1
1
Граф коммитов

29969 Коммитов

Автор SHA1 Сообщение Дата
Tomislav Janjusic
ae30df4bae oshmem/ucx: fixed a build issue
Co-authored with: Artem Polyakov <artemp@mellanox.com>

Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
(cherry picked from commit cb5ff55b27)
2019-12-19 21:20:05 +02:00
Howard Pritchard
f6914ee35c
Merge pull request #7229 from tkordenbrock/topic/v4.0.x/portals4.fix.flowcontrol.bugs
v4.0.x: portals4: fix flow control bugs
2019-12-18 08:35:46 -07:00
George Bosilca
be58cf7982 Fix the communication ordering for all cartesian neighbor collectives.
This work is rooted in the [MPI Forum issue
153](https://github.com/mpi-forum/mpi-issues/issues/153).

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
(cherry picked from commit 86acdee460)
2019-12-17 14:25:22 -08:00
Nathan Hjelm
21221eb70a coll/basic: fix neighbor alltoall message ordering
This commit updates the coll/basic component to correctly order sends
and receives for cartesian communicators with cyclic boundaries. This
addresses an issue identified by mpi-forum/mpi-issues#153. This issue
occurs when the size in any dimension is 1. This gives the same
neighbor in the positive and negative directions. The old code was
sending and receiving in the same order so the -1 buffer contained
the +1 result and vise-versa. The problem is addressed by using
unique tags for each send. This should cover both the case where
overtaking is allowed and is not allowed. The former case will be
possible is a MPI_Cart_create_with_info() call is added to the
standard.

Signed-off-by: Nathan Hjelm <hjelmn@google.com>
(cherry picked from commit 196a91e604)
2019-12-17 14:25:22 -08:00
Howard Pritchard
3f752f1d4f
Merge pull request #7237 from mcoil1/pr/v4.0.x/wbailey2-fixes
v4.0.x/two fixes
2019-12-17 09:22:07 -07:00
Howard Pritchard
c0b27e4f74
Merge pull request #7236 from mcoil1/pr/v4.0.x/several-fixes
v4.0.x/several fixes
2019-12-17 09:19:59 -07:00
Howard Pritchard
af04a9d469
Merge pull request #7215 from janjust/v4.0.x-oshmem-context-fixes
v4.0.x, oshmem:ucx, fix race condition and add context recycling
2019-12-16 13:03:58 -07:00
William Bailey
c01a71fbe9 romio: fix uninitialized variable
Squash compiler warning.

ROMIO is third-party software but has an annoying compiler warning;
this is the minimum distance fix.

Signed-off-by: William Bailey <wbailey2@nd.edu>
(cherry picked from commit 30bda56bce)
2019-12-14 17:18:57 -05:00
William Bailey
bc018dec4a Changed the final URL to https://github.com/westes/flex
Signed-off-by: William Bailey <wbailey2@nd.edu>
(cherry picked from commit caf1d9292c)
2019-12-14 17:17:26 -05:00
Maxwell Coil
07a54b7025 memory/patcher: fix compiler warning
syscall() returns a long, but we are invoking shmat(), which returns
a void*.

Signed-off-by: Maxwell Coil <mcoil@nd.edu>
(cherry picked from commit 52a9cce6f3)
2019-12-14 12:26:47 -05:00
Maxwell Coil
6fdd902d3f romio: Update ADIOI_R_Exchange_data function
Squash compiler warning due to whitespace/brace problems.

The code block from lines 829-839 was improperly indented, which led to
both the code being confusing and a compiler warning. Comparing this code to
the current version in the MPICH repo made it clear that the code was simply
improperly indented. Fixing the indentation both makes the code readable and
squashes the compiler warning.

Signed-off-by: Maxwell Coil <mcoil@nd.edu>
(cherry picked from commit 8c237e2684)
2019-12-14 12:25:51 -05:00
Maxwell Coil
84a67bd6cf libnbc: fixed uninitialized variable
Squash compiler warning.

Signed-off-by: Maxwell Coil <mcoil@nd.edu>
(cherry picked from commit 52241dbbcd)
2019-12-14 12:25:18 -05:00
Maxwell Coil
879a25c239 ompi/dpm/dpm.c: Fix uninititalized variable
Squash compiler warning.

