SLFO-pool/mvapich2
SHA256
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
main
mvapich2/mvapich2.changes

852 lines
36 KiB
Plaintext
Raw Permalink Blame History

-------------------------------------------------------------------
Thu Oct 26 10:42:07 UTC 2023 - Nicolas Morey <nicolas.morey@suse.com>
- Add mvapich2-openpa-add-memory-barriers.patch to fix testsuite issue
on pcc64 (bsc#1216610, bsc#1216612)
-------------------------------------------------------------------
Mon Aug 7 16:57:40 UTC 2023 - Nicolas Morey <nicolas.morey@suse.com>
- Drop support for obsolete TrueScale (bsc#1212146)
-------------------------------------------------------------------
Mon Dec 5 21:11:29 UTC 2022 - Stefan Brüns <stefan.bruens@rwth-aachen.de>
- Reduce constraints to match the actual requirement. Exaggerating
the requirements hurts both this package (time until build can
start) as well as other OBS users (blocking large workers
without need).
- Use a reproducible timestamp instead of removing it altogether.
-------------------------------------------------------------------
Mon Nov 28 08:30:49 UTC 2022 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Update reproducible.patch to remove timestamp generated at compilation time
-------------------------------------------------------------------
Wed Jul 6 15:08:09 UTC 2022 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Add mvapich2-allow-building-with-external-hwloc.patch
to allow building against an external hwloc library
- Build mvapich2 HPC flavors against pmix and hwloc system libraries
-------------------------------------------------------------------
Wed Jun 29 12:54:33 UTC 2022 - Klaus Kämpf <kkaempf@suse.com>
- add pass-correct-size-to-snprintf.patch to fix potential buffer
overflows (required to make 'sundials' testsuite pass)
- Update to mvapich2 2.3.7
* Features and Enhancements (since 2.3.6):
- Added support for systems with Rockport's switchless networks
* Added automatic architecture detection
* Optimized performance for point-to-point operations
- Added support for the Cray Slingshot 10 interconnect
- Enhanced support for blocking collective offload using
Mellanox SHARP
* Scatter and Scatterv
- Enhanced support for non-blocking collective offload using
Mellanox SHARP
* Iallreduce, Ibarrier, Ibcast, and Ireduce
* Bug Fixes (since 2.3.6):
- Removed several deprectated functions
- Thanks to Honggang Li @RedHat for the report
- Fixed a bug where tools like CMake FindMPI would not
detect MVAPICH when compiled without Hydra mpiexec
- Thanks to Chris Chambreau and Adam Moody @LLNL for the report
- Fixed compilation error when building with mpirun and without hydra
- Thanks to James Long @University of Illinois for the report
- Fixed issue with setting RoCE mode correctly without RDMA_CM.
- Thanks to Nicolas Gagnon @Rockport Networks for the report
- Fixed an issue on heterogeneous clusters where QP attributes were
set incorrectly
- Thanks to X-ScaleSolutions for the report and fix
- Fixed a memory leak in improbe on the PSM channel
- Thanks to Gregory Lee @LLNL Beichuan Yan @University of Colorado
for the report
- Added retry logic for PSM connection establishment
- Thanks to Gregory Lee @LLNL for the report and X-ScaleSolutions
for the patch
- Fixed an initialization error when using PSM and gcc's -pg option
- Thanks to Gregory Lee @LLNL for the report and X-ScaleSolutions for
the patch
- Fixed a potential integer overflow when transfering large arrays
- Thanks to Alexander Melnikov for the report and patch
- Fix Url: link
-------------------------------------------------------------------
Wed Feb 16 11:07:12 UTC 2022 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Disable dlopen for verbs library (bsc#1196019)
-------------------------------------------------------------------
Tue Oct 19 15:20:53 UTC 2021 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Move rpm macros to %_rpmmacrodir (bsc#1191386)
-------------------------------------------------------------------
Tue Sep 28 07:20:18 UTC 2021 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Remove obsolete python dependency (bsc#1190996)
-------------------------------------------------------------------
Tue May 18 08:08:52 UTC 2021 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Update to mvapich2 2.3.6
- Enhanced performance for UD-Hybrid code
- Add multi-rail support for UD-Hybrid code
- Enhanced performance for shared-memory collectives
- Enhanced job-startup performance for flux job launcher
- Use PMI2 by default when SLURM is selected as process manager
- Add support to use aligned memory allocations for multi-threaded
applications
- Architecture detection and enhanced point-to-point tuning for
Oracle BM.HPC2 cloud shape
- Add support for GCC compiler v11
- Update hwloc v1 code to v1.11.14
- Update hwloc v2 code to v2.4.2
- Drop obsolete patches:
- fix-missing-return-code.patch as it was fixed upstream
- mvapich2-remove-deprecated-sys_siglist.patch
- rdma_find_network_type-return-MV2_NETWORK_CLASS_UNKNOWN-when-dev_list-is-freed.patch
- Refresh reproducible.patch
-------------------------------------------------------------------
Wed Mar 24 10:06:26 UTC 2021 - Egbert Eich <eich@suse.com>
- Update mvapich2 to 2.3.5.
