*BSD News Article 25253


Return to BSD News archive

Newsgroups: comp.os.386bsd.questions
Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!elroy.jpl.nasa.gov!swrinde!cs.utexas.edu!howland.reston.ans.net!xlink.net!subnet.sub.net!ppcnet.ppc.sub.org!rumtifsl.ruessel.sub.org!strix.ruessel.sub.org!ura
From: ura@strix.ruessel.sub.org (Ulf Andrick)
Subject: Re: 386bsd and Cyrix 486DX40
Organization: Nocturnal Unix System in Kaiserslautern, Germany
References: <CI1x9I.6t@strix.ruessel.sub.org>
Message-ID: <CIEoED.42@strix.ruessel.sub.org>
X-Newsreader: TIN [version 1.1 PL8]
Date: Tue, 21 Dec 1993 22:13:10 GMT
Lines: 21

In <CI1x9I.6t@strix.ruessel.sub.org> Ulf Andrick (ura@strix.ruessel.sub.org) writes:
: I tried to run 386bsd without math emulation on a Cyrix 486DX40.

In fact, it was a package of Cyrix 486CX40 + Coprocessor, as far as I
understand now.

: The kernel was compiled with cpu i386 and options i387, as that was what
: I had before. The recognition of the FPU at boot time failed.
: Is there some known incompatibility of Cyrix 486DX to i486?
: Could the problem be solved by changing to an AMD processor?

The problem WAS solved by changing to an AMD 486DX40. The
performance as measured by Dhrystone 2.1 compiled with the
old gcc 1.?? is also better: 28 kDhrytones/s versus 21.
So, that Cyrix thing doesn't seem to be recommendable at all.

But what does an i486 give in comparison?

-- 
Ulf Andrick
ura@strix.ruessel.sub.org