On Tue, May 23, 2017 at 11:41 PM, David Case <david.case.rutgers.edu> wrote:
> On Tue, May 23, 2017, Van Rompaey Dries wrote:
> >
> > Testing the gfortran compiler:
> > gfortran -fPIC -O0 -o testp testp.f
> > dyld: Library not loaded: @rpath/libgfortran.3.dylib
>
> Just a note: this indicates a problem with your Fortran compiler, not with
> Amber. "testp.f" is just a simple 2-3 line fortran program, and the
> configure
> script is finding that it cannot be compiled.
>
> > Installation through conda finishes fine, but I get errors referring
> > to libgfortran.3.dylib when I try to run sqm, so something’s not quite
> > right there either.
>
>
> .Hai: do you have thoughts on this?
>
I found that gcc and gfortran from conda are not that reliable for amber.
(my conclusion from several trials with them
to install ambertools). And since the gfortran suggested in amber website
just works fine (we have continuous integration for it on travis), so I
tend to always recommend it without trying to debug others :D.
> Does/should a conda install assume
anything about libgfortran?
As long as the suggested gfortran is installed (which is super easy though).
PS: Some links related to "dyld: Library not loaded:
.rpath/libgfortran.3.dylib"
1.
https://github.com/ContinuumIO/anaconda-issues/issues/739
2. This link suggest: brew install gcc49 --with-fortran
https://github.com/Homebrew/homebrew-science/issues/2286
Hai
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 23 2017 - 21:00:04 PDT
