libsvml.so error

Questions regarding the compilation of VASP on various platforms: hardware, compilers and libraries, etc.


Moderators: Global Moderator, Moderator

Post Reply
Message
Author
bandy

libsvml.so error

#1 Post by bandy » Sun Aug 06, 2006 8:33 pm

I have compiled VASP in parallel mode.

The last line to create is:
-----------------------------------------------------
/N/soft/linux-rhel3_AS-ia32/lam-gm-7.1.1-intel/bin/mpif77 -I/N/soft/linux-rhel3_AS-ia32/lam-gm-7.1.1-intel/lib -L/usr/pbs/lib -o vasp libsvml.a libsvml.so main.o base.o mpi.o smart_allocate.o xml.o constant.o jacobi.o main_mpi.o scala.o asa.o lattice.o poscar.o ini.o setex.o radial.o pseudo.o mgrid.o mkpoints.o wave.o wave_mpi.o symmetry.o symlib.o lattlib.o random.o nonl.o nonlr.o dfast.o choleski2.o mix.o charge.o xcgrad.o xcspin.o potex1.o potex2.o metagga.o constrmag.o pot.o cl_shift.o force.o dos.o elf.o tet.o hamil.o steep.o chain.o dyna.o relativistic.o LDApU.o sphpro.o paw.o us.o ebs.o wavpre.o wavpre_noio.o broyden.o dynbr.o rmm-diis.o reader.o writer.o tutor.o xml_writer.o brent.o stufak.o fileio.o opergrid.o stepver.o dipol.o xclib.o chgloc.o subrot.o optreal.o davidson.o edtest.o electron.o shm.o pardens.o paircorrection.o optics.o constr_cell_relax.o stm.o finite_diff.o elpol.o setlocalpp.o fftmpi.o fftmpi_map.o fft3dlib.o -L../vasp.4.lib -ldmy ../vasp.4.lib/linpack_double.o -L/N/soft/linux-rhel3_AS-ia32/intel-mkl-7.2.1.003/lib/32 -lmkl_lapack -L/N/soft/linux-rhel3_AS-ia32/intel-mkl-7.2.1.003/lib/32 -lmkl_p4 -lguide -L/usr/lib -lpthread
---------------------------------------------------
But when I am running vasp:

I am getting the error:
vasp: error while loading shared libraries: libsvml.so: cannot open shared object file: No such file or directory

But I have in vasp.4.6 directory.

Any help will be highly welcome.
Last edited by bandy on Sun Aug 06, 2006 8:33 pm, edited 1 time in total.

lahaye
Jr. Member
Jr. Member
Posts: 98
Joined: Fri Apr 14, 2006 5:08 am
Location: Suwon - Korea

libsvml.so error

#2 Post by lahaye » Mon Aug 07, 2006 1:36 am

Bandy,

Which compiler are you using?
I myself use the intel 9.0 for Linux, which has this library in:

Code: Select all

$base = intel's install directory (of your own choice)
$base/cc/9.0/lib/libsvml.so
Thus libsvml.so is not in the vasp directory. To get this work, I have
in the Makefile:

Code: Select all

LINK    = -L$base/intel/fc/9.0/lib -lsvml
Futhermore, you need to either get his directory included in your
library path with ldconfig, or set LD_LIBRARY_PATH to that
directory.

Does that help?

Rob.
Last edited by lahaye on Mon Aug 07, 2006 1:36 am, edited 1 time in total.

admin
Administrator
Administrator
Posts: 2921
Joined: Tue Aug 03, 2004 8:18 am
License Nr.: 458

libsvml.so error

#3 Post by admin » Mon Aug 07, 2006 8:00 am

please use one of the standard makefiles and load either with
-L./svml (libsvml.a)
or
./libsvml.so (libsvml.so)
please check if the path to this library (./) and the file execution permissions really are correct.
Last edited by admin on Mon Aug 07, 2006 8:00 am, edited 1 time in total.

bandy

libsvml.so error

#4 Post by bandy » Mon Aug 07, 2006 8:15 am

Thanks a lot. I am using linux-rhel3_AS-ia32/intel-ifort8-8.1-028.
with LD_LIBRARY_PATH set in .cshrc and LINK = -L/linux-rhel3_AS-ia32/intel-icc8-8.1-032/lib -lsvml in MAKEFILE, it created working executable.

<span class='smallblacktext'>[ Edited Mon Aug 07 2006, 10:21AM ]</span>
Last edited by bandy on Mon Aug 07, 2006 8:15 am, edited 1 time in total.

bandy

libsvml.so error

#5 Post by bandy » Wed Aug 16, 2006 11:59 pm

I am facing a strange problem regarding execution of parallel vasp. I ran interactively successfully 1-2 times, often I failed to do that. Because of improper setting of environment variables, can anybody give me pbs script to run vasp parallelized with lam-mpi.

Thanks
Last edited by bandy on Wed Aug 16, 2006 11:59 pm, edited 1 time in total.

DavidCGreen

libsvml.so error

#6 Post by DavidCGreen » Thu Aug 17, 2006 1:36 pm

Your PBS script will need to include the relevant setenv lines you have in your .cshrc

Running lam-mpi jobs is relatively easy under PBS (we have PBSPro 7.1)
The key lines you need are:

#PBS -l select=16:ncpus=1
##(16nodes x 1cpu each)

#PBS -l place=scatter
## (spread 'em across cluster)

/usr/pbs/bin/pbs_lamboot
##Gotta start the LAM universe first ... PBS will decide which nodes
##in fact it creates the lamnodes file itself

/usr/pbs/bin/pbs_mpilam -np 16 /path/to/your/code/executable

You might need to tweak these a little bit to suit.
Don't forget the other #PBS directives for retaining output, mailing your, etc etc.

<span class='smallblacktext'>[ Edited Thu Aug 17 2006, 03:42PM ]</span>
Last edited by DavidCGreen on Thu Aug 17, 2006 1:36 pm, edited 1 time in total.

Post Reply