18.2. Open MPI v5.0.x series
This file contains all the NEWS updates for the Open MPI v5.0.x series, in reverse chronological order.
18.2.1. Open MPI version 5.0.0rc10
27 January 2023
MPIR API has been removed
As was announced in summer 2017, Open MPI has removed support of MPIR-based tools beginning with the release of Open MPI v5.0.0.
The new PRRTE based runtime environment supports PMIx-tools API instead of the legacy MPIR API for debugging parallel jobs.
See https://github.com/openpmix/mpir-to-pmix-guide for more information.
zlib is suggested for better user experience
PMIx will optionally use zlib to compress large data streams. This may result in faster startup times and smaller memory footprints (compared to not using compression). The Open MPI community recommends building zlib support with PMIx, regardless of whether you are using an externally-installed PMIx or the PMIx that is installed with Open MPI.
Open MPI no longer builds 3rd-party packages such as Libevent, HWLOC, PMIx, and PRRTE as MCA components and instead:
Relies on external libraries whenever possible, and
Builds the 3rd party libraries only if needed, and as independent libraries, rather than linked into the Open MPI core libraries.
Changes since rc9:
coll/ucc: Added support for Scatter and Iscatter collectives.
Added cache bypass mechanism to the
OFIBTL. This fixes conflicts with Libfabric, which has its own registration cache. This addes a bypass flag which can be used for providers known to have their own registration cache.
common/ompio: implement pipelined read and write operation. This new new code path shows significant performance improvements for reading/writing device buffers compared to the previous implementation, and reduces the memory footprint of
OMPIOby allocating smaller temporary buffers.
32-bit builds have been disabled. Building Open MPI in a 32-bit environment is no longer supported.
MPI-4: MPI_Info_get() and MPI_Infi_get_valuelen() are now deprecated.
MPI-4: Issue a deprecation warning when MPI_Cancel() is called for a non-blocking send request.
Fixed various bugs encountered when running MPI under a debugger.
Cleaned up a number of memory leaks.
Cleaned up global symbol pollution that was leaking out of libmpi.
Removed opal_list_insert(), it was buggy and not used. Thanks to Jinyuan Wang for the contribution.
Many other bug fixes and cleanups.
Various documentation updates.
All other notable updates for v5.0.0:
Updated PMIx to the
v4.2branch - current hash:
Updated PRRTE to the
v3.0branch - current hash:
ULFM Fault Tolerance support has been added. See the ULFM section
CUDAis now supported in the
--mca ompi_display_comm mpi_init/
mpi_finalizehas been added. This enables a communication protocol report: when
MPI_Initis invoked (using the
mpi_initvalue) and/or when
MPI_Finalizeis invoked (using the
The threading framework has been added to allow building OMPI with different threading libraries. It currently supports Argobots, Qthreads, and Pthreads. See the
--with-threadsoption in the
configurecommand. Thanks to Shintaro Iwasaki and Jan Ciesko for their contributions to this effort.
New Thread Local Storage API: Removes global visibility of TLS structures and allows for dynamic TLS handling.
CUDAspecific code was replaced with a generic framework that standardizes various device features such as copies or pointer type detection. This allows for modularized implementation of various devices such as the newly introduced ROCm Accelerator component. The redesign also allows for Open MPI builds to be shipped with
CUDAsupport enabled without requiring
Added load-linked, store-conditional atomics support for AArch64.
Added atomicity support to the
Added support for MPI minimum alignment key to the one-sided
Add ability to detect patched memory to
memory_patcher. Thanks to Rich Welch for the contribution.
MPI-4.0 updates and additions:
MPI Sessionshas been added.
Added partitioned communication using persistent sends and persistent receives.
Added persistent collectives to the
MPI_namespace (they were previously available via the
MPI_Isendrecv()and its variants.
Added support for
Added support for
Added support for
Added error handling for “unbound” errors to
MPI_Win_get_info()compliant to the standard.
Droped unknown/ignored info keys on communicators, files, and windows.
Initial implementations of
Transport updates and improvements
Many MPI one-sided and RDMA emulation fixes for the
This patch series fixs many issues when running with
--mca osc rdma --mca btl tcp, IE - TCP support for one sided MPI calls.
Many MPI one-sided fixes for the
Added support for
acc_single_intrinsicto the one-sided
Removed the legacy
pt2ptone-sided component. Users should use the
rdmaone-sided component instead with the
tcpBTL and/or other BTLs to use MPI one sided-calls via TCP transport.
tcpBTL to use graph solving for global interface matching between peers in order to improve
Changes to the BTL
OFIcomponent to better support the HPE SS11 network.
sm(shared memory) BTL has been removed. The next-generation shared memory BTL
vaderreplaces it, and has been renamed to be
vaderwill still work as an alias).
Update the new
smBTL to not use Linux Cross Memory Attach (CMA) in user namespaces.
Fixed a crash when using the new
smBTL when compiled with Linux Cross Memory Attach (
XPMEM). Thanks to George Katevenis for reporting this issue.
-mca pmloption to only accept one pml, not a list.
