dolfinx::MPI Namespace Reference

MPI support functionality. More...

Classes
class	Comm
	A duplicate MPI communicator and manage lifetime of the communicator. More...
struct	dependent_false

Functions
int	rank (MPI_Comm comm)
	Return process rank for the communicator.
int	size (MPI_Comm comm)
	Return size of the group (number of processes) associated with the communicator.
template<typename T >
graph::AdjacencyList< T >	all_to_all (MPI_Comm comm, const graph::AdjacencyList< T > &send_data)
	Send in_values[p0] to process p0 and receive values from process p1 in out_values[p1].
std::vector< int >	compute_graph_edges (MPI_Comm comm, const std::set< int > &edges)
template<typename T >
graph::AdjacencyList< T >	neighbor_all_to_all (MPI_Comm neighbor_comm, const graph::AdjacencyList< T > &send_data)
	Neighborhood all-to-all. Send data to neighbors. Send in_values[n0] to neighbor process n0 and receive values from neighbor process n1 in out_values[n1].
std::tuple< std::vector< int >, std::vector< int > >	neighbors (MPI_Comm neighbor_comm)
constexpr std::array< std::int64_t, 2 >	local_range (int rank, std::int64_t N, int size)
	Return local range for given process, splitting [0, N - 1] into size() portions of almost equal size. More...
constexpr int	index_owner (int size, std::size_t index, std::size_t N)
	Return which process owns index (inverse of local_range) More...
template<typename T >
constexpr MPI_Datatype	mpi_type ()
	MPI Type.

Detailed Description

MPI support functionality.

Function Documentation

compute_graph_edges()

std::vector< int > dolfinx::MPI::compute_graph_edges	(	MPI_Comm	comm,
		const std::set< int > &	edges
	)

Todo:

Experimental. Maybe be moved or removed.

Compute communication graph edges. The caller provides edges that it can define, and will receive edges to it that are defined by other ranks.

Note

This function involves global communication

Parameters

[in]	comm	The MPI communicator
[in]	edges	Communication edges between the caller and the ranks in edges.

Returns

Ranks that have defined edges from them to this rank

index_owner()

constexpr int dolfinx::MPI::index_owner ( int  size, std::size_t  index, std::size_t  N  )

constexpr

Return which process owns index (inverse of local_range)

Parameters

[in]	size	Number of MPI ranks
[in]	index	The index to determine owning rank
[in]	N	Total number of indices

Returns

The rank of the owning process

local_range()

constexpr std::array<std::int64_t, 2> dolfinx::MPI::local_range ( int  rank, std::int64_t  N, int  size  )

constexpr

Return local range for given process, splitting [0, N - 1] into size() portions of almost equal size.

Parameters

[in]	rank	MPI rank of the caller
[in]	N	The value to partition
[in]	size	The number of MPI ranks across which to partition N

neighbors()

std::tuple< std::vector< int >, std::vector< int > > dolfinx::MPI::neighbors

(

MPI_Comm

neighbor_comm

)

Todo:

Clarify directions

Return list of neighbors for a neighborhood communicator

Parameters

[in]

neighbor_comm

Neighborhood communicator

Returns

source ranks, destination ranks