MPI Point to Point Functions - Message Passing Interface (original) (raw)

• 2018-09-14

In this article

In this section

• MPI_Bsend
  Sends data to a specified process in buffered mode.
• MPI_Bsend_init
  Builds a handle for a buffered send.
• MPI_Cancel
  Cancels a communication request.
• MPI_Get_count
  Gets the number of "top level" elements.
• MPI_Ibsend
  Initiates a buffered mode send operation and returns a handle to the communication operation.
• MPI_Iprobe
  Nonblocking test for a message.
• MPI_Improbe
  Probes for a message in a non-blocking way. Provides a mechanism to receive the specific message that was matched regardless of intervening probe/receive operations. The matched message is de-queued off the receive queue, giving the application an opportunity to decide how to receive the message based on the information returned by the non-blocking matching probe operation. The matched message is then received using the MPI_Mrecv or MPI_Imrecv function.
• MPI_Imrecv
  Performs a non-blocking receive for a message matched by MPI_Mprobe or MPI_Improbe.
• MPI_Irecv
  Initiates a receive operation and returns a handle to the requested communication operation.
• MPI_Irsend
  Initiates a ready mode send operation and returns a request handle that represents the communication operation.
• MPI_Isend
  Initiates a standard mode send operation and returns a handle to the requested communication operation.
• MPI_Issend
  Initiates a synchronous mode send operation and returns a handle to the requested communication operation.
• MPI_Mprobe
  Blocking probes for a message. Provides a mechanism to receive the specific message that was matched regardless of intervening probe/receive operations. The matched message is de-queued off the receive queue, giving the application an opportunity to decide how to receive the message based on the information returned by the matching probe operation. The matched message is then received using the MPI_Mrecv or MPI_Imrecv function.
• MPI_Mrecv
  Performs a blocking receive for a message matched by MPI_Mprobe or MPI_Improbe.
• MPI_Probe
  Blocking test for a message.
• MPI_Recv
  Performs a receive operation and does not return until a matching message is received.
• MPI_Recv_init
  Create a persistent request for a receive.
• MPI_Request_free
  Frees a communication request object.
• MPI_Request_get_status
  Nondestructive test for the completion of a request.
• MPI_Rsend
  Performs a ready mode send operation and returns when the send buffer can be safely reused.
• MPI_Rsend_init
  Creates a persistent request for a ready send.
• MPI_Send
  Performs a standard mode send operation and returns when the send buffer can be safely reused.
• MPI_Send_init
  Creates a persistent request for a standard send.
• MPI_Sendrecv
  Sends and receives a message.
• MPI_Sendrecv_replace
  Sends and receives using a single buffer.
• MPI_Ssend
  Performs a synchronous mode send operation and returns when the send buffer can be safely reused.
• MPI_Ssend_init
  Creates a persistent request for a synchronous send.
• MPI_Start
  Initiates a communication with a persistent request handle.
• MPI_Startall
  Starts a collection of persistent requests.
• MPI_Test
  Tests an outstanding operation for completion.
• MPI_Test_cancelled
  Tests to see if a request was cancelled.
• MPI_Testall
  Tests for the completion of all previously initiated requests.
• MPI_Testany
  Tests for completion of any previdously initiated requests.
• MPI_Testsome
  Tests for completion of some of previously initiated requests.
• MPI_Wait
  Completes an outstanding operation.
• MPI_Waitall
  Completes multiple outstanding operations.
• MPI_Waitany
  Completes one out of several outstanding operations.
• MPI_Waitsome
  Waits for some of given MPI requests to complete.
• MSMPI_Queuelock_acquire
  Acquires the Microsoft MPI library global lock.
• MSMPI_Queuelock_release
  Releases the Microsoft MPI library global lock.
• MSMPI_Waitsome_interruptible
  Waits until at least one of the operations that is associated with active handles in the list has finished, or the call is interrupted by another thread that calls MSMPI_Queuelock_acquire.

Feedback