Summary:	ASTERISK-17890: Signal 11 segmentation fault 0x0809495e in ast_do_masquerade
Reporter:	David Brillert (aragon)	Labels:
Date Opened:	2011-05-19 08:18:26	Date Closed:	2011-07-27 12:58:24
Priority:	Critical	Regression?	No
Status:	Closed/Complete	Components:	Applications/app_meetme
Versions:	1.4.41	Frequency of Occurrence
Related Issues:
Environment:		Attachments:	( 0) debco-meetme-core_show_locks.txt ( 1) lambo-core_show_locks.txt ( 2) meetme_segfault_2011-05-17.txt ( 3) meetme-segfault-24-05-2011_gdb-backtrace.txt
Description:	Asterisk 1.4.41 segfault during meetme. ****** ADDITIONAL INFORMATION ****** bt, bt full, thread apply all full bt is attached in file
Comments:	By: David Brillert (aragon) 2011-05-19 08:23:12 paging/meetme segfault occurs at least once per week using various * versions up to 1.4.41 Backtrace is from non-optimized build. Lots of "out of bounds" in trace but I am hoping a root cause can be found without a valgrind dump.  This is a busy site where paging is often used leading to segfault but a valgrind trace would only be possible if I can lab this up and reproduce... By: Leif Madsen (lmadsen) 2011-05-19 14:26:31 Can you reproduce this on Asterisk 1.8? ~~~~~ Per the Asterisk maintenance timeline page at http://www.asterisk.org/asterisk-versions maintenance (bug) support for the 1.4 and 1.6.x branches has ended. For continued maintenance support please move to the 1.8 branch which is a long term support (LTS) branch. For more information about branch support, please see https://wiki.asterisk.org/wiki/display/AST/Asterisk+Versions By: David Brillert (aragon) 2011-05-19 14:29:13 Hey Leif, I thought 1.4.42 was not released yet?  Still in RC1... I'm not in any position to force upgrade to 1.8 yet, and without looking at the backtrace how can you be certain the problem does not apply to 1.8? By: David Brillert (aragon) 2011-05-25 09:28:06 Another site crashing during meetme paging.  Lots of crashes sometimes 4 per day. Backtrace is almost identical however lacks any signs of out of bounds errors so I am thinking I am lucky and not needing a Valgrind dump to debug. meetme-segfault-24-05-2011_gdb-backtrace.txt is uploaded By: David Brillert (aragon) 2011-05-25 09:38:29 Both sites extensively use park then meetme page for call transfers. I can see the park and page attempts in both backtraces. Both sites use T1 PRI. Site meetme_segfault_2011-05-17.txt uses ACD queuing extensively. Site meetme-segfault-24-05-2011_gdb-backtrace.txt has no ACD queues configured. I am documenting this relation because I see GDB references to app_queue.c in site meetme-segfault-24-05-2011_gdb-backtrace.txt which has no queues in any configuration files. By: David Brillert (aragon) 2011-05-25 09:56:09 Captured core show locks during meetme pages on both sites. core show locks was not captured during segfault.  Only captured to show locking during pages. Attached files for both sites: debco-meetme-core show locks.txt lambo-core show locks.txt By: David Brillert (aragon) 2011-05-25 10:00:30 I cannot test Asterisk 1.8 until our internal beta testing completes sometime near end of June.  All uploaded backtraces and info are for post Asterisk 1.41 SVN. By: Russell Bryant (russell) 2011-07-27 12:58:19.774-0500 Per the Asterisk maintenance timeline page at http://www.asterisk.org/asterisk-versions maintenance (bug) support for the 1.4 and 1.6.x branches has ended. For continued maintenance support please move to the 1.8 branch which is a long term support (LTS) branch. For more information about branch support, please see https://wiki.asterisk.org/wiki/display/AST/Asterisk+Versions If this is still an issue, please open a new issue so it can be re-triaged appropriately. Thanks!