Frederick Bruckman <fb@enteract.com> wrote:

>On Mon, 8 Feb 1999, Paul Goyette wrote:
>
>> It sounds like you're hitting the known "system hangs with multiple
>> disks" problem that has been plaguing the NetBSD/Mac68k kernel since
>> about last September or October.  Several people are desparately trying
>> to find this bug, but so far have been unsuccessful.
>
>However, Guy Santigla is running 1.3.2, so that can't be the problem,
>exactly.

Paul may be correct as this problem does occur in kernels prior to the 
Sept/Oct timeframe. I have a stock 1.3.2 system which is the last one 
that runs somewhat reliably on my system, but I can generate crashs and 
system hangs with it as well.  The crash traces are almost identicial to 
the ones I see with later kernels.

After poking around all weekend, building and crashing a dozen or so 
1.3.2 kernels trying to find this problem I've come to the conclusion 
that it's probably being caused by a stray SCSI disk I/O a stray 
interrupt that's not being properly serviced.  I've actually see the 
stray interrupt while using the ncrscsi driver but couldn't trace it; the 
sbcscsi driver never sees it even though it happens right before the 
system crashes though.

-bob