Signed-off-by: Maxwell Coil <mcoil@nd.edu>
(cherry picked from commit 3ced33c2eb)
2019-12-14 12:24:57 -05:00
Maxwell Coil
3964144ca5 Fix misleading error message with missing #! interpreter
This change fixes the misleading error message. I added a conditional to
determine whether the error is due to a missing file or a bad interpreter.
If it is the latter, a new, more precise error message will be displayed.

Fixes #4528

Signed-off-by: Maxwell Coil <mcoil@nd.edu>
(cherry picked from commit 9b73f6ac83)
2019-12-14 12:22:55 -05:00
Howard Pritchard
59d8d62555
Merge pull request #7116 from ggouaillardet/topic/v4.0.x/f08_bind_c_constants_revamp
v4.0.x: fortran/use-mpi-f08: revamp mpi_f08 constants
2019-12-13 08:07:42 -07:00
Todd Kordenbrock
2c082b6c7c btl-portals4: fix a flow control configure bug
This commit fixes a configure bug that caused flow control to be
disabled regardless of the configure options used.

Signed-off-by: Todd Kordenbrock <thkgcode@gmail.com>
(cherry picked from commit f7e74b6a3d)
2019-12-12 08:43:20 -06:00
Todd Kordenbrock
1f5a79bbd4 mtl-portals4: don't finalize flow control if Portals4 was not initialized
This commit fixes a segfault in mtl-portals4 finalize().  The segfault
occurs if finalize() is called without any calls to add_procs().  This
commit resolves the segfault by skipping the flow control fini() call if
Portals4 was not initialized.

Signed-off-by: Todd Kordenbrock <thkgcode@gmail.com>
(cherry picked from commit e7b867c044)
2019-12-12 08:43:20 -06:00
Howard Pritchard
684c180cce
Merge pull request #7224 from jsquyres/pr/v4.0.x/mpool-basic-base-ptr-fix
v4.0.x: mpool/base: fix basic mpool_base() function
2019-12-12 05:38:49 -07:00
Tomislav Janjusic
b84074d4d0 oshmem:ucx, fix race condition and add context recycling
1) Race condition: Do not add private contexts to active list.
Private contexts are only visible to the user.
2) Recycled contexts: Destroyed contexts are put on an idle list until
finalize, continuous context creation will lead to oom condition.
Instead, check if context from idle list meets new context requirements
and reuse it.

Co-authored with: Artem Y. Polyakov <artemp@mellanox.com>,
                  Manjunath Gorentla Venkata <manjunath@mellanox.com>

Signed-off-by: Tomislav Janjusic <tomislavj@mellanox.com>
(cherry picked from commit bd7cdf718488627e7943aab34275c150baf2284a)
2019-12-10 17:52:31 +02:00
Geoff Paulsen
3652cd93c5
Merge pull request #7227 from wbailey2/pr/4.0.x/two_phase_support-fix
v4.0.x: fcoll/two_phase: Compiler warning for wrong variable type used
2019-12-09 08:56:48 -06:00
William Bailey
71fe9d78e0 fcoll/two_phase: Compiler warning for wrong variable type used
Squash compiler warning. Changed output specifier to match variable type (long int -> long long int).

Signed-off-by: William Bailey <wbailey2@nd.edu>
(cherry picked from commit e2718e0196)
2019-12-08 14:15:29 -05:00
Jeff Squyres
91866789f2 mpool/base: fix basic mpool_base() function
The prior implementation was simply wrong.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
(cherry picked from commit 53ebea12aa)
2019-12-06 10:23:56 -08:00
Gilles Gouaillardet
b004f4c391 schizo/ompi: correctly handle the yield_when_idle option
in schizo/ompi, sets the new OMPI_MCA_mpi_oversubscribe environment
variable according to the node oversubscription state.

This MCA parameter is used to set the default value of the
mpi_yield_when_idle parameter.

This two steps tango is needed so the mpi_yield_when_idle setting
is always honored when set in a config file.