Deprecations and removals:
ORTE, the underlying OMPI launcher has been removed, and replaced with The PMIx Reference RunTime Environment (
PMI support has been removed from Open MPI; now only PMIx is supported. Thanks to Zach Osman for removing config/opal_check_pmi.m4.
Removed transports PML
ikritcomponents. These transports are no longer supported, and are replaced with
Removed all vestiges of Checkpoint Restart (C/R) support.
32 bit atomics are now only supported via C11 compliant compilers.
Explicitly disable support for GNU gcc < v4.8.1 (note: the default gcc compiler that is included in RHEL 7 is v4.8.5).
Various atomics support removed: S390/s390x, Sparc v9, ARMv4 and ARMv5 with CMA support.
The MPI C++ bindings have been removed.
The mpirun options
--amcaoptions have been deprecated.
libompitrace. This library was incomplete and unmaintained. If needed, it is available in the v4/v4.1 series.
The rankfile format no longer supports physical processor locations. Only logical processor locations are supported.
Open MPI now requires HWLOC v1.11.0 or later.
The internal HWLOC shipped with OMPI has been updated to v2.7.1.
Enable –enable-plugins when appropriate.
Documentation updates and improvements:
Open MPI now uses readthedocs.io for all documentation.
Converted man pages to markdown. Thanks to Fangcong Yin for their contribution to this effort.
HACKING.mdfixes - thanks to: Yixin Zhang, Samuel Cho, Robert Langfield, Alex Ross, Sophia Fang, mitchelltopaloglu, Evstrife, Hao Tong and Lachlan Bell for their contributions.
Various CUDA documentation fixes. Thanks to Simon Byrne for finding and fixing these typos.
Build updates and fixes:
Various changes and cleanup to fix, and better support the static building of Open MPI.
Change the default component build behavior to prefer building components as part of the core Open MPI library instead of individual DSOs. Currently, this means the Open SHMEM layer will only build if the UCX library is found.
autogen.plnow supports a
-joption to run multi-threaded. Users can also use the environment variable
autogen.plto support macOS Big Sur. Thanks to @fxcoudert for reporting the issue.
Fixed bug where
autogen.plwould not ignore all excluded components when using the
Fixed a bug the
buildrpm.shwhich would result in an rpm build failure. Thanks to John K. McIver III for reporting and fixing.
C++compiler requirement to build Open MPI.
Updates to improve the handling of the compiler version string in the build system. This fixes a compiler error with clang and armclang.
Added OpenPMIx binaries to the build, including
pmix_info. Thanks to Mamzi Bayatpour for their contribution to this effort.
Open MPI now links to Libevent using
Added support for setting the wrapper C compiler. This adds a new option:
Fixed compilation errors when running on IME file systems due to a missing header inclusion. Thanks to Sylvain Didelot for finding and fixing this issue.
Add support for GNU Autoconf v2.7.x.
Other updates and bug fixes:
Updated Open MPI to use
Fixed Fortran-8-byte-INTEGER vs. C-4-byte-int issue in the
mpi_f08MPI Fortran bindings module. Thanks to @ahaichen for reporting the bug.
Fixed Fortran keyword issue when compiling
oshmem_info. Thanks to Pak Lui for finding and fixing the bug.
Added check for Fortran
ISO_FORTRAN_ENV:REAL16. Thanks to Jeff Hammond for reporting this issue.
Fixed Fortran preprocessor issue with CPPFLAGS. Thanks to Jeff Hammond for reporting this issue.
MPI module: added the mpi_f08 TYPE(MPI_*) types for Fortran. Thanks to George Katevenis for the report and their contribution to the patch.
Fixed a typo in an error string when showing the stackframe. Thanks to Naribayashi Akira for finding and fixing the bug.
Fixed output error strings and some comments in the Open MPI code base. Thanks to Julien Emmanuel for finding and fixing these issues.
uctBTL transport now supports
UCXv1.9 and higher. There is no longer a maximum supported version.
Updated the UCT BTL defaults to allow Mellanox HCAs (
mlx5_0) for compatibility with the one-sided
Fixed a crash during CUDA initialization. Thanks to Yaz Saito for finding and fixing the bug.
MPI_Comm_spawn()support has been fixed.
PowerPC atomics: Force usage of ppc assembly by default.
The default atomics have been changed to be GCC, with C11 as a fallback. C11 atomics incurs sequential memory ordering, which in most cases is not desired.
Various datatype bugfixes and performance improvements.
Various pack/unpack bugfixes and performance improvements.
Various OSHMEM bugfixes and performance improvements.
New algorithm for Allgather and Allgatherv has been added, based on the paper “Sparbit: a new logarithmic-cost and data locality-aware MPI Allgather algorithm”. Default algorithm selection rules are un-changed, to use these algorithms add:
--mca coll_tuned_allgather_algorithm sparbitand/or
--mca coll_tuned_allgatherv_algorithm sparbitto your
mpiruncommand. Thanks to: Wilton Jaciel Loch, and Guilherme Koslovski for their contribution.
Updated the usage of .gitmodules to use relative paths from absolute paths. This allows the submodule cloning to use the same protocol as OMPI cloning. Thanks to Felix Uhl for the contribution.