The ABINIT Users Mailing List ( CLOSED )

[Katalin Gaal-Nagy <katalin.gaal-nagy@physik.uni-regensburg.de>]

Dear Matthieu,

I have a reply of CINECA where I asked for support compiling abinit on 
the SP5 (IBM cluster). The procedure is a little tricky (do the steps 
exactly in that order!).


**********************************************************
./configure CC=xlc FC=xlf90 CXX=xlC

Change on $ABINT_HOME/config/wrappers/wrap-fc:

# Utilities
#SED="@SED@"
SED="/usr/linux/bin/sed"

cp setup_kmesh_cpp.f90 ./src/15gw/.
(a file they gave me ...)
make
**********************************************************
(note: the file they gave me does not differ a lot from the original one, 
there is just a "," ...

 /scratch/cin8183a/abinit]$diff /scratch/cin8183a/abinit/
src/15gw/setup_kmesh_cpp.f90 /scratch/cne1fm2s/abinit/src/15gw/
setup_kmesh_cpp.f90
411c411
<   write(unt,'(2a)')ch10,&
---
>   write(unt,'(2a)'),ch10,&


The compilation went well, at the moment I am running the automatic tests, 
but it seems to be ok.

Thanks for the help!

All the best,
Katalin



On Tue, 6 May 2008, Dr MJ Verstraete wrote:

> Hi Katalin,
>
> IBM is always a pain - they never use the same flags or anything as everyone 
> else. You might try the following configure line. I don't know if it worked 
> in the end, but the compilation certainly finished:
>
> ./configure FC=mpxlf90_r FC_LDFLAGS_OPT="-bmaxstack:250000000 
> -bmaxdata:2000000000 -bloadmap:loadmap" FCFLAGS="-qsuffix=cpp=f" --enable-mpi 
> --with-mpi-cpp
>
> There is also the following in the config examples:
>
> /home/mjv500/CODES/ABINIT/abinit-5.6_mverstra_devel_public/config/build-examples/dirac-ibm.cf
>
> good luck and keep us posted with a working configure line!
>
>
> Matthieu
>
> ================================================================
> Dr. Matthieu Verstraete                 mailto:mjv500@york.ac.uk
> Dept. of Physics, University of York,     tel: +44 1904 43 22 08
> Heslington, YO10 5DD York, United Kingdom fax: +44 1904 43 22 14