www.delorie.com/archives/browse.cgi   search  
Mail Archives: cygwin/2008/12/04/14:35:48

X-Recipient: archive-cygwin AT delorie DOT com
X-Spam-Check-By: sourceware.org
Message-ID: <f79359b60812041135h98484bsded4b78ddb43c467@mail.gmail.com>
Date: Thu, 4 Dec 2008 14:35:02 -0500
From: "Gustavo Seabra" <gustavo DOT seabra AT gmail DOT com>
To: "Eric Blake" <ebb9 AT byu DOT net>
Subject: Re: gcc4/gfortran
Cc: cygwin AT cygwin DOT com
In-Reply-To: <49373437.9040203@byu.net>
MIME-Version: 1.0
References: <f79359b60812030638r7516868dn1524aae80d31b429 AT mail DOT gmail DOT com> <49373437 DOT 9040203 AT byu DOT net>
X-IsSubscribed: yes
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Id: <cygwin.cygwin.com>
List-Unsubscribe: <mailto:cygwin-unsubscribe-archive-cygwin=delorie DOT com AT cygwin DOT com>
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sourceware.org/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com

On Wed, Dec 3, 2008 at 8:36 PM, Eric Blake wrote:
> According to Gustavo Seabra on 12/3/2008 7:38 AM:
>> 1. Is is safe to remove the old gcc (3.*) packages and replace them by
>> symlinks to the new gcc4 executables?
>
> Read the archives.  Dave has mentioned that he is planning on a future
> packaging of the gcc packages that use the alternatives package, so that
> the symlink management of the name gcc can be done automatically to point
> to either gcc-3 or gcc-4.  But at the moment, I'm not sure whether the
> gcc-4 package requires files provided by the gcc package, in which case
> blindly deleting all thing gcc 3.* might break gcc-4.
>

Got it. But I was actually just planning on removing the gcc and g77
executables, and make those names point to gcc4 executables instead.

It actually has nothing to do with disk space: the whole point is
that, when compiling a program, I want to make sure it will *not* use
g77, but gfortran instead. The way it is now, I have to specify
gfortran-4 as the fortran compiler, say by using
FC=/usr/bin/gfortran-4, but one can never be sure exactly how a
specific 'configure' program will find its compilers. So, the removal
of gcc/g77 executables and replacing by a symlink would remove any
possibility for confusion.

>> 2. In this case, which executables should I point the symlink to? For
>> instance, if I were to replace g77 by a symlink to gfortran, which of
>> the 4 gfortran executables should I use:
>>
>>     $ locate gfortran | grep exe
>>     /bin/gfortran-4.exe
>>     /bin/i686-pc-cygwin-gfortran-4.exe
>
> These are identical copies; one is the name preferred when
> cross-compiling, the other when doing native compiles.

Got it, thanks.

> But why worry
> about adding symlinks?  Why not just rely on what the package gave you,
> since it works?  Are you really that low on disk space?  I suppose they
> could be made hardlinks to one another, if someone were to invest the time
> into patching setup.exe to attempt to make hardlinks (instead of its
> current behavior of blindly creating identical copies, even when the tar
> file specifies hardlinks).
>
>>     /usr/bin/gfortran-4.exe
>>     /usr/bin/i686-pc-cygwin-gfortran-4.exe
>
> These two are identical to the ones above - you need to read the manual,
> and remind yourself that /bin and /usr/bin are mount points that visit the
> same directory.  Removing /bin/gfortran-4.exe would simultaneously make
> /usr/bin/gfortran-4.exe disappear.
>
>>
>> 3. Lastly, just a dumb question: why do we get multiple executables in
>> the first place? I noticed that g77 also comes in multiple files:
>>     $ locate g77 | grep exe
>>     /bin/g77.exe
>>     /usr/bin/g77.exe
>>
>> Is that really necessary?
>
> Yes, because that's how the default mount points are set up.

OK, I had missed the point about /bin and /usr/bin actually pointing
to the same directory. Things are a lot clearer now.

Thanks,
-- 
Gustavo Seabra
Postdoctoral Associate
Quantum Theory Project - University of Florida
Gainesville - Florida - USA

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/

- Raw text -


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