1
1
Граф коммитов

8440 Коммитов

Автор SHA1 Сообщение Дата
yohann
27520b99b8 mtl/ofi: add include/exclude list MCA vars.
mtl_ofi_provider_include (resp. mtl_ofi_provider_exclude) can be used
to specify which provider(s) the OFI MTL can select (resp. ignore).

e.g. --mca mtl_ofi_provider_include "psm,sockets"

By default, mtl_ofi_provider_exclude is set to "sockets,mxm".

This deprecates the old MCA var named "mtl_ofi_provider".
2015-08-12 13:52:04 -07:00
Jeff Squyres
e9b7203ece treematch: ensure hwloc support is enabled
This commit does the following:

* s/ompi_check_treematch/ompi_topo_treematch/ (i.e., abide by the
  prefix rule)
* change the value of ompi_topo_treematch_happy from yes/no to 0/1, so
  that we can use -eq for numerical comparisons (vs. string
  comparisons).  It's the little things in life, no?
* Check the valueo f $OPAL_HAVE_HWLOC to ensure that hwloc support is
  enabled.  If not, disqualify treematch from building.
* Fixes a few places that were underquoted
* Convert from "test ... -a ..." to "test ... && test ..."

Fixes open-mpi/ompi#797
2015-08-12 12:23:12 -07:00
Edgar Gabriel
55f0e1a1f8 fix the lustre compilation problems for older lustre versions. Add the prototype for the static function to avoid a warning message. 2015-08-12 09:45:07 -05:00
Jeff Squyres
2409fa166b get_library_version: checking string constants vs. NULL is dead code
The prior code was checking string constants (which are #defines from
configure) against NULL.  They can never be NULL, so the checks were
overly-defensive.  If the preprocessor macros do not exist, we'll get
a different compiler error.  So remove the dead code.

This fixes CID 72349.
2015-08-12 05:35:12 -07:00
Jeff Squyres
3be125afff op base: whitespace cleanup
No logical code changes.
2015-08-12 05:35:11 -07:00
Jeff Squyres
a2addbafed op base: move return statement to correct level
This fixes CID 71945.
2015-08-12 05:35:11 -07:00
Nathaniel Graham
8f4c16da27 Java null handle bugfix
A helper method in Request.java could cause a crash
if the request array that was passed contained nulls.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-08-10 15:54:36 -06:00
Nathan Hjelm
624a4a0f82 Merge pull request #699 from hjelmn/libnbc_fixes
coll/libnbc: rewrite parts of libnbc
2015-08-10 14:51:42 -06:00
Jeff Squyres
87db836800 Merge pull request #788 from yburette/topic/deprioritize_some_providers
mtl/ofi: Deprioritize some OFI providers.
2015-08-10 14:45:59 -04:00
Nathan Hjelm
d42e0968b1 coll/libnbc: rewrite parts of libnbc
This commit rewrites parts of libnbc to fix issues identified by
coverity and myself. The changes are as follows:

 - libnbc function would return invalid error codes (internal to
   libnbc) to the mpi layer. These codes names are of the form
   NBC_. They do not match up with the error codes expected by the mpi
   layer. I purged the use of all these error codes with the exception
   of NBC_OK and NBC_CONTINUE in progress. These codes are used to
   identify when a request handle is complete.

 - Handles and schedules were leaked by all collective routines on
   error. A new routine was added to return a collective handle
   (NBC_Return_handle).

 - Temporary buffers containting in/out neighbors for neighborhood
   collectives were always leaked.

 - Neigborhood collectives contained code to handle MPI_IN_PLACE which
   is never a valid input for the send or receive buffer. Stipped this
   code out.

 - Files were inconsistently named. Most are nbc_isomething.c but one
   was named coll_libnbc_ireduce_scatter_block.c.

 - Made the NBC_Schedule "structure" and object so it can be
   retained/released. This may enable the use of schedule caching at a
   later time. More testing will be needed to ensure the caching code
   works. If it doesn't the code should be stripped out completely.

 - Added code to simply common case of scheduling send/recv +
   barrier.

 - Code cleanup for readability.

The code now passes the clang static analyzer.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2015-08-10 11:53:25 -06:00
George Bosilca
0a91d7af4d Fix issues identified by Coverity. 2015-08-08 16:41:30 -04:00
Howard Pritchard
acf64b20c7 Merge pull request #786 from nrgraham23/add_implements_cloneable
Add explicit implementation of Cloneable
2015-08-08 13:02:43 -06:00
Jeff Squyres
bd5bf4a224 Merge pull request #781 from hppritcha/topic/suppress_picky_warning
mca/topo: suppress picky warning
2015-08-08 06:14:52 -04:00
yohann
88038b5261 mtl/ofi: Deprioritize some OFI providers.
Some OFI providers such as "sockets" are used for debugging
purposes mostly. For these providers, other components usually
offer better performance -- e.g. for sockets, the BTL/TCP would
be a better choice.
Thus, we chose to ignore some providers unless explicitly asked
by the user on the command line:

