ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
97abb7c727
openmpi/config
История
Ralph Castain aec5cd08bd Per the PMIx RFC: 
WHAT:    Merge the PMIx branch into the devel repo, creating a new
               OPAL “lmix” framework to abstract PMI support for all RTEs.
               Replace the ORTE daemon-level collectives with a new PMIx
               server and update the ORTE grpcomm framework to support
               server-to-server collectives

WHY:      We’ve had problems dealing with variations in PMI implementations,
               and need to extend the existing PMI definitions to meet exascale
               requirements.

WHEN:   Mon, Aug 25

WHERE:  https://github.com/rhc54/ompi-svn-mirror.git

Several community members have been working on a refactoring of the current PMI support within OMPI. Although the APIs are common, Slurm and Cray implement a different range of capabilities, and package them differently. For example, Cray provides an integrated PMI-1/2 library, while Slurm separates the two and requires the user to specify the one to be used at runtime. In addition, several bugs in the Slurm implementations have caused problems requiring extra coding.

All this has led to a slew of #if’s in the PMI code and bugs when the corner-case logic for one implementation accidentally traps the other. Extending this support to other implementations would have increased this complexity to an unacceptable level.

Accordingly, we have:

* created a new OPAL “pmix” framework to abstract the PMI support, with separate components for Cray, Slurm PMI-1, and Slurm PMI-2 implementations.

* Replaced the current ORTE grpcomm daemon-based collective operation with an integrated PMIx server, and updated the grpcomm APIs to provide more flexible, multi-algorithm support for collective operations. At this time, only the xcast and allgather operations are supported.

* Replaced the current global collective id with a signature based on the names of the participating procs. The allows an unlimited number of collectives to be executed by any group of processes, subject to the requirement that only one collective can be active at a time for a unique combination of procs. Note that a proc can be involved in any number of simultaneous collectives - it is the specific combination of procs that is subject to the constraint

* removed the prior OMPI/OPAL modex code

* added new macros for executing modex send/recv to simplify use of the new APIs. The send macros allow the caller to specify whether or not the BTL supports async modex operations - if so, then the non-blocking “fence” operation is used, if the active PMIx component supports it. Otherwise, the default is a full blocking modex exchange as we currently perform.

* retained the current flag that directs us to use a blocking fence operation, but only to retrieve data upon demand

