Implement support for route change notifications in the Transport interface.

Implement support for route change notifications in the Transport interface.

BUG=110485
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=122901

[Sergey Ulanov, 2012-02-08 19:11:06]

+Lambros

[Lambros, 2012-02-08 21:43:24]

lgtm

[Wez, 2012-02-21 18:42:33]

http://codereview.chromium.org/9366001/diff/1/remoting/protocol/libjingle_tra...
File remoting/protocol/libjingle_transport_factory.cc (right):

http://codereview.chromium.org/9366001/diff/1/remoting/protocol/libjingle_tra...
remoting/protocol/libjingle_transport_factory.cc:227: NOTREACHED();
Is NOTREACHED() appropriate here?  It'll mean that we report a random route.type
if we ever see a new candidate.type().

http://codereview.chromium.org/9366001/diff/1/remoting/protocol/libjingle_tra...
remoting/protocol/libjingle_transport_factory.cc:232: NOTREACHED();
Similarly here; since these notifications end up being used for audit we should
really crash out if we see invalid data and can't process it.  At this level
it'd make sense to report the fact that we couldn't parse the address to the
event handler, and let it choose how to handle invalid values.

[Sergey Ulanov, 2012-02-21 19:47:19]

http://codereview.chromium.org/9366001/diff/1/remoting/protocol/libjingle_tra...
File remoting/protocol/libjingle_transport_factory.cc (right):

http://codereview.chromium.org/9366001/diff/1/remoting/protocol/libjingle_tra...
remoting/protocol/libjingle_transport_factory.cc:227: NOTREACHED();
On 2012/02/21 18:42:34, Wez wrote:
> Is NOTREACHED() appropriate here?  It'll mean that we report a random
route.type
> if we ever see a new candidate.type().

Good point. Changed this to LOG(FATAL) that would crash in Release builds too.

http://codereview.chromium.org/9366001/diff/1/remoting/protocol/libjingle_tra...
remoting/protocol/libjingle_transport_factory.cc:232: NOTREACHED();
On 2012/02/21 18:42:34, Wez wrote:
> Similarly here; since these notifications end up being used for audit we
should
> really crash out if we see invalid data and can't process it.  At this level
> it'd make sense to report the fact that we couldn't parse the address to the
> event handler, and let it choose how to handle invalid values.

Changed it to LOG(FATAL) too.

[commit-bot: I haz the power, 2012-02-21 21:32:56]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sergeyu@chromium.org/9366001/7005

[commit-bot: I haz the power, 2012-02-21 23:23:12]

Change committed as 122901