1
1
Граф коммитов

2585 Коммитов

Автор SHA1 Сообщение Дата
Jeff Squyres
20349da03b usnic: minor cleanup
This commit was SVN r32367.
2014-07-30 20:56:49 +00:00
Jeff Squyres
f1fb4970a5 usnic: remove all trailing whitespace
Style cleanup only; no code changes.

This commit was SVN r32366.
2014-07-30 20:56:15 +00:00
Jeff Squyres
a6c6aa2815 usnic: remove compatibility with v1.6 series
Update compat.h to only handle compatability between v1.7/v1.8 and
v1.9/2.0 (i.e., the current trunk).  Remove what seems to be the last
vestiages of OMPI/ORTE pollution in the now-OPAL-ized usnic BTL.

Currently use a hard-coded constant for the MCW size (i.e.,
MPI_COMM_WORLD size) for some initialization values in the v1.9/2.0
series; still need to figure out something better there.

This commit was SVN r32365.
2014-07-30 20:55:26 +00:00
Jeff Squyres
aed8b43db2 usnic: remove last vestiges of ORTE process names
These were missed when the BTLs moved to OPAL.

This commit was SVN r32364.
2014-07-30 20:53:41 +00:00
Jeff Squyres
d195b8caf4 usnic: fix a bunch of OMPI_* and ORTE_* constant usage
Somehow a bunch of OMPI_* and ORTE_* constants didn't get renamed in
the BTL move to OPAL.  This commit fixes that.

This commit was SVN r32363.
2014-07-30 20:52:54 +00:00
Jeff Squyres
2447c8479f usnic: do not call ompi_rte_abort()
Adapt to moving down to OPAL: find a PML-registered error callback,
and use that when we don't have a module context and we need to abort.

Failing that, just call exit().

This commit was SVN r32362.
2014-07-30 20:52:06 +00:00
George Bosilca
f39abb9e69 Reverting r32355: a number of processes is not a notion that a low level
communication library should use to initialize itself.

Ralph will champion this change back with an RFC if there is a realistic
need/use case from the community.

This commit was SVN r32361.

The following SVN revision numbers were found above:
  r32355 --> open-mpi/ompi@c903917f47
2014-07-30 20:11:35 +00:00
Nathan Hjelm
a32d93ec20 mca/base: make clang static analyzer happy
cmr=v1.8.3:reviewer=jsquyres

This commit was SVN r32360.
2014-07-30 17:45:28 +00:00
Nathan Hjelm
97fad1dd95 mca/base: ensure component version parameters get deregistered when the
component gets dlclosed

cmr=v1.8.2:reviewer=rhc

This commit was SVN r32359.
2014-07-30 17:45:23 +00:00
Rolf vandeVaart
01c2c1c78c Increase default minimum sm pool size. Discussed in this RFC:
http://www.open-mpi.org/community/lists/devel/2014/07/15274.php

cmr=v1.8.3:reviewer=bosilca

This commit was SVN r32356.
2014-07-30 14:01:08 +00:00
Ralph Castain
c903917f47 Expose the num_procs information to the opal layer as the info is needed in several BTLs
This commit was SVN r32355.
2014-07-30 09:33:41 +00:00
Gilles Gouaillardet
b2dbde2be4 btl/scif : fix compilation
btl/scif could not compile after it was moved down into opal.
this commit fixes the compilation issue *only*

This commit was SVN r32351.
2014-07-30 04:26:50 +00:00
Joshua Ladd
da286d9a84 Fixing the OpenIB receive queue selection logic. Refs trac:4816
This commit was SVN r32350.

The following Trac tickets were found above:
  Ticket 4816 --> https://svn.open-mpi.org/trac/ompi/ticket/4816
2014-07-30 02:17:46 +00:00
Jeff Squyres
10a58992af usnic: add --with-usnic configure switch
If --with-usnic is specified and we can't build the usnic BTL, abort.
If --without-usnic is specified, gracefully skip building the usnic
BTL.  If neither is specified, do the OMPI-default behavior: try to
configure/build the usnic BTL, and if we can't, skip it.
    
