New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 625293 link

Starred by 3 users

Issue metadata

Status: WontFix
Owner:
Closed: Apr 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 2
Type: Bug



Sign in to add a comment

WebRtc: Chrome should support Asymmetric payload types in re-invite.

Reported by rajsripe...@gmail.com, Jul 1 2016

Issue description

UserAgent: Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36

Steps to reproduce the problem:

WebRTC doesn’t support asymmetric payload types for 2 codecs DTMF and WideBand Comfort Noise (CN) in the re-invite.ie it doesn’t allow MP to use different numbers for these two payload types. This happens only on re-invite. The following diagram illustrates this problem.
Consider X and Y are the payload types used by WebRTC and A and B as the payload types used by NonWebRTC client.

WebRTC 					NonWebRTC
================================================
Offer----DTMF PT=X, Wideband CN PT=Y--------->
<--------DTMF PT=A, Wideband CN PT=B------Answer
<--------DTMF PT=A, Wideband CN PT=B------Reinvite

Chrome rejects the Re-invite as it doesn't like a different payload type then what it used. Strangely it doesn't reject the answer for the same reason.

What is the expected behavior?
Chrome accepts the payload types which are different than its own.

What went wrong?
Call is dropped.

Did this work before? No 

Chrome version: 51.0.2704.103  Channel: canary
OS Version: 10.0
Flash Version: Shockwave Flash 22.0 r0
 
Components: Blink>WebRTC
Labels: Te-NeedsFurtherTriage
Cc: blum@chromium.org hta@chromium.org
[triage] Blum, looks like a product question. Can you take a look? Leaving as bug for now (not sure if it's a feature request).

Comment 3 by blum@chromium.org, Jul 12 2016

Cc: pthatcher@chromium.org
From a conceptional point of view I don't understand why Chrome rejects the re-invite. The standards perspective might be different.

Comment 4 by hta@chromium.org, Aug 1 2016

Standards perspective: The inconsistency (accepting the answer and not accepting the re-invite) is wrong, if this is the only difference between them.

Can you attach printouts of the SDP in the offer, answer and re-invite, as well as the precise error that Chrome emits when asked to accept the re-invite? These should be available from Chrome's webrtc-internals page.
Labels: Needs-Feedback
[triage] rajsripeddi@ could you do the steps suggested in #4? Thanks!
[triage] rajsripeddi@: ping regarding the requested feedback.
Status: WontFix (was: Unconfirmed)
Closing in lack of feedback.
Cc: bernard....@gmail.com
Status: Unconfirmed (was: WontFix)
Bernard, can you provide the requested SDP traces?
Cc: juberti@chromium.org

Comment 10 by solm...@gmail.com, Oct 21 2016

Calling setLocalDescription on the original caller fails with:

DOMException: Failed to set local answer sdp: Session error code: ERROR_CONTENT. Session error description: Failed to set local audio description recv parameters..

The following fiddle reproduces the issue: https://jsfiddle.net/gn4wyk9z/4/

Comment 11 by hta@chromium.org, Oct 27 2016

Thanks for the wonderfully short reproduction script!

Comment 12 by hta@chromium.org, Oct 27 2016

When runnign the fiddle, the stderr log in tip-of-tree Chrome says:

[1:14:1027/103806:ERROR:webrtcvoiceengine.cc(1750)] telephone-event payload type changed.
[1:13:1027/103806:ERROR:channel.cc(850)] Failure in SetLocalContent with action 2
[1:13:1027/103806:ERROR:webrtcsession.cc(332)] Failed to set local answer sdp: Session error code: ERROR_CONTENT. Session error description: Failed to set local audio description recv parameters..


Labels: -Needs-Feedback
Status: Untriaged (was: Unconfirmed)
Components: -Blink>WebRTC Blink>WebRTC>Network

Comment 16 by tommi@chromium.org, Mar 13 2017

Owner: solenberg@chromium.org
Status: Assigned (was: Untriaged)
Fredrik - can you take a look and see if we should handle it or if failing is the right thing?

Comment 17 by hta@chromium.org, Mar 13 2017

This is a SDP negotiation thing, and we should handle it.
Labels: Needs-Feedback
I just tested and wasn't able to reproduce this. rajsripeddi@, can you see if it's working for you now? If not, could you attach the SDP?
Status: WontFix (was: Assigned)
Closing since I wasn't able to reproduce.

Sign in to add a comment