[jdev] Re: mu-conference core dumps (jcr 0.2.4 & muc 0.6.0)

[Nick Lee]

Hi All,

I posted a message earlier on muc coredumping. After some
troubleshooting, it seems muc is coredumping in
log_* functions, executing vnsprintf. I'm running on sol9.
Anyone seen this before. About to give up trying to get
muc to work :-(


495       vsnprintf (buffer, 1024, format, args1);
(gdb) where
#0  g_logv (log_domain=0x0, log_level=G_LOG_LEVEL_WARNING,
format=0x46de0 "%s: %s",
    args1=0xef420c78) at gmessages.c:495
#1  0xef6c7784 in g_log (log_domain=0x0, log_level=G_LOG_LEVEL_WARNING, 
    format=0x46de0 "%s: %s") at gmessages.c:574
#2  0x3de68 in log_warn (zone=0x619b0 "jcr_deliver.c:51 (jcr_queue_deliver)", 
    fmt=0x46f48 "packet delivery thread starting.") at jcr_log.c:112
#3  0x3e140 in jcr_queue_deliver (a=0x0) at jcr_deliver.c:51
#4  0xef6e4760 in g_thread_create_proxy (data=0x62028) at gthread.c:538
(gdb) list
490       
491       /* we use a stack buffer of fixed size, because we might get called
492        * recursively.
493        */
494     #ifdef  HAVE_VSNPRINTF
495       vsnprintf (buffer, 1024, format, args1);
496     #else   /* !HAVE_VSNPRINTF */
497       G_VA_COPY (args2, args1);
498       if (printf_string_upper_bound (format, FALSE, args1) < 1024)
499         vsprintf (buffer, format, args2);
(gdb) next

Program received signal SIGSEGV, Segmentation fault.
0xef5024e4 in _doprnt () from /usr/lib/libc.so.1
(gdb)