Releases: pmodels/mpich
v4.0.2
Changes in 4.0.2
-
Fix CUDA configuration logic in yaksa
-
Fix support for dynamic process functionality with PMI2 clients
-
Fix non-zero appnum bug in PMI2 server in Hydra
-
Fix MPI_Op support for types created with MPI_Type_create_f90_xxx
-
Fix building ch4 with Intel compilers on macOS
-
Fix Level Zero properties initialization in MPL. Thanks to Brice
Videau for the report and patch. -
Use standard names for CPU affinity functions with POSIX
threads. Fixes building against Musl libc. Thanks to Mosè Giordano for
the report and patch. -
Add elemental to eq/neq operators in Fortran 2008 binding
-
Workaround for inter-process mutex bug on FreeBSD
v4.0.1
Changes in 4.0.1
-
Multiple fixes for NVIDIA/PGI HPC Compilers support
-
Fix ch4:ofi:gni provider capability set
-
Fix MPI_SESSION_INIT "thread_level" info hint
-
Fix build on macOS with --disable-shared
-
Fix QMPI function definitions
-
Fix support for "host" info hint in MPI_COMM_SPAWN[_MULTIPLE]
-
Fix manpage generation
-
Add missing MPI_F_sync_reg function
-
Add missing const to MPI_Psend_init buffer argument
-
Make Python 3 optional in configure script
-
Remove -Wl,flat_namespace from compile wrappers by default (macOS only)
-
Update UCX module to v1.12.0
-
Update yaksa module to support latest Ampere compute capability
v4.0
Changes in 4.0
-
All MPI-4 APIs have been implemented. Major MPI-4 features include MPI
sessions, partitioned point-to-point communications, events in the MPI tool
information interface, large-count functions, persistent collectives,
MPI_Comm_idup_with_info, MPI_Isendrecv and MPI_Isendrecv_replace,
MPI_Info_get_string, MPI_Comm_split_type with new split_type --
MPI_COMM_TYPE_HW_GUIDED and MPI_COMM_TYPE_HW_UNGUIDED. -
Add QMPI (experimental) support.
-
Add MPIX_Delete_error_{class,code,string}.
-
MPI_Info objects can be accessed before MPI_Init{_thread}.
-
Generate C API interface functions including man page notes and error
checking using Python scripts. -
Generate Fortran (mpif.h, mpi_f08) bindings using Python scripts.
-
Generate collective entrance functions and generate per-algorithm tests.
-
Support explicit --without-cuda configure option.
-
Drop support for UCX version < 1.7.0.
-
Configure now optionally require Python 3 (when F08 is enabled).
-
Multi-NIC support in ch4:ofi.
-
Default to ch4:ofi when configure doesn't have a clear choice. Add message
block at the end of configure to advise user. -
Multiple VCI is fully implemented including the active message fallback paths.
-
Extend IPC to support non-contig datatypes.
-
Add AMD GPU support using HIP.
-
Add generic RNDV callback mechanism with active messages.
-
Refactor ch4 dynamic process functions.
-
Avoid building MPL and hwloc multiple times.
-
Fix MPIX_Query_cuda_support.
-
Many bug fixes and code clean-ups.
v4.0rc3
v4.0rc2
v3.4.3
v4.0a2
v3.4.2
MPICH 4.0a1
v4.0a1 tagging 'v4.0a1'
MPICH 3.4.1
v3.4.1 tagging 'v3.4.1'