[Pw_forum] Re: about parallel calculation under opteron linux cluster with mpich
Jian ZHOU
zjspam at gmail.com
Wed Jun 29 11:39:33 CEST 2005
Dear all,
Now I use the following make.sys. It works well for parallel pw.x. I
changed the LIBS and DFLAGS tags.
# make.sys. Generated from make.sys.in by configure.
CC = mpicc
CCFLAGS = -fast $(DFLAGS) $(IFLAGS)
# See include/defs.h.README for a list of precompilation options
# (possible arguments to -D or -U) and their meaning
DFLAGS = -D__LINUX64 -D__PGI -D__MPI -D__PARA -D__FFTW
-D__USE_INTERNAL_FFTW
FDFLAGS = $(DFLAGS)
IFLAGS = -I. -I../include -I../Modules -I../PW -I../PH
CPP = cpp
CPPFLAGS = -P -traditional $(DFLAGS) $(IFLAGS)
F77 = mpif77
F90 = mpif90
FFLAGS = -fast -r8 $(IFLAGS)
F77FLAGS = $(FFLAGS)
F90FLAGS = $(FFLAGS) $(FDFLAGS)
F77FLAGS_NOOPT = -O0
LD = mpif90
LDFLAGS = $(LIBOBJS) $(LIBS)
LIBOBJS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a
# LIBS must contain the location of all needed external libraries
LIBS = -llapack -lblas
# MYLIB can be one of the following (depending on LIBS):
# blas : compile the local copy of blas routines
# lapack : compile the local copy of lapack routines
# blas_and_lapack : all of the above - use this for a quick test
# or if you don't have an optimized blas/lapack library
# lapack_ibm : compile only lapack routines not present in IBM ESSL
# use this together with IBM ESSL
# lapack_t3e : compile only lapack routines not present in T3E scilib
# use this together with T3E scilib
# lapack_mkl : compile only lapack routines not present in Intel MKL
# use this together with Intel MKL
MYLIB =
AR = ar
ARFLAGS = ruv
RANLIB = echo
On 6/29/05, Jian ZHOU <zjspam at gmail.com> wrote:
> Dear all,
>
> I download the last version 2.1.4 of pwscf, and compile with the
> following the make.sys:
>
> # make.sys. Generated from make.sys.in by configure.
> CC = mpicc
> CCFLAGS = -tp k8-64 -fast -r8 $(DFLAGS) $(IFLAGS)
> # See include/defs.h.README for a list of precompilation options
> # (possible arguments to -D or -U) and their meaning
> DFLAGS = -D__LINUX64 -D__PGI -D__FFTW -D__USE_INTERNAL_FFTW
> FDFLAGS = $(DFLAGS)
> IFLAGS = -I. -I../include -I../Modules -I../PW -I../PH
> CPP = cpp
> CPPFLAGS = -P -traditional $(DFLAGS) $(IFLAGS)
> F77 = mpif77
> F90 = mpif90
> FFLAGS = -fast -r8 -tp k8-64 $(IFLAGS)
> F77FLAGS = $(FFLAGS)
> F90FLAGS = $(FFLAGS) $(FDFLAGS)
> F77FLAGS_NOOPT = -O0
> LD = mpif90
> LDFLAGS = $(LIBOBJS) $(LIBS)
> LIBOBJS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a
> # LIBS must contain the location of all needed external libraries
> LIBS = -llapack -lblas
> # MYLIB can be one of the following (depending on LIBS):
> # blas : compile the local copy of blas routines
> # lapack : compile the local copy of lapack routines
> # blas_and_lapack : all of the above - use this for a quick test
> # or if you don't have an optimized blas/lapack library
> # lapack_ibm : compile only lapack routines not present in IBM ESSL
> # use this together with IBM ESSL
> # lapack_t3e : compile only lapack routines not present in T3E scilib
> # use this together with T3E scilib
> # lapack_mkl : compile only lapack routines not present in Intel MKL
> # use this together with Intel MKL
> MYLIB =
> AR = ar
> ARFLAGS = ruv
> RANLIB = echo
>
>
> For the serial version(replace the mpif90 to pgf90), it works well.
> However, when I use this parallel version, it only created one thread
> even if I told it to run with 2 thread.
>
> I am using a pgi compiler(version 5.2) and mpich under an amd opteron cluster.
>
> When I use defaut make.sys file created by ./configure script, it
> shows a "Segmentation fault" error. So, I make some changes in the
> make.sys.
>
> Does anybody have run the parallel pwscf under opteron linux cluster with mpich?
>
> Thank you in advance.
>
> Best wishes,
>
> Jian
>
More information about the users
mailing list