[petsc-dev] Error while building MAGMA from PETSc

[Barry Smith]

   Have no fear. I was not planning to globalize HAVE_CUBLAS, nor have I.

   Barry

On Aug 15, 2013, at 9:02 PM, Jed Brown <jedbrown at mcs.anl.gov> wrote:

> Barry Smith <bsmith at mcs.anl.gov> writes:
>>     If they have made the bad decision to use magma then they have to
>>     live with the bad ramifications of using magma. Aren't you of the
>>     generation that was raised to understand the consequences of your
>>     actions?
> 
> Some overzealous sysadmin is going to compile the system PETSc with
> support for MAGMA and then the poor user's code is going to break and
> we're going to be the ones debugging.
> 
> I also prefer defining it in the impl file because then it doesn't
> contaminate other compilation units.