ABINIT Discussion Forums

Posted: **Thu Aug 29, 2013 4:41 am**

Hi. I recently installed abinit-7.2.2 and abinit-7.4.1 on our linux cluster with openmpi-1.4.3-intel64-v12.0.4, but I always get the failure message on tests/v2:

Case_13 failed
Case_18 failed
Case_21 failed (too large absolute error : 3.599e-03 , accepted 2.510e-10 )

It seems that these tests have used ANADDB or related subroutines. By the way I have used the following compiling options:
CC="mpicc"
CXX="mpic++"
FC="mpif90"
F77="mpif77"
enable_mpi="yes"
enable_mpi_io="no"
with_mpi_level="2"
with_linalg_flavor="mkl"
with_linalg_incs="-I/opt/intel/composerxe-2011.4.191/include/intel64/lp64"
with_linalg_libs="-L/opt/intel/composerxe-2011.4.191/mkl/lib/intel64 -Wl,--start-group -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -Wl,--end-group -lpthread"
with_dft_flavor="atompaw+libxc+wannier90"
enable_gw_dpc="yes"

I have tried to use the "enable_64bit=yes" but the result is the same.

Thank you.

Posted: **Thu Aug 29, 2013 1:10 pm**

Hi,

with these soft versions :

intel 12.1.4 + mkl 10.3
openmpi 1.6.4

and

with this ac file

Code: Select all

enable_mpi="yes"
enable_mpi_io="yes"
with_mpi_prefix="/home/naps/buildbot/openmpi_intel"

with_trio_flavor="etsf_io+netcdf"
with_dft_flavor="atompaw+bigdft+libxc+wannier90"

with_fft_flavor="fftw3"
with_fft_libs="-L/usr/local/intel/composerxe/mkl/lib/intel64 -Wl,--start-group  -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -Wl,--end-group"
with_linalg_flavor="mkl"
with_linalg_libs="-L/usr/local/intel/composerxe/mkl/lib/intel64 -Wl,--start-group  -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -Wl,--end-group"

enable_clib="yes"
enable_gw_dpc="yes"

all tests succeed

ABINIT Discussion Forums

Test failure on ANADDB

Test failure on ANADDB

Re: Test failure on ANADDB