This commit was SVN r32570.
2014-08-21 18:56:47 +00:00
..
c_get_alignment.m4 Complete the cleanup of this file 2014-05-06 16:57:34 +00:00
c_weak_symbols.m4 configury: disable pshmem if buggy compiler is detected 2014-08-19 05:52:34 +00:00
distscript.csh Backout the ORCA commit. :( 2012-06-27 01:28:28 +00:00
libltdl-preopen-error.diff WARNING: Work on the temp branch being merged here encountered problems with bugs in subversion. Considerable effort has gone into validating the branch. However, not all conditions can be checked, so users are cautioned that it may be advisable to not update from the trunk for a few days to allow MTT to identify platform-specific issues. 2010-09-17 23:04:06 +00:00
ltmain_pgi_tp.diff Patch ltmain.sh in autogen.sh per this Libtool thread: 2010-08-25 19:40:17 +00:00
Makefile.am Sigh - forgot to change name in Makefile.am 2014-05-06 14:35:14 +00:00
Makefile.options Rollback the last two commits. 2008-10-28 17:22:29 +00:00
ompi_check_gni.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_check_libfca.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_check_libhcoll.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_check_lustre.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_check_mx.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_check_mxm.m4 MXM: update min required version for mxm as 2.1+ 2014-07-10 13:16:57 +00:00
ompi_check_plfs.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_check_psm.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_check_pvfs2.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_check_udapl.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
ompi_config_files.m4 Fortran: remove the scripts from the use-mpi-tkr implementation 2014-06-19 13:35:30 +00:00
ompi_config_solaris_threads.m4 Fix some fortran flags variable names, and fortran configury C test 2012-04-23 15:28:26 +00:00
ompi_config_threads.m4 Correct a typo to get the right configure option to match the internal logic, thus making --enable-mpi-thread-multiple to actually work 2014-01-15 03:56:29 +00:00
ompi_configure_options.m4 Move the OPAL_ENABLE_PROGRESS_THREADS definition to opal_configure_options.m4 2014-08-01 18:11:36 +00:00
ompi_contrib.m4 One more level of subsubsubsubsubtitle... 2014-05-07 15:51:20 +00:00
ompi_cxx_find_exception_flags.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_cxx_find_template_parameters.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_cxx_find_template_repository.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_cxx_have_exceptions.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_endpoint_tag.m4 Remove the proc_pml and proc_bml fields from ompi_proc_t and replace with a 2013-08-30 16:54:55 +00:00
ompi_ext.m4 One more level of subsubsubsubsubtitle... 2014-05-07 15:51:20 +00:00
ompi_find_mpi_aint_count_offset.m4 Add file that was forgotten in r30776. Refs trac:4205. 2014-02-20 11:32:57 +00:00
ompi_fortran_check_abstract.m4 Minor configure output consistency updates 2014-01-13 18:28:26 +00:00
ompi_fortran_check_asynchronous.m4 Minor configure output consistency updates 2014-01-13 18:28:26 +00:00
ompi_fortran_check_bind_c.m4 ompi_fortran_check_bind_c.m4: Fix comment in existing .m4 file 2014-06-19 13:39:11 +00:00
ompi_fortran_check_f08_assumed_rank.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_check_ignore_tkr.m4 Thanks to a reminder from Tobias Burunus, commit support for the 2014-01-08 03:46:27 +00:00
ompi_fortran_check_logical_array.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_check_optional_args.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_check_private.m4 Minor configure output consistency updates 2014-01-13 18:28:26 +00:00
ompi_fortran_check_procedure.m4 Missed this one in r30271. 2014-01-13 18:30:15 +00:00
ompi_fortran_check_protected.m4 Make the use of PROTECTED in the mpi_f08 module be optional. 2014-01-13 18:35:42 +00:00
ompi_fortran_check_real16_c_equiv.m4 Per RFC, continue renaming of build tools: 2014-05-06 15:01:34 +00:00
ompi_fortran_check_type.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_check.m4 Per RFC, continue the naming conversion 2014-05-06 23:34:33 +00:00
ompi_fortran_find_ext_symbol_convention.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_find_module_include_flag.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_get_alignment.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_get_handle_max.m4 On OpenBSD (and probably others), expr operates on a 32 bit signed integer, 2012-12-22 16:24:36 +00:00
