Date: Sat, 25 Aug 2001 10:18:43 +0300
From: "Eli Zaretskii" <eliz AT is DOT elta DOT co DOT il>
Sender: halo1 AT zahav DOT net DOT il
To: djgpp-workers AT delorie DOT com
Message-Id: <3405-Sat25Aug2001101842+0300-eliz@is.elta.co.il>
X-Mailer: Emacs 20.6 (via feedmail 8.3.emacs20_6 I) and Blat ver 1.8.9
In-reply-to: <200108250135.VAA16071@envy.delorie.com> (message from DJ Delorie
	on Fri, 24 Aug 2001 21:35:22 -0400)
Subject: Re: gcc-3.0.1 and Win2k
References: <Pine DOT A41 DOT 4 DOT 05 DOT 10108241840300 DOT 67980-100000 AT ieva06 DOT lanet DOT lv> <4634-Fri24Aug2001204242+0300-eliz AT is DOT elta DOT co DOT il> <200108250135 DOT VAA16071 AT envy DOT delorie DOT com>
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

> Date: Fri, 24 Aug 2001 21:35:22 -0400
> From: DJ Delorie <dj AT delorie DOT com>
> 
> > Fine with me, but the reason we use the script name is that we invoke
> > the linker directly in some cases, not through GCC.  Does anyone
> > remember why is this done, why don't we use GCC to link?  DJ?
> 
> I don't recall exactly, but it might be a combination of not trusting
> gcc to use our newly created files (and nothing else) and perhaps gcc
> not doing the right thing back then anyway.

Yes, I think that's it: we wanted to make sure the linker uses crt0.o
and libc.a from the source tree, not from the production
installation.  I think the same can be achieved with GCC, given a
suitable combination of -nostdlib, -L, and -lfoo options.