[asterisk-dev] Possible race condition in Asterisk

Hans Petter Selasky hselasky at c2i.net
Wed Mar 21 09:17:25 MST 2007


Hi,

One of my clients had a crash with Asterisk 1.2.16 on NetBSD:

#0  0xbd9dc0bb in kill () from /usr/lib/libc.so.12
#1  0xbdbb22e5 in pthread__errorfunc () from /usr/lib/libpthread.so.0
#2  0xbdbaf3d9 in pthread_mutex_unlock () from /usr/lib/libpthread.so.0
#3  0xbd896caf in local_hangup (ast=0x805e684) 
at ../include/asterisk/lock.h:542
#4  0x088407c0 in ?? ()
(gdb)

The structure pointed to by "ast" looks like it is freed. At least it is not 
valid. Does anyone here know what this "chan_local" is used for, and what 
might have happened?

My client basically uses SIP, IAX2 and chan_capi.

--HPS


More information about the asterisk-dev mailing list