GitLab

> CHECK for adding duplicate converter inputs.
>
> CHECK to see if we're getting into weird states due to inputs
> being added twice.
>
> BUG=233026
> TEST=none
>
> Review URL: https://chromiumcodereview.appspot.com/14083011

TBR=scherkus
NOTRY=true

Review URL: https://codereview.chromium.org/100583002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238239 0039d316-1c4b-4281-b951-d872f2087c98

BUG=
R=scottmg@chromium.org

Review URL: https://codereview.chromium.org/99783006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238238 0039d316-1c4b-4281-b951-d872f2087c98

vpx_video_decoder does not allow odd videos with odd height/width
to playback since it is not valid YV12. But ffmpeg currently
supports it by merely rounding up (and so does the libvpx library).
Changing vpx_video_decoder to behave the same way as ffmpeg and
getting rid of the Checks.

BUG=315817
TEST=media_unittests

Review URL: https://codereview.chromium.org/92073002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238237 0039d316-1c4b-4281-b951-d872f2087c98

Allows a tab dragged out of a maximized browser to get properly side-snapped (similar to how docking is allowed).

BUG=319205

Review URL: https://codereview.chromium.org/80323002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238236 0039d316-1c4b-4281-b951-d872f2087c98

There's currently a #include cycle between view.h and border.h

Also removes some redundant forward-declares in view.h

BUG=None

Review URL: https://codereview.chromium.org/93843002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238235 0039d316-1c4b-4281-b951-d872f2087c98

This reverts commit 6aa58b85.

Fixed double-close issue when ending recording.

> Revert 237280 "Remove TraceController"
>
> Seems to have broken trace-based telemetry benchmarks on android.
>
> BUG=323749
>
> > Remove TraceController
> >
> > TraceController is obsoleted by TracingController.
> > Changed all remaining clients to use TracingController.
> >
> > BUG=none
> >
> > Review URL: https://codereview.chromium.org/67683003
>
> TBR=wangxianzhu@chromium.org
>
> Review URL: https://codereview.chromium.org/89753004
>
> git-svn-id: svn://svn.chromium.org/chrome/trunk/src@237452 0039d316-1c4b-4281-b951-d872f2087c98

TBR=wangxianzhu@chromium.org

Review URL: https://codereview.chromium.org/99103004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238234 0039d316-1c4b-4281-b951-d872f2087c98

When the Chrome hotdog menu button is focused by pressing Alt-Shift-T or F6 on
Windows/ChromeOS and tabbing over to itr, show the rectangular focus indicator.

BUG=320759

Review URL: https://codereview.chromium.org/83513004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238233 0039d316-1c4b-4281-b951-d872f2087c98

BUG=324158

Review URL: https://codereview.chromium.org/93613002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238231 0039d316-1c4b-4281-b951-d872f2087c98

The "increase preroll capacity on underflow" logic was disabled for MSE
playbacks a while ago because it was causing unwanted garbage
collections and was allowing the capacity to grow too large. This change
restores the capacity increase logic, but restricts the capacity so that
it never grows beyond 3 seconds worth of data. This will avoid the GC
issues encountered before.

This change also resets the capacity on flush so that capacity increases
in one playback period don't negatively effect the seek start time.


BUG=291726
TEST=AudioRendererImplTest.Underflow, AudioRendererImplTest.Underflow_FollowedByFlush

Review URL: https://codereview.chromium.org/88743004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238230 0039d316-1c4b-4281-b951-d872f2087c98

BUG=323727

Review URL: https://codereview.chromium.org/89773002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238228 0039d316-1c4b-4281-b951-d872f2087c98

The test puts a mailbox into the compositor tree and then makes sure
only one commit is scheduled.

Right now the test does this by just doing EndTest() and verifying
that a 2nd commit doesn't happen before the test exits. This causes
a crash though when the mailbox is destroyed before being cleaned
up properly, and then the release callback runs on a bad test
pointer.

Instead, we can have the test make use of CommitPendingForTesting()
to verify another commit was not scheduled, then have it remove the
mailbox from the tree as other tests do.

R=enne
BUG=317854

Review URL: https://codereview.chromium.org/99913003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238227 0039d316-1c4b-4281-b951-d872f2087c98

This fixes a crash when navigating to an URL containing an
internationalized domain name.

