- announce - mailman3.cels.anl.gov

Join us for the MPICH BoF at SC23! Wednesday, 11/15 @ 5:15pm MST
by Raffenetti, Ken 10 Nov '23

10 Nov '23

All, The MPICH team will host an SC23 Birds-of-a-Feather session Wednesday, November 15th at 5:15pm MST. We hope to see you there to discuss the latest developments in MPICH, and to hear presentations from MPICH partners and users. Agenda: * ANL Update – Yanfei Guo * Partner Presentations * Maria Garzaran (Intel) * Hari Subramoni (MVAPICH2) * Heidi Poxon (Amazon Web Services) * Simon Pickartz (ParTec AG) * Jim Dinan (NVIDIA) * Michael Raymond (HPE) * User Presentations * Paul Fischer (UIUC, Nek5000/NekRS) * Junchao Zhang (ANL, PETSc) * Q&A More details at https://github.com/pmodels/mpich/wiki/2023-Supercomputing-BoF. Thanks, The MPICH team

1 0

MPICH 4.1.2 released
by Raffenetti, Ken 08 Jun '23

08 Jun '23

A new stable release of MPICH, 4.1.2, is now available for download. This is the second bug-fix release in the 4.1.x series. It contains a number of fixes for user-reported crashes and build issues. You can find the release on our downloads page (https://www.mpich.org/downloads/) Regards, The MPICH team =============================================================================== Changes in 4.1.2 =============================================================================== # Update UCX module to includes fixes for building with GCC 13 # Update libfabric module to 1.18.0 with additional fixes for building with recent versions of LLVM/Clang # Fix compiler wrapper scripts to be compatible with CUDA memory hooks # Fix MPIX_WAITALL_ENQUEUE to make a copy of the input request array # Fix bug in MPI_ALLREDUCE that could result in ranks receiving different floating point values # Fix potential deadlock when progressing RMA windows # Fix potential crash in MPI_REDUCE with non-zero root and MPI_IN_PLACE # Fix potential crash during probe with libfabric CXI provider # Fix MPI_PARRIVED when the partitioned request is inactive # Fix potential bug when an attribute delete callback deletes another attribute on the same object # Fix build issue in ROMIO Lustre driver # Improve Fortran 2008 binding support detection during configure # Report an error if collective tuning json file fails to open # Several fixes for testsuite programs and build configuration A full list of changes since 4.1.1 can be found here: https://github.com/pmodels/mpich/compare/v4.1.1...v4.1.2

1 0

MPICH 4.1.1 released
by Raffenetti, Ken 06 Mar '23

06 Mar '23

A new stable release of MPICH, 4.1.1, is now available for download. This is the first bug-fix release in the 4.1.x series. This release contains fixes for MPI Sessions, mpif.h Fortran support, MPI_IALLTOALLW, persistent MPI_ALLGATHER, testsuite usage with non-MPICH libraries, and updates to the embedded UCX and yaksa submodules. You can find the release on our downloads page (https://www.mpich.org/downloads/) Regards, The MPICH team =============================================================================== Changes in 4.1.1 =============================================================================== # Update embedded UCX module to 1.13.1. Fixes a build issue with binutils >= 2.39. # Update yaksa module. Support explicit NVCC setting by the user. Fixes a build issue when there is no libtool available in PATH. # Fix ch4:ucx initialization when configured with --enable-ch4-vci-method=implicit. # Fix potential error handler leak during MPI_SESSION_FINALIZE # Fix value of MPI_UNDEFINED in mpif.h binding # Fix MPI_IALLTOALLW with MPI_IN_PLACE # Fix send attribute handling in IPC path # Fix a bug in persistent MPI_ALLGATHER # Fix tests for use with non-MPICH libraries # Add missing MPI_T_ERR_NOT_ACCESSIBLE error code # Fix manpages for MPIX functions A full list of changes since 4.1 can be found here: https://github.com/pmodels/mpich/compare/v4.1...v4.1.1

1 0

[REMINDER] Join the MPICH team at ECP Community BoF Days today!
by Raffenetti, Ken 15 Feb '23

