[asterisk-dev] [Code Review] Fix deadlock between subscription event RWLOCK and dialogs container lock in chan_sip.

[schmidts]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviewboard.asterisk.org/r/1557/#review4728
-----------------------------------------------------------

Ship it!


i have done several load tests and also rtptimeout tests and it works just fine.



/branches/10/channels/chan_sip.c
<https://reviewboard.asterisk.org/r/1557/#comment8910>

    copy and paste error ;)
    this is not the "callback to check rtptimeout and hangup calls if necessary" its more like "callback to check dialogs which need to be destroyed"


- schmidts


On Nov. 7, 2011, 4:39 p.m., rmudgett wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviewboard.asterisk.org/r/1557/
> -----------------------------------------------------------
> 
> (Updated Nov. 7, 2011, 4:39 p.m.)
> 
> 
> Review request for Asterisk Developers, David Vossel and schmidts.
> 
> 
> Summary
> -------
> 
> Timing between dialog destruction and a MWI event sending a message could result in a deadlock.
> 
> Order of events causing deadlock:
> 
> 1a) The event subscription system calls the registered callbacks with its list RWLOCK held.
> 1b) The SIP monitor checks for dialogs needing destruction.  It does an ao2_callback that holds the dialogs container lock while searching for dialogs to destroy.
> 2a) The event subscription SIP callback needs to create a temporary dialog to send out the MWI notification.  That temporary dialog needs to be inserted in the dialogs container so it must wait.
> 2b) The dialog search finds a dialog to destroy and as a result releases the last reference for a peer.  The peer destructor attempts to get the subscription RWLOCK but must wait.
> 3) deadlock
> 
> 
> Residual changes for Asterisk v10 branch after https://reviewboard.asterisk.org/r/1564/ commit
> and associated dialogs callid hash key change fix.
> 
> NOTE THE CHANGE FROM v1.8 to v10 BRANCH.
> 
> * Make check_rtp_timeout() return CMP_MATCH if need to delete dialog from
> dialogs_rtpcheck.  This is an optimization to avoid an unneeded
> lock/unlock and object search when using ao2_unlink.
> 
> * Prevent crash in check_rtp_timeout() if dialog->rtp is NULL.
> 
> * Make pvt_set_needdestroy() protect from possible double entries in
> dialogs_needdestroy.
> 
> Schmidts please note that change_callid_pvt() is different between v1.8
> and v10 for your unleas-the-beast branch.
> 
> 
> Diffs
> -----
> 
>   /branches/10/channels/chan_sip.c 343715 
> 
> Diff: https://reviewboard.asterisk.org/r/1557/diff
> 
> 
> Testing
> -------
> 
> It compiles. :)
> 
> 
> Thanks,
> 
> rmudgett
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-dev/attachments/20111108/b7e5c15c/attachment.htm>