GitLab

Since upgrading the builder to 10.8, it never failed again.

BUG=370551
R=thakis@chromium.org

Review URL: https://codereview.chromium.org/401103006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284668 0039d316-1c4b-4281-b951-d872f2087c98

The ProfileManager can call its callback (which points to CreateProfileHandler::OnProfileCreated) with a NULL profile (https://code.google.com/p/chromium/codesearch#chromium/src/chrome/browser/profiles/profile_manager.cc&l=397). This case was not handled in CreateProfileHandler::ShowProfileCreationError.

BUG=395558

Review URL: https://codereview.chromium.org/405883003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284667 0039d316-1c4b-4281-b951-d872f2087c98

BUG=387738

Review URL: https://codereview.chromium.org/395133002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284666 0039d316-1c4b-4281-b951-d872f2087c98

It is flakey on win bots.
http://build.chromium.org/p/chromium.win/builders/Win%207%20Tests%20x64%20%283%29/builds/15075
http://build.chromium.org/p/chromium.win/builders/Win7%20Tests%20%282%29/builds/30300

TBR=achuith@chromium.org
NOTRY=true
BUG=394454

Review URL: https://codereview.chromium.org/406233002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284665 0039d316-1c4b-4281-b951-d872f2087c98

Reason for revert:
This patch seems to make following browser_tests flakey

PPAPINaClNewlibTest.Graphics2D_FlushOffscreenUpdate
NetInternalsTest.netInternalsHSTSViewAddOverwrite
NetInternalsTest.netInternalsHSTSViewAddDelete
NetInternalsTest.netInternalsHSTSViewAddTwice

http://build.chromium.org/p/chromium.win/builders/XP%20Tests%20%282%29/builds/34734
http://build.chromium.org/p/chromium.win/builders/XP%20Tests%20%281%29/builds/32086
http://build.chromium.org/p/chromium.win/builders/Vista%20Tests%20%281%29/builds/47545


Original issue's description:
> This is jamesr@ code I am landing.
> 
> On Windows the message pump code tried to manage the systemwide timer resolution to fire delayed tasks with better than 15ms resolution but it was buggy:
> 
> 1- A short task that was not followed by any other task will leave the systemwide timer pegged to 1ms
> 
> 2- After we decided to crank up the timer we would 'lease' the timer for 1 second, for no good reason.
> 
> Both issues are detrimental to battery power.
> 
> The source of both problems is that we tried to decide with incomplete information. This patch solves that by having 1 bit for each pending task that requires a high resolution timer and a sum of the number of tasks that require high res timers.
> 
> BUG=153139
> TEST=included here, also see the bug for manual testing.
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=284625

TBR=jamesr@chromium.org,darin@chromium.org,cpu@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=153139

Review URL: https://codereview.chromium.org/407073004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284664 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284663 0039d316-1c4b-4281-b951-d872f2087c98

Please reply to the V8 sheriff svenpanne@chromium.org in case of problems.

TBR=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/412433002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284662 0039d316-1c4b-4281-b951-d872f2087c98

The blink roll contains a fix to the problem, but is stale at the moment.

BUG=395882
TBR=jochen@chromium.org
NOTRY=true
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/405413002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284661 0039d316-1c4b-4281-b951-d872f2087c98

BUG=

Review URL: https://codereview.chromium.org/404913002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284660 0039d316-1c4b-4281-b951-d872f2087c98

The format of dumpsys batterystats has changed. This CL updates the
parsing.

Review URL: https://codereview.chromium.org/402153002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284659 0039d316-1c4b-4281-b951-d872f2087c98

VideoFrame based on StreamTexture is reused on many frames. However, clients (e.g.
compositor, WebGL) inserts a sync point every frame. So VideoFrame on Android
can keep several mega byte sync points unnecessarily.

This CL makes VideoFrame keep only one sync point. When a client set new sync point
to VideoFrame and VideoFrame already keeps the sync point of another client, VideoFrame
waits the previous sync point before inserting a new sync point. A client must provide
a VideoFrame::SyncPointProvider so that VideoFrame can insert or wait a sync point.

