Update mojo sdk to rev 7214b7ec7d27563b2666afad86cf1c5895c56c18

Update mojo sdk to rev 7214b7ec7d27563b2666afad86cf1c5895c56c18

BUG=
TBR=jamesr@chromium.org for third_party/mojo
TBR=dpranke@chromium.org for build

Committed: https://crrev.com/ad6b6cbf77962ba99de7180403d30845484e2640
Cr-Commit-Position: refs/heads/master@{#321388}

[Ken Rockot(use gerrit already), 2015-03-19 00:57:46]

rockot@chromium.org changed reviewers:
+ imcheng@chromium.org

[Ken Rockot(use gerrit already), 2015-03-19 00:57:47]

Please take a look

[Ken Rockot(use gerrit already), 2015-03-19 00:58:04]

(for presentation service changes)

[imcheng, 2015-03-19 01:21:14]

lgtm

Thanks. I reviewed the presentation_service_impl* files.

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
File content/browser/presentation/presentation_service_impl.cc (right):

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
content/browser/presentation/presentation_service_impl.cc:99:
callback.Run(presentation_url, false);;
extra semicolon

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
content/browser/presentation/presentation_service_impl.cc:167:
base::Bind(&PresentationServiceImpl::OnStartOrJoinSessionSucceeded,
This is actually another potential place where callback can get dropped. You can
put a TODO for me and reference crbug.com/468575 here.

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
content/browser/presentation/presentation_service_impl.cc:194:
base::Bind(&PresentationServiceImpl::OnStartOrJoinSessionSucceeded,
This is actually another potential place where callback can get dropped. You can
put a TODO for me and reference crbug.com/468575 here.

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
content/browser/presentation/presentation_service_impl.cc:322:
availability_contexts_.clear();
This is actually another place where callbacks get dropped.
availability_contexts_ and queued_start_session_requests_ both hold references
to mojo callbacks.

[Ken Rockot(use gerrit already), 2015-03-19 16:16:39]

Thanks!

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
File content/browser/presentation/presentation_service_impl.cc (right):

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
content/browser/presentation/presentation_service_impl.cc:99:
callback.Run(presentation_url, false);;
On 2015/03/19 01:21:14, imcheng1 wrote:
> extra semicolon

Done.

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
content/browser/presentation/presentation_service_impl.cc:167:
base::Bind(&PresentationServiceImpl::OnStartOrJoinSessionSucceeded,
On 2015/03/19 01:21:14, imcheng1 wrote:
> This is actually another potential place where callback can get dropped. You
can
> put a TODO for me and reference crbug.com/468575 here.

Done.

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
content/browser/presentation/presentation_service_impl.cc:194:
base::Bind(&PresentationServiceImpl::OnStartOrJoinSessionSucceeded,
On 2015/03/19 01:21:14, imcheng1 wrote:
> This is actually another potential place where callback can get dropped. You
can
> put a TODO for me and reference crbug.com/468575 here.

Done.

https://codereview.chromium.org/1019173002/diff/20001/content/browser/present...
content/browser/presentation/presentation_service_impl.cc:322:
availability_contexts_.clear();
On 2015/03/19 01:21:14, imcheng1 wrote:
> This is actually another place where callbacks get dropped.
> availability_contexts_ and queued_start_session_requests_ both hold references
> to mojo callbacks.

Oops. I've changed this to signal the availability_contexts_ for now and added a
TODO for the queued requests.

[Ken Rockot(use gerrit already), 2015-03-19 16:18:59]

rockot@chromium.org changed reviewers:
+ mlamouri@chromium.org

[Ken Rockot(use gerrit already), 2015-03-19 16:19:00]

Mounir, could you please take a look at PermissionServiceImpl changes?

Note that this is needed because Mojo pipes are now properly closed if a
receiver drops a response callback (i.e. if it becomes impossible to respond).

I just moved the callback from bind arguments into the PendingRequest structure,
and it will respond on destruction if it's still hanging around by then.

[mlamouri (slow - plz ping), 2015-03-19 16:40:25]

On 2015/03/19 at 16:19:00, rockot wrote:
> Mounir, could you please take a look at PermissionServiceImpl changes?
> 
> Note that this is needed because Mojo pipes are now properly closed if a
receiver drops a response callback (i.e. if it becomes impossible to respond).
> 
> I just moved the callback from bind arguments into the PendingRequest
structure, and it will respond on destruction if it's still hanging around by
then.

Why not pass the callback as base::Owned()? Could that be generalised to some
other call sites like the PresentationService?

[mlamouri (slow - plz ping), 2015-03-19 16:45:43]

On 2015/03/19 at 16:40:25, Mounir Lamouri wrote:
> On 2015/03/19 at 16:19:00, rockot wrote:
> > Mounir, could you please take a look at PermissionServiceImpl changes?
> > 
> > Note that this is needed because Mojo pipes are now properly closed if a
receiver drops a response callback (i.e. if it becomes impossible to respond).
> > 
> > I just moved the callback from bind arguments into the PendingRequest
structure, and it will respond on destruction if it's still hanging around by
then.
> 
> Why not pass the callback as base::Owned()? Could that be generalised to some
other call sites like the PresentationService?

Hmm, I guess having a wrapper owning the callback would work better. But it's
probably worth looking at after this CL.

All this added complexity bothers me a lot but I guess I can't block the roll
for that. lgtm

[Ken Rockot(use gerrit already), 2015-03-19 17:07:14]

The CQ bit was checked by rockot@chromium.org

[Ken Rockot(use gerrit already), 2015-03-19 17:07:15]

The patchset sent to the CQ was uploaded after l-g-t-m from imcheng@chromium.org
Link to the patchset: https://codereview.chromium.org/1019173002/#ps60001
(title: "Keep permission service alive if embedder drops requests")

[commit-bot: I haz the power, 2015-03-19 17:07:30]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1019173002/60001

[commit-bot: I haz the power, 2015-03-19 17:54:32]

Committed patchset #4 (id:60001)

[commit-bot: I haz the power, 2015-03-19 17:55:24]

Patchset 4 (id:??) landed as
https://crrev.com/ad6b6cbf77962ba99de7180403d30845484e2640
Cr-Commit-Position: refs/heads/master@{#321388}