GitLab

Brings the onFocusChanged handling exactly inline between ContentView and
AwContents.

NOTRY=true
BUG=285582

Review URL: https://chromiumcodereview.appspot.com/23623020

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221482 0039d316-1c4b-4281-b951-d872f2087c98

When pushing data files avoids matching files whose relative paths do not match a file being pushed from the host.

BUG=284906
TEST=None
NOTRY=True

Review URL: https://chromiumcodereview.appspot.com/23513019

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221477 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221475 0039d316-1c4b-4281-b951-d872f2087c98

http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog_blink.html?url=/trunk&range=157170:157296&mode=html
TBR=
BUG=

Review URL: https://codereview.chromium.org/23819022

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221473 0039d316-1c4b-4281-b951-d872f2087c98

BUG=none

Review URL: https://chromiumcodereview.appspot.com/23532033

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221472 0039d316-1c4b-4281-b951-d872f2087c98

BUG=280550
TEST=PipelineIntegrationTest.MediaSource_MP3

Review URL: https://chromiumcodereview.appspot.com/23454006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221471 0039d316-1c4b-4281-b951-d872f2087c98

Also, add NDEBUG to CFLAGS by default so that asserts not
compiled into release builds, and fix resulting unused
variable warnings by making KernelProxy::Init do more
error checking and return a result.

R=binji@chromium.org

Review URL: https://codereview.chromium.org/22842011

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221470 0039d316-1c4b-4281-b951-d872f2087c98

responses with no addresses.

> [net/dns] Perform A/AAAA queries for AF_UNSPEC resolutions in parallel. 
> 
> The second DnsTransaction is scheduled as a second job on the resolver's
> PrioritizedDispatcher, at the beginning of its priority queue. The two
> DnsTransactions run independently of each other, although if one of them
> finishes with an error, the other one will be scrapped immediately, and
> the Job will fall back to ProcTask. 
> 
> BUG=174992
> 
> Original Review URL: https://chromiumcodereview.appspot.com/19498003
Revert URL: https://codereview.chromium.org/23102009

BUG=174992,277625

Review URL: https://chromiumcodereview.appspot.com/22909037

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221469 0039d316-1c4b-4281-b951-d872f2087c98

Disable actual http://groupcloned.com and add test page
http://groupcloned.com/test/list-images-variable/index.html instead.

The actual groupcloned page keeps failing in the key_mobile_sites page set.
It seems to behave non-deterministically, i.e. it's trying to load different
items on different invocations. This is despite Telemetry trying to get rid
of all randomness. The problem with the non-determinism is that the page often
tries to load resources that haven't been loaded during recording.

BUG=238864
R=tonyg@chromium.org

Review URL: https://codereview.chromium.org/23500008

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221468 0039d316-1c4b-4281-b951-d872f2087c98

TEST=tools/perf/run_measurement page_cycler_bloat --v8-object-stats

Review URL: https://chromiumcodereview.appspot.com/23963002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221467 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221466 0039d316-1c4b-4281-b951-d872f2087c98

1715: Tests: Try to set +x bot on new change.sh file, try 2.
1716: ninja&make/mac: Give loadable_modules type MH_BUNDLE.

BUG=280718
TBR=mark@chromium.org

Review URL: https://codereview.chromium.org/23457004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221465 0039d316-1c4b-4281-b951-d872f2087c98

Note that the new code is still behind a runtime flag
(--brave-new-test-launcher), but compiling tests with support for it
will make further testing possible.

BUG=236893, 79359
R=akalin@chromium.org, enne@chromium.org, erikwright@chromium.org, joi@chromium.org, keybuk@chromium.org, sky@chromium.org, thestig@chromium.org, tommi@chromium.org, wtc@chromium.org, xhwang@chromium.org, yzshen@chromium.org

Review URL: https://codereview.chromium.org/23442019

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221464 0039d316-1c4b-4281-b951-d872f2087c98

Ensure the 1993 NTP logs to UMA Most Visited tiles that use the fallback thumbnail instead of the preferred one.

TBR=sky@chromium.org

Review URL: https://chromiumcodereview.appspot.com/23874005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221463 0039d316-1c4b-4281-b951-d872f2087c98

* The bounds obtained from GetBoundsOverrideAsh is ignored
  (overridden) by the following logic if it returns false,
  so skip early for the condition that returns false.
* Removed the code for non tabbed case in GetBoundsOverrideAsh as it won't be used.

