rhc [Tue, 06 Nov 2012 05:22:07 +0000] rev 20965
Fixes #3387: Please integrate IB performance fix into v1.7
---svn-pre-commit-ignore-below---
r27559 [[BR]]
performance optimization: process completions in the batch manner
rhc [Tue, 06 Nov 2012 05:20:31 +0000] rev 20964
Fixes #3381: Move minor Fortran08 Makefile.am fix to v1.7
---svn-pre-commit-ignore-below---
r27554 [[BR]]
* MPI_Reduce_scatter_block is now implemented, so remove it from the
"not yet implemented" list (it was already in the mpi_api_files
macro, actually)
* Add mistakenly missed palltoall file. Thanks to Brian for
noticing its absence.
rhc [Tue, 06 Nov 2012 05:18:56 +0000] rev 20963
Fixes #3379: Please integrate ConnectIB param into v1.7
---svn-pre-commit-ignore-below---
r27552 [[BR]]
SRQ now supported in ConnectIB
rhc [Tue, 06 Nov 2012 05:17:33 +0000] rev 20962
Fixes #3377: Move autogen fix from trunk to v1.7
---svn-pre-commit-ignore-below---
r27548 [[BR]]
Add autogen.sh to list of files to be distributed, so autogen works off a
tarball
rhc [Tue, 06 Nov 2012 05:16:12 +0000] rev 20961
Fixes #3388: Move fix for VT+CUDA build failure into v1.7 branch
---svn-pre-commit-ignore-below---
r27560 [[BR]]
Do forward '--with-cuda[-libdir]' arguments to the VT configure
hjelmn [Mon, 05 Nov 2012 20:51:10 +0000] rev 20960
Fix leaks in the opal if posix code. Error paths were not calling OBJ_RELEASE on an opal_if_t created with OBJ_NEW.
This affects both trunk and 1.7 and might affect 1.6.
cmr:v1.7
rhc [Mon, 05 Nov 2012 16:59:53 +0000] rev 20959
If the nodename is an IP address, we need to retain the full name (even if keep_fqdn is false) so that the ssh tree spawn can proceed.
cmr:v1.7
jurenz [Mon, 05 Nov 2012 14:53:00 +0000] rev 20958
Do forward '--with-cuda[-libdir]' arguments to the VT configure
miked [Mon, 05 Nov 2012 14:02:37 +0000] rev 20957
performance optimization: process completions in the batch manner
jsquyres [Sat, 03 Nov 2012 20:40:06 +0000] rev 20956
r27471 wasn't entirely correct: the libmpi_f90_so_version should have
been updated to 4:0:3 (not 4:0:1). The commit message stated that we
''changed'' and interface in 1.6.1, but that wasn't quite correct.
The change was a compile-time issue, not a link-time issue. It
changed a parameter from being INOUT to IN, but that doesn't change
the link-time semantics.