Skip to content

update to most recent disco ompi #6

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 622 commits into from
Feb 26, 2016
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
622 commits
Select commit Hold shift + click to select a range
f0e3e16
pmix/base: add missing #include <unistd.h>
ggouaillardet Dec 24, 2015
0b3e3c6
opal/runtime: add missing #include <unistd.h>
ggouaillardet Dec 24, 2015
41c325f
Shift common code for calculating a port count and btl_rank in openib
artpol84 Dec 23, 2015
400af6c
openib addproc improvements:
artpol84 Dec 23, 2015
3031aff
Fix openib process accounting if procs was dynamically added.
artpol84 Dec 24, 2015
8ab28cd
Fix a typo that causes segfaults on multi-node executions
Dec 24, 2015
2d9aa38
btl/openib: fix heterogeneous support
ggouaillardet Dec 25, 2015
a20826e
Fix vader resource leak.
artpol84 Dec 27, 2015
6fd5041
Merge pull request #1264 from artpol84/vader_resource_leak_fix
hjelmn Dec 27, 2015
700a210
Merge pull request #1260 from artpol84/openib_proc_account_fix
hjelmn Dec 27, 2015
47ab2fc
man: fix MPI_Neighbor_alltoall{v,w} prototypes
ggouaillardet Dec 28, 2015
352b05a
rmaps: warn if oversubscribing when manually setting the number of hosts
ggouaillardet Dec 25, 2015
fec973e
configury: test portability
ggouaillardet Dec 28, 2015
4f8c0c5
event/external: add missing include file
ggouaillardet Dec 28, 2015
d416c7f
pmix/external: no more circular dependencies if not building shared DSO
ggouaillardet Dec 28, 2015
55ae676
configury: use --with-pmix option instead of --with-external-pmix
ggouaillardet Dec 28, 2015
1157329
configury: misc pmix112 fixes
ggouaillardet Dec 28, 2015
c757c5c
pmix/external: Fix error handler usage
ggouaillardet Dec 28, 2015
810f244
Add pmix120 component, update the error handling functions in the PMI…
Dec 17, 2015
a5b95a0
Continue work on error notification system
Dec 25, 2015
3300d7c
pmix: rename pmix_munge_module
ggouaillardet Dec 28, 2015
0ca1ee5
configury: misc pmix120 fixes
ggouaillardet Dec 28, 2015
b7b4231
Merge pull request #1261 from ggouaillardet/topic/rmaps_rr_oversubsri…
Dec 28, 2015
b20a219
hwloc/external: abort if hwloc v2 is detected since it is not yet sup…
ggouaillardet Dec 29, 2015
5dfb7ac
Merge pull request #1266 from ggouaillardet/topic/misc_pmix_fixes
Dec 29, 2015
ddf0f27
Fix typo
Dec 29, 2015
12dccaa
Standardize the handling of shutdown in the OOB TCP component
Dec 29, 2015
dc857a9
Merge pull request #1267 from rhc54/topic/oob
Dec 29, 2015
2abb297
Fix Mellanox copyrights with respect to the following PRs:
artpol84 Dec 29, 2015
a04f1cd
Silence some Coverity warnings
Dec 30, 2015
1cdc1c1
Revert "Standardize the handling of shutdown in the OOB TCP component"
Dec 30, 2015
0a6b8d2
Correctly handle connection terminations during finalize so mpirun do…
Dec 30, 2015
d387725
Merge pull request #1270 from rhc54/topic/cleanup
Dec 30, 2015
f53d3c7
Silence warning
Dec 30, 2015
d80b98b
Merge pull request #1268 from artpol84/fix_mlnx_copy
artpol84 Jan 2, 2016
14fdf75
fs/pvfs2: fix typo
ggouaillardet Jan 3, 2016
6d073a8
btl_sm: add a comment explaining why we rename(2)
jsquyres Jan 4, 2016
06ecdb6
osc/pt2pt: use two distinct "namespaces" for tags
ggouaillardet Jan 5, 2016
da309ac
- use a unique pid for each process as requested by the API
edgargabriel Jan 4, 2016
7861a8c
revise the logic in the fbtl plfs avoiding the memcpy operation
edgargabriel Jan 5, 2016
1b0b849
remove the MCA parameter setting the number of hosts in PLFS, since t…
edgargabriel Jan 5, 2016
f8722a9
travis: add a bunch of comments explaining things
jsquyres Jan 5, 2016
d827da5
Merge pull request #1279 from edgargabriel/pr/plfs-fixes-2
edgargabriel Jan 5, 2016
335e8bd
Merge pull request #1280 from jsquyres/pr/travis
jsquyres Jan 5, 2016
e4bdad0
pmix: remove extra wrapper LIBS
jsquyres Jan 5, 2016
213b2ab
dpm: correctly handle procs_cutoff in ompi_dpm_connect_accept()
ggouaillardet Jan 6, 2016
4c1ea4a
dpm: correctly handle procs_cutoff in ompi_dpm_connect_accept()
ggouaillardet Jan 6, 2016
713e3ea
configury: fix pthread_join() call in OPAL_INTL_PTHREAD_TRY_LINK_FORTRAN
ggouaillardet Jan 7, 2016
