Enable WebRTC support in the remoting host.

Enable WebRTC support in the remoting host.

WebRTC is enabled in the host when the host is started with
the --enable-webrtc flag. It also allows no-op authentication, so
it's only allowed in Debug builds.

BUG=547158
Committed: https://crrev.com/de15c037ab8425eecaa7ba7c3f63fc42e99d9f09
Cr-Commit-Position: refs/heads/master@{#364515}

[Sergey Ulanov, 2015-12-09 22:00:14]

Patchset #1 (id:1) has been deleted

[Sergey Ulanov, 2015-12-09 22:00:19]

Patchset #1 (id:20001) has been deleted

[Sergey Ulanov, 2015-12-09 22:00:31]

sergeyu@chromium.org changed reviewers:
+ jamiewalch@chromium.org

[Jamie, 2015-12-10 18:34:15]

LGTM with one suggestion.

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
File remoting/host/host_status_logger.cc (right):

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
remoting/host/host_status_logger.cc:39: if (connected &&
connection_route_type_.count(jid) > 0)
Why this change? And why not just change the DCHECK to consider the >1 case?

[Sergey Ulanov, 2015-12-10 22:18:17]

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
File remoting/host/host_status_logger.cc (right):

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
remoting/host/host_status_logger.cc:39: if (connected &&
connection_route_type_.count(jid) > 0)
On 2015/12/10 18:34:15, Jamie wrote:
> Why this change? And why not just change the DCHECK to consider the >1 case?

Currently WebrtcTransport doesn't report route type so when it's being used
connection_route_type_.count(jid) returns 0 and the DCHECK was failing.

[Sergey Ulanov, 2015-12-10 22:18:48]

The CQ bit was checked by sergeyu@chromium.org

[commit-bot: I haz the power, 2015-12-10 22:20:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1506383004/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1506383004/40001

[Jamie, 2015-12-10 22:29:17]

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
File remoting/host/host_status_logger.cc (right):

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
remoting/host/host_status_logger.cc:39: if (connected &&
connection_route_type_.count(jid) > 0)
On 2015/12/10 22:18:17, Sergey Ulanov wrote:
> On 2015/12/10 18:34:15, Jamie wrote:
> > Why this change? And why not just change the DCHECK to consider the >1 case?
> 
> Currently WebrtcTransport doesn't report route type so when it's being used
> connection_route_type_.count(jid) returns 0 and the DCHECK was failing. 

Do you mean that with WebRTC LogSessionStateChange is called more often, and
connected_ might be true before there are any connection route types? Why the
change from ==1 to >0?

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
remoting/host/host_status_logger.cc:40: AddConnectionTypeToLogEntry(entry.get(),
connection_route_type_[jid]);
Is this line still correct, even if connection_route_type_.count(jid) > 1?

[Sergey Ulanov, 2015-12-10 22:48:05]

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
File remoting/host/host_status_logger.cc (right):

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
remoting/host/host_status_logger.cc:39: if (connected &&
connection_route_type_.count(jid) > 0)
On 2015/12/10 22:29:17, Jamie wrote:
> On 2015/12/10 22:18:17, Sergey Ulanov wrote:
> > On 2015/12/10 18:34:15, Jamie wrote:
> > > Why this change? And why not just change the DCHECK to consider the >1
case?
> > 
> > Currently WebrtcTransport doesn't report route type so when it's being used
> > connection_route_type_.count(jid) returns 0 and the DCHECK was failing. 
> 
> Do you mean that with WebRTC LogSessionStateChange is called more often, and
> connected_ might be true before there are any connection route types? Why the
> change from ==1 to >0?

No. With WebrtcTransport LogSessionStateChange() is never called so
connection_route_type_ is empty. Connection type cannot be reported in that
case.

https://codereview.chromium.org/1506383004/diff/40001/remoting/host/host_stat...
remoting/host/host_status_logger.cc:40: AddConnectionTypeToLogEntry(entry.get(),
connection_route_type_[jid]);
On 2015/12/10 22:29:17, Jamie wrote:
> Is this line still correct, even if connection_route_type_.count(jid) > 1?

connection_route_type_ is a map (not std::multimap) so count() can return only 0
or 1.

[commit-bot: I haz the power, 2015-12-10 23:08:15]

Committed patchset #1 (id:40001)

[commit-bot: I haz the power, 2015-12-10 23:10:10]

Description was changed from

==========
Enable WebRTC support in the remoting host.

WebRTC is enabled in the host when the host is started with
the --enable-webrtc flag. It also allows no-op authentication, so
it's only allowed in Debug builds.

BUG=547158
==========

to

==========
Enable WebRTC support in the remoting host.

WebRTC is enabled in the host when the host is started with
the --enable-webrtc flag. It also allows no-op authentication, so
it's only allowed in Debug builds.

BUG=547158

Committed: https://crrev.com/de15c037ab8425eecaa7ba7c3f63fc42e99d9f09
Cr-Commit-Position: refs/heads/master@{#364515}
==========

[commit-bot: I haz the power, 2015-12-10 23:10:11]

Patchset 1 (id:??) landed as
https://crrev.com/de15c037ab8425eecaa7ba7c3f63fc42e99d9f09
Cr-Commit-Position: refs/heads/master@{#364515}