1
1
Граф коммитов

10183 Коммитов

Автор SHA1 Сообщение Дата
Edgar Gabriel
19b71e4eb6 ompio/fs: add summary of supported file systems
Add the list of supported file systems to the summary output
add the end of configure

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-04-25 14:39:18 -05:00
Jeff Squyres
5bd02f6649 fortran/tkr: Fix Makefile.am warnings
Set mpi.lo dependencies outside of AM conditionals.  Per
https://github.com/open-mpi/ompi/issues/5085, make mpi.lo depend on
whatever we decide the sizeof source files are (which may be empty, if
this compiler does not support the Right Stuff for MPI_SIZEOF, or it
may be mpi-tkr-sizeof.[h|f90]).

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-04-25 10:24:01 -07:00
Gilles Gouaillardet
2abeada060 fortran: build MPI_Sizeof() interface in use-mpi-tkr bindings
whenever possible.

Add the missing OMPI_FORTRAN_BUILD_SIZEOF macro to Fortran
and add a missing dependency.

Thanks Themos Tsikas for reporting this issue.

Refs open-mpi/ompi#5085

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-04-25 15:09:33 +09:00
Xin Zhao
53bdfd1dcb OMPI/OSC/UCX: fix issue in impl of MPI_Win_create_dynamic/MPI_Win_attach/MPI_Win_detach
Signed-off-by: Xin Zhao <xinz@mellanox.com>
2018-04-24 23:09:52 +03:00
Mikhail Kurnosov
787ec8929b Rename function rounddown into ompi_rounddown
FreeBSD 11 `/sys/param.h` has declaration of `rounddown`

Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2018-04-24 07:59:49 +07:00
Mikhail Kurnosov
4cbcff7fcd coll/base: add recursive doubling algorithm for MPI_Reduce_scatter_block
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2018-04-23 11:02:31 +07:00
raafatfeki
91e028f7fd fcoll/dynamic_gen2: Reduce number of realloc calls
keep track of the sizeof the blocklen_per_process and displs_per_process
on the aggregator datastructure to minimze the number of realloc function
calls required in the shuffle_init operation.

Signed-off-by: raafatfeki <fekiraafat@gmail.com>
2018-04-20 10:13:57 -05:00
Jeff Squyres
8251139bac status_set_cancelled: fix F08 binding
The F08 bindings for MPI_STATUS_SET_CANCELLED incorrectly had the
"flag" dummy parameter set as INTENT(OUT) when it really should be
INTENT(IN).

On the one hand, this is technically an ABI change.  On the other
hand, this is an incorrect MPI binding.  On the other hand (that's 3
hands for you fans counting at home), this is such a rarely-used API
-- even in the C bindings -- that I'm guessing no one is using this
API, and therefore no one has noticed this error and it isn't worth
porting back to the release branches.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-04-16 17:33:22 -07:00
Nathan Hjelm
84765001aa io/romio: do not use removed functions
This commit attempts to update the romio io component to not use
functions removed in MPI-3.0 (2012). This is a first cut and will
probably need to be reviewed for correctness.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-04-16 12:06:52 -06:00
Nathan Hjelm
4d876ec6fe io/romio314: fix minmax datatypes
romio assumes that all predefined datatypes are contiguous. Because of
the (terribly named) composed datatypes MPI_SHORT_INT, MPI_DOUBLE_INT,
MPI_LONG_INT, etc this is an incorrect assumption. The simplest way to
fix this is to override the MPI_Type_get_envelope and
MPI_Type_get_contents calls with calls that will work on these
datatypes. Note that not all calls to these MPI functions are
replaced, only the ones used when flattening a non-contiguous
datatype.

References #5009

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-04-16 10:46:38 -06:00
George Bosilca
6ff11267fb
Remove warnings identified by clang.
Plus minor spacing and indentation issues.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2018-04-14 17:14:12 -04:00
George Bosilca
cd683e3eec
Allow OPAL DDT to receive size_t count argument.
Fixes issue #5069, which relates a BigMPI bug with the use of
MPI_Type_vectpor to construct very large datatypes (>2GB).

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2018-04-14 15:32:19 -04:00
Todd Kordenbrock
d646a00cd9
Merge pull request #5054 from tkordenbrock/topic/master/mtl-portals4.finalize.fix
master: mtl-portals4: don't call progress() in finalize() if Portals4 was not initialized
2018-04-12 12:12:05 -05:00
Todd Kordenbrock
90659671bc mtl-portals4: don't call progress() in finalize() 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 progress() loop in
finalize() if the Portals was not initialized.

Signed-off-by: Todd Kordenbrock (thkgcode@gmail.com)
2018-04-10 14:22:32 -05:00
Mikhail Kurnosov
82a3a5bdb5 Fix dynamic decision for Scan and bug in Allreduce
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2018-04-06 11:03:17 +07:00
KAWASHIMA Takahiro
5e12e0f2e5
Merge pull request #5001 from ggouaillardet/topic/javah
configury: use javac vs javah whenever possible.
2018-04-06 11:30:20 +09:00
Jeff Squyres
fc8ebbb0e0 MPI_Comm_spawn_multiple.3in: update Fortran string array notes
Per 0ab6b201fe, note in the MPI_Comm_spawn_multiple.3in man page that
the array_of_commands does not need to be terminated -- it just need
to have exactly "count" entries.  In the Fortran binding, at least,
this is different than in prior released versions of Open MPI (it's
not a backwards incompatibility, since prior versions of Open MPI
required array_of_commands to be blank-string-terminated in Fortran --
this change makes Open MPI be *less* restrictive, and therefore still
backwards compatible).

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-04-05 06:52:46 -07:00
Jeff Squyres
0ab6b201fe mpi/fortran: fix parsing arrays of Fortran strings
MPI defines the "argv" param to Fortran MPI_COMM_SPAWN as being
terminated by a blank string.  While not precisely defined (except
through a non-binding example, Example 10.2, MPI-3.1 p382:6-29), one
can infer that the "array_of_argv" param to Fortran
MPI_COMM_SPAWN_MULTIPLE is also a set of argv, each of which are
terminated by a blank line.

The "array_of_commands" argument to Fortran MPI_COMM_SPAWN_MULTIPLE is
a little less well-defined.  It is *assumed* to be of length "count"
(another parameter to MPI_COMM_SPAWN_MULTIPLE) -- and *not* be
terminated by a blank string.  This is also given credence by the same
example 10.2 in MPI-3.1.

The previous code assumed that "array_of_commands" should also be
terminated by a blank line -- but per the above, this is incorrect.
Instead, we should just parse our "count" number of strings from
"array_of_commands" and *not* look for a blank line termination.

This commit separates these two cases:

* ompi_fortran_argv_blank_f2c(): parse a Fortran array of strings out
  and stop when reaching a blank string.
* ompi_fortran_argv_count_f2c(): parse a Fortran array of strings out
  and stop when "count" number of strings have been parsed.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-04-04 18:56:44 -07:00
Gilles Gouaillardet
5370586d98 configury: use javac vs javah whenever possible
javah is no more available from Java 10, so try
javac -h first (available since Java 8) and fallback on javah

Refs. open-mpi/ompi#5000

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-04-05 10:37:35 +09:00
Gilles Gouaillardet
132ea1a6b0 java: cleanup the list of automatically generated header files
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-04-05 09:27:34 +09:00
Edgar Gabriel
ef28d941d9
Merge pull request #5002 from raafatfeki/pr/coverty-dynamic_gen2-fixes
fcoll/dynamic_gen2: fix coverty warnings
2018-04-04 09:08:42 -05:00
Gilles Gouaillardet
e85fa469f3 coll/tuned: add recursive doubling algo for [ex]scan
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-04-04 14:56:23 +09:00
Gilles Gouaillardet
393376bbd9 coll/basic: move [ex]scan from coll/basic to coll/base
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-04-04 13:41:01 +09:00
Gilles Gouaillardet
65fa0b59c3 coll/tuned: add Rabenseifner algo for [all]reduce
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-04-04 13:25:41 +09:00
Mikhail Kurnosov
177c6ce51f Move algorithms from coll/spacc to coll/base and remove coll/spacc
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2018-04-04 10:21:06 +07:00
raafatfeki
5d99af29cd fcoll/dynamic_gen2: Formatting fixes
Adjust Coding Style to match the 4 space tab rule.
Signed-off-by: raafatfeki <fekiraafat@gmail.com>
2018-04-02 17:25:00 -05:00
raafatfeki
92822613ea fcoll/dynamic_gen2: fix coverty warnings
fix warnings for coverty CID 1433655 and CID 1433654

Signed-off-by: raafatfeki <fekiraafat@gmail.com>
2018-04-02 16:18:07 -05:00
Mikhail Kurnosov
1d2d43bdf0 Fix compile error with dtype
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2018-04-01 08:27:34 +07:00
Edgar Gabriel
c4879ec29f io/ompio: don't reset amode if MODE_SEQUENTIAL is set
the ompio module resets the amode from WRONLY to RDWR in order
to accoomodate data sieving in the two-phase fcoll componet. This
leads however to an error if MPI_MODE_SEQUENTIAL has been requested
by the user, since MODE_SEQUENTIAL is incompatible with MODE_RDWR.
SInce the change to the amode was done after opening the file for
individual file pointers but before opening the file for shared filepointers,
this lead to an error message in the sharedfp component.

Note, that data sieving is never necessary if MODE_SEQUENTIAL is set,
so this should not be a problem for any scenario.

Fixes #4991

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-03-30 07:56:47 -05:00
Mikhail Kurnosov
50ec214d42 Add recursive doubling algorithm for MPI_Scan and MPI_Exscan to coll/base
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2018-03-30 10:12:51 +07:00
raafatfeki
100677721d fcoll/dynamic_gen2: use hindexed constructor on the sender side
instead of using a temporary buffer and copy data into the temp buffer before sending, use a derived datatype to describe the data that needs to be sent during a cycle in the collective I/O operation.

Signed-off-by: raafatfeki <fekiraafat@gmail.com>
2018-03-28 14:37:30 -05:00
Mikhail Kurnosov
bd12e2b1c6 Add recursive doubling algorithm for Scan and Exscan
Implements recursive doubling algorithm for MPI_Scan and MPI_Exscan.
The algorithm preserves order of operations so it can be used both
by commutative and non-commutative operations.

Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2018-03-28 16:27:11 +07:00
Nathan Hjelm
e79debc320 osc/rdma: fix overflow in offset calculation
This commit fixes a bug is osc/rdma that can occur if the total size
of the shared memory segment gets larger than 4 GiB. The bug was
caused by a typo. The type of my_base_offset should have been size_t
not int.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-03-27 09:33:44 -06:00
Nathan Hjelm
f7faacca4e osc/rdma: fix 32-bit builds
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-03-27 09:16:04 -06:00
Jeff Squyres
06af6f1c4c
Merge pull request #4962 from jsquyres/pr/cid-fixes
A bunch of CID fixes
2018-03-26 22:30:31 -04:00
Ralph Castain
f92acd735b
Merge pull request #4965 from rhc54/topic/rank
Fix breakage in ranking system and silence OSC/RDMA warnings
2018-03-26 19:10:36 -05:00
Jeff Squyres
5360035995 topo/treematch: fix CID 1416327
Ensure to free things in the right order so that we don't access
memory after it is freed.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-26 14:26:17 -07:00
Jeff Squyres
08ceb66a19 osc/pt2pt: fix (effectively false positive) CID 1402113
This will almost certainly never happen, but be defensive and
guarantee that we never return an uninitialized variable.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-26 14:26:17 -07:00
Jeff Squyres
9de750a280 io/ompio: fix CID 1269889
Free some memory upon error conditions.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-26 14:21:21 -07:00
Jeff Squyres
dca66b9775 comm_join: fix CID 1323170
Enusre that the port name is always NULL-terminated.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-26 14:21:21 -07:00
Jeff Squyres
6319292170 fcoll/static: fix CID 1413066
local_iov_array is unconditionally allocated, so unconditionally
de-allocate it, too.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-26 14:21:21 -07:00
Jeff Squyres
2968ffa296 fcoll/static: remove useless/dead code
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-26 14:21:21 -07:00
Jeff Squyres
8e925b4f17 fbtl/posix: fix CID 1419954
Ensure to initialized ret_code.  This problem will likely never occur
in practice, but we might as well be defensive about it.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-26 14:21:21 -07:00
Jeff Squyres
124208198c osc/rdma: fix CID 1424327
Fix minor memory leak.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-26 14:21:21 -07:00
Nathan Hjelm
1c75aa82fc use-mpi-f08: fix rma function signatures
The various RMA functions need to have the asynchronous property on
all buffers. This property was missing and some buffers were
incorrectly marked as intent(in). This commit fixes the function
signatures.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-03-26 15:11:07 -06:00
Ralph Castain
3a93b535ec Silence the flood of OSC/RDMA warnings
Fixes #4950

Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2018-03-25 16:12:41 -07:00
Artem Polyakov
77ff99e9ee
Merge pull request #4933 from karasevb/timings_update
timings: added new timing points
2018-03-25 00:10:49 -07:00
Jeff Squyres
871e5c76bc
Merge pull request #4960 from jsquyres/pr/warnings-fixes
Coverity fix + compiler warning fixes
2018-03-23 14:47:56 -05:00
Jeff Squyres
c3adcb05eb Miscellaneous compiler warnings fixes
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-03-23 11:45:30 -07:00
Nathan Hjelm
5f7ff5307e fcoll/two_phase: do not use removed function (MPI_Address)
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-03-23 08:43:24 -06:00
Edgar Gabriel
36747cca67 io/ompio: disable the fcoll timing by default
somehow the flag indicating to gather performance data
on collective io operations has changed to 1 accidentally.
Should be 0 ( false) by default.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-03-21 11:34:35 -05:00
Edgar Gabriel
aae8c6c6ad remove addproc sharedfp component
never got to move this sharedfp component into anything
usable. Can easily be restored if necessary.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-03-21 11:27:01 -05:00
Edgar Gabriel
e703ac2da8 remove plfs components
plfs components are at this point not utilized by anybody as far as I know.
Easy to bring back if we want to.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-03-21 11:27:01 -05:00
Boris Karasev
3796307a57 timings: added new timing points
Signed-off-by: Boris Karasev <karasev.b@gmail.com>
2018-03-21 05:16:25 +02:00
bosilca
bf3dd8af19
Merge pull request #4884 from bosilca/topic/fix_wtime
Improve the range and accuracy of MPI_Wtime.
2018-03-16 14:09:33 +09:00
Nathan Hjelm
7f4872d483 osc/rdma: performance improvments and bug fixes
This commit is a large update to the osc/rdma component. Included in
this commit:

 - Add support for using hardware atomics for fetch-and-op and single
   count accumulate  when using the accumulate lock. This will improve
   the performance of these operations even when not setting the
   single intrinsic info key.

