then you can try to avoid compiling libnetcdf.a entirely
by configuring it with ./configure --enable-static=no
and then only link to the so files, by changing amber's config.h from
NETCDF=$(INCDIR)/netcdf.mod
NETCDFLIB=$(LIBDIR)/libnetcdf.a
NETCDFLIBF=$(LIBDIR)/libnetcdff.a $(LIBDIR)/libnetcdf.a
to
NETCDF=$(INCDIR)/netcdf.mod
NETCDFLIB=$(LIBDIR)/libnetcdf.so
NETCDFLIBF=$(LIBDIR)/libnetcdff.so $(LIBDIR)/libnetcdf.so
On Tue, Oct 20, 2015 at 11:01 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
> > Begin forwarded message:
> >
> > From: David A Case <dacase.scarletmail.rutgers.edu>
> > Date: October 20, 2015 at 20:00:32 PDT
> > To: Ross Walker <ross.rosswalker.co.uk>
> > Subject: Re: [AMBER-Developers] Problems with AmberTools 15 and
> netcdf-fortran-4.2 compile
> >
> > On Tue, Oct 20, 2015, Ross Walker wrote:
> >>
> >> The odd thing is that netcdf-4.3.0 seems to compile fine, it is just
> >> the netcdf-fortan-4.2 that fails. How hard would it be to switch to
> >> netcdf-4.3.0 for everything?
> >
> > There is no netcdf-fortran-4.3, as far as I can tell. Released version
> is
> > 4.2. You could try what is on github, but that seems like a long shot.
> >
> > For Tim: this is not our code, so one would have to grok a pretty complex
> > build script (lots of autotools complexity) to replace libgfortran.a with
> > libgfortran.so. Still, might be worth some effort.
> >
> > ....dac
> >
>
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
>
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Tue Oct 20 2015 - 21:00:03 PDT
