005

Cr-Commit-Position: refs/heads/master@{#314693}

This reverts commit cc6a8bfe.
From review: https://codereview.chromium.org/899973002/

Original issue's description:
> Device Orientation API on Chrome OS
>
> Implement DataFetcherSharedMemory on Chrome OS to provide Device Orientation support. Create SensorManagerChromeOS to observe accelerometer events, and to generate updates to the Device Orientation API.
>
> This is a follow up to: https://codereview.chromium.org/680383007/
>
> With the accelerometer on Chrome OS now exposed to content/ from chromeos/accelerometer, no ash/ implementation is required. No public api in content/public/browser is needed either.
>
> TEST=SensorManagerTest.OrientationBuffer, SensorManagerTest.NeutralOrientation, SensorManagerTest.UpsideDown, SensorManagerTest.BeforeUpsideDownBoundary, SensorManagerTest.LeftEdge, SensorManagerTest.RightEdge, SensorManagerTest.BeforeRightEdgeBoundary
> BUG=342908

TBR=timvolodine@chromium.org,jam@chromium.org
BUG=342908

Review URL: https://codereview.chromium.org/880083005

Cr-Commit-Position: refs/heads/master@{#314692}

Instead of parsing the whole proc stats just to extract two values, we
skim through it just to extract the two values we're interested in.

BUG=453871
TEST=base_unittests --gtest_filter=ProcessMetricsTest.*

Review URL: https://codereview.chromium.org/888193002

Cr-Commit-Position: refs/heads/master@{#314691}

There is a bug where the "Close Window" shortcut for hosted/packaged
apps would disappear after focusing the File menu in a Chrome browser.
This is because the shortcuts had the same key equivalent, and OS X was
choosing to only display one of them. This CL fixes this by keeping a
reference to the conflicting menu items and enabling/disabling the
shortcuts on focus.

BUG=455067

Review URL: https://codereview.chromium.org/898543008

Cr-Commit-Position: refs/heads/master@{#314690}

testGetUpdatedInspectableContexts should only run when tab control is supported.

BUG=452277

Review URL: https://codereview.chromium.org/893233003

Cr-Commit-Position: refs/heads/master@{#314689}

- ! allowed by a tts property will annotate the output with that property
- $descendants(args...) where args... is a comma delimited list of roles, will compute output for the node's descendants that are in the set of roles.
- apply these two new functions to compute utterances for alerts.

TEST=navigate to a page with an alert like
html5demos.com/geo.
BUG=382650

Review URL: https://codereview.chromium.org/816093002

Cr-Commit-Position: refs/heads/master@{#314688}

This removes the need to compare a raw handle against
a predefined value from the caller's code.

BUG=none
TEST=base_unittests

Review URL: https://codereview.chromium.org/868143004

Cr-Commit-Position: refs/heads/master@{#314687}

- Switch to using expand and collapse icons from internal repo
- Start using TintedDrawable for WebsitePreferences.

BUG=454859

Review URL: https://codereview.chromium.org/900743003

Cr-Commit-Position: refs/heads/master@{#314686}

Add a basic settings page with a Polymer paper element which loads at chrome://md-settings (if the enable-md-settings flag is enabled).

BUG=454961

Review URL: https://codereview.chromium.org/893323002

Cr-Commit-Position: refs/heads/master@{#314685}

Long live cc.

R=enne, vmpstr

Review URL: https://codereview.chromium.org/897873005

Cr-Commit-Position: refs/heads/master@{#314684}

BrowserGpuChannelHostFactory::GetGpuChannel() can
return NULL. In particular it already checks
IsLost() and returns NULL if that is the case.

BUG=445945

Review URL: https://codereview.chromium.org/867093004

Cr-Commit-Position: refs/heads/master@{#314683}

  - Fail if VideoToolbox returns a null pointer, instead of crashing.
  - Stop processing as soon as an error occurs.

BUG=453050

Review URL: https://codereview.chromium.org/900053002

Cr-Commit-Position: refs/heads/master@{#314682}

Review URL: https://codereview.chromium.org/873923004

Cr-Commit-Position: refs/heads/master@{#314681}

Replaces the use of MultiChannelResampler with AudioConverter in
FakeMediaSource to allow the cast_sender_app and cast_simulator to
re-mix any-number-of-channels audio into the 2-channel audio for Cast
sending.

Review URL: https://codereview.chromium.org/900913004

Cr-Commit-Position: refs/heads/master@{#314680}

BUG=455349

Review URL: https://codereview.chromium.org/866913006

Cr-Commit-Position: refs/heads/master@{#314679}

On android versions before 5.0, EGL sync objects are
incorrectly destroyed when the associated EGL context is
destroyed. The EGL sync objects error handling is tight to
prevent driver regressions from known issues. This combined
with this triggers DCHECk failures and crashes.

So add another workaround to ignore errors from EGL sync
object calls.

BUG=453857

Review URL: https://codereview.chromium.org/892323002

Cr-Commit-Position: refs/heads/master@{#314678}

The Google style guide states that only one of {virtual,override,final}
should be used for each declaration, since override implies virtual and
final implies both virtual and override.

