George Bosilca 16d9f71d01 Correctly compute the space needed for the args. ...

Add checks to bail out if our precomputed value is less
than needed (we are already at fault).

bot:milestone:v1.10.3
bot:milestone:v2.0
bot🏷️bug
bot:assign: @ggouaillardet

2016-05-21 16:01:16 -04:00

..

attribute

more c99 updates

2015-06-25 10:14:13 -06:00

class

Purge whitespace from the repo

2015-06-23 20:59:57 -07:00

communicator

Merge pull request #1397 from hjelmn/enable_thread_multiple

2016-04-23 08:40:22 -06:00

contrib

Purge whitespace from the repo

2015-06-23 20:59:57 -07:00

datatype

Correctly compute the space needed for the args.

2016-05-21 16:01:16 -04:00

debuggers

Merge pull request #1640 from jsquyres/pr/mpir-cleanup

2016-05-05 21:23:30 -04:00

dpm

ompi/dpm: retrieves OPAL_PMIX_ARCH in heterogeneous mode

2016-02-22 11:01:06 +09:00

errhandler

Per request from Jeff, aggregate all help messages during MPI_Init thru MPI_Finalize as long as the RTE is available

2016-04-15 13:37:22 -07:00

etc

Purge whitespace from the repo

2015-06-23 20:59:57 -07:00

file

io: do not cast way the const modifier when this is not necessary

2015-09-09 09:18:58 +09:00

group

ompi/group: fix sparse group proc reference counting

2016-04-27 15:55:13 -06:00

include

fortran: add missing constants for MPI_WIN_CREATE_FLAVOR and MPI_WIN_MODEL

2016-03-14 10:19:21 +09:00

info

ompi: fixup hostname max length usage

2016-04-25 07:08:23 +02:00

mca

bml/r2: always add btl progress function

2016-05-21 15:54:04 -04:00

message

Purge whitespace from the repo

2015-06-23 20:59:57 -07:00

mpi

Only check the parameters on valid dimensions.

2016-05-21 15:54:04 -04:00

mpiext

configury: remove the --enable-mpi-profiling option

2015-10-13 08:52:35 +09:00

op

op: allow user operations in ompi_3buff_op_reduce

2015-10-02 10:35:21 -06:00

patterns

Purge whitespace from the repo

2015-06-23 20:59:57 -07:00

peruse

more c99 updates

2015-06-25 10:14:13 -06:00

proc

ompi_proc_complete_init_single: make the subroutine public

2016-02-22 11:01:06 +09:00

request

Make the request lock recursive.

2016-04-26 16:16:07 -04:00

runtime

When direct launching applications, we must allow the MPI layer to progress during RTE-level barriers. Neither SLURM nor Cray provide non-blocking fence functions, so push those calls into a separate event thread (use the OPAL async thread for this purpose so we don't create another one) and let the MPI thread sping in wait_for_completion. This also restores the "lazy" completion during MPI_Finalize to minimize cpu utilization.

2016-05-14 16:37:00 -07:00

tools

Merge pull request #1552 from kmroz/wip-hostname-len-cleanup-1

2016-05-02 09:44:18 -04:00

win

win: silence a warning in alloc_window(...)

2016-03-28 14:57:31 +09:00

Makefile.am

use-mpi extensions do not have a .la lib, so the fortran module should not depend on them.

2016-05-03 11:54:35 -04:00