2002-11-26 03:35:26

by Patrick Finnegan

[permalink] [raw]
Subject: Linux Kernel on NCR Worldmark 5100 and 3450 Workstation

Does anyone know anything about getting a Linux kernel to boot on either
of these machines? The Worldmark 5100 is a massive multiprocessor Pentium
Pro machine with two MCA buses. The 3450 is a quad-Pentium workstation
with a single MCA bus.

Thanks for any info.

Pat
--
Purdue Universtiy ITAP/RCS
Information Technology at Purdue
Research Computing and Storage
http://www-rcd.cc.purdue.edu


2002-11-26 08:51:21

by Andrey Panin

[permalink] [raw]
Subject: Re: Linux Kernel on NCR Worldmark 5100 and 3450 Workstation

On Mon, Nov 25, 2002 at 10:45:27PM -0500, Patrick Finnegan wrote:
> Does anyone know anything about getting a Linux kernel to boot on either
> of these machines? The Worldmark 5100 is a massive multiprocessor Pentium
> Pro machine with two MCA buses. The 3450 is a quad-Pentium workstation
> with a single MCA bus.

You can ask James Bottomley <[email protected]>
about these machines, may be they have Voyager nature.

--
Andrey Panin | Embedded systems software developer
[email protected] | PGP key: wwwkeys.eu.pgp.net


Attachments:
(No filename) (563.00 B)
(No filename) (232.00 B)
Download all attachments

2002-11-26 16:25:26

by James Bottomley

[permalink] [raw]
Subject: Re: Linux Kernel on NCR Worldmark 5100 and 3450 Workstation

> Does anyone know anything about getting a Linux kernel to boot on
> either of these machines? The Worldmark 5100 is a massive
> multiprocessor Pentium Pro machine with two MCA buses. The 3450 is a
> quad-Pentium workstation with a single MCA bus.

The WorldMark was the new name for the voyager systems (I even forgot that
they changed it). A (somewhat contorted) boot process is documented here.

The 3450 should work flawlessly as long as it has either a buslogic or D700
SCSI controller (I use a relative of this, the 3455 for my SCSI development
work).

The 5100 should work, but you'll have trouble with the Q720 SCSI controller.
The patches I had to make this card work with the 53c7,8xx driver are defunct
in 2.5 (although they may work for 2.4). I had planned to try to leverage
Richard Hirst's parisc Zalon SCSI work to do this and trick the consistent
memory allocation routines to use the 2Mb memory window behind the MCA bridge,
but I haven't even begun that yet.

I'm just starting to look at support for the secondary microchannel in the
5100---If you follow the mca-sysfs thread on linux-kernel, you'll see all the
necessary abstractions. Unfortunately, I don't actually have a machine with
two MCA busses, only the specs, so I'm flying by simulator only.

The project and its patches are all available here:

http://www.hansenpartnership.com/voyager

James Bottomley