[asterisk-dev] [Code Review] 2696: chan_pjsip: Improve initial INVITE handling, session reference counting, and CANCEL handling
Mark Michelson
reviewboard at asterisk.org
Wed Jul 24 16:24:33 CDT 2013
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviewboard.asterisk.org/r/2696/#review9208
-----------------------------------------------------------
Ship it!
/trunk/res/res_sip/sip_distributor.c
<https://reviewboard.asterisk.org/r/2696/#comment18159>
red blob
- Mark Michelson
On July 24, 2013, 9:23 p.m., Joshua Colp wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviewboard.asterisk.org/r/2696/
> -----------------------------------------------------------
>
> (Updated July 24, 2013, 9:23 p.m.)
>
>
> Review request for Asterisk Developers and Mark Michelson.
>
>
> Repository: Asterisk
>
>
> Description
> -------
>
> The attached change improves initial INVITE handling so it does not crash under certain circumstances.
>
> 1. The session reference count is now incremented when a reference is taken by INVITE session, previously it stole it.
> 2. Most initial INVITE handling now occurs as a serialized task within the session.
> 3. For cases where the remote side is broken and sends a CANCEL before receiving a provisional response we now detect and reject it.
>
>
> Diffs
> -----
>
> /trunk/channels/chan_gulp.c 395284
> /trunk/res/res_sip/sip_distributor.c 395284
> /trunk/res/res_sip_session.c 395284
>
> Diff: https://reviewboard.asterisk.org/r/2696/diff/
>
>
> Testing
> -------
>
> Tested normal scenarios with calls coming in, confirmed happy.
> Wrote a sipp scenario file to test CANCEL situation (before and after provisional response), confirmed happy.
>
>
> Thanks,
>
> Joshua Colp
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-dev/attachments/20130724/601830e3/attachment.htm>
More information about the asterisk-dev
mailing list