ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
30 913 коммитов 3 Ветки 0 Теги
Граф коммитов

30913 Коммитов

This Branch
This Branch
Все ветки
Автор SHA1 Сообщение Дата
Howard Pritchard
d6ac41cbbd OFI: patch OFI MTL for GNI provider 
Uncovered a problem using the GNI provider with the OFI MTL.
See https://github.com/ofiwg/libfabric/issues/6194.

Related to #8001

Signed-off-by: Howard Pritchard <hppritcha@gmail.com>
 2020-08-26 08:25:53 -06:00
Artem Polyakov
dfb0ae748f
Merge pull request #7681 from janjust/master-tls-refactor_v3 
ompi/osc/ucx: remove global TLS tables
 2020-07-31 10:39:53 -07:00
Tommy Janjusic
2c8da2c0a9 Further code reduction and simplifications. 
Co-authored-by: Artem Polyakov <artpol84@gmail.com>

Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
 2020-07-30 20:00:22 +03:00
Tomislav Janjusic
cbfc9a3263 opal/mca/common/ucx: Use new TSD api 
Co-authored-by: Artem Y. Polyakov <artemp@mellanox.com>

Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
 2020-07-30 00:21:26 +03:00
Tomislav Janjusic
72296e12f4 opal/common/ucx: 
-mutex lock/unlock suggestions
-common destructor/cleanup

Co-authored-with: Artem Y. Polyakov <artemp@mellanox.com>

Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
 2020-07-30 00:21:26 +03:00
Tomislav Janjusic
27ba4b612f ompi/osc/ucx: Remove workerpool's global thread storage tables. 
Co-authored-by: Artem Y. Polyakov <artemp@mellanox.com>

Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
 2020-07-30 00:21:26 +03:00
Brian Barrett
41df122083
Merge pull request #7730 from wckzhang/newdefaults 
coll/tuned: Change the default collective algorithm selection
 2020-07-28 15:27:46 -07:00
William Zhang
ce40cfbaa5 coll/tuned: Change the default collective algorithm selection 
The default algorithm selections were out of date and not performing
well. After gathering data from OMPI developers, new default algorithm
decisions were selected for:

    allgather
    allgatherv
    allreduce
    alltoall
    alltoallv
    barrier
    bcast
    gather
    reduce
    reduce_scatter_block
    reduce_scatter
    scatter

These results were gathered using the ompi-collectives-tuning package
and then averaged amongst the results gathered from multiple OMPI
developers on their clusters.

You can access the graphs and averaged data here:
https://drive.google.com/drive/folders/1MV5E9gN-5tootoWoh62aoXmN0jiWiqh3

Signed-off-by: William Zhang <wilzhang@amazon.com>
 2020-07-28 10:41:48 -07:00
Austen Lauria
d0152eb51e
Merge pull request #7940 from awlauria/revert_libevent_commit 
Revert "Address a race condition in libevent select."
 2020-07-28 11:34:59 -04:00
Jeff Squyres
c07d77fbf2
Merge pull request #7957 from bosilca/fix/avx_alignment 
Use the unaligned SSE memory access primitive.
 2020-07-27 15:50:40 -04:00
Artem Polyakov
e5ef80fe8c
Merge pull request #7936 from janjust/master-new-tsd-thread-api 
Master: new thread-specific-data (tsd) api
 2020-07-24 14:58:03 -07:00
Ralph Castain
863a058f8d
Merge pull request #7964 from rhc54/topic/sync 
Sync to PRRTE master
 2020-07-24 14:57:32 -07:00
Ralph Castain
8c0269cd4f
Sync to PRRTE master 
Pickup the FT and libev cleanups

Signed-off-by: Ralph Castain <rhc@pmix.org>
 2020-07-24 14:11:34 -07:00
Tomislav Janjusic
d809f6ba27 New TSD API interface fix for various components 
Co-authored by: Artem Polykaov <artemp@mellanox.com>

Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
 2020-07-24 18:29:40 +03:00
Tomislav Janjusic
cba5a0e117 Rename tsd interface function calls 
Co-authored by: Artem Polykaov <artemp@mellanox.com>
Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
 2020-07-24 18:29:07 +03:00
Tomislav Janjusic
cb1955bb53 Fix renamed interface functions for argo, q, and pthreads 
Co-authored by: Artem Polykaov <artemp@mellanox.com>

Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
 2020-07-24 18:29:07 +03:00
Tomislav Janjusic
07dc86eb3a opal/thread: New TSD API 
Co-authored-by: Artem Polyakov <artemp@mellanox.com>
Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
 2020-07-24 18:29:07 +03:00
Ralph Castain
06c585c316
Merge pull request #7962 from rhc54/topic/sync 
Sync to PMIx and PRRTE master
 2020-07-23 16:22:32 -07:00
Ralph Castain
c0bc89dc50
Sync to PMIx and PRRTE master 
Signed-off-by: Ralph Castain <rhc@pmix.org>
 2020-07-23 12:35:17 -07:00
George Bosilca
c4e88a43a3
Check unaligned ops for correctness. 
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
 2020-07-22 11:26:07 -04:00
Joshua Ladd
366e92ce54
Merge pull request #7860 from vspetrov/hcoll_reduce_scatter 
Coll/Hcoll: reduce_scatter(block) interface
 2020-07-22 09:45:34 -04:00
George Bosilca
b6d71aa893
Use the unaligned SSE memory access primitive. 
Alter the test to validate misaligned data.

Fixes #7954.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
 2020-07-22 01:19:12 -04:00
Jeff Squyres
30ba603c2c
Merge pull request #7953 from cniethammer/configure-leak-fix 
Fix memory leak in configure, which prevents leak sanitizer usage
 2020-07-21 16:34:27 -04:00
Christoph Niethammer
6564c1b942 Fix memory leak in configure, which prevents leak sanitizer usage 
If building Open MPI with sanitizers, e.g
$ configure CC=clang CFLAGS=-fsanitize=address ....
configure test programs are also build with the sanitizers and will
report errors resulting in configure to fail.

Signed-off-by: Christoph Niethammer <niethammer@hlrs.de>
 2020-07-21 21:28:29 +02:00
Aurelien Bouteiller
816acbdfb1
Merge pull request #7840 from abouteiller/mpi-next/init-errh 
MPI-4: Initial error handler
 2020-07-21 11:55:14 -04:00
Joseph Schuchart
60aa97b301
Merge pull request #7948 from devreal/osc-rdma-check-endpoints 
osc/rdma: fail query_btls if no endpoint for non-local peer is found
 2020-07-20 15:14:25 +02:00
bosilca
1139d9ecae
Merge pull request #7931 from bosilca/fix/7928 
Fix the BTL API conversion for the SMCUDA BTL
 2020-07-18 17:35:39 -04:00
Joseph Schuchart
eebc451ec8 osc/rdma: fail query_btls if no endpoint for non-local peer is found 
Signed-off-by: Joseph Schuchart <schuchart@hlrs.de>
 2020-07-16 17:06:35 +02:00
Aurelien Bouteiller
7118755ae8
Add a tester for the initial error handler 
Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
 2020-07-16 03:10:32 -04:00
Aurelien Bouteiller
5f1f7fe313
route errors to self/initial error handler depending upon the state of 
MPI initialization

Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
 2020-07-16 03:10:32 -04:00
Aurélien Bouteiller
bed909c3ba
Read the info key mpi_initial_errhandler from spawn/spawn_multiple 
Signed-off-by: Aurélien Bouteiller <bouteill@icl.utk.edu>

Use the same env to transmit the initial error handler to spawnees

Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
 2020-07-16 03:10:32 -04:00
Aurélien Bouteiller
83d0f92152
Set the initial error handler onto predefined communicators 
Signed-off-by: Aurélien Bouteiller <bouteill@icl.utk.edu>

update to the predefined initial error handler selection

Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
 2020-07-16 03:10:32 -04:00
Aurélien Bouteiller
3cd85a9ec5
Add the initial_errhandler info key to MPI_INFO_ENV and populate the 
value from prun populated paremeters

Signed-off-by: Aurélien Bouteiller <bouteill@icl.utk.edu>

Allow errhandlers to invoke the initial error handler before MPI_INIT

Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>

Indentation

Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
 2020-07-16 03:10:32 -04:00
Aurélien Bouteiller
703b8c356f
Make error_class and error_string callable before/after 
MPI_INIT/FINALIZE

Signed-off-by: Aurélien Bouteiller <bouteill@icl.utk.edu>

make lazy initialization opal unlikely

Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
 2020-07-16 03:10:32 -04:00
Ralph Castain
7702dfcdd2
Merge pull request #7942 from rhc54/topic/init 
Ensure we init and protect values
 2020-07-15 07:59:01 -07:00
George Bosilca
8bc1f3d8fb Don't allow any asynchronous CUDA operations. 
There are 2 reasons for this:
- pending CUDA events are not progressed by this BTL, so anything that becomes
  asychronous will never be completed.
- we use the packed data on the shared memory backing file, and this will be
  returned to the peer process upon return (thus if we copy asynchronously we
  might not copy the right data).

