[Octopus-users] Error: Cannot open variables file

Tue Feb 16 16:14:50 WET 2010

Hi All,

I haven't managed to solve either problem with the PGI or GNU compilers.

I have however realised the mistake I had made with the pathscale compiled
executable: I had set the OCTOPUS_SHARE variable to point to a symbolic link
to my work directory, which was in my home directory and not accessible from
the compute nodes.

Thanks for the help,

Conn

On Tue, Feb 16, 2010 at 12:06 AM, David Strubbe <dstrubbe at berkeley.edu>wrote:

> Conn,
>
> PGI 7 or 9 certainly work for this file. I don't know why version 8.0-2 is
> giving these problems.
>
> David
>
>
> On Mon, Feb 15, 2010 at 2:00 AM, matthieu verstraete <
> Matthieu.Verstraete at ulg.ac.be> wrote:
>
>> Need full config.log, and in particular compiler versions. gfortran
>> before 4.2 is quite bad. That routine has been compiled to death with
>> 4.3 4.4 so it should not be an octopus problem...
>>
>> I stay away from pgi: for some reason they always have bugs, it's a
>> real pain. It may be right about the typecasting though. You can also
>> try with gcc instead of pgcc (keeping pgf90)
>>
>> ciao
>>
>> Matthieu
>>
>> On Sun, Feb 14, 2010 at 10:51 PM, Conn O'Rourke
>> <conn.orourke at googlemail.com> wrote:
>> > Hi All,
>> >
>> > I have decided to try and compile with one of the other compilers on the
>> > system to see if that improves things, but I am coming across a couple
>> of
>> > errors with two seperate compilers. Any help solving either would be
>> great!:
>> >
>> > using the gnu compilers i get:
>> >
>> > 3   -I ../../src/basic -I ../../src/math -I ../../src/species -I
>> > ../../src/ions -I ../../src/grid -I ../../src/beak -I ../../src/poisson
>> -I
>> > ../../src/states -I ../../src/xc -I ../../src/system -I
>> > ../../src/hamiltonian -I ../../src/scf -I ../../src/td -I
>> > ../../src/opt_control -I ../../src/sternheimer -I ../../libxc/src -I
>> > ../../external_libs/qshep  -c  -o mesh.o mesh_oct.f90
>> > /opt/cray/xt-asyncpe/3.5/bin/ftn: INFO: linux target is being used
>> > gfortran: Internal error: Segmentation fault (program f951)
>> > Please submit a full bug report.
>> > See <URL:http://gcc.gnu.org/bugs.html> for instructions.
>> > make[3]: *** [mesh.o] Error 1
>> > make[3]: Leaving directory
>> `/work/e89/e89/cor/bin/octopus-3.2.0/src/grid'
>> > make[2]: *** [all-recursive] Error 1
>> > make[2]: Leaving directory `/work/e89/e89/cor/bin/octopus-3.2.0/src'
>> > make[1]: *** [all-recursive] Error 1
>> > make[1]: Leaving directory `/work/e89/e89/cor/bin/octopus-3.2.0'
>> >
>> > Which I guess is a problem with the compiler, so I tried the pgi
>> compilers
>> > and I get:
>> >
>> > cc -DHAVE_CONFIG_H -I. -I../..  -I../../src/include -I../../src/include
>> > -I../../external_libs/spglib/src -I../../liboct_parser
>> > -I../../external_libs/metis-4.0/ -I../../external_libs/zoltan/include
>> -g
>> > -I/home/e89/e89/cor/work/bin/gsl-1.9/include -I/usr/include -c
>> > operate_ri_vec.c
>> > /opt/cray/xt-asyncpe/3.5/bin/cc: INFO: linux target is being used
>> > PGC-W-0095-Type cast required for this conversion (operate_ri_vec.c:
>> 260)
>> > PGC-W-0095-Type cast required for this conversion (operate_ri_vec.c:
>> 283)
>> > PGC-W-0095-Type cast required for this conversion (operate_ri_vec.c:
>> 285)
>> > PGC-W-0277-Cannot inline function _mm_set1_pd - data type mismatch
>> > (operate_ri_vec.c: 69)
>> > PGC-W-0277-Cannot inline function _mm_setzero_pd - data type mismatch
>> > (operate_ri_vec.c: 88)
>> > PGC-W-0277-Cannot inline function _mm_loadu_pd - data type mismatch
>> > (operate_ri_vec.c: 93)
>> > PGC-W-0277-Cannot inline function _mm_mul_pd - data type mismatch
>> > (operate_ri_vec.c: 93)
>> > PGC-W-0277-Cannot inline function _mm_add_pd - data type mismatch
>> > (operate_ri_vec.c: 93)
>> > PGC-W-0277-Cannot inline function _mm_loadu_pd - data type mismatch
>> > (operate_ri_vec.c: 94)
>> > PGC-W-0277-Cannot inline function _mm_mul_pd - data type mismatch
>> > (operate_ri_vec.c: 94)
>> > PGC-W-0277-Cannot inline function _mm_add_pd - data type mismatch
>> > (operate_ri_vec.c: 94)
>> > PGC-W-0277-Cannot inline function _mm_loadu_pd - data type mismatch