 - Rework how large accumulates are done. They now block on the get
   operation to fix some bugs discovered by an IBM one-sided test. I
   may roll back some of the changes if the underlying bug in the
   original design is discovered. There appear to be no real
   difference (on the hardware this was tested with) in performance so
   its probably a non-issue. References #2530.

 - Add support for an additional lock-all algorithm: on-demand. The
   on-demand algorithm will attempt to acquire the peer lock when
   starting an RMA operation. The lock algorithm default has not
   changed. The algorithm can be selected by setting the
   osc_rdma_locking_mode MCA variable. The valid values are two_level
   and on_demand.

 - Make use of the btl_flush function if available. This can improve
   performance with some btls.

 - When using btl_flush do not keep track of the number of put
   operations. This reduces the number of atomic operations in the
   critical path.

 - Make the window buffers more friendly to multi-threaded
   applications. This was done by dropping support for multiple
   buffers per MPI window. I intend to re-add that support once the
   underlying performance bug under the old buffering scheme is
   fixed.

 - Fix a bug in request completion in the accumulate, get, and put
   paths. This also helps with #2530.

 - General code cleanup and fixes.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-03-15 14:53:53 -06:00
Benoît Legat
00600c7cbb Fix typo in MPI_Cart_shift doc
Signed-off-by: Benoît Legat <benoit.legat@gmail.com>
2018-03-13 15:25:42 +01:00
Edgar Gabriel
da640f98df fcoll/two_phase: data sieving has to occur at offset 0 as well
data sieving has to occur for any offset provided that is larger
or equal zero for this implementation to work correctly.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-03-10 11:23:09 -06:00
George Bosilca
0f0c27a184
Allow MPI_PROC_NULL as neighbor.
Allowing MPI_PROC_NULL as a neighbor in any topology allows us to add
gaps on the send and recv buffers. This does make the traditional
neighbor collective have a similar behavior as the V version, but in
same time it allows the users to skip the step where they prepare the
counts and the displacement array.

For more info please take a look at issue #4675.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2018-03-09 12:20:26 +09:00
Edgar Gabriel
c83b47c266 io/romio314: mark datatypes of size 0 as contiguous
this commit fixes an issue observed with romio314 and the hdf5 1.10.x testsuite.
The ADIOI_Datatype_iscontig() routine in romio314/src/io_romio314_module.c
will now return for a datatype of size 0 that it is contiguous, even if the extent
of the datatype is non-zero. This avoids a segmentation fault observed in the
ADIOI_Flatten routine, and fixes this particular with the hdf5 1.10.x testsuite in
OpenMPI with romio314.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-03-08 09:10:09 -06:00
George Bosilca
9bced03213
Improve the range and accuracy of MPI_Wtime.
As discussed on https://github.com/mpi-forum/mpi-issues/issues/77#issuecomment-369663119
the conversion to double in the MPI_Wtime decrease the range
and accuracy of the resulting timer. By setting the timer to
0 at the first usage we basically maintain the accuracy for
194 days even for gettimeofday.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2018-03-08 14:26:02 +09:00
Nathan Hjelm
5ed2fc2d48 mca/base: add support for additional variable types
This commit adds long, int32_t, uint32_t, int64_t, and uint64_t as
possible MCA variable types.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-03-01 20:42:27 -07:00
bosilca
9944d63de1
Merge pull request #4852 from thananon/pr/ob1_oos_fix
pml/ob1: fixed out of sequence bug.
2018-02-28 13:02:03 -05:00
Thananon Patinyasakdikul
09cba8b30b pml/ob1: fixed out of sequence bug.
This commit fixes #4795

- Fixed typo that sometimes causes deadlock in change of protocol.
- Redesigned out of sequence ordering and address the overflow case of
  sequence number from uint16_t.

Signed-off-by: Thananon Patinyasakdikul <tpatinya@utk.edu>
2018-02-27 13:49:40 -05:00
Valentin Petrov
bf4e694a96 coll/hcoll: Fix return codes
Signed-off-by: Valentin Petrov <valentinp@mellanox.com>
2018-02-22 17:48:29 +02:00
Matias Cabral
0a822f8f99
Merge pull request #4821 from nrspruit/OFI_mtl_multi_event_progress
MTL OFI: Added support for reading multiple CQ events in ofi progress
2018-02-20 14:59:47 -08:00
Jeff Squyres
9ef0f3d83a ompi/monitoring: add .sh versionig to common monitoring lib
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2018-02-20 07:07:23 -08:00
Artem Polyakov
b601dd504a ompi: ompi_mpi_init(): do not export threading level to modex.
For some of our configuration this flag increases per-process contribution
by ~20% while it is not being used currently.

The consumer of this flag was communicator ID calculation logic, but it was
changed in 0bf06de3f1.

Signed-off-by: Artem Polyakov <artpol84@gmail.com>
2018-02-18 01:40:15 +07:00
Spruit, Neil R
e7bff501cd MTL OFI: Added support for reading multiple CQ events in ofi progress
-Updated ompi_mtl_ofi_progress to use an array to read CQ events up to a
threshold that can be set by the Open MPI User.

-Users can adjust the number of events that can be handled in the
ompi_mtl_ofi_progress by setting "--mca mtl_ofi_progress_event_cnt #".

-The default value for the the number of CQ events that can be read in a
single call to ofi progress is 100 which is an average
based off workload usecase anaylsis showing 70-128 as the range of
multiple events returned during ofi progress.

Signed-off-by: Spruit, Neil R <neil.r.spruit@intel.com>
2018-02-15 09:41:14 -05:00
Nathan Hjelm
0e83568466 coll/libnbc: do not take lock in progress if there are no requests
This commit fixes a flaw in the progress function for libnbc. The
function was unconditionally taking a lock even if there are no
requests to process. This lock was showing up in vtune traces of
multi-threaded benchmarks.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-02-13 09:51:01 -07:00
Edgar Gabriel
a3a734b6d2 io/ompio: correctly reset the request
after performing the final OBJ_RELEASE on the request,
reset the user level variable to MPI_REQUEST_NULL.
Otherwise the c_2_f translation step in the fortran
interface fails.

Fixes issue #4807

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-02-13 09:18:25 -06:00
Jeff Squyres
e7f91f8068
Merge pull request #4527 from clementFoyer/osc-no-includes
Remove inter-dependencies between OSC modules.
2018-02-09 15:49:56 -05:00
Nathan Hjelm
da9f833f4a pml/ob1: ignore the eager limit of RDMA-only btls
This commit fixes a flaw in the eager limit check in pml/ob1. The
check was incorrectly checking if RDMA-only BTLs (BTLs without the
send flag) has a valid eager limit. This commit fixes the check by
adding an additional check for the send flag on the BTL module.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-02-07 12:42:44 -07:00
Clement Foyer
f5b4fc05f8 Remove inter-dependencies between OSC modules.
The osc monitoring component needed to include other OSC components
header in order to be able tu access communicator through the
component specific ompi_osc_*_module_t structures. This commit remove
the dependency, and resolve the issue #4523.

Extend the common monitoring API.

  * Now it's possible to translate from local rank to world rank from
    both the communicator and the group.
  * Remove useless hashtable as we directly use the w_group contained
    in window structure.

Add automatic generation at config time.

The templates are expanded at configure time. It creates a new header
file that generates all the variables/functions needed. Adding this
during the autogen automagicaly generates for each of the available
modules the proper functions.

Only keep a generated argv-style array.

Following Jeff's advice, the configure.m4 file generate a simple array
of module variables to be iterated over to find the proper module.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>
2018-02-07 11:52:00 +00:00
Ralph Castain
7ddffc627d
Merge pull request #4776 from rhc54/topic/rte
Correct abstraction break and update ignores
2018-01-31 04:34:05 -08:00
Ralph Castain
8e8a9aecc5 Correct abstraction break - direct reference to ORTE
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2018-01-30 21:19:14 -08:00
Gilles Gouaillardet
34b45cc879 osc/sm: fix the osc_free callback
If component selection fails, then module->bases might be unallocated
when ompi_osc_sm_free() in invoked, so test it before trying to free()
module->bases[0].

Thanks Martin Binder for the report.

Refs open-mpi/ompi#4770

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-01-31 11:23:21 +09:00
Sergey Oblomov
7a5811d0a8 request/state: update state for canceled request
- fixed issue in set state for canceled request

Signed-off-by: Sergey Oblomov <sergeyo@mellanox.com>
2018-01-29 18:26:20 +02:00
Edgar Gabriel
bcf26d419f fs/ufs and fs/lustre: remove erroneous return statement
an erroneous return statement has creeped in commit 1885d99
which leads to some processes not resetting stripe_size
and stripe_count correctly. This can lead in 3.0.x to different
fcoll modules being selected. The impact is not that dramatic on
master and 3.1.x, but could lead to problems as well.

Fixes #4745

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2018-01-24 14:07:21 -06:00
Xin Zhao
72ff2b1135 OMPI/OSC/UCX: adding atomic lock for fetch_and_op and compare_and_swap
Signed-off-by: Xin Zhao <xinz@mellanox.com>
2018-01-18 00:36:22 +02:00
Yossi Itigin
f2851fd502
Merge pull request #4724 from alex-mikheev/topic/ucx_as_default
ompi/oshmem: ucx is selected over yalla/ikrit by default
2018-01-17 17:41:49 +02:00
Alex Mikheev
640e945b9c ompi: pml/ucx: blocking send using ucp_tag_send_nbr
Signed-off-by: Alex Mikheev <alexm@mellanox.com>
2018-01-17 15:54:18 +02:00
Alex Mikheev
ae326546f4
ompi/oshmem: ucx is selected over yalla/ikrit by default
Signed-off-by: Alex Mikheev <alexm@mellanox.com>
2018-01-17 15:08:04 +02:00
Gilles Gouaillardet
cb5dfbe5b1 man: fix indentation of MPI_Comm_spawn[_multiple]
no code change.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-01-15 11:29:15 +09:00
Matias Cabral
8049c06a96
Merge pull request #4580 from matcabral/fix_comments_pr_425_osc_rmda
osc/rmda: fix missing opal_argv_free in mtls search.
2018-01-12 09:20:11 -08:00
Matias A Cabral
009ba475e1 osc/rmda: fix missing opal_argv_free in mtls search.
Use asprintf in description message to avoid missing default values
Signed-off-by: Matias Cabral <matias.a.cabral@intel.com>
2018-01-11 14:29:16 -08:00
Jeff Squyres
1c5664fdec
Merge pull request #4681 from nathanweeks/issue/f08_mpi_errcodes_ignore
Fix type of mpi_f08 MPI_ERRCODES_IGNORE
2018-01-10 12:34:16 -05:00
Gilles Gouaillardet
02f8215b25 ompi: enhance MPI_File_set_view datatype check.
Per MPI 3.1 chapter 13.3 :
"Derived etypes can be constructed by using any of the MPI
datatype constructor routines, provided all resulting typemap
displacements are non-negative and monotonically nondecreasing."
Same restriction applies to ftypes.

add the OMPI_DATATYPE_CHECK_FOR_VIEW() macro that is
check the underlying opal_datatype_t is monotonic, on top
of all checks performed in OMPI_DATATYPE_CHECK_FOR_RECV().

Since checking monotoniciy is expensive, check is only performed
when needed, but the result is cached by ompi_datatype_is_monotonic().

Thanks Wei-keng Liao for the valuable feedback.
Thanks George for the guidance.

Refs. open-mpi/ompi#4682

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2018-01-09 18:05:15 +09:00
Nathan T. Weeks
3158d2c5ed Fix type of mpi_f08 MPI_ERRCODES_IGNORE
Signed-off-by: Nathan T. Weeks <weeks@iastate.edu>
2018-01-07 05:36:41 -08:00
KAWASHIMA Takahiro
710080be63
Merge pull request #4667 from kawashima-fj/pr/f08-pmpi
fortran: Fix PMPI interface bugs in mpi_f08 module
2018-01-05 03:45:10 -06:00
bosilca
ef38ca5663
Merge pull request #4644 from bosilca/topic/treematch
Fix treematch topology assert
2018-01-02 21:21:54 -05:00
Gilles Gouaillardet
2dd345465f ompi/communicator: optimize ompi_comm_split()
set grp_local_rank as MPI_UNDEFINED before invoking
ompi_comm_nexcid() in order to benefit from the optimizations
introduced in open-mpi/ompi@68167ec879

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-12-28 15:34:04 +09:00
Yossi Itigin
697a9437e2
Merge pull request #4666 from alex-mikheev/topic/pml_ucx_recv_fix
ompi: pml ucx: improve recv latency
2017-12-26 20:12:18 +02:00
Alex Mikheev
e7bf0617cf
ompi: pml ucx: improve recv latency
Signed-off-by: Alex Mikheev <alexm@mellanox.com>
2017-12-26 16:24:16 +02:00
KAWASHIMA Takahiro
bd2fe9c324 fortran: Call PMPI from PMPI_Status_set_cancelled_f08
This is a bug which was forgotten to change in c08f97b030.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-12-26 15:53:12 +09:00
KAWASHIMA Takahiro
00e3c7a973 fortran: Align indentation
This change makes comparison of `mpi-f08-interfaces.F90` and
`pmpi-f08-interfaces.F90` easier.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-12-26 14:39:09 +09:00
KAWASHIMA Takahiro
056eb39b12 fortran: Correct type of info_used
It is incorrectly typed as `MPI_Comm` in only `pmpi` in 24f7bd327e.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-12-26 14:39:09 +09:00
KAWASHIMA Takahiro
0c3a534b32 fortran: Use C_PTR for buffer_addr
It was changed to use `C_PTR` in only `mpi` in fc69c0be24.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-12-26 14:39:09 +09:00
KAWASHIMA Takahiro
d4fc404dc6 fortran: Change PMPI_Aint_{add,diff} to functions.
They were incorrectly changed to subroutines in only `pmpi`
in 258d1aa160.