* Enhanced performance for MPI_Allreduce and MPI_Barrier
* Support collective offload using Mellanox's SHARP for Barrier
- Enhanced tuning framework for Barrier using SHARP
* Remove dependency on underlying libibverbs, libibmad, libibumad, and
librdmacm libraries using dlopen
* Add support for Broadcom NetXtreme RoCE HCA
- Enhanced inter-node point-to-point support
* Support architecture detection for Fujitsu A64fx processor
* Enhanced point-to-point and collective tuning for Fujitsu A64fx processor
* Enhanced point-to-point and collective tuning for AMD ROME processor
* Add support for process placement aware HCA selection
- Add "MV2_PROCESS_PLACEMENT_AWARE_HCA_MAPPING" environment variable to
enable process placement aware HCA mapping
* Add support to auto-detect RoCE HCAs and auto-detect GID index
* Add support to use RoCE/Ethernet and InfiniBand HCAs at the same time
* Add architecture-specific flags to improve performance of certain CUDA
operations
- Thanks to Chris Chambreau @LLNL for the report
* Read MTU and maximum outstanding RDMA operations from the device
* Improved performance and scalability for UD-based communication
* Update maximum HCAs supported by default from 4 to 10
* Enhanced collective tuning for Frontera@TACC, Expanse@SDSC,
Ookami@StonyBrook, and bb5@EPFL
* Enhanced support for SHARP v2.1.0
* Generalize code for GPU support
- Obsolete: wrapper-revert-ldflag-order-change.patch.
- Replace: mvapich2-fix-double-free.patch by
rdma_find_network_type-return-MV2_NETWORK_CLASS_UNKNOWN-when-dev_list-is-freed.patch
-------------------------------------------------------------------
Thu Feb 18 04:46:04 UTC 2021 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Re-add mvapich2-fix-double-free.patch as the bug was
somehow be reintroduced (bsc#1144000)
- Add mvapich2-remove-deprecated-sys_siglist.patch to
fix compilation errors with newer glibc
-------------------------------------------------------------------
Sun Nov 29 13:44:44 UTC 2020 - Egbert Eich <eich@suse.com>
- HPC: Fix environment module settings for MANPATH.
-------------------------------------------------------------------
Sat Jul 25 07:03:57 UTC 2020 - Egbert Eich <eich@suse.com>
- For HPC builds check for gnu compiler suite version >= 10 and
set -fallow-argument-mismatch.
-------------------------------------------------------------------
Thu Jul 23 15:53:00 UTC 2020 - Egbert Eich <eich@suse.com>
- Add build support for gcc8/9/10 to HPC build (bsc#1174439).
-------------------------------------------------------------------
Wed Jun 3 08:13:07 UTC 2020 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Update so mvapich2 2.3.4
- See CHANGELOG for fixes and new features
- Add fix-missing-return-code.patch to fix compilation errors
- Add 0001-Drop-Real-16.patch to disable Real(16) support on Armv7
- Add wrapper-revert-ldflag-order-change.patch to revert LDFLAGS order
change done in 2.3.4.
This allows legacy builds to work without adding a -lmpi flag
-------------------------------------------------------------------
Tue Jan 21 08:17:00 UTC 2020 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Update to mvapich2 2.3.3 (jsc#SLE-8497)
- See CHANGELOG for fixes and new features
-------------------------------------------------------------------
Fri Sep 6 08:10:21 UTC 2019 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Update to mvapich2 2.3.2 (jsc#SLE-8544)
- See CHANGELOG for fixes and new features
- Drop mvapich2-make-sure-ibv_get_device_list-returned-one-before-freeing-it.patch
as it was fixed upstream.
- Use FAT LTO objects in order to provide proper static library.
