GitLab

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278019 0039d316-1c4b-4281-b951-d872f2087c98

This broke virtual/gpu/fast/canvas/canvas-draw-canvas-on-canvas-shadow.html
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20Blink&tests=virtual/gpu/fast/canvas/canvas-draw-canvas-on-canvas-shadow.html


> roll skia to db6346a5b1d76b150920a5dce47c436fee9c6281
> 
> TBR=
> 
> Review URL: https://codereview.chromium.org/341603006

TBR=reed@chromium.org

Review URL: https://codereview.chromium.org/340023002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278018 0039d316-1c4b-4281-b951-d872f2087c98

BUG=114092

Review URL: https://codereview.chromium.org/308833002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278017 0039d316-1c4b-4281-b951-d872f2087c98

BUG=374520

Review URL: https://codereview.chromium.org/336313003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278015 0039d316-1c4b-4281-b951-d872f2087c98

On failure, future transport writes should synchronously return. This is
important on Chrome OS and Linux where we have a separate NSS task runner.
If we query the transport each time (in hopes that it will return the error
code) it becomes an asynchronous error and so the state machine keeps pumping
itself in response to the state change. (It alternates between "write pending"
and "write failed".)

Add a test that asserts we do not keep trying to write to the transport in a
loop.

This fixes one of the infinite loops in bug #381160, but not the other.

BUG=381160

Review URL: https://codereview.chromium.org/337823002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278014 0039d316-1c4b-4281-b951-d872f2087c98

BUG=382931
TEST=Build sandbox twice. The second time it says "no work to do".

Review URL: https://codereview.chromium.org/332403005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278013 0039d316-1c4b-4281-b951-d872f2087c98

I still have to rename MOJO_WAIT_... and MojoWaitFlagsState, but I'll do
that separately.

R=sky@chromium.org

Review URL: https://codereview.chromium.org/345463003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278012 0039d316-1c4b-4281-b951-d872f2087c98

This duplicates the accessibility enums between accessibility and automation converting to camel casing. It also cleans up some of the idl

For the sake of record, attempts to share the same enum ran into the following issues:
- the docs generator assumes a one-to-one (save camel/snake/unix casing) correspondance between filenames and idl namespaces.
- the handlebar templating language does not allow for logical string comparisons/equality checks or passing in-line string literals, so attempting to in-line names was unsuccessful.
- references to the enums (which were cased for the sake of cpp enums), were incorrect for the docs (which want camel casing).

A local cl with the change is still available  but suffers
from odd filename, casing conflicts, and AX* naming issues.

BUG=309681

Review URL: https://codereview.chromium.org/335923003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278011 0039d316-1c4b-4281-b951-d872f2087c98

Suspect that this is causing an increase in OOM errors.

> base: Increase emulated discardable memory limit.
>
> This increases the limit to 128MB but keeps the limit under moderate
> pressure low so the total amount of memory used by chrome is not
> significantly increased.
>
> BUG=368078
>
> Review URL: https://codereview.chromium.org/266993003

TBR=willchan@chromium.org
BUG=372158

Review URL: https://codereview.chromium.org/341773002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278010 0039d316-1c4b-4281-b951-d872f2087c98

This time, add an annotation to ignore the leak in system libraries.

BUG=176888,375126

Review URL: https://codereview.chromium.org/321223004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278008 0039d316-1c4b-4281-b951-d872f2087c98

PepperPacketSocketFactory has been updated to handle transient errors
properly in https://codereview.chromium.org/336113002/. This is
correpsonding change for ChromiumPacketSocketFactory. Particularly
ChromiumPacketSocketFactory now tries resending each packet after
EHOSTUNREACH.

Review URL: https://codereview.chromium.org/339503006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278007 0039d316-1c4b-4281-b951-d872f2087c98

ProfileSyncComponentsFactoryImpl now creates an AttachmentService with
an AttachmentUploadImpl instead of a fake.

Moved the logic for computing the sync server URL into
ProfileSyncService.

Currently, a new AUI is created for each AttachmentService.  In the
future, we will create one AUI that is shared among AttachmentService
instances.

ProfileSyncComponentsFactoryImplTest now uses TestBrowserThreadBundle
instead of the deprecated TestBrowserThread. PSCFIT now needs an IO
thread (for AUI's URLFetcher).

BUG=
TBR=finnur@chromium.org

Review URL: https://codereview.chromium.org/334713008

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278006 0039d316-1c4b-4281-b951-d872f2087c98

BUG=344769

Review URL: https://codereview.chromium.org/330283002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278005 0039d316-1c4b-4281-b951-d872f2087c98

BUG=none
R=robertshield@chromium.org

Review URL: https://codereview.chromium.org/338353002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278004 0039d316-1c4b-4281-b951-d872f2087c98

After the landing https://codereview.chromium.org/291093005,
browser_tab_strip_controller.cc doesn't need the command line switch
related headers anymore. So we can remove these.

