From: eplmst AT lu DOT erisoft DOT se (Martin Stromberg)
Message-Id: <199902121842.TAA27240@juno.erisoft.se>
Subject: Re: Carry flag
To: dj AT delorie DOT com (DJ Delorie)
Date: Fri, 12 Feb 1999 19:42:22 +0100 (MET)
Cc: djgpp-workers AT delorie DOT com (DJGPP-WORKERS)
In-Reply-To: <199902121833.NAA20598@envy.delorie.com> from "DJ Delorie" at Feb 12, 99 01:33:27 pm
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com

> > I'm not sure I follow you. Are you saying that _only_ those _three_
> > lines should be added to dpmi.h (and dos.h)?
> 
> That's the most commonly used flag, so yes.
> 
> > If so, then I don't understand why we shouldn't add the whole bunch of
> > them (parity, zero, sign, ...).
> 
> I think we shouldn't add those definitions *twice*.  We've gone many
> years without them at all, too.

Ok, let's put them in dpmi.h?

> > Personally I think the asm.h (not machine/asm.h) is the right way. Why
> > shouldn't dpmi.h include that file?
> 
> A whole new file just to define some flags?  I'd rather find a good
> place for them to live in an existing file.
> 
> Unless, of course, you want to create a <cpu.h> that has bitfield and
> structure definitions for *all* the cpu registers (flags, tss,
> descriptor tables, fpu, etc).

That sounds good. Considering how dpmi.h looks, most of it (the new
file) can be created by things moved in from dpmi.h. Or are there some
big complicated omissions? 


As I said, I'm not an assebly man, but I'm willing to create a first
version, which other persons who knows more can augment
(hint-hint-nugde-youknowwhatImean-wink-wink-hint).


Right,

							MartinS