Get rid of virtual Display::CreateScheduler.

Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Committed: https://crrev.com/8eb035c0f0b18c55ddd76dad607649e596035995
Cr-Commit-Position: refs/heads/master@{#400529}

[danakj, 2016-06-02 01:39:05]

Description was changed from

==========
Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for
https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
==========

to

==========
Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for
https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[danakj, 2016-06-02 01:53:40]

Description was changed from

==========
Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for
https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for
https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[danakj, 2016-06-02 01:53:40]

danakj@chromium.org changed reviewers:
+ boliu@chromium.org, sievers@chromium.org

[danakj, 2016-06-02 01:54:04]

The CQ bit was checked by danakj@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-02 01:55:14]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/2029323004/20001

[commit-bot: I haz the power, 2016-06-02 02:09:02]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-02 02:09:02]

Dry run: Try jobs failed on following builders:
  linux_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/8...)

[Fady Samuel, 2016-06-02 02:49:43]

mus lgtm

[enne (OOO), 2016-06-02 18:15:20]

https://codereview.chromium.org/2029323004/diff/20001/cc/surfaces/display.cc
File cc/surfaces/display.cc (right):

https://codereview.chromium.org/2029323004/diff/20001/cc/surfaces/display.cc#...
cc/surfaces/display.cc:61: if (begin_frame_source) {
I'm not sure I understand this logic.  Should this conditional line just be
removed?

[danakj, 2016-06-02 18:27:22]

https://codereview.chromium.org/2029323004/diff/20001/cc/surfaces/display.cc
File cc/surfaces/display.cc (right):

https://codereview.chromium.org/2029323004/diff/20001/cc/surfaces/display.cc#...
cc/surfaces/display.cc:61: if (begin_frame_source) {
On 2016/06/02 18:15:20, enne wrote:
> I'm not sure I understand this logic.  Should this conditional line just be
> removed?

Oh, this replaces the InitializeSynchronous(), see the comment in the header
file. When this is null, no scheduler is created.

[enne (OOO), 2016-06-02 18:58:10]

lgtm in general

https://codereview.chromium.org/2029323004/diff/20001/cc/surfaces/display.cc
File cc/surfaces/display.cc (right):

https://codereview.chromium.org/2029323004/diff/20001/cc/surfaces/display.cc#...
cc/surfaces/display.cc:61: if (begin_frame_source) {
On 2016/06/02 at 18:27:21, danakj wrote:
> On 2016/06/02 18:15:20, enne wrote:
> > I'm not sure I understand this logic.  Should this conditional line just be
> > removed?
> 
> Oh, this replaces the InitializeSynchronous(), see the comment in the header
file. When this is null, no scheduler is created.

Ohhhh ok ok, then can you remove the DCHECKs about begin frame source inside the
block?

[boliu, 2016-06-02 18:59:54]

rs lgtm

[danakj, 2016-06-03 01:28:35]

PTAL enne, jbauman especially.

https://codereview.chromium.org/2029323004/diff/20001/cc/surfaces/display.cc
File cc/surfaces/display.cc (right):

https://codereview.chromium.org/2029323004/diff/20001/cc/surfaces/display.cc#...
cc/surfaces/display.cc:61: if (begin_frame_source) {
On 2016/06/02 18:58:09, enne wrote:
> On 2016/06/02 at 18:27:21, danakj wrote:
> > On 2016/06/02 18:15:20, enne wrote:
> > > I'm not sure I understand this logic.  Should this conditional line just
be
> > > removed?
> > 
> > Oh, this replaces the InitializeSynchronous(), see the comment in the header
> file. When this is null, no scheduler is created.
> 
> Ohhhh ok ok, then can you remove the DCHECKs about begin frame source inside
the
> block?

Soo... why remove 2 DCHECKs when we can change who creates and owns everything?
YAS.

So now the creator of Display makes:
- BeginFrameSource
- DisplayScheduler
- OutputSurface
- TextureMailboxDeleter

They pass ownership of all 4 to Display. Now Display doesn't need to know about
TaskRunner.

