GitLab

Skia makes no guarantees about the contents of a newly allocated
SkBitmap. Consequently, it's possible that the bitmap allocated for the
overscroll effect isn't zeroed before the effect itself is painted.
Prevent related issues by always clearing the allocated SkBitmap's
contents (to fully transparent) before painting the effect.

BUG=408225

Review URL: https://codereview.chromium.org/484373004

Cr-Commit-Position: refs/heads/master@{#292280}

We just added a new error code to ChromeOS's update_engine so also
update histograms.xml.

BUG=chromium:406546
TEST=Validated the xml, pretty print succeeded.

Review URL: https://codereview.chromium.org/514003002

Cr-Commit-Position: refs/heads/master@{#292279}

BUG=408294
TBR=dpranke

Review URL: https://codereview.chromium.org/510153003

Cr-Commit-Position: refs/heads/master@{#292278}

Picks up https://codereview.chromium.org/508343002

BUG=82385
TBR=rnk@chromium.org

Review URL: https://codereview.chromium.org/515773002

Cr-Commit-Position: refs/heads/master@{#292277}

Revert of Adding Select Action Bar Unit test case for input (patchset #3 of https://codereview.chromium.org/506423003/)

Reason for revert:
Failed compile on Android builders.

Original issue's description:
> [Android] Add unit tests for SelectActionBar appearance
>
> The SelectActionBar currently lacks test coverage. Add several test
> cases covering the appearance of the bar when interacting with an input
> field.
>
> Committed: https://chromium.googlesource.com/chromium/src/+/0bc56569817a24acea46748c599d77b59157d84d

BR=jdduke@chromium.org,aurimas@chromium.org,ajith.v@samsung.com
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/514493004

Cr-Commit-Position: refs/heads/master@{#292276}

(This removes the one place where we did. Sadly, there's no way to
assert that a lock isn't held by the current thread.)

R=darin@chromium.org

Review URL: https://codereview.chromium.org/488003006

Cr-Commit-Position: refs/heads/master@{#292275}

R=wtc@chromium.org

Review URL: https://codereview.chromium.org/511543006

Cr-Commit-Position: refs/heads/master@{#292274}

This reverts HidService to a singleton instance, adding two gross
and presently hard-to-avoid properties:

1. Its instance getter takes a UI message loop proxy
2. Its lifetime is bound by the message loop of its origin thread

The purpose of this is to preserve the necessary threading discipline
required by both the chrome.hid API implementation and the
permission_broker DBus client while ensuring that the HidService is
created and destroyed on the same thread.

BUG=401234

Review URL: https://codereview.chromium.org/464753002

Cr-Commit-Position: refs/heads/master@{#292273}

BUG=403597

Review URL: https://codereview.chromium.org/490193003

Cr-Commit-Position: refs/heads/master@{#292272}

BUG=144491

Review URL: https://codereview.chromium.org/440813002

Cr-Commit-Position: refs/heads/master@{#292271}

Instad of generating exported validateFooRequest(),
validateFooResponse() methods, a pointer to a validation
functions is added to the generated Stub and Proxy class
prototypes. For example, given an interface Foo, with
client interface FooClient:

  FooStub.prototype.validators = [validateFooRequest];
  FooProxy.prototype.validators = [validateFooResponse];

Note that if interface Foo has no callbacks, then
FooProxy.prototype.validators is an empty array.

The router validates incoming messages using these functions
and validateMessageHeader().

BUG=407181

Review URL: https://codereview.chromium.org/488173006

Cr-Commit-Position: refs/heads/master@{#292270}

Sysfs related functionality should have never been a separate Platform. There
should just be one Platform at a time.

BUG=

Review URL: https://codereview.chromium.org/509043002

Cr-Commit-Position: refs/heads/master@{#292269}

BUG=405642

Review URL: https://codereview.chromium.org/490033002

Cr-Commit-Position: refs/heads/master@{#292268}

BUG=407874,407861

Review URL: https://codereview.chromium.org/507873002

Cr-Commit-Position: refs/heads/master@{#292267}

Revert of Re-enable memory tests. Fixed sign-in issue. (patchset #3 of https://codereview.chromium.org/491863003/)

Reason for revert:
Failing on bots. Looks like we forgot to commit the archive.

Traceback (most recent call last):
  File "/b/build/slave/Android_Nexus5_Perf/build/src/tools/perf/../telemetry/telemetry/value/failure.py", line 39, in _GetExcInfoFromMessage
    raise Exception(message)
Exception: Page set archive doesn't exist.

http://chromegw/i/chromium.perf/builders/Android%20Nexus5%20Perf/builds/1480/steps/memory.mobile_memory/logs/stdio

Original issue's description:
> Re-enable memory tests. Fixed sign-in issue.
>
> Re-recording other tests may be required.
>
> BUG=370977
>
> Committed: https://chromium.googlesource.com/chromium/src/+/383ffaab6b5fac34b950d9faaab637cc75907aca

TBR=skyostil@chromium.org,starodub@google.com
NOTREECHECKS=true
NOTRY=true
BUG=370977

Review URL: https://codereview.chromium.org/511023002

Cr-Commit-Position: refs/heads/master@{#292266}

Currently only one site that deletes a tile removes it from the
recycle twin. Change the others to also, or DCHECK showing they don't
need to.