e.g. --mca mtl_ofi_provider sockets
2015-08-07 16:09:51 -07:00
Edgar Gabriel
d719497f82 Performance tuning: increase the priority of the sm sharedfp component to ensure that it is selected if it can run. 2015-08-07 16:32:53 -05:00
Edgar Gabriel
9e29edf15c remove a erroneous paranthesis which prevents the compilation of the lustre adio 2015-08-07 15:22:41 -05:00
Edgar Gabriel
1293d9c69b free memory correctly in case of an error. Fixes CID 131540 and CID 1315419 2015-08-07 13:30:50 -05:00
Nathaniel Graham
8fcc317a57 Add explicit implementation of Cloneable
Added Cloneable to the implemented interface list as per
Coverity suggestion.  The required methods were already
implemented, but it was not explicitly stated.  This is
an intent revealing change.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-08-07 12:16:06 -06:00
Edgar Gabriel
0aa3049bfc Performance tuning: change the default behavior of ompio to *not* segment individual read/write operations.
In most cases, performance seems to be better if not segmented.
2015-08-07 13:06:39 -05:00
Edgar Gabriel
db5af26de7 Performance tuning. make sure we catch if the user wants to set the default fileview and replace it with our optimized default file view. Otherwise, performance will suffer. file_get_view should still return the correct filetype, not our optimized default file view. This is the correct version compared to ffa67b9693, which unfortunately broke
some test cases in mpi_test_suite. Thanks for @ggouaillardet for reporting this!
2015-08-07 12:49:58 -05:00
Edgar Gabriel
6f6c01ee8d free the datatypes that were created using type_dup during file_set_view 2015-08-07 11:50:25 -05:00
Edgar Gabriel
1ae4f8c7e6 Revert "Performance tuning. make sure we catch if the user wants to set the default fileview and replace it with"
This reverts commit ffa67b9693.
2015-08-07 09:53:07 -05:00
Gilles Gouaillardet
907c095f66 Merge pull request #779 from edgargabriel/topic/fcoll_fixes
Topic/fcoll fixes
2015-08-07 09:14:31 +09:00
Howard Pritchard
10aac8037f mca/topo: suppress picky warning
When configured with --enable-picky

topo_base_lazy_init.c compiles with a warning:

  CC       base/topo_base_lazy_init.lo
base/topo_base_lazy_init.c:46:67: warning: implicit conversion from enumeration type 'enum mca_base_register_flag_t' to different enumeration type 'mca_base_open_flag_t' (aka 'enum mca_base_open_flag_t') [-Wenum-conversion]
        err = mca_base_framework_open (&ompi_topo_base_framework, MCA_BASE_REGISTER_DEFAULT);

This commit fixes this implicit conversion problem.

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
2015-08-05 16:11:04 -06:00
Rolf vandeVaart
cb84a85d17 Build always and fix return value 2015-08-05 17:23:55 -04:00
Edgar Gabriel
16d4171f6b the individual component should call internal ompio functions directly. The reason is that otherwise
the redirection to the ompi_file_t structure (and back to the ompio internal structure) is ambiguise and wrong
for the shared file pointer scenario.
2015-08-05 14:31:11 -05:00
Edgar Gabriel
02a4eb2f13 add the ompi_file_t pointer correctly on the ompio file handle for the sm and individual component. 2015-08-05 14:28:27 -05:00
Jeff Squyres
a36d7e6026 treematch: __FUNCTION__ -> __func__ fixes 2015-08-05 05:39:38 -07:00
Jeff Squyres
a0ebbee6ef libnbc: __FUNCTION__ -> __func__ fixes 2015-08-05 05:27:23 -07:00
Gilles Gouaillardet
3d1780f1a2 sharedfp: set f_fh when opening a shared file 2015-08-05 15:07:21 +09:00
Howard Pritchard
36d7855280 Merge pull request #763 from nrgraham23/status_x_java_bindings
Status x java bindings
2015-08-03 09:26:16 -06:00
Gilles Gouaillardet
2fa8f4d53a Revert "fortran: optimize testall and waitall when MPI_STATUSES_IGNORE is used"
Optimization was incorrect, so revert this commit.

This reverts commit 83375bcc5c.
2015-08-03 08:45:12 +09:00
Jeff Squyres
047eccef8d Merge pull request #725 from bosilca/treematch
Add a new topo module: Treematch
2015-07-31 15:17:54 -04:00
Howard Pritchard
8649a9f6ef Merge pull request #757 from roblatham00/lustre-excl-open-fix
hint processing should not open files
2015-07-31 12:16:14 -06:00
Howard Pritchard
f18988c06f Merge pull request #764 from nrgraham23/Datatype.java_comment_fixes
Commenting fixes for Datatype.java
2015-07-31 12:03:04 -06:00
rhc54
a9b10cfbf0 Merge pull request #761 from jithinjosepkl/master
Fix warnings in direct (pml-cm,mtl-ofi) build
2015-07-31 09:15:30 -07:00
Edgar Gabriel
ffa67b9693 Performance tuning. make sure we catch if the user wants to set the default fileview and replace it with
our optimized default file view. Otherwise, performance will suffer. file_get_view should still return the correct filetype, not our optimized default file view
2015-07-30 19:15:00 -05:00
Edgar Gabriel
93a303ba89 Performance tuning: make sure the individual component is selected for 1 and 2 process communicators (important for some benchmarks) 2015-07-30 17:31:16 -05:00
Edgar Gabriel
9b2a7e41f0 make sure the final number of aggregators is recorded correctly when not using
our aggregator selection logic.
2015-07-30 17:24:01 -05:00
Nathaniel Graham
57b95fff2e Commenting fixes for Datatype.java
Some of the @code tags were incorrect.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-30 15:17:55 -06:00
Nathaniel Graham
17c606b35c White Space Fixes
Github revealed some wierd spacing issues.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-30 15:07:19 -06:00
Nathaniel Graham
2fecb06059 Status X Functons
Includes java bindings for MPI_GET_ELEMENTS_X and
MPI_STATUS_SET_ELEMENTS_X.  This PR also adds the
Count object which represents MPI_Count.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-30 15:00:13 -06:00
Howard Pritchard
ac4197fb1b Merge pull request #760 from nrgraham23/request_get_status_java
Java binding for MPI_REQUEST_GET_STATUS
2015-07-30 13:57:06 -06:00
Rob Latham
6e9cbe397f hint processing should not open files
move opening of files from hint processing and into open routines.