Refs. open-mpi/ompi#6433

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
(cherry-picked from cc97c0f611)
2019-12-02 17:18:26 -05:00
Howard Pritchard
d4dd837a3c
Merge pull request #7194 from edgargabriel/pr/two-phase-aggr-calc-32bits-bug-v4.0.x
fcoll/two_phase: fix error in calculating aggregators in 32bit mode
2019-11-27 12:20:34 -07:00
Howard Pritchard
41717c8b73
Merge pull request #7193 from edgargabriel/pr/simple-aggr-mode-fix-v4.0.x
common/ompio: fix calculation in simple-grouping option
2019-11-26 09:26:26 -07:00
Edgar Gabriel
02da54c174 fcoll/two_phase: fix error in calculating aggregators in 32bit mode
In fcoll_two_phase_supprot_fns.c: calculation of the aggregator index
failed for large offsets on 32bit machine, due to improper handling of
64bit offsets.

Fixes Issue #7110

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
(cherry picked from commit ea1355beae)
2019-11-25 09:06:36 -06:00
Edgar Gabriel
39acc3a251 common/ompio: fix calculation in simple-grouping option
This is based on a bug reported on the mailing list using a netcdf testcase.
The problem occurs if processes are using a custom file view, but on some
of them it appears as if the default file view is being used. Because of that,
the simple-grouping option lead to different number of aggregators used on different
processes, and ultimately to a deadlock. This patch fixes the problem by not using
the file_view size anymore for the calculation in the simple-grouping option,
but the contiguous chunk size (which is identical on all processes).

Fixes issue #7109

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
(cherry picked from commit ad5d0df4e9)
2019-11-25 09:04:13 -06:00
Howard Pritchard
7cc5841b9a
Merge pull request #7180 from devreal/btl-ugni-deadlock-v4.0.x
uGNI: Fix potential deadlock when processing outstanding transfers (v4.0.x)
2019-11-20 05:03:52 -07:00
Joseph Schuchart
a346756bf4 uGNI: Fix potential deadlock when processing outstanding transfers
Signed-off-by: Joseph Schuchart <schuchart@hlrs.de>
(cherry picked from commit c09ca039b4)
2019-11-19 22:39:21 +01:00
Howard Pritchard
d57bea0e6d
Merge pull request #7176 from hppritcha/topic/pr_7096_to_v4.0.x
opal_check_alps: fix configure output
2019-11-19 12:35:39 -07:00
Geoff Paulsen
40b396c0e8
Merge pull request #7175 from Akshay-Venkatesh/topic/v4.0.x-openib-cx6-detect
OPAL/MCA/BTL/OPENIB: Detect ConnectX-6 HCAs
2019-11-19 06:45:17 -06:00
Jeff Squyres
87c0178ed4 opal_check_alps: fix configure output
There was a path where OPAL_CHECK_ALPS would exit its testing but
still leave `opal_check_cray_alps_happy` blank.  Fix that by setting
it to "no".

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
(cherry picked from commit 26705efad0)
2019-11-19 05:07:32 -07:00
Akshay Venkatesh
db3e563749 OPAL/MCA/BTL/OPENIB: Detect ConnectX-6 HCAs
Signed-off-by: Akshay Venkatesh <akvenkatesh@nvidia.com>
2019-11-18 17:04:54 -08:00
Geoff Paulsen
a2304ebaae
Merge pull request #7152 from hppritcha/topic/btl_uct_fixes_for_v4.0.x
Topic/btl uct fixes for v4.0.x
2019-11-08 14:25:43 -06:00
Geoff Paulsen
0e25083700
Merge pull request #7145 from devreal/shmem_memheap_alloc_band_v4.0.x
Shmem: use bitwise and instead of logical and to check for allocator capabilities (v4.0.x)
2019-11-08 13:52:41 -06:00
Howard Pritchard
59b24ab4f7 btl/uct: add UCT API version check to configury
related to #7128

The UCX crew is no longer guaranteeing that the UCT API is going to be frozen,
so this is kind of a whack-a-mole problem trying to keep the BTL UCT working
with various changing UCT APIs.

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
(cherry picked from commit 9d345d9aa0)
2019-11-07 10:01:52 -07:00
Nathan Hjelm
55e01220cd btl/uct: fix compilation for UCX 1.7.0
Ref #7128

Signed-off-by: Nathan Hjelm <hjelmn@google.com>
(cherry picked from commit a3026c016a)
2019-11-07 10:00:33 -07:00
Nathan Hjelm
47ec3e4d2b btl/uct: add support for OpenUCX v1.8 API changes
OpenUCX broke the UCT API again in v1.8. This commit updates
btl/uct to fix compilation with current OpenUCX master
(future v1.8). Further changes will likely be needed for
the final release.