- Add 0001-Drop-real128.patch to fix compilation on armv7
-------------------------------------------------------------------
Fri Aug 2 08:56:21 UTC 2019 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Add mvapich2-make-sure-ibv_get_device_list-returned-one-before-freeing-it.patch
to Fix segfault when ib_uverbs is not loaded (bsc#1144000)
-------------------------------------------------------------------
Mon May 27 15:14:10 UTC 2019 - Bernhard Wiedemann <bwiedemann@suse.com>
- Add reproducible.patch to sort readdir to make package build reproducible
(boo#1041090)
-------------------------------------------------------------------
Mon May 13 07:59:22 UTC 2019 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Update to mvapich2 2.3.1
- See CHANGELOG for fixes and new features
- Refreshed patches against the new version:
- 0001-Drop-GCC-check.patch
- mvapich2-arm-support.patch
- mvapich2-s390_get_cycles.patch
- Drop mvapich2-fix-double-free.patch as it was merged upstream
-------------------------------------------------------------------
Thu May 2 08:00:51 UTC 2019 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
- Add mvapich2-fix-double-free.patch to fix a segfault
when running on a machine with no RDMA hardware (bsc#1133797)
-------------------------------------------------------------------
Wed Mar 20 19:01:00 UTC 2019 - Ana Guerrero Lopez <aguerrero@suse.com>
- Add patch to remove obsolete GCC check (bnc#1129421). It also patches
autogen.sh to get the autotools working in SLE12SP4.
* 0001-Drop-GCC-check.patch
- Force to re-run autotools to generate properly the files after
patching src/binding/cxx/buildiface
-------------------------------------------------------------------
Sun Nov 18 14:21:27 UTC 2018 - eich@suse.com
- Add macro _hpc_mvapich2_modules for modules support (bsc#1116458).
-------------------------------------------------------------------
Mon Sep 10 12:15:10 UTC 2018 - nmoreychaisemartin@suse.com
- Remove bashism in postun scriptlet
-------------------------------------------------------------------
Wed Sep 5 06:54:33 UTC 2018 - nmoreychaisemartin@suse.com
- Fix handling of mpi-selector during updates (bsc#1098653)
-------------------------------------------------------------------
Sun Aug 19 09:21:45 UTC 2018 - eich@suse.com
- macros.hpc-mvapich2:
replace %%compiler_family by %%hpc_compiler_family
-------------------------------------------------------------------
Mon Jul 16 16:13:32 UTC 2018 - msuchanek@suse.com
- Use sched_yield instead of pthread_yield (boo#1102421).
- drop mvapich2-pthread_yield.patch
-------------------------------------------------------------------
Mon Jun 18 13:12:15 UTC 2018 - nmoreychaisemartin@suse.com
- Add missing bsc and fate references to changelog
-------------------------------------------------------------------
Tue Jun 12 07:07:37 UTC 2018 - nmoreychaisemartin@suse.com
- Disable HPC builds for SLE12 (fate#323655)
-------------------------------------------------------------------
Sun Mar 25 13:07:18 UTC 2018 - kasimir_@outlook.de
- Change mvapich2-arm-support.patch to provide missing functions for
armv6hl
-------------------------------------------------------------------
Fri Feb 9 11:29:26 UTC 2018 - cgoll@suse.com
- Fix summary in module files (bnc#1080259)
-------------------------------------------------------------------
Tue Jan 30 20:04:01 UTC 2018 - eich@suse.com
- Use macro in mpivars.(c)sh to be independent of changes to the module
setup for the compiler (boo#1078364).
-------------------------------------------------------------------
Fri Jan 5 19:03:44 UTC 2018 - eich@suse.com
- Switch from gcc6 to gcc7 as additional compiler flavor for HPC on SLES.
- Fix library package requires - use HPC macro (boo#1074890).
-------------------------------------------------------------------
Fri Oct 6 14:06:37 UTC 2017 - nmoreychaisemartin@suse.com
- Add conflicts between the macros-devel packages
-------------------------------------------------------------------
Thu Oct 5 13:30:48 UTC 2017 - nmoreychaisemartin@suse.com
- Add BuildRequires to libibmad-devel for older release (SLE <= 12.2, Leap <= 42.2)
-------------------------------------------------------------------
Fri Sep 12 15:38:14 UTC 2017 - eich@suse.com
- Add HPC specific build targets using environment modules
(FATE#321712).
