[Octopus-users] make of octopus-3.2.0pre (3): one success and one failure

[matthieu verstraete]

It could be that you input file contains non ascii characters (are you
using a japanese keyboard?) which are accepted by one compiler but
crash the other... They might even be invisible if you are using a
remote vi or something. Do the tests run?

Matthieu

On Tue, Oct 20, 2009 at 2:48 AM, Motohiko Tanaka <mtanaka at nifs.ac.jp> wrote:
> Dear Matthieu and Octopus develpers,
>
>   Thank you very much for your comments. I have created a program
> containing:
>      subroutine dummy1 ()
>      end subroutine dummy1
> and compiled it by two compilers:
>
> (1) PGI:
> % pgf95 -c test.f90
>  which makes no message, and its object file is:
>
> %nm test.o
> 0000000000000006 t __dummy1_END
> 0000000000000000 T dummy1_
>                 U pgf90_compiled
>
> (2) AIX 5/Hitachi:
> % f90 -c test.f90
> f90: compile start : test.f90
>
> *OFORT90 V02-00-/A entered.
> *program name = DUMMY1
> *end of compilation : DUMMY1
> *program units = 0001, no diagnostics generated.
>
> This time an object file is created in contrast to the case
> of src/xc/functional_list.F90 (program units= 1, not 0), and
> it looks like
> % nm -X64 test.o
> .DUMMY1              T           0
> DUMMY1               D         152          24
> DUMMY1               d         136           8
> TOC                  d         128
> _Sec_Int_Literal     d         128           8
> _Sec_Int_Literal     d         144           4
> test.f90             f           -
>
> ************
> One success:
> ************
>    Following the above, I have treated functional_list.F90
> in the same way as you showed me during the "make" step of
> octopus-3.2.0pre, which creates an object file successfully:
> % nm -X64 functionals_list.o
> .functional_list     T           0
> TOC                  d          64
> _Sec_Int_Literal     d          64           8
> _Sec_Int_Literal     d          88           4
> _Sec_String_Literal  d          72           8
> _Sec_String_Literal  d          96          15
> functional_list      D         112          24
> functional_list      d          80           8
> functionals_list_oct.f90 f           -
>
> ************
> One failure:
> ************
>     Now the run by octopus-3.2.0pre really starts - it is great.
> But, it stops in a short while whose messages are:
>
>                            Running octopus
>
> Version                : 3.2.0pre1
> Revision               : 5907 [par_vec.F90 2009-09-24]
> Build time             : Sat Oct 17 09:37:40 JST 2009
> Configuration options  : max-dim=3 mpi
> Optional libraries     : mpi2 metis libnbc
> Architecture           : powerpc
> C compiler             : f90 -noparallel
> C compiler flags       : No flags information available.
> Fortran compiler       : f90 -i,AS -nohugeary -nosave -noparallel -msg=w
> Fortran compiler flags : No flags information available.
>
>                 The octopus is swimming in sri (AIX)
>
>
>            Calculation started on 2009/10/20 at 09:28:47
>
>
> ************************** Calculation Mode **************************
> Input: [CalculationMode = gs]
> **********************************************************************
>
>  KCHF361R invalid character : is found in input data.
>  0x00000001000039a4  _hf_err+0x674
>  0x000000010001e864  _hf_ioerr+0x340
>  0x0000000100077b3c  _hf_erric+0xb4
>  0x000000010007eb7c  ._hf_cvis+0x80c
>  0x0000000100078048  _hf_cnvi+0x4d0
>  0x000000010045a0f0  ldi_othr+0x6b0
>  0x000000010045c8f0  ldi_ldi+0x6b4
>  0x0000000100453768  __hf_ldi+0x280
>  0x00000001004534a8  _hf_ldi+0x48
>  0x0000000100452460  species_m%%read_from_default_file+0x180
>  0x000000010044d660  species_m%%species_read+0xd60
>  0x00000001005078b8  geometry_m%%geometry_init_species+0x7b8
>  0x0000000100506744  geometry_m%%geometry_init+0x44
>  0x00000001009cab9c  system_m%%system_init+0x19c
>  0x00000001000dc184  run_m%%run_init+0x104
>  0x00000001000027ac  octopus+0x22ac
>  0x000000010000031c  __start+0x9c
>
> ****************************** Species *******************************
> Reading pseudopotential from file:
>      'O.psf'
>      Calculating atomic pseudo-eigenfunctions for species O ....
>      Done.
> Info: l =  0 component used as local potential
>
> **************************** FATAL ERROR *****************************
> *** Fatal Error (description follows)
> *--------------------------------------------------------------------
> * From node =    0
> *--------------------------------------------------------------------
> * Memory limit set in the input file was passed
> **********************************************************************
>
> I have used exactly the same input files as for the case of
> the run on PGI/Linux which runs OK. The error message "invalid
> character : is found in input data" is incomprehensible to me.
>
> Motohiko Tanaka
>



-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
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