[Webkit-unassigned] [Bug 169919] [WebRTC] Fix final mid of a transceiver

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=169919

Alejandro G. Castro <alex at igalia.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #305006|review?                     |review-
              Flags|                            |

--- Comment #2 from Alejandro G. Castro <alex at igalia.com> ---
Comment on attachment 305006
  --> https://bugs.webkit.org/attachment.cgi?id=305006
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=305006&action=review

Thanks for the patch, can you explain how we can reproduce this issue?

> Source/WebCore/ChangeLog:3
> +        [WebRTC] Fix final mid of a transceiver

Try to be more descriptive explaining the issue with the title, instead of explaining how you fix it, such as: Problem when setting final mid for the ...

> Source/WebCore/ChangeLog:9
> +        (WebCore::MediaEndpointPeerConnection::setLocalDescriptionTask):

Add here the comments about the changes you did.

> Source/WebCore/Modules/mediastream/MediaEndpointPeerConnection.cpp:374
>              RTCRtpTransceiver* transceiver = matchTransceiver(transceivers, [&mediaDescription] (RTCRtpTransceiver& current) {
> -                return current.provisionalMid() == mediaDescription.mid;
> +                return current.mid().isNull() && current.sender().trackKind() == mediaDescription.type;
>              });
>              if (transceiver)
> -                transceiver->setMid(transceiver->provisionalMid());
> +                transceiver->setMid(mediaDescription.mid);

I don't understand the solution, the matchTransceiver function makes sure the provisionalMid and the mediaDescription mid are the same value, and the transceiver exist, so it should be safe to assign the provisionalMid in that situation. Also your change of the matchTransceiver function is not matching correctly because there could be multiple tracks with the same type.

I think if the payload of a mediaDescription does not match and their mids match the problem should be somewhere else.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20170322/9e4ddfc1/attachment.html>