Re: [AMBER] problems with NAB and SHIFTS

From: Jason Swails <jason.swails.gmail.com>
Date: Wed, 16 Jun 2010 09:15:51 -0400

On Wed, Jun 16, 2010 at 3:56 AM, Jakob Toudahl Nielsen <jtn.chem.au.dk>wrote:

> Thank you all for your help leading me through this very difficult
> installation!
>
> I been talking to our system administrator who suggested that it is
> best to use gfortran as the driver of ld. i.e. a typical
> link-statement could be:
>
> gfortran -o program.exe mod1.o mod2.o -L/path/to/library -lfoo -lbar
> - where it should be able to find the following:
> ls -l /path/to/library/libfoo.*
> ls -l /path/to/library/libbar.*
>
> I will also expand the path to the LD_LIBRARY_PATH, before I guess
> there was no specific info regarding gfortran:
>
> >> "echo $LD_LIBRARY_PATH":
> >>
> /com/intel/Compiler/11.1/064/mkl/lib/em64t:/com/intel/Compiler/11.1/064/lib/intel64
>

Why not use the intel compilers? You can use the iccvars.sh and
ifortvars.sh scripts to set your environment variables (like
LD_LIBRARY_PATH) for you.

Good luck!
Jason


>
> I need your help on where in the test/installation-scripts to apply
> these changes, i.e. where is "gfortran" and "ld" called. Should I
> change something in "/src/config.h"? such as modifying the FLIBS?
>
> ...
> FLIBS= $(LIBDIR)/arpack.a $(LIBDIR)/lapack.a $(LIBDIR)/blas.a -lgfortran
> FLIBSF= $(LIBDIR)/arpack.a $(LIBDIR)/lapack.a $(LIBDIR)/blas.a
> ...
>
> or
>
> ...
> # Information about Fortran compilation:
>
> FC=gfortran
> ...
>
> thanks in advance, best regards,
> Jakob
>
> information regarding versions:
>
> >> gcc --version
> gcc (GCC) 3.4.6 20060404 (Red Hat 3.4.6-10)
>
> >> gfortran --version
> GNU Fortran (GCC) 4.1.2 20071124 (Red Hat 4.1.2-42)
>
>
>
> Citing case <case.biomaps.rutgers.edu>:
>
> > On Tue, Jun 15, 2010, Jakob Toudahl Nielsen wrote:
> >
> >> gfortran is installed, running test for nab leads to the same error
> >> messages corcerning lgfortran:
> >>
> >> /usr/bin/ld: cannot find -lgfortran
> >
> > OK. First, there is no reason to work more with SHIFTS: you have to get
> nab
> > working first.
> >
> > You should look at the output of the configure script in AmberTools.
> Does
> > is say something like "Unable to compile mixed C/Fortran code"? This
> should
> > be a fatal error, but is not in the distributed version of configure. So
> > if your configure script is failing here, we need to get that fixed
> first.
> >
> > What do "gcc --version" and "gfortran --version" report? As was
> mentioned
> > in an earlier post, finding libgfortran.a, and setting LD_LIBRARY_PATH to
> > that directory, might fix this. (But if gcc and gfortran are arising
> from
> > the same package, that ought to automatically occurr...)
> >
> > ....dac
> >
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
>
> --
> Jakob Toudahl Nielsen, post doc
> Laboratory for Biomolecular NMR Spectroscopy
> inSPIN, Center for Insoluble Protein Structures
> Department of Chemistry, University of Aarhus
> Langelandsgade 140, DK-8000 Aarhus C, Denmark
> Phone: +458942 5528 or +452993 8502 (cell)
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>



-- 
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jun 16 2010 - 06:30:04 PDT
Custom Search