Page 1 of 1

ver8.2.2 make error related to BigDFT

Posted: Wed Apr 12, 2017 1:48 pm
by weitong
I am making Abinit ver 8.2.2, at the end of "make", I got this error:

-L/opt/intel/composer_xe_2015.1.133/compiler/lib/intel64/ -L/opt/intel/composer_xe_2015.1.133/mkl/lib/intel64/ -L/opt/intel/composer_xe_2015.1.133/ipp/../compiler/lib/intel64/ -L/opt/intel/composer_xe_2015.1.133/ipp/lib/intel64/ -L/opt/intel/composer_xe_2015.1.133/tbb/lib/intel64/gcc4.4/ -L/usr/lib/gcc/x86_64-redhat-linux/4.8.5/../../../ -L/lib64 -L/lib/ -L/usr/lib64 -L/usr/lib -lmpi_usempif08 -lmpi_usempi_ignore_tkr -lmpi_mpifh -lmpi -lifport -lifcore -limf -lsvml -lm -lipgo -lirc -lpthread -lirc_s -ldl
/opt/intel/mkl/lib/intel64/libmkl_blacs_lp64.a(blacs_gridinfo_.o): In function `blacs_gridinfo_':
../../../../scalapack/BLACS/SRC/MPI/blacs_gridinfo_.c:(.text+0x0): multiple definition of `blacs_gridinfo_'
/home/weitong/Download/abinit-8.2.2/tmp_openmpi_intel15-2/fallbacks/exports/lib/libbigdft-1.a(blacs_fake.o):/home/weitong/Download/abinit-8.2.2/tmp_openmpi_intel15-2/fallbacks/sources/bigdft-1.7.1.23/src/modules/blacs_fake.f90:21: first defined here
make[4]: *** [multibinit] Error 1
make[4]: *** Waiting for unfinished jobs....
make[4]: *** [abinit] Error 1
make[4]: Leaving directory `/home/weitong/Download/abinit-8.2.2/tmp_openmpi_intel15-2/src/98_main'
make[3]: *** [all-recursive] Error 1
make[3]: Leaving directory `/home/weitong/Download/abinit-8.2.2/tmp_openmpi_intel15-2/src'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory `/home/weitong/Download/abinit-8.2.2/tmp_openmpi_intel15-2'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/weitong/Download/abinit-8.2.2/tmp_openmpi_intel15-2'
make: *** [multi] Error 2



Here is my configure.ac file:
CC="mpicc"
CXX="mpicxx"
FC="mpif90"
enable_64bit_flags="yes"
prefix="/usr/local/abinit8.2_openmpi_intel"
enable_mpi="yes"
enable_mpi_io="yes"
with_mpi_prefix="/usr/local/openmpi_intel15"
with_trio_flavor="netcdf+etsf_io"
with_fft_flavor="fftw3"
with_fft_incs="-I/usr/local/fftw3_intel15_openmpi/include/"
with_fft_libs="-L/usr/local/fftw3_intel15_openmpi/lib -lfftw3 -lfftw3_mpi -lfftw3_threads"
with_linalg_flavor="mkl"
with_linalg_libs="-L/opt/intel/mkl/lib/intel64 -Wl,--start-group -lmkl_intel_lp64 -lmkl_scalapack_lp64 -lmkl_blacs_lp64 -lmkl_blas95_lp64 -lmkl_lapack95_lp64 -lmkl_sequential -lmkl_core -Wl,--end-group"
with_dft_flavor="atompaw+libxc+wannier90+bigdft"
enable_gw_dpc="yes"
enable_memory_profiling="no"
enable_openmp="no"
enable_maintainer_checks="no"
enable_mpi="yes"
enable_mpi_io="yes"


If I deselec bigdft with
with_dft_flavor="atompaw+libxc+wannier90"

I got everything OK through "make".

Re: ver8.2.2 make error related to BigDFT  [SOLVED]

Posted: Thu Apr 13, 2017 2:22 pm
by weitong
I got the solution.

Remove " -lmkl_scalapack_lp64" from "with_linalg_libs".