ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
24 839 коммитов 3 Ветки 0 Теги
Граф коммитов

2527 Коммитов

Автор SHA1 Сообщение Дата
Rolf vandeVaart
7da614c75e Add ability for user to empty the CUDA IPC registration cache when it is full 2015-09-17 16:42:16 -04:00
Gilles Gouaillardet
975b6fd51b hwloc: do not count not allowed cores in df_search_cores 2015-09-17 13:10:34 +09:00
Ralph Castain
1b7930ad52 Silence some warnings and address Coverity issues 2015-09-16 07:58:22 -07:00
rhc54
5597416fe0 Merge pull request #897 from rhc54/topic/oob 
Remove the last involvement of the OOB system from the MPI layer
 2015-09-15 14:40:21 -07:00
Ralph Castain
c1bbbb5e2f Remove the last involvement of the OOB system from the MPI layer, remove the no-longer-needed usock/oob component, and have procs no longer open the RML, OOB, ROUTED, and GRPCOMM frameworks as PMIx now provides all required app-mpirun cmds 2015-09-15 13:08:35 -07:00
Rolf vandeVaart
555f14a479 Merge pull request #893 from rolfv/pr/more-verbose-fix 
Cleanup handle verbose messages
 2015-09-15 15:45:52 -04:00
George Bosilca
0e7e14449f Typo in the modex_recv. 2015-09-14 18:00:02 -04:00
Rolf vandeVaart
34fe2188cd Cleanup handle verbose messages 2015-09-14 11:01:25 -04:00
Gilles Gouaillardet
d5af5d106c btl/sm: mca_btl_sm_sendi: do not set *descriptor when descriptor is NULL 2015-09-14 14:04:40 +09:00
rhc54
33f5e4c766 Merge pull request #892 from rhc54/topic/pmix 
Fix the no-disconnect test by resolving a segfault on free - opal_dss…
 2015-09-11 16:01:42 -07:00
Ralph Castain
22d7c0081a Fix the no-disconnect test by resolving a segfault on free - opal_dss.unload will return the remaining unpacked portion of a buffer. As such, it cannot return the pointer to that info as it might be partway inside of a malloc'd region. So copy the data out of the buffer. 2015-09-11 13:01:35 -07:00
Rolf vandeVaart
90dd1d264b Fix cuda verbosity messages 2015-09-11 15:44:36 -04:00
Ralph Castain
dc5796b8a1 Revert "Revert "Fix the handling of cpusets so we get the correct cpuset for each local peer. Add the ability to indicate that a modex request is "optional" so we don't call the server if we don't find the value. Take advantage of that to allow the MPI layer to decide that the lack of locality info indicates non-local"" 
Fix the locality computation by correctly computing the vpid of the local peer

This reverts commit open-mpi/ompi@6a8fad49e5.
 2015-09-11 08:29:51 -07:00
Ralph Castain
6a8fad49e5 Revert "Fix the handling of cpusets so we get the correct cpuset for each local peer. Add the ability to indicate that a modex request is "optional" so we don't call the server if we don't find the value. Take advantage of that to allow the MPI layer to decide that the lack of locality info indicates non-local" 
This reverts commit f94f3cda214ab937c46802896fb53b84bec6cc3a.
 2015-09-11 02:01:25 -07:00
Ralph Castain
e0a52354d4 Sync to PMIx master at open-mpi/pmix@89680d6663 
Includes changes to support BigEndian machines
 2015-09-10 20:47:40 -07:00
Ralph Castain
a2a15cea8a Fix the s1 component so direct launch is supported for SLURM 2015-09-10 16:07:37 -07:00
rhc54
3430f154fc Merge pull request #885 from hppritcha/topic/pmix_not_pmix1xx_u16_prob 
pmix/~pmix1xx: use u32 for OPAL_PMIX_LOCAL_SIZE
 2015-09-10 15:38:54 -07:00
Nathan Hjelm
899bf548a2 opal/hwloc: fix topology detection when socket is above numa 
The OPAL_PROC_ON_* definitions have been changed from values to
flags. This should not cause any problems as these values were already
used as flags throughout the code base. Note, there will be a
difference between localities produced by the new code and the
old. For example, if a machine does not have a level-3 but two cores
share a level-1 or level-2 cache cache the level-3 bit will not be set
in the locality and OPAL_PROC_ON_LOCAL_L3CACHE will return 0. Before
this change it would have returned 1.