This is MPICH commit 92f1c69f0de8 and 22a77dceda11

see https://trac.mpich.org/projects/mpich/ticket/2261
Ref: https://github.com/open-mpi/ompi/issues/158

Signed-off-by: Pavan Balaji <balaji@anl.gov>
2015-07-30 12:25:20 -05:00
Nysal Jan K.A
5a7bd898f9 Fix an invalid memory access in mrecv & imrecv
After the call to PML mrecv/imrecv, the message handle
is set to MPI_MESSAGE_NULL. Use the cached communicator
while invoking the error handler.
2015-07-30 19:52:40 +05:30
Jithin Jose
bc4e8b7e73 Fix warnings in direct (pml-cm,mtl-ofi) build
Signed-off-by: Jithin Jose <jithin.jose@intel.com>
2015-07-29 15:49:37 -07:00
Howard Pritchard
ea6f5b31fd Merge pull request #749 from nrgraham23/additional_comm_java_bindings
Additional java bindings for the Comm class
2015-07-29 16:17:54 -06:00
Edgar Gabriel
477083bca3 the memory chunk that has to be allocated for the llapi_get_stripe function seems to have changed compared to earlier version. This implementation now follows the code snipplet from the man pages. 2015-07-29 17:13:55 -05:00
Edgar Gabriel
217dcca853 - the memory chunk that has to be allocated for the llapi_get_stripe function seems to have changed compared to earlier version. This implementation now follows the code snipplet from the man pages.
- implementation of file_get_size and set_size
2015-07-29 17:10:39 -05:00
yohann
6eba52a121 mtl/ofi: add missing return. 2015-07-29 14:14:34 -07:00
Nathaniel Graham
9b5786d954 Java binding for MPI_REQUEST_GET_STATUS
This adds the java binding for MPI_REQUEST_GET_STATUS.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-29 12:04:46 -06:00
Howard Pritchard
569764cd98 Merge pull request #759 from nrgraham23/status_java_bindings
Java bindings for Status methods
2015-07-29 10:36:56 -06:00
Nathaniel Graham
efe69c89a7 Java bindings for Status methods
Includes java bindings for the MPI_STATUS_SET_ELEMENTS and
MPI_STATUS_SET_CANCELLED.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-29 09:47:03 -06:00
Ralph Castain
023936e84b Silence coverity warnings 2015-07-29 07:28:08 -07:00
Gilles Gouaillardet
83375bcc5c fortran: optimize testall and waitall when MPI_STATUSES_IGNORE is used 2015-07-29 10:00:27 +09:00
Edgar Gabriel
a3327fe299 Merge pull request #756 from edgargabriel/pr/nb-sharedfp-splitcoll2
- make the split collective shared file pointer operations work
2015-07-28 19:53:27 -05:00
Edgar Gabriel
3780089ce0 clean up the usage of opal_output vs. printf 2015-07-28 18:27:31 -05:00
Howard Pritchard
377bad18bd Merge pull request #747 from hppritcha/topic/ofi_progress_fix
mtl/ofi: don't inline ofi progress method
2015-07-28 09:42:01 -06:00
Edgar Gabriel
f09b834cab remove the C++ interfaces for the new non-blocking collective I/O functions. They are not defined in MPI 3.1 2015-07-28 09:28:37 -05:00
Edgar Gabriel
824d488709 - make the split collective shared file pointer operations work
- minor code restructering in io/ompio required for that.
2015-07-28 09:05:05 -05:00
Edgar Gabriel
e380f8c235 - fix the delete priority of the ompio component
- some application use MPI_File_delete as a collective function (e.g. IOR), which I think is not really covered by the standard. Right now, one process succeeds and theother ones return an error code. Fix that by not returning no error if the file that we try to delete does not exist anymore, to make these applications work.
2015-07-27 15:53:40 -05:00
Edgar Gabriel
3fb0614566 mark the request as ACTIVE 2015-07-27 12:43:45 -05:00
Edgar Gabriel
5e166c81a1 Merge pull request #745 from edgargabriel/pr/sharedfp-sm-logic3
Pr/sharedfp sm logic3
2015-07-27 12:04:53 -05:00
Howard Pritchard
f5c43c1185 mtl/ofi: retain inline progress function
Retain inline progress function for ofi
mtl, but have a non-inlined progress function
which is registered with the opal progress
mechanism.

 @jithinjosepkl

I've bad news about the psm provider.  I still notice
segfaults - not always - but frequently at finalize
when using the psm provider.  I don't notice this
when using the sockets provider.

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
2015-07-27 09:16:52 -06:00
Gilles Gouaillardet
9c77c6b66d fortran: fix f08 bindings
only define the unique fortran symbol depending on
 - CAPS
 - PLAIN
 - SINGLE_UNDERSCORE
 - DOUBLE_UNDERSCORE
and bind the f08 symbol to the uniquely defined C symbol.

Use real data structures to make the code simpler.
(perl script written by Jeff)
2015-07-27 16:28:57 +09:00
Gilles Gouaillardet
318a1a40a4 coll/libnbc: ireduce_scatter_block
silence malloc(0) warning reported by Lisandro
2015-07-27 16:23:08 +09:00
Gilles Gouaillardet
7af5238595 f08: misc Makefile fixes 2015-07-27 16:19:44 +09:00
George Bosilca
e239de581b Create a new topology framework using the TreeMatch library developped
at Inria Bordeaux. This allows us to take advantage of the remap
capability of MPI to rearrange the ranks beased on the weights
povided by the application.

Fix the indentation and protect with __DEBUG__ one fprintf.

Add the Cecill-B license to the imported library.

Fix a compiler warning.

Restrict the TreeMatch dependencies.