A lot of changes in this patchset though, so PTAL everyone.

[danakj, 2016-06-03 01:28:50]

The CQ bit was checked by danakj@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-03 01:29:44]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/2029323004/60001

[Fady Samuel, 2016-06-03 01:31:51]

I like this cleanup! mus lgtm

[commit-bot: I haz the power, 2016-06-03 01:42:31]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-03 01:42:33]

Dry run: Try jobs failed on following builders:
  cast_shell_linux on tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/cast_shell_linu...)
  chromeos_daisy_chromium_compile_only_ng on tryserver.chromium.linux
(JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromeos_daisy_...)

[jbauman, 2016-06-03 02:34:28]

This generally seems like it should work.

https://codereview.chromium.org/2029323004/diff/60001/cc/surfaces/display.cc
File cc/surfaces/display.cc (right):

https://codereview.chromium.org/2029323004/diff/60001/cc/surfaces/display.cc#...
cc/surfaces/display.cc:73: new BackToBackBeginFrameSource(task_runner));
Where is this code to create a BackToBackBeginFrameSource going to live?

[boliu, 2016-06-03 17:15:28]

a_w still lgtm

[danakj, 2016-06-03 17:26:32]

https://codereview.chromium.org/2029323004/diff/60001/cc/surfaces/display.cc
File cc/surfaces/display.cc (right):

https://codereview.chromium.org/2029323004/diff/60001/cc/surfaces/display.cc#...
cc/surfaces/display.cc:73: new BackToBackBeginFrameSource(task_runner));
On 2016/06/03 02:34:27, jbauman wrote:
> Where is this code to create a BackToBackBeginFrameSource going to live?

Oh I forgot this, it needs to be added to GpuProcessTransformFactory.

[enne (OOO), 2016-06-03 18:33:16]

https://codereview.chromium.org/2029323004/diff/60001/cc/surfaces/display.h
File cc/surfaces/display.h (right):

https://codereview.chromium.org/2029323004/diff/60001/cc/surfaces/display.h#n...
cc/surfaces/display.h:113: DisplayScheduler* SchedulerForTesting() { return
scheduler_.get(); }
Is this still needed?

[danakj, 2016-06-03 21:34:12]

So I BackToBack vs Synthetic BFS is kinda problematic.

Right now OutputSurface owns a Synthetic BFS. Display may choose to ignore it
and own a separate BFS.

I don't really want to make 2 BFS and give one to each thing.. so I think what
I'll do is make a "IgnoreVsync" option on Synthetic BFS, and turn BackToBack BFS
into that. Then we just need the one, and when teh setting is true we can just
tell it to ignore vsync.

https://codereview.chromium.org/2029323004/diff/60001/cc/surfaces/display.h
File cc/surfaces/display.h (right):

https://codereview.chromium.org/2029323004/diff/60001/cc/surfaces/display.h#n...
cc/surfaces/display.h:113: DisplayScheduler* SchedulerForTesting() { return
scheduler_.get(); }
On 2016/06/03 18:33:16, enne wrote:
> Is this still needed?

Ah it was for constructors. But I deleted the class.

[danakj, 2016-06-15 21:47:29]

The CQ bit was checked by danakj@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-15 21:48:07]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/80001

[commit-bot: I haz the power, 2016-06-15 22:03:09]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-15 22:03:10]

Dry run: Try jobs failed on following builders:
  chromeos_daisy_chromium_compile_only_ng on tryserver.chromium.linux
(JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromeos_daisy_...)
  chromeos_x86-generic_chromium_compile_only_ng on tryserver.chromium.linux
(JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromeos_x86-ge...)

