[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: f2c obsolete?
From: |
Michael Goffioul |
Subject: |
Re: f2c obsolete? |
Date: |
Wed, 31 Oct 2007 21:18:22 +0100 |
On 10/31/07, John W. Eaton <address@hidden> wrote:
> | Not really. An additional -subsystem:console flag needs also to be
> | passed to the linker. This is required to tell the linker that the entry
> | point is main and not WinMain. By default, it looks for WinMain
> | except if main is present in one of the object files. When compiling
> | a fortran program with f2c, the main symbol is part of the libf2c
> | (which calls MAIN__). This defeats the autodetection in the linker,
> | which looks for WinMain and fails.
> |
> | Where should this flag belong?
>
> I don't know and I'm a bit confused about when it is needed. Is it
> only necessary to use this option when compiling and linking a Fortran
> (only) program?
This is only needed when linking a fortran program; in the context
of octave, it would only be used by the configure script. However, I
chose to not touch octave configure script and to handle the additional
flags myself like:
F77=fort77 FFLAGS="-Wc,-MD -Wl,-subsystem:console" \
FLIBS=-lf2c ./configure
This works fine.
Michael.
- Re: f2c obsolete?, (continued)
- Re: f2c obsolete?, John W. Eaton, 2007/10/24
- Re: f2c obsolete?, Thomas Treichl, 2007/10/24
- Re: f2c obsolete?, John W. Eaton, 2007/10/24
- Re: f2c obsolete?, Thomas Treichl, 2007/10/25
- Re: f2c obsolete?, John W. Eaton, 2007/10/25
- Re: f2c obsolete?, Thomas Treichl, 2007/10/25
- Re: f2c obsolete?, Thomas Treichl, 2007/10/27
- Re: f2c obsolete?, John W. Eaton, 2007/10/31
- Re: f2c obsolete?, Michael Goffioul, 2007/10/25
- Re: f2c obsolete?, John W. Eaton, 2007/10/31
- Re: f2c obsolete?,
Michael Goffioul <=