Strictly speaking, this change involves binary incompatibility.
But nobody used these subroutines and nobody will be affected because
these subroutines were useless (didn't return a calculated value).

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-12-26 14:39:09 +09:00
KAWASHIMA Takahiro
9240967b8f fortran: Remove ASYNCHRONOUS from mpi_f08 pmpi
It was removed from only `mpi` as a bug fix in db41d749c1.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-12-26 14:39:09 +09:00
KAWASHIMA Takahiro
4cd917a3fd
Merge pull request #4659 from yanagibashi/pr/fortran2008
Add missing Fortran 2008 binding subroutines
2017-12-22 01:53:36 -06:00
Tsubasa Yanagibashi
3f4b373856 Add missing Fortran 2008 binding subroutines
added missing Fortran 2008 binding pmpi_{*} subroutines to Open MPI.

Signed-off-by: Tsubasa Yanagibashi <fj2505dt@aa.jp.fujitsu.com>
2017-12-22 13:45:53 +09:00
Aravind Gopalakrishnan
fb68726baf MTL OFI: Allow retries in MTL progress for interrupted syscalls
This fixes a regression in sockets provider which could return -EINTR value
from fi_cq_read() due to a syscall being interrupted. The error value is
currently interpreted as fatal condition. Relax the rule so that we can retry
fi_cq_read() operation.

Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
2017-12-20 14:58:49 -08:00
George Bosilca
38455845db
Fix asserts.
In both cases we were comparing with the wrong size, it should be either
the number of local processes or the number of nodes, and not the size
of the communicator.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-12-20 11:51:35 -05:00
George Bosilca
808f865e9d
Force all output to use OMPI infrastructure.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-12-20 11:50:51 -05:00
Jed Brown
533800070e MPI_Attr_get: doc fix: MPI_Comm_create_attr -> MPI_Comm_get_attr
MPI_Comm_create_attr does not exist.

Signed-off-by: Jed Brown <jed@jedbrown.org>
2017-12-17 07:44:22 -07:00
Gilles Gouaillardet
794cc09d3e mpiext: fix path to Fortran 2008 modules
OMPI_FORTRAN_USEMPIF08_MOD macro was removed in open-mpi/ompi@791bcee6c0
so this macro is now manually expanded to mpi/fortran/use-mpi-f08/mod

Thanks to Nathan T. Weeks for reporting

Refs open-mpi/ompi#3605

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-12-11 11:02:43 +09:00
Matias Cabral
2c86b8723d
Merge pull request #4510 from matcabral/mtl_psm2_shadow_vars
New flag for MCA parameters that allows a behaving with a default value of "unset".
2017-12-04 12:25:37 -08:00
Howard Pritchard
b160cf6339
Merge pull request #4533 from hppritcha/topic/ofi_mtl_mprobe_fixes
mtl/ofi: fix problem with mprobe/mrecv
2017-12-04 09:11:47 -07:00
Howard Pritchard
2233e44848
Merge pull request #4534 from hppritcha/topic/fix_a_segv_in_request
pml/cm: check for request comp. before completing bsend
2017-12-04 09:09:41 -07:00
Gilles Gouaillardet
2f5b1e9fe0
Merge pull request #4551 from ggouaillardet/topic/communicator_mutex_c_lock
Make usage of ompi_communicator_t, ompi_file_t and ompi_win_t mutex consistent
2017-12-04 09:20:52 +09:00
Edgar Gabriel
1f151be6d2 io/ompio: introduce a new function to retrieve mca parameter values
ompio has the unique problem, that mca parameters set in the io/ompio component
have to be accessible from other frameworks as well. This is mostly done to avoid
a replication in the parameter names and to reduce the number of mca parameters that
and end-user has to worry about.

This commit introduces a generic function to retrieve ompio mca parameters, the function pointer
is stored on the file handle. It replaces two functions that used the same concept already for
one parameter each.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-12-01 10:00:23 -06:00
Gilles Gouaillardet
b8e77ba759 mpi/c: use OPAL_THREAD[UN]LOCK() instead of opal_mutex_[un]lock()
in order to keep consistency between ompi_communicator_t, ompi_file_t
and ompi_win_t.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-12-01 16:06:32 +09:00
Gilles Gouaillardet
1ba4c185bc ompi/communicator: destruct ompi_communicator_t's c_lock in the destructor
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-12-01 16:06:27 +09:00
Gilles Gouaillardet
5f1a967351 ompi/file: rename ompi_file_t's f_mutex into f_lock
in order to use a consistent name between ompi_file_t,
ompi_win_t and ompi_communicator_t

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-12-01 16:06:22 +09:00
bosilca
5cb72aa568
Merge pull request #4552 from hjelmn/asm_cleanup2
Add atomic fetch-and-op and compare-exchange functions
2017-11-30 22:29:38 -05:00
Nathan Hjelm
7893248c5a opal/asm: add fetch-and-op atomics
This commit adds support for fetch-and-op atomics. This is needed
because and and or are irreversible operations so there needs to be a
way to get the old value atomically. These are also the only semantics
supported by C11 (there is not atomic_op_fetch, just
atomic_fetch_op). The old op-and-fetch atomics have been defined in
terms of fetch-and-op.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-11-30 10:41:23 -07:00
Nathan Hjelm
1282e98a01 opal/asm: rename existing arithmetic atomic functions
This commit renames the arithmetic atomic operations in opal to
indicate that they return the new value not the old value. This naming
differentiates these routines from new functions that return the old
value.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-11-30 10:41:22 -07:00
Nathan Hjelm
9d0b3fe9f4 opal/asm: remove opal_atomic_bool_cmpset functions
This commit eliminates the old opal_atomic_bool_cmpset functions. They
have been replaced by the opal_atomic_compare_exchange_strong
functions.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-11-30 10:41:22 -07:00
Gilles Gouaillardet
f1778d2778 communicator: remove the USE_MUTEX_FOR_COMMS macro
It should have always been #define'd in order to correctly handle the
multi-threaded case.

Also fix indentation in ompi/mpi/c/comm_get_errhandler.c

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-11-30 14:29:11 +09:00
Gilles Gouaillardet
7c3e675479 fix communicator's c_lock usage
- initialize c_lock in the ompi_communicator_t constructor
 - USE_OPAL_THREAD_[UN]LOCK(c_lock)
 - #ifdef USE_MUTEX_FOR_COMMS protect c_lock access

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-11-30 14:27:59 +09:00
Nathan Hjelm
45db3637af osc/rdma: bug fixes
This commit fixes the following bugs:

 - Allow a btl to be used for communication if it can communicate with
   all non-self peers and it supports global atomic visibility. In
   this case CPU atomics can be used for self and the btl for any
   other peer.

 - It was possible to get into a state where different threads of an
   MPI process could issue conflicting accumulate operations to a
   remote peer. To eliminate this race we now update the peer flags
   atomically.

 - Queue up and re-issue put operations that failed during a BTL
   callback. This can occur during an accumulate operation. This was
   an unhandled error case.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-11-29 12:43:58 -07:00
Nathan Hjelm
67e26b6e5a
Merge pull request #4482 from matcabral/osc_rdma_skip_mtls
osc/rdma: mca parameter to list MTLs that lower osc rdma priority
2017-11-29 09:34:33 -07:00
Nathan Hjelm
647b40f3f2
Merge pull request #4442 from bosilca/topic/ob1_pvar
Topic/ob1 pvar
2017-11-29 09:31:07 -07:00
Ralph Castain
5e109fba6a Take the smallest change approach to fixing the hang on disconnect of connect/accept operations by replacing the call to pmix.disconnect with a simple pmix.fence as this is all OMPI currently requires
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-11-29 03:30:40 -08:00
Ralph Castain
7ad6886a30 Add a new OMPI rte component to support direct-launch using PMIx.
Cleanup several places where abstraction violations crept into OMPI layer (direct reference of ORTE). Add some missing includes that were exposed by this change.

Note that this compiles, but I haven't tested it for execution yet. Handing it over to Noah Evans for completion

Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-11-28 12:05:01 -08:00
bosilca
c4c5df4d1c
Merge pull request #4476 from hjelmn/errhandler_fixes
ompi/errhandler: make set/get actually thread safe
2017-11-27 19:43:42 -05:00
Nathan Hjelm
6b68d1cfc8 ompi/errhandler: make set/get actually thread safe
The current versions of these functions have a fatal flaw. If a
errhandler set and free call is made by another thread while the
thread calling get is between the cmpset and retain then we will
retain an invalid object. Fixing this by just using locking. This is
not a critical path so this should be ok.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-11-27 15:14:17 -07:00
Howard Pritchard
3285325884 pml/cm: check for request comp. before completing bsend.
Turns out there are edge cases where an MTL's isend
method may end up marking a send request complete prior
to returning to the CM code.  The would end up causing
problems in the bsend path since the ompi_request_complete
would end up getting invoked a second time on this request.
This ended up causing segfaults, etc. in ompi_request_complete .

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
2017-11-27 12:27:43 -07:00
Ralph Castain
3906aaf41a Silence warnings
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-11-25 11:50:18 -08:00
Howard Pritchard
cd48eccbae mtl/ofi: fix problem with mprobe/mrecv
At least with some providers (sockets and GNI), the mprobe/mrecv
ofi mtl methods were incorrect.  For these two providers at least
one must supply the original tag and mask bits used with the
prior FI_PEEK | FI_CLAIM request that had been used to probe for
the message.

These providers take a strict interpretation of the following sentence
from the libfabric fi_tagged man page:

```
 Claimed messages can only be retrieved using a subsequent, paired receive  operation  with  the  FI_CLAIM  flag  set.
```

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
2017-11-24 08:11:18 -07:00
Edgar Gabriel
75ab006ec0 io/ompio: add a new option to disable amode overwriting
ompio has historically changed the WRONLY flag provided by the applicaiton
to RDWR to allow for the data sieving optimization within the two-phase I/O
fcoll component. This change did not have a performance impact
on regular UNIX file systems, but seems to hurt performance on NFS (and maybe Lustre?)

So provide an option that allows to keep the WRONLY option, and raise an error
if tha fcoll/two-phase would actually like to use the data sieving.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-11-17 13:13:38 -06:00
Matias A Cabral
1fad59465f New flag for MCA parameters that allows a behaving with a default value
of "unset".
mtl/psm2: Update some shadow mca parameters to use the default "unset".
mtl/psm2: Add new shadow parameter to allow specifying the service level.

Signed-off-by: Matias A Cabral <matias.a.cabral@intel.com>
2017-11-16 16:28:50 -08:00
Matias Cabral
d1869a725a
Merge pull request #4467 from matcabral/master
mtl/ofi: Set data and control progress options default values to FI_PROGRESS_UNSPEC
2017-11-13 07:35:39 -08:00
Jeff Squyres
a8686a6813 mtl ofi: squelch compiler warnings
gcc 5.2 complains:

```
mtl_ofi_component.c: In function ‘ompi_mtl_ofi_finalize’:
mtl_ofi_component.c:613:5: warning: suggest parentheses around assignment used as truth value [-Wparentheses]
     if (ret = fi_close((fid_t)ompi_mtl_ofi.fabric)) {
     ^
```

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-11-11 05:07:11 -08:00
Jeff Squyres
5a6ddf42d6 mtl ofi: it is not an error to return no data from fi_getinfo()
Before this commit, the presence of usNIC devices -- which will
(currently) return no data when fi_getinfo() is queried for tagged
matching providers -- would cause an error message to be displayed.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-11-11 05:07:11 -08:00
Jeff Squyres
f910f554f7 mtl ofi: show the positive value of the error
The value of ret is negative (e.g., -61), but it is displayed in the
help message as `%zd`, which renders as unsigned (i.e., a giant
positive value).  So make sure to negate the negative value before
rendering it (e.g., so we display "61", not "4294967235").

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-11-11 05:07:11 -08:00
Jeff Squyres
e8c13ef286 mtl ofi: fix trivial comment whitespace
No code or logic changes.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-11-11 05:07:10 -08:00
Jeff Squyres
bed1930df8 mtl ofi: fix formatting of help message
No code or logic changes.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-11-11 05:07:05 -08:00
bosilca
3f43e2c8df
Merge pull request #4419 from thananon/pr_newob1
pml/ob1: match callback will now queue wrong sequence frag and return.
2017-11-10 20:10:53 -05:00
Matias A Cabral
80c8858c5a osc/rdma: add an mca parameter to list MTLs for which osc pt2pt should have
higher priority than rdma and default to psm2.

