[asterisk-dev] [Code Review] 2794: bridge_native_rtp: Fix bugs caused by HOLD/UNHOLD/UPDATE_RTP_PEER

[jrose]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviewboard.asterisk.org/r/2794/
-----------------------------------------------------------

Review request for Asterisk Developers, Joshua Colp, Matt Jordan, Mark Michelson, and rmudgett.


Bugs: ASTERISK-22217
    https://issues.asterisk.org/jira/browse/ASTERISK-22217


Repository: Asterisk


Description
-------

Issuing hold/unhold would lead to odd behavior. Between two chan_sip devices, a hold could cause an endless chain of updates while with pjsip a similar chain would begin but then end somewhat randomly. This patch fixes that by no longer tweaking the RTP glue on both sides of the call for every HOLD/UNHOLD/UPDATE_RTP_PEER frame.

Overall this restores holding behavior with chan_sip to how it worked with 11 (which includes some bugs I've noticed). chan_pjsip seems to be working as well, but it doesn't have the same bugs chan_sip exhibited with 11 (chan_sip does), which I guess is a good thing.

This patch doesn't fully address the test failures noted in the bug report. Asterisk 12 will needs a new version of the test with Asterisk 12 specific sipp scenarios to deal with that.


Diffs
-----

  /trunk/bridges/bridge_native_rtp.c 397424 

Diff: https://reviewboard.asterisk.org/r/2794/diff/


Testing
-------

Polycom to Polycom holds and unholds from both sides with directmedia using chan_sip
Polycom to Polycom holds and unholds from both sides with directmedia using chan_pjsip
Polycom to JITSI softphone same scenario (pjsip)
Polycom to Digium phone same scenario (pjsip)


Thanks,

jrose

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-dev/attachments/20130823/02b2323f/attachment.htm>