About the compilation of V6.0.4

option, parallelism,...

Moderators: jbeuken, Jordan, pouillon

Forum rules
Please have a look at ~abinit/doc/config/build-config.ac in the source package for detailed and up-to-date information about the configuration of Abinit 8 builds.
For a video explanation on how to build Abinit 7.x for Linux, please go to: http://www.youtube.com/watch?v=DppLQ-KQA68.
IMPORTANT: when an answer solves your problem, please check the little green V-like button on its upper-right corner to accept it.
Post Reply
haiou
Posts: 12
Joined: Fri Dec 18, 2009 4:11 pm

About the compilation of V6.0.4

Post by haiou » Mon Jul 05, 2010 5:46 pm

Hi,
I get a problem when making compilation of abinit V6.0.4 on Xeon with Intel 10.1.
./configure FC=mpif90 CC=mpicc --enable-mpi=yes, and it works well. When I make it, the following error occurs,
Fortcom: Error: Xallgather_mpi.F90: This name does not have a type, and must have an explicit type
[MPI_COMM_NULL]

But I compile abinit V5.8.4 successfully with the above method on the same machine. Can anyone tell me why? Thanks a lot!

User avatar
jbeuken
Posts: 353
Joined: Tue Aug 18, 2009 9:24 pm
Contact:

Re: About the compilation of V6.0.4

Post by jbeuken » Wed Jul 07, 2010 1:57 pm

Hi,

I'm just tried with 6.0.4 on Xeon L5420 with intel 10.1 and openmpi 1.3.1

./configure --with-config-file=host.ac

in host.ac :

enable_optim="yes"
with_fc_optflags="-O2"
enable_mpi="yes"
with_mpi_prefix="/cvos/shared/apps/openmpi/intel/64/1.3.1/"
enable_gw_dpc="yes"
enable_nightly="yes"
enable_test_abirules="no"

make works...

regards,

jmb
------
Jean-Michel Beuken
Computer Scientist

haiou
Posts: 12
Joined: Fri Dec 18, 2009 4:11 pm

Re: About the compilation of V6.0.4

Post by haiou » Fri Jul 09, 2010 6:30 pm

Hi, jbeuken
I have tried your method but i didn't work. Thank you all the same.

User avatar
pouillon
Posts: 650
Joined: Wed Aug 19, 2009 10:08 am
Location: Spain
Contact:

Re: About the compilation of V6.0.4

Post by pouillon » Fri Jul 23, 2010 3:18 pm

You can try as well to pass the --with-mpi-level=1 option to configure.
Yann Pouillon
Simune Atomistics
Donostia-San Sebastián, Spain

Post Reply