Context: the Intel Omni-path driver (hfi1) has verbs support, so the openib
btl is available to use. However, at a bad performance. Without this
change osc rdma using btl openib is the default choice when running on Intel
Omni-path, with a lower performance than osc pt2pt over mtl psm2.

Signed-off-by: Matias A Cabral <matias.a.cabral@intel.com>
2017-11-09 11:54:11 -08:00
George Bosilca
409638bdf4 Keep the out-of-sequence fragment ordered.
Rework the logic to handle the out-of-sequence fragments on the receiver
side. A large number of OOS messages are still arriving even in single
threaded scenarios.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-11-08 14:27:13 -05:00
Matias Cabral
b76bb42ac1 mtl/ofi: Set data and control progress options default
values to FI_PROGRESS_UNSPEC so each provider will use its default.

Signed-off-by: Matias Cabral <matias.a.cabral@intel.com>
2017-11-08 08:24:33 -08:00
Thomas Naughton
86bb6f8bac
Merge pull request #4444 from naughtont3/tjn-fix-plm-monitoring-configury
configury: single quote to avoid trouble with BSD
2017-11-08 10:56:37 -05:00
Jeff Squyres
5ac1ea4512 ompi_info: add ipv6 config param
Allow ompi_info to show whether IPv6 support was enabled in Open MPI
or not.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-11-08 07:02:27 -08:00
Edgar Gabriel
0c3aa44ba3
Merge pull request #4454 from edgargabriel/topic/large-file-bug-fix
io/ompio: fix a bug in handling large write/read operations
2017-11-07 09:46:47 -06:00
Gilles Gouaillardet
18d3897479 fs/ufs: add some more error codes on file_open
set proper error codes in mca_fs_ufs_file_open by mapping the errno value to
the MPI error code.

Refs. open-mpi/ompi#4443

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-11-07 11:09:32 +09:00
Edgar Gabriel
c9bb049d00 io/ompio: fix a bug in handling large write/read operations
This is a bug fix based on a problem reported on the mailing list.
For very large read/write operations, ompio breaks the operation
down into multiple cycles. The problem was that
one of the variables required to maintain its values
across the different cycles did not do that, and because
of that the calculations of the memory offsets was wrong.

Fixes issue #4453

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-11-06 11:48:13 -06:00
Edgar Gabriel
91881c4fdc fs/lustre: update component to match fs/ufs
the fs/lustre component has missed out on a number of updates to the fs/ufs component.
This commit tries to import all the changes performed on the fs/ufs component
w.r.t to the file_open operation, including updates on how the amode is set,
error is propegated and setting the fs_block_size value (which is required for
locking purposes).

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-11-03 16:11:46 -05:00
Edgar Gabriel
1885d99ac7 fs/ufs: set proper error codes on file_open
set proper error codes in mca_fs_ufs_file_open by mapping the errno value to
the MPI error code. Fixes an issue reported on the mailing by Wei-keng Liao

Fixes Issue #4443

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-11-03 16:06:31 -05:00
Thomas Naughton
c5dc41ee1a configury: single quote to avoid trouble with BSD
Signed-off-by: Thomas Naughton <naughtont@ornl.gov>
2017-11-03 11:34:28 -04:00
George Bosilca
d261282029
OB1 pvars should be linked with the OB1 component.
If not the pvars will remain valid after the OB1 PML is unloaded, and
any access will segfault (the callbacks associated with the pvar will
point to the memory of the dlclosed module).

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-11-03 01:18:26 -04:00
bosilca
3e1f771045
Merge pull request #4403 from naughtont3/tjn-fix-plm-monitoring-configury
fix PML monitoring configury to compile DSOs
2017-11-02 19:08:20 -04:00
bosilca
63e8a8c608
Merge pull request #4431 from hjelmn/asm_cleanup
opal: rename opal_atomic_cmpset* to opal_atomic_bool_cmpset*
2017-11-02 18:45:56 -04:00
Matias Cabral
c8aa22ee22
Merge pull request #4304 from aravindksg/master
Fix OFI MTL to recognize correct CQ empty scenario and improve error reporting
2017-11-02 11:56:57 -07:00
George Bosilca
b2b3da3046
Do not access the frag after returning it.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-10-31 16:39:23 -04:00
Nathan Hjelm
3ff34af355 opal: rename opal_atomic_cmpset* to opal_atomic_bool_cmpset*
This commit renames the atomic compare-and-swap functions to indicate
the return value. This is in preperation for adding support for a
compare-and-swap that returns the old value. At the same time the
return type has been changed to bool.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-10-31 12:47:23 -06:00
Ralph Castain
27f3d417ca Revert the MPI_Init fence operations to use volatile bool instead of thread macros.
The problem is that the waiting thread is cycling using OMPI_LAZY_WAIT_FOR_COMPLETION so it can exercise opal_progress. This probably isn't as critical for the modex step, but definitely necessary for the barrier at the end of mpi_init. The problem this creates is that the lazy macro exits as soon as "active" becomes false, and then we destruct the lock.

However, wakeup_thread sets "active" to false - and then calls the condition broadcast to wakeup any waiting threads. So there is a race condition between that broadcast and the lock destruct.

Add OPAL_ACQUIRE_OBJECT and OPAL_POST_OBJECT memory barriers to help protect against thread race conditions on some platforms

Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-10-31 08:09:02 -07:00
Aravind Gopalakrishnan
285fc42b4e Fix OFI MTL to recognize correct CQ empty scenario
Currently, the progress function is incorrectly interpreting any error
value other than a positive value or -FI_EAVAIL to mean CQ is empty.
CQ is empty only if fi_cq_read() call returned -EAGAIN error
code. Fix that here.

While at it, fix help text output for calls made to OFI API.

Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
2017-10-30 12:13:44 -07:00
Thananon Patinyasakdikul
e3b267a8fe pml/ob1: match callback will now queue wrong sequence frag and return.
In multithreaded case, it is expensive to release the lock, call the slow match
and retake the lock again just to queue the frag. This patch will eliminate number of
lock taken by queueing the frag right away and return.

Signed-off-by: Thananon Patinyasakdikul <tpatinya@utk.edu>
2017-10-30 12:01:50 -04:00
Aravind Gopalakrishnan
bea4503f95 Move help text output regarding PSM2_CUDA envvar to component init phase
The messages should be printed only in the event of CUDA builds and in the
presence of supporting hardware and when PSM2 MTL has actually been selected
for use. To this end, move help text output to component init phase.

Also use opal_setenv/unsetenv() for safer setting, unsetting of the environment
variable and sanitize the help text message.

Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
2017-10-26 16:01:01 -07:00
Thomas Naughton
86d282d6dd fix PML monitoring configury to compile DSOs
Signed-off-by: Thomas Naughton <naughtont@ornl.gov>
2017-10-26 15:53:11 -04:00
Ralph Castain
cf3bc4f55b Merge pull request #4346 from matcabral/psm2_mtl_mq_thread_fix
MTL PSM2: add a thread lock while peeking and completing the psm2 requests.
2017-10-24 16:41:29 -05:00
Ralph Castain
0353be9704 Update MPI init to properly skip barriers
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-10-23 19:28:34 -07:00
Ralph Castain
6ea3c8a0bd Update the interlib example to show an alternative method for model declaration. Add a missing range value to the OPAL layer. Make it easier to see OMPI model callbacks
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-10-23 11:27:42 -07:00
Matias Cabral
b81bcd4b0d MTL PSM2: add a thread lock while peeking and completing the psm2
requests.
Reviewed-by: Gopalakrishnan, Aravind <aravind.gopalakrishnan@intel.com>
Signed-off-by: Matias Cabral <matias.a.cabral@intel.com>
2017-10-20 14:46:48 -07:00
Edgar Gabriel
be0de21e6f fs/ufs and fbtl/posix: cleanup lock management
This commit looks large, but its really mostly a cleanup step.
1. introduce proper error handling for the return values of fcntl and the fbtl_posix_lock function
2. rename a parameter to more accurately reflect what it does
3. introduce an mca parameter in the fs/ufs component that allows to control
   what the level of locking the user would like to enforce
4. move the initialization of the fs_block_size parameter from fs/ufs into the
   common/ompio component. An fs component might be allowed to overwrite this
   value, but none of the actual fs components do that.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-10-19 14:56:28 -05:00
Edgar Gabriel
e62f9d2e52 fs/ufs: ensure that the never-lock flag is set if not on NFS
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-10-19 13:32:40 -05:00
Edgar Gabriel
f66c55f77a fbtl/posix: fixes in the offset calculation and for aio operations
our own internal testsuite passes now correctly. More testing to follow.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-10-19 13:32:39 -05:00
Edgar Gabriel
a3c638bc38 fbtl/posix: add support for file locking for the non-blocking operations
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-10-19 13:32:38 -05:00
Edgar Gabriel
415e76514d fbtl/posix: make the code compile
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-10-19 13:32:37 -05:00
Edgar Gabriel
f5e158c869 fbtl/posix: first cut in adding locking support
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-10-19 13:32:37 -05:00
Gilles Gouaillardet
9771c575f5 Merge pull request #4352 from edgargabriel/pr/sem_close_fix
sharedfp/sm: close the named semaphore
2017-10-19 17:04:43 +09:00
Edgar Gabriel
4d995bd4eb sharedfp/sm: close the named semaphore
in case a named semaphore is used, it is necessary to close the semaphore to remove
all sm segments. sem_unlink just removes the name references once all proceeses have closed
the sem.

Fixes issue: #4336

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>

sharedfp/sm: unlink only needs to be called by one process

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-10-18 10:37:30 -05:00
Aurelien Bouteiller
3ef23f41a3
Bugfix a crash when a comm cannot be initialized
Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
2017-10-18 11:32:37 -04:00
Valentin Petrov
1e311b2619 coll/hcoll: dtype fallback optimization
If hcoll fails to create mpi derived type let's set zero_dte on this dtype.
    This will save cycles on subsequent collective calls with the same derived
    type since we will not try to create hcoll type again.

Signed-off-by: Valentin Petrov <valentinp@mellanox.com>
2017-10-06 10:29:29 +03:00
Valentin Petrov
06ef344630 coll/hcoll: extends dtypes support
Adds support for legacy MPI_UB/LB types (old apps may use it) as
    well as for BOOL/WCHAR.

Signed-off-by: Valentin Petrov <valentinp@mellanox.com>
2017-10-06 10:29:29 +03:00
Geoff Paulsen
be7b0af5d9 Merge pull request #3609 from markalle/pr/single_type_with_LB_UB
single_predefined_type with MPI_LB/UB
2017-10-04 15:13:09 -05:00
Mark Allen
e24d5ccb7e single_predefined_type with MPI_LB/UB
The ompi_datatype_get_single_predefined_type_from_args() recurses down
into a constructed type to identify what base datatype it's built from
if it's built from a single type.  But if the type has MPI_LB/MPI_UB,
for example
    lens[0] = 1;
    lens[1] = 1;
    disps[0] = 0;
    disps[1] = 0;
    types[0] = MPI_LB;
    types[1] = MPI_INT;
    MPI_Type_create_struct(2, lens, disps, types, &mydt);
then this function will see the base type MPI_LB as differing from MPI_INT
and will identify mydt as not being constructed from a single base type, so
the type will be rejected for calls like MPI_Accumulate.

I think those "meta data" types shouldn't result in rejection like that, and
the above mydt should still be identified as having a single base type
of MPI_INT.

Addition: boslica wanted another change discussed here
    https://github.com/open-mpi/ompi/pull/3609
relating to the calculation for "count" after identifying the
predefined_type that was being used.

Signed-off-by: Mark Allen <markalle@us.ibm.com>
2017-10-03 19:08:18 -04:00
George Bosilca
bdbea63a1c
Update the MPI standard reference.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-10-03 16:48:50 -04:00
George Bosilca
a3ac67be0d
Remove double include.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-10-02 21:33:40 -04:00
Gilles Gouaillardet
9492766dbd romio: disable Fortran support
romio314 is a just a component that does not require Fortran bindings,
so simply disable Fortran support to prevent warnings about deprecated flags

Fixes open-mpi/ompi#4281

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-10-02 21:44:33 +09:00
Matias Cabral
d9b2c94d4a Merge pull request #4286 from aravindksg/master
Use opal_show_help to warn about PSM2_CUDA envvar setting
2017-10-01 11:10:19 -07:00
George Bosilca
2a2db13b32
Gracefully deal with a get returning 1 (complete right away).
Kudos to @EmmanuelBRELLE for spotting it.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-10-01 02:24:02 -04:00
Aravind Gopalakrishnan
f8a2b7f6bf Use opal_show_help to warn about PSM2_CUDA envvar setting
If Open MPI is configured with CUDA, then user also should be using a CUDA build of
PSM2 and therefore be setting PSM2_CUDA environment variable to 1 while using
CUDA buffers for transfers. If we detect this setting to be missing, force set
it. If user wants to use this build for regular (Host buffer) transfers, we
allow the option of setting PSM2_CUDA=0, but print a warning
message to user that it is not a recommended usage scenario.

Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
2017-09-29 17:04:10 -07:00
George Bosilca
92e4ecb618
Fix the offset computation.
Also ensure the marked array is correctly freed in all cases.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-27 11:42:21 -04:00
George Bosilca
0ceed71368
Fix Coverity warnings in treematch topo.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-26 22:24:01 -04:00
Edgar Gabriel
4c0d347412 Merge pull request #4230 from edgargabriel/topic/no-smart-fview
io/ompio: add a new grouping option avoiding communication
2017-09-26 10:56:06 -05:00
bosilca
f44e674992 Merge pull request #4074 from bosilca/topic/coverity
Fix coverity complaints.
2017-09-25 15:05:31 -04:00
Guillaume Mercier
4e7c130c31
Add correct reordering computation in partially distributed case.
Replaced matching array with k and bcast with scatter.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
Signed-off-by: Guillaume Mercier <mercier@labri.fr>
2017-09-25 13:10:11 -04:00
George Bosilca
3dd1d8cb53
Delay the first check for the HWLOC topology.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-25 13:09:57 -04:00
George Bosilca
28046b37df
Always succesfully return.
As the reordering is an optional step, if any operation during the
reorder fails we can return the duplicata of the original communicator
associated with the topology information.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-25 12:53:20 -04:00
George Bosilca
219a96fa69
Prevent memory leaks.
Reorder the code to simplify the memory management.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-25 12:53:20 -04:00
George Bosilca
64bff0e326
Disable monitoring if we compile statically.
Protect all components against compilation on static builds.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-25 12:18:23 -04:00
George Bosilca
458ccc12e1
Move the profiling library in common/monitoring
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-25 12:18:23 -04:00
Clément FOYER
f334607c34
Simplify the communicator's name caching management (#6)
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>
2017-09-25 12:18:23 -04:00
bosilca
a680b3ac6d Merge pull request #3853 from clementFoyer/master
OMPI monitoring: Simplify the communicator's name caching management + misc test changes
2017-09-25 12:14:36 -04:00
yohann
1f8cabc890 mtl/ofi: Fix provider selection.
This allows mtl_ofi_provider_include to work with layered providers as well.
e.g. --mca mtl_ofi_provider_include "providerX;ofi_rxm"

Signed-off-by: yohann <yohann.burette@intel.com>
2017-09-20 16:00:50 -07:00
Gilles Gouaillardet
b9315edb85 configury: remove the --disable-mpi-io option
Fixes open-mpi/ompi#2185

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-09-20 14:39:09 +09:00
Edgar Gabriel
76a8c67575 io/ompio: add a new grouping option avoiding communication
the new grouping option simple+ performs all calculations used
for the aggregator selection as if the default file view would be used,
thus avoiding communication in file_set_view all together. This mode
is useful for applications that do not set a file view, but use
explicit offset operations on the default file view.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-09-18 12:30:34 -05:00
Ralph Castain
ed508010b4 Remove stale tools
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-09-18 07:30:47 -07:00
Ralph Castain
3c914a7a97 Complete the fix of the ORTE DVM. We will now use "prun" instead of "orterun -hnp foo" to execute jobs. This provides the feature of automatic discovery of the orte-dvm so you don't need to manually enter URI's or contact file locations. All IO is forwarded to prun.
Still in the "needs to be done" category:

* mapping/ranking/binding options aren't correctly supported

* if the DVM encounters some errors (e.g., not enough resources for the job), the resulting error is globally set and impacts any subsequent job submission

Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-09-16 13:13:07 -07:00
Ralph Castain
3f8908871b Since the DVM is now tied to prun, don't build the DVM either unless prun can be built
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-09-13 11:55:10 -07:00
Brian Barrett
637ebf60f9 atomics: Remove requirement of 64 bit atomics
Remove two of the three  instances of components requiring
64 bit atomics, even on 32 bit systems.  The SM OSC component
also uses 64 bit atomics, but is a more complicated fix that
will follow this one.  Currently, no one is testing on
platforms that don't provide 64 bit atomics (even in 32 bit
mode), but with the removal of the non-inline assembly for
IA32, the older compilers on Absoft's test systems now
result in no practical way to call cmpxchg8 in 32 bit mode.
At that point, these failures started popping up.

Signed-off-by: Brian Barrett <bbarrett@amazon.com>
2017-09-11 19:50:10 -07:00
Nathan Hjelm
7cdda24206 osc/sm: do not require 64-bit atomic math
This commit fixes a compile issue on 32-bit systems that do not
support 64-bit atomic math. The active target path was using 64-bit
atomics exclusively to support PSCW. This commit updates the code to
use either 32 or 64-bit atomic math depending on what is available.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-09-11 14:10:38 -10:00
Nathan Hjelm
4bba8774f4 monitoring: fix MPI_T regression
The monitoring code causes MPI_T based tools to segfault when
monitoring is disabled. This happens because the performance
variables remain registered after the common/monitoring
component is dlclosed due to a missing variable registration
flag. This commit adds the necessary flag to all the registered
performance variables.

The issue on github is #4162. Close when applied to master.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-09-06 14:24:35 -06:00
bosilca
dc538e9675 Merge pull request #1177 from bosilca/topic/large_msg
Topic/large msg
2017-09-05 13:30:19 -04:00
Gilles Gouaillardet
ecb6b81a05 mpi: correctly handle MPI_IN_PLACE by memchecker in neighborhood collectives
MPI_IN_PLACE is not a valid send buffer for neighborhood collectives, so do not
invoke memchecker in this case.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-09-04 11:21:32 +09:00
Gilles Gouaillardet
66c9485e77 MPI_Isend: memchecker do not mark send buffer as unaccessible after pml isend invokation
Today's MPI standard mandates the send buffer remains accessible during the send operation.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-09-04 11:21:32 +09:00
Gilles Gouaillardet
af8242a121 pml/ob1: have memchecker make recv buffer defined again when mca_pml_ob1_recv completes
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-09-04 11:18:05 +09:00
Gilles Gouaillardet
6ee9366243 MPI_Wait: correctly handle MPI_STATUS_IGNORE in MEMCHECKER
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-09-04 11:18:05 +09:00
Aravind Gopalakrishnan
2e83cf15ce Add support for GPU buffers for PSM2 MTL
PSM2 enables support for GPU buffers and CUDA managed memory and it can
directly recognize GPU buffers, handle copies between HFIs and GPUs.
Therefore, it is not required for OMPI to handle GPU buffers for pt2pt cases.
In this patch, we allow the PSM2 MTL to specify when
it does not require CUDA convertor support. This allows us to skip CUDA
convertor init phases and lets PSM2 handle the memory transfers.

This translates to improvements in latency.
The patch enables blocking collectives and workloads with GPU contiguous,
GPU non-contiguous memory.

Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
2017-09-01 16:59:03 -07:00
George Bosilca
866899e836
Always abide to the RDMA pipeline limit.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-01 18:52:48 -04:00
George Bosilca
050bd3b6d7
Make the pipeline depth an int instead of a size_t. While
they are supposed to be unsigned, casting them to a signed
value for all atomic operations is as errorprone as handling
them as signed entities.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-09-01 18:52:48 -04:00
Clement Foyer
9a8fc1b9f1 Simplify the communicator's name caching management
Remove useless over-initialization

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>
2017-08-29 12:52:47 +02:00
Yossi Itigin
14a93a5992 pml_ucx: fix tag/context_id layout and upper bounds.
Signed-off-by: Yossi Itigin <yosefe@mellanox.com>
2017-08-27 17:15:48 +03:00
Josh Hursey
ad87aa2674 Merge pull request #4121 from jjhursey/explore/dlopen-local
mca: Dynamic components link against project lib
2017-08-25 13:15:51 -05:00
Joshua Hursey
49c40f05d4 mpi/java: Remove dlopen() workaround
* See discussion on Issue #3705 regarding why this is no longer needed.

Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
2017-08-24 11:56:17 -04:00
Joshua Hursey
e1d079544b mca: Dynamic components link against project lib
* Resolves #3705
 * Components should link against the project level library to better
   support `dlopen` with `RTLD_LOCAL`.
 * Extend the `mca_FRAMEWORK_COMPONENT_la_LIBADD` in the `Makefile.am`
   with the appropriate project level library:
```
MCA components in ompi/
       $(top_builddir)/ompi/lib@OMPI_LIBMPI_NAME@.la
MCA components in orte/
       $(top_builddir)/orte/lib@ORTE_LIB_PREFIX@open-rte.la
MCA components in opal/
       $(top_builddir)/opal/lib@OPAL_LIB_PREFIX@open-pal.la
MCA components in oshmem/
       $(top_builddir)/oshmem/liboshmem.la"
```

Note: The changes in this commit were automated by the script in
the commit that proceeds it with the `libadd_mca_comp_update.py`
script. Some components were not included in this change because
they are statically built only.

Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
2017-08-24 11:56:16 -04:00
Ralph Castain
e02c39385a Merge branch 'master' into topic/modex 2017-08-22 20:06:35 -07:00
George Bosilca
50f471e31e
Cleanup a set of warnings reported by Ralph.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-22 23:00:18 -04:00
Ralph Castain
d80b0c7990 If the HWLOC shared memory system is unable to connect, then fallback to providing the topology via XML. Do not automatically provide the XML to every process as that defeats the purpose of the shared memory system. Instead, use PMIx_Query_info_nb to get the info from the server when required.
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-08-22 18:12:26 -07:00
Jeff Squyres
ea5093fc14 mpi/info_delete: fix return code
Per MPI-3.1, ensure to raise an MPI exception with value
MPI_ERR_INFO_NOKEY if we try to MPI_INFO_DELETE a key that does not
exist.  Thanks to @dalcinl (Lisando Dalcin) for raising the issue.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-08-22 08:56:40 -07:00
Gilles Gouaillardet
a3e31fa8d0 ompi/communicator: plug a memory leak in ompi_comm_init()
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-08-21 11:47:11 +09:00
Edgar Gabriel
99c7482dd8 Merge pull request #3739 from cniethammer/sharedfp_sm_file_dir
Create file for file backed shared memory in process job session dir.
2017-08-15 11:53:30 -05:00
Edgar Gabriel
8fe1c63e25 io/ompio: change the increment for cost based aggr. selection
- change the increment used to test various no. of aggregators
  to avoid using only power of two numbers
- convert some paratemers in the cost function from integers to
  to floats for providing smoother and more consistent results
- set the FVIEW_IS_SET flag on the file *only* if the user
  has set anything else than the default file view.

Signed-off-by: Edgar Gabriel <gabriel@cs.uh.edu>
2017-08-15 09:50:41 -05:00
Edgar Gabriel
f258036e06 fcoll/two_phase: adjust aggregator selection to new mapby flag on MPI_COMM_WORLD
adjust how the aggregator nodes are selected depending on whether processes
have been mapped by node or anything else.

Signed-off-by: Edgar Gabriel <gabriel@cs.uh.edu>
2017-08-15 09:50:41 -05:00
Edgar Gabriel
92eff9050c communicator/comm_init.c: add a new flag indicating binding policy
Check for the binding policy used. We are only interested in
whether mapby-node has been set right now (could be extended later)
and only on MPI_COMM_WORLD, since for all other sub-communicators
it is virtually impossible to identify their layout across nodes
in the most generic sense. This is used by OMPIO for deciding which
ranks to use for aggregators

Signed-off-by: Edgar Gabriel <gabriel@cs.uh.edu>
2017-08-15 09:50:41 -05:00
Edgar Gabriel
b3f59c76e1 io/ompio: new simple aggr. selection algorithm
add a new aggregator selection algorithm based on the performance
model described in:

Shweta Jha, Edgar Gabriel,
'Performance Models for Communication in Collective I/O Operations'
Proceedings of the 17th IEEE/ACM Symposium
on Cluster, Cloud and Grid Computing, Workshop on Theoretical
Approaches to Performance Evaluation, Modeling and Simulation, 2017.

Signed-off-by: Edgar Gabriel <gabriel@cs.uh.edu>
2017-08-15 09:50:41 -05:00
Jeff Squyres
791bcee6c0 ompi/fortran: remove proof-of-concept mpi_f08 module
This module was always intended to be a proof of concept, and was far
from complete.  If/when someone implemented F08 descriptor support for
the mpi_f08 module, this commit can either be restored or used as
reference material.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-08-10 06:19:17 -07:00
Gilles Gouaillardet
dfe7b2be3f fortran/use-mpi-f08-desc: add a missing include file
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-08-09 13:19:22 +09:00
Gilles Gouaillardet
2c71c27882 fortran2008: fix mpiext example
in order to solve an egg and the chicken problem,  in which mpiext need mpi-f08-types.mod
and/but use-mpi-f08[-desc] needs mpiext, add an extra step
- build fortran 2008 modules only
- build fortran 2008 mpi extensions
- and then build fortran 2008 bindings

Fixes open-mpi/ompi#3605

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-08-09 13:19:22 +09:00
bosilca
9b43de112c Merge pull request #4014 from bosilca/topic/treematch
Topic/treematch
2017-08-08 11:28:22 -04:00
Nathan Hjelm
76320a8ba5 opal: rename opal_atomic_init to opal_atomic_lock_init
This function is used to initalize and opal atomic lock. The old name
was confusing.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-08-07 14:15:11 -06:00
Joshua Ladd
c27beea3a1 Merge pull request #3962 from karasevb/ucx_detect
configure: detect UCX support by default
2017-08-03 16:33:57 -04:00
Nathan Hjelm
29b059e4eb Merge pull request #3971 from plesn/yield_srun
fix srun latency, change default yield_when_idle=0
2017-08-03 07:49:00 -06:00
George Bosilca
3d27e0d3a4
Add support for hwloc 2.0 API.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-03 02:47:04 -04:00
Guillaume Mercier
569239ec44
Check if topo weighted in case of partially distrib case
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-03 00:47:46 -04:00
George Bosilca
1d7cca75a1
Fix a typo in the copyright.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-03 00:47:10 -04:00
George Bosilca
e4db9e574f
Fix all warnings.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-03 00:47:02 -04:00
George Bosilca
c2927d7e91
Update to the latest version provided by Guillaume.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-03 00:46:48 -04:00
George Bosilca
6c8ea09cc5
Use OPAL random generator.
This fix is related to issue #1877, and prevents the OMPI library from
messing the user level random values.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-03 00:46:37 -04:00
George Bosilca
5542559130
Cleaning and optimizations.
Including variable renaming and loop merging.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-03 00:46:28 -04:00
George Bosilca
bc634dbcb0
Make sure the gather is called in all cases, and not
simply based on some local state. This is the second
part of the patch proposed for open-mpi/ompi#1183.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-08-03 00:46:17 -04:00
Brian Barrett
1ec3fd38be Revert "Topic/treematch" 2017-08-02 14:40:55 -07:00
bosilca
d6048af915 Merge pull request #3960 from bosilca/topic/treematch
Update OMPI support for topologies and reordering.
2017-08-02 12:47:23 -04:00
Ralph Castain
f39ce67982 Merge pull request #3951 from rhc54/topic/hwloc2
Update to hwloc 2.0.0a
2017-08-01 15:18:31 -06:00
KAWASHIMA Takahiro
3eac4b0c9a communicator: Refine ompi_comm_set error check
The `ompi_comm_set` function never sets `NULL` to its first argument
`ncomm`.  So `NULL` check is unnecessary in its callers. Furthermore,
`NULL` check may obscure a real return code when an error occurs
if the variable is initialized to a `NULL` value.

