Issues compiling yambo 5.0 on HPC

[pyadav]

Dear Andrea,

Thank you for your reply. Yes it seems that it's not able to unzip the tar file as I got the following message

Code: Select all

tar xvfz libxc-2.2.3.tar.gz

gzip: stdin: unexpected end of file
tar: Child returned status 1
tar: Error is not recoverable: exiting now
-------------------------------------
tar -zxvf libxc-2.2.3.tar.gz
gzip: stdin: unexpected end of file
tar: Child returned status 1
tar: Error is not recoverable: exiting now

How can I get rid of it?, Should I replace the tar files with freshly downloaded ones? and where can I get these files?

best,
pushpendra

[andrea.ferretti]

Dear Marzieh,

unless I am mistaken, it looks all good to me.
In terms of possible causes to be further investigated
* the MPI issue does not look promising. Can you check with your sys-admin ?
* otherwise, I would try to either use external hdf5 and recompile internal netcdf and netcdff,
* or simply to compile the whole IO stack (netcdf, netcdff, hdf5) internally (just drop the related lines from configure invocation)

take care
Andrea

[pyadav]

Dear Andrea,

When I'm downloading a fresh libxc-2.2.3.tar.gz file, it successfully unzips but when I put it in the archive folder and trying to build the code it shows the error "libxc build failed". Later on when I'm trying to unzip the same "libxc-2.2.3.tar.gz" in the archive, it shows the error

Code: Select all

gzip: stdin: unexpected end of file
tar: Child returned status 1
tar: Error is not recoverable: exiting now

How can I fix this?

Thanks,
Pushpendra

[Marzieh]

Dear Andrea

I recompiled yambo with these modules an configuration:

Code: Select all

module --force  purge
module load releases/2019b
module load intel/2019b
module load libxc/4.3.4-iccifort-2019.5.281
module load HDF5/1.10.5-iimpi-2019b
module load netCDF/4.7.1-iimpi-2019b
module load netCDF-Fortran/4.5.2-iimpi-2019b

Code: Select all

./configure FC="ifort" PFC="mpiifort" F77="ifort" CC="mpiicc" CXX="mpiicpc" \
--enable-hdf5-par-io \
--enable-mpi \
--enable-open-mp \
--enable-dp \
--with-blas-libs="-lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread -lm" \
--with-lapack-libs="-lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread -lm" \
--with-fft-includedir="-I${MKLROOT}/include/fftw" \
--with-fft-libs="-lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread -lm" \
--with-blacs-libs="-lmkl_scalapack_lp64 -lmkl_blacs_intelmpi_lp64 -lmkl_core -liomp5 -lpthread -lm" \
--with-scalapack-libs="-lmkl_scalapack_lp64 -lmkl_blacs_intelmpi_lp64 -lmkl_core -liomp5 -lpthread -lm" \
--with-netcdf-libs="-L$EBROOTNETCDF/lib64/" \
--with-netcdf-path="$EBROOTNETCDF/" \
--with-netcdf-libdir="$EBROOTNETCDF/lib64/" \
--with-netcdf-includedir="$EBROOTNETCDF/include/" \
--with-netcdff-libs="-L$EBROOTNETCDFMINFORTRAN/lib/" \
--with-netcdff-path="$EBROOTNETCDFMINFORTRAN/" \
--with-netcdff-libdir="$EBROOTNETCDFMINFORTRAN/lib/" \
--with-netcdff-includedir="$EBROOTNETCDFMINFORTRAN/include/" \
--with-hdf5-libs="-L$EBROOTHDF5/lib/" \
--with-hdf5-path="$EBROOTHDF5/" \
--with-hdf5-libdir="$EBROOTHDF5/lib/" \
--with-hdf5-includedir="$EBROOTHDF5/include/"

The output of the configuaration was as:

