[asterisk-bugs] [Asterisk 0014859]: [patch] lock is not released on channel masquarade

[Asterisk Bug Tracker]

A NOTE has been added to this issue. 
====================================================================== 
https://issues.asterisk.org/view.php?id=14859 
====================================================================== 
Reported By:                atis
Assigned To:                tilghman
====================================================================== 
Project:                    Asterisk
Issue ID:                   14859
Category:                   Functions/func_lock
Reproducibility:            always
Severity:                   minor
Priority:                   normal
Status:                     feedback
Asterisk Version:           1.4.19 
Regression:                 No 
SVN Branch (only for SVN checkouts, not tarball releases): N/A 
SVN Revision (number only!):  
Request Review:              
====================================================================== 
Date Submitted:             2009-04-08 10:39 CDT
Last Modified:              2009-08-11 19:54 CDT
====================================================================== 
Summary:                    [patch] lock is not released on channel masquarade
Description: 
Scenario: 

Queue -> Local channel -> answer_macro -> TRYLOCK -> bridge (masquarade)
-> TRYLOCK

First TRYLOCK creates lock correctly, but unlock is called outside of the
same thread, so it gives permission denied. Respectively lock remains even
if channel is dead or masqueraded, and second attempt to TRYLOCK gives
failure.


====================================================================== 

---------------------------------------------------------------------- 
 (0108924) atis (reporter) - 2009-08-11 19:54
 https://issues.asterisk.org/view.php?id=14859#c108924 
---------------------------------------------------------------------- 
Tilghman, that's exactly my point. what I'm trying to achieve is not allow
several instances of answer macro to execute simultaneously (another hard
issue around here)

It is not invalid to use it there, just current behavior is buggy. The
lock should either be moved to other channel upon masquerade - so that it
can be destroyed in correct thread (as I've done in my patch), or it should
be set on both channels and destroyed if one of them dies.

I haven't fully read Your patch and don't yet have idea how You are
solving it, but I can test it. 

P.S. It's hard to remember issue after several months :) 

Issue History 
Date Modified    Username       Field                    Change               
====================================================================== 
2009-08-11 19:54 atis           Note Added: 0108924                          
======================================================================