compilation of yambo-3-4-1-rev61

Having trouble compiling the Yambo source? Using an unusual architecture? Problems with the "configure" script? Problems in GPU architectures? This is the place to look.

Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, andrea marini, Daniele Varsano, Conor Hogan, Nicola Spallanzani

Forum rules
If you have trouble compiling Yambo, please make sure to list:
(1) the compiler (vendor and release: e.g. intel 10.1)
(2) the architecture (e.g. 64-bit IBM SP5)
(3) if the problems occur compiling in serial/in parallel
(4) the version of Yambo (revision number/major release version)
(5) the relevant compiler error message
Post Reply
senersen
Posts: 13
Joined: Mon Aug 01, 2016 2:12 pm
Location: Spain

compilation of yambo-3-4-1-rev61

Post by senersen » Thu Aug 11, 2016 2:42 pm

I am trying to compile an older version of yambo, namely yambo-3.4.1-rev61 in parallel, lthough i have a newer version already but does not work with abinit somehow.

I want it to work with abinit-7.10.4. I made a script that got me until some point but i am stuck at >>>[Linking yambo]<<< part. I do not know what to do next. Someone suggested me to use -static-intel but i even do not know how to do that i am that much ignorent of these things.

Here is the script

#!/bin/bash
make clean
module purge
module load Yambo/4.0.2-rev.90-intel-2016a-iimpi
./configure CC="mpiicc" MPICC="mpiicc" FC="mpif90" CPP="cpp" F77="mpiifort" CPPFLAGS="" FCFLAGS="$FCFLAGS" CFLAGS="$CFLAGS" FFLAGS="$FCFLAGS" LDFLAGS="$LDFLAGS -Wl,-rpath=$MKLROOT/lib/intel64" --enable-netcdf-LFS --with-blas="-L/$MKLROOT/lib/intel64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core" --with-lapack="-L/$MKLROOT/lib/intel64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core" --with-blacs="-L/$MKLROOT/lib/intel64 -lmkl_blacs_intelmpi_lp64" --with-scalapack="-L/$MKLROOT/lib/intel64 -lmkl_scalapack_lp64" --with-fftw="$EBROOTFFTW" --with-netcdf-lib="-lnetcdff -lnetcdf"
LANG=C
LC_ALL=C
make yambo interfaces


This is the list of modules

1) GCCcore/4.9.3 11) zlib/1.2.8-intel-2016a
2) binutils/2.25-GCCcore-4.9.3 12) Szip/2.1-intel-2016a
3) icc/2016.1.150-GCC-4.9.3-2.25 13) HDF5/1.8.16-intel-2016a
4) ifort/2016.1.150-GCC-4.9.3-2.25 14) cURL/7.47.0-intel-2016a
5) iccifort/2016.1.150-GCC-4.9.3-2.25 15) netCDF/4.4.0-intel-2016a
6) impi/5.1.2.150-iccifort-2016.1.150-GCC-4.9.3-2.25 16) netCDF-Fortran/4.4.2-intel-2016a
7) iimpi/8.1.5-GCC-4.9.3-2.25 17) QuantumESPRESSO/5.4.0-intel-2016a
8) imkl/11.3.1.150-iimpi-8.1.5-GCC-4.9.3-2.25 18) FFTW/3.3.4-intel-2016a-PFFT-20150905
9) intel/2016a 19) ETSF_IO/1.0.4-intel-2016a
10) libxc/3.0.0-intel-2016a 20) Yambo/4.0.2-rev.90-intel-2016a-iimpi

I also attach the log file.

Can you help me compiling this version in parallel?

Thanks in advance
You do not have the required permissions to view the files attached to this post.
Nano-bio Spectroscopy Group European Theoretical Spectroscopy Facility (ETSF) Dpto. Física de Materiales, UPV/EHU
Centro Joxe Mari Korta Avda. de Tolosa, 72 2º piso 20018 Donostia - San Sebastián, Spain
Tel: +34-943018840 Fax: +34-943018390

User avatar
Daniele Varsano
Posts: 4198
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: compilation of yambo-3-4-1-rev61

Post by Daniele Varsano » Thu Aug 11, 2016 5:21 pm

Dear Senersen,

3.4.1 is not supported anymore. It is impossible to us to maintain older versions, moreover in the newer versions there are many bug fixes.