BUG=324855
R=robertshield@chromium.org

Review URL: https://codereview.chromium.org/99623003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238226 0039d316-1c4b-4281-b951-d872f2087c98

Review URL: https://codereview.chromium.org/99453002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238225 0039d316-1c4b-4281-b951-d872f2087c98

joth out
Some new folks in

TBR=darin@chromium.org

Review URL: https://codereview.chromium.org/99653002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238224 0039d316-1c4b-4281-b951-d872f2087c98

The in-process command buffer expects all of its callers to be on a single
thread. Prior to this CL, we were bouncing back to the shell thread to call
swap buffers. Now we call SwapBuffers from sample_app's thread.

R=davemoore@chromium.org

Review URL: https://codereview.chromium.org/99323005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238223 0039d316-1c4b-4281-b951-d872f2087c98

BUG=324353

Review URL: https://codereview.chromium.org/95713002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238222 0039d316-1c4b-4281-b951-d872f2087c98

defines the entire site chrome, rather than spreading it over several files.

BUG=320339
R=mkearney@chromium.org, yoz@chromium.org

Review URL: https://codereview.chromium.org/92143003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238221 0039d316-1c4b-4281-b951-d872f2087c98

Forgot to delete this function in http://crrev.com/73463006.

BUG=

Review URL: https://codereview.chromium.org/90153003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238219 0039d316-1c4b-4281-b951-d872f2087c98

R=pkasting@chromium.org

Review URL: https://codereview.chromium.org/62873007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238218 0039d316-1c4b-4281-b951-d872f2087c98

When cached timestamps are available, and toolchain is up-to-date,
reduces the runtime to ~.2s.

R=maruel@chromium.org
TBR=cpu@chromium.org
BUG=323300

Review URL: https://codereview.chromium.org/95983002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238217 0039d316-1c4b-4281-b951-d872f2087c98

This change simplifies some functions, removes some dead code and
fixes up pylint style violations.

BUG=

Review URL: https://codereview.chromium.org/98803004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238216 0039d316-1c4b-4281-b951-d872f2087c98

This is a step forward on removing ui_cocoa_third_party_toolkits target
from ui.gyp.

BUG=299841
TEST=None
R=tapted@chromium.org,ben@chromium.org

Review URL: https://codereview.chromium.org/83073012

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238215 0039d316-1c4b-4281-b951-d872f2087c98

BUG=

Review URL: https://codereview.chromium.org/99493002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238214 0039d316-1c4b-4281-b951-d872f2087c98

Remove BitmapPlatformDeviceData, and merge its functionality into the appropriate BitmapPlatformDevice. It was a little twisty maze of #ifdefs, ostensibly created to allow successful assignment and copy construction of the device. However, I could not find any such use, and the BitmapPlatformDevices are all marked DISALLOW_COPY_AND_ASSIGN!

Instead, we use a subclass of SkPixelRef to handle ownership of the platform-specific bitmaps. This allows correct Skia semantics: the SkBitmap used for device drawing can safely outlive the device, since it is refcounted. Such a subclass already existed for Windows; this simply implements it for Cairo as well, and uses it in all cases.

TBR=brettw

Review URL: https://codereview.chromium.org/95773002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238213 0039d316-1c4b-4281-b951-d872f2087c98

Changes sequence of docked animations when evicting windows from dock. Windows that are restored into docked area are slid from below rather than from above as before. Additionally minimizing animation for the windows that no longer fit in the dock is slowed down in hopes of making it less confusing.
To allow this a new ScopedLayerAnimationSettings::LockTransitionDuration method is introduced which locks the current animation transition duration until the ScopedLayerAnimationSettings object that invoked it goes out of scope.

BUG=323188
TEST=ash_unittests --gtest_filter=WindowAnimationsTest.LockAnimationDuration

Review URL: https://codereview.chromium.org/82573002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238212 0039d316-1c4b-4281-b951-d872f2087c98

The TODO was relevant for the SetProperty() call, which was removed. So the TODO
should be removed too.

BUG=none
TBR=sky@chromium.org

Review URL: https://codereview.chromium.org/100333002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238211 0039d316-1c4b-4281-b951-d872f2087c98

BUG=None

Review URL: https://codereview.chromium.org/54043003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238210 0039d316-1c4b-4281-b951-d872f2087c98

base64.b64encode not base64.encode because the latter inserts a \n, and add a
User-Agent header.

R=yoz@chromium.org

Review URL: https://codereview.chromium.org/98183003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238209 0039d316-1c4b-4281-b951-d872f2087c98

This would deadlock in some error cases when the failing thread tried to acquire the lock to report an error, but the main thread was in the process of shutting down (due to a previous error) and blocking on the worker threads finishing while holding thc lock.

This moves the worker pool shutdown to outside the lock, and adds some extra shutdown assertions.