-------------------------------------------------------------------
Tue Sep 12 09:25:01 UTC 2017 - nmoreychaisemartin@suse.com
- Drop unnecessary dependency to xorg-x11-devel
-------------------------------------------------------------------
Mon Sep 11 09:19:37 UTC 2017 - nmoreychaisemartin@suse.com
- Only requires verbs libraries for verbs build.
libibverbs devel causes a SEGV when run in a chroot using the
psm or psm2 conduits
- Add testuite packages for all build flavours
-------------------------------------------------------------------
Thu Jul 13 14:59:13 UTC 2017 - nmoreychaisemartin@suse.com
- Add LD_LIBRARY_PATH to mpivars.sh and mpivars.csh
-------------------------------------------------------------------
Thu Jul 13 07:54:38 UTC 2017 - nmoreychaisemartin@suse.com
- Disable rpath in pkgconfig files
-------------------------------------------------------------------
Wed Jul 5 15:12:46 UTC 2017 - nmoreychaisemartin@suse.com
- Remove redondant configure options already passed by %configure
-------------------------------------------------------------------
Mon Jun 26 12:24:30 UTC 2017 - nmoreychaisemartin@suse.com
- Change install dir to allow multiple flavor to be installed
at the same time (bsc#934090)
- Fix bsc#1045955
- Fix mvapich2-psm package to use libpsm (TrueScale)
- Add mvapich2-psm2 package using libpsm2 (OmniPath)
-------------------------------------------------------------------
Mon Jun 26 09:37:42 UTC 2017 - nmoreychaisemartin@suse.com
- Use _multibuild to build the various mvapich2-flavours
-------------------------------------------------------------------
Fri Jun 23 15:40:51 UTC 2017 - nmoreychaisemartin@suse.com
- Replace dependency from libibmad-devel to infiniband-diags-devel
-------------------------------------------------------------------
Wed Jun 14 10:01:07 UTC 2017 - nmoreychaisemartin@suse.com
- Have mvapich2 and mvapich2-psm conflicts between them
- Cleanup spec file
- Remove mvapich2-testsuite RPM
-------------------------------------------------------------------
Thu Jun 8 13:55:32 UTC 2017 - nmoreychaisemartin@suse.com
- Reenable arm compilation
- Rename and cleanup mvapich-s390_get_cycles.patch to
mvapich2-s390_get_cycles.patch for coherency
- Cleanup mvapich2-pthread_yield.patch
- Add mvapich2-arm-support.patch to provide missing functions for
armv7hl and aarch64
-------------------------------------------------------------------
Thu Jun 8 11:38:36 UTC 2017 - nmoreychaisemartin@suse.com
- Remove version dependencies to libibumad, libibverbs and librdmacm
-------------------------------------------------------------------
Tue May 16 16:29:41 UTC 2017 - nmoreychaisemartin@suse.com
- Fix mvapich2-testsuite packaging
- Disable build on armv7
-------------------------------------------------------------------
Wed Mar 29 08:06:23 CEST 2017 - pth@suse.de
- Make dependencies on libs now coming from rdma-core versioned.
-------------------------------------------------------------------
Tue Nov 29 13:08:18 CET 2016 - pth@suse.de
- Create environment module (bsc#1004628).
-------------------------------------------------------------------
Wed Nov 23 11:00:43 CET 2016 - pth@suse.de
- Fix URL.
