1
1
Граф коммитов

1781 Коммитов

Автор SHA1 Сообщение Дата
Brian Barrett
f9564e7b69 * fix typo in the spawn function
This commit was SVN r1941.
2004-08-06 21:49:41 +00:00
Brian Barrett
5506cab8b2 * Bump up the priority of COFS so that it is higher than RSH (for now, anyway)
This commit was SVN r1940.
2004-08-06 21:49:13 +00:00
Jeff Squyres
88a2858e21 Add some comments about the whole "common shared library" thing
because a) it's confusing, and b) I'm sure it's only the first.  :-)

This commit was SVN r1939.
2004-08-06 21:40:23 +00:00
Brian Barrett
e8c5a60cc9 * add (useless?) timing comment
This commit was SVN r1938.
2004-08-06 21:18:34 +00:00
Ralph Castain
bdccb58d63 update unit test - prep to renaming.
This commit was SVN r1937.
2004-08-06 21:01:29 +00:00
Ralph Castain
ac7d64af5a Update this unit test - will be moved and further updated shortly.
This commit was SVN r1936.
2004-08-06 20:54:42 +00:00
Tim Prins
42593fc947 Setting svn ignores
This commit was SVN r1935.
2004-08-06 20:52:32 +00:00
Rich Graham
e92447153f add some dummy code so that we can build.
This commit was SVN r1934.
2004-08-06 20:40:09 +00:00
Weikuan Yu
944432a352 save lots of type and name changes
This commit was SVN r1933.
2004-08-06 20:28:23 +00:00
Ginger Young
5e22c1edfe Adding preliminary ompi daemon
This commit was SVN r1932.
2004-08-06 20:23:57 +00:00
Jeff Squyres
2060091238 - Creation of the common code repository for cross-component /
horizontal integration
- Move the mpool_sm_mmap.[ch] files to a new sm common code area, and
  rename to common_sm_mmap.[ch]
- update sm ptl and mpool to refer to the sm common code area for the
  mmap code

This commit was SVN r1930.
2004-08-06 19:35:57 +00:00
Tim Prins
259b2d9c35 A temporary workaround to avoid segfaults when the tcp oob is cleaning up.
This simply removes the event handler deletions from the clean up code.

This commit was SVN r1929.
2004-08-06 19:12:26 +00:00
Tim Prins
c592935f89 More bug fixes. The finalize functions in the oob were not being called,
which lead to some race conditions.
There are still cases where the tcp oob will segfault.

This commit was SVN r1928.
2004-08-06 17:23:37 +00:00
Edgar Gabriel
a35a6ff778 alltoallv: nothing has really changed.
allreduce: rearranging the final waitall, since one request was
           accidentally not initialized before calling waitall.

This commit was SVN r1927.
2004-08-06 15:47:40 +00:00
Edgar Gabriel
ba165b3d8a initializing two variables used in a dummy allreduce
This commit was SVN r1926.
2004-08-06 15:46:44 +00:00
Edgar Gabriel
b8d6fd5d50 introducing a 'hidden' communicator flag. The communicators marked hidden will not be exported to totalview (e.g. internal communicators of the MagPIe coll framework).
This commit was SVN r1925.
2004-08-06 14:33:23 +00:00
Jeff Squyres
a721849e84 While waiting for a lengthy compile, I noticed a bunch of old remnants
from the "LAM/MPI" -> "Open MPI" name change that didn't convert
properly.  So this commit fixes all instances of "OMPI/MPI" to "Open
MPI".

This commit was SVN r1924.
2004-08-06 14:30:18 +00:00
Brian Barrett
e3788309f2 * fix make dist. argh!
This commit was SVN r1922.
2004-08-05 22:52:11 +00:00
Edgar Gabriel
578c313a61 fixing a minor issue in intercomm_create.
adapting group_incl/excl/range_incl/range_excl to return
MPI_GROUP_EMPTY in case the new group size is zero.
group_union/difference/intersection still to be adapted to handle that scenario.