Fixes CSCuq13889.

This commit was SVN r32349.
2014-07-30 02:11:04 +00:00
Joshua Ladd
7a694cb95e Reverting r32346. Will do it the correct way and update the patch in Refs trac:32346
This commit was SVN r32348.

The following SVN revision numbers were found above:
  r32346 --> open-mpi/ompi@5be6ff07d5

The following Trac tickets were found above:
  Ticket 32346 --> https://svn.open-mpi.org/trac/ompi/ticket/32346
2014-07-29 23:23:44 +00:00
Joshua Ladd
5be6ff07d5 This fixes the OpenIB BTL receive queue selection logic in the trunk. Custom patch for 1.8.2 is provided in Refs trac:4816
This commit was SVN r32346.

The following Trac tickets were found above:
  Ticket 4816 --> https://svn.open-mpi.org/trac/ompi/ticket/4816
2014-07-29 21:42:20 +00:00
Nathan Hjelm
b8c3b01643 Fix some missed ompi->opal renamings
This commit was SVN r32345.
2014-07-29 18:59:59 +00:00
Nathan Hjelm
b10a29fd96 btl/vader: Fix typo in add_procs introduced by the btl move
This commit was SVN r32342.
2014-07-29 16:07:59 +00:00
Rolf vandeVaart
e39c673dd2 Fix typo.
This commit was SVN r32339.
2014-07-29 13:21:24 +00:00
Ralph Castain
d674e22433 Remove stale include as header no longer exists. Add missing header
This commit was SVN r32336.
2014-07-29 01:24:27 +00:00
Nathan Hjelm
0e47441333 Remove unused files
This commit was SVN r32335.
2014-07-28 22:01:16 +00:00
Nathan Hjelm
dffbcb3803 User opal_process_info.cpuset to determine if the process is bound
This commit was SVN r32334.
2014-07-28 22:00:18 +00:00
Nathan Hjelm
1407c1f501 Remove RML code from common/sm
The only user of this code was coll/sm. I implemented a basic replacement
for the removed code. This gets the trunk compiling again with
--disable-dlopen.

This commit was SVN r32333.
2014-07-28 22:00:12 +00:00
Nathan Hjelm
603ba71b0d Ignore components broken by the BTL move.
common/ofacm is only used by the iboffload code in ompi. This code does
not currently work so it is safe to ignore these components until it is
fixed.

This commit was SVN r32331.
2014-07-28 21:24:18 +00:00
Ryan Grant
caa10a5faf Portals fixes after latest move
This commit was SVN r32330.
2014-07-28 19:25:03 +00:00
George Bosilca
68c4ecbe06 Even less ompi usages down here.
This commit was SVN r32328.
2014-07-26 22:21:08 +00:00
George Bosilca
3d5578d6df No vestiges of the RTE or OMPI left.
This commit was SVN r32327.
2014-07-26 22:19:49 +00:00
George Bosilca
4d3d421b06 Cleanup (not tested).
This commit was SVN r32326.
2014-07-26 22:19:07 +00:00
George Bosilca
f217661ee0 Use opal_process_info whenever possible. Some other minor cleanups.
This commit was SVN r32325.
2014-07-26 21:48:23 +00:00
George Bosilca
a3feb627cf Move some of the ompi_process_info down in OPAL.
This commit was SVN r32324.
2014-07-26 21:43:34 +00:00
George Bosilca
85b89795ff Rename all the win32 files to use the correct layer name.
This commit was SVN r32323.
2014-07-26 20:17:09 +00:00
George Bosilca
de2cb71790 Rename the help file.
This commit was SVN r32322.
2014-07-26 20:16:28 +00:00
Ralph Castain
76d82b885f Correctly dereference the thread object
This commit was SVN r32321.
2014-07-26 17:01:27 +00:00
Ralph Castain
fbed15072c Add missing rename of file
This commit was SVN r32318.
2014-07-26 01:21:07 +00:00
Ralph Castain
552c9ca5a0 George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-)
WHAT:    Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL

All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies.  This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP.  Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose.  UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs.  A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic.

