[petsc-dev] KNC build?

[Richard Mills]

I'm in complete agreement with you, Victor.  But we'll have to convince
someone at NERSC to set things up to work this way on Babbage, which
probably isn't going to happen since it's not a "production" resource.

--Richard

On Tue, Nov 4, 2014 at 3:58 PM, Victor Eijkhout <eijkhout at tacc.utexas.edu>
wrote:

>
>  On Nov 4, 2014, at 5:20 PM, Richard Mills <richardtmills at gmail.com>
> wrote:
>
> I noticed that you are using the wrong compiler wrappers.  Instead of
> 'mpicc', 'mpif90', and 'mpicxx' (which will invoke GNU compilers), you
> probably want 'mpiicc', 'mpiifort', and 'mpiicpc' (invoke the Intel
> compilers).
>
>
> May I remark that that’s a silly setup? It’s far more convenient to have
> the mpi compiler script to have the same name always, regardless of what
> compiler and MPI you have loaded.
>
>  I understand that you’re already using the software modules system, so
> there is no reason not to do it that way: the search path is dynamically
> set so you get the right script automatically.
>
>  Btw, if you haven’t replaced the environment modules by lmod (developed
> at TACC and backwards compatible but far better) you’re behind the times….
>
>  Victor.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20141104/c339fc7c/attachment.html>