"""Auxiliary functions and variables for dealing with MPI multiprocessing.
Warning:
These functions are mostly no-ops unless MPI is properly installed and python code
was started using :code:`mpirun` or :code:`mpiexec`. Please refer to the
documentation of your MPI distribution for details.
.. autosummary::
:nosignatures:
mpi_send
mpi_recv
mpi_allreduce
.. codeauthor:: David Zwicker <david.zwicker@ds.mpg.de>
"""
from __future__ import annotations
import os
import sys
import traceback
from typing import TYPE_CHECKING, TypeVar
import numpy as np
if TYPE_CHECKING:
from .typing import NumericArray
T = TypeVar("T")
# Initialize assuming that we run serial code if `numba_mpi` is not available
initialized: bool = False
"""bool: Flag determining whether mpi was initialized (and is available)"""
size: int = 1
"""int: Total process count"""
rank: int = 0
"""int: ID of the current process"""
class _OperatorRegistry:
"""Collection of operators that MPI supports."""
_name_ids: dict[str, int]
_ids_operators: dict[int, MPI.Op]
def __init__(self):
self._name_ids = {}
self._ids_operators = {}
def register(self, name: str, op: MPI.Op):
op_id = int(MPI._addressof(op))
self._name_ids[name] = op_id
self._ids_operators[op_id] = op
def id(self, name_or_id: int | str) -> int:
if isinstance(name_or_id, int):
return name_or_id
return self._name_ids[name_or_id]
def operator(self, name_or_id: int | str) -> MPI.Op:
if isinstance(name_or_id, str):
name_or_id = self._name_ids[name_or_id]
return self._ids_operators[name_or_id]
def __getattr__(self, name: str):
try:
return self._name_ids[name]
except KeyError:
msg = f"MPI operator `{name}` not registered"
raise AttributeError(msg) from None
MPIOperator = _OperatorRegistry()
# read state of the current MPI node
try:
from mpi4py import MPI
except ImportError:
# package `numba_mpi` could not be loaded
if int(os.environ.get("PMI_SIZE", "1")) > 1:
# environment variable indicates that we are in a parallel program
sys.exit(
"WARNING: Detected multiprocessing run, but could not import python "
"package `numba_mpi`"
)
else:
initialized = MPI.Is_initialized()
size = MPI.COMM_WORLD.size
rank = MPI.COMM_WORLD.rank
MPIOperator.register("MAX", MPI.MAX)
MPIOperator.register("MIN", MPI.MIN)
MPIOperator.register("SUM", MPI.SUM)
parallel_run: bool = size > 1
"""bool: Flag indicating whether the current run is using multiprocessing"""
is_main: bool = rank == 0
"""bool: Flag indicating whether the current process is the main process (with ID 0)"""
[docs]
def mpi_send(data: NumericArray, dest: int, tag: int) -> None:
"""Send data to another MPI node.
Args:
data: The data being send
dest (int): The ID of the receiving node
tag (int): A numeric tag identifying the message
"""
MPI.COMM_WORLD.send(data, dest=dest, tag=tag)
[docs]
def mpi_recv(data: NumericArray, source: int, tag: int) -> None:
"""Receive data from another MPI node.
Args:
data: A buffer into which the received data is written
source (int): The ID of the sending node
tag (int): A numeric tag identifying the message
"""
data[...] = MPI.COMM_WORLD.recv(source=source, tag=tag)
[docs]
def mpi_bcast(data: T, root: int = 0) -> T:
"""Broadcast data from root node to all other MPI nodes.
Args:
data: The data being sent
root (int): The ID of the sending node
"""
return MPI.COMM_WORLD.bcast(data, root=root) # type: ignore
[docs]
def mpi_allreduce(data, operator: int | str):
"""Combines data from all MPI nodes.
Note that complex datatypes and user-defined reduction operators are not properly
supported in numba-compiled cases.
Args:
data:
Data being send from this node to all others
operator:
The operator used to combine all data. Possible options are summarized in
the IntEnum :class:`numba_mpi.Operator`.
Returns:
The accumulated data
"""
if not parallel_run:
# in a serial run, we can always return the value as is
return data
if isinstance(data, np.ndarray):
# synchronize an array
out = np.empty_like(data)
MPI.COMM_WORLD.Allreduce(data, out, op=MPIOperator.operator(operator))
return out
# synchronize a single value
return MPI.COMM_WORLD.allreduce(data, op=MPIOperator.operator(operator))
[docs]
def mpi_excepthook(exc_type, exc_value, exc_tb):
"""Print uncaught exceptions with rank information and abort the MPI job.
This function is intended to be assigned to :data:`sys.excepthook` in MPI runs.
It formats the traceback on the local rank, writes the error to stderr, flushes
stdio streams, and then calls :meth:`MPI.Comm.Abort` to terminate all ranks.
Args:
exc_type: Exception class of the uncaught exception.
exc_value: Exception instance.
exc_tb: Traceback object of the uncaught exception.
"""
msg = "".join(traceback.format_exception(exc_type, exc_value, exc_tb))
print(f"\n[FATAL] rank={rank}/{size}\n{msg}", file=sys.stderr, flush=True)
try:
sys.stderr.flush()
sys.stdout.flush()
finally:
# we need to make sure MPI is imported
from mpi4py import MPI
MPI.COMM_WORLD.Abort(1)