Summary:ASTERISK-15476: Segfault under
Reporter:Iñaki Baz Castillo (ibc)Labels:
Date Opened:2010-01-20 04:56:57.000-0600Date Closed:2011-06-07 14:00:44
Versions:Frequency of
Environment:Attachments:( 0) gdb.out
Description:I've running Asterisk for more than 40 days with an aggresive SIPp scenario. No crashes at all.

Now it's in production with much less traffic (same configuration). After 3 weeks it has crashed:

kernel: asterisk[17723]: segfault at 0000000000000020 rip 000000000044a405 rsp 0000000041858320 error 6

I attach the complete gdb output.


I cannot upgrade to 1.4.28 due to existing issues with newer versions. I just would like to know if this segfault is a known issue. In that case it would be great if I could be pointed to the commit fixxing it.
Comments:By: Russell Bryant (russell) 2010-01-20 10:02:38.000-0600

There have been many releases since  Please try the latest version before reporting a crash.

By: Iñaki Baz Castillo (ibc) 2010-01-20 10:22:52.000-0600

Thanks for so useful reply. I already explained that I cannot upgrade the version and also that I tested this release during 5 weeks in a testing scenario with really high traffic all the day. This means that I *cannot* reproduce the problem so checking a new version is useless (except if I test a new version on production, not a very good idea...).

Also, which new version should I test? 1.4.26/27 in which chanspy and AMI don't work properly? every last 1.4.XX version has cool regressions which make it unsuitable for production.

I just wanted to know if this crash is a known bug, no more.

Thanks a lot.

By: Russell Bryant (russell) 2010-01-20 10:36:23.000-0600

1.4.29 is the latest version, not 1.4.28, 27, or 26.  If you are experiencing a problem with 1.4.29 that makes it so you can not use it, I would love to know about it so that we can get it fixed as soon as possible.