So, my interpretation of this backtrace is that Kamailio simply crashed during what would otherwise be ordinary unwinding of TM hash table allocation. I suppose that means I should try to figure out why Kamailio shut down in the first place, i.e. from the logging.
On 04/23/2015 03:26 AM, Alex Balashov wrote:
Hello,
I just got this crash:
(gdb) where #0 0x00007fc3c1319625 in raise () from /lib64/libc.so.6 #1 0x00007fc3c131ae05 in abort () from /lib64/libc.so.6 #2 0x0000000000619d31 in fm_free (qm=0x7fc1b6ffa000, p=0x1e11a02400000000, file=0x7fc3bf8ebbfd "tm: h_table.c", func=0x7fc3bf8ebed8 "free_cell", line=162) at mem/f_malloc.c:588 #3 0x00007fc3bf8300a1 in free_cell (dead_cell=0x7fc1b71da540) at h_table.c:162 #4 0x00007fc3bf831705 in free_hash_table () at h_table.c:448 #5 0x00007fc3bf85577a in tm_shutdown () at t_funcs.c:123 #6 0x0000000000590d79 in destroy_modules () at sr_module.c:811 #7 0x000000000049bb43 in cleanup (show_status=1) at main.c:569 #8 0x000000000049d10b in shutdown_children (sig=15, show_status=1) at main.c:711 #9 0x000000000049f6e1 in handle_sigs () at main.c:802 #10 0x00000000004a6fbf in main_loop () at main.c:1757 #11 0x00000000004ab8bf in main (argc=11, argv=0x7fff1dbcda68) at main.c:2578
The fact that shutdown handlers were called makes me wonder if this is not so much a crash as a foreseen condition that I am not properly handling, like OOM. Can anyone shed any light?
Thanks,