3.4.2 and 4.0 should work with abinit-7.x, so please report what kind of problem you encountered and we will try to help you.

Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

senersen
Posts: 13
Joined: Mon Aug 01, 2016 2:12 pm
Location: Spain

Re: compilation of yambo-3-4-1-rev61

Post by senersen » Fri Aug 12, 2016 11:25 am

Dear Daniele

All right i gave up on yambo 3.4.1. Lets talk about yambo 4.0.4.

I use abinit 7.10.4, i patched it for yambo. i applied a2y command of yambo 4.0.4 which gave no error and created the SAVE folder and filled it with the necessary files. Now i try yambo command and it says
[sener@ganbo60 yambopatchli]$ yambo

[ERROR][NetCDF] NetCDF: Numeric conversion not representable
[sener@ganbo60 yambopatchli]$

By the way yambo is configured with this script;

#!/bin/bash
module load intel/2016a
module load libxc
module load HDF5
module load netCDF
module load netCDF-Fortran
module load QuantumESPRESSO
module load FFTW
module load ETSF_IO
./configure \
CC="mpiicc" MPICC="mpiicc" FC="mpiifort" FCCPP="fpp" F77="mpiifort" \
CPPFLAGS="" FCFLAGS="$FCFLAGS" CFLAGS="$CFLAGS" FFLAGS="$FCFLAGS" \
LDFLAGS="$LDFLAGS -Wl,-rpath=$MKLROOT/lib/intel64" \
--with-blas-libs="-L/$MKLROOT/lib/intel64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core" --with-lapack-libs="-L/$MKLROOT/lib/intel64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core" \
--with-blacs-libs="-L/$MKLROOT/lib/intel64 -lmkl_blacs_intelmpi_lp64" --with-scalapack-libs="-L/$MKLROOT/lib/intel64 -lmkl_scalapack_lp64" \
--with-iotk-path="$EBROOTQUANTUMESPRESSO/espresso-$EBVERSIONQUANTUMESPRESSO/S3DE/iotk/" \
--with-libxc-path="$EBROOTLIBXC" --with-fft-path="$EBROOTFFTW" \
--with-etsf-io-path="$EBROOTETSF_IO" \
--enable-netcdf-LFS --enable-open-mp --enable-time-profile \
--with-netcdf-libs="-lnetcdff -lnetcdf" --with-hdf5-path="$EBROOTHDF5"
LANG=C
LC_ALL=C
make all
Nano-bio Spectroscopy Group European Theoretical Spectroscopy Facility (ETSF) Dpto. Física de Materiales, UPV/EHU
Centro Joxe Mari Korta Avda. de Tolosa, 72 2º piso 20018 Donostia - San Sebastián, Spain
Tel: +34-943018840 Fax: +34-943018390

andrea.ferretti
Posts: 214
Joined: Fri Jan 31, 2014 11:13 am

Re: compilation of yambo-3-4-1-rev61

Post by andrea.ferretti » Fri Aug 12, 2016 11:52 am

Dear Senersen,

can you post the config.log file also for yambo 4.0.4 ?
thanks
Andrea
Andrea Ferretti, PhD
CNR-NANO-S3 and MaX Centre
via Campi 213/A, 41125, Modena, Italy
Tel: +39 059 2055322; Skype: andrea_ferretti
URL: http://www.nano.cnr.it

senersen
Posts: 13
Joined: Mon Aug 01, 2016 2:12 pm
Location: Spain

Re: compilation of yambo-3-4-1-rev61

Post by senersen » Fri Aug 12, 2016 2:52 pm

Dear Andrea

You can find the log file in the attachment.

