int mca_pml_ob1_send(const void *buf,
size_t count,
ompi_datatype_t * datatype,
int dst,
int tag,
mca_pml_base_send_mode_t sendmode,
ompi_communicator_t * comm)
{
mca_pml_ob1_comm_proc_t *ob1_proc = mca_pml_ob1_peer_lookup (comm, dst);
ompi_proc_t *dst_proc = ob1_proc->ompi_proc;
mca_bml_base_endpoint_t* endpoint = mca_bml_base_get_endpoint (dst_proc);
mca_pml_ob1_send_request_t *sendreq = NULL;
int16_t seqn;
int rc;
if (OPAL_UNLIKELY(NULL == endpoint)) {
return OMPI_ERR_UNREACH;
}
if (OPAL_UNLIKELY(MCA_PML_BASE_SEND_BUFFERED == sendmode)) {
/* large buffered sends *need* a real request so use isend instead */
ompi_request_t *brequest;
rc = mca_pml_ob1_isend (buf, count, datatype, dst, tag, sendmode, comm, &brequest);
if (OPAL_UNLIKELY(OMPI_SUCCESS != rc)) {
return rc;
}
ompi_request_wait_completion (brequest);
ompi_request_free (&brequest);
return OMPI_SUCCESS;
}
seqn = (uint16_t) OPAL_THREAD_ADD32(&ob1_proc->send_sequence, 1);
/**
* The immediate send will not have a request, so they are
* intracable from the point of view of any debugger attached to
* the parallel application.
*/
if (MCA_PML_BASE_SEND_SYNCHRONOUS != sendmode) {
rc = mca_pml_ob1_send_inline (buf, count, datatype, dst, tag, seqn, dst_proc,
endpoint, comm);
if (OPAL_LIKELY(0 <= rc)) {
return OMPI_SUCCESS;
}
}
if (OPAL_LIKELY(!ompi_mpi_thread_multiple)) {
sendreq = mca_pml_ob1_sendreq;
mca_pml_ob1_sendreq = NULL;
}
if( OPAL_UNLIKELY(NULL == sendreq) ) {
MCA_PML_OB1_SEND_REQUEST_ALLOC(comm, dst, sendreq);
if (NULL == sendreq)
return OMPI_ERR_TEMP_OUT_OF_RESOURCE;
}
sendreq->req_send.req_base.req_proc = dst_proc;
sendreq->rdma_frag = NULL;
MCA_PML_OB1_SEND_REQUEST_INIT(sendreq,
buf,
count,
datatype,
dst, tag,
comm, sendmode, false);
PERUSE_TRACE_COMM_EVENT (PERUSE_COMM_REQ_ACTIVATE,
&sendreq->req_send.req_base,
PERUSE_SEND);
MCA_PML_OB1_SEND_REQUEST_START_W_SEQ(sendreq, endpoint, seqn, rc);
if (OPAL_LIKELY(rc == OMPI_SUCCESS)) {
ompi_request_wait_completion(&sendreq->req_send.req_base.req_ompi);
rc = sendreq->req_send.req_base.req_ompi.req_status.MPI_ERROR;
}
if (OPAL_UNLIKELY(ompi_mpi_thread_multiple || NULL != mca_pml_ob1_sendreq)) {
MCA_PML_OB1_SEND_REQUEST_RETURN(sendreq);
} else {
mca_pml_ob1_send_request_fini (sendreq);
mca_pml_ob1_sendreq = sendreq;
}
return rc;
}
int mca_pml_ob1_send(void *buf,
size_t count,
ompi_datatype_t * datatype,
int dst,
int tag,
mca_pml_base_send_mode_t sendmode,
ompi_communicator_t * comm)
{
mca_pml_ob1_comm_t* ob1_comm = comm->c_pml_comm;
ompi_proc_t *dst_proc = ompi_comm_peer_lookup (comm, dst);
mca_bml_base_endpoint_t* endpoint = (mca_bml_base_endpoint_t*)
dst_proc->proc_endpoints[OMPI_PROC_ENDPOINT_TAG_BML];
mca_pml_ob1_send_request_t *sendreq = NULL;
int16_t seqn;
int rc;
if (OPAL_UNLIKELY(MCA_PML_BASE_SEND_BUFFERED == sendmode)) {
/* large buffered sends *need* a real request so use isend instead */
ompi_request_t *brequest;
rc = mca_pml_ob1_isend (buf, count, datatype, dst, tag, sendmode, comm, &brequest);
if (OPAL_UNLIKELY(OMPI_SUCCESS != rc)) {
return rc;
}
/* free the request and return. don't care if it completes now */
ompi_request_free (&brequest);
return OMPI_SUCCESS;
}
if (OPAL_UNLIKELY(NULL == endpoint)) {
return OMPI_ERR_UNREACH;
}
seqn = (uint16_t) OPAL_THREAD_ADD32(&ob1_comm->procs[dst].send_sequence, 1);
/**
* The immediate send will not have a request, so they are
* intracable from the point of view of any debugger attached to
* the parallel application.
*/
if (MCA_PML_BASE_SEND_SYNCHRONOUS != sendmode) {
rc = mca_pml_ob1_send_inline (buf, count, datatype, dst, tag, seqn, dst_proc,
endpoint, comm);
if (OPAL_LIKELY(0 <= rc)) {
return OMPI_SUCCESS;
}
}
#if !OMPI_ENABLE_THREAD_MULTIPLE
sendreq = mca_pml_ob1_sendreq;
if( OPAL_UNLIKELY(NULL == sendreq) )
#endif /* !OMPI_ENABLE_THREAD_MULTIPLE */
{
MCA_PML_OB1_SEND_REQUEST_ALLOC(comm, dst, sendreq);
if (NULL == sendreq)
return OMPI_ERR_TEMP_OUT_OF_RESOURCE;
#if !OMPI_ENABLE_THREAD_MULTIPLE
mca_pml_ob1_sendreq = sendreq;
#endif /* !OMPI_ENABLE_THREAD_MULTIPLE */
}
sendreq->req_send.req_base.req_proc = dst_proc;
sendreq->rdma_frag = NULL;
MCA_PML_OB1_SEND_REQUEST_INIT(sendreq,
buf,
count,
datatype,
dst, tag,
comm, sendmode, false);
PERUSE_TRACE_COMM_EVENT (PERUSE_COMM_REQ_ACTIVATE,
&sendreq->req_send.req_base,
PERUSE_SEND);
MCA_PML_OB1_SEND_REQUEST_START_W_SEQ(sendreq, endpoint, seqn, rc);
if (OPAL_LIKELY(rc == OMPI_SUCCESS)) {
ompi_request_wait_completion(&sendreq->req_send.req_base.req_ompi);
rc = sendreq->req_send.req_base.req_ompi.req_status.MPI_ERROR;
}
#if OMPI_ENABLE_THREAD_MULTIPLE
MCA_PML_OB1_SEND_REQUEST_RETURN(sendreq);
#else
mca_pml_ob1_send_request_fini (sendreq);
#endif
return rc;
}