[danakj, 2016-06-16 19:03:27]

The CQ bit was checked by danakj@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-16 19:04:28]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/100001

[danakj, 2016-06-16 21:38:45]

OK this patch now preserves using BackToBack when vsync is disabled, by making
that BFS out in GpuProcessTransportFactory. I rebased onto newer things, and
fixed a bunch of self-nits. PTAL

[danakj, 2016-06-16 21:39:16]

Patchset #8 (id:140001) has been deleted

[danakj, 2016-06-16 21:39:37]

The CQ bit was checked by danakj@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-16 21:40:36]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/160001

[jbauman, 2016-06-16 22:00:17]

lgtm

[no sievers, 2016-06-16 22:02:09]

android lgtm

[commit-bot: I haz the power, 2016-06-16 22:05:07]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-16 22:05:10]

Dry run: Try jobs failed on following builders:
  linux_chromium_chromeos_ozone_rel_ng on tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[danakj, 2016-06-16 22:37:10]

Oh, because I changed SurfaceDisplayOutputSurface::BindToClient, this patch has
to change a little bit.

I made it so SurfaceDisplayOutputSurface::BindToClient does
RegisterSurfaceFactoryClient() before setting up its own client.

But this patch has Display's constructor doing RegisterBeginFrameSource. That
means that when SurfaceDisplayOutputSurface registers itself, it will
immediately hear about the BFS.

Since it is now doing that before setting up its own client, it is DCHECKing
that its too soon.

So I moved the call to RegisterBeginFrameSource to Display::Initialize, so the
SurfaceDisplayOutputSurface can make it happen after it has its client set up.
This is the time when the Display will start ticking so that seems okay. I added
comments about it to display.cc with the change.

[danakj, 2016-06-16 22:37:20]

The CQ bit was checked by danakj@chromium.org

[danakj, 2016-06-16 22:37:21]

The patchset sent to the CQ was uploaded after l-g-t-m from enne@chromium.org,
fsamuel@chromium.org, boliu@chromium.org, sievers@chromium.org,
jbauman@chromium.org
Link to the patchset: https://codereview.chromium.org/2029323004/#ps180001
(title: "displaytest: dcheck")

[commit-bot: I haz the power, 2016-06-16 22:38:11]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/180001

[danakj, 2016-06-16 22:39:29]

The CQ bit was checked by danakj@chromium.org

[danakj, 2016-06-16 22:39:30]

The patchset sent to the CQ was uploaded after l-g-t-m from
fsamuel@chromium.org, enne@chromium.org, sievers@chromium.org,
jbauman@chromium.org, boliu@chromium.org
Link to the patchset: https://codereview.chromium.org/2029323004/#ps200001
(title: "displaytest: dcheck")

[enne (OOO), 2016-06-16 22:39:42]

On 2016/06/16 at 22:37:10, danakj wrote:

> So I moved the call to RegisterBeginFrameSource to Display::Initialize, so the
SurfaceDisplayOutputSurface can make it happen after it has its client set up.
This is the time when the Display will start ticking so that seems okay. I added
comments about it to display.cc with the change.

lgtm

It makes sense to me that display shouldn't tick anything until after it's been
initialized.

[danakj, 2016-06-16 22:39:50]

Patchset #9 (id:180001) has been deleted

[commit-bot: I haz the power, 2016-06-16 22:40:18]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/200001

[danakj, 2016-06-17 00:36:08]

The CQ bit was unchecked by danakj@chromium.org

[danakj, 2016-06-17 00:36:16]

The CQ bit was checked by danakj@chromium.org

[commit-bot: I haz the power, 2016-06-17 00:36:46]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/200001

[commit-bot: I haz the power, 2016-06-17 04:08:31]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-17 04:08:33]

Try jobs failed on following builders:
  linux_android_rel_ng on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[danakj, 2016-06-17 18:19:18]

The CQ bit was checked by danakj@chromium.org

[commit-bot: I haz the power, 2016-06-17 18:20:00]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/200001

[commit-bot: I haz the power, 2016-06-17 20:16:04]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-17 20:16:05]

Try jobs failed on following builders:
  linux_android_rel_ng on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[boliu, 2016-06-17 20:21:17]

On 2016/06/17 20:16:05, commit-bot: I haz the power wrote:
> Try jobs failed on following builders:
>   linux_android_rel_ng on tryserver.chromium.android (JOB_FAILED,
>
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)


not webview this time!

