GitLab

Previous instrumentations showed that AccountIdFetcher::Start
alone is responsible for 20.5 janks per hour in UI thread, with 203ms average
run time. I need to instrument
the code inside it to find out which part causes jank.

This is a mechanical change that adds instrumentation required to locate the
source of jankiness (i.e. a long-running fragment of code executed as a part of
the task that causes jank) in the code. See the bug for details on what kind of
jank we are after.
A number of similar CLs were landed, and none of them caused issues. They've
helped to find and fix janky code. The code of the instrumentation is highly
optimized and is not expected to affect performance. The code simply creates a
diagnostic task which is identical to ones created by PostTask or IPC message
handlers. The task gets created only in developer build and in Canary channel.

BUG=422460

Review URL: https://codereview.chromium.org/685033006

Cr-Commit-Position: refs/heads/master@{#302649}

Introduces the metadata required to link together the desktop tree and a views::WebView. Additionally, implements the js side logic to resolve these links.

A views::WebView now gets assigned an AXRole of webView and its immediate child resolves to the rootWebArea it hosts. This resolution is done dynamically at runtime using the underlying AX tree id which itself is mapped using AXTreeIDRegistry back in the browser. If the hosting webView has yet to load its child rootWebArea, enableFrame will be called to and a callback waiting for the subroot's data; this auto load occurs as soon as a webView is seen during a tree update and results in a childrenChanged event once its child rootWebArea is available. When a caller asks for the parent of a rootWebArea, a dynamic lookup is also performed to retrieve a hosting webView, if any.

Review URL: https://codereview.chromium.org/667713006

Cr-Commit-Position: refs/heads/master@{#302648}

BUG=none

Review URL: https://codereview.chromium.org/685793002

Cr-Commit-Position: refs/heads/master@{#302647}

The Google C++ style guide states:

  Explicitly annotate overrides of virtual functions or virtual
  destructors with an override or (less frequently) final specifier.
  Older (pre-C++11) code will use the virtual keyword as an inferior
  alternative annotation. For clarity, use exactly one of override,
  final, or virtual when declaring an override.

To better conform to these guidelines, the following constructs have
been rewritten:

- if a base class has a virtual destructor, then:
    virtual ~Foo();                   ->  ~Foo() override;
- virtual void Foo() override;        ->  void Foo() override;
- virtual void Foo() override final;  ->  void Foo() final;

This patch was automatically generated. The clang plugin can generate
fixit hints, which are suggested edits when it is 100% sure it knows how
to fix a problem. The hints from the clang plugin were applied to the
source tree using the tool in https://codereview.chromium.org/598073004.

BUG=417463
R=jochen@chromium.org

Review URL: https://codereview.chromium.org/678093002

Cr-Commit-Position: refs/heads/master@{#302646}

BUG=424387

Review URL: https://codereview.chromium.org/692723002

Cr-Commit-Position: refs/heads/master@{#302645}

BUG=427584

Review URL: https://codereview.chromium.org/693893004

Cr-Commit-Position: refs/heads/master@{#302644}

BUG=415774

Review URL: https://codereview.chromium.org/699963002

Cr-Commit-Position: refs/heads/master@{#302643}

Review URL: https://codereview.chromium.org/699183002

Cr-Commit-Position: refs/heads/master@{#302642}

This adds several performance tests for CSS Animations
initiated in various ways.

BUG=421733

Review URL: https://codereview.chromium.org/665833002

Cr-Commit-Position: refs/heads/master@{#302641}

When decoder is draining, the current access unit is pointing to the config change AU.
If something interrupts the draining, a surface change for example,
CreateMediaCodecBridge() will fail because the current AU is not an i-frame.
This will cause an unnecessary browser seek.
To avoid this, we should call OnDecoderDrained() first to let video decoder know that config change AU is consumed.

The issue is captured by the DecoderDrainInterruptedBySurfaceChange test on nexus 9.
And this change reenables that test, so no additional test is added.

BUG=429889

Review URL: https://codereview.chromium.org/699763002

Cr-Commit-Position: refs/heads/master@{#302640}

This CL changes the account reconcilor to use the new cookie store
API that allows clients to adds callback for cookie changed events.
It adds logic to the chrome sign-in client to register for cookie
changed events on the IO thread and call the cookie changed callback
on the main thread.

Note that in order to access the cookie store, we need to use the
URL context which can only be used on the IO thread. Subscribing for
cookie change events requires thus a thread jump to the IO thread on
all platforms.

