I know that gcc8 is unsupported for cuda, but I believe that ambertools
doesn't need cuda for serial and parallel installations.
Using compatibility flag is possible workaround, but you should recompile
openmpi itself, thus eliminating auto-updates from upstream repo for this
package leaving control of update process to yourself. This is a bit
confusing and upstream patch for ambertools would be great.
пт, 15 февр. 2019 г., 3:54 Stephan Schott schottve.hhu.de:
> Hi Viktor,
> OpenMPI 4 worked for me by using gcc 7 and including the
> "--enable-mpi1-compatibility"
> during compilation. gcc 8 is still not a supported compiler for CUDA afaik,
> so you should use gcc 7 with amber anyway.
> Hope it helps,
>
> El jue., 14 feb. 2019 a las 19:36, viktor drobot (<linux776.gmail.com>)
> escribió:
>
> > Hello, dear AMBER developers and users!
> >
> > OpenMPI-4.0.0 released on 12.11.2018. This major update introduced
> > long-waited removals and cleanups proposed almost 23 years ago. Because
> > of this compilation of AmberTools ver. 18 (and, supposedly, 16 and
> > earlier) is broken now. Just tested on GCC8 compilers suite and Arch
> > Linux x64 (compilation with 3.1.3 version is fine). Due to deprecations
> > in OpenMPI standard some MPI-enabled modules such as sander, mdgx and
> > others are not able to be built.
> >
> > Some advices and possible fix-ups are described on FAQ page of OpenMPI
> > project (https://www.open-mpi.org/faq/?category=mpi-removed) but finding
> > and fixing things by hand is somewhat difficult and error-prone process,
> > developers know their own code much better. I wonder to know can we hope
> > that in nearest future some patches for AmberTools ver. 18 will be
> > released?
> >
> > Thank you!
> >
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
> --
> Stephan Schott Verdugo
> Biochemist
>
> Heinrich-Heine-Universitaet Duesseldorf
> Institut fuer Pharm. und Med. Chemie
> Universitaetsstr. 1
> 40225 Duesseldorf
> Germany
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Feb 14 2019 - 23:00:03 PST