BUG=332077

Review URL: https://codereview.chromium.org/339023002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278003 0039d316-1c4b-4281-b951-d872f2087c98

This corresponds to the Chromecast change at this URL:
https://eureka-internal-review.googlesource.com/#/c/19857/

R=gwendal@chromium.org, isherman@chromium.org, robertshield@chromium.org, semenzato@chromium.org, asvitkine@chromium.org
BUG=384533

Review URL: https://codereview.chromium.org/335913002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278002 0039d316-1c4b-4281-b951-d872f2087c98

Make a copy of the std::vector before modifying it.

BUG=385724, 385868

Review URL: https://codereview.chromium.org/339223004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278001 0039d316-1c4b-4281-b951-d872f2087c98

BUG=370600

Review URL: https://codereview.chromium.org/330953002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278000 0039d316-1c4b-4281-b951-d872f2087c98

Should be landed after:
https://codereview.chromium.org/323513003/
https://codereview.chromium.org/314293004/

which fix related shutdown issues.

BUG=367409,373249

Review URL: https://codereview.chromium.org/319053006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277998 0039d316-1c4b-4281-b951-d872f2087c98

Rewrite safe_browsing_util::GetHashIndex() and GetUrlHashIndex() to use
sensible return types rather than transitioning to/from int.  Pull them
into database_manager.cc, which was the only place they were used.

BUG=none

Review URL: https://codereview.chromium.org/337723004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277997 0039d316-1c4b-4281-b951-d872f2087c98

this regressed when we switched to the aura build for linux.

BUG=none

Review URL: https://codereview.chromium.org/331333002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277996 0039d316-1c4b-4281-b951-d872f2087c98

Otherwise, we won't be able to initialize it again.

BUG=chrome-os-partner:29674
TEST=Run the unittest on nyan.

Review URL: https://codereview.chromium.org/339863002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277995 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277994 0039d316-1c4b-4281-b951-d872f2087c98

The thumbnail of an image file is a squared image. This is implemented as an
image in a square container. The non square image is cropped
by the square container, but the hidden part of the image causes unwated
position gap of the dragging thumbnail.

Previously we handle the gap by adding 1000px padding for the thumbnail. But the
padding causes another bug when the browser is shown in non 100% scale. This CL
stops to add 1000px padding. Instead, to handle the unwanted gap, the CL let
Files.app copy the image to a square canvas. For the solution, we don't need to
crop an image by a square canvas because the canvas itself is square, and thus we
are free from position gap problem.

BUG=339179
TEST=manually
R=mtomasz@chromium.org

Review URL: https://codereview.chromium.org/338253003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277993 0039d316-1c4b-4281-b951-d872f2087c98

TBR=

Review URL: https://codereview.chromium.org/341603006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277992 0039d316-1c4b-4281-b951-d872f2087c98

Revert "Improve the no-op reconcilor so that it doesn't "hang" after the first execution, but instead successfully completes the reconciliation." (https://codereview.chromium.org/309843002/)

The original CL causes crashes about 20 secs after launching the linux build of chromeos.

This reverts commit 660a4d00.

Conflicts:
	chrome/browser/signin/account_reconcilor_unittest.cc

Original issue's description:
> Improve the no-op reconcilor so that it doesn't "hang" after the first execution, but instead successfully completes the reconciliation.
> Improve the UMAHistogramHelper so that it can nicely take snapshots between tests, allowing the test writer to easily test ONLY the changes logged in "this" test.
> Write a new unit test that executes the reconcilor twice, and make all reconcilor unit tests execute with and without the flag.
>
> BUG=357693
> TBR=phajdan.jr@chromium.org
>
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=277605

BUG=357693,386045
TEST=None
TBR=mlerman@chromium.org

Review URL: https://codereview.chromium.org/344513005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277990 0039d316-1c4b-4281-b951-d872f2087c98

- All files from webkit/renderer/compositor_bindings have been moved to content/renderer/compositor_bindings
- Webkit_compositor_support is moved in content_renderer
- Webkit_compositor_bindings_unittests is moved in content_unittests
- A dummy target has been left for webkit_compositor_bindings_unittests so as to not break the build.

TBR=nduca

BUG=265753

Review URL: https://codereview.chromium.org/317163002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277988 0039d316-1c4b-4281-b951-d872f2087c98

If I've never used them, they must not be useful.

BUG=none

Review URL: https://codereview.chromium.org/344513003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277987 0039d316-1c4b-4281-b951-d872f2087c98

On 32-bit platform, the expected result for this test is address of
64-bit, while the actual result is address of 32-bit. So they don't
match with each other.