The TreeMatch software is released under BSD3 (as indicated by their
copyright information @
https://gforge.inria.fr/scm/viewvc.php/COPYING?view=markup&root=treematch).

Update the README.
2015-07-25 13:30:42 -04:00
Jeff Squyres
3e6694f7ea sharedfp: whitespace cleanup
No code changes.

Replace tabs with spaces and do other whitespace cleanup (via emacs).
2015-07-25 05:46:37 -07:00
Jeff Squyres
868a84d4da sharedfp: have sm_data->mutex always point to the right mutex
Even if the mutex is actually located in
sm_data->sm_offset_ptr->mutex, have sm_data->mutex point to it.  This
avoids a few #if blocks that are otherwise identical.
2015-07-25 05:42:57 -07:00
Jeff Squyres
5e159feb5a get_library_version: always use "X.Y.Z" version number format
Fix one last place where we would write only "X.Y" if Z was 0 (we now
always write all three X, Y, Z values, even if Z is zero).
2015-07-25 08:01:36 -04:00
Nathaniel Graham
59c43f4669 Additional java bindings for the Comm class
Added bindings for MPI_COMM_CREATE_GROUP and MPI_COMM_DUP_WITH_INFO.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-24 16:14:35 -06:00
Nathaniel Graham
f87c41e926 Fix a compile issue in MPI.java
A conflict was not fully resolved in the previous merge.  This
fixes the issue that was created.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-24 11:54:52 -06:00
Nathaniel Graham
0d7a38552e Formatting fixes for Java files.
Fixed tab indenting as well as some of the block comments
at the top of a few of the files.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-24 11:02:13 -06:00
Howard Pritchard
5f14273e32 Merge pull request #743 from nrgraham23/java_win_name_bindings
Java bindings for window name get and set
2015-07-24 10:53:14 -06:00
Edgar Gabriel
4f85e0d833 add the configure logic to check for sem_open and sem_init.
Change the code to rely on HAVE_SEM_OPEN etc. instead of my internal macro.
2015-07-24 10:23:43 -05:00
Edgar Gabriel
d1d23054c6 rename the sm_offset structure to mca_sharedfp_sm_offset to obey to the Open MPI naming convention 2015-07-24 10:10:41 -05:00
Edgar Gabriel
c91cb67787 fix a bug in the unnamed semaphore section that was introduced when I tried to unify the named and unnamed semaphore logic. 2015-07-24 10:05:07 -05:00
Edgar Gabriel
57c301f25a remove an erroneous free statement. 2015-07-24 09:44:27 -05:00
Nathaniel Graham
2a156a1c2e Java bindings for window name get and set
Includes bindings for MPI_WIN_GET_NAME and MPI_WIN_SET_NAME.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-23 18:01:19 -06:00
Nathaniel Graham
a1b47a333a Java Environment Variable Bindings
Added the bindings for MPI_GET_VERSION and MPI_GET_LIBRARY_VERSION.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-23 14:52:01 -06:00
Howard Pritchard
967907fea4 Merge pull request #731 from nrgraham23/alltoallw_functions
Alltoallw functions
2015-07-22 12:21:01 -06:00
Nathaniel Graham
1197d2a893 Adding missed copyrights.
I did not add the LANL copyrights to the files I touched
while fixing the javadoc errors.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-22 11:04:41 -06:00
Jeff Squyres
6929aca1b7 topo/basic: also remove .windows from Makefile.am 2015-07-22 09:20:43 -04:00
Jeff Squyres
24ca887bd8 topo/basic: remove stale (empty) .windows file 2015-07-22 09:10:50 -04:00
Nathaniel Graham
11e1f09c25 White space fixes
Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-21 16:39:59 -06:00
Nathaniel Graham
7548e342dd Java bindings for alltoallw functions.
Includes bindings for MPI_ALLTOALLW and MPI_IALLTOALLW.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-21 16:29:22 -06:00
Gilles Gouaillardet
46a87cabf0 fix comment in ompi/mpi/c/group_{excl,incl}.c
no code change
2015-07-21 10:29:55 +09:00
Edgar Gabriel
b484784dca make ompio return gracefully in case something goes wrong early in file_open. 2015-07-20 10:03:16 -05:00
Edgar Gabriel
86c3000e18 fix the delete selection logic in io/base. With the previous version, there was a mismatch
in the version number and no component was selected for file_delete.
2015-07-20 10:01:30 -05:00
Gilles Gouaillardet
bd60ce1655 fortran: add missing non blocking i/o files to Makefile
fixes open-mpi/ompi@01f752a504
2015-07-16 11:27:14 +09:00
Howard Pritchard
466c8b0159 Merge pull request #697 from edgargabriel/pr/nb-coll-part2
pr/nb collective I/O part2
2015-07-14 14:00:39 -06:00
Edgar Gabriel
e355db005e fix the logic for setting stripe size and stripe count in the lustre fs module. Takes now also the MPI_Info object into consideration. 2015-07-14 10:53:19 -05:00
Ralph Castain
683efcb850 Rename the current opal_event_base to opal_sync_event_base in preparation for adding an async progress thread to opal. No functional changes made here - just a simple rename. 2015-07-11 10:08:19 -07:00
Jeff Squyres
efc4c93d7a MPI_Scatter.3in: fix typo
Thanks to Akshay Venkatesh for noticing the mistake.
2015-07-10 20:11:39 -07:00
Howard Pritchard
4a3ca9d9b8 Merge pull request #710 from nrgraham23/javadoc_warnings_fix
Javadoc warnings fix
2015-07-10 19:08:38 -06:00
Nathaniel Graham
6bca7486c5 Added suppression statements.
There are a few places where adding the @param for the variable
javadoc wants does not make sense, so I added suppression statements
in those areas.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-10 16:04:22 -06:00
Rolf vandeVaart
06dbcc1b25 Merge pull request #708 from rolfv/pr/cuda-macro-fix-minor
Minor rewrite of text and comments
2015-07-10 15:32:03 -04:00
Nathaniel Graham
580251e96a Fixes for javadoc warnings.
Added missings @params and @throws as well as adding definitions for the @throws tags.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-10 12:59:45 -06:00
Rolf vandeVaart
5ee2505174 Minor rewrite of text and comments 2015-07-10 13:50:06 -04:00
Nathaniel Graham
7f6d141bd6 More one sided java bindings.
Bindings for the MPI_WIN_FLUSH_LOCAL, MPI_WIN_FLUSH_LOCAL_ALL, MPI_WIN_ALLOCATE, MPI_WIN_ALLOCATE_SHARED, and MPI_COMM_SPLIT_TYPE.  Also added several necessary constants.