In addition the OPAL_PROC_ON_LOCAL_* macros have been simplified.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2015-09-10 14:17:45 -06:00
Howard Pritchard
2bbf22e2d0 pmix/~pmix1xx: use u32 for OPAL_PMIX_LOCAL_SIZE 
Looks like in ess_pmi_module.c u32 is being used
for retrieving OPAL_PMIX_LOCAL_SIZE, while s1/s2/cray
pmix components were storing as u16.

This commit fixes this problem.

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
 2015-09-10 11:41:39 -07:00
Ralph Castain
f94f3cda21 Fix the handling of cpusets so we get the correct cpuset for each local peer. Add the ability to indicate that a modex request is "optional" so we don't call the server if we don't find the value. Take advantage of that to allow the MPI layer to decide that the lack of locality info indicates non-local 2015-09-10 10:25:30 -07:00
Jeff Squyres
f7d90abf42 usnic: update for new add_procs() downcall behavior 2015-09-10 08:55:55 -06:00
Jeff Squyres
2f2d5ff855 btl.h: update comment for new add_procs behavior 2015-09-10 08:55:55 -06:00
Nathan Hjelm
2041aac4e4 btl/openib: add support for dynamic add_procs 
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2015-09-10 08:55:55 -06:00
Nathan Hjelm
40067f7ec4 btl/tcp: add support for dynamic add_procs 
This commit makes two changes to the tcp btl:

 - If a tcp proc does not exist when handling a new connection create
   a new proc and use it. The current implementation uses the
   opal_proc_by_name() function to get the opal_proc_t then calls
   add_procs on all btl modules. It may be sufficient to just call
   add_procs until an endpoint is created so this may change somewhat.

 - In add_procs add a check for an existing endpoint before creating
   one.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2015-09-10 08:55:55 -06:00
Nathan Hjelm
536aba1172 btl/portals4: add send flag to btl_flags 2015-09-10 08:55:55 -06:00
Nathan Hjelm
408da16d50 ompi/proc: add proc hash table for ompi_proc_t objects 
This commit adds an opal hash table to keep track of mapping between
process identifiers and ompi_proc_t's. This hash table is used by the
ompi_proc_by_name() function to lookup (in O(1) time) a given
process. This can be used by a BTL or other component to get a
ompi_proc_t when handling an incoming message from an as yet unknown
peer.

