RE: SIGSEGV, Segmentation fault.

[Jan-Henrik Haukeland]

Igor, it will help if you can tell us a bit more about the
environment you have e.g. what kind of OS and glibc version 
you are using and which version of monit you are running or 
are you running the bleeding edge version from CVS? (I did 
some changes to the log.c file in CVS lately)
Even a copy of your monitrc control file would be nice.

I have not seen this problem before, so I'm curious about 
what and where the cause may be. (Well, it's almost certainly
in the vsnprintf statement that builds the message buffer sent
to syslog, this is also why syslog is mentioned in the 
stack-trace from gbd. So, the response you got from Drepper 
is probably correct, that is, the problem is most likely in the 
monit code.

But it's a strange problem because it must have turned up recently,
if not you will have reported this earlier I guess. So, have 
you done some strange things on your system lately, for instance 
upgraded to RedHat 9.0? I have heard that they have a new thread 
technology which are not backward compatible etc. Also a significant 
question; do you also get SIGSEGV if you are *not* running monit 
through the gdb debugger? 

In other words all info you can provide may help :)


Regards

Jan-Henrik

Ps. I'll be back on Friday and will give it a stab, if
it's not already fixed by then by the other guys.