Signed-off-by: Nathaniel Graham ngraham@lanl.gov
2015-07-10 11:40:18 -06:00
Howard Pritchard
e9d6c7a910 Merge pull request #693 from nrgraham23/additional_java_bindings
Additional java bindings
2015-07-10 11:10:29 -06:00
Edgar Gabriel
95c144c4b6 make the f08 interface actually compile. 2015-07-09 10:58:14 -05:00
Edgar Gabriel
601e5efca8 add the manpages for the new MPI-3.1 non-blocking collective I/O operations. 2015-07-09 10:58:14 -05:00
Edgar Gabriel
19994ff60d add fortran interfaces to the new nb collective I/O functions 2015-07-09 10:58:14 -05:00
Edgar Gabriel
01f752a504 make the c interface functions compile and work and start with the fortran interface functions. 2015-07-09 10:58:13 -05:00
Edgar Gabriel
f2af8e94ff - first cut on the io interface changes
- add the C interfaces for the new non-blocking collective I/O functions of MPI 3.1
2015-07-09 10:58:13 -05:00
Rolf vandeVaart
77950374a5 Merge pull request #695 from rolfv/pr/fix-previous-cuda-macro-change
Fix errors from previous commit
2015-07-08 12:59:10 -04:00
yosefe
103cac5bd9 yalla: fix mxm configuration parsing.
Take configuration from MXM_MPI_xx instead of MXM_PML_xx, same as mtl
mxm.
2015-07-08 19:18:23 +03:00
Rolf vandeVaart
1229dd3ed6 Fix errors from previous commit 2015-07-08 11:46:52 -04:00
Rolf vandeVaart
2797f5b8fc Merge pull request #694 from rolfv/pr/update-cuda-macro
Some minor cleanup and renaming of query function.
2015-07-08 11:09:17 -04:00
Rolf vandeVaart
6372baa7b9 Some minor cleanup and renaming of query function. 2015-07-08 10:17:35 -04:00
Jeff Squyres
d5e16839e9 mpi_f08: restore C/Fortran compatibility with F08 BUFFER_DETACH
The C int/Fint interoperability code was erroneously removed when this
function was moved to the use-mpi-f08 directory.
2015-07-08 03:55:50 -07:00
Nathaniel Graham
26c528a627 Copyright additions
Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-07 16:14:21 -06:00
Howard Pritchard
68c69d07ff add copyrights to changed java related files
Signed-off-by: Howard Pritchard <howardp@lanl.gov>
2015-07-07 16:10:58 -06:00
Nathaniel Graham
982a232955 Additional Java Bindings
Java bindings for the following functions: MPI_RACCUMULATE, MPI_GET_ACCUMULATE, MPI_RGET_ACCUMULATE, MPI_WIN_LOCK_ALL, MPI_WIN_UNLOCK_ALL, MPI_WIN_SYNC, MPI_WIN_FLUSH, MPI_WIN_FLUSH_ALL, MPI_COMPARE_AND_SWAP, and MPI_FETCH_AND_OP.  Also includes Java bindings for the Operations MPI_REPLACE and MPI_NO_OP.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-07-07 16:04:17 -06:00
Jeff Squyres
8ff60c47cc mpi_f08: move the f08 version of buffer_detach to use-mpi-f08
@ggouaillardet was right -- we should have put the
ompi_buffer_detach_f08() function in the use-mpi-f08 directory to
begin with.  Putting it in the mpif-h directory made it complicated as
to whether the function would be built or not (e.g., whether weak
symbols were supported or not, whether the profiling layer was
disabled or not, ...etc.).

Just put it in the use-mpi-f08 directory and always build it (when the
mpi_f08 module is built, of course), and keep it simple.
2015-07-07 07:38:09 -07:00
Gilles Gouaillardet
9e89985f3d restore whitespaces into the pdf files 2015-07-07 09:17:00 +09:00
Howard Pritchard
fb9c218e5d Merge pull request #684 from nrgraham23/java_bindings
ompi/java: add MPI_Rget and MPI_Rput java bindings
2015-07-06 15:25:06 -06:00
Jeff Squyres
e28f016793 fortran only compile f08 buffer_detach once
Since there is no profiling version of the f08 buffer_detach function
(or, more specifically, the Fortran compile does the name mangling of
MPI and PMPI to the back-end C function for us), ensure that it is
only compiled once.

Also, per Gilles' observation, the f08-related #pragmas are no longer
relevant.
2015-07-06 09:52:27 -07:00
Jeff Squyres
4694b16256 mpi_f08: further fixes for MPI_BUFFER_DETACH
A few minor fixes in addition to
open-mpi/ompi@bd94da7de4 and
open-mpi/ompi@fc69c0be24.
2015-07-04 05:27:22 -07:00
Jeff Squyres
bd94da7de4 mpi_f08: fix MPI_BUFFER_DETACH interfaces
Add an mpi_f08-specific implementation for MPI_BUFFER_DETACH.

