mpi.h error in netcdf
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
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
-
- Posts: 10
- Joined: Sat May 03, 2014 3:14 pm
mpi.h error in netcdf
Hi,
after installation Nedcdf-4.3.2 in ubuntu 12.04 64 bit with ifort composer 2013, im blocked in the case of ../include/netcdf_par.h:13:17: fatal error: mpi.h: No such file or directory
the file of mpi.h is in directory /usr/include/mpi,
regards
after installation Nedcdf-4.3.2 in ubuntu 12.04 64 bit with ifort composer 2013, im blocked in the case of ../include/netcdf_par.h:13:17: fatal error: mpi.h: No such file or directory
the file of mpi.h is in directory /usr/include/mpi,
regards
Pr.Chemam Faical
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
- Daniele Varsano
- Posts: 4198
- Joined: Tue Mar 17, 2009 2:23 pm
- Contact:
Re: mpi.h error in netcdf
Dear Chemam,
it looks mpi library are not correctly seen:
Just to understand, the problem is found in the netcdf or in yambo compilation?
in order to have a first look can you post your config.log file?
Best,
Daniele
it looks mpi library are not correctly seen:
Just to understand, the problem is found in the netcdf or in yambo compilation?
in order to have a first look can you post your config.log file?
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/
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/
-
- Posts: 10
- Joined: Sat May 03, 2014 3:14 pm
Re: mpi.h error in netcdf
Hi,
the problem is in netcdf compilation, after compilation with only ''make'', or ''make check install''
the config.log is in attached.
regards
the problem is in netcdf compilation, after compilation with only ''make'', or ''make check install''
the config.log is in attached.
regards
You do not have the required permissions to view the files attached to this post.
Pr.Chemam Faical
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
- Daniele Varsano
- Posts: 4198
- Joined: Tue Mar 17, 2009 2:23 pm
- Contact:
Re: mpi.h error in netcdf
Dear Chemam,
this is not a yambo issue, but netcdf problem, and you should eventually ask to netcdf people.
Anyway: please note that
1. Parallel compilation is not needed for the netcdf
2. the fortran compiler has not been recognized, in general you can try to compile by using ./configure FC=gfortran
3. Moreover it looks me that 4.3.2 is the c library, next you have to install a fortran library (after compiling the c libraries):
https://www.unidata.ucar.edu/downloads/ ... /index.jsp
4. If you have problem you can try to installa an older release of netcdf (netcdf 3.6.3 is working for yambo) where c and fortran source were not separated.
Best,
Daniele
this is not a yambo issue, but netcdf problem, and you should eventually ask to netcdf people.
Anyway: please note that
1. Parallel compilation is not needed for the netcdf
2. the fortran compiler has not been recognized, in general you can try to compile by using ./configure FC=gfortran
3. Moreover it looks me that 4.3.2 is the c library, next you have to install a fortran library (after compiling the c libraries):
https://www.unidata.ucar.edu/downloads/ ... /index.jsp
4. If you have problem you can try to installa an older release of netcdf (netcdf 3.6.3 is working for yambo) where c and fortran source were not separated.
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/
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/
-
- Posts: 10
- Joined: Sat May 03, 2014 3:14 pm
Re: mpi.h error in netcdf
Hi, again,
thank you very much for u help,
ive used the netcdf3.6.3 its work, make , make check and make install no error
now the problem is the yambo linking with netcdf, isn't appear.
like this
#
# [VER] 3.4.1 r.3187
#
# [SYS] linux@x86_64
# [SRC] /home/chemam/yambo-3.4.1
# [BIN] /home/chemam/yambo-3.4.1/bin
# [FFT] Goedecker Fast Fourier transform with 0 cache
#
# [ ] Double precision
# [X] Redundant compilation
# [X] MPI
# [ ] OpenMP
# [X] PW (4.0) support
# [ ] ETSF I/O support
# [ ] SCALAPACK
# [ ] NETCDF/HDF5/Large Files
# [XX ] Built-in BLAS/LAPACK/LOCAL
#
# [ CPP ] gcc -E -P
# [ C ] gcc -g -O2 -D_C_US -D_FORTRAN_US
# [MPICC] mpicc -g -O2 -D_C_US -D_FORTRAN_US
# [ F90 ] gfortran -O3 -mtune=native
# [MPIF ] mpif90 -O3 -mtune=native
# [ F77 ] gfortran -O3 -mtune=native
# [Cmain]
# [NoOpt] -O0 -mtune=native
#
# [ MAKE ] make
# [EDITOR] vi
ive used this instruction ./configure --with-netcdf-include=/home/chemam/netcdf-3.6.3/include/ --with-netcdf-lib=/home/chemam/netcdf-3.6.3/lib/ --with-iotk=/home/chemam/espresso-5.0.2/iotk/ --with-p2y=4.0
best regards
thank you very much for u help,
ive used the netcdf3.6.3 its work, make , make check and make install no error
now the problem is the yambo linking with netcdf, isn't appear.