Code: Select all

# [VER] 5.0.0 r.19466
#
# - GENERAL CONFIGURATIONS -
#
# [SYS] linux@x86_64
# [SRC] /home/ucl/modl/afekete/src/yambo-5.0.0
# [BRN]
# [CMP] /home/ucl/modl/afekete/src/yambo-5.0.0
# [TGT] /home/ucl/modl/afekete/src/yambo-5.0.0
# [BIN] /home/ucl/modl/afekete/src/yambo-5.0.0/bin
# [LIB] /home/ucl/modl/afekete/src/yambo-5.0.0/lib/external
#
# [EDITOR] vim
# [ MAKE ] make
#
# [X] Double precision
# [X] Keep object files
# [X] Run-Time timing profile
# [-] Run-Time memory profile
#
# - SCRIPTS -
#
# [-] YDB: Yambo DataBase
# [-] YAMBOpy: Yambo Python scripts
#
# - PARALLEL SUPPORT -
#
# [X] MPI
# [X] OpenMP
#
# - LIBRARIES [E=external library; I?=internal library (c=to be compiled / f=found already compiled); X=system default; -=not used;] -
#
# > I/O: (NETCDF with large files support)
#
# [ - ] FUTILE  :
#
# [ - ] YAML    :
#
# [ If] IOTK    : /home/ucl/modl/afekete/src/yambo-5.0.0/lib/external/intel/mpiifort/lib/libiotk.a (QE hdf5-support)
#                 -I/home/ucl/modl/afekete/src/yambo-5.0.0/lib/external/intel/mpiifort/include/
# [ - ] ETSF_IO :
#
# [ E ] NETCDF  : -L/opt/sw/arch/easybuild/2019b/software/netCDF-Fortran/4.5.2-iimpi-2019b/lib/ -lnetcdff -L/opt/sw/arch/easybuild/2019b/software/netCDF/4.7.1-iimpi-2019b/lib64/ -lnetcdf
#                 -I/opt/sw/arch/easybuild/2019b/software/netCDF-Fortran/4.5.2-iimpi-2019b/include/ -I/opt/sw/arch/easybuild/2019b/software/netCDF/4.7.1-iimpi-2019b/include/
# [ E ] HDF5    : -L/opt/sw/arch/easybuild/2019b/software/HDF5/1.10.5-iimpi-2019b/lib/ -lhdf5hl_fortran -lhdf5_fortran -lhdf5_hl -lhdf5 -lz -lm -ldl -lcurl
#                 -I/opt/sw/arch/easybuild/2019b/software/HDF5/1.10.5-iimpi-2019b/include/
#
# > MATH: (FFTW MKL)
#
# [ E ] FFT       : -lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread -lm
#
# [ E ] BLAS      : -lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread -lm
# [ E ] LAPACK    : -lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread -lm
# [ E ] SCALAPACK : -lmkl_scalapack_lp64 -lmkl_blacs_intelmpi_lp64 -lmkl_core -liomp5 -lpthread -lm
# [ E ] BLACS     : -lmkl_scalapack_lp64 -lmkl_blacs_intelmpi_lp64 -lmkl_core -liomp5 -lpthread -lm
# [ - ] PETSC     :
#
# [ - ] SLEPC     :
#
#
# > OTHERs
#
# [ If] LibXC     : /home/ucl/modl/afekete/src/yambo-5.0.0/lib/external/intel/mpiifort/lib/libxcf90.a /home/ucl/modl/afekete/src/yambo-5.0.0/lib/external/intel/mpiifort/lib/libxc.a
#                   -I/home/ucl/modl/afekete/src/yambo-5.0.0/lib/external/intel/mpiifort/include
# [ E ] MPI       : -L/opt/sw/arch/easybuild/2019b/software/impi/2018.5.288-iccifort-2019.5.281/intel64/lib -Xlinker --enable-new-dtags -Xlinker -rpath -Xlinker /opt/sw/arch/easybuild/2019b/software/impi/2018.5.288-iccifort-2019.5.281/intel64/lib/release_mt -Xlinker -rpath -Xlinker /opt/sw/arch/easybuild/2019b/software/impi/2018.5.288-iccifort-2019.5.281/intel64/lib -Xlinker -rpath -Xlinker /opt/intel/mpi-rt/2017.0.0/intel64/lib/release_mt -Xlinker -rpath -Xlinker /opt/intel/mpi-rt/2017.0.0/intel64/lib -lmpifort -lmpi -lmpigi -ldl -lrt -lpthread
#                   -I/opt/sw/arch/easybuild/2019b/software/impi/2018.5.288-iccifort-2019.5.281/intel64/include
# [ Ic] Ydriver   : 1.0.0
#
# - COMPILERS -
#
# FC kind = intel ifort version 19.0.5.281
# MPI kind= Intel(R) MPI Library 2018 Update 5 for Linux* OS
#
# [ CPP ] mpiicc -E -ansi -D_HDF5_LIB -D_HDF5_IO -D_PAR_IO -D_MPI -D_FFTW -D_FFTW_OMP  -D_SCALAPACK  -D_DOUBLE -D_OPENMP -D_TIMING     -D_P2Y_QEXSD_HDF5
# [ FPP ] fpp -free -P -D_HDF5_LIB -D_HDF5_IO -D_PAR_IO -D_MPI -D_FFTW -D_FFTW_OMP  -D_SCALAPACK  -D_DOUBLE -D_OPENMP -D_TIMING
# [ CC  ] mpiicc -O2 -D_C_US -D_FORTRAN_US
# [ FC  ] mpiifort -assume bscc -O3 -g -ip    -qopenmp
# [ FCUF] -assume bscc -O0 -g
# [ F77 ] mpiifort -assume bscc -O3 -g -ip
# [ F77U] -assume bscc -O0 -g
# [Cmain] -nofor_main

