delorie.com/archives/browse.cgi   search  
Mail Archives: cygwin/2000/06/16/20:01:23

Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT sourceware DOT cygnus DOT com>
List-Archive: <http://sourceware.cygnus.com/ml/cygwin/>
List-Post: <mailto:cygwin AT sourceware DOT cygnus DOT com>
List-Help: <mailto:cygwin-help AT sourceware DOT cygnus DOT com>, <http://sourceware.cygnus.com/ml/#faqs>
Sender: cygwin-owner AT sourceware DOT cygnus DOT com
Delivered-To: mailing list cygwin AT sourceware DOT cygnus DOT com
From: Chris Faylor <cgf AT cygnus DOT com>
Date: Fri, 16 Jun 2000 20:00:37 -0400
To: cygwin AT sourceware DOT cygnus DOT com
Subject: Re: New GCC available for testing (TRY 3) -- fixes -mno-cygwin problem
Message-ID: <20000616200037.A12597@cygnus.com>
Reply-To: cygwin AT sourceware DOT cygnus DOT com
Mail-Followup-To: cygwin AT sourceware DOT cygnus DOT com
References: <20000616173012 DOT A10540 AT cygnus DOT com> <VA DOT 00000439 DOT 0172a584 AT bk-satellite>
Mime-Version: 1.0
User-Agent: Mutt/1.2i
In-Reply-To: <VA.00000439.0172a584@bk-satellite>; from bkeener@thesoftwaresource.com on Fri, Jun 16, 2000 at 06:28:08PM -0400

On Fri, Jun 16, 2000 at 06:28:08PM -0400, Brian Keener wrote:
>>> Come on!  What does "fails" mean???  I am not psychic.
>
>You are correct, my mistake and apologies I seem to have missed a beat a 
>forgot to include some diagnostics - command line is as follows:
>c++ -c -mno-cygwin hello.cc
>c++ -o hello -mno-cygwin hello.o
>
>Program is as follows:
>
>#include <iostream>
>int main() {
>        cout << "Hello world!" <<endl;
>        return 0;
>}
>
>Output is as follows (at least the last screen I was able to see:
>
>/Usr/lib/gcc-lib/i686-pc-cygwin/2.95.2/libstdc++.a(iovfscanf.o)(.text+0x5
>d1):iovfscanf.c: undefined reference to `_imp___ctype_'
>/usr/lib/gcc-lib/i686-pc-cygwin/2.95.2/libstdc++.a(iovfscanf.o)(.text+0x8
>5c):iovfscanf.c: undefined reference to `_imp___ctype_'
>[snip]

Yep.  Those are standard errors resulting from the fact that libstdc++.a
is compiled via g++ rather than 'g++ -mno-cygwin'.

>Now as I said in one of my other messages - this may be the way it was 
>intended to work because if I replace:
>
>c++ -o hello -mno-cygwin hello.o
>with
>c++ -o hello -mno-cygwin -L/usr/lib/gcc-lib/i386-mingw32/2.95.2 hello.o
>
>the errors above do not occur and I get an executable program, which as I 
>said I believe is understandable.

Yep.

>>> Btw, I am not interested in any other specs files than the one that I
>>> released.
>
>Kind of figured that out which is why my last few attempts have been 
>limited to your most recent specs file, although since some of what you 
>did (I thought was based on some input from Dale) I thought his specs 
>file might give a little insight and make it work.  Someone as ignorant 
>as myself and a non C/C++ developer that is only trying to learn had no 
>business trying to help in providing input in this anyways.

I had tried to use Dale's specs file but I've finally realized that it
didn't fit well with what I'm trying to do.  There are still pieces of
his stuff in the specs file that I released but some parts had to be
removed since his stuff seemed to use msvcrt a lot and that is not
currently part of the cygwin release.

Again, if either Mumit comes back and wants to fix this or someone else
wants to volunteer to handle the gcc release, I'll be happy to give them
my sources.  I am not interested in spending a lot of time trying to get
-mno-cygwin working in every conceivable configuration.  This is only on
the very periphery of what I am interested in and responsible for.

>My apologies and I will bow out now and let the gurus have at it.

Thanks for your reports.

cgf

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019