[asterisk-dev] [Code Review] Fix some issues in the provisional response keepalive code.

[schmidts]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviewboard.asterisk.org/r/1648/#review5095
-----------------------------------------------------------

Ship it!


looks good. just a very minor thing so the code would be easier to follow.


/branches/1.8/channels/chan_sip.c
<https://reviewboard.asterisk.org/r/1648/#comment9366>

    imho adding a comment here that this object will be removed from the scheduler after this function thats why the unref is called.
    it would make it easier to understand the concept 


- schmidts


On Dec. 31, 2011, 12:58 p.m., Russell Bryant wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviewboard.asterisk.org/r/1648/
> -----------------------------------------------------------
> 
> (Updated Dec. 31, 2011, 12:58 p.m.)
> 
> 
> Review request for Asterisk Developers.
> 
> 
> Summary
> -------
> 
> 1) Fix a sip_pvt reference leak.  If the scheduler callback returned 0, it did not release the reference to the associated sip_pvt.
>     
> 2) Add locking in the scheduler callback.  This code runs from the scheduler, which is executed in the SIP monitor thread.  A channel thread will be working with the sip_pvt at the same time.  Use sip_pvt_lock_full to ensure that the ast_channel/sip_pvt locking order is honored.
> 
> 
> This addresses bug ASTERISK-18979.
>     https://issues.asterisk.org/jira/browse/ASTERISK-18979
> 
> 
> Diffs
> -----
> 
>   /branches/1.8/channels/chan_sip.c 349445 
> 
> Diff: https://reviewboard.asterisk.org/r/1648/diff
> 
> 
> Testing
> -------
> 
> Currently running under heavy load on the reporter's production system.
> 
> 
> Thanks,
> 
> Russell
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-dev/attachments/20120102/fdb4f756/attachment.htm>