Also, `NULL` check is added in the `ompi_comm_set` function to
avoid segmentation fault in an out-of-memory condition.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-07-31 20:26:51 +09:00
KAWASHIMA Takahiro
ebc4eb347c Merge pull request #3701 from kawashima-fj/pr/non-pml-persistent
ompi/request: Support non-PML persistent requests
2017-07-31 02:36:17 -05:00
Edgar Gabriel
d93dae326e Merge pull request #3959 from edgargabriel/topic/performance-fixes
Topic/performance fixes
2017-07-27 09:51:57 -05:00
Piotr Lesnicki
3fa7aabf89 fix srun latency, change default yield_when_idle=0
This changes the default to 0, to avoid yields during progress in srun.

In mpirun, ompi_mpi_yield_when_idle is set to 1 if oversubscribed
otherwise 0. But the default is 1 though, and it is used in srun.
Now srun and mpirun have the same latency in non-oversubscribed cases.

Signed-off-by: Piotr Lesnicki <piotr.lesnicki@atos.net>
2017-07-27 09:41:48 +02:00
Guillaume Mercier
a66dc811b2
Check if topo weighted in case of partially distrib case 2017-07-26 11:54:24 -04:00
George Bosilca
8a7f0baee0
Fix call to opal_hwloc_base_get_topology.
Make sure the HWLOC topology is available as early as possible, so that
we can fail graciously.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-07-26 11:54:24 -04:00
George Bosilca
6061454055
Fix a typo in the copyright.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-07-26 11:54:24 -04:00
George Bosilca
911850d82e
Fix all warnings.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-07-26 11:54:24 -04:00
George Bosilca
2c00c4209a
Update to the latest version provided by Guillaume.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-07-26 11:54:23 -04:00
George Bosilca
fc21ffadc9
Cleaning and optimizations.
Including variable renaming and loop merging.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-07-26 11:54:23 -04:00
George Bosilca
081f9bc8db
Use OPAL random generator.
This fix is related to issue #1877, and prevents the OMPI library from
messing the user level random values.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-07-26 11:54:23 -04:00
George Bosilca
fbe6c22b90
Make sure the gather is called in all cases, and not
simply based on some local state. This is the second
part of the patch proposed for open-mpi/ompi#1183.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-07-26 11:52:47 -04:00
Ralph Castain
7a83fdb9bb Update to hwloc 2.0.0a with shmem support.
Update to support passing of HWLOC shmem topology to client procs
Update use of distance API per @bgoglin
Have the openib component lookup its object in the distance matrix
Bring usnic up-to-date
Restore binding for hwloc2

Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-07-25 20:26:22 -07:00
Boris Karasev
d917d54ddc configure: detect UCX support by default
Adds detecting UCX from following paths: "/usr /usr/local /opt/ucx"

Signed-off-by: Boris Karasev <karasev.b@gmail.com>
2017-07-25 23:48:49 +03:00
Joshua Ladd
8f5cb4c459 Merge pull request #3690 from xinzhao3/topic/ompi-osc-ucx
Topic/ompi-osc-ucx: Add ucx implementation for ompi osc
2017-07-25 16:26:45 -04:00
Edgar Gabriel
ca1462a889 common/ompio: adjust location of fcoll_base_file_select
adjust the location on where the fcoll_base_file_select function is
colled to ensure that all fs level parameters are correctly set.

io/ompio: minor fixes to initialization of the stripe_size and an if statement in the
simple_grouping option.

Signed-off-by: Edgar Gabriel <gabriel@cs.uh.edu>
2017-07-25 10:43:38 -05:00
Edgar Gabriel
450ccd439b fcoll/base: adjust selection table
adjust the fcoll selection table to achieve the following:
 - two_phase should not advertise itself on lustre file systems
 - two_phase should advertise itself on sequential file systems (stripe_size == 0 )
 - priority for dynamic, static and individual is reduced. This will lead to
   two_phase being selected in scenarios where two or more components indicate
   willingness to run.

Signed-off-by: Edgar Gabriel <gabriel@cs.uh.edu>
2017-07-25 10:37:22 -05:00
Gilles Gouaillardet
e054f870a8 Merge pull request #3901 from Zzzoom/nbc_remove_progress_lock
coll/libnbc: demote progress_lock to regular flag
2017-07-25 09:24:08 +09:00
Carlos Bederián
1767b218fb coll/libnbc: demote progress_lock to regular flag
Signed-off-by: Carlos Bederián <bc@famaf.unc.edu.ar>
2017-07-24 20:19:55 -03:00
Gilles Gouaillardet
60aa9cfcb6 hwloc: add support for hwloc v2 API
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-07-20 17:39:44 +09:00
Gilles Gouaillardet
9f29f3bff4 hwloc: since WHOLE_SYSTEM is no more used, remove useless
checks related to offline and disallowed elements

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-07-20 17:39:21 +09:00
Xin Zhao
2aa5292dbf Add UCX component for ompi/mca/osc for MPI one-sided communication.
Signed-off-by: Xin Zhao <xinz@mellanox.com>
2017-07-19 19:45:40 +03:00
Edgar Gabriel
b363a0f4db Merge pull request #3908 from edgargabriel/pr/lustre-header-update
fs/lustre: update lustre header file used in the component
2017-07-18 09:57:11 -05:00
Gilles Gouaillardet
1b46fe2d9a pml/ob1: fix mca_pml_ob1_progress_needed usage
correctly use OPAL_ATOMIC_ADD32() that returns the *new* value
and *not* the previous one.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-07-18 09:30:57 +09:00
Nathan Hjelm
5a5edfce88 Merge pull request #3876 from hppritcha/topic/add_psm2_get_stats
mtl/psm2: add pvar support for PSM2 MQ stats
2017-07-17 16:02:51 -05:00
Edgar Gabriel
13b14f5efe Merge pull request #3906 from edgargabriel/pr/lazy_open_fix
common/ompio: fix the lazy_open flag
2017-07-17 10:58:13 -05:00
Edgar Gabriel
931c2021c7 Merge pull request #3905 from edgargabriel/pr/fix-grouping-options
io/ompio: fix grouping option
2017-07-17 10:55:57 -05:00
Edgar Gabriel
bc8f642211 fs/lustre: update lustre header file used in the component
liblustreapi.h is at this point deprecated. Switch to lustreapi.h instead

fixes issue #3223

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-07-17 10:33:52 -05:00
Nathan Hjelm
9b702fb9bd ompi: clean up topo helper functions
This commit removes the communicator topo helper functions in favor
of functions in mca/topo/base.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-07-17 10:30:42 -05:00
Edgar Gabriel
8e17827a13 common/ompio: fix the lazy_open flag
fixes an erroneous error code being returned when activating
the mca_io_ompio_sharedfp_lazy_open flag with MPI_MODE_APPEND.

fixes issue #3904

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-07-17 09:44:34 -05:00
Edgar Gabriel
4bdddfb74b io/ompio: fix grouping option
changing the value of mca_io_ompio_grouping_option lead to a segfault due to
a double-free problem. Remove the erroneous free statements that have been introduced
and add a note ensuring that we are not re-adding them back at that spot.

fixes issue #3903

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-07-17 09:38:10 -05:00
Yossi Itigin
b0692c6836 Merge pull request #3833 from yosefe/topic/pml-yalla-dt-size-fix
pml/yalla: fix getting size of a continuous type.
2017-07-16 11:19:16 +03:00
Howard Pritchard
701a1d0218 mtl/psm2: add pvar support for PSM2 MQ stats
Add pvars for PSM2 MQ stats to help in analyzing performance
of Omnipath.

Tested (modestly) using modified OSU pt2pt benchmarks.

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
2017-07-14 10:31:35 -06:00
Ryan Grant
0ce8590e7c Merge pull request #3837 from tkordenbrock/topic/master/get.retry.timeout
master: mtl-portals4: add timeout to rendezvous get fragments
2017-07-13 09:59:54 -06:00
Nathan Hjelm
6fb81f20e4 mtl/psm2: create mca variables to shadow PSM2 environment variables
This commit enables MCA support for the following PSM2 environment
variables: PSM2_DEVICES, PSM2_MEMORY, PSM2_MQ_SENDREQS_MAX,
PSM2_MQ_RECVREQS_MAX, PSM2_MQ_RNDV_HFI_THRESH,
PSM2_MQ_RNDV_SHM_THRESH, PSM2_RCVTHREAD, PSM2_SHAREDCONTEXTS,
PSM2_SHAREDCONTEXTS_MAX, and PSM2_TRACEMASK. These variable can be set
by MCA if they are not already set in the environment.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-07-13 09:48:46 -06:00
Christoph Niethammer
e1fc6ae304 Change filename for shared_fm file to include comm cid instead of masterjobid.
Signed-off-by: Christoph Niethammer <niethammer@hlrs.de>
2017-07-13 01:38:40 +02:00
Jeff Squyres
ccf17808b6 Merge pull request #3258 from markalle/pr/symbol_name_pollution
symbol name pollution
2017-07-12 16:19:25 -05:00
Nathan Hjelm
3c0e94afab mpi/neighbor_allgatherv: fix copy&paste error and add helpers
This commit adds a helper function to get the inbound and outbound
neighbor count and updates the neighbor_allgatherv bindings to use the
correct count when checking the input parameters.

Fixes #2324

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-07-12 12:54:24 -06:00
Nathan Hjelm
bf1c863b96 osc/pt2pt: make progress in flush*_local
There is no reason not to progress OSC during the MPI_Win_flush_local
and MPI_Win_flush_all_local calls. This fixes #3750.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-07-12 09:09:58 -06:00
Gilles Gouaillardet
7a866f754c topo/treematch: fix topo_treematch_distgraph_create
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-07-12 10:27:45 +09:00
Gilles Gouaillardet
72cfbb665c ompi/attributes: revamp attribute handling.
we now have 12 cases to deal (4 writers and 3 readers) :

1. C `void*` is written into the attribute value, and the value is read into a C `void*` (unity)
2. C `void*` is written, Fortran `INTEGER` is read
3. C `void*` is written, Fortran `INTEGER(KIND=MPI_ADDRESS_KIND)` is read
4. Fortran `INTEGER` is written, C `void*` is read
5. Fortran `INTEGER` is written, Fortran `INTEGER` is read (unity)
6. Fortran `INTEGER` is written, Fortran `INTEGER(KIND=MPI_ADDRESS_KIND)` is read
7. Fortran `INTEGER(KIND=MPI_ADDRESS_KIND)` is written, C `void*` is read
8. Fortran `INTEGER(KIND=MPI_ADDRESS_KIND)` is written, Fortran `INTEGER` is read
9. Fortran `INTEGER(KIND=MPI_ADDRESS_KIND)` is written, Fortran `INTEGER(KIND=MPI_ADDRESS_KIND)` is read (unity)
10. Intrinsic is written, C `void*` is read
11. Intrinsic is written, Fortran `INTEGER` is read
12. Intrinsic is written, Fortran `INTEGER(KIND=MPI_ADDRESS_KIND)` is read

MPI-2 Fortran "integer representation" has type `INTEGER(KIND=MPI_ADDRESS_KIND)` as clarified
at https://github.com/mpiwg-rma/rma-issues/issues/1

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-07-12 10:27:45 +09:00
Nathan Hjelm
e73ab93ebf pml/ob1: do not access fragment after calling btl rget
This commit fixes a bug that occurs when the btl callback happens before
the rget returns. In this case the fragment has been returned and is no
longer valid. This commit saves the size before calling rget. This is
valid since the BTL is not allowed to change the read size.

Fixes #3821

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-07-11 15:59:40 -06:00
Joshua Hursey
c81795cbda Revert "Fix MPI_SIZEOF for gfortran 4.8"
This reverts commit 5de3d5dde6.

 Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
2017-07-11 11:09:17 -05:00
Mark Allen
552216f9ba scripted symbol name change (ompi_ prefix)
Passed the below set of symbols into a script that added ompi_ to them all.

