Date: Thu, 17 Oct 1996 09:14:20 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To: "John M. Aldrich" <fighteer AT cs DOT com>
Cc: grendel AT ananke DOT amu DOT edu DOT pl, djgpp-workers AT delorie DOT com
Subject: Re: Install thingy
In-Reply-To: <3265A6E6.7F6@cs.com>
Message-Id: <Pine.SUN.3.91.961017090937.24334I-100000@is>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


On Wed, 16 Oct 1996, John M. Aldrich wrote:

> > Isn't Ctrl-Break handled by BIOS and intercepted by DOS which is, in turn,
> > intercepted by our signal handler?
> 
> As I recall, Ctrl-Break as handled by CWSDPMI completely kills the
> program, bypassing any signal handlers the program may have installed. 

This is incorrect.  Ctrl-Break winds up in the same signal handler as 
Ctrl-C, and therefore also causes the program to exit with the same exit 
code (unless you catch the signal, of course).  It is true that these two 
keystrokes are handled differently on a PC, but the DJGPP setup and the 
signal handling machinery take care to tunnel them both into the same 
place.  The only difference in their handling in the default handler is 
that Ctrl-Break causes a full stack trace to be printed, whereas Ctrl-C 
doesn't, but that's a feature designed to allow you to abort a stray 
program and know where did it go bananas.