Last time I played with the async progress I used the TCP BTL, and it
worked. The disclaimer here is that it was few month ago (about a year
to be really honest) ...
george.
On Nov 24, 2009, at 12:11 , Barrett, Brian W wrote:
Jeff -
I was going to look into the ROMIO problems this weeke
Jeff -
I was going to look into the ROMIO problems this weekend. I started looking
at them on the train between Portland and Seattle after the Forum (trying to
get async progress to work for some tests of the RMA code), but got too
confused. We're doing requests absolutely incorrectly with ROMIO
Per the call today, I've added a Big Hairy Warning if you --enable-
progress-threads. It's a configure change, so I'll "sleep 6h; svn
ci ".
Note that I can't compile the trunk right now if --enable-progress-
threads is used:
CC base/io_base_request.lo
base/io_base_request.c:254:
