To generate the OpenMP/MPI hybrid parallelized version, only thing you have to do
is to specify CC and LIB in the makefile in 'source'.
To proceed the installation of the OpenMP/MPI version, move to the directory,
'source', and specify CC and LIB in the makefile, for example,
as follows:
For icc
CC = mpicc -openmp -O3 -I/usr/local/include
LIB = -L/usr/local/lib -lfftw3 -llapack -lblas -lg2c -static
For pgcc
CC = mpicc -mp -O3 -I/usr/local/include
LIB = -L/usr/local/lib -lfftw3 -llapack -lblas -lg2c -static
Of course, the specification depends on your computer environment.
Also, it is noted that older versions of icc and pgcc do not support
the compiler option for OpenMP.
After specifying CC and LIB appropriately, then install as follows:
% make install
When the compilation is completed normally, then you can
find the executable file, openmx, in the directory, 'work'.
To make the execution of OpenMX efficient, you can change a compiler and
compile options appropriate for your computational environment,
which can generate an optimized executable file.
It should be noted that the compilation of only OpenMP without MPI is
also possible.