[Pw_forum] Want test crash

[mashaojie163]

Dear Sir:
     did you run the whole test serially or in parallel ??
> note that only the transport part of WanT support parallelization at the 
> moment
I think the error is resulted in because  I compiled want and excute want in parallel. So now I compile want  and excute it seriallly
1) I compiled it by  ./configue F90=pgf90 ;  make all 
everything is OK
 2)  and I excute DFT calculation in parallel 
 I modify PARA_PREFIX="mpirun -np 2 -machinefile ../nodefile"
./run.sh dft
3)  I excute want serially
PARA_PREFIX=
./run.sh want

everything is OK until band.x crash. The error in band.out is 

  DFT-data fmt automaticaly detected: qexml

 *** from PE    :     0
 *** in routine : lattice_read_ext
 *** error msg. : reading lattice
 *** error code :     1
 *** aborting ***
4)  ./run.sh check
running CHECK... 

######################### File: disentangle.out 

Variable           This File       Reference    % Difference

iteration                  1               1        0.000000    ok
omega_i            26.785130       26.785104        0.000001    ok
omega_i_avrg        6.696282        6.696276        0.000001    ok
status             CONVERGED       CONVERGED


######################### File: wannier.out 

Variable           This File       Reference    % Difference

iteration               8500            8500        0.000000    ok
center_sum         28.627843       28.627818        0.000001    ok
omega_i            26.785130       26.785104        0.000001    ok
omega_d             0.000006        0.000006        0.000000    ok
omega_od            1.842708        1.842709        0.000001    ok
omega_tot          28.627843       28.627818        0.000001    ok
omega_avrg          7.156961        7.156955        0.000001    ok
onsite_hr           5.215476        5.215477        0.000000    ok
status           END_OF_ITER     END_OF_ITER



Then I ecute test02.
1) .PARA_PREFIX="mpirun -np 2 -machinefile ../nodefile" 
  ./run_Cu.sh dft
  every thing is ok
2) .PARA_PREFIX=
  ./run_Cu.sh want
running DISENTANGLE calculation... CRASH
 done
running WANNIER calculation... CRASH
 done
running BANDS calculation... CRASH
 done
running DOS calculation... CRASH
 done
running PLOT calculation... CRASH
 done

The same error as above occur  in disentangle.out wannier.out ...etc.

  DFT-data fmt automaticaly detected: qexml

 *** from PE    :     0
 *** in routine : lattice_read_ext
 *** error msg. : reading lattice
 *** error code :     1
 *** aborting ***
Why the error occur?
----- Original Message ----- 
From: "Andrea Ferretti" <ferretti.andrea at unimore.it>
To: <pw_forum at pwscf.org>
Sent: Monday, July 02, 2007 10:12 PM
Subject: (ÈðÐÇÌáʾ-´ËÓʼþ¿ÉÄÜÊÇÀ¬»øÓʼþ)Re: [Pw_forum] Want test crash


> Dear Mashao,
> 
>> Thank you for your reponse.
>> I excute ./run.sh all . and the other calculations  
>> terminated normally until the bands.x excute.
> 
> ok...
> did you run the whole test serially or in parallel ??
> note that only the transport part of WanT support parallelization at the 
> moment
> 
> if the test crashed when doing bands.x calculaiton, it means it was 
> successful in e.g. wannier localization, is it true ??
> 
> 
>> Another question: I can not compile want successfully with 
>> intel compiler. 
>> So I compiled it with PGI compiler. I have compiled  
>> PWSCF with intel compiler successfully!
>> 
> 
> please, send also a log of the compilation failure with the intel 
> compiler; moving to PGI, try first compiling serially (which should be 
> the defualt for WanT configure)
> 
> --
> Andrea Ferretti
> National Research Center S3, CNR-INFM  ( http://s3.infm.it )
> Dip. Fisica, Univ. di Modena e Reggio E. (Italy)
> Tel: +39 059 2055301;  Fax: +39 059 374794;  Skype: andrea_ferretti
> URL: http://www.nanoscience.unimo.it
> 
> Please, if possible, don't send me MS Word or PowerPoint attachments
> Why? See:  http://www.gnu.org/philosophy/no-word-attachments.html
> 
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>