like this
#
# [VER] 3.4.1 r.3187
#
# [SYS] linux@x86_64
# [SRC] /home/chemam/yambo-3.4.1
# [BIN] /home/chemam/yambo-3.4.1/bin
# [FFT] Goedecker Fast Fourier transform with 0 cache
#
# [ ] Double precision
# [X] Redundant compilation
# [X] MPI
# [ ] OpenMP
# [X] PW (4.0) support
# [ ] ETSF I/O support
# [ ] SCALAPACK
# [ ] NETCDF/HDF5/Large Files
# [XX ] Built-in BLAS/LAPACK/LOCAL
#
# [ CPP ] gcc -E -P
# [ C ] gcc -g -O2 -D_C_US -D_FORTRAN_US
# [MPICC] mpicc -g -O2 -D_C_US -D_FORTRAN_US
# [ F90 ] gfortran -O3 -mtune=native
# [MPIF ] mpif90 -O3 -mtune=native
# [ F77 ] gfortran -O3 -mtune=native
# [Cmain]
# [NoOpt] -O0 -mtune=native
#
# [ MAKE ] make
# [EDITOR] vi
ive used this instruction ./configure --with-netcdf-include=/home/chemam/netcdf-3.6.3/include/ --with-netcdf-lib=/home/chemam/netcdf-3.6.3/lib/ --with-iotk=/home/chemam/espresso-5.0.2/iotk/ --with-p2y=4.0
best regards
Pr.Chemam Faical
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
- Daniele Varsano
- Posts: 4198
- Joined: Tue Mar 17, 2009 2:23 pm
- Contact:
Re: mpi.h error in netcdf
Dear Chemam,
Please verify that netcdf and yambo are compiled with the same compiler, next please post the config.log file and we will try to spot the problem.
Regards,
Daniele
Please verify that netcdf and yambo are compiled with the same compiler, next please post the config.log file and we will try to spot the problem.
Regards,
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/
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/
-
- Posts: 10
- Joined: Sat May 03, 2014 3:14 pm
Re: mpi.h error in netcdf
Hi,
the config.log is in attached
regards
the config.log is in attached
regards
You do not have the required permissions to view the files attached to this post.
Pr.Chemam Faical
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
- Daniele Varsano
- Posts: 4198
- Joined: Tue Mar 17, 2009 2:23 pm
- Contact:
Re: mpi.h error in netcdf
Dear Chemam,
I can't see anything strange in your configure, the netcdf are ignored so it is not easy to spot the problem.
As told before be sure you are using the same compiler for compiling netcdf and configuring yambo. In order to look exactly to which compiler your mpif90 is linked you can type mpif90 --version
You can also try to add in your configure --with-netcdf-link="-lnetcdff -lnetcdf" , but I do not think it will solve the problem.
You can also add --enable-msgs-comps, in order to add more verbosity in the config.log, this could help.
Finally, please note that netcdf library are not mandatory, you can run the code without using that, the advantage of netcdf is that you can move databases in machines with different architectures, and you need them to run the tutorials. The tutorials can be also run without netcdf but you have to run the ground state calculations first, and use the interface.
Best,
Daniele
I can't see anything strange in your configure, the netcdf are ignored so it is not easy to spot the problem.
As told before be sure you are using the same compiler for compiling netcdf and configuring yambo. In order to look exactly to which compiler your mpif90 is linked you can type mpif90 --version
You can also try to add in your configure --with-netcdf-link="-lnetcdff -lnetcdf" , but I do not think it will solve the problem.
You can also add --enable-msgs-comps, in order to add more verbosity in the config.log, this could help.
Finally, please note that netcdf library are not mandatory, you can run the code without using that, the advantage of netcdf is that you can move databases in machines with different architectures, and you need them to run the tutorials. The tutorials can be also run without netcdf but you have to run the ground state calculations first, and use the interface.
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/
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/
-
- Posts: 10
- Joined: Sat May 03, 2014 3:14 pm
Re: mpi.h error in netcdf
Hi,
if ive understand that after runnig yambo in terminal
u have this instruction, yambo: cannot access CORE database (SAVE/*db1 and/or SAVE/*wf), but u need to install netcdf to escape
this instruction.
me now im in this case, i cant undesrtand how to procede after.
i see the tutorial but no thing in my mind,
Best regards
if ive understand that after runnig yambo in terminal
u have this instruction, yambo: cannot access CORE database (SAVE/*db1 and/or SAVE/*wf), but u need to install netcdf to escape
this instruction.
me now im in this case, i cant undesrtand how to procede after.
i see the tutorial but no thing in my mind,
Best regards
Pr.Chemam Faical
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
Laboratoire de la Physique Appliquée et Théorique,
Université de Tebessa, Algérie
- Daniele Varsano
- Posts: 4198
- Joined: Tue Mar 17, 2009 2:23 pm
- Contact:
Re: mpi.h error in netcdf
Dear Chemam,
yes, if you are running the tutorial you need the netcdf linked in order to run using the databases stored in the SAVE directory,
Anyway there are also the input for the ground state calculations under the PWSCF directory. If you run the ground state using pwscf (quantum espresso),
and generate your SAVE directory using p2y, you will have in the SAVE directos the s.db and s.wf databases (instead of ns.*) if p2y (and Yambo) are compiled without nectdf support and all the calculations can be run without using the netcdf.
Anyway the error you mention appears when you run yambo without having the SAVE directory in the directory you are running.
Are you running the tutorial or you system of interest?
In the latter case, did you perform the ground state calculation and run the interface (p2y or a2y)? Yambo have to be run in the directory where the SAVE directory is present. The SAVE directory is generated by the p2y, or a2y, interface. If you can't solve it, please report all the step you did before running yambo.
Best,
Daniele
yes, if you are running the tutorial you need the netcdf linked in order to run using the databases stored in the SAVE directory,
Anyway there are also the input for the ground state calculations under the PWSCF directory. If you run the ground state using pwscf (quantum espresso),
and generate your SAVE directory using p2y, you will have in the SAVE directos the s.db and s.wf databases (instead of ns.*) if p2y (and Yambo) are compiled without nectdf support and all the calculations can be run without using the netcdf.
Anyway the error you mention appears when you run yambo without having the SAVE directory in the directory you are running.
Are you running the tutorial or you system of interest?
In the latter case, did you perform the ground state calculation and run the interface (p2y or a2y)? Yambo have to be run in the directory where the SAVE directory is present. The SAVE directory is generated by the p2y, or a2y, interface. If you can't solve it, please report all the step you did before running yambo.
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/
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/