Note that if processing a symbol named "foo" the script turns
    foo  into  ompi_foo
but doesn't turn
    foobar  into  ompi_foobar

But beyond that the script is blind to C syntax, so it hits strings and
comments etc as well as vars/functions.

    coll_base_comm_get_reqs
    comm_allgather_pml
    comm_allreduce_pml
    comm_bcast_pml
    fcoll_base_coll_allgather_array
    fcoll_base_coll_allgatherv_array
    fcoll_base_coll_bcast_array
    fcoll_base_coll_gather_array
    fcoll_base_coll_gatherv_array
    fcoll_base_coll_scatterv_array
    fcoll_base_sort_iovec
    mpit_big_lock
    mpit_init_count
    mpit_lock
    mpit_unlock
    netpatterns_base_err
    netpatterns_base_verbose
    netpatterns_cleanup_narray_knomial_tree
    netpatterns_cleanup_recursive_doubling_tree_node
    netpatterns_cleanup_recursive_knomial_allgather_tree_node
    netpatterns_cleanup_recursive_knomial_tree_node
    netpatterns_init
    netpatterns_register_mca_params
    netpatterns_setup_multinomial_tree
    netpatterns_setup_narray_knomial_tree
    netpatterns_setup_narray_tree
    netpatterns_setup_narray_tree_contigous_ranks
    netpatterns_setup_recursive_doubling_n_tree_node
    netpatterns_setup_recursive_doubling_tree_node
    netpatterns_setup_recursive_knomial_allgather_tree_node
    netpatterns_setup_recursive_knomial_tree_node
    pml_v_output_close
    pml_v_output_open
    intercept_extra_state_t
    odls_base_default_wait_local_proc
    _event_debug_mode_on
    _evthread_cond_fns
    _evthread_id_fn
    _evthread_lock_debugging_enabled
    _evthread_lock_fns
    cmd_line_option_t
    cmd_line_param_t
    crs_base_self_checkpoint_fn
    crs_base_self_continue_fn
    crs_base_self_restart_fn
    event_enable_debug_output
    event_global_current_base_
    event_module_include
    eventops
    sync_wait_mt
    trigger_user_inc_callback
    var_type_names
    var_type_sizes

Signed-off-by: Mark Allen <markalle@us.ibm.com>
2017-07-11 02:13:23 -04:00
Todd Kordenbrock
5ecd905358 mtl/portals4: move opal_timer_base_get_usec() out of the fast path
Rearrange the receive frag timeout logic to avoid calling
opal_timer_base_get_usec() in read_msg().  Instead set it at the first
retry.

Signed-off-by: Todd Kordenbrock <thkgcode@gmail.com>
2017-07-09 22:12:45 -05:00
Todd Kordenbrock
37766d770d mtl/portals4: if frag retry fails, then fail the entire receive
If the a frag cannot be retried because the ni_fail_type is other than
PTL_NI_DROPPED, then set the return type and jump to callback_error.
This sets MPI_ERROR and completes the receive.

Signed-off-by: Todd Kordenbrock <thkgcode@gmail.com>
2017-07-09 22:12:31 -05:00
Piotr Lesnicki
99453e6b10 mtl/portals4: get retransmission REPLY code
Signed-off-by: Todd Kordenbrock <thkgcode@gmail.com>
2017-07-09 22:12:25 -05:00
Piotr Lesnicki
06b15cebbf mtl/portals4: add timeout to get retransmit
Signed-off-by: Todd Kordenbrock <thkgcode@gmail.com>
2017-07-09 22:12:08 -05:00
Yossi Itigin
0522179efc pml/yalla: use opal_datatype_span() to get config type length.
Signed-off-by: Yossi Itigin <yosefe@mellanox.com>
2017-07-10 01:25:42 +03:00
Yossi Itigin
e94c6b16f0 pml/yalla: fix getting size of a continuous type.
pull request #3765 introduced a bug where the extent of a type is used
instead of its size.

Signed-off-by: Yossi Itigin <yosefe@mellanox.com>
2017-07-07 19:03:54 +03:00
Chris Ward
5de3d5dde6 Fix MPI_SIZEOF for gfortran 4.8
Add copyrights.

Revise the README to take out the 'most notably' statement about GNU Fortran 4.8

Signed-off-by: Chris Ward <tjcw@uk.ibm.com>
2017-07-07 13:47:35 +01:00
Gilles Gouaillardet
fc11c37223 Merge pull request #3646 from ggouaillardet/spacc-fix-coverity-warnings
coll/spacc: misc fixes
2017-07-06 11:39:14 +09:00
Mikhail Kurnosov
44acc92104 Fix buffer overflow
Add check for bounds of sindex[] and rindex[].

Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2017-07-06 10:49:08 +09:00
Gilles Gouaillardet
5fceca235b coll/spacc: silence more coverity warnings in mca_coll_spacc_allreduce_intra_redscat_allgather()
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-07-06 10:49:08 +09:00
Mikhail Kurnosov
2f0f476642 Silence spacc coverity warnings
1. Add assert for opal_hibit return value: comm_size is always > 1.
2. Modified verbose output (dead-code warning).

Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
2017-07-06 10:49:08 +09:00
Ralph Castain
31130a4bee Replace syntax with something less strictly C99
Fixes #3809

Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-07-05 16:54:36 -07:00
Gilles Gouaillardet
d1c5955b73 coll/base: optimize handling of zero-byte datatypes in mca_coll_base_alltoallv_intra_basic_inplace()
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-06-30 09:47:08 +09:00
Ralph Castain
bd4a6fee22 Attempt to detect when we are direct-launched without the necessary PMI support, and thus are incorrectly identified as being "singleton". Advise the user on the required PMI(x) support and error out.
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-06-29 15:26:53 -07:00
Gilles Gouaillardet
7e5e5fe887 Merge pull request #3719 from ggouaillardet/topic/libnbc_revamp
coll/libnbc: revisit NBC_Handle usage
2017-06-29 11:13:58 +09:00
Nathan Hjelm
022c658bbf osc/rdma: rework locking code to improve behavior of unlock
This commit changes the locking code to allow the lock release to be
non-blocking. This helps with releasing the accumulate lock which may
occur in a BTL callback.

Fixes #3616

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-06-27 15:29:51 -06:00
George Bosilca
f8ffec926e
Protect the monitoring infrastructure initialization. 2017-06-27 18:35:24 +02:00
Clément FOYER
c885ee3f3c Fix Coverity warning CID 1413323 (#3764)
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>
2017-06-27 12:39:31 +02:00
bosilca
d55b666834 Topic/monitoring (#3109)
Add a monitoring PML, OSC and IO. They track all data exchanges between processes,
with capability to include or exclude collective traffic. The monitoring infrastructure is
driven using MPI_T, and can be tuned of and on any time o any communicators/files/windows.
Documentations and examples have been added, as well as a shared library that can be
used with LD_PRELOAD and that allows the monitoring of any application.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>


* add ability to querry pml monitorinting results with MPI Tools interface
using performance variables "pml_monitoring_messages_count" and
"pml_monitoring_messages_size"

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>

* Fix a convertion problem and add a comment about the lack of component
retain in the new component infrastructure.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>

* Allow the pvar to be written by invoking the associated callback.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>

* Various fixes for the monitoring.
Allocate all counting arrays in a single allocation
Don't delay the initialization (do it at the first add_proc as we
know the number of processes in MPI_COMM_WORLD)

Add a choice: with or without MPI_T (default).

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>

* Cleanup for the monitoring module.
Fixed few bugs, and reshape the operations to prepare for
global or communicator-based monitoring. Start integrating
support for MPI_T as well as MCA monitoring.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>

* Adding documentation about how to use pml_monitoring component.

Document present the use with and without MPI_T.
May not reflect exactly how it works right now, but should reflects
how it should work in the end.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Change rank into MPI_COMM_WORLD and size(MPI_COMM_WORLD) to global variables in pml_monitoring.c.
Change mca_pml_monitoring_flush() signature so we don't need the size and rank parameters.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>

* Improve monitoring support (including integration with MPI_T)

Use mca_pml_monitoring_enable to check status state. Set mca_pml_monitoring_current_filename iif parameter is set
Allow 3 modes for pml_monitoring_enable_output: - 1 : stdout; - 2 : stderr; - 3 : filename
Fix test : 1 for differenciated messages, >1 for not differenciated. Fix output.
Add documentation for pml_monitoring_enable_output parameter. Remove useless parameter in example
Set filename only if using mpi tools
Adding missing parameters for fprintf in monitoring_flush (for output in std's cases)
Fix expected output/results for example header
Fix exemple when using MPI_Tools : a null-pointer can't be passed directly. It needs to be a pointer to a null-pointer
Base whether to output or not on message count, in order to print something if only empty messages are exchanged
Add a new example on how to access performance variables from within the code
Allocate arrays regarding value returned by binding

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add overhead benchmark, with script to use data and create graphs out of the results
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix segfault error at end when not loading pml
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Start create common monitoring module. Factorise version numbering
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix microbenchmarks script
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Improve readability of code

NULL can't be passed as a PVAR parameter value. It must be a pointer to NULL or an empty string.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add osc monitoring component

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add error checking if running out of memory in osc_monitoring

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Resolve brutal segfault when double freeing filename
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Moving to ompi/mca/common the proper parts of the monitoring system
Using common functions instead of pml specific one. Removing pml ones.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add calls to record monitored data from osc. Use common function to translate ranks.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix test_overhead benchmark script distribution

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix linking library with mca/common

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add passive operations in monitoring_test

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix from rank calculation. Add more detailed error messages

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix alignments. Fix common_monitoring_get_world_rank function. Remove useless trailing new lines

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix osc_monitoring mget_message_count function call

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Change common_monitoring function names to respect the naming convention. Move to common_finalize the common parts of finalization. Add some comments.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add monitoring common output system

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add error message when trying to flush to a file, and open fails. Remove erroneous info message when flushing wereas the monitoring is already disabled.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Consistent output file name (with and without MPI_T).

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Always output to a file when flushing at pvar_stop(flush).

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Update the monitoring documentation.
Complete informations from HowTo. Fix a few mistake and typos.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Use the world_rank for printf's.
Fix name generation for output files when using MPI_T. Minor changes in benchmarks starting script

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Clean potential previous runs, but keep the results at the end in order to potentially reprocess the data. Add comments.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add security check for unique initialization for osc monitoring

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Clean the amout of symbols available outside mca/common/monitoring

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Remove use of __sync_* built-ins. Use opal_atomic_* instead.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Allocate the hashtable on common/monitoring component initialization. Define symbols to set the values for error/warning/info verbose output. Use opal_atomic instead of built-in function in osc/monitoring template initialization.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Deleting now useless file : moved to common/monitoring

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add histogram ditribution of message sizes

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add histogram array of 2-based log of message sizes. Use simple call to reset/allocate arrays in common_monitoring.c

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add informations in dumping file. Separate per category (pt2pt/osc/coll (to come)) monitored data

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add coll component for collectives communications monitoring

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix warning messages : use c_name as the magic id is not always defined. Moreover, there was a % missing. Add call to release underlying modules. Add debug info messages. Add warning which may lead to further analysis.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix log10_2 constant initialization. Fix index calculation for histogram array.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add debug info messages to follow more easily initialization steps.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Group all the var/pvar definitions to common_monitoring. Separate initial filename from the current on, to ease its lifetime management. Add verifications to ensure common is initialized once only. Move state variable management to common_monitoring.
monitoring_filter only indicates if filtering is activated.
Fix out of range access in histogram.
List is not used with the struct mca_monitoring_coll_data_t, so heritate only from opal_object_t.
Remove useless dead code.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix invalid memory allocation. Initialize initial_filename to empty string to avoid invalid read in mca_base_var_register.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Don't install the test scripts.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix missing procs in hashtable. Cache coll monitoring data.
    * Add MCA_PML_BASE_FLAG_REQUIRE_WORLD flag to the PML layer.
    * Cache monitoring data relative to collectives operations on creation.
    * Remove double caching.
    * Use same proc name definition for hash table when inserting and
      when retrieving.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Use intermediate variable to avoid invalid write while retrieving ranks in hashtable.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add missing release of the last element in flush_all. Add release of the hashtable in finalize.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Use a linked list instead of a hashtable to keep tracks of communicator data. Add release of the structure at finalize time.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Set world_rank from hashtable only if found

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Use predefined symbol from opal system to print int

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Move collective monitoring data to a hashtable. Add pvar to access the monitoring_coll_data. Move functions header to a private file only to be used in ompi/mca/common/monitoring

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix pvar registration. Use OMPI_ERROR isntead of -1 as returned error value. Fix releasing of coll_data_t objects. Affect value only if data is found in the hashtable.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add automated check (with MPI_Tools) of monitoring.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix procs list caching in common_monitoring_coll_data_t

    * Fix monitoring_coll_data type definition.
    * Use size(COMM_WORLD)-1 to determine max number of digits.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add linking to Fortran applications for LD_PRELOAD usage of monitoring_prof

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add PVAR's handles. Clean up code (visibility, add comments...). Start updating the documentation

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix coll operations monitoring. Update check_monitoring accordingly to the added pvar. Fix monitoring array allocation.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Documentation update.
Update and then move the latex and README documentation to a more logical place

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Aggregate monitoring COLL data to the generated matrix. Update documentation accordingly.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix monitoring_prof (bad variable.vector used, and wrong array in PMPI_Gather).

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add reduce_scatter and reduce_scatter_block monitoring. Reduce memory footprint of monitoring_prof. Unify OSC related outputs.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add the use of a machine file for overhead benchmark

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Check for out-of-bound write in histogram

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Fix common_monitoring_cache object init for MPI_COMM_WORLD

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add RDMA benchmarks to test_overhead
Add error file output. Add MPI_Put and MPI_Get results analysis. Add overhead computation for complete sending (pingpong / 2).

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add computation of average and median of overheads. Add comments and copyrigths to the test_overhead script

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add technical documentation

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Adapt to the new definition of communicators

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Update expected output in test/monitoring/monitoring_test.c

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add dumping histogram in edge case

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Adding a reduce(pml_monitoring_messages_count, MPI_MAX) example

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add consistency in header inclusion.
Include ompi/mpi/fortran/mpif-h/bindings.h only if needed.
Add sanity check before emptying hashtable.
Fix typos in documentation.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* misc monitoring fixes

* test/monitoring: fix test when weak symbols are not available
* monitoring: fix a typo and add a missing file in Makefile.am
and have monitoring_common.h and monitoring_common_coll.h included in the distro
* test/monitoring: cleanup all tests and make distclean a happy panda
* test/monitoring: use gettimeofday() if clock_gettime() is unavailable
* monitoring: silence misc warnings (#3)

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

* Cleanups.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>

* Changing int64_t to size_t.
Keep the size_t used accross all monitoring components.
Adapt the documentation.
Remove useless MPI_Request and MPI_Status from monitoring_test.c.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add parameter for RMA test case

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Clean the maximum bound computation for proc list dump.
Use ptrdiff_t instead of OPAL_PTRDIFF_TYPE to reflect the changes from commit fa5cd0dbe5.

