From: sandmann AT clio DOT rice DOT edu (Charles Sandmann)
Message-Id: <10104260336.AA18836@clio.rice.edu>
Subject: Re: win2000/ntvdm/djgpp (fwd)
To: eliz AT is DOT elta DOT co DOT il
Date: Wed, 25 Apr 2001 22:36:21 -0500 (CDT)
Cc: theowl AT freemail DOT c3 DOT hu (The Owl), djgpp-workers AT delorie DOT com
In-Reply-To: <6480-Wed25Apr2001233942+0300-eliz@is.elta.co.il> from "Eli Zaretskii" at Apr 25, 2001 11:39:43 PM
X-Mailer: ELM [version 2.5 PL2]
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> > pmode int and pmode selector.
> 
> Having thought about this, I must admit this worries me a bit.  What
> will CWSDPMI do when the application issues a PM Int 21h with the PSP
> selector in EBX?  

Um... the wrong thing :-(

> According to what I see in CWSDPMI's sources, it
> simply chains to the real-mode handler of Int 21h, i.e. to DOS.  But
> DOS doesn't know what to do with a PM selector of our PSP, so it might
> become confused, with potentially disastrous consequences (file
> operations will begin to fail, memory allocations might not work,
> function 4Ch could hang, etc.).

Yes, this would be bad.

> So could we please talk about this some more?  Why do we need to issue
> a PM interrupt, and why do we need to use the PM selector of our PSP?
> Isn't __dpmi_int with the real-mode PSP segment good enough?

I'm guessing it needs to be the PM selector on the basis of the 
previous discussions.  A test would be required.  We could also just do
the setting if the DOS/Windows version made sense :-P