Thanks in advance,

sener
You do not have the required permissions to view the files attached to this post.
Nano-bio Spectroscopy Group European Theoretical Spectroscopy Facility (ETSF) Dpto. Física de Materiales, UPV/EHU
Centro Joxe Mari Korta Avda. de Tolosa, 72 2º piso 20018 Donostia - San Sebastián, Spain
Tel: +34-943018840 Fax: +34-943018390

senersen
Posts: 13
Joined: Mon Aug 01, 2016 2:12 pm
Location: Spain

Re: compilation of yambo-3-4-1-rev61

Post by senersen » Tue Aug 16, 2016 11:05 am

Do you need /config/report file as well?

sener
Nano-bio Spectroscopy Group European Theoretical Spectroscopy Facility (ETSF) Dpto. Física de Materiales, UPV/EHU
Centro Joxe Mari Korta Avda. de Tolosa, 72 2º piso 20018 Donostia - San Sebastián, Spain
Tel: +34-943018840 Fax: +34-943018390

User avatar
Daniele Varsano
Posts: 4198
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: compilation of yambo-3-4-1-rev61

Post by Daniele Varsano » Tue Aug 16, 2016 11:12 am

Dear Sener,
unfortunately in these days I have very limited connection so I cannot have a look to the files you posted in details. Hopefully someone will help you soon. In the meanwhile you can try to recompile the code without linking your compiled netcdf libraries ?(just skip all the netcdf and hdf option). In this way yambo will download automatically and compile the needed netcdf libs, just in case the problem arise for some incompatibility between yambo and the netcdf previously compiled.

Best

Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

andrea.ferretti
Posts: 214
Joined: Fri Jan 31, 2014 11:13 am

Re: compilation of yambo-3-4-1-rev61

Post by andrea.ferretti » Tue Aug 16, 2016 12:08 pm

Dear Sener,

the error message comes from the netcdf library itself, and seems to point to some internal conversions from undefined numbers
(p/m inf for floats, integers that are too large for 32 bits, etc)
It seems the problem is quite known in the web for other cases...

Concerning yambo, it would be useful to have a small test case reproducing the problem...
* is your system very large ? (perhaps the problem is related to the system size)
* have you tried compiling yambo without netcdf LFS ?
* in order to better understand where the problem is, could you try to issue
yambo -D (it should stop when trying to read the DB that causes the problem)
and report the output ?
* Last but not least, did you try the e2y interface ? (you should have it compiled)

take care
Andrea
Andrea Ferretti, PhD
CNR-NANO-S3 and MaX Centre
via Campi 213/A, 41125, Modena, Italy
Tel: +39 059 2055322; Skype: andrea_ferretti
URL: http://www.nano.cnr.it

senersen
Posts: 13
Joined: Mon Aug 01, 2016 2:12 pm
Location: Spain

Re: compilation of yambo-3-4-1-rev61

Post by senersen » Wed Aug 17, 2016 3:32 pm

Dear Andrea,

The system i am trying is the one that i used to work on with an old version like 3.3.0. I had no problem with that version like these. Therefore, i assume it should not be too large for yambo to handle it.

I did not try to compile yambo without netcdf LFS, i am going to do that now and let you know what the result is.

I issued yambo -D and got the following output;
[RD./SAVE//ns.db1]------------------------------------------
Bands : 110
K-points : 102
G-vectors [RL space]: 89233
Components [wavefunctions]: 79586
Symmetries [spatial+T-rev]: 12
Spinor components : 1
Spin polarizations : 1
Temperature [ev]: 0.000000
Electrons : 43.00000
WF G-vectors : 89144
Max atoms/species : 4
No. of atom species : 3
Magnetic symmetries : no
- S/N 003197 --------------------------- v.04.00.04 r.0000 -
[RD./SAVE//ns.wf]-------------------------------------------
Fragmentation :yes
Bands in each block : 110
Blocks : 1
- S/N 003197 --------------------------- v.04.00.04 r.0000 -
[RD./SAVE//ns.kb_pp]----------------------------------------
Fragmentation :yes
- S/N 003197 --------------------------- v.04.00.04 r.0000 -


I did not try e2y interface as i believe it works only with the files with an extension .nc which my abinit does not create so missing etsf configuration.
Nano-bio Spectroscopy Group European Theoretical Spectroscopy Facility (ETSF) Dpto. Física de Materiales, UPV/EHU
Centro Joxe Mari Korta Avda. de Tolosa, 72 2º piso 20018 Donostia - San Sebastián, Spain
Tel: +34-943018840 Fax: +34-943018390

Post Reply