Date: Mon, 6 Dec 1999 16:13:14 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Dave Bloodgood <dabldgd AT home DOT com>
cc: djgpp AT delorie DOT com, Peter Johnson <locke AT mcs DOT net>
Subject: Re: NT4 DPMI BUG (303h, Allocate Real-Mode Callback)--ATTN: DJ Deloie, Shawn Hargreaves, Eli Zaretskii
In-Reply-To: <004e01bf3ff3$23853b00$0201a8c0@pv1.ca.home.com>
Message-ID: <Pine.SUN.3.91.991206160744.21067E-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Mon, 6 Dec 1999, Dave Bloodgood wrote:

> Some months ago you ( Eli ) and I had a "discussion" about a similar problem
> with exception handling in windows NT - in particular, I can't install 
> a custom exception handler - without getting a similar sounding
> Dr Watson session. Is there a chance that this is a the heart of that
> problem also ?

It doesn't sound like this is the same problem.

IIRC, the problem you were having was with installing a signal handler 
(for SIGFPE, I think).  If I'm right, that problem was due to the fact 
that SIGFPE is generated by invalidating the segment whose selector is 
loaded into the DS register when DJGPP programs run, and NT seems to use 
that selector for some bookkeeping while it processes the exception 
caused by the invalid selector.

The problem in this thread is with defining a real-mode callback, not 
with fake exceptions used by the DJGPP signal-handling machinery.

However, if I'm mistaken, please remind what was the problem in your 
case.