[mpich2-dev] wrapping library calls

[Rob Latham]

On Tue, Nov 24, 2009 at 06:17:33AM -0600, Brian Smith wrote:

> The libraries in comm/fast are compiled using the XL compilers with 
> minimal asserts and a lot of debug turned off. They tend to make a 
> reasonable improvement for point-to-point operations and not much of an 
> improvement for collectives. We generally recommend you make sure your app 
> runs fine using the comm/bin scripts first, then, yes, for benchmark 
> results you can recompile against the comm/fast libraries.

At least in V1R3, I/O codes using the fast libraries often abort.
Kazutomo tells me it's related to something involving alignment of
floating point accesses?   I should try again with V1R4.

==rob

-- 
Rob Latham
Mathematics and Computer Science Division
Argonne National Lab, IL USA