- Update to mvapich 2.2 GA. Changes since rc1:
MVAPICH2 2.2 (09/07/2016)
* Features and Enhancements (since 2.2rc2):
- Single node collective tuning for Bridges@PSC, Stampede@TACC and other
architectures
- Enable PSM builds when both PSM and PSM2 libraries are present
- Add support for HCAs that return result of atomics in big endian notation
- Establish loopback connections by default if HCA supports atomics
* Bug Fixes (since 2.2rc2):
- Fix minor error in use of communicator object in collectives
- Fix missing u_int64_t declaration with PGI compilers
- Fix memory leak in RMA rendezvous code path
MVAPICH2 2.2rc2 (08/08/2016)
* Features and Enhancements (since 2.2rc1):
- Enhanced performance for MPI_Comm_split through new bitonic algorithm
- Enable graceful fallback to Shared Memory if LiMIC2 or CMA transfer fails
- Enable support for multiple MPI initializations
- Unify process affinity support in Gen2, PSM and PSM2 channels
- Remove verbs dependency when building the PSM and PSM2 channels
- Allow processes to request MPI_THREAD_MULTIPLE when socket or NUMA node
level affinity is specified
- Point-to-point and collective performance optimization for Intel Knights
Landing
- Automatic detection and tuning for InfiniBand EDR HCAs
- Warn user to reconfigure library if rank type is not large enough to
represent all ranks in job
- Collective tuning for Opal@LLNL, Bridges@PSC, and Stampede-1.5@TACC
- Tuning and architecture detection for Intel Broadwell processors
- Add ability to avoid using --enable-new-dtags with ld
- Add LIBTVMPICH specific CFLAGS and LDFLAGS
* Bug Fixes (since 2.2rc1):
- Disable optimization that removes use of calloc in ptmalloc hook
detection code
- Fix weak alias typos (allows successful compilation with CLANG compiler)
- Fix issues in PSM large message gather operations
- Enhance error checking in collective tuning code
- Fix issues with UD based communication in RoCE mode
- Fix issues with PMI2 support in singleton mode
- Fix default binding bug in hydra launcher
- Fix issues with Checkpoint Restart when launched with mpirun_rsh
- Fix fortran binding issues with Intel 2016 compilers
- Fix issues with socket/NUMA node level binding
- Disable atomics when using Connect-IB with RDMA_CM
- Fix hang in MPI_Finalize when using hybrid channel
- Fix memory leaks
-------------------------------------------------------------------
Tue Nov 15 14:04:50 CET 2016 - pth@suse.de
- Update to version 2.2rc1 (fate#319240). Changes since 2.1:
MVAPICH2 2.2rc1 (03/29/2016)
* Features and Enhancements (since 2.2b):
- Support for OpenPower architecture
- Optimized inter-node and intra-node communication
- Support for Intel Omni-Path architecture
- Thanks to Intel for contributing the patch
- Introduction of a new PSM2 channel for Omni-Path
- Support for RoCEv2
- Architecture detection for PSC Bridges system with Omni-Path
- Enhanced startup performance and reduced memory footprint for storing
InfiniBand end-point information with SLURM
- Support for shared memory based PMI operations
- Availability of an updated patch from the MVAPICH project website
with this support for SLURM installations
- Optimized pt-to-pt and collective tuning for Chameleon InfiniBand
systems at TACC/UoC
- Enable affinity by default for TrueScale(PSM) and Omni-Path(PSM2)
channels
- Enhanced tuning for shared-memory based MPI_Bcast
- Enhanced debugging support and error messages
- Update to hwloc version 1.11.2
* Bug Fixes (since 2.2b):
- Fix issue in some of the internal algorithms used for MPI_Bcast,
MPI_Alltoall and MPI_Reduce
- Fix hang in one of the internal algorithms used for MPI_Scatter
- Thanks to Ivan Raikov@Stanford for reporting this issue
- Fix issue with rdma_connect operation
- Fix issue with Dynamic Process Management feature
- Fix issue with de-allocating InfiniBand resources in blocking mode
- Fix build errors caused due to improper compile time guards
- Thanks to Adam Moody@LLNL for the report
- Fix finalize hang when running in hybrid or UD-only mode
- Thanks to Jerome Vienne@TACC for reporting this issue
- Fix issue in MPI_Win_flush operation
- Thanks to Nenad Vukicevic for reporting this issue
- Fix out of memory issues with non-blocking collectives code
- Thanks to Phanisri Pradeep Pratapa and Fang Liu@GaTech for
reporting this issue
- Fix fall-through bug in external32 pack
- Thanks to Adam Moody@LLNL for the report and patch
- Fix issue with on-demand connection establishment and blocking mode
- Thanks to Maksym Planeta@TU Dresden for the report
- Fix memory leaks in hardware multicast based broadcast code
- Fix memory leaks in TrueScale(PSM) channel
- Fix compilation warnings
MVAPICH2 2.2b (11/12/2015)
* Features and Enhancements (since 2.2a):
- Enhanced performance for small messages
- Enhanced startup performance with SLURM