13f9bb9
Use PMI2 constants for consistency
nysal Jan 7, 2016
0a1b735
use the actual preadv and pwritev functions if available. That's what…
edgargabriel Jan 7, 2016
fab1eca
grpcomm: fix bugs in grpcomm algorithms
hjelmn Jan 7, 2016
73ca195
Merge pull request #1288 from hjelmn/grpcomm_fixes
Jan 7, 2016
1384559
Update hwloc to v1.11.2
hjelmn Jan 7, 2016
15007b4
linux: use mntent.h instead of manually parsing /proc/mounts
hjelmn Jan 7, 2016
faeca56
Merge pull request #1289 from hjelmn/hwloc_fix
hjelmn Jan 7, 2016
ac34c0e
Merge pull request #1287 from edgargabriel/posix-fbtl-update
edgargabriel Jan 8, 2016
73daf58
pmix: do not include automatically generated include/private/autogen/…
ggouaillardet Jan 7, 2016
84e4fb3
Fix race condition in UDCM where service thread sees that
artpol84 Jan 6, 2016
0fb7b07
opal/progress: fix non debug builds
ggouaillardet Jan 9, 2016
aefc2ed
Merge pull request #1283 from artpol84/udcm_race_fix
hjelmn Jan 9, 2016
332019b
Silence warning
Jan 10, 2016
d26cc3f
ompi/group: do no decrement parent group proc pointers in destruct
hjelmn Jan 11, 2016
b6366e5
Merge pull request #1294 from hjelmn/group_fix
hjelmn Jan 11, 2016
955fe85
pmix/pmix120: add missing include file
ggouaillardet Jan 12, 2016
f2d4a8d
Replace a bit counting loop with a call to an efficient population co…
Jan 12, 2016
dd4d49c
Merge pull request #1278 from ggouaillardet/poc/osc_pt2pt
hjelmn Jan 12, 2016
3858bc8
Adding support for dynamic endpoint creation
Dec 17, 2015
4c43fb2
orte_rmaps_base_map_job: set OPAL_BIND_ALLOW_OVERLOAD when needed
ggouaillardet Jan 13, 2016
26e882c
Merge pull request #1300 from ggouaillardet/topic/oversubscribe
Jan 13, 2016
eb65b5f
Merge pull request #1297 from timattox/use_hwloc_bitmap_weight
Jan 13, 2016
270cc11
opal hotel: only delete events that have not yet fired
jsquyres Jan 8, 2016
e5cf2db
Merge pull request #1291 from jsquyres/pr/hotel-fix
jsquyres Jan 13, 2016
1d38430
opal: replace opal_convert_jobid_to_string with opal_snprintf_jobid
ggouaillardet Jan 14, 2016
afa62d8
Addressing reviewers' comments for https://github.com/open-mpi/ompi-r…
jladd-mlnx Jan 14, 2016
18c5a21
Fix typo in error handling flow.
jladd-mlnx Jan 14, 2016
958de82
hwloc_base_util.c: Remove newly unused variable 'i'.
Jan 14, 2016
ef24f71
Merge pull request #1303 from timattox/remove_unused_var
Jan 15, 2016
c43d4fd
find_common_syms: trivial updates
jsquyres Jan 16, 2016
96f94f8
fortran: whitelist deliberate common symbols
jsquyres Jan 16, 2016
60ffe71
common syms: whitelist bison-generated common symbols
jsquyres Jan 16, 2016
6c96cb1
find_common_syms: arrgh -- re-add the x bit
jsquyres Jan 16, 2016
0f5fcf9
usnic: fix common symbol
jsquyres Jan 16, 2016
348ac50
usnic: explain why we still have OPAL_HAVE_HWLOC
jsquyres Jan 16, 2016
4dad5de
Silence a couple of warnings - strncpy returns a char*, not an int
Jan 16, 2016
fc6b260
Protect against PMIx-based requests that don't come thru the MPI comm…
Jan 16, 2016
b172b85
Merge pull request #1285 from ggouaillardet/topic/pmix_dist_fix
Jan 17, 2016
d282e94
add the new dynamic_gen2 component, designed to coexist for now with …
edgargabriel Jan 12, 2016
4bbb22b
add a new field to the ompio data structure (stripe_count) and set it…
edgargabriel Jan 12, 2016
2bdd6ba
correctly free some buffers, and ensure that lustre_stripe_size and s…
edgargabriel Jan 14, 2016
2bcae84
further debugging
edgargabriel Jan 14, 2016
39d5c8c
further bug fixes silencing a compiler warning and fixing a memory ov…
edgargabriel Jan 15, 2016
26c57ef
separate the size of the buffer used for the shuffle step and the siz…
edgargabriel Jan 15, 2016
56e11bf
initialize the stripe_size variable as well.
edgargabriel Jan 15, 2016
a9ca370
improve the communicaton abstraction. This commit also allows all agg…
edgargabriel Jan 16, 2016
7d6b75f
orte_util_snprintf_jobid: return ORTE_SUCCESS or ORTE_ERROR
ggouaillardet Jan 18, 2016
c0f8f2c
ompi/dpm: correctly handle sentinels in construct peers
ggouaillardet Jan 18, 2016
bd04192
Merge pull request #1234 from ggouaillardet/poc/travis_gcc5
jsquyres Jan 18, 2016
2adbe27
mpi: have MPI_Wtick() return the period (and not the frequency) if OP…
