[Home]

Summary:ASTERISK-09127: segmentation fault in find_callno using iax modem
Reporter:Frank Sautter (xylome)Labels:
Date Opened:2007-03-28 05:06:44Date Closed:2007-07-09 21:20:43
Priority:CriticalRegression?No
Status:Closed/CompleteComponents:Core/General
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:( 0) segfault.txt
Description:after some random time asterisk crashes with segfault

****** ADDITIONAL INFORMATION ******

Core was generated by `asterisk -cvvvvv'.
Program terminated with signal 11, Segmentation fault.
#0  find_callno (callno=20660, dcallno=0, sin=0xb757433c, new=1, lockpeer=1, sockfd=9) at /usr/src/asterisk/include/asterisk/lock.h:535
535             __MTX_PROF(pmutex);

(gdb) bt
#0  find_callno (callno=20660, dcallno=0, sin=0xb757433c, new=1, lockpeer=1, sockfd=9) at /usr/src/asterisk/include/asterisk/lock.h:535
#1  0xb773b232 in socket_process (thread=0x81a92d0) at chan_iax2.c:6630
#2  0xb774539b in iax2_process_thread (data=0x81a92d0) at chan_iax2.c:7872
#3  0x080f1470 in dummy_start (data=0x81a0558) at utils.c:545
#4  0xb7e50240 in start_thread () from /lib/tls/i686/cmov/libpthread.so.0
ASTERISK-1  0xb7c7c3de in clone () from /lib/tls/i686/cmov/libc.so.6

(gdb) frame 1
#1  0xb773b232 in socket_process (thread=0x81a92d0) at chan_iax2.c:6630
6630                    fr->callno = find_callno(ntohs(mh->callno) & ~IAX_FLAG_FULL, dcallno, &sin, new, 1, fd);

(gdb) frame 2
#2  0xb774539b in iax2_process_thread (data=0x81a92d0) at chan_iax2.c:7872
7872                            socket_process(thread);

(gdb) frame 3
#3  0x080f1470 in dummy_start (data=0x81a0558) at utils.c:545
545             ret = a.start_routine(a.data);

(gdb) frame 4
#4  0xb7e50240 in start_thread () from /lib/tls/i686/cmov/libpthread.so.0

(gdb) frame 5
ASTERISK-1  0xb7c7c3de in clone () from /lib/tls/i686/cmov/libc.so.6
Comments:By: Serge Vecher (serge-v) 2007-03-28 08:41:55

please attach bts next time. Was DONT OPTIMIZE flag turned on under Compiler Options in menuselect? If not, you'll have to redo the bts.
By: Frank Sautter (xylome) 2007-03-28 09:48:26

compiled new svn-checkout (R59290) with "DONT_OPTIMIZE" flag on.
backtrace attached
By: Russell Bryant (russell) 2007-03-28 12:15:17

Is this trunk or the 1.4 branch?  The information int he bug report suggests both ...
By: Frank Sautter (xylome) 2007-03-29 01:29:20

this is revision 59290 svn-trunk as of yesterday. sorry for the confusion.


By: Russell Bryant (russell) 2007-03-29 12:35:43

The attached patch reverts some code that I introduced into chan_iax2 into the trunk that may be causing problems.  Please try this patch and let me know if it improves the situation.
By: Russell Bryant (russell) 2007-03-29 13:31:36

Actually, I just committed a change to this code in trunk that may fix your issue.   So, try the latest trunk without any patches and let me know how it goes.
By: Serge Vecher (serge-v) 2007-03-29 13:57:11

removing "naughty" patch and documenting the revision of the target commit.

in trunk as of rev. 59368
By: Frank Sautter (xylome) 2007-03-30 08:31:13

the new trunk version seems to be stable so far. i will test it further over the coming weekend.
By: Frank Sautter (xylome) 2007-04-02 03:02:24

i think this bug can be closed as the system was running stable throughout the weekend
By: Serge Vecher (serge-v) 2007-04-02 10:06:52

thanks for reporting back xylome.