Summary:ASTERISK-05966: Asterisk 1.2.1 making core dumps sporadically. Seems to be cause by IAX2?
Reporter:Kenneth Holm (saitech)Labels:
Date Opened:2006-01-03 16:32:32.000-0600Date Closed:2006-03-10 23:33:12.000-0600
Versions:Frequency of
Environment:Attachments:( 0) coredump.txt
( 1) coredump2.txt
( 2) sipdebug.txt
Description:Some times im getting a coredump, im unable to find out why and when. I have not been able to reproduce the error. Im getting a core dump every 48 hour, approx. But if im unlucky, i have got 2-3 core dumps in 1 hour. Mostly this is only happened fo a few times, but it seems like a certain think, that my box are doing core dumps after 48-72 hours.

Beside theese core dumps, my box is also hanging every 10 minute. In my debug log, im getting spammed with this error:

chan_sip.c: Failed to grab lock.. trying again

It hangs for about 5-15 seconds, where im unable to reveice or make any calls, netiher SIP nor IAX.
I have tracked this down in the sourcecode, and it seems to happen when im receiving a sip message, and im only getting 0 or 1 headers, and not 2. Im suspecting this to be a memory problem in my hardware, though i cant find any with HP's own memory test on the server.


Im using Asterisk 1.2.1 on a Gentoo Linux with a 2.6.14 kernel
My hardware is a HP DL360 G3
Comments:By: Tilghman Lesher (tilghman) 2006-01-03 16:47:29.000-0600

"Failed to grab lock" is not an error message; it's a debug message.

The crash you uploaded is a simple race condition, which I believe has already been reported elsewhere.

By: Kenneth Holm (saitech) 2006-01-03 16:53:23.000-0600

Though its only a debug message, the function triggering this debug message, is making my asterisk box to hang for quite some time. Do you have any idea why. I didn't have this problem on a similar server, with same setup. Though it was aterisk 1.2beta1

By: Kenneth Holm (saitech) 2006-01-04 01:43:49.000-0600

I have now compiled Asterisk with "dont-optimize". And ive got a new core dump. It seems to be different from the one before. Is this still a race condition? If so, could you please explain to me, what could cause a race condition, and what is a race condition?

By: Mark Spencer (markster) 2006-02-19 23:24:19.000-0600

Please confirm this bug is fixed in latest SVN trunk / 1.2

By: Tilghman Lesher (tilghman) 2006-03-10 23:33:12.000-0600

Reporter failed to respond in a timely manner.  Assuming fixed.  Reopen if this is still a problem.