BUG=350925

Review URL: https://codereview.chromium.org/312803002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284658 0039d316-1c4b-4281-b951-d872f2087c98

The sandbox policy is too restrictive.

BUG=395578
R=mseaborn@chromium.org

Review URL: https://codereview.chromium.org/408813003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284656 0039d316-1c4b-4281-b951-d872f2087c98

BUG=389319

Review URL: https://codereview.chromium.org/404273002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284655 0039d316-1c4b-4281-b951-d872f2087c98

In Gallery, the size of the slide image's canvas is scaled by
window.devicePixelRatio. For example, if the devicePixelRatio is 2, the doubled
size canvas is prepared and it is scaled to displaying size so that the pixels
of canvas fit to the device pixels.

Previously the canvas is scaled by using CSS transformation. Thus the element
position is calculated by the original canvas size, and then it is scaled. This
makes the position calculation of canvas complex, and it was regressed at the
crrev.com/398263002.

At this CL, canvas is scaled by using CSS width/height property instead of
bringing back the complex position calculation.

BUG=395062
TEST=manually

Review URL: https://codereview.chromium.org/402863002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284654 0039d316-1c4b-4281-b951-d872f2087c98

The unit test failure indicates a serious Document leak, where all
WebCore::Document loaded in Chrome is leaking.
This CL adds warning comments to the baseline to avoid regressions.

BUG=392121
NOTRY=true

Review URL: https://codereview.chromium.org/393123003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284653 0039d316-1c4b-4281-b951-d872f2087c98

Increase performance of RenderPass drawing in compositor and Video-WebGL copy.

BUG=375501

Review URL: https://codereview.chromium.org/371463009

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284652 0039d316-1c4b-4281-b951-d872f2087c98

In bot mode, test_launcher would normally retry tests up to 3 times.
This is sufficient to keep large test suites like browser_tests under
1% flakiness rate if the flakiness of individual tests is also within
1%. However, some tests especially in browser_tests are up to 33%
flaky, making the browser_tests suite 40% flaky.

Subsequent top-level retries effectively double the runtime of all
bots running these tests, and we need a solution to shorten the try
job runtime.

A longer-term solution is the ignorer bot (see the bug), but until
it's deployed, a quick hack is to increase the number of retries so
statistically we should see about 1% flakiness of browser_tests at the
expense of a slight runtime increase.

R=sergiyb@chromium.org
BUG=390600, 395189

Review URL: https://codereview.chromium.org/393283006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284650 0039d316-1c4b-4281-b951-d872f2087c98

It currently just uses 1.0. Instead, fetch it with
ui::GetScaleFactorForNativeView and invalidate it when the scale factor
changes.

This is a small fix for a simple HiDPI bug, but eventually we will be
able to scrap CachedLabel: The performance improvement it offers will be
redundant with http://crbug.com/240037.

BUG=368493

Review URL: https://codereview.chromium.org/405373002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284649 0039d316-1c4b-4281-b951-d872f2087c98

Originally, AppListDialogContentsView extended WidgetDelegate directly,
and was not being detected as a dialog correctly on some platforms.
Updated AppListDialogContentsView to now be a DialogDelegate.

BUG=388637

Review URL: https://codereview.chromium.org/405873003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284648 0039d316-1c4b-4281-b951-d872f2087c98

This matches current definition in the design doc[1] and unifies phase name between Chromium and Blink. Telemetry already expects both "i" and "I" values [2].

It used to be "I" in Chromium but it was changed in https://chromiumcodereview.appspot.com/12252058/

[1] https://docs.google.com/a/chromium.org/document/d/1CvAClvFfyA5R-PhYUmn5OOQtYMH4h6I0nSsKchNAySU/edit#
[2] https://code.google.com/p/chromium/codesearch#chromium/src/tools/telemetry/telemetry/timeline/trace_event_importer.py&q=%22elif%20phase%20==%20'I'%20or%20phase%20==%20'i':%22&sq=package:chromium&type=cs&l=252

BUG=None

Review URL: https://codereview.chromium.org/399423005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284644 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284643 0039d316-1c4b-4281-b951-d872f2087c98

Revert of Revert of Update V8 to version 3.28.30 (based on bleeding_edge revision r22489). (https://codereview.chromium.org/405313002/)

Reason for revert:
See https://code.google.com/p/chromium/issues/detail?id=395882

The reason for the test failures in blink has been fixed with the last blink roll.