I think bot stopped doing stack symbolization after switching to swarming. But
all the tests are failing a DCHECK:

[FATAL:display.cc(340)] Check failed: false.

[danakj, 2016-06-17 20:35:31]

On 2016/06/17 20:21:17, boliu wrote:
> On 2016/06/17 20:16:05, commit-bot: I haz the power wrote:
> > Try jobs failed on following builders:
> >   linux_android_rel_ng on tryserver.chromium.android (JOB_FAILED,
> >
>
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)
> 
> 
> not webview this time!
> 
> I think bot stopped doing stack symbolization after switching to swarming. But
> all the tests are failing a DCHECK:
> 
> [FATAL:display.cc(340)] Check failed: false.

OK so..

OutputSurfaceWithoutParent and ExternalBeginFrameSource both are implementing
CompositorImpl::VSyncObserver.

- ExternalBeginFrameSource uses it to make a begin frame.
- OutputSurfaceWithoutParent passes the vsync params to its client, which is the
Display. That used to be an empty method, now it DCHECKs.

I'll make OutputSurfaceWithoutParent stop being a VSyncObserver at all.

[danakj, 2016-06-17 20:36:38]

On Fri, Jun 17, 2016 at 1:21 PM, <boliu@chromium.org> wrote:

> On 2016/06/17 20:16:05, commit-bot: I haz the power wrote:
> > Try jobs failed on following builders:
> > linux_android_rel_ng on tryserver.chromium.android (JOB_FAILED,
> >
>
>
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...
> )
>
>
> not webview this time!
>
> I think bot stopped doing stack symbolization after switching to swarming.
> But
> all the tests are failing a DCHECK:
>
> [FATAL:display.cc(340)] Check failed: false.
>

Oh, it keeps showing as a purple bot which doesn't look like it was my
patch. >_< Thanks!


>
> https://codereview.chromium.org/2029323004/
>

-- 
You received this message because you are subscribed to the Google Groups
"Chromium-reviews" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[danakj, 2016-06-17 21:07:23]

The CQ bit was checked by danakj@chromium.org

[danakj, 2016-06-17 21:07:24]

The patchset sent to the CQ was uploaded after l-g-t-m from
fsamuel@chromium.org, enne@chromium.org, sievers@chromium.org,
jbauman@chromium.org, boliu@chromium.org
Link to the patchset: https://codereview.chromium.org/2029323004/#ps240001
(title: "displaytest: unusedvar")

[commit-bot: I haz the power, 2016-06-17 21:07:45]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2029323004/240001

[commit-bot: I haz the power, 2016-06-17 22:41:46]

Description was changed from

==========
Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for
https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for
https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[commit-bot: I haz the power, 2016-06-17 22:41:49]

Committed patchset #11 (id:240001)

[commit-bot: I haz the power, 2016-06-17 22:43:39]

Description was changed from

==========
Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for
https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
Get rid of virtual Display::CreateScheduler.

This method is virtual for tests to override the scheduler. Instead
have tests pass a different scheduler to a constructor.

In order to make this work, we must set up the scheduler in the
Display's constructor. For us to do that, we need to pass the
BeginFrameSource to the constructor also, which means the Display's
OutputSurface does not give a BeginFrameSource to the Display during
BindToClient. Instead, the code that makes the OutputSurface and the
Display must figure out the BeginFrameSource and pass it to the
Display (and maybe to the OutputSurface if it needs it).

This is some cleanup followup work for
https://codereview.chromium.org/2036563002/

R=enne@chromium.org, fsamuel@chromium.org, jbauman@chromium.org
BUG=487471
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Committed: https://crrev.com/8eb035c0f0b18c55ddd76dad607649e596035995
Cr-Commit-Position: refs/heads/master@{#400529}
==========

[commit-bot: I haz the power, 2016-06-17 22:43:41]

Patchset 11 (id:??) landed as
https://crrev.com/8eb035c0f0b18c55ddd76dad607649e596035995
Cr-Commit-Position: refs/heads/master@{#400529}