count - count of elements in buffer (integer)
datatype - type of elements in buffer (handle)
source - source rank (integer)
tag - communication tag (integer)
comm - communicator for message (handle)
buf - buffer (array)
rq - request handle for this persistent receive (handle)
The MPI_Recv_init() function is used to create a persistent request that will perform a receive into a buffer when MPI_Start() or MPI_Startall() is used. When that happens, an asynchronous receive will be posted. The receive will complete when a matching message is posted by a function similar to MPI_Send() and a function like MPI_Wait() is called for the request rq . If a message matches the required source , tag , and communicator, comm , then it will be copied into the buffer buf . If the message is larger than the buffer, it will be truncated and the MPI_ERROR field of the status will be set to MPI_ERR_TRUNCATE when the request completes. The datatype and count determine exactly how the message is interpreted. Note that for performance reasons this MPI library does not check that the sent and received datatypes match.
If tag is MPI_ANY_TAG then any tag will match. Similarly, if source is MPI_ANY_SOURCE then so will a message from any source. The original tag and source are stored in the MPI_TAG and MPI_SOURCE fields of the status respectively. One additional field holds the total size of the message in bytes.
When the request is complete, it may be restarted with another call to MPI_Start() or MPI_Startall(). When the persistent request is no longer needed, it should be freed with MPI_Request_free().
If the source is MPI_PROC_NULL then no receive is ever posted, and MPI_SUCCESS is always returned. In this case, when the completed request rq is waited on, the status is filled with MPI_ANY_TAG MPI_PROC_NULL MPI_SUCCESS and zero size in bytes.
The C++ version of this function passes the resulting persistent request as the return value.
The communicator must be a valid one (not MPI_COMM_NULL) and the request must not be MPI_REQUEST_NULL. PMPI_Recv_init() is the profiling version of this function.
All MPI routines except for MPI_Wtime and MPI_Wtick return an error code. The the current MPI error handler is invoked if the return value is not MPI_SUCCESS. The default error handler aborts, but this may be changed with by using the MPI_Errhandler_set() function. The predefined error handler MPI_ERRORS_RETURN may be used to cause error values to be returned instead. Note that MPI does not guarentee that an MPI program can continue past an error. In this implementation, all errors except MPI_ERR_INTERN or MPI_ERR_OTHER should always be recoverable.
In C, the error code is passed as the return value. In FORTRAN, all functions have an parameter ierr which returns the error code. MPI C++ functions do not directly return an error code. However, C++ users may want to use the MPI::ERRORS_THROW_EXCEPTIONS handler. This will throw an MPI::Exception with the corresponding error code. To prevent exceptions from being raised from within C and Fortran code, they will see all error return values as MPI_ERR_PENDING when this handler is chosen. In this implementation, call MPI::throw_exception() to throw the correct exception if this occurs.
MPI_SUCCESS - No error;
MPI_ERR_PENDING - Pending exception;
MPI_ERR_COMM - Invalid communicator;
MPI_ERR_COUNT - Invalid element count;
MPI_ERR_BUFFER - Invalid buffer;
MPI_ERR_TAG - Invalid tag;
MPI_ERR_TYPE - Invalid data type;
MPI_ERR_RANK - Invalid destination;
MPI_ERR_REQUEST - Invalid request;
MPI_ERR_INTERN - Out of Memory.
Company Info |
Product Index |
Category Index |
Copyright © Lockless Inc All Rights Reserved.