ompi/request: Fix a persistent request creation bug

According to the MPI-3.1 p.52 and p.53 (cited below), a request
created by `MPI_*_INIT` but not yet started by `MPI_START` or
`MPI_STARTALL` is inactive therefore `MPI_WAIT` or its friends
must return immediately if such a request is passed.

The current implementation hangs in `MPI_WAIT` and its friends
in such case because a persistent request is initialized as
`req_complete = REQUEST_PENDING`. This commit fixes the
initialization.

Also, this commit fixes internal requests used in `MPI_PROBE`
and `MPI_IPROBE` which was marked wrongly as persistent.

MPI-3.1 p.52:

We shall use the following terminology: A null handle is a handle
with value MPI_REQUEST_NULL. A persistent request and the handle
to it are inactive if the request is not associated with any ongoing
communication (see Section 3.9). A handle is active if it is neither
null nor inactive. An empty status is a status which is set to return
tag = MPI_ANY_TAG, source = MPI_ANY_SOURCE, error = MPI_SUCCESS, and
is also internally configured so that calls to MPI_GET_COUNT,
MPI_GET_ELEMENTS, and MPI_GET_ELEMENTS_X return count = 0 and
MPI_TEST_CANCELLED returns false. We set a status variable to empty
when the value returned by it is not significant. Status is set in
this way so as to prevent errors due to accesses of stale information.

MPI-3.1 p.53:

One is allowed to call MPI_WAIT with a null or inactive request
argument. In this case the operation returns immediately with empty
status.

Signed-off-by: KAWASHIMA Takahiro <[email protected]>
(back-ported from commit 6510800)

Author: kawashima-fj

ompi/mca/pml/ob1/pml_ob1_iprobe.c

@@ -36,7 +36,7 @@ int mca_pml_ob1_iprobe(int src,
     recvreq.req_recv.req_base.req_ompi.req_type = OMPI_REQUEST_PML;
     recvreq.req_recv.req_base.req_type = MCA_PML_REQUEST_IPROBE;
 
-    MCA_PML_OB1_RECV_REQUEST_INIT(&recvreq, NULL, 0, &ompi_mpi_char.dt, src, tag, comm, true);
+    MCA_PML_OB1_RECV_REQUEST_INIT(&recvreq, NULL, 0, &ompi_mpi_char.dt, src, tag, comm, false);
     MCA_PML_OB1_RECV_REQUEST_START(&recvreq);
 
     if( REQUEST_COMPLETE( &(recvreq.req_recv.req_base.req_ompi)) ) {
@@ -66,7 +66,7 @@ int mca_pml_ob1_probe(int src,
     recvreq.req_recv.req_base.req_ompi.req_type = OMPI_REQUEST_PML;
     recvreq.req_recv.req_base.req_type = MCA_PML_REQUEST_PROBE;
 
-    MCA_PML_OB1_RECV_REQUEST_INIT(&recvreq, NULL, 0, &ompi_mpi_char.dt, src, tag, comm, true);
+    MCA_PML_OB1_RECV_REQUEST_INIT(&recvreq, NULL, 0, &ompi_mpi_char.dt, src, tag, comm, false);
     MCA_PML_OB1_RECV_REQUEST_START(&recvreq);
 
     ompi_request_wait_completion(&recvreq.req_recv.req_base.req_ompi);

ompi/request/request.h

@@ -142,13 +142,14 @@ typedef struct ompi_predefined_request_t ompi_predefined_request_t;
  * performance path (since requests may be re-used, it is possible
  * that we will have to initialize a request multiple times).
  */
-#define OMPI_REQUEST_INIT(request, persistent)        \
-    do {                                              \
-        (request)->req_complete = REQUEST_PENDING;    \
-        (request)->req_state = OMPI_REQUEST_INACTIVE; \
-        (request)->req_persistent = (persistent);     \
-        (request)->req_complete_cb  = NULL;           \
-        (request)->req_complete_cb_data = NULL;       \
+#define OMPI_REQUEST_INIT(request, persistent)                  \
+    do {                                                        \
+        (request)->req_complete =                               \
+            (persistent) ? REQUEST_COMPLETED : REQUEST_PENDING; \
+        (request)->req_state = OMPI_REQUEST_INACTIVE;           \
+        (request)->req_persistent = (persistent);               \
+        (request)->req_complete_cb  = NULL;                     \
+        (request)->req_complete_cb_data = NULL;                 \
     } while (0);