The mpi namespace, contains wrapper for many MPI functions. More...

Classes
class	communicator
	A thin wrapper of MPI_Comm that supports most MPI calls. More...

class	contiguous_type
	A move-only wrapper for a contiguous MPI_Datatype. More...

class	environment
	Class that sets up and finalizes the MPI environment. More...

class	request
	The request class is a light, move-only wrapper around the MPI_Request handle. More...

struct	status
	The status struct is a light wrapper around the MPI_Status struct. More...

struct	type_impl
	A struct that is used to determine the MPI_Datatype of a specified type. More...

struct	type_impl< char >

struct	type_impl< double >

struct	type_impl< float >

struct	type_impl< int >

struct	type_impl< long >

struct	type_impl< long double >

struct	type_impl< long long >

struct	type_impl< std::complex< double > >

struct	type_impl< std::complex< float > >

struct	type_impl< unsigned >

struct	type_impl< unsigned char >

struct	type_impl< unsigned long >

struct	type_impl< unsigned long long >

struct	type_impl< unsigned short >

class	window
	This class wraps the MPI_Window class with RAII functionality. More...

Enumerations
enum	thread_type { serialized = MPI_THREAD_SERIALIZED, funneled = MPI_THREAD_FUNNELED, single = MPI_THREAD_SINGLE, multiple = MPI_THREAD_MULTIPLE }
	This enum specifies the threading type to be used when creating an MPI environment.

Functions
constexpr bool	is_gpu_aware ()
	Return if GPU aware functionality is available.

int	map_rank_to_device_id (MPI_Comm comm, int num_devices)
	Maps each MPI rank to a single device id in a round robin manner. More...

std::vector< status >	wait_all (std::vector< request > &req)
	Allows a rank to wait on multiple request handles. More...

bool	requires_host_buffer (const std::shared_ptr< const Executor > &exec, const communicator &comm)
	Checks if the combination of Executor and communicator requires passing MPI buffers from the host memory.

double	get_walltime ()
	Get the rank in the communicator of the calling process. More...

Detailed Description

The mpi namespace, contains wrapper for many MPI functions.

Function Documentation

double gko::experimental::mpi::get_walltime ( )

inline

Get the rank in the communicator of the calling process.

Parameters

comm	the communicator

int gko::experimental::mpi::map_rank_to_device_id	(	MPI_Comm	comm,
		int	num_devices
	)

Maps each MPI rank to a single device id in a round robin manner.

Parameters

comm	used to determine the node-local rank, if no suitable environment variable is available.
num_devices	the number of devices per node.

std::vector<status> gko::experimental::mpi::wait_all ( std::vector< request > & req )

inline

Allows a rank to wait on multiple request handles.

Parameters

req	The vector of request handles to be waited on.