Subject: Re: machine check on 1.4
To: None <port-alpha@netbsd.org>
From: Kevin P. Neal <kpneal@pobox.com>
List: port-alpha
Date: 09/25/1999 15:24:18 
On Sat, Sep 25, 1999 at 11:35:55AM -0700, Jason Thorpe wrote:
> On Sat, 25 Sep 1999 13:08:41 -0400 
>  "Kevin P. Neal" <kpneal@pobox.com> wrote:
> 
>  > unexpected machine check
>  > 
>  > mces = 0x1
>  > vector = 0x670
>  > param = 0xfffffc0000006000
>  > pc = 0x120005444
>  > ra = 0x1200056d8
>  > curproc = 0xfffffc00010d67b8
>  > 
>  > pid = 12646, comm = gzip
>  > 
>  > panic: machine check
>  > Stopped in gzip at Debugger+0x4: ret zero,(ra)
>  > 
>  > (typed in by hand obviously)
> 
> I suspect you have a bad memory board, and what happened is that the
> hardware encountered an uncorrectable ECC error.  Note the PC and RA
> are both in userspace.
> 
> What kind of machine is this?

AlphaStation 200 4/233, about 2-3 years old. 

Is there a reasonable way to check the memory?
-- 
Kevin P. Neal                                http://www.pobox.com/~kpn/

Seen on bottom of IBM part number 1887724:
DO NOT EXPOSE MOUSE PAD TO DIRECT SUNLIGHT FOR EXTENDED PERIODS OF TIME.
u die). +The information then is pretty much diagnostic only, to aid in fault location and faulty FRU location. This is all designed to feed into the Error Logging stuff that Dompaq (that's my new name for "'Digital', uh, I mean 'Compaq'") holds *very* proprietary (they don't even ship the source for this with their OS source CD). +Dompaq service techs are only interested in the output of this error logging tool. Therefore, any work used to improve *BSD/Linux AXP kernel implementations would be to assist the those who do their own service. Ergo, it was a "would be nice, not high priority"... sorry about that..