ggouaillardet Jan 20, 2016
34d861d
orte/oob: Fix issue #1301
igor-ivanov Jan 20, 2016
1671604
Merge pull request #1307 from edgargabriel/fcoll-dynamic_gen2
edgargabriel Jan 20, 2016
9b8d769
will rivist the addproc component later in spring, right now it is co…
edgargabriel Jan 20, 2016
655b4be
find_common_syms: update for OS X symbol naming
jsquyres Jan 20, 2016
f627608
OSHMEM/UCX: implements atomic support
alex-mikheev Jan 21, 2016
b253d4e
fix CID 1349739, CID 1349738, CID 1349736 and (probably) CID 1349740 …
edgargabriel Jan 21, 2016
ae3df29
Add the 1.10.2 NEWS items
Jan 21, 2016
6180386
osc/rdma: disable put aggregation when using threads
hjelmn Jan 21, 2016
b921831
ompi/datatype: make datatype pack thread safe
hjelmn Jan 21, 2016
0fe4818
Merge pull request #1318 from hjelmn/osc_rdma_fixes
hjelmn Jan 22, 2016
243d973
Merge pull request #1316 from hjelmn/datatype_pack_threads
hjelmn Jan 22, 2016
49d2f44
osc/rdma: use correct endpoint for local state
hjelmn Jan 22, 2016
89c7fea
Merge pull request #1315 from alex-mikheev/topic/oshmem_ucx_atomic
mike-dubman Jan 22, 2016
70787d1
Merge pull request #1319 from hjelmn/osc_rdma_fix
hjelmn Jan 22, 2016
e398bf6
Adding entry points for Allgatherv, iAllgatherv, Reduce, and iReduce.
Jan 22, 2016
5e2a2c0
BufFix for coll/hcoll: coll_request must be set to ACTIVE when alloced
Jan 21, 2016
45da311
osc/rdma: fix hang when performing large unaligned gets
hjelmn Jan 23, 2016
ad3aa38
Merge pull request #1322 from jladd-mlnx/topic/BufFix-for-coll-hcoll-…
mike-dubman Jan 25, 2016
500e904
Merge pull request #1320 from hjelmn/osc_rdma_fix
hjelmn Jan 25, 2016
aec3060
Merge pull request #1313 from igor-ivanov/pr/issue-1301
hjelmn Jan 25, 2016
9c93df5
revampt the pvfs2 configure logic
edgargabriel Jan 25, 2016
86765d7
Merge pull request #1325 from edgargabriel/pvfs2-configure-logic-3
edgargabriel Jan 25, 2016
69e3c6f
Merge pull request #1321 from jladd-mlnx/topic/add-allgatherv-reduce
jladd-mlnx Jan 26, 2016
15e26da
pmix configury: add missing PMIX_CHECK_ICC_VARARGS function
ggouaillardet Jan 26, 2016
53185e7
Merge pull request #1326 from ggouaillardet/topic/pmix_check_icc_varargs
Jan 26, 2016
704f14f
f08: do not BIND(C) to subroutines with LOGICAL parameters
ggouaillardet Jan 26, 2016
6d14955
hwloc: have opal_hwloc_base_get_pu search for HWLOC_OBJ_PU when mpiru…
ggouaillardet Jan 26, 2016
722aab9
- extend opal_path_nfs to retrieve the file system type
edgargabriel Jan 26, 2016
b4a725c
need to check for the parent dir as well, since the file might not ex…
edgargabriel Jan 26, 2016
a19c265
osc/rdma: fix typo in ompi_osc_rdma_complete_atomic
hjelmn Jan 26, 2016
e564c69
Merge pull request #1330 from hjelmn/osc_rdma_fix
hjelmn Jan 27, 2016
69ba2a9
ddt: fix support of MPI_COMBINER_RESIZED in __ompi_datatype_create_fr…
ggouaillardet Jan 28, 2016
f5a53b5
pmix: fix Makefile.am to correctly exclude autogenerated file from ta…
ggouaillardet Jan 28, 2016
3f7fff5
Merge pull request #1331 from edgargabriel/solaris-statfs-fix
edgargabriel Jan 29, 2016
d529951
hwloc: correctly count cores with at least one allowed PU
ggouaillardet Jan 29, 2016
4de4a26
usnic: ensure all messages are sent on the data channel
jsquyres Jan 30, 2016
d624e0d
usnic: fix wraparound sequence number issue
jsquyres Jan 30, 2016
db825ab
usnic: don't overrun the fi_av_insert() EQ
jsquyres Jan 30, 2016
797d502
usnic: better av_eq_num default value handling
jsquyres Jan 30, 2016
c2615a4
usnic: change retrans timeout to 5ms
jsquyres Jan 30, 2016
9f3ed00
usnic: minor updates from code review
jsquyres Feb 1, 2016
910eca7
Merge pull request #1327 from ggouaillardet/poc/mpi_xxx_dup_yyy_no_bind
jsquyres Feb 1, 2016
728a97c
use-mpi-f08: remove duplicates from Makefile.am
ggouaillardet Feb 2, 2016
cda094a
mpi_f08: correctly implements MPI_{COMM,TYPE,WIN}_{DUP,NULL_{COPY,DEL…
ggouaillardet Feb 2, 2016
1470420
btl/ugni: fix race condition when adding endpoint to wait list
hjelmn Feb 2, 2016
cd11fc3
btl/ugni: fix race condition that causes completions to be dropped
hjelmn Feb 2, 2016
0715802
verbs_usnic: do not build by default
jsquyres Feb 2, 2016
d7264aa
