[OpenSIPS-Users] opensips tm timer core dump

[Kennard_White at logitech.com]

Hi Bogden,

Thanks for explaining the child processes involved -- I misunderstood what
was happening.

Unfortunately, I don't have the core anymore. My recollection is that I
couldn't print anything useful due to compiler optimization. That said,
this should re-create pretty easily, and I'll get more dumps next time it
happens.

Regards,
Kennard



From:	Bogdan-Andrei Iancu <bogdan at voice-system.ro>
To:	OpenSIPS users mailling list <users at lists.opensips.org>
Date:	10/05/2010 01:41 AM
Subject:	Re: [OpenSIPS-Users] opensips tm timer core dump
Sent by:	users-bounces at lists.opensips.org



Hi Kennard,

The core was generated by process 22255:
    [22238]: INFO:core:handle_sigs: child process 22255 exited by a
signal 11

and this process also reported mem problems:
    [22255]: ERROR:tm:new_t: out of mem

Can you print the "tl" or "ptr" variables in frame 0?

Regards,
Bogdan

Kennard_White at logitech.com wrote:
>
> Running against opensips HEAD, I got a segfault in the tm timer code.
> I believe this is triggered by running out of shared memory.
>
>
> The stack trace:
>
> (gdb) where
> #0 0x00007fe8f8d96212 in insert_timer_unsafe (new_tl=0x7fe8f66337b0,
> list_id=WT_TIMER_LIST, ext_timeout=<value optimized out>) at timer.c:731
> #1 set_1timer (new_tl=0x7fe8f66337b0, list_id=WT_TIMER_LIST,
> ext_timeout=<value optimized out>) at timer.c:904
> #2 0x00007fe8f8d78ac8 in t_release_transaction (trans=0x7fe8f6633730)
> at t_funcs.c:122
> #3 0x00007fe8f8d808e5 in t_unref (p_msg=<value optimized out>)
> at t_lookup.c:1152
> #4 0x0000000000483ae5 in exec_post_req_cb ()
> #5 0x000000000046c1e4 in receive_msg ()
> #6 0x00000000004bc77c in udp_rcv_loop ()
> #7 0x000000000042de9c in main ()
>
> The offending code (I believe):
> if (tl->time_out==ptr->time_out) {
> tl->ld_tl = ptr->ld_tl
> ptr->ld_tl = 0;
> tl->ld_tl->ld_tl = tl; <-- SEG FAULT HERE (according to trace)
> } else {
> tl->ld_tl = tl;
> }
>
> Unfortunately, due to optimization I cannot dump anything useful, and
> I'm not convinced the actual fault is on the line indicated. Note that
> the core dump is not one of the processes that reported out of memory.
> Maybe one of the other processes left the timer list in a corrupt state?
>
> The log file:
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22255]:
> ERROR:tm:sip_msg_cloner: no more share memory
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22255]:
> ERROR:tm:new_t: out of mem
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22255]:
> ERROR:tm:t_newtran: new_t failed
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22254]:
> WARNING:core:fm_malloc: Not enough free memory, will atempt
defragmenation
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22254]:
> ERROR:tm:sip_msg_cloner: no more share memory
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22254]:
> ERROR:tm:new_t: out of mem
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22254]:
> ERROR:tm:t_newtran: new_t failed
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22238]:
> INFO:core:handle_sigs: child process 22255 exited by a signal 11
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22238]:
> INFO:core:handle_sigs: core was generated
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22238]:
> INFO:core:handle_sigs: terminating due to SIGCHLD
> Sep 29 11:43:36 org-sip01 /var/run/openser/opensips-pres[22256]:
> INFO:core:sig_usr: signal 15 received
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Users mailing list
> Users at lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>


--
Bogdan-Andrei Iancu
OpenSIPS Bootcamp
15 - 19 November 2010, Edison, New Jersey, USA
www.voice-system.ro


_______________________________________________
Users mailing list
Users at lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.opensips.org/pipermail/users/attachments/20101007/d30fe101/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: graycol.gif
Type: image/gif
Size: 105 bytes
Desc: not available
Url : http://lists.opensips.org/pipermail/users/attachments/20101007/d30fe101/attachment.gif