Re: [AMBER-Developers] can people try an AmberTools/Amber build using the current master branch?

From: Scott Brozell <sbrozell.comcast.net>
Date: Wed, 26 May 2021 21:08:31 -0400

Hi,

On Wed, May 26, 2021 at 02:40:54PM -0400, David A Case wrote:
> On Fri, May 21, 2021, Xiongwu Wu wrote:
>
> >I tried according to your procedure with cmake/3.18.0 and intel/17.0.1
> >compiler. Make install stopped with the error message listed below.
>
> >[ 28%] Building Fortran object AmberTools/src/sqm/CMakeFiles/sqm_common.dir/elementOrbitalIndex.F90.o
> >/u/wxw/ambers/amber20_src/AmberTools/src/sqm/elementOrbitalIndex.F90:12.6:
> >
> > use UtilitiesModule, only: Upcase
> > 1
> >Fatal Error: File 'utilitiesmodule.mod' opened at (1) is not a GNU Fortran module file
>
> This seems odd, since you are using Intel compilers. I know things work
> on more recent (2019) Intel compilers. Maybe someone on the list has
> experience with Intel 17.
>
> Can you try this: type "VERBOSE=1 make -j1 " after the cmake step. This
> will give more information about exactly what is going on.

./configure_cmake.py --prefix /tmp/junk/amber20 --compiler INTEL
intel/17.0.7 on Red Hat Enterprise Linux Server release 7.9 (Maipo)
worked for me: 134 serial exes built; no test errors; 9 unignored
test failures all of which look minor.

I don't see any abnormal messages wrt elementOrbitalIndex or utilitiesmodule.

scott



_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Wed May 26 2021 - 18:30:02 PDT
Custom Search