Cray-X1 report

[Robert Latham]

Hi

Just wanted to report that early testing on the Cray X1 at ORNL has
gone pretty well.  nc_test passes in the single processor case as long
as you up MPI_TYPE_MAX (4096 worked for me).    There are oddities in
the multiprocessor case, but they look more like operator error than
anything pnetcdf-specific.

I've added a README.CRAY-X1 document to the pnetcdf tree documenting
the (few) steps needed to build and run.

==rob

-- 
Rob Latham
Mathematics and Computer Science Division    A215 0178 EA2D B059 8CDF
Argonne National Labs, IL USA                B29D F333 664A 4280 315B