R=enne, vmpstr

Review URL: https://codereview.chromium.org/513903002

Cr-Commit-Position: refs/heads/master@{#292265}

* Prep for athena on chrome.

Another notable change:
* pass app_id to AppActivity. it's known at creation time.

Note: this depends on https://codereview.chromium.org/505273002/

BUG=397167

Review URL: https://codereview.chromium.org/501093003

Cr-Commit-Position: refs/heads/master@{#292264}

The ABS_X and ABS_Y axes are not part of the MT protocol and should
therefore not be used to update MT slot positions.

The symptom of this was secondary touches warping to the position of the
first touch in the X and/or Y axis, since ABS_X/Y always corresponds to
the first touch.

BUG=407386
TEST=chrome on pixel with --ash-touch-hud
NOTRY=true

Review URL: https://codereview.chromium.org/506303002

Cr-Commit-Position: refs/heads/master@{#292263}

John understands the infrastructure at a deep level and will provide
good code reviews and nominate others appropriately.

BUG=none
R=piman@chromium.org

Review URL: https://codereview.chromium.org/510173002

Cr-Commit-Position: refs/heads/master@{#292262}

Adds a toolchain variable "concurrent_links". The value of this variable in the default toolchain is used as the depth of the link pool for the build.

Remove the "pool" stuff from the tool definitions that was never hooked up. The link pool is hardcoded in GN and is controlled by the above variable.

R=scottmg@chromium.org

Review URL: https://codereview.chromium.org/508763002

Cr-Commit-Position: refs/heads/master@{#292261}

Android devtools tries open socket twice on same UnixDomainServerSocket
instance. Instead of destroying UnixDomainServerSocket instance, not assigning
internal listen socket on socket open error is cheaper and makes more sense.

BUG=408224

Review URL: https://codereview.chromium.org/515653003

Cr-Commit-Position: refs/heads/master@{#292260}

sizeWithFont: and drawInRect:withFont:lineBreakMode:alignment: were
deprecated in iOS 7. This CL roll GTM to a revision where those methods
are no longer used (when building against iOS 7 or later).

BUG=405578

Review URL: https://codereview.chromium.org/507333003

Cr-Commit-Position: refs/heads/master@{#292259}

While attempting to resolve a flaky browser_test, it became clear that
AudioMirroringManager was not outliving its use.  In the original change
that introduced this class, it was instantiated and owned by
BrowserMainLoop as a matter of convenience.  However, with some
debugging, it's clear that it must outlive objects that can outlive
BrowserMainLoop (e.g., WebContentsAudioInputStream).

Side notes: I've checked the feasibility of other solutions, confirming
that the shutdown of AudioManager does NOT guarantee complete teardown
of an AudioInputStream, so it's not sufficient to simply change the
destruction order of the objects in BrowserMainLoop to resolve this
problem.  As AudioMirroringManager provides a browser-wide service and
owns no objects, it seems reasonable for it to exist as a global
LazyInstance.

BUG=396413

Review URL: https://codereview.chromium.org/499483003

Cr-Commit-Position: refs/heads/master@{#292258}

Currently chrome creates a file path when uploading a photo taken through camera
However, this doesn't always work.
Instead, we should use content URI. See more details in the crbug.

BUG=405593

Review URL: https://codereview.chromium.org/489053003

Cr-Commit-Position: refs/heads/master@{#292257}

Plugins are not enabled on android or ios.

BUG=none
R=brettw@chromium.org

Review URL: https://codereview.chromium.org/508823004

Cr-Commit-Position: refs/heads/master@{#292256}

BUG=398951

Review URL: https://codereview.chromium.org/507133002

Cr-Commit-Position: refs/heads/master@{#292255}

BUG=408276

Review URL: https://codereview.chromium.org/513973002

Cr-Commit-Position: refs/heads/master@{#292254}

We want to see if it continues to cause flake and noise on the bots. If so, feel free to revert.

BUG=336558
TEST=None.
R=tonyg

Review URL: https://codereview.chromium.org/510143003

Cr-Commit-Position: refs/heads/master@{#292253}

Revert of Adding Select Action Bar Unit test case for input (patchset #3 of https://codereview.chromium.org/506423003/)

Reason for revert:
See previous review message.

Original issue's description:
> [Android] Add unit tests for SelectActionBar appearance
>
> The SelectActionBar currently lacks test coverage. Add several test
> cases covering the appearance of the bar when interacting with an input
> field.
>
> Committed: https://chromium.googlesource.com/chromium/src/+/0bc56569817a24acea46748c599d77b59157d84d

TBR=jdduke@chromium.org,aurimas@chromium.org,ajith.v@samsung.com
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/510953003

Cr-Commit-Position: refs/heads/master@{#292252}

Have ui::KeycodeConverter::NativeKeycode use evdev codes when
compiled appropriately.

TEST=Manually tested on link_freon.
BUG=362698

Review URL: https://codereview.chromium.org/488123004

Cr-Commit-Position: refs/heads/master@{#292251}

This allows us to output the exception message even if no test has explicitly failed.

