Date: Sat, 13 Jan 2001 19:58:38 +0200
From: "Eli Zaretskii" <eliz AT is DOT elta DOT co DOT il>
Sender: halo1 AT zahav DOT net DOT il
To: "Mark E." <snowball3 AT bigfoot DOT com>
Message-Id: <9628-Sat13Jan2001195838+0200-eliz@is.elta.co.il>
X-Mailer: Emacs 20.6 (via feedmail 8.3.emacs20_6 I) and Blat ver 1.8.6
CC: djgpp-workers AT delorie DOT com
In-reply-to: <3A604169.19279.186908@localhost> (snowball3@bigfoot.com)
Subject: Re: Where does gcc -o foo make foo.exe
References: <3A5F49FE DOT 11625 DOT 45D878 AT localhost> (snowball3 AT bigfoot DOT com) <3A604169 DOT 19279 DOT 186908 AT localhost>
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

> From: "Mark E." <snowball3 AT bigfoot DOT com>
> Date: Sat, 13 Jan 2001 11:52:09 -0500
> 
> > > Right. But DJGPP overrides LINK_COMMAND_SPEC in gcc's djgpp.h to use 
> > > djgpp.djl (which sets the target to coff-go32) and then call stubify on the 
> > > result. Any chance we can now finally agree to dispense with this?
> > 
> > Why would we want to dispense with that?
> 
> It would cut out an unneccessary step during linking.

If you mean the stubify step, then I think before we toss it, we need
to have a satisfactory solution to these two problems:

  - "gcc -o foo" should create both foo and foo.exe, to make Autoconf
    and Make happy;

  - the stub used to produce the .exe program should be the same stub
    distributed with the installed djdev (I didn't forget the GO32STUB
    feature, I just don't think it's reliable enough, because it
    depends on DJGPP.ENV).

It would also be nice to solve the problem with extra 2KB of garbage
in the executables produced by ld.exe, but that's a minor nuisance
(and maybe it's already solved, I don't remember).