[Octopus-devel] PFFT test

[David Strubbe]

Ok, you are partly right. The out-of-bounds errors disappear on make clean,
but the test still fails, starting with r8560. I am compiling with -g -O0
-fbounds-check.

Starting test run ...
/auto/dstrubbe/tigger/octopus/testsuite/components/../../bin//oct-test_mpi
Executing: cd /tmp/octopus.2siCgT;  /usr/bin/mpiexec -n 2
/auto/dstrubbe/tigger/octopus/testsuite/components/../../bin//oct-test_mpi
> out
        Elapsed time:     12.5 s

Finished test run.

 Execution                              :        [   OK   ]

-----------------------------------------

Match FFT           :

   Calculated value : 11.2069804832571
   Reference value  : 0.00035839
   Difference       : 11.2066220932571
   Tolerance        : 0.005

 FFT                                    :        [  FAIL  ]

-----------------------------------------



Using input file : ./10-hartree_pfft.02-fft_corrected.inp

Starting test run ...
/auto/dstrubbe/tigger/octopus/testsuite/components/../../bin//oct-test_mpi
Executing: cd /tmp/octopus.2siCgT;  /usr/bin/mpiexec -n 2
/auto/dstrubbe/tigger/octopus/testsuite/components/../../bin//oct-test_mpi
> out
        Elapsed time:     10.1 s

Finished test run.

 Execution                              :        [   OK   ]

-----------------------------------------

Match FFT Corrected :

   Calculated value : 0.4898658297960
   Reference value  : 0.00154839
   Difference       : 0.488317439796
   Tolerance        : 0.005

 FFT Corrected                          :        [  FAIL  ]


On Tue, Nov 15, 2011 at 12:09 AM, Micael Oliveira <micael at teor.fis.uc.pt>wrote:

> Hi David,
>
>  I'm using gfortran with the out-of-bounds check, and all the tests pass
> in my machine. Nevertheless, I often get this type of errors after changing
> a data type. It usually disappears by doing make clean. Could this be the
> same kind of problem?
>
>  Cheers
>
> Micael
>
>
>
>
> On 11/14/2011 08:47 PM, David Strubbe wrote:
>
>> I will shortly set up the tigger buildslave to use PFFT so it can check
>> the PFFT test, but when I run it manually I notice that recent changes
>> have made it fail. (On Friday it was passing.) See below.
>>
>> [12:44:48][tigger:components] ../oct-run_regression_test.pl
>> <http://oct-run_regression_**test.pl <http://oct-run_regression_test.pl>>
>> -D ../../bin/ -f 10-hartree_pfft.test
>>
>>  ***** Hartree *****
>>
>> Using workdir    : /tmp/octopus.bZtNpP
>> Using executable :
>> /auto/dstrubbe/tigger/octopus/**testsuite/components/../../**
>> bin//oct-test_mpi
>> Using test file  : 10-hartree_pfft.test
>>
>>
>> Using input file : ./10-hartree_pfft.01-fft.inp
>>
>> Starting test run ...
>> /auto/dstrubbe/tigger/octopus/**testsuite/components/../../**
>> bin//oct-test_mpi
>> Executing: cd /tmp/octopus.bZtNpP;  /usr/bin/mpiexec -n 2
>> /auto/dstrubbe/tigger/octopus/**testsuite/components/../../**
>> bin//oct-test_mpi
>>  > out
>> At line 139 of file fourier_space_inc.F90
>> Fortran runtime error: Array reference out of bounds for array 'op',
>> upper bound of dimension 1 exceeded (54 > 53)
>> At line 139 of file fourier_space_inc.F90
>> Fortran runtime error: Array reference out of bounds for array 'op',
>> upper bound of dimension 1 exceeded (54 > 53)
>> ------------------------------**------------------------------**
>> --------------
>> mpiexec has exited due to process rank 0 with PID 17809 on
>> node tigger exiting without calling "finalize". This may
>> have caused other processes in the application to be
>> terminated by signals sent by mpiexec (as reported here).
>> ------------------------------**------------------------------**
>> --------------
>> Elapsed time:     18.2 s
>>
>>
>>
>> Test run failed with exit code 512.
>>
>>  Execution                              : [  FAIL  ]
>>
>>
>>
>> Using input file : ./10-hartree_pfft.02-fft_**corrected.inp
>>
>> Starting test run ...
>> /auto/dstrubbe/tigger/octopus/**testsuite/components/../../**
>> bin//oct-test_mpi
>> Executing: cd /tmp/octopus.bZtNpP;  /usr/bin/mpiexec -n 2
>> /auto/dstrubbe/tigger/octopus/**testsuite/components/../../**
>> bin//oct-test_mpi
>>  > out
>> At line 139 of file fourier_space_inc.F90
>> Fortran runtime error: Array reference out of bounds for array 'op',
>> upper bound of dimension 1 exceeded (33 > 32)
>> ------------------------------**------------------------------**
>> --------------
>> mpiexec has exited due to process rank 0 with PID 17896 on
>> node tigger exiting without calling "finalize". This may
>> have caused other processes in the application to be
>> terminated by signals sent by mpiexec (as reported here).
>> ------------------------------**------------------------------**
>> --------------
>> Elapsed time:      4.0 s
>>
>>
>>
>> Test run failed with exit code 512.
>>
>>  Execution                              : [  FAIL  ]
>>
>>
>>
>> ______________________________**_________________
>> Octopus-devel mailing list
>> Octopus-devel at tddft.org
>> http://www.tddft.org/mailman/**listinfo/octopus-devel<http://www.tddft.org/mailman/listinfo/octopus-devel>
>>
>
> ______________________________**_________________
> Octopus-devel mailing list
> Octopus-devel at tddft.org
> http://www.tddft.org/mailman/**listinfo/octopus-devel<http://www.tddft.org/mailman/listinfo/octopus-devel>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.tddft.org/pipermail/octopus-devel/attachments/20111115/0db2ea7c/attachment.html>