BUG=NONE

Review URL: https://codereview.chromium.org/695553002

Cr-Commit-Position: refs/heads/master@{#302639}

Review URL: https://codereview.chromium.org/698293002

Cr-Commit-Position: refs/heads/master@{#302638}

This patch makes more explicit the set of dependencies
allowed from //extensions onto //ui.

//extensions was allowing dependencies on all of //ui
by default. This is unfortunate for people who want to
introduce new code into //extensions only to be shot
down because they depend on pieces of ui that aren't
allowed.

BUG=None

Review URL: https://codereview.chromium.org/702513002

Cr-Commit-Position: refs/heads/master@{#302637}

This patch ensures that we can the raster scale to min contents scale.
We already did this, but later could modify the value again to be lower
than MinContentsScale. Now, the clamp happens near the end of the
function.

R=danakj
BUG=429397

Review URL: https://codereview.chromium.org/699053003

Cr-Commit-Position: refs/heads/master@{#302636}

Currently _refresh handler caches data uses tip of tree
and caches data under the remote's master commit ID if
run without a 'commit' paramter.

This CL changes the handler to use the last cached 'master'
commit ID instead, so that refresh tasks for non-chromium-git
data can be run between refresh cycles.

BUG=None
R=kalman@chromium.org
NOTRY=True

Review URL: https://codereview.chromium.org/684113003

Cr-Commit-Position: refs/heads/master@{#302635}

TEST=components_unittests --gtest_filter=BookmarkIndexTest.ChangeURL
R=sky@chromium.org
BUG=429042

Review URL: https://codereview.chromium.org/688473005

Cr-Commit-Position: refs/heads/master@{#302634}

In an effort to remove rasterize_time from the impl stats, I'm using
commit_to_activate_duration in its place, which should give a better
indication of the commit to activation time (raster + management time).

BUG=

Review URL: https://codereview.chromium.org/689503002

Cr-Commit-Position: refs/heads/master@{#302633}

net: remove matteomarescotti.name from the HSTS list by request.
Cr-Commit-Position: refs/heads/master@{#302632}

Remove unnecessary #includes from //components/keyed_service/content
after recent refactoring.

BUG=419366

Review URL: https://codereview.chromium.org/701463003

Cr-Commit-Position: refs/heads/master@{#302631}

subprocess.call() is documented as accepting the same arguments as
subprocess.Popen(). This lets us simplify some usages.

NOTRY=true

Review URL: https://codereview.chromium.org/698103003

Cr-Commit-Position: refs/heads/master@{#302630}

BUG=429300

Review URL: https://codereview.chromium.org/695073002

Cr-Commit-Position: refs/heads/master@{#302629}

First attempt in https://codereview.chromium.org/701853003/ does not update Linux 64bit.

The WebRtc roll in VP9 dependency https://codereview.chromium.org/703443004/.
WebRtc submit: https://code.google.com/p/webrtc/source/detail?r=7588

"linux-release/sizes/chrome-bss/bss": {"reva": 302608, "revb": 302608, "type":
     "absolute", "better": "lower", "improve": 501584, "regress": 554384, "sha1": "63
     34ce61"},

BUG=424235
TBR=andrewhayden@chromium.org,tommi@chromium.org

Review URL: https://codereview.chromium.org/698313002

Cr-Commit-Position: refs/heads/master@{#302628}

Picture layer impl has been changed to use pile tiling size as layer
bounds while creating picture layer tilings. Therefore, perftests should
not create inifinite piles as they result in picture layer impls
creating infinite tilings which exhaust memory.

BUG=428844

Review URL: https://codereview.chromium.org/684273003

Cr-Commit-Position: refs/heads/master@{#302627}

BUG=430001
TBR=perkj@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/700913002

Cr-Commit-Position: refs/heads/master@{#302626}

BUG=None
TEST=None
NOTRY=True

Review URL: https://codereview.chromium.org/699213002

Cr-Commit-Position: refs/heads/master@{#302625}

This is accomplished by the introduction of the DownloadMetadataManager,
which maintains state about the most recent binary download.

BUG=389123, 386915, 389643

Review URL: https://codereview.chromium.org/663023007

Cr-Commit-Position: refs/heads/master@{#302624}

BUG=430096
TBR=nasko@chromium.org for render_widget_host_view_guest.*
TBR=kenrb@chromium.org for browser_plugin_messages.h
TBR=lazyboy@chromium.org for overall.

Review URL: https://codereview.chromium.org/690263003

Cr-Commit-Position: refs/heads/master@{#302623}

During the GPU Channel set up, we queue Cursor related IPC calls and
calls related to Widget delegate creation is done only after the channel
is established. This has a effect that the cursor related IPC calls are
handled first and then calls related to Widget delegate creation. However
ResetCursor expects WindowDelegate for the widget to be already
present. Now, we force the creation of WidgetDelegate in OnCursorSet.

