Lorenzo, I have done a change to the message handler, it is now using an atomic memory access operation where it previously did not. I have just pushed that change as part of the helgrind branch.
I would appreciate if you could give it a try and send me (via private mail) a helgrind log. I don't think that this change is sufficient to resolve the issue. But I would like to take a granular approach so that I can be sure to find the real root cause. Thanks, Rainer On Tue, 2008-09-23 at 14:04 +0200, Rainer Gerhards wrote: > ---------- Forwarded message ---------- > From: Lorenzo M. Catucci <[EMAIL PROTECTED]> > Date: Tue, Sep 23, 2008 at 2:00 PM > Subject: Re: Fwd: [Valgrind-users] helgrind points to race in > pthread_mutex_lock? > To: Rainer Gerhards <[EMAIL PROTECTED]> > > RG> Lorenzo, > RG> > RG> any chance you could do what is described below? (I guess "no", but I > RG> thought I ask ;)). > RG> > > It's running! I didn't recompile, since I didn't find anything new within > the helgrind branch. > > I'm enclosing the very first "race" message, let me know if this does seem > better. > > Thank you, and keep on course. Yours, > > lorenzo _______________________________________________ rsyslog mailing list http://lists.adiscon.net/mailman/listinfo/rsyslog