osc/pt2pt: various threading fixes
hjelmn Feb 2, 2016
519fffb
osc/pt2pt: eager sends are always active if MPI_MODE_NOCHECK is used
hjelmn Feb 2, 2016
a016c17
Merge pull request #1338 from hjelmn/ugni_threading
hjelmn Feb 2, 2016
2cf9b26
verbs_usnic: previous commit missed a symbol
jsquyres Feb 2, 2016
d812695
verbs: fix typo
jsquyres Feb 2, 2016
615b27c
Merge pull request #1339 from hjelmn/osc_pt2pt_fixes
hjelmn Feb 2, 2016
89eea51
usnic: fix calculation for number of blocks
jsquyres Feb 3, 2016
6eac6a8
osc/sm: create datafile into the per proc directory in order to make …
ggouaillardet Feb 3, 2016
68912d0
Fix the grpcomm operations at scale. Restore the direct component to …
Feb 1, 2016
f38ad4a
Merge pull request #1335 from rhc54/topic/gcom
Feb 4, 2016
5b9c82a
osc/pt2pt: bug fixes
hjelmn Feb 4, 2016
a1e784d
Merge pull request #1341 from hjelmn/osc_pt2pt_fixes
hjelmn Feb 5, 2016
5124039
usnic: ensure to init module->av_eq_num
jsquyres Feb 6, 2016
dac2fe1
usnic: ensure to use ntohl() for network-order values
jsquyres Feb 5, 2016
87dbe6c
usnic: add high-verbose reachability messages
jsquyres Feb 5, 2016
8d0a592
usnic: update a few verbose reachability messages
jsquyres Feb 5, 2016
7c574a3
Typo.
bosilca Feb 7, 2016
3fbad2e
Transfer across the -host number of slots
Feb 8, 2016
8558def
opal.pc.in: fix typo; use the write AC_SUBST'ed variable
jsquyres Feb 8, 2016
7c99115
opal/dss: fix comparison of OPAL_VALUE types
ggouaillardet Feb 9, 2016
902b477
ompi_mpi_params.c: set mpi_add_procs_cutoff default to 0
jsquyres Feb 4, 2016
d537ee9
Merge pull request #1340 from jsquyres/pr/decrease-mpi_add_procs_cutoff
jsquyres Feb 9, 2016
7850517
brucks: rename the "brks" component to be "brucks"
jsquyres Feb 9, 2016
030a5f2
sentinel: use type uintptr_t for sentinel
ggouaillardet Feb 6, 2016
b55b9e6
sentinel: fix sentinel to proc_name conversion
ggouaillardet Feb 6, 2016
96310f4
sentinel: fix 32 bits arch
ggouaillardet Feb 8, 2016
39b44d0
Merge pull request #1345 from ggouaillardet/topic/sentinel_proc_name_…
hjelmn Feb 11, 2016
8b05f30
opal/memory: Move Memory Allocation Hooks usage from openib
igor-ivanov Feb 10, 2016
d98616b
Merge pull request #1337 from ggouaillardet/poc/f08_fn
jsquyres Feb 11, 2016
b3d8ead
mtl/ofi: Fix dynamic add_procs.
yburette Feb 12, 2016
67ce4a0
mtl/ofi: FI_AV_MAP support only.
yburette Feb 12, 2016
064a67f
Fix MPI_Get_address (MPI_BOTTOM, ...)
hjelmn Feb 11, 2016
aa9e5a1
Add support for Singularity containers, including a .m4 file for chec…
Feb 11, 2016
7bc62e8
Merge pull request #1356 from hjelmn/get_address
jsquyres Feb 13, 2016
52acd5b
Merge pull request #1354 from rhc54/topic/sing
Feb 13, 2016
233bd08
Protect against a non-privileged port connecting to us when we are ru…
Feb 12, 2016
06c3dfc
Refactor the ORTE DVM code so that external codes can submit multiple…
Dec 16, 2015
1f00a11
Merge pull request #1334 from rhc54/topic/dvm
Feb 13, 2016
59cc1f0
Merge pull request #1357 from rhc54/topic/oob
Feb 13, 2016
68c36ea
Fix two annoying warnings in our UCX support.
bosilca Feb 14, 2016
56425a5
Fix issue identified by Lisandro Dalcin regarding the lack
bosilca Feb 14, 2016
3510706
Correct ordering when checking for privileged ports
Feb 14, 2016
656de55
Merge pull request #1361 from rhc54/topic/oob2
Feb 14, 2016
d9eefef
Merge pull request #1351 from igor-ivanov/pr/issue-1336
igor-ivanov Feb 15, 2016
9dc79f4
Initialize these 2 common symbols.
bosilca Feb 15, 2016
14f0390
Release child object when we are recording someone's relatives.
Feb 15, 2016
7c20d3d
Merge pull request #1365 from rhc54/topic/dvm
Feb 16, 2016
7de01b3
ompi/init: fix abstraction violation
ggouaillardet Feb 16, 2016
22eddfe
mtl/ofi: update copyright dates.
yburette Feb 16, 2016
7fe395c
mtl/ofi: cleanup
yburette Feb 13, 2016
123a39a
btl/openib: fix regression in XRC support
hjelmn Feb 16, 2016
201c280
btl/openib: fix error in param check in mca_btl_openib_put
hjelmn Feb 16, 2016
bf83603
btl/openib/udcm: fix XRC support
hjelmn Feb 16, 2016
1e26f9c
test/dss: update tests to make them usable again, and run them
ggouaillardet Feb 8, 2016
5043100
Modify the IOF subsystem to handle per-job directives for redirecting…
