mojo: fail gracefully if ServiceRegistry can't find ServiceConnector

mojo: fail gracefully if ServiceRegistry can't find ServiceConnector

Previously, the newly added test would crash because the ServiceRegistry asserted it had a ServiceConnector for anything allowed through by AllowIncomingConnection.  Nothing actually prevents this from happening, so this patch makes sure it is handled gracefully by placing the InterfacePtr into an error state.

BUG=None

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=278240

[tim (not reviewing), 2014-06-18 02:49:57]

Thanks both for the help today.  Trung, please review. Dave FYI  (although I'd
accept approval from either of you!)

[DaveMoore, 2014-06-18 13:23:54]

lgtm

[viettrungluu, 2014-06-18 14:52:06]

LGTM (with question, possibly for Dave).

https://codereview.chromium.org/343473005/diff/60001/mojo/public/cpp/applicat...
File mojo/public/cpp/application/lib/service_registry.cc (right):

https://codereview.chromium.org/343473005/diff/60001/mojo/public/cpp/applicat...
mojo/public/cpp/application/lib/service_registry.cc:62:
name_to_service_connector_.find(service_name) ==
I wonder if we shouldn't reverse the order of the checks, i.e., not call
application_->AllowIncomingConnection() on nonexistent service names.

(Just a thought. I'll defer to Dave's opinion on this.)

[tim (not reviewing), 2014-06-18 20:18:20]

https://codereview.chromium.org/343473005/diff/60001/mojo/public/cpp/applicat...
File mojo/public/cpp/application/lib/service_registry.cc (right):

https://codereview.chromium.org/343473005/diff/60001/mojo/public/cpp/applicat...
mojo/public/cpp/application/lib/service_registry.cc:62:
name_to_service_connector_.find(service_name) ==
On 2014/06/18 14:52:06, viettrungluu wrote:
> I wonder if we shouldn't reverse the order of the checks, i.e., not call
> application_->AllowIncomingConnection() on nonexistent service names.
> 
> (Just a thought. I'll defer to Dave's opinion on this.)

I like that better than what I had. It seems more in line with the expectation
of AllowIncomingConnection that it only makes decisions for realistic
connections. Also, the way the test was structured was to increment a
num_connected_ bool as a side effect of an overridden AllowIncomingConnection,
which was weird in this case where the ++ happened but the connection didn't.
Now I can add an additional assertion to the unit test.  Done.

[tim (not reviewing), 2014-06-18 20:18:24]

The CQ bit was checked by tim@chromium.org

[commit-bot: I haz the power, 2014-06-18 20:20:28]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/tim@chromium.org/343473005/100001

[commit-bot: I haz the power, 2014-06-19 01:58:56]

Change committed as 278240