ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
24 367 коммитов 3 Ветки 0 Теги
Граф коммитов

24367 Коммитов

This Branch
This Branch
Все ветки
Автор SHA1 Сообщение Дата
Edgar Gabriel
a9ca37059a improve the communicaton abstraction. This commit also allows all aggregators to work simultaniously, instead of the slightly staggered way of the previous version. 2016-01-17 09:48:49 -06:00
Edgar Gabriel
56e11bfc97 initialize the stripe_size variable as well. 2016-01-17 09:48:49 -06:00
Edgar Gabriel
26c57ef374 separate the size of the buffer used for the shuffle step and the size of the buffer used for a pwritev operation. 2016-01-17 09:48:49 -06:00
Edgar Gabriel
39d5c8c281 further bug fixes silencing a compiler warning and fixing a memory overrun 2016-01-17 09:48:49 -06:00
Edgar Gabriel
2bcae84e11 further debugging 2016-01-17 09:48:49 -06:00
Edgar Gabriel
2bdd6ba17a correctly free some buffers, and ensure that lustre_stripe_size and stripe_count are always read from the file system. 2016-01-17 09:48:49 -06:00
Edgar Gabriel
4bbb22bd0b add a new field to the ompio data structure (stripe_count) and set it correctly on pvfs2 and lustre. 2016-01-17 09:48:49 -06:00
Edgar Gabriel
d282e94b67 add the new dynamic_gen2 component, designed to coexist for now with the original dynamic component 2016-01-17 09:48:49 -06:00
rhc54
b172b8599b Merge pull request #1285 from ggouaillardet/topic/pmix_dist_fix 
pmix: do not include automatically generated include/private/autogen/…
 2016-01-16 20:49:41 -08:00
Ralph Castain
fc6b260146 Protect against PMIx-based requests that don't come thru the MPI comm_spawn interface 2016-01-16 13:36:06 -08:00
Ralph Castain
4dad5de8ff Silence a couple of warnings - strncpy returns a char*, not an int 2016-01-16 09:44:52 -08:00
Jeff Squyres
348ac507c2 usnic: explain why we still have OPAL_HAVE_HWLOC 
Put in a comment explaining why btl_usnic_compat.h still defines
OPAL_HAVE_HWLOC, even though master/v2.x no longer does.
 2016-01-16 04:11:05 -08:00
Jeff Squyres
0f5fcf9029 usnic: fix common symbol 2016-01-16 03:55:27 -08:00
Jeff Squyres
6c96cb1ad0 find_common_syms: arrgh -- re-add the x bit 
Previous commit accidentally removed the x bit from this script.  This
commit puts it back.
 2016-01-16 03:53:43 -08:00
Jeff Squyres
60ffe713b8 common syms: whitelist bison-generated common symbols 
Bison generates some common symbols that we can't do anything about,
so whitelist them.
 2016-01-16 03:53:14 -08:00
Jeff Squyres
96f94f8228 fortran: whitelist deliberate common symbols 
The Fortran library has a number of common symbols that are
deliberate, so whitelist them.
 2016-01-16 03:53:14 -08:00
Jeff Squyres
c43d4fd915 find_common_syms: trivial updates 
Look for "common_sym_whitelist.txt" files (not
"common_sym_whitelist").  Also, skip blank lines in the
whitelistfiles, too.
 2016-01-16 03:53:14 -08:00
rhc54
ef24f710a7 Merge pull request #1303 from timattox/remove_unused_var 
hwloc_base_util.c: Remove newly unused variable 'i'.
 2016-01-15 00:35:14 -08:00
Tim Mattox
958de82471 hwloc_base_util.c: Remove newly unused variable 'i'. 2016-01-14 16:35:47 -05:00
Joshua Ladd
18c5a21562 Fix typo in error handling flow. 2016-01-14 22:28:54 +02:00
Joshua Ladd
afa62d8ca1 Addressing reviewers' comments for https://github.com/open-mpi/ompi-release/pull/891 2016-01-14 19:22:27 +02:00
Gilles Gouaillardet
1d38430e43 opal: replace opal_convert_jobid_to_string with opal_snprintf_jobid 2016-01-14 10:39:03 +09:00
Jeff Squyres
e5cf2db3b7 Merge pull request #1291 from jsquyres/pr/hotel-fix 
opal hotel: only delete events that have not yet fired
 2016-01-13 14:51:51 -05:00
Jeff Squyres
270cc11156 opal hotel: only delete events that have not yet fired 
The eviction callback, for convenience (and to avoid code
duplication), use to call opal_hotel_checkout().  However,
opal_hotel_checkout() deletes the eviction event -- which is fine to
do when opal_hotel_checkout() is invoked by the application.  But when
it's invoked by the same event that it's deleting, it can cause Bad
Things to happen.

For simplicity, instead of invoking opal_hotel_checkout() from the
eviction callback, just duplicate the checkout logic into the eviction
callback function (and skip the delete-the-evict-event part).

For good measure, put a comment in all three places where the checkout
logic occurs (because it's inlined): don't change this logic without
changing all 3 places.