BUG=380492

Review URL: https://codereview.chromium.org/310323003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277986 0039d316-1c4b-4281-b951-d872f2087c98

BUG=

Review URL: https://codereview.chromium.org/331283002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277985 0039d316-1c4b-4281-b951-d872f2087c98

This can happen if we receive an IdleNotification after OnRenderProcessShutdown.

BUG=367711

Review URL: https://codereview.chromium.org/256803012

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277984 0039d316-1c4b-4281-b951-d872f2087c98

BUG=

Review URL: https://codereview.chromium.org/339273002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277983 0039d316-1c4b-4281-b951-d872f2087c98

Events are now sent to the window manager (assumed to be at connection
id 1 for now). The window manager can then do what it wants with
them. For now I made it bounce the event back to the server so
everything works as it did.

Until we get FIFO across pipes I'm putting these messages on the
ViewManager/ViewManagerClient interface. That'll change.

BUG=384433
TEST=covered by tests
R=ben@chromium.org

Review URL: https://codereview.chromium.org/338353006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277982 0039d316-1c4b-4281-b951-d872f2087c98

Changes:
1) Run clang-format through source and header files.
2) Rename data member variables to use unix_hacker_ style.
2) Rename methods to use CamelCase style.
3) Rename file name to web_permissions.

BUG=331299
TEST=content_unittests, content_shell. No functional changes
TBR=dpranke@chromium.org

Review URL: https://codereview.chromium.org/322713002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277981 0039d316-1c4b-4281-b951-d872f2087c98

This cl seems to cause heap use after free on the Asan bots.
LargePage
FindLongString

http://build.chromium.org/p/chromium.memory/builders/Linux%20ASan%20LSan%20Tests%20%282%29/builds/3837/steps/browser_tests/logs/LargePage

http://build.chromium.org/p/chromium.memory/builders/Linux%20ASan%20LSan%20Tests%20%283%29/builds/4211/steps/browser_tests/logs/FindLongString
> cc: In SyncFromActiveLayer, drop all tiles not in the recorded_viewport.
> 
> Currently we do a walk over all tiles in the pending tree's pile, but
> there can be 1 million x 1 million tiles on some pages, which makes this
> method take multiple _seconds_ to complete.
> 
> These loops were added in r184525 which gave the tradeoffs that led to
> them, which are that we want to only use a single pile for all the tiles
> on this layer, but we don't want raster tiles on the active layer (when
> this activates) that can't be rastered by the pile they are attached to.
> 
> Instead of walking every pile-tile, to find the tiles that are not
> present in the current recording and invalidate them, we expand
> invalidations outside the interest rect to cover the full recording tiles,
> and we expand invalidation inside the interest rect to include any raster
> tiles that don't have a recording (such as in the offscreen animating gif
> case). We give this expanded invalidation to the pending layer, causing
> it to drop rastered tiles from the active tree that intersect with an
> unrecorded area.
> 
> R=enne
> BUG=371839
> 
> Review URL: https://codereview.chromium.org/294163009

TBR=danakj@chromium.org

Review URL: https://codereview.chromium.org/347493002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277980 0039d316-1c4b-4281-b951-d872f2087c98

BUG=none
TEST=compile

Review URL: https://codereview.chromium.org/335353002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277979 0039d316-1c4b-4281-b951-d872f2087c98

Pretty straightforward, nothing really blocks it, so we can move it
easily from ui_unittests target to gfx_unittests now.

BUG=331829
TEST=ui_unittests, gfx_unittests with/out --test-launcher-bot-mode
TBR=danakj@chromium.org,ben@chromium.org

Review URL: https://codereview.chromium.org/332353004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277978 0039d316-1c4b-4281-b951-d872f2087c98

Since only one of GL_TEXTURE_2D and GL_TEXTURE_EXTERNAL_OES would be used. Switch active texture according to the target type and unbound the texture after use.
    
BUG=chromium:383414
TEST=Run the unittest on link and peach-pit

Review URL: https://codereview.chromium.org/324253005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277977 0039d316-1c4b-4281-b951-d872f2087c98

If a frame overlaps appendWindowStart, partial append window filtering
is supported for the track, and the frame is appended to the track
buffer, this change updates the track buffer's last frame duration with
the frame's original duration instead of the possibly trimmed duration.
If the trimmed duration were used, then the discontinuity detection
logic could detect a discontinuity on the next frame for the track where
there would otherwise not be a discontinuity.

BUG=381114
R=acolwell@chromium.org
TEST=FrameProcessorTest.PartialAppendWindowFilterNoDiscontinuity

Review URL: https://codereview.chromium.org/344513002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277976 0039d316-1c4b-4281-b951-d872f2087c98

(previously JsCollectGarbageAction and InteractAction).

BUG=361809

Review URL: https://codereview.chromium.org/338883005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277975 0039d316-1c4b-4281-b951-d872f2087c98