[petsc-dev] Double BLAS in PETSc complex

[Matthew Knepley]

On Wed, Aug 28, 2013 at 2:36 PM, Jed Brown <jedbrown at mcs.anl.gov> wrote:

> Satish Balay <balay at mcs.anl.gov> writes:
>
> > On Wed, 28 Aug 2013, Matthew Knepley wrote:
> >
> >> The last error I have is needing the double version of BLAS when we are
> >> using complex. How do we handle that?
> >
> > Hm - I'm not sure.. Perhaps we need to add in a new prefix for
> > Blas(PetscReal) [and corresponding macros?] to petscblaslapack.h?
> > The #ifdef login in there will get more complex...
>
> What if we _always_ defined real and complex prototypes, so the only
> #ifdef logic would be in precision.  Then we define the scalar versions
> something like this:
>
> #define BLAS_Scalar_GEMM  SCALAR_SWITCH(BLAS_Real_GEMM,BLAS_Complex_GEMM)
>

This sounds better.

   Matt

-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20130828/df961342/attachment.html>