Message-Id:	<199612272249.XAA13864@math.amu.edu.pl>
Comments:	Authenticated sender is <grendel AT ananke DOT amu DOT edu DOT pl>
From:	"Mark Habersack" <grendel AT ananke DOT amu DOT edu DOT pl>
Organization:	Home, sweet home (Poznan, Poland)
To:	"John M. Aldrich" <fighteer AT cs DOT com>
Date:	Fri, 27 Dec 1996 23:48:47 +0100
MIME-Version:	1.0
Subject:	Re: DPMI incorporation...
Reply-to:	grendel AT ananke DOT amu DOT edu DOT pl
CC:	djgpp AT delorie DOT com

Once upon a time (on 27 Dec 96 at 0:31) John M. Aldrich said:

> > I forgot about the second reason, the end target machine is a 286.
> 
> Not possible with DJGPP, unless you completely rewrite libc and
> create a whole new set of machine description files to recompile
> gcc.  GNU does not support 16-bit hardware and never will.  The 286,
> while still capable of running in a 16-bit quasi-DPMI, is not a
> 32-bit processor and cannot be used like one.
DPMI was designed for i386+ only and it would never run on iAPX286- 
machines. DPMI heavily relies on features found only in i386+ 
machines (like paged memory).

> Of course, it is theoretically possible to make gcc support a 286,
> but it would take a LOT of work.  I don't know if anyone has managed
> to accomplish it.
I don't think it'd make sense, anyway. The best use for 286 CPUs 
today is in Tekram (or similar) cacheing controllers or in advanced 
monitors. Thus 286 has virtually been 'reduced' to a level of a 
processor in embedded systems (although i386EX is taking over the 
area).

****************************************************************
You ask for my love on a strength of a kiss,
But can't you just play for experience?
I'm a poet, I'm a minstral, I don't need your chains
Romance lies in ruin, let debauch you bewail, let it rain...
---
Visit http://ananke.amu.edu.pl/~grendel

- Raw text -