Feb 16, 2016
7a0605f
Merge pull request #1368 from rhc54/topic/iof
Feb 17, 2016
e0de442
Remove debug
Feb 17, 2016
2a728f3
Merge pull request #1367 from hjelmn/xrc_fixes
hjelmn Feb 17, 2016
d544e0e
Merge pull request #1347 from ggouaillardet/topic/dss_tests
jsquyres Feb 17, 2016
31841b4
ras/alps: squelch common symbol warnings
hppritcha Feb 17, 2016
8f9508c
Further enhance the support for Singularity containers. Extend the "p…
Feb 17, 2016
72c7558
Merge pull request #1371 from hppritcha/topic/alps_common_syms
hppritcha Feb 17, 2016
4f4ea96
btl/openib/udcm: fix local XRC connections
hjelmn Feb 17, 2016
3223673
Fix parsing of envvars in MCA files
hjelmn Feb 17, 2016
4a8fbb5
Merge pull request #1373 from hjelmn/xrc_fixes
hjelmn Feb 18, 2016
92a15cc
Merge pull request #1374 from hjelmn/tune_fix
hjelmn Feb 18, 2016
dc4d3ed
Merge pull request #1372 from rhc54/topic/sing
Feb 18, 2016
efb0eff
Plug a leak in the PMIx subsystem
Feb 18, 2016
cc9b131
Make UD OOB memory registrations a multiple of page size
nysal Feb 18, 2016
2745610
Merge pull request #1377 from rhc54/topic/pmix
Feb 18, 2016
c18af0d
Merge pull request #1378 from nysal/issue_1363
nysal Feb 18, 2016
60a7bc2
Enable the PMIx notification callback system. This currently is only …
Jan 11, 2016
142e38c
Merge pull request #1358 from rhc54/topic/notification
Feb 18, 2016
7b73c86
memchecker.h: fix memchecker no-data case
jsquyres Feb 18, 2016
1748f44
Stop a segfault that results in zombied processes by checking for NUL…
Feb 18, 2016
4dc73d7
btl/openib: XRC fix bug that could cause an invalid SRQ# to be used
hjelmn Feb 18, 2016
371df45
btl/openib: fix locking bugs with XRC ib_addr lock
hjelmn Feb 18, 2016
69de442
orterun: allow DDT if options contain :'s
hjelmn Feb 18, 2016
6e68d75
Cleanup some valgrind complaints about jumps with uninitialized value…
Feb 18, 2016
820b178
Merge pull request #1380 from hjelmn/xrc_get_fix
hjelmn Feb 19, 2016
27e7b6e
Merge pull request #1381 from hjelmn/ddt_colon_fix
hjelmn Feb 19, 2016
bfd4254
Merge pull request #1382 from rhc54/topic/cleanup
Feb 19, 2016
2031bb6
btl/openib: XRC save SRQ#s on the loopback endpoint
hjelmn Feb 19, 2016
e57ce1e
Merge pull request #1384 from hjelmn/xrc_get_fix
hjelmn Feb 19, 2016
ad79012
first cut on the version which overlaps the communication/computation…
edgargabriel Feb 3, 2016
268d525
change the tag to be a positive value. handle 0-byte situations corre…
edgargabriel Feb 3, 2016
4f40031
add the dynamic_gen2 component into the fcoll selection table.
edgargabriel Feb 19, 2016
e63836c
clean up the mca parameter handling of the component. Add new paramet…
edgargabriel Feb 19, 2016
92d1b99
optimize the shuffle step:
edgargabriel Feb 19, 2016
b33db51
Merge pull request #1387 from edgargabriel/dynamic_gen2-overlap
edgargabriel Feb 19, 2016
0c72ba8
Cleanup the output-filename options so they work as expected. Have th…
Feb 19, 2016
1f7e2d7
Merge pull request #1388 from rhc54/topic/iof
Feb 19, 2016
8c92a17
Minor memory leak
Feb 19, 2016
3ad5943
mtl/ofi: Prevent possible memory leak.
yburette Feb 20, 2016
404987e
mtl/ofi: Fix mismatching types.
yburette Feb 20, 2016
bd47062
mtl/ofi: Fix error handling.
yburette Feb 20, 2016
59b6d04
mtl/ofi: Check allocated pointer.
yburette Feb 20, 2016
309e23a
Fix minor typo
Feb 20, 2016
d653cf2
Convert the orte_job_data pointer array to a hash table so it doesn't…
Feb 16, 2016
b499d4b
Merge pull request #1391 from rhc54/topic/dvm
Feb 21, 2016
64b7728
Fix typo - do not look at daemon job when considering completion of l…
Feb 21, 2016
77f800b
Tools don't create the orte_job_data table, so don't remove jobs from it
Feb 22, 2016
1df4457
Merge pull request #1392 from rhc54/topic/dvm
Feb 22, 2016
a4aa4c9
ompi_proc_complete_init_single: make the subroutine public
ggouaillardet Feb 19, 2016
308bbcb
ompi/dpm: retrieves OPAL_PMIX_ARCH in heterogeneous mode
ggouaillardet Feb 19, 2016
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
15 changes: 11 additions & 4 deletions .gitignore
Original file line number Diff line number Diff line change
Expand Up @@ -37,6 +37,7 @@
*.exe
*.log
*.trs
*.sapp
*~
*\\#

