Xref: news2.mv.net comp.os.msdos.djgpp:1293
From: <SANDMANN AT arion DOT rice DOT edu>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: CWSDPMI on non-386
Date: Tue, 20 Feb 1996 20:58:52
Organization: Rice University, Houston, Texas
Lines: 14
Message-ID: <312a360c.SANDMANN@arion.rice.edu>
References: <4gcqsi$era AT superb DOT csc DOT ti DOT com> <3129fa3d DOT sandmann AT clio DOT rice DOT edu> <Dn34K3 DOT IIM AT cfanews DOT harvard DOT edu>
Reply-To: SANDMANN AT arion DOT rice DOT edu
NNTP-Posting-Host: arion.rice.edu
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp


> It's no big deal; the program wasn't going to run in any case.  
Well, it IS a big deal, a message is much better than a hang, especially
if you don't know why it hung.  It's just one of those things that fell
between the cracks.
> If this is fixed, will a djgpp program survive long enough on a non-386+
> system to attempt to call cwsdpmi and trigger a useful message?
Yes, the stub is 16-bit clean (in theory at least :-) up to the dpmi 
testing and loading stage.  At that point no suitable DPMI would be found,
so it would call CWSDPMI, which says "80386 required", so the DPMI load
fails, the stub sees no DPMI and says "no dpmi" and exits.
Since we spent some time discussing and testing all of this on an XT type
machine, it's very disappointing that I botched it someplace along the 
way :-(