Hello,
the input is not very important, since nwchem from debian
crashes before parsing it and even before printing anything
to stdout.
Anyhow the first test I tried is nwchem.nw in nwchem-$version/src.
It is very simple:
#----------------------------------------------------------
start h2o
geometry autosym
O 0.0 0.0 -0.02
H -0.74 0.0 -0.76
H 0.74 0.0 -0.76
end
basis
H library cc-pvdz
O library cc-pvdz
end
driver
clear
end
scf; print low; end
task scf optimize
#-------------------------------------------------------------
Fulvio
On 06/11/2014 21:33, Michael Banck wrote:
Hi,
please also keep the Bug address CC'd.
On Thu, Nov 06, 2014 at 08:16:01PM +0100, fulvio ciriaco wrote:
today I compiled nwchem-6.5 with:
a. mpich and openblas: works up to first scf iteration of H2O test,
than fails
solving a linear system in openblas
b. mpich and nwchem own blas: tried a few tests, seems to work fine.
So, apparently it is a problem with either openmpi or scalapack.
Hopefully I will find some time to test the other options.
Hrm, ok. Did you attach the input file you are using already? I don't
remember seeing it so far.
Michael
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
