Thanks Rainer. I've attached the Valgrind log (rsyslog-7.3.10.log) from a version built from Git master as of almost exactly 24 hours ago, and built with the debug symbols intact this time.
I've also attached the GDB backtrace from the previous email (again this time with debug symbols), as I have it and it might be helpful. -- Kristian Van Der Vliet -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Rainer Gerhards Sent: 10 April 2013 13:15 To: rsyslog-users Subject: Re: [rsyslog] 7.3.10 (master) segfault in imuxsock The valgrind/debug log is great! Would it be possible to do one with git master branch? I hope to be able to llok at it this evening/tomorrow morning, and havoing both would be great. Rainer On Wed, 2013-04-10 at 12:09 +0000, Van Der Vliet, Kristian wrote: > >I've been working on an AMQP plugin for rsyslog, and had the basic > >version working against a version of rsyslog 7.3.6. Now I've had to > >move on to building and packaging my changes "properly" for our own > >use, so I've moved my changes into a copy of the current Git master > >branch (7.3.10). Everything builds just fine but I've got a segfault > >in imuxsock. The segfault is unrelated to my changes as it happens > >whether or not my omamqp plugin is enabled or not. > > I'm now ready to start sacrificing chickens. > > In order to eliminate a problem with the development build I moved > back to 7.2.6. While I'm here I've also tried both libestr 0.1.5 and > 0.1.4 just to eliminate any potential issues with libestr 0.1.5. > > I've pared the configuration file down to the absolute barest minimum > (attached) and rebuilt with --enable-memcheck & --enable-valgrind > > No matter what I've tried I still get a segfault in imuxsock as soon > as it receives the first message I.e. me running "logger" in a > different session. Valgrind indicates it's a null pointer dereference > in > MsgSetInputName() which makes no sense at all, and leads me to suspect > it's either a stack smash elsewhere or a really really obvious and > dumb problem I've overlooked. > > The Valgrind log of rsyslogd 7.2.6 (with libestr 0.1.5) running the > above configuration is also attached. > > Has anyone ever seen anything like this before or have any idea what > the problem may be? If I can get this fixed I can continue testing my > AMQP output plugin, and then hopefully get the patches upstream. > > _______________________________________________ > rsyslog mailing list > http://lists.adiscon.net/mailman/listinfo/rsyslog > http://www.rsyslog.com/professional-services/ > What's up with rsyslog? Follow https://twitter.com/rgerhards NOTE > WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE THAT. _______________________________________________ rsyslog mailing list http://lists.adiscon.net/mailman/listinfo/rsyslog http://www.rsyslog.com/professional-services/ What's up with rsyslog? Follow https://twitter.com/rgerhards NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE THAT.
rsyslog-7.3.10.log
Description: Binary data
rsyslog.bt
Description: Binary data
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________ rsyslog mailing list http://lists.adiscon.net/mailman/listinfo/rsyslog http://www.rsyslog.com/professional-services/ What's up with rsyslog? Follow https://twitter.com/rgerhards NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE THAT.
