Subject: Re: 1.5 Release documentation and Power-tec/PowerROM problems
To: Chris Gilbert <chris@buzzbee.freeserve.co.uk>
From: Kjetil B. Thomassen <kjetil@thomassen.priv.no>
List: port-arm32
Date: 11/05/2000 21:33:04
On Sat 04 Nov, Chris Gilbert wrote:
> On Sat, 04 Nov 2000, Mike Pumford wrote:
> > > > No this is a different problem. PR 10539 applies to cards which sti=
ll
> > > > have the original ROMS. The PowerROM issue is different as the Powe=
rROM
> > > > uses the same device ID's as PowerTec SCSI cards and then misprobed=
 by
> > > > the kernel. Som additional work in the podulebus probe code will be
> > > > required to deal with this issue. I don't see this making it into 1=
.5.
> > >
> > > It has now been proven that it does not fix that problem as expected.
> >
> > Maybe a FAQ update indicating what to do if this occurs would be good. =
I
> > think an initial hack would be to have a podule override which would al=
low
> > an overide of the podule ID's when booting. As I already said a proper =
fix
> > is to beef up the podulebus probe code so that it can tell these cards
> > apart.
>=20
> You can pass in the boot option to force a card to be something, eg:
> podule1=3D0x00000002

Yes, this is correct. Manufacturer 0x0000, product 0x0002.

> will force podule 1 to be an Acorn SCSI 1 card, IE it don't probe it.

and, it also probes the Mikrotek ScanMaker E3 scanner and Philips
CDD2600 CD-R drive correctly on that Acorn card.

Excellent. Now we have a workaround for the PowerROM.

> See podules.h in sys/arch/arm32/podulebus for the list of podule cards an=
d
> manufacturers.  (Note I only found this out a week or so ago when delving=
=20
> into the guts of podulebus)

Excellent work.

> > > It may have been that kernel.
> > >
> > > Could you please compile a Risc PC install kernel for me so that I ma=
y
> > > be able to install NetBSD 1.5_BETA.
> > >
> > > After some more testing, I will raise a few PRs so that we can get on
> > > with getting more of the stuff working.
> >
> > I can indeed but it won't be available till at least Monday. Building a=
n
> > install kernel requires a full 1.5_BETA install and it will take a day =
or
> > so for my machine to build and install that.
>=20
> Yep and builds are painful on RPC's.  Perhaps if I can twist my kintic in=
to=20
> doing some things it might speed up (using a RAPIDE card has speeded lots=
 of=20
> stuff up, makes CVS only take about 30 or so minutes on the main src tree=
,=20
> perhaps I should time it)

Will NFS be able to improve the speed? I will set up another NetBSD box
with a 4x2GB CCD (10 ms drives), and if that is my SPARCstation 2, it
should deliver very fast performance on the 20 MHz SCSI bus.

I am patiently waiting for an install kernel that I can test.

What about netbooting a Risc PC from NetBSD/sparc or NetBSD/i386?
Is that possible, and if yes, how?

I'm getting quite a lot of ports around here in my 40 m=B2 flat, as I now
have:
- Acorn Archimedes A310
- Acorn R140
- Acorn Risc PC with SA-110
- SPARCstation 2
- 486DX/33
- HP 9000/712
- My portable work pc, a Dell Latitude LS-400

Not all of them are running NetBSD at the moment, but I expect to have
them up and running NetBSD in the months to come if possible.

Whenever Mike has an install kernel ready for me I can proceed with the
testing on NetBSD/arm32

Thanks for your help so far!

Kjetil B.
http://www.thomassen.priv.no/
http://home.eunet.no/~kjetilbt/MyRiscPC/
mailto:kjetil@thomassen.priv.no