Re: [AMBER] many failures on "make test"

From: Hai Nguyen <nhai.qn.gmail.com>
Date: Mon, 23 Sep 2019 18:03:26 -0400

On Mon, Sep 23, 2019 at 4:57 PM David Mathog <mathog.caltech.edu> wrote:

> On 2019-09-23 13:47, Hai Nguyen wrote:
> > On Mon, Sep 23, 2019 at 4:19 PM David Mathog <mathog.caltech.edu>
> > wrote:
> >
> >> Greetings.
> >>
> >> Just installed AmberTools which ended up in a directory amber18. We
> >> do
> >> not at this time have a license for Amber 18. Did this:
> >>
> >>
> >> TOPDIR=/usr/common/modules/el7/x86_64/software/amber/19-CentOS-vanilla
> >> mkdir -p $TOPDIR
> >> export PATH=/opt/ompi401/bin:$PATH #or it cannot figure out which
> >> mpi, fails in fftw
> >> ./configure --python-install local -mpi --with-python
> >> /usr/bin/python
> >> --prefix $TOPDIR gnu
> >>
> >
> > Please try again without the `--prefix` option. It might not setup
> > properly
> > for testing cpptraj.
>
> Is it possible to install to the --prefix directory after running the
> test?
>

Unfortunately not yet. For other questions, I will leave them to other
developers.

Cheers
Hai


> Also, I tried in a rebuild earlier to get it to see the netcdf installed
> on CentOS. (Because the first error was netcdf related, and I thought
> that might be the problem.) It wants
> the netcdf.mod file, which is here:
>
> /usr/lib64/gfortran/modules/netcdf.mod
>
> The only way I could get this to configure was by
>
> #as root
> ln -s /usr/lib64/gfortran/modules/netcdf.mod /usr/include/netcdf.mod
>
> then
>
> TOPDIR=/usr/common/modules/el7/x86_64/software/amber/19-CentOS-vanilla
> export PATH=/opt/ompi401/bin:$PATH #or it cannot figure out which
> mpi, fails in fftw s
> ./configure --python-install local -mpi \
> --with-python /usr/bin/python --prefix $TOPDIR \
> --with-netcdf /usr gnu
>
> all the other attempts constructed odd paths for the fortran test
> program which sees if netcdf is present. Is there a way to do that
> which does not require adding that link? The problem is that the build
> seems to expect a particular netcdf directory structure, and it can
> either find
> the mod file, or the headers and libraries, but not both!
>
> Thanks,
>
> David Mathog
> mathog.caltech.edu
> Manager, Sequence Analysis Facility, Biology Division, Caltech
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Sep 23 2019 - 15:30:02 PDT
Custom Search