From:	sandmann AT clio DOT rice DOT edu (Charles Sandmann)
Message-Id:	<10109051343.AA18271@clio.rice.edu>
Subject:	Re: Win2K breakpoints
To:	eliz AT is DOT elta DOT co DOT il (Eli Zaretskii)
Date:	Wed, 5 Sep 2001 08:43:33 -0500 (CDT)
Cc:	djgpp-workers AT delorie DOT com
In-Reply-To:	<Pine.SUN.3.91.1010905130913.13798B-100000@is> from "Eli Zaretskii" at Sep 05, 2001 01:10:26 PM
X-Mailer:	ELM [version 2.5 PL2]
Mime-Version:	1.0
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

> On Tue, 4 Sep 2001, Charles Sandmann wrote:
> 
> > If we assume that NT flavors always have handles 0..3
> 
> Is that what happens with other DPMI hosts?  In other words, is it 
> reasonable to expect that?

CWSDPMI does 0..3.  Many others do also, but I can't say they all do.
No, it's not reasonable to expect this (I haven't tested it either).
But it's not reasonable to expect the DPMI provider to neglect to 
give you the handle either ...  The only down side of this fix would
be if the handle is bogus, and they don't issue handle numbers 0..3,
then we can't check or free the handles, and you would get the same
behavior we currently see under Win2K (we would replace one bogus 
handle with another).  But if it's already broken ...

The dbgcom.c code currently doesn't check/notify on failure status 
with a bad handle.  When I convert the second assembly block to calling
the standard DPMI calls I'll add those checks.

- Raw text -