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
 2020-07-15 01:37:09 -04:00
George Bosilca
0e32b0acef Avoid a lock if no CUDA IPC operations are pending. 
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
 2020-07-15 01:35:34 -04:00
Ralph Castain
a574addce9
Ensure we init and protect values 
Scrub the entire ompi_rte.c file to initialize and protect values
received from PMIx.

Signed-off-by: Ralph Castain <rhc@pmix.org>
 2020-07-14 15:25:14 -07:00
Austen Lauria
67d90166cf Revert "Address a race condition in libevent select." 
We do not want to be patching upstream components anymore.
The proper method is to get this merged upstream, then
pull it in the next upstream release.

This reverts commit c39fb5758a772c062e20db9b42f2b06805884802.

Signed-off-by: Austen Lauria <awlauria@us.ibm.com>
 2020-07-14 16:23:21 -04:00
George Bosilca
fd4ca394e2 Make the smcuda BTL great again. 
It has been broken for months because of the lack of initialization of the
HWLOC library. The smcuda process creating the backing file (local rank 0)
uses opal_cache_line_size to align the objects in the backing file, and the
opal_cache_line_size is initialized by default to 128. Later on, when the rest
of the processes attach the same backing file, HWLOC has been called and the
cache size has now been updated to the correct value. If this value is
different than the default one (and they are as most cache sizes are 64 bytes
right now) the objects in the backing file will be misaligned.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
 2020-07-14 01:48:08 -04:00
George Bosilca
96e8cbe25f First step on fixing the BTL API conversion for the SMCUDA BTL 
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
 2020-07-13 14:46:10 -04:00
Joshua Ladd
aa8f7f4ede
Merge pull request #7893 from bureddy/cuda-ucx 
UCX: initialize cuda from ucx pml component
 2020-07-13 14:18:48 -04:00
bosilca
1f237f5fc9
Merge pull request #7419 from bosilca/topic/avx512 
Add support for AVX512/AVX2/SSE/MMX
 2020-07-13 11:56:50 -04:00
Devendar Bureddy
2547e24c55 UCX: initialize cuda from ucx pml component 
Signed-off-by: Devendar Bureddy <devendar@mellanox.com>
 2020-07-12 18:41:40 +03:00
Nathan Hjelm
d0c0cb7144
Merge pull request #7913 from hjelmn/btl_base_atomics_are_awesome 
btl: change argument type of BTL receive callbacks
 2020-07-11 12:13:26 -06:00
dongzhong
14b3c70628
Add supports for MPI_OP using AVX512, AVX2 and MMX 
Add logic to handle different architectural capabilities
Detect the compiler flags necessary to build specialized
versions of the MPI_OP. Once the different flavors (AVX512,
AVX2, AVX) are built, detect at runtime which is the best
match with the current processor capabilities.

Add validation checks for loadu 256 and 512 bits.
Add validation tests for MPI_Op.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
Signed-off-by: dongzhong <zhongdong0321@hotmail.com>
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
 2020-07-10 21:25:35 -04:00
Howard Pritchard
677b662295
Merge pull request #7912 from tomhers/fix_opal_ofi_compile_bug 
BTL/OFI: Fix missing include file.
 2020-07-10 14:43:38 -06:00
Jeff Squyres
b62545ef4c
Merge pull request #7921 from jsquyres/pr/fix-fortran8integer-vs-c4int-issue 
mpi_f08: fix Fortran-8-byte-INTEGER vs. C-4-byte-int issue
 2020-07-10 09:15:55 -04:00
Jeff Squyres
62d1f89622
Merge pull request #7917 from jmciver/fix/rpmbuild-rpm-build-parameter 
Fix buildrpm.sh "-r" option used for RPM options specification
 2020-07-09 20:58:07 -04:00
Brian Barrett
4a6e1629e8
Merge pull request #7906 from dancejic/multi 
common/ofi: added address format check to fix provider selection
 2020-07-09 16:59:02 -07:00