Expand Down Expand Up @@ -300,10 +301,13 @@ opal/mca/hwloc/hwloc*/hwloc/include/private/autogen/config.h

opal/mca/installdirs/config/install_dirs.h

opal/mca/pmix/pmix1xx/pmix/include/pmix/autogen/config.h
opal/mca/pmix/pmix1xx/pmix/include/pmix/pmix_common.h
opal/mca/pmix/pmix1xx/pmix/include/private/autogen/config.h
opal/mca/pmix/pmix1xx/pmix/include/private/autogen/config.h.in
opal/mca/pmix/pmix112/pmix/include/pmix/autogen/config.h
opal/mca/pmix/pmix112/pmix/include/private/autogen/config.h
opal/mca/pmix/pmix112/pmix/include/private/autogen/config.h.in
opal/mca/pmix/pmix120/pmix/include/pmix/autogen/config.h
opal/mca/pmix/pmix120/pmix/include/private/autogen/config.h
opal/mca/pmix/pmix120/pmix/include/private/autogen/config.h.in


opal/tools/opal-checkpoint/opal-checkpoint
opal/tools/opal-checkpoint/opal-checkpoint.1
Expand Down Expand Up @@ -364,6 +368,7 @@ orte/test/mpi/parallel_w64
orte/test/mpi/pubsub
orte/test/mpi/read_write
orte/test/mpi/reduce-hang
orte/test/mpi/ring
orte/test/mpi/segv
orte/test/mpi/simple_spawn
orte/test/mpi/slave
Expand Down Expand Up @@ -590,6 +595,8 @@ test/event/signal-test
test/event/event-test
test/event/time-test

test/monitoring/monitoring_test

test/mpi/environment/chello

test/runtime/parse_context
Expand Down
81 changes: 81 additions & 0 deletions .travis.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,81 @@
# Use "required" for sudo, because we want to use the "trusty" Debian
# distro, which is (currently) only available in the legacy Travis
# infrastructure (i.e., if we put "sudo: false" to use the new container-
# based Travis infrastructure, then "trusty" is not available). We
# need the "trusty" distro because it has more recent versions of the
# GNU Autotools (i.e., autogen.pl will fail if you use the regular
# distro because the GNU Autotools are too old).
sudo: required
dist: trusty
language: c

# Iterate over 2 different compilers
compiler:
- gcc
- clang

# Iterate over 2 different OSs
os:
- linux
- osx

addons:
# For Linux, make sure we have some extra packages that we like to
# build with
apt:
packages:
- autoconf
- automake
- libtool
- libnl-3-200
- libnl-3-dev
- libnl-route-3-200
- libnl-route-3-dev
- libibverbs-dev
- librdmacm-dev
sources:
- ubuntu-toolchain-r-test

env:
global:
- AM_MAKEFLAGS="-j4"
- CPPFLAGS="-I$HOME/bogus/include"
- LDFLAGS="-I$HOME/bogus/lib"
- LD_LIBRARY_PATH="$HOME/bogus/lib"
matrix:
- GCC_VERSION=default
- GCC_VERSION=5

# Install dependencies for the verbs and usnic providers. Open MPI is
# not currently using the verbs provider in Libfabric, so we might as
# well not build it.
before_install:
- if [[ "GCC_VERSION" == "5" ]]; then COMPILERS="CC=gcc-5 CXX=g++-5 FC=gfortran-5"; fi
- export CONFIGURE_ARGS="--prefix=$HOME/bogus $COMPILERS" DISTCHECK_CONFIGURE_FLAGS="$CONFIGURE_ARGS"
- export DISTCHECK_CONFIGURE_FLAGS="$CONFIGURE_ARGS"
- if [[ "$TRAVIS_OS_NAME" == "linux" ]]; then git clone https://github.com/ofiwg/libfabric.git ; fi
- if [[ "$TRAVIS_OS_NAME" == "linux" ]] && [[ "$GCC_VERSION" == "5" ]] ; then sudo apt-get --assume-yes install gcc-5 g++-5 gfortran-5; fi
- if [[ "$TRAVIS_OS_NAME" == "linux" ]]; then cd libfabric && ./autogen.sh && ./configure --prefix=$HOME/bogus --enable-usnic --disable-verbs $COMPILERS && make install && cd .. ; fi
- if [[ "$TRAVIS_OS_NAME" == "osx" ]] && [[ "$GCC_VERSION" == "5" ]] ; then brew update; brew unlink gcc ; brew install gcc; fi

# Note that we use "make -k" to do the entire build, even if there was a
# build error in there somewhere. This prevents us from needing to submit
# to Travis, see the first error, fix that first error, submit again, ...etc.
install:
- m4 --version
- autoconf --version
- automake --version
- if [[ "$TRAVIS_OS_NAME" == "linux" ]]; then libtool --version; else glibtool --version; fi
- ./autogen.pl
- if [[ "$TRAVIS_OS_NAME" == "linux" && "$CC" == "gcc" ]]; then ./configure $CONFIGURE_ARGS --with-libfabric=$HOME/bogus --with-usnic --with-verbs; else ./configure $CONFIGURE_ARGS; fi
- make -k

# We only need to distcheck on one OS / compiler combination (this is just
# a minor optimization to make the overall set of builds faster).
script:
- if [[ "$TRAVIS_OS_NAME" == "linux" && "$CC" == "gcc" ]]; then make distcheck; else make check; fi

matrix:
exclude:
- env: GCC_VERSION=5
compiler: clang
2 changes: 1 addition & 1 deletion AUTHORS
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,7 @@ [email protected] Ethan Mallove Sun, Oracle
[email protected] Eugene Loh Sun, Oracle
[email protected] Edgar Gabriel HLRS, UH, UTK
[email protected] Graham Fagg UTK
gilles[email protected] Gilles Gouaillardet RIST
gilles@rist.or.jp Gilles Gouaillardet RIST
[email protected] Ginger Young LANL
[email protected] Gleb Natapov Voltaire
[email protected] Galen Shipman LANL
Expand Down
167 changes: 166 additions & 1 deletion NEWS
Original file line number Diff line number Diff line change
Expand Up @@ -67,6 +67,161 @@ Master (not on release branches yet)
Please consider Score-P (score-p.org) as an external replacement.