15 Feb '23

All, The MPICH team will host a session today as part of the ECP Community BoF Days (https://www.exascaleproject.org/event/2023-ecp-community-bof-days/) We will discuss MPICH release updates as well as hear from some of our partners. The event is virtual and registration is free at this link: https://uoregon.zoom.us/meeting/register/tJMlde-gqjItEte5PxaSSXa9iO275T9SRu… Date and Time: February 15th, 2:00pm - 3:30pm Central Agenda: Intro - Ken Raffenetti (ANL) MPICH Update - Yanfei Guo (ANL) Partner Updates (10 minutes each) - MVAPICH2 Update - Hari Subramoni (OSU) - Intel Update - Gengbin Zheng - NVIDIA Update - Jim Dinan - ParTec Update - Simon Pickartz - HPE/Cray - Krishna Kandalla Q&A This information is also available on our wiki: https://github.com/pmodels/mpich/wiki/2023-ECP-Community-BoF Cheers, The MPICH team

1 0

Join the MPICH team at ECP Community BoF Days next week!
by Raffenetti, Ken 09 Feb '23

09 Feb '23

All, The MPICH team will host a session as part of next week's ECP Community BoF Days (https://www.exascaleproject.org/event/2023-ecp-community-bof-days/) We will discuss MPICH release updates as well as hear from some of our partners. The event is virtual and registration is free at this link: https://uoregon.zoom.us/meeting/register/tJMlde-gqjItEte5PxaSSXa9iO275T9SRu… Date and Time: February 15th, 2:00pm - 3:30pm Central Agenda (subject to change): Intro - Ken Raffenetti (ANL) MPICH Update - Yanfei Guo (ANL) Partner Updates (10 minutes each) - MVAPICH2 Update - Hari Subramoni (OSU) - Intel Update - Gengbin Zheng - NVIDIA Update - Jim Dinan - ParTec Update - TBD Q&A This information is also available on our wiki: https://github.com/pmodels/mpich/wiki/2023-ECP-Community-BoF Cheers, The MPICH team

1 0

MPICH 4.1 released
by Raffenetti, Ken 27 Jan '23

27 Jan '23

A new stable release of MPICH, 4.1, is now available for download. This is the first stable release in the 4.1.x series. Major features in this release include: * Standalone libpmi * Independent use of MPICH testsuite * Experimental MPIX stream prototypes to support explicit threading context and asynchronous enqueue to GPU streams. Both CUDA and HIP are supported. * Significant refactor in PMI and Hydra You can find the release on our downloads page (https://www.mpich.org/downloads/) Regards, The MPICH team =============================================================================== Changes in 4.1 =============================================================================== # Thread-cs in ch4 changed to per-vci. # Testsuite (test/mpi) is configured separately from mpich configure. # Added options in autogen to accelerate CI builds, including using pre-built sub-modules. Added -yaksa-depth option to generate shallower yaksa pup code for faster build and smaller binaries. # Support singleton init using hydra. # On OSX, link option flat_namespace is no longer turned on by default. # Generate mpi.mod Fortran interfaces using Python 3. For many compilers, including gfortran, flags such as -fallow-mismatched-args is no longer necessary. # Fixed message queue debugger interface in ch4. # PMI (src/pmi) is refactored as a subdir and can be separately distributed. # Added MPIX_Comm_get_failed. # Experimental MPIX stream API to enable explicit thread contexts. # Experimental MPIX gpu enqueue API. It currently only supports CUDA streams. # Delays GPU resource allocation in yaksa. # CH3 nemesis ofi netmod is removed. # New collective algorithms. All collective algorithms are listed in src/mpi/coll/coll_algorithms.txt # Removed hydra2. We will port unique features of hydra2, including tree-launching, to hydra in the future release. # Added in-repository wiki documentation. # Added stream workq to support optimizations for enqueue operations. # Better support for large count APIs by eliminating type conversion issues. # Hydra now uses libpmi (src/pmi) for handling PMI messages. # Many bug fixes and enhancements. A full list of changes since 4.0.3 can be found here: https://github.com/pmodels/mpich/compare/v4.0.3...v4.1