Per MPI-3.1:3.6, p45, the buffer argument is ignored in
MPI_BUFFER_DETACH for mpif.h and the mpi module.  But in the mpi_f08
module, the buffer argument is treated like it is in the C binding.
2015-07-03 08:56:34 -07:00
bosilca
77367ca02c Merge pull request #687 from rolfv/pr/fix-smcuda-perfprob
Add the ability use different size buffers for host and CUDA buffers
2015-07-02 18:42:41 -04:00
Rolf vandeVaart
30a872b478 Add the ability to send host buffers through one sized staging buffers and CUDA buffers through different sized buffers. Fixes performance issues 2015-07-02 11:11:15 -04:00
Gilles Gouaillardet
6356c4b113 mpif: fix typo in interface name
* MPI_Comm_dup_with_info
 * MPI_Comm_split_type
 * PMPI_Comm_split_type
2015-07-02 17:13:07 +09:00
Gilles Gouaillardet
fa50e0d712 mpi_f08: fix typo in binding name
replace MPI_Win_flush_all_local with MPI_Win_flush_local_all
2015-07-02 15:47:38 +09:00
Nathaniel Graham
bb5699e912 ompi/java: add MPI_Rget and MPI_Rput java bindings
Wrote tests for ompi tests, but will commit later.

Signed-off-by: Nathaniel Graham <ngraham@lanl.gov>
2015-06-30 12:05:59 -06:00
Gilles Gouaillardet
c4025f55a4 mpi_f08: add MPI_Type_size_x and misc PMPI_xxx bindings 2015-06-30 16:31:14 +09:00
Gilles Gouaillardet
18fd30a173 mpi_f08: fix MPI_FREE_MEM binding 2015-06-30 14:19:50 +09:00
Gilles Gouaillardet
13950c3ec1 mpi_f08: fix MPI_WIN_{ATTACH,DETACH} bindings
fixes INTENT from open-mpi/ompi@9600e2bc63
2015-06-30 12:28:23 +09:00
Gilles Gouaillardet
86bbdd2ebe mpi_f08: fix MPI_COMM_SPAWN_MULTIPLE binding 2015-06-30 11:47:45 +09:00
Gilles Gouaillardet
a4cc83f4f7 mpiext/affinity: fix missing function type 2015-06-30 09:37:27 +09:00
Rolf vandeVaart
dfa4b5e474 Merge pull request #680 from rolfv/pr/add-cuda-macro
Add a macro to indicate CUDA-aware support
2015-06-29 14:24:19 -04:00
Jeff Squyres
f6882a85bb mpiext_affinity_str.c: minor stylistic updates
No real functional changes:

* Reduce #if's a little -- have a single "no hwloc" and "hwloc"
  section.
* Make a common subroutine (no_hwloc_support()) for when we don't have
  any hwloc support
2015-06-29 09:36:10 -07:00
Jeff Squyres
fc69c0be24 mpi_f08: fix MPI_BUFFER_DETACH binding 2015-06-29 09:22:30 -07:00
Rolf vandeVaart
e1983ff46a Add a macro to indicate CUDA-aware support 2015-06-29 12:08:27 -04:00
Jeff Squyres
91d3b5f555 errcode.h: fix typo: errnum, not errno
Thanks to Åke Sandgren for pointing out the error.
2015-06-29 08:32:46 -07:00
Gilles Gouaillardet
6994d742fd mpiext/affinity: #if OPAL_HAVE_HWLOC most parts of the code
that allow this extension to compile if configure'd with --without-hwloc
2015-06-29 09:54:03 +09:00
Gilles Gouaillardet
e0dd9bd2e7 Fix java bindings disaster
that commit fixes :
- open-mpi/ompi@224f97b009
- open-mpi/ompi@ade7de549c
- open-mpi/ompi@0836344673
- open-mpi/ompi@7e0581c853
- open-mpi/ompi@d948a43993
2015-06-29 09:51:36 +09:00
Gilles Gouaillardet
d948a43993 java: fix typo in MPI_Win_attach binding 2015-06-28 09:27:19 +09:00
Gilles Gouaillardet
98bf02485d fortran: add Fortran 2008 bindings for attr_fn (MPI_COMM_DUP_FN and friends)
Fixes open-mpi/ompi#460
2015-06-26 15:21:30 +09:00
Gilles Gouaillardet
e701bf35cc MPI_Win_{get,set}_info : add man pages 2015-06-26 15:14:25 +09:00
Gilles Gouaillardet
6e2951e8d7 MPI_Win_{attach,detach} : add man pages 2015-06-26 15:14:25 +09:00
Gilles Gouaillardet
f7bbad96be MPI_Win_create_dynamic : add man page 2015-06-26 15:14:24 +09:00
Gilles Gouaillardet
7e0581c853 MPI_Win_{get,set}_info : add Java bindings 2015-06-26 15:10:15 +09:00
Gilles Gouaillardet
0836344673 MPI_Win_{attach,detach} : add Java bindings 2015-06-26 15:10:14 +09:00
Gilles Gouaillardet
ade7de549c MPI_Win_create_dynamic : add Java bindings 2015-06-26 15:10:14 +09:00
Ralph Castain
2fa8167383 Remove manpages when doing distclean in the OMPI extensions area 2015-06-25 21:01:17 -07:00
Jeff Squyres
13425e759c bml r2: very minor cleanups
Delete stale comments, use C99 struct initialization.
2015-06-25 15:54:16 -07:00
Rolf vandeVaart
5aa11a2c9f Fix library name 2015-06-25 18:07:01 -04:00
Rolf vandeVaart
2750a6e237 Fix filename name 2015-06-25 17:41:44 -04:00
Rolf vandeVaart
995723ca6f Fix compilation of cuda extensions
Fix compilation
2015-06-25 17:19:14 -04:00
rhc54
d01f37c063 Merge pull request #667 from rhc54/topic/cudaext
Add CUDA MPI extension for Rolf - because @jsquyres  said it was OKAY
2015-06-25 13:41:54 -07:00
Ralph Castain
265cd14f60 Purge whitespace 2015-06-25 13:27:56 -07:00
Ralph Castain
a390a8f5ef Add a CUDA MPI extension to provide the MPIx_CUDA_SUPPORT macro 2015-06-25 13:07:35 -07:00
rhc54
8b62b63786 Merge pull request #666 from rhc54/topic/libfab
Add a common/libfabric component
2015-06-25 12:46:34 -07:00
rhc54
5e6670343e Merge pull request #673 from jsquyres/pr/mpi-extension-defaults
MPI extensions: build extensions by default
2015-06-25 12:38:26 -07:00
Jeff Squyres
fca193d316 MPI extensions: build extensions by default
* affinity: will build unless disabled.
* cr: will build if FT is enabled, unless it is disabled.  It will
  also complain/abort if you --with-mpi-ext=cr, but FT is disabled.
* example: will only build if --with-mpi-ext=example (and .ompi_ignore
  is removed)
2015-06-25 11:28:58 -07:00
Yohann Burette
7fd5ded327 mtl/ofi: message truncation is now indicated by FI_ETRUNC. 2015-06-25 11:06:41 -07:00
Yohann Burette
483ff23db1 mtl/ofi: cancels are now tracked by an error entry. 2015-06-25 11:06:41 -07:00
Ralph Castain
ea0e21bb06 Add a common/libfabric component to the opal layer where we can place common functions 2015-06-25 11:04:00 -07:00
Nathan Hjelm
ee36d813dc Merge pull request #657 from hjelmn/c99
more c99 updates
2015-06-25 11:21:09 -06:00
Howard Pritchard
f45914db9b Merge pull request #670 from hppritcha/topic/ownership_update
ownership: update ownership files
2015-06-25 11:02:45 -06:00
Nathan Hjelm
4d92c9989e more c99 updates
This commit does two things. It removes checks for C99 required
headers (stdlib.h, string.h, signal.h, etc). Additionally it removes
definitions for required C99 types (intptr_t, int64_t, int32_t, etc).

