[Serusers] SER coredumped in the final_response_handler

[szj]

   I used the SER 0.10.99 from CVS code. I have
tested it several days. But I found that it often
crashed due to unknow reason.
   I debugged it with the coredump ser produced.
it reports:

Loaded symbols for /lib/libnss_dns.so.2
#0  0x00a83755 in relay_reply (t=0xb60abbd8, p_msg=0xffffffff, branch=0, 
msg_status=408, cancel_bitmap=0xbffced88)
    at t_reply.c:996
996             DBG("DEBUG: relay_reply: branch=%d, save=%d, relay=%d, 
msg=%s\n",
(gdb) where
#0  0x00a83755 in relay_reply (t=0xb60abbd8, p_msg=0xffffffff, branch=0, 
msg_status=408, cancel_bitmap=0xbffced88)
    at t_reply.c:996
#1  0x00a771d8 in timer_routine (ticks=1104, attr=0x0) at timer.c:251
#2  0x080825d9 in timer_ticker () at timer.c:146
#3  0x0805ce12 in main_loop () at main.c:1097
#4  0x0805e5da in main (argc=3, argv=0xbffcef04) at main.c:1569
 
  Fortunately I started the SER with logging functions.
The log says:

 8(5775) DEBUG: add_to_tail_of_timer[7]: 0xb60abcf0
18(5775) DEBUG: retransmission_handler : done
18(5775) DEBUG: timer routine:0,tl=0xb60abd00 next=(nil)
18(5775) DEBUG: final_response_handler:stop retr. and send CANCEL 
(0xb60abbd8)
18(5775) ->>>>>>>>> T_code=180, new_code=408
27(5784)  0(5757) child process 5775 exited by a signal 11
 0(5757) core was generated
 0(5757) INFO: terminating due to SIGCHLD
19(5776) INFO: signal 15 received
19(5776) Memory status (pkg):
19(5776) qm_status (0x8108d20):
19(5776)  heap size= 1027664
19(5776)  used= 28000, used+overhead=55840, free=971824
19(5776)  max used (+overhead)= 58848

   I have two sip ua, the call must go through the SER.
It will crash about several setup procedure.

   I wonder what happens to the SER.

   Thanks for your kind help.
   Best Regards

Sun Zongjun