[asterisk-bugs] [Asterisk 0010406]: [patch] Asterisk stops processing calls

[noreply at bugs.digium.com]

A NOTE has been added to this issue. 
====================================================================== 
http://bugs.digium.com/view.php?id=10406 
====================================================================== 
Reported By:                callguy
Assigned To:                Corydon76
====================================================================== 
Project:                    Asterisk
Issue ID:                   10406
Category:                   Channels/chan_sip/General
Reproducibility:            always
Severity:                   major
Priority:                   normal
Status:                     feedback
Asterisk Version:           1.4.11  
SVN Branch (only for SVN checkouts, not tarball releases): N/A  
SVN Revision (number only!):  
Disclaimer on File?:        N/A 
Request Review:              
====================================================================== 
Date Submitted:             08-08-2007 11:44 CDT
Last Modified:              10-01-2007 14:59 CDT
====================================================================== 
Summary:                    [patch] Asterisk stops processing calls
Description: 
Asterisk stops processing calls in the sip channel at random intervals.
This appears to be related to reloading chan_sip.so. When this happens the
console partially locks up (show channels becomes unresponsive) and shortly
after sip processing ceases to function. 

Running bt attached. 
====================================================================== 

---------------------------------------------------------------------- 
 Corydon76 - 10-01-07 14:59  
---------------------------------------------------------------------- 
Okay, this SECOND deadlock was exposed when we fixed the FIRST deadlock, so
both patches must be applied.  Note that the first patch has already been
applied to 1.4 SVN, so if you run SVN, you'll only need to apply the second
patch.

Basically what we've done is to remove the recursive use of ast_write,
doing deadlock avoidance on the two channels involved in a proxy, which
should permit locks to pass in the future.

I believe this only occurred when you performed an attended transfer to an
Agent channel. 

Issue History 
Date Modified   Username       Field                    Change               
====================================================================== 
10-01-07 14:59  Corydon76      Note Added: 0071272                          
======================================================================