[asterisk-bugs] [JIRA] (ASTERISK-26197) chan_sip: Deadlock during hangup

[Oleg Schelicalnov (JIRA)]

    [ https://issues.asterisk.org/jira/browse/ASTERISK-26197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=243752#comment-243752 ] 

Oleg Schelicalnov edited comment on ASTERISK-26197 at 6/9/18 1:42 AM:
----------------------------------------------------------------------

I gets deadlock looks same which also happens in hangup. It easily reproduced when maxcalls variable is set to low value 200 or 300 calls and AMI originates call at 70 cps.

Version 13.21.0


was (Author: olegsch):
gdb info from deadlocked atersik

> chan_sip: Deadlock during hangup
> --------------------------------
>
>                 Key: ASTERISK-26197
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-26197
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: Channels/chan_sip/General
>    Affects Versions: 11.22.0, 13.18.4
>         Environment: CentOS release 6.7 (Final) x64
>            Reporter: Dmitrii kondrashin
>         Attachments: backtrace.txt, core-show-fd.txt, core-show-locks.txt, core-show-taskprocessors.txt, core-show-threads.txt, dumps.txt
>
>
> Periodically, the asterisk stop to handle call(SIP messages) correctly. Сorrect handling is restored only after restarting asterisk.
> It seems that the lock is created while asterisk tries to read from the tls socket, after that the thread crushes without release  of lock, and  all other threads are waiting for the release of this lock.
> Asterisk was compiled without srtp module and tls is disabled in asterisk's configs. 
> I've attached  the output of commands "core show locks", "core show taskprocessors", "core show threads", "core show fd", and backtrace-file prepared using gdb. 
> Build Options: DONT_OPTIMIZE, DEBUG_THREADS, LOADABLE_MODULES, DEBUG_FD_LEAKS



--
This message was sent by Atlassian JIRA
(v6.2#6252)