Finally, also add a line in the docs for opal_hotel_init() warning
users from calling opal_hotel_checkout() from their eviction
callback.
 2016-01-13 10:59:06 -08:00
rhc54
eb65b5f97e Merge pull request #1297 from timattox/use_hwloc_bitmap_weight 
Replace a bit counting loop with an efficient population count
 2016-01-13 09:05:13 -08:00
rhc54
26e882c1c3 Merge pull request #1300 from ggouaillardet/topic/oversubscribe 
orte_rmaps_base_map_job: set OPAL_BIND_ALLOW_OVERLOAD when needed
 2016-01-13 09:00:02 -08:00
Gilles Gouaillardet
4c43fb2a50 orte_rmaps_base_map_job: set OPAL_BIND_ALLOW_OVERLOAD when needed 2016-01-13 17:13:36 +09:00
Tomislav Janjusic
3858bc8e62 Adding support for dynamic endpoint creation 
Signed-off-by: Tomislav Janjusic <tomislavj@mngx-apl-01.mtl.labs.mlnx>
Signed-off-by: Tomislavj Janjusic <tomislavj@mellanox.com>
Signed-off-by: Joshua Ladd <jladd.mlnx@gmail.com>
 2016-01-12 22:17:03 +02:00
Nathan Hjelm
dd4d49cbbb Merge pull request #1278 from ggouaillardet/poc/osc_pt2pt 
osc/pt2pt: use two distinct "namespaces" for tags
 2016-01-12 09:49:31 -07:00
Tim Mattox
f2d4a8d266 Replace a bit counting loop with a call to an efficient population count routine 2016-01-12 10:48:56 -05:00
Gilles Gouaillardet
955fe85cb6 pmix/pmix120: add missing include file 2016-01-12 11:35:32 +09:00
Nathan Hjelm
b6366e52a8 Merge pull request #1294 from hjelmn/group_fix 
ompi/group: do not decrement parent group proc pointers in destruct
 2016-01-11 13:49:23 -07:00
Nathan Hjelm
d26cc3fece ompi/group: do no decrement parent group proc pointers in destruct 
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2016-01-11 12:56:11 -07:00
Ralph Castain
332019b43a Silence warning 2016-01-10 09:59:36 -08:00
Nathan Hjelm
aefc2ed3e8 Merge pull request #1283 from artpol84/udcm_race_fix 
Fix race condition in UDCM
 2016-01-09 08:19:20 -07:00
Gilles Gouaillardet
0fb7b07a71 opal/progress: fix non debug builds 
this bug was introduced in open-mpi/ompi@64b695669a

Thanks Pavel (Pasha) Shamis for reporting this issue
 2016-01-09 15:47:40 +09:00
Artem Polyakov
84e4fb308b Fix race condition in UDCM where service thread sees that 
`cm_message_event_active == 1` but main thread has already stopped
processing messages and thus we will have the situation where one
message was left unhandled leading to a hang.
 2016-01-08 23:56:21 +06:00
Gilles Gouaillardet
73daf58ee5 pmix: do not include automatically generated include/private/autogen/config.h into dist tarball 
Thanks Siegmar Gross for the initial report of this issue
 2016-01-08 13:18:15 +09:00
Edgar Gabriel
ac34c0ec51 Merge pull request #1287 from edgargabriel/posix-fbtl-update 
use the actual preadv and pwritev functions if available. That's what…
 2016-01-07 19:48:51 -06:00
Nathan Hjelm
faeca5663c Merge pull request #1289 from hjelmn/hwloc_fix 
Update hwloc to 1.11.2 + Fix /proc/mounts issue.
 2016-01-07 16:14:55 -07:00
Nathan Hjelm
15007b4e2b linux: use mntent.h instead of manually parsing /proc/mounts 
setmntent() doesn't support root_fd, but manual parsing of
/proc/mounts is fragile, and actually buggy for very long mount lines
(see open-mpi/hwloc#142 (comment)).

Since we only openat("/proc/mounts") there, just manually concatenate
the fsroot_path and use setmntent().

Thanks to Nathan Hjelm for the report.

(Cherry-picked from open-mpi/hwloc@d2d07b9a22)

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2016-01-07 12:55:03 -07:00
Nathan Hjelm
1384559fcd Update hwloc to v1.11.2 
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2016-01-07 12:33:12 -07:00
rhc54
73ca19581e Merge pull request #1288 from hjelmn/grpcomm_fixes 
grpcomm: fix bugs in grpcomm algorithms
 2016-01-07 10:24:03 -08:00
Nathan Hjelm
fab1eca536 grpcomm: fix bugs in grpcomm algorithms 
This commit fixes multiple issues in the bruck's and recursive
doubling grpcomm algorithms. The following changes are included:

 - Use the existing bitmap implementation instead of implementing a
   new one. There were bugs in the implementation that caused an
   overrun of the bitmap array.

 - Clean up the algorithms to eliminate errors.

 - Send as little extra data as possible in the bruck's
   algorithm.

The changes were testest with various numbers of ortes varying from 1
to 4096.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2016-01-07 10:12:08 -07:00
Edgar Gabriel
0a1b735eed use the actual preadv and pwritev functions if available. That's what the fbtl interfaces have been designed for. 2016-01-07 08:29:17 -06:00
Nysal Jan K.A
13f9bb9202 Use PMI2 constants for consistency 2016-01-07 11:46:22 +05:30
Gilles Gouaillardet
713e3ea2e5 configury: fix pthread_join() call in OPAL_INTL_PTHREAD_TRY_LINK_FORTRAN 2016-01-07 10:20:20 +09:00
Gilles Gouaillardet
4c1ea4a171 dpm: correctly handle procs_cutoff in ompi_dpm_connect_accept() 
this commit includes missing bits from open-mpi/ompi@213b2abde4
 2016-01-07 09:11:03 +09:00
Gilles Gouaillardet
213b2abde4 dpm: correctly handle procs_cutoff in ompi_dpm_connect_accept() 2016-01-06 16:21:13 +09:00
Jeff Squyres
e4bdad09c1 pmix: remove extra wrapper LIBS 
These extra libs are now no longer necessary.

Fixes open-ompi/ompi#1281.
 2016-01-05 12:09:53 -08:00