Segmentation fault vasp 5 (serial version)
Posted: Sun Jan 30, 2011 8:00 pm
Dear Sir,
I used intel ifort compiler 12 (Intel(R) Fortran Compiler XE 12.0 Update 2 for Linux) and GotoBLAS2 to compile vasp 5.2.2 with the following makefile
FC=ifort -static
...
CPP = $(CPP_) -DHOST=\"LinuxIFC\" \
-Dkind8 -DCACHE_SIZE=12000 -DPGF90 -Davoidalloc -DNGXhalf \
# -DRPROMU_DGEMV -DRACCMU_DGEMV
...
# parallel goto version requires sometimes -libverbs
BLAS= ../GotoBLAS2/libgoto2.a
# LAPACK, simplest use vasp.5.lib/lapack_double
LAPACK= ../vasp.5.lib/lapack_double.o
...
Afte that, the compiled vasp can run for a small cell, e.g. hcp-AlN (4 atoms). But when I use it for calculation with more atoms (32 atoms). It print out the following error message,
> vasp522
vasp.5.2.2 15Apr09 complex
POSCAR found : 2 types and 32 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
POSCAR, INCAR and KPOINTS ok, starting setup
FFT: planning ...( 1 )
reading WAVECAR
entering main loop
N E dE d eps ncg rms rms(c)
Segmentation fault
>
Could you please give some suggestion?
Many thanks for your consideration of this matter.
Best,
Ruifeng
I used intel ifort compiler 12 (Intel(R) Fortran Compiler XE 12.0 Update 2 for Linux) and GotoBLAS2 to compile vasp 5.2.2 with the following makefile
FC=ifort -static
...
CPP = $(CPP_) -DHOST=\"LinuxIFC\" \
-Dkind8 -DCACHE_SIZE=12000 -DPGF90 -Davoidalloc -DNGXhalf \
# -DRPROMU_DGEMV -DRACCMU_DGEMV
...
# parallel goto version requires sometimes -libverbs
BLAS= ../GotoBLAS2/libgoto2.a
# LAPACK, simplest use vasp.5.lib/lapack_double
LAPACK= ../vasp.5.lib/lapack_double.o
...
Afte that, the compiled vasp can run for a small cell, e.g. hcp-AlN (4 atoms). But when I use it for calculation with more atoms (32 atoms). It print out the following error message,
> vasp522
vasp.5.2.2 15Apr09 complex
POSCAR found : 2 types and 32 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
POSCAR, INCAR and KPOINTS ok, starting setup
FFT: planning ...( 1 )
reading WAVECAR
entering main loop
N E dE d eps ncg rms rms(c)
Segmentation fault
>
Could you please give some suggestion?
Many thanks for your consideration of this matter.
Best,
Ruifeng