- Support for PMIX_Iallgather and PMIX_Ifence
- Support to enable affinity with asynchronous progress thread
- Enhanced support for MPIT based performance variables
- Tuned VBUF size for performance
- Improved startup performance for QLogic PSM-CH3 channel
- Thanks to Maksym Planeta@TU Dresden for the patch
* Bug Fixes (since 2.2a):
- Fix issue with MPI_Get_count in QLogic PSM-CH3 channel with very large
messages (>2GB)
- Fix issues with shared memory collectives and checkpoint-restart
- Fix hang with checkpoint-restart
- Fix issue with unlinking shared memory files
- Fix memory leak with MPIT
- Fix minor typos and usage of inline and static keywords
- Thanks to Maksym Planeta@TU Dresden for the patch and suggestions
- Fix missing MPIDI_FUNC_EXIT
- Thanks to Maksym Planeta@TU Dresden for the patch
- Remove unused code
- Thanks to Maksym Planeta@TU Dresden for the patch
- Continue with warning if user asks to enable XRC when the system does not
support XRC
MVAPICH2 2.2a (08/17/2015)
* Features and Enhancements (since 2.1 GA):
- Based on MPICH 3.1.4
- Support for backing on-demand UD CM information with shared memory
for minimizing memory footprint
- Reorganized HCA-aware process mapping
- Dynamic identification of maximum read/atomic operations supported by HCA
- Enabling support for intra-node communications in RoCE mode without
shared memory
- Updated to hwloc 1.11.0
- Updated to sm_20 kernel optimizations for MPI Datatypes
- Automatic detection and tuning for 24-core Haswell architecture
* Bug Fixes (since 2.1 GA):
- Fix for error with multi-vbuf design for GPU based communication
- Fix bugs with hybrid UD/RC/XRC communications
- Fix for MPICH putfence/getfence for large messages
- Fix for error in collective tuning framework
- Fix validation failure with Alltoall with IN_PLACE option
- Thanks for Mahidhar Tatineni @SDSC for the report
- Fix bug with MPI_Reduce with IN_PLACE option
- Thanks to Markus Geimer for the report
- Fix for compilation failures with multicast disabled
- Thanks to Devesh Sharma @Emulex for the report
- Fix bug with MPI_Bcast
- Fix IPC selection for shared GPU mode systems
- Fix for build time warnings and memory leaks
- Fix issues with Dynamic Process Management
- Thanks to Neil Spruit for the report
- Fix bug in architecture detection code
- Thanks to Adam Moody @LLNL for the report
-------------------------------------------------------------------
Fri Oct 14 11:28:41 CEST 2016 - pth@suse.de
- Create and include modules file for Mvapich2 (bsc#1004628).
- Remove mvapich2-fix-implicit-decl.patch as the fix is upstream.
- Adapt spec file to the changed micro benchmark install directory.
-------------------------------------------------------------------
Sun Jul 24 14:24:59 UTC 2016 - p.drouand@gmail.com
- Update to version 2.1
* Features and Enhancements (since 2.1rc2):
- Tuning for EDR adapters
- Optimization of collectives for SDSC Comet system
- Based on MPICH-3.1.4
- Enhanced startup performance with mpirun_rsh
- Checkpoint-Restart Support with DMTCP (Distributed MultiThreaded
CheckPointing)
- Thanks to the DMTCP project team (http://dmtcp.sourceforge.net/)
- Support for handling very large messages in RMA
- Optimize size of buffer requested for control messages in large message
transfer
- Enhanced automatic detection of atomic support
- Optimized collectives (bcast, reduce, and allreduce) for 4K processes
- Introduce support to sleep for user specified period before aborting
- Disable PSM from setting CPU affinity
- Install PSM error handler to print more verbose error messages
- Introduce retry mechanism to perform psm_ep_open in PSM channel
* Bug-Fixes (since 2.1rc2):
- Relocate reading environment variables in PSM
- Fix issue with automatic process mapping
- Fix issue with checkpoint restart when full path is not given
- Fix issue with Dynamic Process Management
- Fix issue in CUDA IPC code path
- Fix corner case in CMA runtime detection
* Features and Enhancements (since 2.1rc1):
- Based on MPICH-3.1.4
- Enhanced startup performance with mpirun_rsh
- Checkpoint-Restart Support with DMTCP (Distributed MultiThreaded
CheckPointing)
- Support for handling very large messages in RMA
- Optimize size of buffer requested for control messages in large message
transfer
- Enhanced automatic detection of atomic support
- Optimized collectives (bcast, reduce, and allreduce) for 4K processes
- Introduce support to sleep for user specified period before aborting
- Disable PSM from setting CPU affinity
- Install PSM error handler to print more verbose error messages
- Introduce retry mechanism to perform psm_ep_open in PSM channel
* Bug-Fixes (since 2.1rc1):
- Fix failures with shared memory collectives with checkpoint-restart
- Fix failures with checkpoint-restart when using internal communication
