18.2.14. MPI_Allreduce

MPI_Allreduce, MPI_Iallreduce, MPI_Allreduce_init - Combines values from all processes and distributes the result back to all processes.

18.2.14.1. SYNTAX

18.2.14.1.1. C Syntax

#include <mpi.h>

int MPI_Allreduce(const void *sendbuf, void *recvbuf, int count,
                  MPI_Datatype datatype, MPI_Op op, MPI_Comm comm)

int MPI_Iallreduce(const void *sendbuf, void *recvbuf, int count,
                   MPI_Datatype datatype, MPI_Op op, MPI_Comm comm,
                   MPI_Request *request)

int MPI_Allreduce_init(const void *sendbuf, void *recvbuf, int count,
                       MPI_Datatype datatype, MPI_Op op, MPI_Comm comm,
                       MPI_Info info, MPI_Request *request)

18.2.14.1.2. Fortran Syntax

USE MPI
! or the older form: INCLUDE 'mpif.h'
MPI_ALLREDUCE(SENDBUF, RECVBUF, COUNT, DATATYPE, OP, COMM, IERROR)
     <type>  SENDBUF(*), RECVBUF(*)
     INTEGER COUNT, DATATYPE, OP, COMM, IERROR

MPI_IALLREDUCE(SENDBUF, RECVBUF, COUNT, DATATYPE, OP, COMM, REQUEST, IERROR)
     <type>  SENDBUF(*), RECVBUF(*)
     INTEGER COUNT, DATATYPE, OP, COMM, REQUEST, IERROR

MPI_ALLREDUCE_INIT(SENDBUF, RECVBUF, COUNT, DATATYPE, OP, COMM, INFO, REQUEST, IERROR)
     <type>  SENDBUF(*), RECVBUF(*)
     INTEGER COUNT, DATATYPE, OP, COMM, INFO, REQUEST, IERROR

18.2.14.1.3. Fortran 2008 Syntax

USE mpi_f08
MPI_Allreduce(sendbuf, recvbuf, count, datatype, op, comm, ierror)
     TYPE(*), DIMENSION(..), INTENT(IN) :: sendbuf
     TYPE(*), DIMENSION(..) :: recvbuf
     INTEGER, INTENT(IN) :: count
     TYPE(MPI_Datatype), INTENT(IN) :: datatype
     TYPE(MPI_Op), INTENT(IN) :: op
     TYPE(MPI_Comm), INTENT(IN) :: comm
     INTEGER, OPTIONAL, INTENT(OUT) :: ierror

MPI_Iallreduce(sendbuf, recvbuf, count, datatype, op, comm, request,
             ierror)
     TYPE(*), DIMENSION(..), INTENT(IN), ASYNCHRONOUS :: sendbuf
     TYPE(*), DIMENSION(..), ASYNCHRONOUS :: recvbuf
     INTEGER, INTENT(IN) :: count
     TYPE(MPI_Datatype), INTENT(IN) :: datatype
     TYPE(MPI_Op), INTENT(IN) :: op
     TYPE(MPI_Comm), INTENT(IN) :: comm
     TYPE(MPI_Request), INTENT(OUT) :: request
     INTEGER, OPTIONAL, INTENT(OUT) :: ierror

MPI_Allreduce_init(sendbuf, recvbuf, count, datatype, op, comm, info, request,
             ierror)
     TYPE(*), DIMENSION(..), INTENT(IN), ASYNCHRONOUS :: sendbuf
     TYPE(*), DIMENSION(..), ASYNCHRONOUS :: recvbuf
     INTEGER, INTENT(IN) :: count
     TYPE(MPI_Datatype), INTENT(IN) :: datatype
     TYPE(MPI_Op), INTENT(IN) :: op
     TYPE(MPI_Comm), INTENT(IN) :: comm
     TYPE(MPI_Info), INTENT(IN) :: info
     TYPE(MPI_Request), INTENT(OUT) :: request
     INTEGER, OPTIONAL, INTENT(OUT) :: ierror

18.2.14.2. INPUT PARAMETERS

  • sendbuf: Starting address of send buffer (choice).

  • count: Number of elements in send buffer (integer).

  • datatype: Datatype of elements of send buffer (handle).

  • op: Operation (handle).

  • comm: Communicator (handle).

  • info: Info (handle, persistent only).

18.2.14.3. OUTPUT PARAMETERS

  • recvbuf: Starting address of receive buffer (choice).

  • request: Request (handle, non-blocking only).

  • ierror: Fortran only: Error status (integer).

18.2.14.4. DESCRIPTION

Same as MPI_Reduce except that the result appears in the receive buffer of all the group members.

