Well, if there's a code problem, it may be in userland, in RAIDframe.  When I used raidctl
directed at nonexistant partitions, it hung with a misdirected IRQ.  When I use it with the
proper /dev/sd*c partitions, it works great.  Newfs is wise enough to choke on /dev/sd1a
without IRQ errors.

Bottom line is that I can do a 'newfs raid0c' now, and it works great.  The bug would lie in
raidctl's error checking, right?

dhop

Jason Thorpe wrote:

> On Wed, 15 Dec 1999 10:07:44 -0800
>  David Hopper <dhop@nwlink.com> wrote:
>
>  > Okay, cool.  It looks like I've wasted everybody's time.  The IRQ was failing when
>  > processes were directed to /dev/sd1a and /dev/sd2a, but /dev/sd1c and /dev/sd2c refer
>  > to the disks, as they should under NetBSD-alpha.
>  >
>  > Next time I'll give it a week or so before I post  (hah!),
>  > dhop
>
> Huh?  What do you mean?  A non-existent partition shouldn't cause the lossage
> you were seeing...
>
>         -- Jason R. Thorpe <thorpej@nas.nasa.gov>