buffers of different size
- Fix undeclared variable error when --disable-cxx is specified with
configure
- Fix segfault seen during connect/accept with dynamic processes
- Fix errors with large messages pack/unpack operations in PSM channel
- Fix for bcast collective tuning
- Fix assertion errors in one-sided put operations in PSM channel
- Fix issue with code getting stuck in infinite loop inside ptmalloc
- Fix assertion error in shared memory large message transfers
- Fix compilation warnings
* Features and Enhancements (since 2.1a):
- Based on MPICH-3.1.3
- Flexibility to use internal communication buffers of different size for
improved performance and memory footprint
- Improve communication performance by removing locks from critical path
- Enhanced communication performance for small/medium message sizes
- Support for linking Intel Trace Analyzer and Collector
- Increase the number of connect retry attempts with RDMA_CM
- Automatic detection and tuning for Haswell architecture
* Bug-Fixes (since 2.1a):
- Fix automatic detection of support for atomics
- Fix issue with void pointer arithmetic with PGI
- Fix deadlock in ctxidup MPICH test in PSM channel
- Fix compile warnings
* Features and Enhancements (since 2.0):
- Based on MPICH-3.1.2
- Support for PMI-2 based startup with SLURM
- Enhanced startup performance for Gen2/UD-Hybrid channel
- GPU support for MPI_Scan and MPI_Exscan collective operations
- Optimize creation of 2-level communicator
- Collective optimization for PSM-CH3 channel
- Tuning for IvyBridge architecture
- Add -export-all option to mpirun_rsh
- Support for additional MPI-T performance variables (PVARs)
in the CH3 channel
- Link with libstdc++ when building with GPU support
(required by CUDA 6.5)
* Bug-Fixes (since 2.0):
- Fix error in large message (>2GB) transfers in CMA code path
- Fix memory leaks in OFA-IB-CH3 and OFA-IB-Nemesis channels
- Fix issues with optimizations for broadcast and reduce collectives
- Fix hang at finalize with Gen2-Hybrid/UD channel
- Fix issues for collectives with non power-of-two process counts
- Make ring startup use HCA selected by user
- Increase counter length for shared-memory collectives
- Use download Url as source
- Some other minor improvements
- Add mvapich2-fix-implicit-decl.patch
-------------------------------------------------------------------
Thu Oct 9 13:32:28 CEST 2014 - pth@suse.de
- Don't provide the full source uri as build servis can't handle it.
-------------------------------------------------------------------
Wed Oct 8 17:12:27 CEST 2014 - pth@suse.de
- Only run autogen.sh if the distribution has a new enough automake.
-------------------------------------------------------------------
Wed Sep 24 17:06:22 CEST 2014 - pth@suse.de
- Update to mvapich2 2.0 GMC:
* Features and Enhancements (since 2.0rc2):
- Consider CMA in collective tuning framework
* Bug-Fixes (since 2.0rc2):
- Fix bug when disabling registration cache
- Fix shared memory window bug when shared memory collectives
are disabled.
- Fix mpirun_rsh bug when running mpmd programs with no arguments
- Exclude Aarch64 for the time being as asm/timex.h seems to be missing
from the glibc kernel headers.
-------------------------------------------------------------------
Tue Jun 3 11:24:34 CEST 2014 - pth@suse.de
- Update to OFED 3.12 final.
-------------------------------------------------------------------
Mon May 26 13:02:24 CEST 2014 - pth@suse.de
- Update to 2.0rc2:
* Features and Enhancements (since 2.0rc1):
- CMA support is now enabled by default
- Optimization of collectives with CMA support
- RMA optimizations for shared memory and atomic operations
- Tuning RGET and Atomics operations
- Tuning RDMA FP-based communication
- MPI-T support for additional performance and control variables
- The --enable-mpit-pvars=yes configuration option will now
enable only MVAPICH2 specific variables
- Large message transfer support for PSM interface
- Optimization of collectives for PSM interface
- Updated to hwloc v1.9
* Bug-Fixes (since 2.0rc1):
- Fix multicast hang when there is a single process on one node
and more than one process on other nodes
- Fix non-power-of-two usage of scatter-doubling-allgather algorithm
- Fix for bcastzero type hang during finalize
- Enhanced handling of failures in RDMA_CM based
connection establishment
- Fix for a hang in finalize when using RDMA_CM
- Finish receive request when RDMA READ completes in RGET protocol
- Always use direct RDMA when flush is used
- Fix compilation error with --enable-g=all in PSM interface
- Fix warnings and memory leaks
-------------------------------------------------------------------
Thu May 15 16:01:50 CEST 2014 - pth@suse.de
- mvapich2-psm-devel requires infinipath-psm-devel.