This commit was SVN r32317.
2014-07-26 00:47:28 +00:00
Ralph Castain
408474d2d8 Remove some CRLFs so git quits changing them
This commit was SVN r32278.
2014-07-22 16:56:24 +00:00
Ralph Castain
c1bb5b68d0 It is possible to have a "standard" progress thread, so simplify the usage of the opal_progress_thread code.
This commit was SVN r32277.
2014-07-22 16:55:23 +00:00
George Bosilca
0b72cdabfd This is basically a revert of r32236. What looked as a good patch
turned out to be a nightmare, as the pointers to the classes are
located in shared libraries memory areas, and are not accesible
after the shared library is unloaded. Thus, OPAL cannot cleanup
the left-over classes from the other shared libraries.

This commit was SVN r32248.

The following SVN revision numbers were found above:
  r32236 --> open-mpi/ompi@59017433e1
2014-07-15 18:21:01 +00:00
George Bosilca
7c21491858 Fix the indentation. Allow for deregistration of OPAL params.
This commit was SVN r32242.
2014-07-15 05:20:26 +00:00
George Bosilca
2f6bc76dc1 Be symmetric to the opal_init.
This commit was SVN r32241.
2014-07-15 05:05:26 +00:00
George Bosilca
ca76fb69df Start with a sane value.
This commit was SVN r32240.
2014-07-15 05:04:55 +00:00
George Bosilca
77c74e8872 Don't iniaitlize twice the event framework (it is already initialized
the init_tool).

This commit was SVN r32239.
2014-07-15 05:04:29 +00:00
George Bosilca
ed3d98a76d Up to strlen and not to sizeof. This is guaranteed to work as
in the worst case we just forced a \0 at the end of the string.

This commit was SVN r32238.
2014-07-15 05:03:06 +00:00
George Bosilca
a648fcdeb0 Upon close reset the search_dirs.
This commit was SVN r32237.
2014-07-15 05:02:19 +00:00
George Bosilca
59017433e1 Allow for class allocation/deallocation.
This commit was SVN r32236.
2014-07-15 05:01:36 +00:00
Mike Dubman
0db23b0210 BUILD: support new automake
new automake requires subdirs-object directive, to resolve this:

09:43:37 automake: warning: possible forward-incompatibility.
09:43:37 automake: At least a source file is in a subdirectory, but the 'subdir-objects'
09:43:37 automake: automake option hasn't been enabled.  For now, the corresponding output
09:43:37 automake: object file(s) will be placed in the top-level directory.  However,
09:43:37 automake: this behaviour will change in future Automake versions: they will
09:43:37 automake: unconditionally cause object files to be placed in the same subdirectory
09:43:37 automake: of the corresponding sources.
09:43:37 automake: You are advised to start using 'subdir-objects' option throughout your
09:43:37 automake: project, to avoid future incompatibilities.
09:43:37 tools/otfmerge/Makefile.common:13: warning: source file '$(OTFMERGESRCDIR)/otfmerge.c' is in a subdirectory,
09:43:37 tools/otfmerge/Makefile.common:13: but option 'subdir-objects' is disabled

cmr=v1.8.2:reviewer=ompi-rm1.8

This commit was SVN r32225.
2014-07-12 12:38:15 +00:00
Mike Dubman
e342a11c2e opal envlist mca: implement Jeff`s quibbles
fixed by Elena, reviewed by Miked

This commit was SVN r32216.
2014-07-11 07:23:20 +00:00
Ralph Castain
60da1456d9 Silence unused var warning
This commit was SVN r32187.
2014-07-09 22:37:22 +00:00
Ralph Castain
354a1d07a6 Silence warning about set-but-unused var
This commit was SVN r32186.
2014-07-09 22:37:05 +00:00
Mike Dubman
6efc9a7329 opal mca: externalize delimiter for base_env_list mca parameters
fixed by Elena, reviewed by MikeD

This commit was SVN r32182.
2014-07-09 18:55:49 +00:00