>> > (operate_ri_vec.c: 95)
>> > PGC-W-0277-Cannot inline function _mm_mul_pd - data type mismatch
>> > (operate_ri_vec.c: 95)
>> > PGC-W-0277-Cannot inline function _mm_add_pd - data type mismatch
>> > (operate_ri_vec.c: 95)
>> > PGC-W-0277-Cannot inline function _mm_loadu_pd - data type mismatch
>> > (operate_ri_vec.c: 96)
>> > PGC-W-0277-Cannot inline function _mm_mul_pd - data type mismatch
>> > (operate_ri_vec.c: 96)
>> > PGC-W-0277-Cannot inline function _mm_add_pd - data type mismatch
>> > (operate_ri_vec.c: 96)
>> > PGC-S-0094-Illegal type conversion required (operate_ri_vec.c: 99)
>> > PGC/x86-64 Linux 8.0-2: compilation completed with severe errors
>> > make[2]: *** [operate_ri_vec.o] Error 2
>> > make[2]: Leaving directory
>> `/work/e89/e89/cor/bin/octopus-3.2.0/src/beak'
>> > make[1]: *** [install-recursive] Error 1
>> > make[1]: Leaving directory `/work/e89/e89/cor/bin/octopus-3.2.0/src'
>> > make: *** [install-recursive] Error 1
>> >
>> > Cheers!
>> >
>> > Conn
>> >
>> > On Sun, Feb 14, 2010 at 12:18 PM, Conn O'Rourke
>> > <conn.orourke at googlemail.com> wrote:
>> >>
>> >> Hi Alberto,
>> >>
>> >> Thanks for the reply. I remembered to do a make install and it is still
>> >> unable to find it.
>> >>
>> >> All the best,
>> >> Conn
>> >>
>> >> On Sun, Feb 14, 2010 at 11:26 AM, Alberto Castro <acastro at bifi.es>
>> wrote:
>> >>>
>> >>> The most common cause of this problem is forgetting to do "make
>> install".
>> >>>
>> >>> 2010/2/14 Conn O'Rourke <conn.orourke at googlemail.com>:
>> >>> > Dear Octopus Users,
>> >>> >
>> >>> > I have compiled Octopus using the pathscale 2.1.56 compilers on a
>> Cray
>> >>> > XT4 -
>> >>> > AMD Opteron.
>> >>> >
>> >>> > I get an executable but when I try to run it I get the following
>> error:
>> >>> >
>> >>> > *** Fatal Error (description follows)
>> >>> > Cannot open variables file:
>> >>> >
>> >>> > I have performed a make clean before configuration and also tried
>> >>> > setting
>> >>> > OCTOPUS_SHARE to point in the appropriate direction, with no
>> success.
>> >>> >
>> >>> > Any other suggestions/workarounds would be greatly appreciated.
>> >>> >
>> >>> > Thanks in advance,
>> >>> > Conn
>> >>> >
>> >>> > _______________________________________________
>> >>> > Octopus-users mailing list
>> >>> > Octopus-users at tddft.org
>> >>> > http://www.tddft.org/mailman/listinfo/octopus-users
>> >>> >
>> >>> >
>> >>>
>> >>>
>> >>>
>> >>> --
>> >>> ===============================
>> >>> Dr. Alberto Castro
>> >>>
>> >>> Instituto de Biocomputación y Física de Sistemas Complejos (BIFI),
>> >>> Corona de Aragón 42,
>> >>> 50009 Zaragoza, España.
>> >>>
>> >>> Temporarily at:
>> >>>
>> >>> Departamento de Física Teórica, Facultad de Ciencias,
>> >>> Universidad de Zaragoza,
>> >>> Pedro Cerbuna 12,
>> >>> 50009 Zaragoza, España.
>> >>>
>> >>> Tel:   +34 976761000 3332
>> >>> Fax:  +34 976761264
>> >>> skype: alberto_c_barrigon
>> >>> email: acastro at bifi.es
>> >>> ===============================
>> >>
>> >
>> >
>> > _______________________________________________
>> > Octopus-users mailing list
>> > Octopus-users at tddft.org
>> > http://www.tddft.org/mailman/listinfo/octopus-users
>> >
>> >
>>
>>
>>
>> --
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> Prof. Matthieu Verstraete
>>
>> Universite de Liège
>> Institut de Physique, Bat. B5, 3/7
>> Allée du 6 aout, 17
>> B- 4000 Sart Tilman, Liège
>> Belgium
>>
>> Associate member European Theoretical Spectroscopy Facility (ETSF
>> http://www.etsf.eu)
>>
>> Phone : +32 4 366 90 17
>> Fax   : +32 4 366 36 29
>>
>> Mail : matthieu.verstraete at ulg.ac.be
>>          matthieu.jean.verstraete at gmail.com
>>
>> _______________________________________________
>> Octopus-users mailing list
>> Octopus-users at tddft.org
>> http://www.tddft.org/mailman/listinfo/octopus-users
>>
>
>
> _______________________________________________
> Octopus-users mailing list
> Octopus-users at tddft.org
> http://www.tddft.org/mailman/listinfo/octopus-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.tddft.org/pipermail/octopus-users/attachments/20100216/5e876cb3/attachment.html 