Abinit 6.8.1 autotests failures
Posted: Wed Aug 03, 2011 4:40 pm
I compile Abinit 6.8.1 with ifort 11.1 and mvapich 2.1.4 and MKL library. The compile succeeded, but there are always two tests failure.
unitary/tmp-test_Linux_20110803/report:Case_fourdp_01 failed
unitary/tmp-test_Linux_20110803/report:Case_fftgw_01 failed
All options are default except mpi and enable_64bits.
tfftgw_01.out:
tfourdp_01.out:
unitary/tmp-test_Linux_20110803/report:Case_fourdp_01 failed
unitary/tmp-test_Linux_20110803/report:Case_fftgw_01 failed
All options are default except mpi and enable_64bits.
tfftgw_01.out:
Code: Select all
.Version 6.8.1 of FFTPROF
.(MPI version, prepared for a x86_64_linux_intel11.1 computer)
.Copyright (C) 1998-2011 ABINIT group .
FFTPROF comes with ABSOLUTELY NO WARRANTY.
It is free software, and you are welcome to redistribute it
under certain conditions (GNU General Public License,
see ~abinit/COPYING or http://www.gnu.org/copyleft/gpl.txt).
ABINIT is a project of the Universite Catholique de Louvain,
Corning Inc. and other collaborators, see ~abinit/doc/developers/contributors.txt .
Please read ~abinit/doc/users/acknowledgments.html for suggested
acknowledgments of the ABINIT effort.
For more information, see http://www.abinit.org .
.Starting date : Wed 3 Aug 2011.
- ( at 17h 6 )
Tool for profiling and testing the FFT libraries used in ABINIT.
Allowed options are:
fourdp --> Test FFT transforms of density and potentials on the full box.
fourwf --> Test FFT transforms of wavefunctions using the zero-pad algorithm.
gw_fft --> Test the FFT transforms used in the GW code.
all --> Test all FFT routines.
>>> Enter the string defining the routines to be tested:
>>> Enter cutoff energy ecut in Hartree:
>>> Enter lattice vectors rprimd in Bohr:
Real(R)+Recip(G) space primitive vectors, cartesian coordinates (Bohr,Bohr^-1):
R(1)= 20.0000000 0.0000000 0.0000000 G(1)= 0.0500000 0.0000000 0.0000000
R(2)= 0.0000000 20.0000000 0.0000000 G(2)= 0.0000000 0.0500000 0.0000000
R(3)= 0.0000000 0.0000000 20.0000000 G(3)= 0.0000000 0.0000000 0.0500000
Unit cell volume ucvol= 8.0000000E+03 bohr^3
Unit cell volume ucvol= 8.0000000E+03 bohr^3
Angles (23,13,12)= 9.00000000E+01 9.00000000E+01 9.00000000E+01 degrees
Angles (23,13,12)= 9.00000000E+01 9.00000000E+01 9.00000000E+01 degrees
>>> Enter number of calls for each test:
For input ecut= 2.000000E+01 best grid ngfft= 90 90 90
max ecut= 2.498244E+01
==== FFT mesh ====
FFT mesh divisions ........................ 90 90 90
Augmented FFT divisions ................... 91 91 90
FFT algorithm ............................. 112
FFT cache size ............................ 16
For input ecut= 2.000000E+01 best grid ngfft= 90 90 90
max ecut= 2.498244E+01
==== FFT mesh ====
FFT mesh divisions ........................ 90 90 90
Augmented FFT divisions ................... 91 91 90
FFT algorithm ............................. 412
FFT cache size ............................ 16
For input ecut= 2.000000E+01 best grid ngfft= 90 90 90
max ecut= 2.498244E+01
==== FFT mesh ====
FFT mesh divisions ........................ 90 90 90
Augmented FFT divisions ................... 91 91 90
FFT algorithm ............................. 312
FFT cache size ............................ 0
tfourdp_01.out:
Code: Select all
.Version 6.8.1 of FFTPROF
.(MPI version, prepared for a x86_64_linux_intel11.1 computer)
.Copyright (C) 1998-2011 ABINIT group .
FFTPROF comes with ABSOLUTELY NO WARRANTY.
It is free software, and you are welcome to redistribute it
under certain conditions (GNU General Public License,
see ~abinit/COPYING or http://www.gnu.org/copyleft/gpl.txt).
ABINIT is a project of the Universite Catholique de Louvain,
Corning Inc. and other collaborators, see ~abinit/doc/developers/contributors.txt .
Please read ~abinit/doc/users/acknowledgments.html for suggested
acknowledgments of the ABINIT effort.
For more information, see http://www.abinit.org .
.Starting date : Wed 3 Aug 2011.
- ( at 17h 6 )
Tool for profiling and testing the FFT libraries used in ABINIT.
Allowed options are:
fourdp --> Test FFT transforms of density and potentials on the full box.
fourwf --> Test FFT transforms of wavefunctions using the zero-pad algorithm.
gw_fft --> Test the FFT transforms used in the GW code.
all --> Test all FFT routines.
>>> Enter the string defining the routines to be tested:
>>> Enter cutoff energy ecut in Hartree:
>>> Enter lattice vectors rprimd in Bohr:
Real(R)+Recip(G) space primitive vectors, cartesian coordinates (Bohr,Bohr^-1):
R(1)= 20.0000000 0.0000000 0.0000000 G(1)= 0.0500000 0.0000000 0.0000000
R(2)= 0.0000000 20.0000000 0.0000000 G(2)= 0.0000000 0.0500000 0.0000000
R(3)= 0.0000000 0.0000000 20.0000000 G(3)= 0.0000000 0.0000000 0.0500000
Unit cell volume ucvol= 8.0000000E+03 bohr^3
Unit cell volume ucvol= 8.0000000E+03 bohr^3
Angles (23,13,12)= 9.00000000E+01 9.00000000E+01 9.00000000E+01 degrees
Angles (23,13,12)= 9.00000000E+01 9.00000000E+01 9.00000000E+01 degrees
>>> Enter number of calls for each test:
For input ecut= 2.000000E+01 best grid ngfft= 90 90 90
max ecut= 2.498244E+01
==== FFT mesh ====
FFT mesh divisions ........................ 90 90 90
Augmented FFT divisions ................... 91 91 90
FFT algorithm ............................. 112
FFT cache size ............................ 16
For input ecut= 2.000000E+01 best grid ngfft= 90 90 90
max ecut= 2.498244E+01
==== FFT mesh ====
FFT mesh divisions ........................ 90 90 90
Augmented FFT divisions ................... 91 91 90
FFT algorithm ............................. 412
FFT cache size ............................ 16
For input ecut= 2.000000E+01 best grid ngfft= 90 90 90
max ecut= 2.498244E+01
==== FFT mesh ====
FFT mesh divisions ........................ 90 90 90
Augmented FFT divisions ................... 91 91 90
FFT algorithm ............................. 312
FFT cache size ............................ 0