On Sun, Nov 18, 2007 at 07:20:25AM -0800, Paul Goyette wrote:

> On Sun, 18 Nov 2007, Andrew Doran wrote:
> 
> >I think this is hardware related. What kind of CPU do you have? Here are
> >some of the nastier kernel bugs in -current I know about:
> >
> >- I've seen two reports of an assertion failure in uvm_fault.c.
> 
> I've had this happen on two completely different motherboards, so I am 
> quite sure it's not hardware related!

Right, I'm not suggesting that's hardware related. I think the issues that
Thomas is encountering could be.

> Repeatable for me just by exitting from an xterm window.  Doesn't matter
> if the X display is local or remote.  Funny thing, if I 'kill -KILL' the
> xterm process, it dies gracefully, but exitting from xterm kills the
> kernel.  Once I got a reference to sme_events_worker in the panic message,
> but that doesn't happen every time.  (Usually no additional messages at
> all.)
> 
> I've got the whole 4GB crash file and kernel available if anyone wants 
> to look into it further!  :)  (And gzipping the crash file saves very 
> little, less than 10%.)

I'll see about adding a bit more debugging around the code that is failing.

Thanks,
Andrew