ompi_fortran_get_kind_value.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_get_sizeof.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_fortran_get_value_true.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
ompi_interix.m4 Per RFC: 2014-05-05 22:35:23 +00:00
ompi_microsoft.m4 Per RFC, continue renaming of build tools: 2014-05-06 15:01:34 +00:00
ompi_setup_contrib.m4 Per RFC: 2014-05-05 22:35:23 +00:00
ompi_setup_cxx.m4 Per RFC, continue the joyous fun of the renaming exercise 2014-05-06 20:13:37 +00:00
ompi_setup_fc.m4 Per RFC, continue the joyous fun of the renaming exercise 2014-05-06 20:13:37 +00:00
ompi_setup_java.m4 Per RFC: 2014-05-05 22:35:23 +00:00
ompi_setup_mpi_ext.m4 Per RFC: 2014-05-05 22:35:23 +00:00
ompi_setup_mpi_fortran.m4 fortran: Revert r32354 2014-07-30 22:48:49 +00:00
ompi_setup_mpi_profiling.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
opal_case_sensitive_fs_setup.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
opal_check_attributes.m4 Per RFC, continue cleanup with minor change to one file 2014-05-06 18:30:55 +00:00
opal_check_broken_qsort.m4 Continue the renaming, fix ompi_show_subsubtitle 2014-05-07 15:45:47 +00:00
opal_check_cma.m4 Cleanup some autoconf messages - thanks to Paul Hargrove for noting them 2014-08-05 14:48:42 +00:00
opal_check_compiler_version.m4 The bulk of the remaining renaming changes, in one final glorious "blob". Thanks to Jeff for some help chasing down a few spots. Per chat with Jeff, we decided to cleanup a few things that were historical in nature: 2014-05-07 21:48:53 +00:00
opal_check_compiler_works.m4 Per RFC, continue the renaming effort 2014-05-06 21:16:29 +00:00
opal_check_cuda.m4 Chnage some CUDA configure code and macro names per review request by jsquyres in ticket #3880. 2013-12-06 14:35:10 +00:00
opal_check_db.m4 Improve/expand the database configure logic for future use 2014-03-11 03:01:15 +00:00
opal_check_icc.m4 Per RFC, continue with build system renaming 2014-05-06 19:22:27 +00:00
opal_check_ident.m4 Per RFC, continue with build system renaming 2014-05-06 19:22:27 +00:00
opal_check_offsetof.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
opal_check_openfabrics.m4 Rename the .m4 files to reflect the move to opal so non-MPI projects properly see them 2014-08-01 17:03:34 +00:00
opal_check_os_flavors.m4 Ensure we properly check for sockaddr_in before checking for tcp 2014-03-20 00:17:29 +00:00
opal_check_package.m4 Per RFC, continue with build system renaming 2014-05-06 19:37:10 +00:00
opal_check_pmi.m4 Per the PMIx RFC: 2014-08-21 18:56:47 +00:00
opal_check_portals4.m4 Rename the .m4 files to reflect the move to opal so non-MPI projects properly see them 2014-08-01 17:03:34 +00:00
opal_check_portals.m4 Rename the .m4 files to reflect the move to opal so non-MPI projects properly see them 2014-08-01 17:03:34 +00:00
opal_check_ps.m4 Per RFC, continue the renaming effort 2014-05-06 21:16:29 +00:00
opal_check_pthread_pids.m4 Clean up variables used in configure with OPAL_VAR_SCOPE. 2013-12-06 23:40:34 +00:00
opal_check_ugni.m4 Per RFC, continue the renaming process 2014-05-06 20:53:55 +00:00
opal_check_vendor.m4 Per RFC, continue the joyous fun of the renaming exercise 2014-05-06 20:13:37 +00:00
opal_check_verbs.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
opal_check_visibility.m4 Per RFC, continue renaming of build tools: 2014-05-06 15:01:34 +00:00
opal_check_withdir.m4 Per RFC, next major step in cleaning up the build system naming patterns: rename files containing things used by the OPAL layer to be opal_foo.m4 instead of ompi_foo.m4. The ALPS plm is currently checking UGNI, so shift the check_ugni.m4 to orte for now. 2014-05-06 03:20:16 +00:00
opal_config_asm.m4 opal_config_asm.m4: fix typo in output message 2014-08-11 13:28:17 +00:00
opal_config_files.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
opal_config_pthreads.m4 silence warnings 2014-08-11 07:41:54 +00:00
opal_config_subdir_args.m4 Per RFC, continue pecking away at the build system renaming 2014-05-06 16:27:38 +00:00
opal_config_subdir.m4 Per RFC, continue pecking away at the build system renaming 2014-05-06 16:27:38 +00:00
opal_config_threads.m4 Per RFC, continue the renaming effort 2014-05-06 21:16:29 +00:00
opal_configure_options.m4 Move the OPAL_ENABLE_PROGRESS_THREADS definition to opal_configure_options.m4 2014-08-01 18:11:36 +00:00
