[NTLUG:Discuss] X 3.3.6 + Redhat 6.2 = Memory Prob

[lee@brave.com]

$700!?!  Figuring out libraries?  Feh, that's too much work!  :-(

I also noticed that GAIM crashes right after startup.  I'm installing
the latest glib and gtk+ and see if that's the culprit.  Uhhh, I'm on
the right track, right?

Kevin Brannen wrote:
> 
> lee at brave.com wrote:
> >
> > Running X 3.3.6 on a relatively fresh install of RH 6.2.
> >
> > X consumes more and more memory over about a 8 hr period and then starts
> > to thrash swap; only way out is to kill X.
> >
> > I had been running RH 6.0/1 with the X 3.3.6 server for quite a while with
> > no problem like this, I'm sure there's something with the latest RH
> > version which is causing the problem.  ...
> >
> > Anyone have a clue about what might be going on with this?  Help?  :-)
> 
> I suppose "you have a memory leak" wouldn't be of much help would it?
> :-)  My X server is usually between 8-15M, and that's after a week of
> usage, with a little browsing and a lot of xterm/gvim work.
> 
> If it's the same X server, then it must be something else the X server
> depends on.  Look at the "What's new"/"Changes" list between the 6.0 &
> 6.2.  Maybe a new libc?  (like from glibc 2.0 to 2.1?  or a newer
> network lib?)  It'll be a leak in some supporting library.  The real
> way to find it would be to get a tool like Purify and use that on the
> X server (I think there's a company that makes one for Linux called
> Sentinal (~$700), IIRC but it's been a long time).  With such a tool,
> you'd have the answer pretty quickly.  There are other malloc debug
> tools out there.  If you have the desire to track this one down, get
> one of those, recompile your X server, run it for awhile (2-4 hours),
> then shut X down, and look at the log.  Hopefully it would give you a
> clue, but no guarantees like with Purify/Sentinal.
> 
> HTH,
> Kevin
> 
> _______________________________________________
> http://ntlug.org/mailman/listinfo/discuss