Parallel version error for EM64t/Linux

Message

okuno · #1 Post by **okuno** » Tue Apr 25, 2006 11:47 am

Dear Vasp users.

I'm tryng to use Vasp for mpi version of Vasp4.6.28 on Linux EM64t platform
with FFTW. The fortran compiler is IFC8.0.
I could compile the Vasp for mpi, but when I run vasp, the calculation stopped
with the message,

p2_19530: p4_error: interrupt SIGSEGV: 11
rm_l_2_19531: (10.910156) net_send: could not write to fd=5, errno = 32
p2_19530: (10.910156) net_send: could not write to fd=5, errno = 32
p1_20798: p4_error: interrupt SIGx: 13

What is the problem for running the vasp?

Parts of my Makefils is as follow,

.SUFFIXES: .inc .f .f90 .F

SUFFIX=.f90

#-----------------------------------------------------------------------
# fortran compiler and linker
#-----------------------------------------------------------------------
FC=ifc
# fortran linker
FCL=$(FC)

#----------------------------------------------

FFLAGS = -FR -lowercase -assume byterecl -I/fftw3.0/include

#-----------------------------------------------------------------------

OFLAG=-O3 -xW -tpp7

OFLAG_HIGH = $(OFLAG)
OBJ_HIGH =

OBJ_NOOPT =
DEBUG = -FR -O0
INLINE = $(OFLAG)

BLAS=/opt/intel/mkl721/LIB/mkl721_em64/libmkl_solver.a /opt/intel/mkl721/LIB/mkl721_em64/libguide.a /opt/intel/mkl721/LIB/mkl721_em64/libmkl_em64t.a /opt/intel_fce_80/LIB/intel_fce_80/libsvml.a

LAPACK= /opt/intel/mkl721/LIB/mkl721_em64/libmkl_lapack.a

#-----------------------------------------------------------------------

LIB = -L../vasp.4.lib -ldmy \
../vasp.4.lib/linpack_double.o $(LAPACK) \
$(BLAS)

FC=mpif90
FCL=$(FC)

CPP_ = ./preprocess <$*.F | /usr/bin/cpp -P -C -traditional >$*$(SUFFIX)

CPP = $(CPP_) -DMPI -DHOST=\"LinuxIFC\" -DIFC \
-Dkind8 -DNGZhalf -DCACHE_SIZE=4000 -DPGF90 -Davoidalloc \
-DMPI_BLOCK=500 \
-DRPROMU_DGEMV -DRACCMU_DGEMV

#CPP = $(CPP) -DMPI -DNGZhalf -DMPI_CHAIN -DMPI_BLOCK=500

#-----------------------------------------------------------------------
# location of SCALAPACK
# if you do not use SCALAPACK simply uncomment the line SCA
#-----------------------------------------------------------------------

BLACS=$(HOME)/archives/SCALAPACK/BLACS/
SCA_=$(HOME)/archives/SCALAPACK/SCALAPACK

SCA= $(SCA_)/libscalapack.a \
$(BLACS)/LIB/blacsF77init_MPI-LINUX-0.a $(BLACS)/LIB/blacs_MPI-LINUX-0.a $(BLACS)/LIB/blacsF77init_MPI-LINUX-0.a

SCA=

#-----------------------------------------------------------------------
# libraries for mpi
#-----------------------------------------------------------------------

LIB = -L../vasp.4.lib -ldmy \
../vasp.4.lib/linpack_double.o $(LAPACK) \
$(SCA) $(BLAS)

# FFT: fftmpi.o with fft3dlib of Juergen Furthmueller
#FFT3D = fftmpi.o fftmpi_map.o fft3dlib.o

# fftw.3.0.1 is slighly faster and should be used if available
FFT3D = fftmpiw.o fftmpi_map.o fft3dlib.o /fftw3.0/lib/libfftw3.a

and so on.

#2 Post by **admin** » Tue Apr 25, 2006 1:23 pm

where does this segmentation fault occur?

okuno · #3 Post by **okuno** » Wed Apr 26, 2006 10:53 am

Dear admin.

Thank you for reply.

I can run the Vasp by myself.
I found previous

http://cms.mpi.univie.ac.at/vasp-forum/ ... .php?3.741

I changed the file fftmpiw.F as previous job's potst,
and I could run the parallel Vasp on Em64t with FFTW3.0.

% diff fftmpiw.F.orig fftmpiw.F
56c56,58
< INTEGER :: planx, plany, planz
---
> ! Integer kind for plans shou
ld be large enough for a pointer.
> integer, parameter :: pk = se
lected_int_kind(18)
> INTEGER(pk) :: planx, plany, planz