[petsc-dev] PETSc VecView with MPIIO and complex numbers broken?

Barry Smith bsmith at mcs.anl.gov
Mon Mar 2 23:36:00 CST 2015


  Dave,

   I don't know what the cause of the error is. If I did I would have just fixed it.

   One should always do a configure and make and make alltests with complex before submitting anything for review. Of course, lazy bastards like Matt often skip it and screw up the tests.


   Barry

> On Mar 2, 2015, at 10:39 PM, Dave May <dave.mayhem23 at gmail.com> wrote:
> 
> Hi Barry,
> 
> I fixed all the PetscPrintf's to have the (double) cast.
> I am wondering if this could this be an uninitialized variable issue?
> 
> In that code, I have the following declaration
>   const PetscScalar test_values[] = { 0.311256, 88.068, 11.077444, 9953.62, 7.345, 64.8943, 3.1458, 6699.95, 0.00084, 0.0647 };
> which is then used to build a Vec object.
> I am not explicitly setting the imaginary part of the vector to zero when test_values is declared.
> 
> Apologize for the crappy code -  I really never ever use complex numbers in petsc.
> 
> If this is in fact the issue, please note that the same time of declaration appears at the top of
>   petsc / src / dm / examples / tutorials / ex15.c
> 
> 
> Cheers
>    Dave
> 
> 
> 
> 
> On 2 March 2015 at 20:31, Barry Smith <bsmith at mcs.anl.gov> wrote:
> 
> http://ftp.mcs.anl.gov/pub/petsc/nightlylogs/archive/2015/03/01/examples_master_arch-opensolaris-cmplx-pkgs-dbg_n-gage.log
> 
> ******* Testing: testexamples_C *******
> 4,7c4,7
> <   min(a)   = +8.40e-04 [loc 8]
> <   max(a)   = +9.95e+03 [loc 3]
> <   min(b)   = +8.40e-04 [loc 8]
> <   max(b)   = +9.95e+03 [loc 3]
> ---
> >   min(a)   = -4.21e+207 [loc 6]
> >   max(a)   = +8.10e-320 [loc 5]
> >   min(b)   = -4.21e+207 [loc 6]
> >   max(b)   = +8.10e-320 [loc 5]
> /export/home/petsc/petsc.clone-3/src/vec/vec/examples/tests
> Possible problem with with ex46, diffs above
> =========================================
> 4,7c4,7
> <   min(a)   = +8.40e-04 [loc 8]
> <   max(a)   = +9.95e+03 [loc 3]
> <   min(b)   = +8.40e-04 [loc 8]
> <   max(b)   = +9.95e+03 [loc 3]
> ---
> >   min(a)   = -4.21e+207 [loc 6]
> >   max(a)   = +8.10e-320 [loc 5]
> >   min(b)   = -4.21e+207 [loc 6]
> >   max(b)   = +8.10e-320 [loc 5]
> /export/home/petsc/petsc.clone-3/src/vec/vec/examples/tests
> Possible problem with with ex46_2, diffs above
> =========================================
> 4,7c4,7
> <   min(a)   = +8.40e-04 [loc 8]
> <   max(a)   = +9.95e+03 [loc 3]
> <   min(b)   = +8.40e-04 [loc 8]
> <   max(b)   = +9.95e+03 [loc 3]
> ---
> >   min(a)   = -4.21e+207 [loc 6]
> >   max(a)   = +8.10e-320 [loc 5]
> >   min(b)   = -4.21e+207 [loc 6]
> >   max(b)   = +8.10e-320 [loc 5]
> /export/home/petsc/petsc.clone-3/src/vec/vec/examples/tests
> Possible problem with with ex46_3, diffs above
> =========================================
> 6,9c6,9
> <   min(a)   = +8.40e-04 [loc 8]
> <   max(a)   = +9.95e+03 [loc 3]
> <   min(b)   = +8.40e-04 [loc 8]
> <   max(b)   = +9.95e+03 [loc 3]
> ---
> >   min(a)   = -4.21e+207 [loc 6]
> >   max(a)   = +8.10e-320 [loc 5]
> >   min(b)   = -4.21e+207 [loc 6]
> >   max(b)   = +8.10e-320 [loc 5]
> /export/home/petsc/petsc.clone-3/src/vec/vec/examples/tests
> Possible problem with with ex46_mpiio, diffs above
> =========================================
> 
> 




More information about the petsc-dev mailing list