[asterisk-bugs] [Asterisk 0016432]: racecondition leading to deadlock in chan_local

[Asterisk Bug Tracker]

A NOTE has been added to this issue. 
====================================================================== 
https://issues.asterisk.org/view.php?id=16432 
====================================================================== 
Reported By:                fnordian
Assigned To:                
====================================================================== 
Project:                    Asterisk
Issue ID:                   16432
Category:                   Channels/chan_local
Reproducibility:            random
Severity:                   major
Priority:                   normal
Status:                     acknowledged
Asterisk Version:           SVN 
JIRA:                       SWP-528 
Regression:                 No 
Reviewboard Link:            
SVN Branch (only for SVN checkouts, not tarball releases): 1.6.1 
SVN Revision (number only!): 230432 
Request Review:              
====================================================================== 
Date Submitted:             2009-12-11 10:12 CST
Last Modified:              2009-12-14 12:29 CST
====================================================================== 
Summary:                    racecondition leading to deadlock in chan_local
Description: 
Hi,

there's a race in chan_local between local_hangup and queuing a dtmf
(ast_senddigit ... local_queue_frame), which can cause hanging channels and
a hanging ast_senddigit.

This is possible because local_digit_*-functions assume that they were
called on an unlocked channel, which is not always the cases, e.g.
manager_play_dtmf uses a locked channel to call ast_senddigit. With this
wrong assumption, deadlock_avoidance code is not called.
====================================================================== 

---------------------------------------------------------------------- 
 (0115212) lmadsen (administrator) - 2009-12-14 12:29
 https://issues.asterisk.org/view.php?id=16432#c115212 
---------------------------------------------------------------------- 
I've acknowledged this for now, but it will be up to a developer to
determine if this backtrace is useful or not -- typically without
DONT_OPTIMIZE enabled, the backtrace doesn't provide enough information. If
you can either reproduce this, or if it happens often enough that you can
enable DONT_OPTIMIZE until you can get this to happen again, that would be
useful. 

Issue History 
Date Modified    Username       Field                    Change               
====================================================================== 
2009-12-14 12:29 lmadsen        Note Added: 0115212                          
======================================================================