Subject: Re: Status of SCSI SYS_INST?
To: Terry Lambert <terry@lambert.org>
From: Jason Thorpe <thorpej@nas.nasa.gov>
List: port-hp300
Date: 12/09/1996 11:23:15
On Mon, 9 Dec 1996 10:47:11 -0700 (MST) 
 Terry Lambert <terry@lambert.org> wrote:

 > Is there any status on the SCSI SYS_INST?

First, lets clarify...

There is no such thing as the "SCSI SYS_INST".  SYS_INST is a single
program with both SCSI and HP-IB disk drivers.  It uses the same
SCSI disk driver as the boot program, SYS_UBOOT.

Secondly...

There is no updated status.  I sense you're upset because I posted about
some interrupt related changes, but haven't yet addressed "the SYS_INST
bug".  Rest assured that I'll post when I've had time to address it (see
below).

<gripe>
I'm getting a little frustrated, here.  While I understand that you
want to be able to install onto your SCSI disk, I feel compelled to
point out that I am a _volunteer_, and I work on what I can, when I have
time to do so.

Regarding the interrupt changes that I posted about yesterday, I had
a compelling need for them right then.  I've been spending late evenings
doing Real Job(tm) work from home (because I like to get out of my office
occasionally), and the fact that my PPP link got flaky when my backups
ran was actually a real hinderence.

Rest assured that when I have time to pull my HP disk enclosure apart,
somehow mount the 3.5'' disk in the 5.25'' bracket, put the enclosure
back together, and get my spare hp340 working again, I'll work on the
SYS_INST bug; I had hoped to do that this weekend, but I simply didn't
have time to do so.
</gripe>

Ok, now that I've got that off my chest...

You've been suggesting a method for dumping a "disk image" onto the disk.
While that's a fine idea, I don't particularly want the "disk image
including disklabel and boot block" method to be Offically Supported
by NetBSD, mainly because it locks you in to the disklabel that was
used by whomever created the image.  I consider that a non-solution.

Here's what I want to have...

A program that prompts the user for disklabel parameters (sort of like
SYS_INST does, but since this program will be in a hosted environment,
feel free to make it nicer), writes the disklabel to disk (be careful
about byte order and structure packing!), writes the boot block to disk
(be careful; the hp300 boot block is a bit tricky!), and then writes
the standard miniroot image to the `b' partition.

That's what I'd like to have.  It's really just a SMOP.  And, it can be
done on any system, at any time (including a FreeBSD i386 system, Terry :-).
However, I simply don't have the time to do it right now, and I'm counting
on contributors for this one.

Ok, thanks for listening...

Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939