2.0.0
------

**********************************************************************
* Open MPI is now fully MPI-3.1 compliant
**********************************************************************


- Enhancements to reduce the memory footprint for jobs at scale.
A new MCA parameter - mpi_add_procs_cutoff is available to set
the threshold for using this feature.
- Enhancements to MPI RMA. Open MPI now maps MPI RMA operations on
to native RMA operations for those networks which support this
capability.
- Add support for PMIx - Process Management Interface for Exascale.
Version 1.1.2 of PMIx is included internally in this release.
- Add support for PLFS file systems in Open MPI I/O.
- Add support for UCX transport.
- Improved support for MPI_THREAD_MULTIPLE.
- Update Java interfaces to MPI-3.1
- Simplify build process for Cray XC systems. Add support for
using native SLURM.
- Update the internal copy of ROMIO to that which shipped in MPICH
3.1.4.
- Update internal copy of libevent to v2.0.22.
- Update internal copy of hwloc to v1.11.1.
- Removed checkpoint/restart code due to loss of maintainer. :-(
- Add a new MCA parameter - opal_common_verbs_want_fork_support. This
replaces the btl_openib_want_fork_support parameter.
- Add a -tune command line option to simplify setting many environment
variables and mca parameters.
- Removed --enable-mpi-profiling configure option.
- Add --with-platform-patches-dir configure option.
- Add a new mca parameter default-dash-to-host to offer a mirror of
the default_hostfile.
- Add a --with-pmi-libdir configure option for environments that install
PMI libs in a non-default location.
- Add the ability to specify the number of desired slots in the --host
option.
- Remove support for OS X Leopard.
- Remove support for Cray XT systems.
- Remove VampirTrace.
- Remove support for Myrinet/MX.
- Remove legacy collective module:ML.
- Remove support for Alpha processors.
- Numerous fixes/improvements to orte-dvm. Special thanks to Mark
Santcroos for his help.
- Silence a compiler warning in PSM MTL. Thanks to Adrian Reber for
reporting this.
- Add missing Fortran bindings for MPI_Win_allocate. Thanks to Christoph
Niethammer for reporting and fixing.
- Replace use of alloca with malloc for certain datatype creation
functions. Thanks to Bogdan Sataric for reporting this.
- Fix use of MPI_LB and MPI_UB in creation of of certain MPI datatypes.
Thanks to Gus Correa for helping to fix this.
- Implement a workaround for a libtool problem. Thanks to Eric
Schnetter for reporting and fixing.
- Improve hcoll library detection in configure. Thanks to David
Shrader and Ake Sandgren for reporting this.
- Fix problems with XRC detection in OFED 3.12 and older releases.
Thanks to Paul Hargrove for his analysis of this problem.
- Fix an issue with MCA parameters for Java bindings. Thanks to
Takahiro Kawashima and Siegmar Gross for reporting this issue.
- Fix a performance problem for large messages for Cray XC systems.
Thanks to Jerome Vienne for reporting this.
- Fix an issue with MPI_Win_lock_all. Thanks to Thomas Jahns for
reporting.
- Fix an issue with passing a parameter to configure multiple times.
Thanks to QuesarVII for reporting and supplying a fix.
- Add support for ALPS resource allocation system on Cray CLE 5.2 and
later. Thanks to Mark Santcroos.
- Correction to HACKING file. Thanks to Maximilien Levesque.
- Fix an issue with user supplied reduction operator functions.
Thanks to Rupert Nash for reporting this.
- Fix an issue with an internal list management function. Thanks to
Adrian Reber for reporting this.
- Fix a problem with MPI-RMA PSCW epochs. Thanks to Berk Hess for
reporting this.
- Fix a problem in neighbor collectives. Thanks to Lisandro Dalcin
for reporting.
- Fix a problem when using persistent requests in the Java bindings.
Thanks to Nate Chambers for reporting.
- Add additional flavors to MPI_Comm_split_type. Thanks to Nick
Andersen for supplying this enhancement.
- Improve closing of file descriptors during job launch phase. Thanks
to Piotr Lesnicki for reporting and providing this enhancement.
- Fix a problem in MPI_Get_accumulate and MPI_Rget_accumulate when
using Portals4. Thanks to Nicolas Chevalier for reporting.
- Use correct include file for lstat prototype in ROMIO. Thanks to
William Throwe for finding and providing a fix.
- Fortran related fixes to handle Intel 2016 compiler. Thanks to
Fabrice Roy for reporting this.
- Fix an issue with use of DL-related macros in opal library. Thanks to
Scott Atchley for finding this.
- Fix a Fortran linkage issue. Thanks to Macro Atzeri for finding and
suggesting a fix.
- Fix an issue with parsing mpirun command line options which contain
colons. Thanks to Lev Given for reporting.
- Fix a problem with Open MPI's package config files. Thanks to
Christoph Junghans for reporting.
- Fix a typo in the MPI_Intercomm_merge man page. Thanks To Harald
Servat for reporting and correcting.
- Update man pages for non-blocking sends per MPI 3.1 standard.
Thanks to Alexander Pozdneev for reporting.