1 0

MPICH 4.1rc4 released
by Raffenetti, Ken 23 Jan '23

23 Jan '23

A new release candidate of MPICH, 4.1rc4, is now available for download. This release contains fixes for shared memory initialization and MPI_Allgatherv calls on intracommunicators of size=1. As a reminder, only critical bugs will be fixed before the 4.1 GA release. Please give it a try and report any issues! You can find the release on our downloads page (https://www.mpich.org/downloads/) Regards, The MPICH team =============================================================================== Changes in 4.1 =============================================================================== # Thread-cs in ch4 changed to per-vci. # Testsuite (test/mpi) is configured separately from mpich configure. # Added options in autogen to accelerate CI builds, including using pre-built sub-modules. Added -yaksa-depth option to generate shallower yaksa pup code for faster build and smaller binaries. # Support singleton init using hydra. # On OSX, link option flat_namespace is no longer turned on by default. # Generate mpi.mod Fortran interfaces using Python 3. For many compilers, including gfortran, flags such as -fallow-mismatched-args is no longer necessary. # Fixed message queue debugger interface in ch4. # PMI (src/pmi) is refactored as a subdir and can be separately distributed. # Added MPIX_Comm_get_failed. # Experimental MPIX stream API to enable explicit thread contexts. # Experimental MPIX gpu enqueue API. It currently only supports CUDA streams. # Delays GPU resource allocation in yaksa. # CH3 nemesis ofi netmod is removed. # New collective algorithms. All collective algorithms are listed in src/mpi/coll/coll_algorithms.txt # Removed hydra2. We will port unique features of hydra2, including tree-launching, to hydra in the future release. # Added in-repository wiki documentation. # Added stream workq to support optimizations for enqueue operations. # Better support for large count APIs by eliminating type conversion issues. # Hydra now uses libpmi (src/pmi) for handling PMI messages. # Many bug fixes and enhancements. A full list of changes since 4.0.3 can be found here: https://github.com/pmodels/mpich/compare/v4.0.3...v4.1rc4

1 0

MPICH 4.1rc3 released
by Raffenetti, Ken 19 Jan '23

19 Jan '23

A new release candidate of MPICH, 4.1rc3, is now available for download. This release contains a fix for the MPIR_proctable implementation in Hydra. Thanks to the DDT folks for the report. At this stage, only critical bugs will be fixed before the 4.1 GA release. Please give it a try! You can find the release on our downloads page (https://www.mpich.org/downloads/) Regards, The MPICH team =============================================================================== Changes in 4.1 =============================================================================== # Thread-cs in ch4 changed to per-vci. # Testsuite (test/mpi) is configured separately from mpich configure. # Added options in autogen to accelerate CI builds, including using pre-built sub-modules. Added -yaksa-depth option to generate shallower yaksa pup code for faster build and smaller binaries. # Support singleton init using hydra. # On OSX, link option flat_namespace is no longer turned on by default. # Generate mpi.mod Fortran interfaces using Python 3. For many compilers, including gfortran, flags such as -fallow-mismatched-args is no longer necessary. # Fixed message queue debugger interface in ch4. # PMI (src/pmi) is refactored as a subdir and can be separately distributed. # Added MPIX_Comm_get_failed. # Experimental MPIX stream API to enable explicit thread contexts. # Experimental MPIX gpu enqueue API. It currently only supports CUDA streams. # Delays GPU resource allocation in yaksa. # CH3 nemesis ofi netmod is removed. # New collective algorithms. All collective algorithms are listed in src/mpi/coll/coll_algorithms.txt # Removed hydra2. We will port unique features of hydra2, including tree-launching, to hydra in the future release. # Added in-repository wiki documentation. # Added stream workq to support optimizations for enqueue operations. # Better support for large count APIs by eliminating type conversion issues. # Hydra now uses libpmi (src/pmi) for handling PMI messages. # Many bug fixes and enhancements. A full list of changes since 4.0.3 can be found here: https://github.com/pmodels/mpich/compare/v4.0.3...v4.1rc3