This patch was automatically generated with an OS=android build using a
variation of https://codereview.chromium.org/598073004.

BUG=417463
R=caitkp@chromium.org

Review URL: https://codereview.chromium.org/893303003

Cr-Commit-Position: refs/heads/master@{#314677}

CaptivePortalBrowserTest.InterstitialTimerNavigateAwayWhileLoading was
disabled a few days ago, this one has the same issue.

BUG=453875
TBR=thestig@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/902583005

Cr-Commit-Position: refs/heads/master@{#314676}

Sets list item selection and focus correctly when edit is committed in
name/phone/email lists on chrome://settings/autofillEditAddress page.

"Correctly" means the just-edited list item remains the selected item in
the list, and the next focusable element after the list gets focus. The
list no longer has focus, but if the user gives focus back to the list
using tab/shift+tab then the just-edited list item will receive focus.

This is different than the default InlineEditableItemList behavior
because most addresses will have only a single name/phone/email, so it's
better to move to the next field type rather than to the "Add another"
placeholder of the same field type. I tried to keep the changes
contained to autofill_options_list.js because that's the only page that
needs this focus behavior. Generalizing the code into
inline_editable_list.js might result in more regressions to fix
elsewhere.

The fix implementation is to disable list events during list operations
in autofillOptions.ValuesListItem.onEditCommitted_, so that focus is not
returned back to the list that was just tabbed away from. Then
onEditCommitted_ calls the new
InlineEditableItemList.SelectIndexWithoutFocusing method to set the
selected list item without setting focus back to the list.

Same fix for both regressions.

https://crbug.com/440760 was caused by me.
https://crbug.com/443491 was from a CL by hcarmona@, but we agreed I
should take it because I'm more familiar with this code.

BUG=440760,443491

Review URL: https://codereview.chromium.org/819193003

Cr-Commit-Position: refs/heads/master@{#314675}

This fixes a bunch of small nits:
* JSCompile is now enabled for this app.
* No need to manually focus the PNaCl plugin.
* Distinguishes between left and right modifiers for PNaCl events.
* Debug log is hidden by default, and selectable when shown.
* Minor CSS tweaks and code clean-up.

Review URL: https://codereview.chromium.org/900983003

Cr-Commit-Position: refs/heads/master@{#314674}

BUG=454940

Review URL: https://codereview.chromium.org/902443002

Cr-Commit-Position: refs/heads/master@{#314673}

This includes a fix for Python 2.7.7, which removed the
HTTPSConnection._set_hostport function.

BUG=410545
R=sbc@chromium.org

Review URL: https://codereview.chromium.org/904513005

Cr-Commit-Position: refs/heads/master@{#314672}

Review URL: https://codereview.chromium.org/892203004

Cr-Commit-Position: refs/heads/master@{#314671}

Currently in the webapp, we didin't disconnect the native messaging port
after a share and therefore keeping the remote-assistance process alive.

Summary of changes:
1. Disconnect the native message port when the host status changes to
   DISCONNECTED or ERROR.
2. Remove remoting.hostSession global and wrap host_screen.js in an
   anonymous function.

BUG=416676

Review URL: https://codereview.chromium.org/895723006

Cr-Commit-Position: refs/heads/master@{#314670}

Follow-up to crrev.com/895523004/

BUG=

Review URL: https://codereview.chromium.org/896853002

Cr-Commit-Position: refs/heads/master@{#314669}

setIgnoreViewportTagScaleLimits copies the default limits into the UA
constraints. This was happening before the defualt limits were set during
initialization so it was clamping page scale at [1,1].

BUG=453377

Review URL: https://codereview.chromium.org/860283005

Cr-Commit-Position: refs/heads/master@{#314668}

TBR=bruening@google.com
BUG=455417
NOTRY=true

Review URL: https://codereview.chromium.org/896333002

Cr-Commit-Position: refs/heads/master@{#314667}

Trying base_unittests first to check that isolation works correctly.

BUG=454828
TBR=phajdan.jr@chromium.org

Review URL: https://codereview.chromium.org/897023002

Cr-Commit-Position: refs/heads/master@{#314666}

When in using single thread with browser compositor, make reflector use
main thread MessageLoop instead.

Test with forced SW mirrored display on both Linux-cros build and Pixel.
BUG=453732

Review URL: https://codereview.chromium.org/899073003

Cr-Commit-Position: refs/heads/master@{#314665}

At this point, "--run-layout-test" is the flag that is actually functional, and using "--dump-render-tree" simply causes the "--run-layout-test" to be appended to the command line.

A subsequent Blink patch will change the scripts in Blink to use "--run-layout-test" instead of "--dump-render-tree".

BUG=453252

Review URL: https://codereview.chromium.org/899103002

Cr-Commit-Position: refs/heads/master@{#314664}

Reason for revert:
4000+ browser test are failing on ChromeOS after this patch and this is the only patch that touches ChromeOS in the blamelist.

