*BSD News Article 31028


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!constellation!mimbres.cs.unm.edu!ees1a0.engr.ccny.cuny.edu!scisun!psinntp!psinntp!news.intercon.com!panix!zip.eecs.umich.edu!newsxfer.itd.umich.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uknet!demon!minnow.render.com!news!dfr
From: dfr@render.com (Doug Rabson)
Newsgroups: comp.os.386bsd.development
Subject: Re: 386 w/o V86? (was Re: More Details on the 386BSD Release 1.0 CD-ROM)
Date: 26 May 94 09:38:45
Organization: RenderMorphics Ltd.
Lines: 20
Message-ID: <DFR.94May26093845@minnow.render.com>
References: <jmonroyCq1qK0.5vJ@netcom.com> <1994May25.040649.28711@njitgw.njit.edu>
	<jmonroyCqCrwC.L5K@netcom.com> <2rvq8a$9h7@inews.intel.com>
	<2rvtq4$dha@omega.gmd.de>
NNTP-Posting-Host: minnow.render.com
In-reply-to: veit@borneo.gmd.de's message of 25 May 1994 16:19:16 GMT

In article <2rvtq4$dha@omega.gmd.de> veit@borneo.gmd.de (Holger Veit) writes:
> [...]
> 
> Today, they all have, but it is possible that some early versions of the chip
> had bugs concerning this mode, that reduced their usability w.r.t. v86 mode.
> The notorious 'Single Sigma' chip comes to mind, but this was a bug related to
> 32 bit operability.

I think that some of the early 386s had bugs in their 32 bit multiply
instructions.  I remember I had to upgrade a machine before I could
run Xenix386 on it.  Those were the days - running Unix and GNU Emacs
on a 1 megabyte machine (might have been two - my memory is a bit
dim).  It worked pretty well as long as you quit out from emacs before
trying to compile anything.


--
Doug Rabson, RenderMorphics Ltd.	Mail:  dfr@render.com
					Phone: +44 71 266 5090
					FAX:   +44 71 266 1623