1 0

MPICH 4.1rc2 released
by Raffenetti, Ken 13 Jan '23

13 Jan '23

A new release candidate of MPICH, 4.1rc2, is now available for download. This release contains fixes for a few user reported issues, as well as improvements to the release packaging process. At this stage, only critical bugs will be fixed before the 4.1 GA release. Please give it a try! You can find the release on our downloads page (https://www.mpich.org/downloads/) Regards, The MPICH team =============================================================================== Changes in 4.1 =============================================================================== # Thread-cs in ch4 changed to per-vci. # Testsuite (test/mpi) is configured separately from mpich configure. # Added options in autogen to accelerate CI builds, including using pre-built sub-modules. Added -yaksa-depth option to generate shallower yaksa pup code for faster build and smaller binaries. # Support singleton init using hydra. # On OSX, link option flat_namespace is no longer turned on by default. # Generate mpi.mod Fortran interfaces using Python 3. For many compilers, including gfortran, flags such as -fallow-mismatched-args is no longer necessary. # Fixed message queue debugger interface in ch4. # PMI (src/pmi) is refactored as a subdir and can be separately distributed. # Added MPIX_Comm_get_failed. # Experimental MPIX stream API to enable explicit thread contexts. # Experimental MPIX gpu enqueue API. It currently only supports CUDA streams. # Delays GPU resource allocation in yaksa. # CH3 nemesis ofi netmod is removed. # New collective algorithms. All collective algorithms are listed in src/mpi/coll/coll_algorithms.txt # Removed hydra2. We will port unique features of hydra2, including tree-launching, to hydra in the future release. # Added in-repository wiki documentation. # Added stream workq to support optimizations for enqueue operations. # Better support for large count APIs by eliminating type conversion issues. # Hydra now uses libpmi (src/pmi) for handling PMI messages. # Many bug fixes and enhancements. A full list of changes since 4.0.3 can be found here: https://github.com/pmodels/mpich/compare/v4.0.3...v4.1rc2

1 0

MPICH 4.1rc1 released
by Raffenetti, Ken 21 Dec '22

21 Dec '22

A new release candidate of MPICH, 4.1rc1, is now available for download. Major features in this release include: * Standalone libpmi * Independent use of MPICH testsuite * Experimental MPIX stream prototypes to support explicit threading context and asynchronous enqueue to GPU streams. Both CUDA and HIP are supported. * Significant refactor in PMI and Hydra At this stage, only critical bugs will be fixed before the 4.1 GA release. Please give it a try! You can find the release on our downloads page (https://www.mpich.org/downloads/ <https://www.mpich.org/downloads/>). Regards, The MPICH team =============================================================================== Changes in 4.1 =============================================================================== # Thread-cs in ch4 changed to per-vci. # Testsuite (test/mpi) is configured separately from mpich configure. # Added options in autogen to accelerate CI builds, including using pre-built sub-modules. Added -yaksa-depth option to generate shallower yaksa pup code for faster build and smaller binaries. # Support singleton init using hydra. # On OSX, link option flat_namespace is no longer turned on by default. # Generate mpi.mod Fortran interfaces using Python 3. For many compilers, including gfortran, flags such as -fallow-mismatched-args is no longer necessary. # Fixed message queue debugger interface in ch4. # PMI (src/pmi) is refactored as a subdir and can be separately distributed. # Added MPIX_Comm_get_failed. # Experimental MPIX stream API to enable explicit thread contexts. # Experimental MPIX gpu enqueue API. It currently only supports CUDA streams. # Delays GPU resource allocation in yaksa. # CH3 nemesis ofi netmod is removed. # New collective algorithms. All collective algorithms are listed in src/mpi/coll/coll_algorithms.txt # Removed hydra2. We will port unique features of hydra2, including tree-launching, to hydra in the future release. # Added in-repository wiki documentation. # Added stream workq to support optimizations for enqueue operations. # Better support for large count APIs by eliminating type conversion issues. # Hydra now uses libpmi (src/pmi) for handling PMI messages. # Many bug fixes and enhancements. A full list of changes since 4.0.3 can be found here: https://github.com/pmodels/mpich/compare/v4.0.3...v4.1rc1

1 0