Signed-off-by: Nathan Hjelm <hjelmn@me.com>
2015-06-25 10:14:13 -06:00
Howard Pritchard
e49a37c034 ownership: update ownership files
per discussions at OMPI devel workshop

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
2015-06-25 10:04:42 -06:00
Jeff Squyres
0bb3fd0a10 coll hierarch: remove last stale file 2015-06-25 08:40:50 -07:00
Nathan Hjelm
e561f150d6 Merge pull request #658 from hjelmn/mpit_fixes
Fix definition of MPI_T_pvar_get_index
2015-06-24 16:40:25 -07:00
Nathan Hjelm
4552afff06 Fix definition of MPI_T_pvar_get_index
The definition of MPI_T_pvar_get_index was incorrect. This commit
fixes the definition and adds a missing return code.

Signed-off-by: Nathan Hjelm <hjelmn@me.com>
2015-06-24 17:31:26 -06:00
George Bosilca
dc1b125b12 There is no destructor for the base requests. 2015-06-24 14:29:45 -07:00
bosilca
1b8556f926 Merge pull request #653 from hjelmn/moar_ob1_fixes
pml/ob1: fix bugs in static request objects
2015-06-24 14:28:11 -07:00
Nathan Hjelm
458cd9d611 Merge pull request #629 from hjelmn/aint_math
Add support for MPI-3.1 MPI_Aint functions
2015-06-24 10:52:44 -07:00
Nathan Hjelm
6dd9658d73 man: remove F77 notes from MPI_Aint_add/MPI_Aint_diff man pages
Signed-off-by: Nathan Hjelm <hjelmn@me.com>
2015-06-24 11:49:48 -06:00
Jeff Squyres
258d1aa160 mpi-f08: corrections to new MPI_Aint_* subroutines
(cherry picked from commit 8f252e65b02bfe20cacee0ece8c835246eea3bd9)
2015-06-24 10:42:01 -07:00
Ralph Castain
869041f770 Purge whitespace from the repo 2015-06-23 20:59:57 -07:00
Nathan Hjelm
9a8a87611e pml/ob1: fix bugs in static request objects
This commit fixes several bugs in the static request objects used by
ob1 for blocking send/receive operations.

 - Fix memory leak when using MPI_THREAD_MULTIPLE. Requests were
   allocated off the free list but were destructed and NOT returned.

 - Fix double-destruct of static objects. There is no reason to
   CONSTRUCT/DESTUCT the static object for each send/receive
   operation. This adds overhead and no benefit. To keep the code
   clean helper functions have been added to finalize ob1 send/receive
   requests.

 - Remove now unnecessary include of alloca.h.

Signed-off-by: Nathan Hjelm <hjelmn@me.com>
2015-06-23 11:00:45 -06:00
Nathan Hjelm
ac51acb3e1 Merge pull request #651 from hjelmn/fix_thread_multiple_check
pml/ob1: do not use OPAL_ENABLE_MULTI_THREADS to determine thread multiple support
2015-06-22 21:45:43 -06:00
Nathan Hjelm
284dd6babe pml/ob1: do not use OPAL_ENABLE_MULTI_THREADS to determine thread multiple support
OPAL_ENABLE_MULTI_THREADS is always on. The correct value to check is
OMPI_ENABLE_THREAD_MULTIPLE.

