On Mon, 21 Jun 2004 19:53:53 +0200, Marcel Janssen wrote: > Anyways, you're right that the glx stuff causes the problem. This is why > I disabled it in the first place (i didn't have a need for it anyways). > Would be nice if it could work though. It turns out not to be the glx module... or rather, not just that. Not loading the glx module just seemed to reduce the memory usage a little, staving off failure. I posted a bug about it: http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=126475 I don't know what's going on, but it does seem there's a nasty bug in there somewhere. Very strange. I'll see what comes of it, thanks for the help.