It also fixes a syntax error in the base Linux build file.

BUG=
R=scottmg@chromium.org

Review URL: https://codereview.chromium.org/100193002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238207 0039d316-1c4b-4281-b951-d872f2087c98

BUG=None
TBR=craigdh@chromium.org

Review URL: https://codereview.chromium.org/100323002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238206 0039d316-1c4b-4281-b951-d872f2087c98

It is possible to copy a key-event using the copy-constructor, without needing
a separate method for it. So remove the redundant code for copying a KeyEvent.

BUG=none
R=estade@chromium.org, sky@chromium.org

Review URL: https://codereview.chromium.org/97963002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238205 0039d316-1c4b-4281-b951-d872f2087c98

* Minor indenting/wrapping/comment changes
* Remove misplaced standalone "#" (how did this compile?)
* Make definition order match declaration order
* Style guide: Consts come before constructors
* Eliminate unnecessary #if
* Pass MIDIPermissionInfoBarDelegate the necessary |contents_unique_id|; this
  matches other infobars and will be necessary after the refactor, since the
  InfoBarService will no longer be provided
* Explicitly declare destructors
* MIDIPermissionInfoBarDelegate constructor need not be protected
* EXPECT_TRUE(false) -> ADD_FAILURE()
* Eliminate unnecessary using directives
* Bail out of TranslateInfoBarDelegate::Create() earlier to avoid unnecessary
  infobar creation
* Eliminate unnecessary temp
* BrowserWindowControllerTest can just use a SimpleAlertInfoBar instead of
  defining its own custom infobar

BUG=none
TEST=none
R=miguelg@chromium.org, sky@chromium.org

Review URL: https://codereview.chromium.org/89443002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238203 0039d316-1c4b-4281-b951-d872f2087c98

Soon the patch to boost AES-GCM cipher suites when AES-NI is supported will
land. Also, Google frontend servers will be configured with an equal
preference for ChaCha20 and AES-GCM, meaning that they will respect the
client's preference between those two families of ciphers.

Thus this change enables ChaCha20. Initially these cipher suites will be used
for all HTTPS connections to Google until the equal preference is configured.
At that point they'll only be used when hardware AES-GCM support isn't enabled.

BUG=310768

Review URL: https://codereview.chromium.org/92703002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238202 0039d316-1c4b-4281-b951-d872f2087c98

R=maruel@chromium.org, thakis@chromium.org

Review URL: https://codereview.chromium.org/99653004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238199 0039d316-1c4b-4281-b951-d872f2087c98

quic_server and internal QUIC server. Made client side fixes
(per rch's comments in https://codereview.chromium.org/97433002/)

R=rch@chromium.org

Review URL: https://codereview.chromium.org/99203003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238198 0039d316-1c4b-4281-b951-d872f2087c98

Hebrew keyboard shortcuts for 10.9 mirror brackets and parens
differently than 10.8.  Convert to a more general mapping table which
doesn't make assumptions about the codes for the mirrored elements.

BUG=319182

Review URL: https://codereview.chromium.org/78783006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238197 0039d316-1c4b-4281-b951-d872f2087c98

fork chromium. The first patchset below is equivalent to what we have
in KLP branch (minus machine generated makefiles). 

This is to unfork the implementation.

Note that the print messages and print_web_view_helper are same (or
very slightly different) from their counterpart in
chrome/renderer/printing and chrome/common/print_messages. The plan is
to start componentizing printing immediately after and remove the
duplicate code

TBR=danakj@chromium.org
BUG=322276

Review URL: https://codereview.chromium.org/54963005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238196 0039d316-1c4b-4281-b951-d872f2087c98

There is a kConstrainedWindowKey in both ash and aura::client namespace, which is
pretty confusing when reading the code. So remove the ash version.

BUG=none
R=sky@chromium.org

Review URL: https://codereview.chromium.org/97893002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238195 0039d316-1c4b-4281-b951-d872f2087c98

There are some use cases that we don't allow user to close the web dialog until they have finish the UI flow and confirmed some important information. Currently, WebDialogDelegate can't control the closing of the dialog if user presses the "x" button on dialog title bar or presses Esc shortcut key. I am adding a CanClose() method to allow the subclass to block the dialog close if it needs to do so. 

BUG=322614

Review URL: https://codereview.chromium.org/82913007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238193 0039d316-1c4b-4281-b951-d872f2087c98

Doing the 'self' exclusion in the loop was too late.

BUG=

Review URL: https://codereview.chromium.org/94033003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@238192 0039d316-1c4b-4281-b951-d872f2087c98