Additionally, this commit adds a new MCA variable to control the new
add_procs behavior: mpi_add_procs_cutoff. If the number of ranks in
the process falls below the threshold a ompi_proc_t is created for
every process. If the number of ranks is above the threshold then a
ompi_proc_t is only created for the local rank. The code needed to
generate additional ompi_proc_t's for a communicator is not yet
complete.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2015-09-10 08:55:54 -06:00
Nathan Hjelm
6f8f2325ed btl: btls are now required to set the send flag if supported 
This commit updates each non-compliant btl to send the
MCA_BTL_FLAGS_SEND flag in the btl_flags field if send is
supported. This fixes a problem identified after the latest bml/r2
update which excplicitly checks for the send flag.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2015-09-10 08:55:54 -06:00
Ralph Castain
4c47c498ac Sync to latest PMIx master 
Allow the blocking send and recv to keep trying
 2015-09-09 11:48:47 -07:00
Matias Cabral
f360eebfeb Merge pull request #855 from matcabral/btl_openib_mtu 
Fix for openib btl mca command line parameter btl_openib_mtu being ignored
 2015-09-09 11:22:00 -07:00
Gilles Gouaillardet
7f0ed74d24 pmix1xx: fix CPPFLAGS when DSO are not built 2015-09-09 14:20:12 +09:00
rhc54
f6b6b9a9ca Merge pull request #877 from rhc54/topic/s1s2 
Cleanup s1 and s2 components
 2015-09-08 19:20:59 -07:00
Ralph Castain
1cdb86b8c7 Cleanup s1 and s2 components, and ensure that mpirun and orteds only use non-direct-launch pmix components. 2015-09-08 18:37:09 -07:00
Gilles Gouaillardet
6e6a3e965c pml: do not cast way the const modifier when this is not necessary 
update the pml framework and mpi c bindings
 2015-09-09 09:18:57 +09:00
rhc54
3a446c9797 Merge pull request #876 from rhc54/topic/hnp 
Fix segfault upon job error
 2015-09-08 15:10:51 -07:00
rhc54
47f437608d Merge pull request #875 from rhc54/topic/dynamics 
Stop a segfault in the test by correctly passing all the argv during spawn
 2015-09-08 14:35:42 -07:00
Ralph Castain
459f169e06 Fix segfault upon job error 
Silence some unnecessary error-logs
 2015-09-08 14:03:06 -07:00
Ralph Castain
ae7156cabb Stop a segfault in the test by correctly passing all the argv during spawn 2015-09-08 13:42:46 -07:00
Rolf vandeVaart
188c30a01a Merge pull request #867 from rolfv/pr/openib-hwloc-verbosity 
Add some verbosity to help debug hwloc issues
 2015-09-08 14:43:35 -04:00
rhc54
8053357fcc Merge pull request #873 from rhc54/topic/static 
Add the libs required for PMIx to support static builds (and trim all excess whitespace)
 2015-09-08 11:28:47 -07:00
Rolf vandeVaart
2e64a69fa9 Add some verbosity to help debug hwloc issues 2015-09-08 10:50:22 -07:00
Ralph Castain
291afe502f Add the libs required for PMIx to support static builds 
Remove unneeded CPPFLAGS
 2015-09-08 10:21:06 -07:00
Jeff Squyres
bc9e5652ff whitespace: purge whitespace at end of lines 
Generated by running "./contrib/whitespace-purge.sh".
 2015-09-08 09:47:17 -07:00
Ralph Castain
e6add86e4f Deal with connect/accept between two jobs from different mpirun's. Somewhat optimize connect/accept by using MPI bcast to distribute the participants instead of another PMIx lookup. Cleanup some Coverity issues. 2015-09-07 09:19:24 -07:00
Ralph Castain
37c3ed68e7 Cleanup connect/disconnect and bring comm_spawn back online! 2015-09-06 10:27:39 -07:00
Jeff Squyres
f782a7640e usnic: minor re-order of Makefile.am sources 
Put the hwloc.c file alphabetically in the list.
 2015-09-05 05:02:00 -07:00
rhc54
665b30376a Merge pull request #868 from rhc54/topic/hwloc 
Remove OPAL_HAVE_HWLOC qualifier and error out if --without-hwloc is given
 2015-09-04 17:58:07 -07:00
Ralph Castain
d97bc29102 Remove OPAL_HAVE_HWLOC qualifier and error out if --without-hwloc is given 2015-09-04 16:54:40 -07:00
rhc54
d45ccda813 Merge pull request #866 from rhc54/topic/updatepmix 
Update PMIx support
 2015-09-04 11:09:36 -07:00
Ralph Castain
f6948c2bb4 Sync with PMIx master 43e45c3. Get multi-node publish/lookup/unpublish working 2015-09-04 10:07:17 -07:00
Rolf vandeVaart
ebfd00b66e While debugging user problems, these extra verbosity statements would be helpful 2015-09-03 17:15:39 -04:00