[asterisk-dev] [Code Review] 3267: pjsip: avoid edge case potential crash in answer()
Joshua Colp
reviewboard at asterisk.org
Wed Feb 26 09:52:08 CST 2014
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviewboard.asterisk.org/r/3267/#review10958
-----------------------------------------------------------
/branches/12/channels/chan_pjsip.c
<https://reviewboard.asterisk.org/r/3267/#comment20585>
This log message doesn't fit in with the rest of the messages in PJSIP land, and this can occur for non-header reasons. To copy/paste my own commit message:
If a response to an initial incoming INVITE results in a transport error
the INVITE transaction is removed from the INVITE session. Any attempts
to answer the INVITE session after this results in a crash as it requires
the INVITE transaction to exist. This change explicitly locks the dialog
and checks to ensure that the INVITE transaction exists before answering.
- Joshua Colp
On Feb. 25, 2014, 7:45 p.m., Scott Griepentrog wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviewboard.asterisk.org/r/3267/
> -----------------------------------------------------------
>
> (Updated Feb. 25, 2014, 7:45 p.m.)
>
>
> Review request for Asterisk Developers.
>
>
> Repository: Asterisk
>
>
> Description
> -------
>
> When accidentally compiling asterisk against a rogue pjproject installation that had a slightly different definition pjsip_inv_session structure, the invite_tsx structure could appear null when answer() is called. This led to a crash because ast_sip_session_send_response would be called with an uninitialized packet.
>
> This patch corrects the uninitialized packet to prevent the crash, and adds a diagnostic message to aid in discovering the cause of the problem (in this case, remove /usr/local/include/pj* to resolve conflicting structure).
>
>
> Diffs
> -----
>
> /branches/12/channels/chan_pjsip.c 408931
>
> Diff: https://reviewboard.asterisk.org/r/3267/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Scott Griepentrog
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-dev/attachments/20140226/65e2bc42/attachment-0001.html>
More information about the asterisk-dev
mailing list