1250356 - Update WebRTC code to webrtc.org stable branch 49

Status: RESOLVED FIXED

(Core :: WebRTC, defect, P1)

Comment 1

[Randell Jesup [:jesup] (needinfo me)]

Attachment: extmap_direction_support_and_cleanup

Adds basic directionality support for extmaps, and also removes RID from the audio extmaps.  Uses SendOnly for both audio-level and RID; note that this means in firefox-firefox calls, RID will be rejected at the SDP level.  This does not fix that RID is enabled in non-simulcast calls.

Comment 2

[Byron Campen [:bwc]]

Comment on attachment 8795400 [details] [diff] [review]
extmap_direction_support_and_cleanup

Review of attachment 8795400 [details] [diff] [review]:
-----------------------------------------------------------------

::: media/webrtc/signaling/src/sdp/SdpHelper.cpp
@@ +756,5 @@
> +
> +           ((i->direction == SdpDirectionAttribute::Direction::kSendrecv ||
> +             i->direction == SdpDirectionAttribute::Direction::kRecvonly) &&
> +            (j->direction == SdpDirectionAttribute::Direction::kSendrecv ||
> +             j->direction == SdpDirectionAttribute::Direction::kSendonly)))) {

You can probably simplify this a lot by using the flags that make up the direction enum:

https://dxr.mozilla.org/mozilla-central/source/media/webrtc/signaling/src/sdp/SdpAttribute.h#184

Comment 3

[Randell Jesup [:jesup] (needinfo me)]

nico: your try had compile issues, and also didn't disable the test.
Try this:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=42c72403157ec15202d91a07135e92bb78922b6a

Comment 4

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/e10e9f0e5ca2
Webrtc updated to upstream branch 49; pull made 2016-02-22 by pkerr rs=jesup
https://hg.mozilla.org/integration/mozilla-inbound/rev/126348e718d0
rollup of changes for webrtc after applying webrtc.org v49 update r=pkerr,ng,pehrsons,etc

Comment 5

[Joel Maher ( :jmaher ) (UTC -8) (PTO'ish until July 5)]

as a note, there is a slight increase in the number of constructors measured during compilation:
== Change summary for alert #4650 (as of December 28 2016 00:45 UTC) ==

Regressions:

  2%  compiler_metrics num_constructors linux32 opt       97 -> 99.33
  2%  compiler_metrics num_constructors linux32 pgo       97 -> 99.25
  2%  compiler_metrics num_constructors linux64 pgo       97 -> 99.25
  1%  compiler_metrics num_constructors linux32 debug     178 -> 180.17
  1%  compiler_metrics num_constructors linux64 debug     178 -> 180.17

For up to date results, see: https://treeherder.mozilla.org/perf.html#/alerts?id=4650

We don't require anything here, this is just an FYI.

Comment 6

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/e10e9f0e5ca2
https://hg.mozilla.org/mozilla-central/rev/126348e718d0

Comment 7

[Ryan VanderMeulen [:RyanVM]]

I'm thinking this could probably use some QA testing at some point during the 53 cycle as well.

Comment 8

[Andrei Vaida [:avaida]]

(In reply to Ryan VanderMeulen [:RyanVM] from comment #7)
> I'm thinking this could probably use some QA testing at some point during
> the 53 cycle as well.

Jesup, is there anything manual QA can do here to help? We need to be aware of testing needs as soon as possible, especially since this is already in Beta 53.