- Remove redundent requires for the devel-static package.
-------------------------------------------------------------------
Wed May 7 15:40:33 UTC 2014 - stefan.fent@suse.com
- remove typo in mvapich-s390_get_cycles.patch
-------------------------------------------------------------------
Tue Apr 29 13:47:06 CEST 2014 - pth@suse.de
- Remove bogus 0 from spec.
-------------------------------------------------------------------
Mon Apr 28 12:30:12 CEST 2014 - pth@suse.de
- Remove all additional mvapich specific CFLAGS and extra LIBS.
-------------------------------------------------------------------
Fri Apr 25 09:41:47 CEST 2014 - pth@suse.de
- Fix ExclusiveArch
- Only PSM needs explicit configuration so drop the else branch
in configure call.
- mvapich2 now builds in parallel so tell make.
-------------------------------------------------------------------
Thu Apr 24 21:27:06 CEST 2014 - pth@suse.de
- Build Mvapich2 for Qlogic from its own mvapich2-psm.spec.
-------------------------------------------------------------------
Wed Apr 23 18:04:36 CEST 2014 - pth@suse.de
- Add mvapich2-pthread_yield.patch to define GNU_SOURCE before
including pthread.h to get pthread_yield declared.
-------------------------------------------------------------------
Wed Apr 23 14:48:07 CEST 2014 - pth@suse.de
- Don't require libibcommon as it's gone with OFED 3.12.
-------------------------------------------------------------------
Wed Apr 16 15:50:22 UTC 2014 - stefan.fent@suse.com
- add asm code from kernel to properly implement get_cycles on
s390 and s390x (bnc #870424) (mvapich-s390_get_cycles.patch)
-------------------------------------------------------------------
Mon Apr 7 14:49:22 CEST 2014 - pth@suse.de
- Fix spec so that testsuite builds correctly.
-------------------------------------------------------------------
Sat Apr 5 20:28:49 CEST 2014 - pth@suse.de
- Update config.* to make it build on ppc64le.
-------------------------------------------------------------------
Thu Mar 27 12:56:50 CET 2014 - pth@suse.de
- Regenerate autotool files to get ppc64le recognized.
- The predefined platform macros for s390 are lower case not upcase.
-------------------------------------------------------------------
Wed Mar 26 16:15:45 CET 2014 - pth@suse.de
- Finally got the syntax for conditionals in spec right...
- Add a dummy implementation of get_cycles for s390x.
- Update to 2.0rc1 as this is a MPI-3 implementation. For
detailed changes see.
- Fix options passed to mpi-selector
-------------------------------------------------------------------
Tue Mar 25 14:42:43 CET 2014 - pth@suse.de
- Include the two COPYRIGHT files in the package.
- BuildRequire kernel-headers on s390x.
- Fix spec file
-------------------------------------------------------------------
Wed Mar 5 14:04:47 CET 2014 - pth@suse.de
- Compile with support for PSM on ix86 (fate#315889).
- mvapich2 has a testsuite, so run it from a separate spec file.
-------------------------------------------------------------------
Mon Feb 10 13:13:39 CET 2014 - pth@suse.de
- Update to 1.9:
- Remove mvapich2-1.0.2-non-void-rtn.patch as the changes are in
the upstream source.
- Reformat BuildRequires
-------------------------------------------------------------------
Fri Jan 24 19:15:39 CET 2014 - pth@suse.de
- Update to OFED 3.12 daily.
-------------------------------------------------------------------
Fri Feb 29 00:00:00 CET 2008 - - jjolly@suse.de
- Update to 1.0.2 from OFED 1.3 GA release
- Minor changes to return value patch
-------------------------------------------------------------------
Thu Jan 31 00:00:00 CET 2008 - - jjolly@suse.de
- Update to 1.0.1 from OFED 1.3 rc2
- Fixed several 'undefined return value' compile errors
-------------------------------------------------------------------
Tue Jul 10 00:00:00 CET 2007 - - hvogel@suse.de
- Initial Package, Version 0.9.8
Reference in New Issue View Git Blame Copy Permalink