Seems like with the switch enabled, we are creating the GetTimezoneResolver that leaks the g_browser_process->system_request_context()

Sample failure:
http://build.chromium.org/p/chromium.chromiumos/builders/Linux%20ChromiumOS%20Tests%20%28dbg%29%281%29/builds/37149

Original issue's description:
> https://codereview.chromium.org/898503002
> Timezone auto update option should be available by default.
>
> This CL makes "timezone auto update" feature available (but > switched off by
> default) to users. And adds flag to disable this feature.
>
> BUG=416494
> TEST=manual
>
> Committed:
> https://crrev.com/13bb9e13b4719c96112ca8770d2c7b48c23e150d
> Cr-Commit-Position: refs/heads/master@{#314615}

TBR=alemate
TEST=MANUAL
NOTRY=true
NOTREECHECKS=true
NOPRESUBMIT=true

Review URL: https://codereview.chromium.org/899113002

Cr-Commit-Position: refs/heads/master@{#314663}

BUG=454822
TBR=glider@chromium.org

Review URL: https://codereview.chromium.org/901043002

Cr-Commit-Position: refs/heads/master@{#314662}

The default theme's background image is a subtle texture pattern that we can
scale without being easily noticed.
When using the default theme, let the layer get scaled during live resize
and redraw at the end of resize.

TabStripView is changed to no longer be a subclass of
BackgroundGradientView because we don't want to scale its layer (it may
draw a drop arrow).
It is also inefficient for it to call cr_recursivelySetNeedsDisplay.

With this optimization, the following views no longer redraw repeatedly
during live resize (when using default theme):
    ToolbarView
    BookmarkBarToolbarView
    DownloadShelfView
    OmniboxPopupTopSeparatorView

BUG=453996

Review URL: https://codereview.chromium.org/867273003

Cr-Commit-Position: refs/heads/master@{#314661}

Tracking areas in Cocoa are a horrible mess. 10.5 semi-deprecated
NSWindow cursorRects, and introduced [NSResponder cursorUpdate:]. Both
are supported and they react with each other in strange ways.
Intercepting events for toolkit-views SetCapture seems to further
complicate things.

After numerous dead-ends, this CL implements SetCursor() by overriding
cursorUpdate: on NativeWidgetMac's NSWindow override. This works because
cursorUpdate: messages are forwarded along the responder chain (starting
from varied and somewhat unpredictable starting points), and usually end
up at the NSWindow.

Rather than adding more NSTrackingArea cruft, the one in
BridgedContentView is updated to handle cursor updates. This ensures the
cursor is reset to "normal" whenever it leaves the content area.

Installing a global event monitor also resets the mouse cursor to an
arrow. I tried tracing this with NSObjCMessageLoggingEnabled, but the
reset isn't done with Objective-C. To fix, "remind" AppKit what the
cursor should be after setting capture.

See code comments for all the other traps.

Testing tracking rectangles with simulated events is pretty much a lost
cause, but NativeWidgetMacTest.SetCursor() is added which gets some
coverage by sending an event directly to [NSWindow cursorUpdate:], and
using the ui::test::EventGenerator to forward events to toolkit-views to
actually request new cursors to set.

BUG=454698

Review URL: https://codereview.chromium.org/891003004

Cr-Commit-Position: refs/heads/master@{#314660}

Follow-up on comment from https://codereview.chromium.org/889013002/
TBR=ppi@chromium.org

Review URL: https://codereview.chromium.org/902593003

Cr-Commit-Position: refs/heads/master@{#314659}

R=kalman@chromium.org
BUG=455196
TEST=visual

Review URL: https://codereview.chromium.org/900993002

Cr-Commit-Position: refs/heads/master@{#314658}

After having transitioned all the tracing headers from base/debug/ to
base/trace_event, this CL addresses the namespace move.
In principle, this CL should only change the namespace of the
base/trace_event files but the namespace used by the tracing clients.
In order to achieve this, namespace aliases are appended to the
trace_event headers, to make it so that clients can still refer to
base::debug::TraceFoo, with that being aliased to
base::trace_event::TraceFoo.
The upcoming CLs will gradually migrate the clients to use the
base::trace_event namespace and will remove the ns aliases.
Unfortunately, this CL has also to update few tracing clients,
in particular the ones having forward declarations. Forward
declarations, in fact, cannot be aliased as the compiler sees them
before the alias itself.

See crrev.com/837303004 and the related bug for motivations and design doc.

BUG=451032
TBR=skyostil@chromium.org,jam@chromium.org

Review URL: https://codereview.chromium.org/882673002

Cr-Commit-Position: refs/heads/master@{#314657}

The existing comment references the old name of finalize_apk_action.gypi, this
updates it to its current name.

This also reorders the input list to be alphabetized.

Review URL: https://codereview.chromium.org/899063002

Cr-Commit-Position: refs/heads/master@{#314656}

BUG=454967
NOTRY=true
TBR=aboxhall@chromium.org

Review URL: https://codereview.chromium.org/904523003

Cr-Commit-Position: refs/heads/master@{#314655}

BUG=453188

Review URL: https://codereview.chromium.org/898483006

Cr-Commit-Position: refs/heads/master@{#314654}