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add communicator-specific monitored collective data reset

Signed-off-by: Clement Foyer <clement.foyer@inria.fr>

* Add monitoring scripts to the 'make dist'
Also install them in the build and the install directories.

Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
2017-06-26 18:21:39 +02:00
Nathan Hjelm
31ab83362a osc/rdma: cleanup local peer setup and fix a bug
The data endpoint was not being set correctly for local peers in some
cases. This commit fixes the bug and cleans the associated code to
simplify the logic.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-06-22 13:28:45 -06:00
Christoph Niethammer
7f1347677d Create file for file backed shared memory in process job session dir.
Prevents file collisions and can also be cleaned by orte-clean properly.

Signed-off-by: Christoph Niethammer <niethammer@hlrs.de>
2017-06-22 08:25:34 +02:00
Ralph Castain
814e858082 Merge pull request #3696 from rhc54/topic/pmix3
Update PMIx and integration glue
2017-06-20 10:38:29 -07:00
Ralph Castain
952726c121 Update to latest PMIx master - equivalent to 2.0rc2. Update the thread support in the opal/pmix framework to protect the framework-level structures.
This now passes the loop test, and so we believe it resolves the random hangs in finalize.

Changes in PMIx master that are included here:

* Fixed a bug in the PMIx_Get logic
* Fixed self-notification procedure
* Made pmix_output functions thread safe
* Fixed a number of thread safety issues
* Updated configury to use 'uname -n' when hostname is unavailable

Work on cleaning up the event handler thread safety problem
Rarely used functions, but protect them anyway
Fix the last part of the intercomm problem
Ensure we don't cover any PMIx calls with the framework-level lock.
Protect against NULL argv comm_spawn

Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-06-20 09:02:15 -07:00
George Bosilca
1f291c8728
Add the fragment to the unexpected frags only after extracting the
pml_proc.
2017-06-20 16:03:52 +02:00
Gilles Gouaillardet
9ba85b85e1 coll/libnbc: revisit NBC_Handle usage
make NBC_Handle (almost) an internal structure created
by NBC_Schedule_request()
use a local variable instead of what was previously handle->tmpbuf

Refs open-mpi/ompi#3487

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-06-20 17:24:16 +09:00
Gilles Gouaillardet
68ac95003f coll/base: fix zero size datatype handling in mca_coll_base_alltoallv_intra_basic_inplace()
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-06-20 14:36:35 +09:00
Nathan Hjelm
0c8c7e50d0 Merge pull request #3682 from hjelmn/comm_assertions
ompi: add support for new communicator info assertions
2017-06-19 09:49:59 -06:00
Edgar Gabriel
70107b3e52 Merge pull request #3703 from edgargabriel/pr/cart-comm-file-open-fix
Pr/cart comm file open fix
2017-06-15 15:03:38 -05:00
Edgar Gabriel
3b0b8fa12c io/ompio: update cartesian based grouping strategy
update the cartesian communicator based grouping strategy to match the other
algorithms used in the aggregator selection process.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-06-15 14:05:54 -05:00
Edgar Gabriel
bd6b430798 common/ompio: remove function call to cart_based_grouping
the cart_based_grouping aggregator strategy was not correctly updated
during the last major rewrite of the aggregator selection algorithm.
It is also not supposed to be called from file_open (but from
file_set_view).

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
2017-06-15 14:04:03 -05:00
George Bosilca
e9d533e62e
Fix warnings from non-debug mode.
Thanks Ralph for the report.
2017-06-13 16:57:42 -04:00
Gilles Gouaillardet
72c7329462 configury: use 'uname -n' when 'hostname' is not available
the 'hostname' command might not be available on some platforms
such as Fedora Core 26, so mimick config/libtool.m4 and fallback
to 'uname -n' if needed

Refs. #3680

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-06-12 15:04:32 +09:00
KAWASHIMA Takahiro
b5b6b22848 Merge pull request #3678 from kawashima-fj/pr/signal-abort-delay
Apply `opal_abort_delay` to the OPAL signal handler
2017-06-12 10:35:11 +09:00
Joshua Hursey
80a91dc244 io/romio314: Add work around support for missing MPI_File ops
* Add work around support for the following missing ops in ROMIO 3.1.4
    - `MPI_File_iread_at_all`
    - `MPI_File_iwrite_at_all`
    - `MPI_File_iread_all`
    - `MPI_File_iwrite_all`

Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
2017-06-09 14:42:59 -05:00
Joshua Hursey
29609631a2 mpi/c: Protect some IO functions not widely implemented
* Protects us from segv when ROMIO 314 is selected and one of the
   following operations is called:
   - MPI_File_iread_at_all
   - MPI_File_iwrite_at_all
   - MPI_File_iread_all
   - MPI_File_iwrite_all

Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
2017-06-09 11:42:26 -05:00
Nathan Hjelm
db2204f2f3 ompi: add support for new communicator info assertions
This commit adds code to allow support for the info assertions added
by mpi-forum/mpi-issues#11. The assertions added are:
mpi_assert_no_any_tag, mpi_assert_no_any_source,
mpi_assert_exact_length, and mpi_assert_allow_overtaking.

This commit also adds support for the mpi_assert_no_any_source and
mpi_assert_allow_overtaking info keys to the ob1 pml.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-06-08 15:52:12 -06:00
KAWASHIMA Takahiro
362445d486 Use same prefix format for [host:pid]
Hostname and PID are output as a message prefix in many places in
our code. Their printf-formats were either `[%s:%d]` or `[%s:%05d]`.
This commit changes `[%s:%d]` to `[%s:%05d]`. The latter was more
widely used in our code (including OPAL output system and the signal
handler).

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-06-08 19:35:03 +09:00
KAWASHIMA Takahiro
6b91eddc8b Apply opal_abort_delay to the signal handler
This commit expands the effect of the MCA parameter `opal_abort_delay`
to the OPAL signal handler. This allows attaching of a debugger on
segmentation fault etc. before quitting the job.

The sleep code is moved to the `opal_delay_abort` function from the
`ompi_mpi_abort` and `oshmem_shmem_abort` functions for code cleanup.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-06-08 19:34:48 +09:00
Mark Allen
aeb2c02d2f Type_create_darray with mix of BLOCK/CYCLIC
Example (using MPI_ORDER_C so the below has 6 rows of 4 ints to parcel out)
    size = 4;
    rank = 0;
    ndims=2;
    gsizes[0] = 6;
    gsizes[1] = 4;
    distribs[0] = MPI_DISTRIBUTE_CYCLIC;
    distribs[1] = MPI_DISTRIBUTE_BLOCK;
    dargs[0] = 2;
    dargs[1] = 2;
    psizes[0] = 2;
    psizes[1] = 2;
    MPI_Type_create_darray(size, rank, ndims,
        gsizes, distribs, dargs, psizes,
        MPI_ORDER_C, MPI_INT, &mydt);

Expectation for the layout:
   inner dimension (1) is
       4 items (ints) distributed block over 2 ranks with 2 items each
       eg for rank 0: [ x x . . ]
   outer dimension (0) is:
       6 items (the above [ x x . .]) cyclic over 2 ranks with 2 items each
       eg for rank 0:
           [ x x . . ]    :  offset=0 bytes=8
           [ x x . . ]    :  ofset=16 bytes=8
           [ . . . . ]
           [ . . . . ]
           [ x x . . ]    :  offset=64 bytes=8
           [ x x . . ]    :  offset=80 bytes=8

Or more specifically a stream of ints 0,1,2,3,4,5,6,7 sent into that
type should be
    [ 0 1 . . ]
    [ 2 3 . . ]
    [ . . . . ]
    [ . . . . ]
    [ 4 5 . . ]
    [ 6 7 . . ]
The data was laying out though as
    [ 0 1 2 3 ]
    [ . . . . ]
    [ . . . . ]
    [ . . . . ]
    [ 4 5 6 7 ]
    [ . . . . ]
because the recursive construction inside the block() function (which
creates the smaller row datatype [ x x . . ]) wasn't setting the extent
of that type.

Signed-off-by: Mark Allen <markalle@us.ibm.com>
2017-06-07 16:53:03 -04:00
Jeff Squyres
44aef39b24 Merge pull request #3641 from ggouaillardet/topic/fortran_strings
fortran/base: rename strings.h into fortran_base_strings.h
2017-06-05 15:31:08 -04:00
Ralph Castain
dea9ef2020 Merge pull request #3637 from hjelmn/osc_sm_info_fix
osc/sm: fix SEGV in new info usage
2017-06-05 05:45:21 -07:00
KAWASHIMA Takahiro
0cbdbe32f7 ompi/request: Support non-PML persistent requests
This commit adds the `req_start` member to the `ompi_request_t` struct.
The `MPI_START` and `MPI_STARTALL` routines call this callback function
instead of `MCA_PML_CALL(start(...))`. So components that return
persistent request must set this member to their request objects.

`mca_pml_base_module_t::pml_start` is not deleted because
`MCA_PML_CALL(start(...))` is still used elsewhere across OMPI.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-06-02 13:08:17 +09:00
KAWASHIMA Takahiro
c8d38d31c6 Merge pull request #3618 from kawashima-fj/pr/java-doc-man
java: Detect `javadoc` path and improve `mpijavac` man page
2017-06-02 10:24:05 +09:00
Gilles Gouaillardet
08526e8adc fortran/base: rename strings.h into fortran_base_strings.h
rename ompi/mpi/fortran/base/strings.h so it does not get pulled
when /usr/include/strings.h is expected.

Refs open-mpi/ompi#3639

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-06-02 09:46:20 +09:00
Josh Hursey
1665d771a6 Merge pull request #3635 from wlepera/fix/ibm/155305
MPI_Sendreceive_replace data error with > 2k msg
2017-06-01 14:38:01 -05:00
Jeff Squyres
d520c24f3a predefined MPI object padding: set to fixed number of bytes (#3634)
Convert the predefined MPI object padding to a fixed number of bytes
(vs. a multiple of sizeof(void*)) so that the padding is the same size
between 32 and 64 bit builds.  I.e., we won't have a situation where
we've run out of padding in 32 bit builds but still have more space
available in 64 bit builds.

Fixes #3610

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2017-06-01 15:28:23 -04:00
Nathan Hjelm
d10e6455a0 osc/sm: fix SEGV in new info usage
This commit moves the info subscribe for the blocking_fence to after
the global_state is allocated and moves setting win->w_osc_module to
before the info subscribe for alloc_shared_contig. This fixes a SEGV
caught by MTT.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-06-01 12:32:30 -06:00
William LePera
a7c9c4aef3 MPI_Sendreceive_replace data error with > 2k msg (RTC 155305)
Signed-off-by: William LePera <lepera@us.ibm.com>
2017-06-01 13:08:58 -04:00
Gilles Gouaillardet
5e9be7667b Merge pull request #3600 from ggouaillardet/topic/osc_rdma_get_segment
osc/rdma: fix osc_rdma_get_remote_segment() length parameter
2017-06-01 13:09:14 +09:00
Nathan Hjelm
e1a997c0cb Merge pull request #3593 from hjelmn/bug_3575
osc/rdma: fix typo in ompi_osc_rdma_lock_acquire_exclusive
2017-05-31 08:54:40 -06:00
KAWASHIMA Takahiro
76b1f80664 java: Use correct date/version in mpijava man page
`mpijavac.1` should be generated at `make`-time...

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-05-31 17:24:49 +09:00
KAWASHIMA Takahiro
63f0945dcc java: Detect the path of javadoc in configure
Without this change, the directory of `javadoc` command must be
included in the `PATH` environment variable at `make`-time.
Paths of `javac`, `javah`, and `jar` commands are detected in
`configure`. So the path of `javadoc` also should be detected.

Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
2017-05-31 14:26:14 +09:00
Ralph Castain
ed4078e2dd Protect against the condition where the port string is actually NULL
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-05-28 20:51:09 -07:00
Gilles Gouaillardet
e622ca8c1c osc/rdma: fix osc_rdma_get_remote_segment() length parameter
a buffer defined by (buf, count, dt)
will have data starting at buf+offset and ending len bytes later with
len = opal_datatype_span(&dt.super, count, &offset);

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
2017-05-29 11:08:03 +09:00
Ralph Castain
9f60cd0fe7 Update the connect/accept support so we check to see if we have the proper infrastructure and RTE support, including whether we have ompi-server available if the connect/accept spans multiple applications. Print pretty help messages in all cases where we do not have support
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
2017-05-27 10:47:08 -07:00
Nathan Hjelm
b83c5dbee5 osc/rdma: fix typo in ompi_osc_rdma_lock_acquire_exclusive
Fixes #3575

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-05-26 14:21:08 -06:00
Nathan Hjelm
c7e6294f31 Merge pull request #3589 from hjelmn/cxx_glue
mpi/cxx: remove nonexistent function from cxx glue
2017-05-26 11:24:05 -06:00
Nathan Hjelm
ee9093c373 mpi/cxx: remove nonexistent function from cxx glue
This commit removes a nonexistent function that was causing build
problems under certain environments.

Reference #3442

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2017-05-26 10:40:19 -06:00
Josh Hursey
4bfb0fcddd Merge pull request #3577 from markalle/pr/osc_rdma_rangecheck
fix for buffer length check (rdma osc w/ odd datatypes)
2017-05-26 10:44:33 -05:00