Original issue's description:
> Revert of Update V8 to version 3.28.30 (based on bleeding_edge revision r22489). (https://codereview.chromium.org/407733005/)
> 
> Reason for revert:
> Seems to be causing xss-DENIED-method-with-iframe-proto.html to time out on multiple bots; see http://crbug.com/395882 .
> 
> Original issue's description:
> > Update V8 to version 3.28.30 (based on bleeding_edge revision r22489).
> > 
> > Please reply to the V8 sheriff svenpanne@chromium.org in case of problems.
> > 
> > TBR=svenpanne@chromium.org
> > 
> > Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=284420
> 
> TBR=svenpanne@chromium.org,machenbach@chromium.org
> NOTREECHECKS=true
> NOTRY=true
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=284572

TBR=svenpanne@chromium.org,pkasting@chromium.org
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/411453003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284642 0039d316-1c4b-4281-b951-d872f2087c98

Also removes the unused tpm_token_name_ member from TPMTokenLoader.

BUG=NONE
R=stevenjb@chromium.org

Review URL: https://codereview.chromium.org/408853004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284641 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284640 0039d316-1c4b-4281-b951-d872f2087c98

'WebURLRequest::TargetType' has been dropped in Blink in favor of
'RequestContext' and 'FrameType', matching the Fetch specification's
changes, and allowing both more granularity in various Blink-side checks
and more clarity around the thing that caused a request, and the frame
that the request affects.

This CL should have zero practical effect, as it's simply hoisting the
translation work that 'WebURLRequest::setTargetType' is already doing
up into the content layer to bring us one step closer to removing the
concept entirely.

BUG=390497

Review URL: https://codereview.chromium.org/409583002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284639 0039d316-1c4b-4281-b951-d872f2087c98

When an X.509 certificate appears invalid because it is expired or not yet 
valid, but we have reason to believe the clock is wrong, help users find out 
how to fix their clock.

A future CL will run a check for the time at startup (and periodically
after?) and raise an infobar with similar functionality, as well.

BUG=349653

Review URL: https://codereview.chromium.org/397713007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284638 0039d316-1c4b-4281-b951-d872f2087c98

The api will be used by Easy Unlock component app. Currently it contains
methods needed for establishing secure communication channel on top of
(unprotected) bluetooth.

BUG=390230

Review URL: https://codereview.chromium.org/391513002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284637 0039d316-1c4b-4281-b951-d872f2087c98

BUG=395468

Review URL: https://codereview.chromium.org/409543005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284636 0039d316-1c4b-4281-b951-d872f2087c98

BUG=394273

Review URL: https://codereview.chromium.org/399003002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284635 0039d316-1c4b-4281-b951-d872f2087c98

BUG=

Review URL: https://codereview.chromium.org/401053002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284634 0039d316-1c4b-4281-b951-d872f2087c98

1. android uses stlport in ndk library,
   and this library does not have same demangled name with libc++ stl.
   so we can get rid of tc-std-* from android policies.
2. tc-stl-node and tc-stl-vector have same stackstrace.
   so I changed std-node to std-other to hook all other std mallocs.
3. tc-toplevel-string regex has changed.

Bug=None
NOTRY=True

Review URL: https://codereview.chromium.org/387313003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284633 0039d316-1c4b-4281-b951-d872f2087c98

It needs to depend on the allocator gyp target to pick up the
/NODEFAULTLIBS flags for libcmt.lib and libcmtd.lib and avoid
link errors about multiple definitions of allocation related
symbols.

BUG=
TEST=extensions_unittests links in non-components Debug build on Win

Review URL: https://codereview.chromium.org/394353008

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284632 0039d316-1c4b-4281-b951-d872f2087c98

The file names now match the class name.

BUG=none
TEST=none
R=estade@chromium.org

Review URL: https://codereview.chromium.org/404553008

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284631 0039d316-1c4b-4281-b951-d872f2087c98

VisualStudio does not understand it \u escape sequences like \u00E8 in
UTF8 strings. The fix is to use Wide strings and convert them into UTF8
using base::WideToUTF8() utility.