1.10.2
------

**********************************************************************
* OSHMEM is now 1.2 compliant
**********************************************************************

- Fix NBC_Copy for legitimate zero-size messages
- Fix multiple bugs in OSHMEM
- Correctly handle mpirun --host <user>@<ip-address>
- Centralize two MCA params to avoid duplication between OMPI and
OSHMEM layers: opal_abort_delay and opal_abort_print_stack
- Add support for Fujitsu compilers
- Add UCX support for OMPI and OSHMEM
- Correctly handle oversubscription when not given directives
to permit it. Thanks to @ammore1 for reporting it
- Fix rpm spec file to not include the /usr directory
- Add Intel HFI1 default parameters for the openib BTL
- Resolve symbol conflicts in the PSM2 library
- Add ability to empty the rgpusm cache when full if requested
- Fix another libtool bug when -L requires a space between it
and the path. Thanks to Eric Schnetter for the patch.
- Add support for OSHMEM v1.2 APIs
- Improve efficiency of oshmem_preconnect_all algorithm
- Fix bug in buffered sends support
- Fix double free in edge case of mpirun. Thanks to @jsharpe for
the patch
- Multiple one-sided support fixes
- Fix integer overflow in the tuned "reduce" collective when
using buffers larger than INT_MAX in size
- Fix parse of user environment variables in mpirun. Thanks to
Stefano Garzarella for the patch
- Performance improvements in PSM2 support
- Fix NBS iBarrier for inter-communicators
- Fix bug in vader BTL during finalize
- Improved configure support for Fortran compilers
- Fix rank_file mapper to support default --slot-set. Thanks
to Matt Thompson for reporting it
- Update MPI_Testsome man page. Thanks to Eric Schnetter for
the suggestion
- Fix missing resize of the returned type for subarray and
darray types. Thanks to Keith Bennett and Dan Garmann for
reporting it
- Fix Java support on OSX 10.11. Thanks to Alexander Daryin
for reporting the problem
- Fix some compilation issues on Solaris 11.2. Thanks to
Paul Hargrove for his continued help in such areas


1.10.1
------

Expand Down Expand Up @@ -118,6 +273,12 @@ Master (not on release branches yet)
to our attention.
- Fix linking issues on some platforms (e.g., SLES 12).
- Fix hang on some corner cases when MPI applications abort.
- Add missing options to mpirun man page. Thanks to Daniel Letai
for bringing this to our attention.
- Add new --with-platform-patches-dir configure option
- Adjust relative selection priorities to ensure that MTL
support is favored over BTL support when both are available
- Use CUDA IPC for all sized messages for performance


1.10.0
Expand All @@ -126,7 +287,11 @@ Master (not on release branches yet)
** version numbering scheme. The v1.10.x release series is based on
** the v1.8.x series, but with a few new features. v2.x will be the
** next series after the v1.10.x series, and complete the transition
** to the new version numbering scheme. See README for more details.
** to the new version numbering scheme. See README for more details
** on the new versioning scheme.
**
** NOTE: In accordance with OMPI version numbering, the v1.10 is *not*
** API compatible with the v1.8 release series.

- Added libfabric support (see README for more details):
- usNIC BTL updated to use libfabric.
Expand Down
14 changes: 12 additions & 2 deletions README
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@ Copyright (c) 2004-2008 High Performance Computing Center Stuttgart,
University of Stuttgart. All rights reserved.
Copyright (c) 2004-2007 The Regents of the University of California.
All rights reserved.
Copyright (c) 2006-2015 Cisco Systems, Inc. All rights reserved.
Copyright (c) 2006-2016 Cisco Systems, Inc. All rights reserved.
Copyright (c) 2006-2011 Mellanox Technologies. All rights reserved.
Copyright (c) 2006-2012 Oracle and/or its affiliates. All rights reserved.
Copyright (c) 2007 Myricom, Inc. All rights reserved.
Expand Down Expand Up @@ -952,6 +952,17 @@ NETWORKING SUPPORT / OPTIONS
which covers most cases. This option is only needed for special
configurations.

--with-verbs-usnic
This option will activate support in Open MPI for disabling a
dire-sounding warning message from libibverbs that Cisco usNIC
devices are not supported (because Cisco usNIC devices are supported
through libfabric, not libibverbs). This libibverbs warning can
also be suppressed by installing the "no op" libusnic_verbs plugin
for libibverbs (see https://github.com/cisco/libusnic_verbs, or
download binaries from cisco.com). This option is disabled by
default because it causes libopen-pal.so to depend on libibverbs.so,
which is undesirable to many downstream packagers.

--with-usnic
Abort configure if Cisco usNIC support cannot be built.

Expand Down Expand Up @@ -1561,7 +1572,6 @@ Here's how we apply those rules specifically to Open MPI:
* libmpi_cxx
* libmpi_java
* liboshmem
* liboshmem_java

===========================================================================

Expand Down
16 changes: 7 additions & 9 deletions VERSION
Original file line number Diff line number Diff line change
Expand Up @@ -91,22 +91,20 @@ libopen_rte_so_version=0:0:0
libopen_pal_so_version=0:0:0
libmpi_java_so_version=0:0:0
liboshmem_so_version=0:0:0
liboshmem_java_so_version=0:0:0

# "Common" components install standalone libraries that are run-time
# linked by one or more components. So they need to be versioned as
# well. Yuck; this somewhat breaks the
# components-don't-affect-the-build-system abstraction.

# OMPI layer
libmca_common_cuda_so_version=0:0:0
libmca_common_ofacm_so_version=0:0:0
libmca_common_sm_so_version=0:0:0
libmca_common_ugni_so_version=0:0:0
libmca_common_verbs_so_version=0:0:0

# ORTE layer
libmca_orte_common_alps_so_version=0:0:0

# OPAL layer
libmca_opal_common_cuda_so_version=0:0:0
libmca_opal_common_libfabric_so_version=0:0:0

# ORTE layer
libmca_common_alps_so_version=0:0:0
libmca_opal_common_sm_so_version=0:0:0
libmca_opal_common_ugni_so_version=0:0:0
libmca_opal_common_verbs_so_version=0:0:0
Loading