Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!constellation!convex!convex!darwin.sura.net!howland.reston.ans.net!gatech!swrinde!network.ucsd.edu!news.cerf.net!crash!fredbox!cyb!loodvrij
From: loodvrij%cyb@fredbox.cts.com
Newsgroups: comp.os.386bsd.development
Subject: Re: porting to other architectures
Message-ID: <2a6k6B1w165w@cyb.cojones.com>
Date: Tue, 22 Jun 93 23:00:12 ADT
References: <1993Jun21.042648.1752@demo1.comlab.ox.ac.uk>
Distribution: world
Organization: The Cacophonous Yodellers' BBS - (907) 338-4942
Lines: 33
blt@comlab.ox.ac.uk (Brian Thompson) writes:
> I might also be interested in doing a port for the ARM series of processors,
> for which there isn't an affordable port. ARM based machines (i.e. the Acorn
> Archimedes) are very quick indeed, and are very cheap for the power also.
>
> A PD Unix for that box would be a good thing. Porting a version already
> ported to a 32bit machine with a linear address space and a paged memory
> mapping unit wouldn't do any harm either.
>
> Comments?
Yes, indeedy.
Back when I was in the UK, and had access to an arc, a friend and I made a
start at this very project. We got about as far as almost-but-not-quite
porting gas (after all, one needs an assembler), and making a cross-compiler.
I'll see if I can dig out the gas stuff if you like, I think I have it on
tape somewheres.
On the down side: The ARM is a *lovely* chip, but the MEMC just sucks. Only
128 pages. At 3 pages per process, that doesn't equal a hell of a lot of
resident processes...
Anyhow, let me know how you get on...
Bruce
--
Bruce J. Keeler (907) 337-8193 or 269-4253 | "We are the knights who say: |
Internet: loodvrij%cyb@fredbox.cts.com OR | COBOL !!!!!" |
Internet: loodvrij@cyb.cojones.com | [ PGP key available ] |
Geek code 0.3: G{CS,E} d-- -p+ c++ -l+ m* s++/+ !g w$$^% t+++ r- !x