BUG=395401

Review URL: https://codereview.chromium.org/403233002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284629 0039d316-1c4b-4281-b951-d872f2087c98

BUG=None
TEST=Run the vda_unitest on GLX and EGL machines.

Review URL: https://codereview.chromium.org/400353007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284628 0039d316-1c4b-4281-b951-d872f2087c98

[ServiceWorker] Propagates ServiceWorker fetched response's URL and wasFetchedViaServiceWorker flag.

If the request was fetched via a ServiceWorker, the URL of the response could be different from the URL of the original request. 
The URL must be checked in the renderer process for CSP and CORS.

So in this patch I introduce was_fetched_via_service_worker flag and original_url_via_service_worker which are set in ServiceWorkerURLRequestJob and propagated to the renderer process.


This change depends on http://crrev.com/370733002 and http://crrev.com/378473002.

BUG=373120

Review URL: https://codereview.chromium.org/375513002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284627 0039d316-1c4b-4281-b951-d872f2087c98

Revert of Disable exceptions on Windows also in the shared_library build (https://codereview.chromium.org/404853002/)

Reason for revert:
Speculative revert to try to fix chrome_elf_unittests on Win7 x64.
http://build.chromium.org/p/chromium.win/builders/Win%207%20Tests%20x64%20%282%29/builds/18370
  AddDllsFromRegistryToBlacklist
  LoadBlacklistedLibrary

I'll reland if that's not it.


Original issue's description:
> Disable exceptions on Windows also in the shared_library build
> 
> They were previously only disabled in static_library builds. This is believed
> to have been due to problems with old MSVC versions, but it should work now.
> 
> This change is relevant for Clang, which currently doesn't support exceptions
> on Windows.
> 
> Last time I attempted to do this, it broke some tests due to a debug assertion.
> Gtest had previously been catching the exception, but with exceptions disabled
> it crashed. Those tests were actually broken in static_library Debug builds
> too, but we don't seem to have bots for that. After Blink r176189, we no longer
> hit that debug assertion, so the tests now pass.
> 
> BUG=82385
> TEST=blink_platform_unittests --gtest_filter=DateTimeFormatTest.CommonPattern,
>      content_browsertests --gtest_filter=RenderViewImplTest.SetEditableSelectionAndComposition
>      webkit_unit_tests
> 
> R=thakis@chromium.org
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=284519

TBR=maruel@chromium.org,thakis@chromium.org,hans@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=82385

Review URL: https://codereview.chromium.org/410613002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284626 0039d316-1c4b-4281-b951-d872f2087c98

On Windows the message pump code tried to manage the systemwide timer resolution to fire delayed tasks with better than 15ms resolution but it was buggy:

1- A short task that was not followed by any other task will leave the systemwide timer pegged to 1ms

2- After we decided to crank up the timer we would 'lease' the timer for 1 second, for no good reason.

Both issues are detrimental to battery power.

The source of both problems is that we tried to decide with incomplete information. This patch solves that by having 1 bit for each pending task that requires a high resolution timer and a sum of the number of tasks that require high res timers.

BUG=153139
TEST=included here, also see the bug for manual testing.

Review URL: https://codereview.chromium.org/395913006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284625 0039d316-1c4b-4281-b951-d872f2087c98

Subsequent calls to EmbedRoot are now delegated to the window manager. This allows an application that was not created via the embed flow to obtain a connection to the view manager by asking the view manager init service to embed it somewhere. The init service just forwards this to the window manager. This is effectively like asking the window manager to create a toplevel window.

R=sky@chromium.org
BUG=none

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=284322

Review URL: https://codereview.chromium.org/403083002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284624 0039d316-1c4b-4281-b951-d872f2087c98

invalidation_util.cc has the following include:

invalidation_util.cc is in the target
sync_internal_api. sync_internal_api doesn't depend on
cacheinvalidation_proto_cpp which is what generates types.pb.h.

BUG=395846
TEST=none
R=tim@chromium.org

Review URL: https://codereview.chromium.org/410503002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@284623 0039d316-1c4b-4281-b951-d872f2087c98