Josh Hope wrote:
> 
> >Could you get a kernel trace after the next freeze?  In other words, hit
> >the programmer's switch to drop into the debugger and type 'trace' at the
> >'db>' prompt.  If this doesn't work, you've got something waiting with
> >interrupts off for some reason.
> 
> Yes, I could do a trace. Immediately after the trace, everything froze, 
> and the programmer's switch refused to work again. Do you need to see the 
> trace?

Uh, yes, please do.  Sorry I didn't make that clearer :-)

> >Is this machine networked?  If so, can you ping it?
> 
> Unfortunately, the answer to both questions is "no"

Oh well.

> >Perhaps with a little more info, we can figure this one out.
> 
> I certainly hope so! It's so strange to just "freeze" after a 6 day 
> uptime! And before that I had a 20 day uptime, broken only by a power 
> outage. So, as you can see, it was an EXTREMELY stable machine.

Hmmm...perhaps something was corrupted as a result of the power outage?
You have done a full fsck, haven't you?

> I already am missing NetBSD. I find it hard to live without. I'd really 
> like to get this rectified :)

:-)

> The weird thing is: I changed exactly nothing with my machine at least 
> 24-48 hours before it just started freezing. I didn't even download 
> anything, compile anything, etc. It was just so extremely strange!

Well, hopefully, the trace output will give someone a clue where to start
looking.  BTW, you probably need just the function names, I don't think
their arguments are all that useful (although please include them if you
have time).

Later.

-- 
Colin Wood                                 cwood@ichips.intel.com
Component Design Engineer - MD6                 Intel Corporation
-----------------------------------------------------------------
I speak only on my own behalf, not for my employer.