This commit was SVN r1921.
2004-08-05 22:51:34 +00:00
Edgar Gabriel
60f4cac622 changing the default rank in MPI_GROUP_EMPTY to MPI_UNDEFINED
This commit was SVN r1920.
2004-08-05 22:49:37 +00:00
Edgar Gabriel
6a8bebacd4 bug fix: set comm->c_topo to NULL in the allocator as well, for debug
purposes (although it did not cause any problems) we memset the
    c_coll structure to zero as well.
enabling comm_create for inter-communicators

This commit was SVN r1919.
2004-08-05 22:48:47 +00:00
Rich Graham
3b7ed9136f more changes.
This commit was SVN r1917.
2004-08-05 21:59:24 +00:00
Brian Barrett
9243100d79 * Merge in pcm extension from the tmp/brbarret-pcm branch. Extends the
PCM to have startup functionality (COFS should still work fine).

This commit was SVN r1916.
2004-08-05 21:45:35 +00:00
Tim Woodall
d4f42ec944 don't use BSD KQUEUE until this can be debugged
This commit was SVN r1914.
2004-08-05 21:37:34 +00:00
Weikuan Yu
f4ac830175 get chained DMA done, to test and add threading support
This commit was SVN r1913.
2004-08-05 20:57:25 +00:00
Prabhanjan Kambadur
a9e22a500d Adding the feature to lookup the requested modules thorugh mca_base_param_lookup
This commit was SVN r1912.
2004-08-05 20:31:14 +00:00
Laura Casswell
850a4987c7 forgot to close files
This commit was SVN r1911.
2004-08-05 19:58:57 +00:00
Laura Casswell
8167336e65 added changes to make more uniform in reporting
This commit was SVN r1910.
2004-08-05 19:52:01 +00:00
Laura Casswell
06686b08d4 added macro to test and report
This commit was SVN r1909.
2004-08-05 19:50:54 +00:00
Tim Prins
097ad1a1bc minor bug fix. close the listen socket in the correct place.
updates to oob test.

This commit was SVN r1908.
2004-08-05 19:37:48 +00:00
Tim Woodall
1b6a2a59c1 maintain reference count on communicator for duration of request
This commit was SVN r1907.
2004-08-05 19:35:10 +00:00
David Daniel
90ec5bada0 Adding directory for service framework
This commit was SVN r1906.
2004-08-05 19:33:17 +00:00
Laura Casswell
2fbdeb6f65 Adding calls to general test routines
This commit was SVN r1905.
2004-08-05 19:22:47 +00:00
Brian Barrett
1868364c85 * Make sure the .h file makes it into the dist tarball
This commit was SVN r1904.
2004-08-05 18:59:32 +00:00
Edgar Gabriel
cd2a23898b bug fixes for the MPI-2 extended collective operations. All inter-collectives
seem to pass now the test without coredumps etc.

Checking for MPI_Alltollw(intra)...working
Checking for MPI_Exscan:
         testing MPI_MAX...........working
         testing MPI_MIN...........working
         testing MPI_SUM...........working
         testing MPI_PROD..........working
         testing MPI_LAND..........working
         testing MPI_LOR...........working
         testing MPI_LXOR..........working
         testing MPI_BAND..........working
         testing MPI_BOR...........working
         testing MPI_BXOR..........working
Checking for MPI_Bcast.............working
Checking for MPI_Allreduce.........working
Checking for MPI_Reduce............working
Checking for MPI_Barrier...........working
Checking for MPI_Gather............working
Checking for MPI_Gatherv...........working
Checking for MPI_Scatter...........working
Checking for MPI_Scatterv..........working
Checking for MPI_Allgather.........working
Checking for MPI_Allgatherv........working
Checking for MPI_Alltoall..........working
Checking for MPI_Alltoallv.........working
Checking for MPI_Alltollw(inter)...working
Checking for MPI_Reduce_scatter....working
Checking for MPI_Bcast.............working
Checking for MPI_Allreduce.........working
Checking for MPI_Reduce............working
Checking for MPI_Barrier...........working
Checking for MPI_Gather............working
Checking for MPI_Gatherv...........working
Checking for MPI_Scatter...........working
Checking for MPI_Scatterv..........working
Checking for MPI_Allgather.........working
Checking for MPI_Allgatherv........working
Checking for MPI_Alltoall..........working
Checking for MPI_Alltoallv.........working
Checking for MPI_Alltollw(inter)...working
Checking for MPI_Reduce_scatter....working