* removed fullscreen controller check because it seems to be
  redundant with window->IsFullscreen() check. (IsFullscerenForBrowser() is always false if window->IsFullscreen() is false)

BUG=272460
R=skuhne@chromium.org, sky@chromium.org

Review URL: https://codereview.chromium.org/23567007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221461 0039d316-1c4b-4281-b951-d872f2087c98

Review URL: https://chromiumcodereview.appspot.com/23953002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221460 0039d316-1c4b-4281-b951-d872f2087c98

Also add a little extra flair.

BUG=278904
NOTRY=true

Review URL: https://chromiumcodereview.appspot.com/23611006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221458 0039d316-1c4b-4281-b951-d872f2087c98

Merges them to a single 'variations' target to simplify things.

BUG=none

Review URL: https://chromiumcodereview.appspot.com/23604034

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221457 0039d316-1c4b-4281-b951-d872f2087c98

Apparently under windows requesting focus can trigger capture lost,
which caused all sorts of problems with tab dragging. When you've
dragged enough to initiate a real tab drag we save focus. The code
never expected saving focus to trigger a capture lost (which stops tab
dragging), which triggered the crash. The fix is to make the code
allow for saving focus to delete the TabDragController.

I'm also converting from storing a raw View* for the previously focused view to a ViewStorage id. This way if the previously focused view is deleted no crashes.

I had to add a bit of infrastructure for the test.

BUG=275931
TEST=covered by test now.
R=ben@chromium.org

Review URL: https://chromiumcodereview.appspot.com/23523018

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221456 0039d316-1c4b-4281-b951-d872f2087c98

BUG=273137
TBR=lambroslambros@chromium.org

Review URL: https://codereview.chromium.org/23596010

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221455 0039d316-1c4b-4281-b951-d872f2087c98

> Remove compositing blacklisting of 10.7 and 10.6.
> 
> Also disable vsync to see if this improves performance.
> 
> BUG=266186
> TBR=kbr@chromium.org, zmo@chromium.org
> 
> Review URL: https://chromiumcodereview.appspot.com/23477037

TBR=ccameron@chromium.org

Review URL: https://codereview.chromium.org/23537025

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221446 0039d316-1c4b-4281-b951-d872f2087c98

This patch moves the test for valid handle to a more appropriate place. Currently the expectation itself in the test can cause a race condition because it implicitly looks up a shared mem pointer from a different thread.

BUG=284959
NOTRY=true

Review URL: https://chromiumcodereview.appspot.com/23986002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221444 0039d316-1c4b-4281-b951-d872f2087c98

When a cancel is requested for in-flight Truncate operation,
we don't (can't) stop running the Truncate, let it run until
the end, but then return following values:
- ABORT error for Truncate (while it hasn't been really aborted)
- OK for the cancel operation (while the operation hasn't been really canceled)

This feels weird since the cancel wasn't actually processed.

Instead this CL changes the return values:
- The original status code for Truncate
- INVALID_OPERATION for the cancel operation unles the original status code is ABORT

I verified that this doesn't break existing JS behavior.

BUG=none

Review URL: https://chromiumcodereview.appspot.com/23537020

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221443 0039d316-1c4b-4281-b951-d872f2087c98

Watch for filtering pref changes in ManagedModeInterstitial and unblock requests if they become allowed.

BUG=283379

Review URL: https://chromiumcodereview.appspot.com/23533014

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221442 0039d316-1c4b-4281-b951-d872f2087c98

Header: (80%) #333
Message: (80%) #333
Context Message: (50%) #7f7f7f

BUG=276004

Review URL: https://chromiumcodereview.appspot.com/23530018

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221441 0039d316-1c4b-4281-b951-d872f2087c98

Map glDiscardFramebufferEXT to glInvalidateFramebuffer on OpenGL ES
3.0. Both entry points have the same functionality; just the name was
changed between EXT_framebuffer_discard and ES 3.0.

BUG=274334

Review URL: https://chromiumcodereview.appspot.com/23855003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221440 0039d316-1c4b-4281-b951-d872f2087c98

For discussion, see:
https://chromereviews.googleplex.com/9784017/

BUG=255157
NOTRY=True

Review URL: https://chromiumcodereview.appspot.com/23690004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221439 0039d316-1c4b-4281-b951-d872f2087c98

This removes passing allowed_extensions all the way through, since it
was ignored anyways. It also initializes the workarounds from the
commandline during FeatureInfo construction so it can be used to
make decisions before the first context is created and initialized.

