Reduce the scope of CGL current contexts

Reduce the scope of CGL current contexts

There are a few crash reports coming in where the destructor
~ScopedCGLSetCurrentContext is being implicated inside
CompositorSwapBuffers. Reduce the scope of the 
ScopedCGLSetCurrentContext in CompositorSwapBuffers.

At a minimum, this will make the crash reports reveal which
place is causing problems. At a maximum, this may fix the
issue (in particular, LayoutLayers may call displayIfNeeded,
which may mess with the current context, potentially causing
crashes, as ScopedCGLSetCurrentContext does not expect
this).

BUG=245900
NOTRY=true

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=260448

[ccameron, 2014-03-28 10:01:48]

I'd like to squeeze this one in before the branch is cut this afternoon.

https://codereview.chromium.org/216733002/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/render_widget_host_view_mac.mm (right):

https://codereview.chromium.org/216733002/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/render_widget_host_view_mac.mm:1391: {
Of note is that this should potentially be moved up to the
SetIOSurfaceWithContextCurrent block unconditionally. I'm leaving this as-is to
minimize altered behavior.

[Ken Russell (switch to Gerrit), 2014-03-28 16:18:41]

LGTM

https://codereview.chromium.org/216733002/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/render_widget_host_view_mac.mm (right):

https://codereview.chromium.org/216733002/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/render_widget_host_view_mac.mm:1340:
DCHECK(!CGLGetCurrentContext());
Have you run a debug build with this DCHECK in place for some period of time to
see whether it's triggered?

[ccameron, 2014-03-28 16:22:49]

Thanks!

https://codereview.chromium.org/216733002/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/render_widget_host_view_mac.mm (right):

https://codereview.chromium.org/216733002/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/render_widget_host_view_mac.mm:1340:
DCHECK(!CGLGetCurrentContext());
On 2014/03/28 16:18:41, Ken Russell wrote:
> Have you run a debug build with this DCHECK in place for some period of time
to
> see whether it's triggered?

Yes, but, you bring up a good point -- I haven't run a build long enough. I'll
want to add a bunch of CHECKs to ensure this and some other things (like that
the expected context is current when ~ScopedCGLSetCurrentContext) are true. I'll
put those in a separate patch.

[ccameron, 2014-03-28 16:22:54]

The CQ bit was checked by ccameron@chromium.org

[commit-bot: I haz the power, 2014-03-28 16:27:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ccameron@chromium.org/216733002/10001

[commit-bot: I haz the power, 2014-03-28 17:41:22]

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

[commit-bot: I haz the power, 2014-03-28 17:41:22]

Try jobs failed on following builders:
  tryserver.chromium on linux_chromium_rel

[ccameron, 2014-03-28 17:44:30]

The CQ bit was checked by ccameron@chromium.org

[commit-bot: I haz the power, 2014-03-28 17:46:04]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ccameron@chromium.org/216733002/10001

[Ken Russell (switch to Gerrit), 2014-03-28 17:57:31]

On 2014/03/28 17:41:22, I haz the power (commit-bot) wrote:
> Try jobs failed on following builders:
>   tryserver.chromium on linux_chromium_rel

Chris, please file a bug about the flaky test that caused your try job to fail.

[Avi (use Gerrit), 2014-03-28 18:00:43]

On 2014/03/28 17:57:31, Ken Russell wrote:
> On 2014/03/28 17:41:22, I haz the power (commit-bot) wrote:
> > Try jobs failed on following builders:
> >   tryserver.chromium on linux_chromium_rel
> 
> Chris, please file a bug about the flaky test that caused your try job to
fail.

You don't need to file a bug, but email details to chrome-troopers@google.com.
They want to hear about all flaky tests.

[Ken Russell (switch to Gerrit), 2014-03-28 18:03:34]

On 2014/03/28 18:00:43, Avi wrote:
> On 2014/03/28 17:57:31, Ken Russell wrote:
> > On 2014/03/28 17:41:22, I haz the power (commit-bot) wrote:
> > > Try jobs failed on following builders:
> > >   tryserver.chromium on linux_chromium_rel
> > 
> > Chris, please file a bug about the flaky test that caused your try job to
> fail.
> 
> You don't need to file a bug, but email details to
mailto:chrome-troopers@google.com.
> They want to hear about all flaky tests.

In this particular case it looks like something the Telemetry team would be most
adept at investigating.

[ccameron, 2014-03-28 18:25:13]

On 2014/03/28 18:03:34, Ken Russell wrote:
> On 2014/03/28 18:00:43, Avi wrote:
> > On 2014/03/28 17:57:31, Ken Russell wrote:
> > > On 2014/03/28 17:41:22, I haz the power (commit-bot) wrote:
> > > > Try jobs failed on following builders:
> > > >   tryserver.chromium on linux_chromium_rel
> > > 
> > > Chris, please file a bug about the flaky test that caused your try job to
> > fail.
> > 
> > You don't need to file a bug, but email details to
> mailto:chrome-troopers@google.com.
> > They want to hear about all flaky tests.
> 
> In this particular case it looks like something the Telemetry team would be
most
> adept at investigating.

Filed http://crbug.com/357691 on this. I'm going to kill of the test, since it's
blocking both of tryjobs I'm pushing right now (one on Mac one on Linux). And
I've seen this one flake in the past.

[ccameron, 2014-03-28 18:32:59]

The CQ bit was unchecked by ccameron@chromium.org

[ccameron, 2014-03-28 18:33:27]

The CQ bit was checked by ccameron@chromium.org

[ccameron, 2014-03-28 18:33:27]

The CQ bit was unchecked by ccameron@chromium.org

[ccameron, 2014-03-28 18:35:01]

The CQ bit was checked by ccameron@chromium.org

[ccameron, 2014-03-28 21:45:33]

The CQ bit was unchecked by ccameron@chromium.org

[ccameron, 2014-03-28 21:45:54]

The CQ bit was checked by ccameron@chromium.org

[cmp, 2014-03-30 17:01:12]

The CQ bit was unchecked by cmp@chromium.org

[cmp, 2014-03-30 17:05:18]

The CQ bit was checked by cmp@chromium.org

[commit-bot: I haz the power, 2014-03-30 17:06:03]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ccameron@chromium.org/216733002/50001

[commit-bot: I haz the power, 2014-03-30 17:29:43]

Change committed as 260448