This was causing a crash during startup.

Review URL: https://codereview.chromium.org/698793002

Cr-Commit-Position: refs/heads/master@{#302622}

This test was using the wrong mtime as its reference for freshness,
and the index file directory was being created in the wrong order on
creation, which tended to make the first load of an index appear
stale, which was particularly pernicious for a test.

R=jkarlin@chromium.org
BUG=255775

Review URL: https://codereview.chromium.org/705433002

Cr-Commit-Position: refs/heads/master@{#302621}

The jank almost disappeared, either because it was fixed, or its numbers were absorbed by other instrumentations.
At the current moment in Win, instrumentation for OAuth2AccessTokenFetcherImpl::OnURLFetchComplete shown 1.65 118ms jph, and remaining own jankiness of URLFetcherCore::OnCompletedURLRequest is 2.87.

BUG=422577
TBR=toyoshim@chromium.org, dconnelly@chromium.org, pkasting@chromium.org, zea@chromium.org, yoz@chromium.org, mmenke@chromium.org, dcheng@chromium.org, mattm@chromium.org, jhawkins@chromium.org, rogerta@chromium.org

Review URL: https://codereview.chromium.org/701463004

Cr-Commit-Position: refs/heads/master@{#302620}

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/4ab22e3..7a10fb6

CQ_EXTRA_TRYBOTS=tryserver.blink:linux_blink_rel,linux_blink_dbg
TBR=egdaniel@google.com

Review URL: https://codereview.chromium.org/702713002

Cr-Commit-Position: refs/heads/master@{#302619}

R=aa@chromium.org

Review URL: https://codereview.chromium.org/699123003

Cr-Commit-Position: refs/heads/master@{#302618}

BUG=chromium:400619

Review URL: https://codereview.chromium.org/657023008

Cr-Commit-Position: refs/heads/master@{#302617}

The WebRtc roll in VP9 dependency https://codereview.chromium.org/703443004/.
WebRtc submit: https://code.google.com/p/webrtc/source/detail?r=7588

"linux-release/sizes/chrome-bss/bss": {"reva": 302608, "revb": 302608, "type":
     "absolute", "better": "lower", "improve": 501584, "regress": 554384, "sha1": "63
     34ce61"},

BUG=424235
R=tommi@chromium.org
TBR=andrewhayden@chromium.org

Review URL: https://codereview.chromium.org/701853003

Cr-Commit-Position: refs/heads/master@{#302616}

Benefits of this refactoring are:
1. Removed a redundant adapter.
2. Removed an unnecessary interface conversion (A->B->A).
3. Removed from ClientSessionTestingHost references to unrelated class LocalSessionBridge.
4. Reduced number of public classes.

BUG=383418

Review URL: https://codereview.chromium.org/700833002

Cr-Commit-Position: refs/heads/master@{#302615}

Replace the hardcoded version number used for the WebView's user agent
with the real version number of the corresponding Chrome build. This
uses the version information from src/chrome/VERSION; there is a pending
bug to move this out of chrome/ (http://crbug.com/345538) which will not
be impeded by this change as the location of the version info is
obtained from build/util/version.gypi.

BUG=297522

Review URL: https://codereview.chromium.org/695313002

Cr-Commit-Position: refs/heads/master@{#302614}

Add a new gyp variable: use_libpci (default value is 1).
Add a new define when use_libpci==1: USE_LIBPCI.
Don't call specific PCI code in gpu_info_collector_linux.cc if USE_LIBPCI isn't defined.

BUG=427401
R=piman@chromium.org

Review URL: https://codereview.chromium.org/692893004

Cr-Commit-Position: refs/heads/master@{#302613}

Blink roll dcb6aa9bef3b496069e234ed4dc52d1e8c636776: 80d183d0f13494ea1c51686c688d0f4d4e81aa81 (svn 184812:184821)

TBR=dcheng

Review URL: https://codereview.chromium.org/699133002

Cr-Commit-Position: refs/heads/master@{#302612}

BUG=401729

Review URL: https://codereview.chromium.org/701603002

Cr-Commit-Position: refs/heads/master@{#302611}

MSAA+IA2
ROLE_SYSTEM_EQUATION

AX:
AXGroup
AXDocumentMath
'math'

Expectations for math.html is skipped in this patch, since it need a blink side patch too.

BUG=429476

Review URL: https://codereview.chromium.org/695133002

Cr-Commit-Position: refs/heads/master@{#302610}