Review URL: https://chromiumcodereview.appspot.com/23660006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221438 0039d316-1c4b-4281-b951-d872f2087c98

When on the last focusable view in a Container FocusManager would
start searching from Widget. Problem is child Widgets share the same
FocusManager, meaning focus in a child widget wraps to the parent,
which is unexpected in most situations. In theory there could be
situations where we should wrap to the parent, but I don't think we
have any.

BUG=276213
TEST=covered by test
R=dmazzoni@chromium.org

Review URL: https://chromiumcodereview.appspot.com/23475012

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221437 0039d316-1c4b-4281-b951-d872f2087c98

Clean up URLs from activity log when a user cleans history via the history UI or using the extensions api.


BUG=253367

Review URL: https://chromiumcodereview.appspot.com/23785002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221436 0039d316-1c4b-4281-b951-d872f2087c98

This makes the dependency between the two classes explicit and more manageable. Pure refactoring, no functional change.

BUG=279268
(for chromeos/chromeos.gyp)
TBR=stevenjb@chromium.org

Review URL: https://chromiumcodereview.appspot.com/23522004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221435 0039d316-1c4b-4281-b951-d872f2087c98

Tiny convenience to allow stopping with CTRL+C / sigint.

BUG=

Review URL: https://chromiumcodereview.appspot.com/23450019

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221434 0039d316-1c4b-4281-b951-d872f2087c98

All textures are currently GL_CLAMP_TO_EDGE. This is good for tiles,
but not good for the background linen texture, which needs to be
repeated.

Add a mechanism to specify the texture wrap mode.

BUG=133097

Review URL: https://chromiumcodereview.appspot.com/22529002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221433 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221431 0039d316-1c4b-4281-b951-d872f2087c98

* Eliminate unused CYCLE_BACKWARD_LINEAR action -- linear cycling only goes
  forward.
* Consolidate the window cycling code into helper functions

BUG=none
TEST=ash_unittests Accelerator*

Review URL: https://chromiumcodereview.appspot.com/23654011

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221430 0039d316-1c4b-4281-b951-d872f2087c98

Call GetDefaultOutputDeviceID() from GetDefaultOutputStreamParameters() instead of specifying an empty string.

BUG=285437
R=henrika@chromium.org

Review URL: https://codereview.chromium.org/23440020

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221429 0039d316-1c4b-4281-b951-d872f2087c98

This is in a lead-up to destroying UI resources when visibility is lost.

BUG=279438

Review URL: https://chromiumcodereview.appspot.com/23475021

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221428 0039d316-1c4b-4281-b951-d872f2087c98

This adds UMA-uploaded histograms of sampling intervals for key points of data:
* Total time spent in a CFRunLoop
* Total time waiting in CFRunLoop
* Total time servicing MessagePump::Delegate (the MessageLoop)
* The MessageLoop queue size and queueing delay before servicing

It adds 1 second sampling for 15 seconds at startup, only for the main thread
(NSApplication-based) run loops.

The data will be used to determine if adding scheduling-like behavior to the
MessagePump will result in more efficient processing of the MessageLoop work.

An unforunate side effect of this change is exposing another method on the
MessagePump::Delegate interface, but there does not appear to be a better way
to do this.

BUG=264886
R=jar@chromium.org, mark@chromium.org

Review URL: https://codereview.chromium.org/22911026

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221427 0039d316-1c4b-4281-b951-d872f2087c98

On Android, Telemetry sets the chrome command line file before running any pages.
If the file doesn't already exist, it tries to write it as a normal user, rather
than root, which fails. This patch explicitly checks if the file exists. If it
doesn't exist it writes the file as root.

BUG=284468

Review URL: https://chromiumcodereview.appspot.com/23726018

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221426 0039d316-1c4b-4281-b951-d872f2087c98

When clearing the contents of a framebuffer, also invalidate/discard it
if the GL implementation supports this functionality. This improves
performance especially on tiling architecture GPUs since the graphics
driver does not need to restore the previous contents of the framebuffer
before proceeding with rendering.

Note that normally we don't clear the contents of framebuffers whose
contents we know we will completely re-render. This patch will still
invalidate such framebuffers to let the driver know that the previous
contents are not relevant.

BUG=274334

Review URL: https://chromiumcodereview.appspot.com/23601013

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@221425 0039d316-1c4b-4281-b951-d872f2087c98