opal_ensure_contains_optflags.m4 Per RFC, next major step in cleaning up the build system naming patterns: rename files containing things used by the OPAL layer to be opal_foo.m4 instead of ompi_foo.m4. The ALPS plm is currently checking UGNI, so shift the check_ugni.m4 to orte for now. 2014-05-06 03:20:16 +00:00
opal_find_type.m4 Per RFC, continue renaming project 2014-05-07 01:00:06 +00:00
opal_functions.m4 Per RFC, continue with renaming 2014-05-07 04:51:45 +00:00
opal_get_version.m4 The final step of the RFC: convert the <foo>libdir and friends to fit their respective code areas, and equate them all at the top. Note that we can't entirely separate things as the opal_install_dirs framework can't handle separated locations for the various trees. 2014-05-08 02:01:35 +00:00
opal_get_version.m4sh The final step of the RFC: convert the <foo>libdir and friends to fit their respective code areas, and equate them all at the top. Note that we can't entirely separate things as the opal_install_dirs framework can't handle separated locations for the various trees. 2014-05-08 02:01:35 +00:00
opal_lang_link_with_c.m4 Per RFC, continue the joyous fun of the renaming exercise 2014-05-06 20:13:37 +00:00
opal_load_platform.m4 Per RFC, continue the renaming effort 2014-05-06 21:16:29 +00:00
opal_mca_priority_sort.pl Per RFC, next major step in cleaning up the build system naming patterns: rename files containing things used by the OPAL layer to be opal_foo.m4 instead of ompi_foo.m4. The ALPS plm is currently checking UGNI, so shift the check_ugni.m4 to orte for now. 2014-05-06 03:20:16 +00:00
opal_mca.m4 Remove stale orcm reference 2014-07-11 19:00:28 +00:00
opal_save_version.m4 Per RFC, next major step in cleaning up the build system naming patterns: rename files containing things used by the OPAL layer to be opal_foo.m4 instead of ompi_foo.m4. The ALPS plm is currently checking UGNI, so shift the check_ugni.m4 to orte for now. 2014-05-06 03:20:16 +00:00
opal_search_libs.m4 opal_search_libs.m4: improvements from Dave Goodell's review 2014-08-08 22:22:03 +00:00
opal_setup_cc.m4 Per RFC, continue the joyous fun of the renaming exercise 2014-05-06 20:13:37 +00:00
opal_setup_component_package.m4 Per RFC, continue the renaming effort 2014-05-06 21:16:29 +00:00
opal_setup_cxx.m4 Per RFC, continue the joyous fun of the renaming exercise 2014-05-06 20:13:37 +00:00
opal_setup_ft.m4 Per RFC, another round of changes 2014-05-07 03:16:59 +00:00
opal_setup_java.m4 Per RFC: 2014-05-05 22:35:23 +00:00
opal_setup_libltdl.m4 The bulk of the remaining renaming changes, in one final glorious "blob". Thanks to Jeff for some help chasing down a few spots. Per chat with Jeff, we decided to cleanup a few things that were historical in nature: 2014-05-07 21:48:53 +00:00
opal_setup_wrappers.m4 The bulk of the remaining renaming changes, in one final glorious "blob". Thanks to Jeff for some help chasing down a few spots. Per chat with Jeff, we decided to cleanup a few things that were historical in nature: 2014-05-07 21:48:53 +00:00
opal_strip_optflags.m4 Per RFC, continue the build system renaming 2014-05-06 20:30:37 +00:00
opal_try_assemble.m4 Per RFC, continue the build system renaming 2014-05-06 20:30:37 +00:00
orte_check_alps.m4 Per RFC, continue the renaming process 2014-05-06 20:53:55 +00:00
orte_check_loadleveler.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
orte_check_lsf.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
orte_check_sge.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
orte_check_slurm.m4 Consolidate all the opal/orte/ompi .m4 files back to the top-level 2012-12-19 00:00:36 +00:00
orte_check_tm.m4 Last step under current RFC: OMPI_CHECK_WITHDIR -> OPAL_CHECK_WITHDIR 2014-05-01 15:38:07 +00:00
orte_config_files.m4 Sigh - remove the c++ cruft from here too 2014-02-10 17:32:02 +00:00
orte_configure_options.m4 Per RFC: 2014-05-05 22:35:23 +00:00
orte_setup_debugger_flags.m4 Per RFC, continue the build system renaming 2014-05-06 20:30:37 +00:00
orte_setup_java.m4 Per RFC: 2014-05-05 22:35:23 +00:00
oshmem_config_files.m4 Java - define macro OSHMEM_SETUP_JAVA_BINDINGS 2014-06-11 21:49:31 +00:00
oshmem_configure_options.m4 configury: disable pshmem if buggy compiler is detected 2014-08-19 05:52:34 +00:00
oshmem_setup_java.m4 Java - define macro OSHMEM_SETUP_JAVA_BINDINGS 2014-06-11 21:49:31 +00:00
pkg.m4 add pkg.m4 for PKG_CHK_MODULES, etc. macro 2014-08-05 21:43:58 +00:00