Signed-off-by: Nathan Hjelm <hjelmn@google.com>
(cherry picked from commit 526775dfd7)
2019-11-07 10:00:21 -07:00
Joseph Schuchart
ad86d043cf Shmem: use bitwise and instead of logical and to check for allocator capabilities
Signed-off-by: Joseph Schuchart <schuchart@hlrs.de>
(cherry picked from commit 9f2c6a42c3)
2019-11-06 08:46:06 +01:00
Gilles Gouaillardet
02c79ac0c8 fortran/use-mpi-f08: misc fixes
- fix typos from open-mpi/ompi@b10a60a5a9
 - remove remaining references to OMPI_PROTECTED from open-mpi/ompi@df6d763a53

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>

(cherry picked from commit open-mpi/ompi@fda4d040da)
2019-11-06 10:10:22 +09:00
Gilles Gouaillardet
0ab61c9b74 fortran/use-mpi-f08: remove unused references to OMPI_PROTECTED
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>

(back-ported from commit open-mpi/ompi@df6d763a53)
2019-11-06 10:10:22 +09:00
Gilles Gouaillardet
23ed2f44a2 fortran/use-mpi-f08: revamp constant declarations
In order to work around an issue with flang based compilers,
avoid declaring bind(C) constants and use plain Fortran parameter
instead.

For example,
type(MPI_Comm), bind(C, name="ompi_f08_mpi_comm_world") OMPI_PROTECTED :: MPI_COMM_WORLD
is changed to
type(MPI_Comm), parameter :: MPI_COMM_WORLD = MPI_Comm(OMPI_MPI_COMM_WORLD)

Note that in order to preserve ABI compatibility, ompi/mpi/fortran/use-mpi-f08/constants.{c,h}
have been kept even if its symbols are no more referenced by Open MPI.

Refs. open-mpi/ompi#7091

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>

(back-ported from commit open-mpi/ompi@b10a60a5a9)
2019-11-06 10:10:18 +09:00
KAWASHIMA Takahiro
56d2865cf6 fortran/use-mpi-f08: Add C++ datatypes and MPI_NO_OP
Though the MPI standard does not have `MPI_CXX_COMPLEX`, `mpi.h`,
`mpif.h`, and `mpi.mod` have it. So I added it for consistency.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>

(cherry picked from commit open-mpi/ompi@63ecf01610)
2019-11-06 09:49:17 +09:00
KAWASHIMA Takahiro
792b5a01a5 fortran/use-mpi-f08: Remove unnecessary ;
Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>

(cherry picked from commit open-mpi/ompi@e0c5bad195)
2019-11-06 09:48:58 +09:00
Gilles Gouaillardet
628883b38b fortran/use-mpi-f08: add MPI C types
Refers open-mpi/ompi#5801

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>

(cherry picked from commit open-mpi/ompi@69f1a19c5d)
2019-11-06 09:48:11 +09:00
Geoff Paulsen
524960dcdd
Merge pull request #7119 from devreal/grequestx-progress-v4.0.x
Ensure that grequestx continuously make progress (v4.0.x)
2019-11-01 14:12:48 -05:00
Howard Pritchard
608502ff85
Merge pull request #7102 from edgargabriel/pr/v4.0.x-romio321-status-set-elements-fix
MPIR_Status_set_bytes: fix for large count sizes
2019-11-01 13:08:35 -06:00
Geoff Paulsen
dbb873f46f
Merge pull request #7047 from edgargabriel/pr/v4.0.x-hdf5-2gb-bug
comomn_ompio_file_read/write: fix 2GB limiting issue
2019-11-01 14:06:46 -05:00
Geoff Paulsen
55527abbc7
Merge pull request #7125 from sam6258/smiller_rsh_chdir_v4.0.x
plm/rsh: Add chdir option to change directory before orted exec
2019-10-29 15:59:09 -05:00
Scott Miller
8eae54fd27 plm/rsh: Add chdir option to change directory before orted exec
Signed-off-by: Scott Miller <scott.miller1@ibm.com>
(cherry picked from commit c1b8599528)

Conflicts:
	orte/mca/plm/rsh/plm_rsh_module.c
2019-10-29 15:49:41 -04:00