Signed-off-by: Nathan Hjelm <hjelmn@me.com>
2015-06-22 19:17:23 -06:00
Andrew Friedley
2c9be59b37 Add new PSM2 MTL.
This new MTL runs over PSM2 for Omni Path.  PSM2 is a descendant of PSM
with changes to support more ranks and some MPI-3 features like mprobe.

PSM2 will only support Omni Path networks; PSM only supports True Scale.
Likewise, the existing PSM MTL will continue to be maintained for True
Scale, while the PSM2 MTL is developed and maintained for Omni Path.
2015-06-22 07:55:46 -07:00
Gilles Gouaillardet
0bd765eddd fix NBC_Copy for legitimate zero size messages
this fixes a regression from open-mpi/ompi@9a70765f27
2015-06-22 09:51:25 +09:00
Edgar Gabriel
dedeee9771 finishing the changes for the non-blocking and split cpllective I/O operations. Everything except for the
interface changes to the io framework is done.
2015-06-18 06:22:41 -05:00
Edgar Gabriel
3b11a8b61c making the current work compile. 2015-06-18 05:56:51 -05:00
Edgar Gabriel
cc219281ba checkpoint of the current work, since I need to resync wioth master to fix the compilation problems 2015-06-18 05:20:07 -05:00
Edgar Gabriel
100515e321 remove split collective interfaces from fcoll and their fake implemenations. Not required anymore 2015-06-18 05:20:07 -05:00
Edgar Gabriel
19cac73a9b first part of the changes trequired to support non-blocking colelctive io operations 2015-06-18 05:20:07 -05:00
Gilles Gouaillardet
0f08070a1c ompio: fix misc memory leaks
as identified by Coverity with CIDs 72147-72149, 731275 and 1269872
2015-06-17 11:17:54 +09:00
Gilles Gouaillardet
0f17cdfc57 fcoll: fix misc memory leaks
as reported by Coverity with CIDs 72293,72294 and 1269894
2015-06-17 11:17:52 +09:00
Nathan Hjelm
c33b786dd9 Merge pull request #620 from hjelmn/ompi_coverity
ompi coverity fixes
2015-06-16 06:10:40 -06:00
rhc54
9a8bda0b72 Merge pull request #637 from jithinjosepkl/pr/pml-cm-opt
pml-cm bug fixes
2015-06-15 19:25:09 -07:00
rhc54
ea78f8226d Merge pull request #631 from jsquyres/pr/remove-embedded-libfabric
remove embedded libfabric
2015-06-15 18:14:33 -07:00
Jithin Jose
7ccde09a09 Do opal_convertor_copy_and_prepare_for_send for buffered send mode as
MCA_PML_CM_HVY_SEND_REQUEST_BSEND_ALLOC calls opal_convertor_pack
directly.

Signed-off-by: Jithin Jose <jithin.jose@intel.com>
2015-06-15 17:12:50 -07:00
Howard Pritchard
224f97b009 mpi/java: suppress common symbols warning
Signed-off-by: Howard Pritchard <howardp@lanl.gov>
2015-06-15 17:13:40 -06:00
Jeff Squyres
cc66745e7a mtl/ofi: convert to use external libfabric
Use the new OPAL_CHECK_LIBFABRIC macro.
2015-06-15 15:17:06 -07:00
Gilles Gouaillardet
ee3a1da28a pml/ob1:mca_pml_ob1_recv_request_put_frag silence a warning
proc local variable is used only in heterogeneous mode
2015-06-15 10:00:53 +09:00
George Bosilca
67b70bb47a Add multi-threaded support. 2015-06-12 14:22:17 -07:00
George Bosilca
b2cf74cabc A first cut at a possible solution for the missing requests
from the message queues (a debugging feature). With this approach
all blocking (single threaded) requests are allocated from the main
freelist, so they will be accounted for during the message queues
investigation).
2015-06-12 14:22:17 -07:00
Ryan Grant
eec120678c Merge pull request #614 from tkordenbrock/topic/portals4.triggered.collectives
coll-portals4: implement collective operations using Portals4 triggered operations
2015-06-11 08:20:55 -06:00
Jithin Jose
7cfbfc4c89 Initialize convertor in pml-cm-send and recv.
Signed-off-by: Jithin Jose <jithin.jose@intel.com>
2015-06-10 09:39:31 -07:00
Jeff Squyres
fbaf6888f8 ompi/include/Makefile.am: rm mpi_portable_platform.h first
We've seen this a few times (e.g.,
http://www.open-mpi.org/community/lists/users/2015/06/27057.php
reported via @siegmargross).  I'm not entirely sure why it happens --
the best I can come up with is a poorly-synchronized network
filesystem and/or a bug in "make".  For example: this code hasn't
changed in forever, and it only happens to users *sometimes*.

Regardless, avoid the error altogether by removing the file before
making the sym link (it should be a sym link anyway -- if there's
something there, it should be safe to remove it before we re-create
the sym link that should be there in the first place).

(cherry picked from commit 0edd265ea045e649c9489e3cb8fdb657800d95c3)
2015-06-09 17:59:32 -07:00
Todd Kordenbrock
b725186768 mtl-portals4: Verify the result of PtlPTAlloc()
The Portals4 MTL allocates two Portals IDs requesting specific
well-known IDs and assumes that those IDs are allocated.  If those IDs
are in use, PtlPTAlloc() will allocate a different ID.  This commit
verifies that the requested IDs were allocated.
2015-06-09 14:43:50 -05:00
Nathan Hjelm
090922887b win_get_attr: fix coverity issues
CID 71734 Self assignment (NO_EFFECT)

This code has no effect. The original author of the offending code
does not remember why the self-assignment is there. Fortran
MPI_Win_get_attr tests are working with or without it so remove the
code.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2015-06-09 09:34:26 -06:00