BUG=407976
R=kbr@chromium.org
TBR=kbr@chromium.org

Review URL: https://codereview.chromium.org/515513003

Cr-Commit-Position: refs/heads/master@{#292250}

Before this refactoring, the ES management was left
to the ES parser implementation.
It is now moved to the EsParser class and ES management
now relies explicitly on an ES byte queue.

BUG=None

Review URL: https://codereview.chromium.org/497203004

Cr-Commit-Position: refs/heads/master@{#292249}

The SelectActionBar currently lacks test coverage. Add several test
cases covering the appearance of the bar when interacting with an input
field.

Review URL: https://codereview.chromium.org/506423003

Cr-Commit-Position: refs/heads/master@{#292248}

The SelectActionBar currently lacks test coverage. Add several test
cases covering the appearance of the bar when interacting with an input
field.

Review URL: https://codereview.chromium.org/506423003

Cr-Commit-Position: refs/heads/master@{#292247}

BUG=

Review URL: https://codereview.chromium.org/508233002

Cr-Commit-Position: refs/heads/master@{#292246}

BUG=279351
For chrome/browser/ui/webui/chromeos/network_config_message_handler.cc:
TBR=xiyuan@chromium.org

Review URL: https://codereview.chromium.org/492383002

Cr-Commit-Position: refs/heads/master@{#292245}

As described in bug 396413, it was possible for a posted task to be run
after CastReceiverImpl/FrameReceiver were destroyed.  This change fixes
the bug by having FrameReceiver post a cancelable task (i.e., using its
WeakPtr) instead.

BUG=396413

Review URL: https://codereview.chromium.org/491953003

Cr-Commit-Position: refs/heads/master@{#292244}

This is a reland of
https://codereview.chromium.org/489373005/
which was reverted because it caused failures in interactive_ui_tests on mac
on OmniboxViewTest.AcceptKeywordBySpace:
https://build.chromium.org/p/chromium.mac/builders/Mac%2010.6%20Tests%20(dbg)(1)/builds/52199
See [1].

It turns out these failures were because the test uses a bookmark with no
title and BookmarkProvider doesn't set classifications for these correctly.
(It set the title as having a classification where, without any text, it
doesn't make sense to have any classification.)

This change fixes this issue and adds a test in BookmarkProvider for it.

I will annotate the *second* patchset I upload with the differences
between the previously-reviewed changelist and this changelist.
(The first patchset has debug lines; ignore it.)

Otherwise, the changelist description remains the same, which I will paste
below [2].

I am marking this change as
TBR=sky
(because he previously reviewed this change and the new parts are
 not under his purview.)

---

[1]

OmniboxViewTest.AcceptKeywordBySpace (run #1):
[ RUN      ] OmniboxViewTest.AcceptKeywordBySpace
[12075:263:0822/122851:ERROR:location_bar_view_mac.mm(592)] Not implemented reached in virtual void LocationBarViewMac::EndOriginChipAnimations(bool)
[12075:263:0822/122852:FATAL:autocomplete_match.cc(516)] Check failed: classifications.empty().
0   libbase.dylib                       0x1ca75c5f base::debug::StackTrace::StackTrace() + 63
1   libbase.dylib                       0x1ca75cbb base::debug::StackTrace::StackTrace() + 43
2   libbase.dylib                       0x1cb21862 logging::LogMessage::~LogMessage() + 82
3   libbase.dylib                       0x1cb2058b logging::LogMessage::~LogMessage() + 43
4   interactive_ui_tests                0x11fb469d AutocompleteMatch::ValidateClassifications(std::basic_string\u003Cunsigned short, base::string16_char_traits, std::allocator\u003Cunsigned short> > const&, std::vector\u003CAutocompleteMatch::ACMatchClassification, std::allocator\u003CAutocompleteMatch::ACMatchClassification> > const&) const + 317
5   interactive_ui_tests                0x11fb453b AutocompleteMatch::Validate() const + 107
6   interactive_ui_tests                0x11fbbb51 AutocompleteResult::Validate() const + 113
7   interactive_ui_tests                0x0d6f5a3f AutocompleteController::UpdateResult(bool, bool) + 815
8   interactive_ui_tests                0x0d6f55b7 AutocompleteController::Start(AutocompleteInput const&) + 1799
...

[2]

This used to be controlled by a field trial created in
https://codereview.chromium.org/184663002

After evaluating this change, we've decided to launch it.
This change turns the flag on by default and removes the field trial code.
It also removes the index_urls_ parameters everywhere because those will
always be true forevermore.

I tested this interactively.
Also, all the unit tests still apparently pass.  And yes some of these do
exercise this feature.

BUG=157204,378854

Review URL: https://codereview.chromium.org/510533002

Cr-Commit-Position: refs/heads/master@{#292243}

BUG=407863

Review URL: https://codereview.chromium.org/512493003

Cr-Commit-Position: refs/heads/master@{#292242}

BUG=408189

Review URL: https://codereview.chromium.org/509103002

Cr-Commit-Position: refs/heads/master@{#292241}