The previous error was solved but I do not have any values for E-Eo [eV] and Sc|Eo [eV] in o.qp

Code: Select all

#    K-point            Band               Eo [eV]            E-Eo [eV]          Sc|Eo [eV]
#
        1                  1                 -39.73587                   NaN                NaN
        1                  2                 -19.57856                   NaN                NaN
        1                  3                 -1.099598                   NaN                NaN
        1                   4                -0.251880                   NaN                NaN
        1                   5                -0.205098                   NaN                NaN
....

Could you please tell me where I made mistake?

Best,
Marzieh

[Daniele Varsano]

Dear Marzieh,
can you check in the report file if you have NaN also in the HF part of the calculation, or in the Sc only?
Otherwise, If you attach your report file we can have a look.

Best,
Daniele

[Marzieh]

Dear Daniele,
I have attached the generated files.

Best,
Marzieh

[Daniele Varsano]

Dear Marzieh,

the problem is in the correlation part of the self energy. Not easy to spot, but note the parallelism is incomplete.

Can you remove the ./SAVE/ndb.QP and ./SAVE/ndb.pp* files,
and run again the code after setting:

SE_CPU= "1 1 2" # [PARALLEL] CPUs for each role
SE_ROLEs= "q qp b" # [PARALLEL] CPUs roles (q,qp,b)

X_and_IO_CPU= "1 1 1 2 1" # [PARALLEL] CPUs for each role
X_and_IO_ROLEs= "q g k c v" # [PARALLEL] CPUs roles (q,g,k,c,v)

Note that this is just a test, in your input file you have "NGsBlkXp= 1 RL" , and this make your screening unreliable.

Best,
Daniele

[Marzieh]

Dear Daniele,
I did what you suggested me. But I still have that problem.
Best,
Marzieh

[Marzieh]

Dear Daniele,

Since I did not see these parts during run by yambo 5.0.0 (in both parallel and without parallel):

Code: Select all

[WF-HF/Rho] Performing Wave-Functions I/O from ./SAVE
[WF-GW] Performing Wave-Functions I/O from ./SAVE

It seems yambo 5.0.0 can not read files from ./SAVE.
( As I mentioned before I use abinit (of course a2y) to generate files.

So I had to compile yambo 4.5.3 on the cluster. It works.
I would like to ask you if there are main differences between yambo-5.0.0 and yambo4.5.03 especially in nonlinear optics? because I am going to calculate nlo.

Best,
Marzieh

[claudio]

Dear Marzieh

there are not particular differences in the non-linear optics part from the 4.5.3 and the 5.0, as far as I remember,
you can continue with the 4.5.3.
In the while we are trying to solve all compilation problems of the 5.0

best
Claudio

NB: as you see the databases of the 4.5.3 are not compatible with the 5.0