Example 1: A routine that computes the product of a vector and an array that are distributed across a group of processes and returns the answer at all nodes (compare with Example 2, with MPI_Reduce, below).

SUBROUTINE PAR_BLAS2(m, n, a, b, c, comm)
REAL a(m), b(m,n)    ! local slice of array
REAL c(n)            ! result
REAL sum(n)
INTEGER n, comm, i, j, ierr

! local sum
DO j= 1, n
  sum(j) = 0.0
  DO i = 1, m
    sum(j) = sum(j) + a(i)*b(i,j)
  END DO
END DO

! global sum
CALL MPI_ALLREDUCE(sum, c, n, MPI_REAL, MPI_SUM, comm, ierr)

! return result at all nodes
RETURN

Example 2: A routine that computes the product of a vector and an array that are distributed across a group of processes and returns the answer at node zero.

SUBROUTINE PAR_BLAS2(m, n, a, b, c, comm)
REAL a(m), b(m,n)    ! local slice of array
REAL c(n)            ! result
REAL sum(n)
INTEGER n, comm, i, j, ierr

! local sum
DO j= 1, n
  sum(j) = 0.0
  DO i = 1, m
    sum(j) = sum(j) + a(i)*b(i,j)
  END DO
END DO

! global sum
CALL MPI_REDUCE(sum, c, n, MPI_REAL, MPI_SUM, 0, comm, ierr)

! return result at node zero (and garbage at the other nodes)
RETURN

18.2.14.5. USE OF IN-PLACE OPTION

When the communicator is an intracommunicator, you can perform an all-reduce operation in-place (the output buffer is used as the input buffer). Use the variable MPI_IN_PLACE as the value of sendbuf at all processes.

Note that MPI_IN_PLACE is a special kind of value; it has the same restrictions on its use as MPI_BOTTOM.

Because the in-place option converts the receive buffer into a send-and-receive buffer, a Fortran binding that includes INTENT must mark these as INOUT, not OUT.

18.2.14.6. WHEN COMMUNICATOR IS AN INTER-COMMUNICATOR

When the communicator is an inter-communicator, the reduce operation occurs in two phases. The data is reduced from all the members of the first group and received by all the members of the second group. Then the data is reduced from all the members of the second group and received by all the members of the first. The operation exhibits a symmetric, full-duplex behavior.

When the communicator is an intra-communicator, these groups are the same, and the operation occurs in a single phase.

18.2.14.7. NOTES ON COLLECTIVE OPERATIONS

The reduction functions ( MPI_Op ) do not return an error value. As a result, if the functions detect an error, all they can do is either call MPI_Abort or silently skip the problem. Thus, if you change the error handler from MPI_ERRORS_ARE_FATAL to something else, for example, MPI_ERRORS_RETURN , then no error may be indicated.

18.2.14.8. ERRORS

Almost all MPI routines return an error value; C routines as the return result of the function and Fortran routines in the last argument.

Before the error value is returned, the current MPI error handler associated with the communication object (e.g., communicator, window, file) is called. If no communication object is associated with the MPI call, then the call is considered attached to MPI_COMM_SELF and will call the associated MPI error handler. When MPI_COMM_SELF is not initialized (i.e., before MPI_Init/MPI_Init_thread, after MPI_Finalize, or when using the Sessions Model exclusively) the error raises the initial error handler. The initial error handler can be changed by calling MPI_Comm_set_errhandler on MPI_COMM_SELF when using the World model, or the mpi_initial_errhandler CLI argument to mpiexec or info key to MPI_Comm_spawn/MPI_Comm_spawn_multiple. If no other appropriate error handler has been set, then the MPI_ERRORS_RETURN error handler is called for MPI I/O functions and the MPI_ERRORS_ABORT error handler is called for all other MPI functions.

Open MPI includes three predefined error handlers that can be used:

  • MPI_ERRORS_ARE_FATAL Causes the program to abort all connected MPI processes.

  • MPI_ERRORS_ABORT An error handler that can be invoked on a communicator, window, file, or session. When called on a communicator, it acts as if MPI_Abort was called on that communicator. If called on a window or file, acts as if MPI_Abort was called on a communicator containing the group of processes in the corresponding window or file. If called on a session, aborts only the local process.

  • MPI_ERRORS_RETURN Returns an error code to the application.

MPI applications can also implement their own error handlers by calling:

Note that MPI does not guarantee that an MPI program can continue past an error.

See the MPI man page for a full list of MPI error codes.

See the Error Handling section of the MPI-3.1 standard for more information.