Jeff Squyres e36038bb17 We know that --enable-progress-threads doesn't work. But this allows ...

it to at least compile.  If you actually get to the point of invoking
the openib btl progress thread, you'll get a big opal_output warning
that it is pretty much guaranteed not to work.

This commit was SVN r15628.

2007-07-26 00:58:56 +00:00

..

base

reapply r15517 and r15520, which were removed in r15527 so that I could get

2007-07-20 02:34:29 +00:00

elan

Remove all warnings.

2007-07-23 21:06:25 +00:00

gm

reapply r15517 and r15520, which were removed in r15527 so that I could get

2007-07-20 02:34:29 +00:00

mvapi

reapply r15517 and r15520, which were removed in r15527 so that I could get

2007-07-20 02:34:29 +00:00

mx

reapply r15517 and r15520, which were removed in r15527 so that I could get

2007-07-20 02:34:29 +00:00

ofud

reapply r15517 and r15520, which were removed in r15527 so that I could get

2007-07-20 02:34:29 +00:00

openib

We know that --enable-progress-threads doesn't work. But this allows

2007-07-26 00:58:56 +00:00

portals

* Provide flag if the BTL can do RDMA, but requires a prepare_{src,dst}

2007-07-11 21:21:40 +00:00

self

Rename btl_rdma_offset to btl_pipeline_send_length.

2007-06-21 07:12:40 +00:00

sm

Move modex from pml base to general ompi runtime, sicne it's used by more

2007-07-09 17:16:34 +00:00

tcp

Add MCA parameter to enable/disable Nagle's algorithm on the TCP BTL.

2007-07-25 12:21:00 +00:00

template

fix template btl

2007-07-13 20:06:22 +00:00

udapl

add a missing subroutine prototype

2007-07-24 21:07:57 +00:00

btl.h

cleanup btl.h comments

2007-07-25 17:26:23 +00:00

Makefile.am

Clean up a couple of configure things:

2007-05-15 04:23:48 +00:00