[petsc-dev] Enabling single precision for MUMPS and SUPERLU

[Jed Brown]

On Fri, Sep 21, 2012 at 10:02 AM, Barry Smith <bsmith at mcs.anl.gov> wrote:

>   Who the hell did that? :-)
>

You did, naturally. ;-)

http://petsc.cs.iit.edu/petsc/petsc-dev/rev/18da6d3cac89


> > and even BLAS/Lapack.
>
>    Ok, you convinced me. I'm fine with the macros. Maybe better names? Are
> the umfpack names ok, they don't have PETSc in them.
>

I don't know why you chose those names. They are private to umfpack.c so
it's only a problem if Umfpack itself starts using the umfpack_UMF_ prefix.
What would be like PETSc declaring Petsc_PETSC_MatCreate() which seems
unlikely enough that I don't mind leaving it (mostly to preserve usefulness
of "hg annotate").
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120921/58be28a1/attachment.html>