Matthew MacManes wrote:
I would be happy to help troubleshoot, but I am not much of a programmer to
know how. The hang is reproducible, and -mca btl ^sm is about 15% faster.
if you want to shoot me some instructions off list, I can give it a go.
The application that I am working with, primarily, is ABySS:
http://www.bcgsc.ca/platform/bioinfo/software/abyss
How about this? File a trac ticket for each issue (hang with more
FIFOs, ^sm is 15% faster) describing in the simplest terms possible how
to reproduce each problem (which OMPI rel, simple test code if possible
so people don't need to come up to speed on ABySS, etc.).