This commit was SVN r1903.
2004-08-05 18:52:14 +00:00
Tim Woodall
c5626a6300 acquire peer lock before destroying it
This commit was SVN r1902.
2004-08-05 18:35:42 +00:00
Edgar Gabriel
eb4cec3566 updating the comm-creating functions to the new flow-chart
This commit was SVN r1901.
2004-08-05 16:32:13 +00:00
Edgar Gabriel
2e43e4980e a couple of changes:
- because we had the scenario, that fragments for a communicator, which was not 
   not yet set up on all procs, arrived and caused problem, we introduced
   a comm_activate function call, which executes a kind of barrier (using
   the allreduce functions used for the comm_cid allocation).
   Setting up the coll-component has moved *after* this barrier, since 
   some coll-modules (e.g. the MagPIe component) might want to communicate
   using this communicator already (e.g. a comm_split).
-  adding a new file comm_dyn.c, which basically abstracts the required functionality
   for connect-accept, and therefore is the 'magic' code (from the MPI point of view)
   for all dynamically created communicators.

This commit was SVN r1900.
2004-08-05 16:31:30 +00:00
Edgar Gabriel
b4a77ae5e0 changing the interface of two of the *new* functions from size_t to int,
since it is easier to specify the data type for an int (in both
OOB terms as well as in MPI terms), if we have to send the result
data length around.

This commit was SVN r1899.
2004-08-05 16:27:13 +00:00
Tim Prins
d2df4acb65 Get rid of some annoying compiler warnings.
This commit was SVN r1898.
2004-08-05 15:30:36 +00:00
Brian Barrett
e30b55565c * fix stupid typo
This commit was SVN r1896.
2004-08-05 14:51:56 +00:00
Brian Barrett
2f507dfc67 * Add some doxy comment to shut up Doxygen when doing LLM files
This commit was SVN r1895.
2004-08-05 14:48:33 +00:00
Brian Barrett
3be27734c0 * add Doxy comments for the init / finalize functions
This commit was SVN r1894.
2004-08-05 14:35:38 +00:00
Brian Barrett
dadb047103 * Add LLM (the rte resource allocator) framework
* Add hostfile component for the LLM (reads hostfiles, returns array of
  node identifiers

NOTES: 
  - This will require the full autogen / configure / make.
  - You now need flex to build Open MPI from Subversion.  The versions
    available on most Linux boxen and OS X is more than new enough.  You
    do *not* need flex to build from a nightly or release tarball.

This commit was SVN r1890.
2004-08-05 14:01:45 +00:00
Jeff Squyres
2399ac3cf8 Check for case where we *have* <stdbool.h>, but it doesn't *work*
(e.g., Portland compilers don't have/know about <stdbool.h>, but
AC_CHECK_HEADERS will typically find the GNU/gcc <stdbool.h>).
Reviewed by Brian.

This commit was SVN r1889.
2004-08-05 11:12:25 +00:00
Tim Prins
457a4b632c Some more changes to the oob:
All the user interface functions are now in mca/oob/base/base.h
Anyone who uses the oob should just include this file.

All component related functions have been moved to mca/oob/oob.h

The reason for this change was to make the user interface more clear.

This commit was SVN r1884.
2004-08-04 23:42:51 +00:00
Weikuan Yu
2478c6bc37 setup put. To do a chained QDMA for notification. Also PML changes is needed for this stuff to work
This commit was SVN r1883.
2004-08-04 23:18:45 +00:00
Laura Casswell
549a7cfba8 change order of includes so will compile
This commit was SVN r1881.
2004-08-04 22:47:25 +00:00
Denis Dimick
18cbe897d0 Added note about parallel builds
This commit was SVN r1880.
2004-08-04 22:25:47 +00:00
Tim Woodall
f100d08dd5 resolved another problem introduced with global name change
This commit was SVN r1879.
2004-08-04 22:09:40 +00:00