Fix MPI_Alltoallv in coll/tuned.

Correctly handle the corner case in MPI_Alltoallv when
some tasks have no data to transfer and some other tasks
do have data to transfer.

This test case is covered in ibm/collective/alltoallv_somezeros
from the ompi-tests repo.

cmr=v1.8.2:reviewer=bosilca

This commit was SVN r31985.

ompi/mca/coll/tuned/coll_tuned_util.c

@@ -9,6 +9,8 @@
  *                         University of Stuttgart.  All rights reserved.
  * Copyright (c) 2004-2005 The Regents of the University of California.
  *                         All rights reserved.
+ * Copyright (c) 2014      Research Organization for Information Science
+ *                         and Technology (RIST). All rights reserved.
  * $COPYRIGHT$
  * 
  * Additional copyrights may follow
@@ -90,6 +92,82 @@ int ompi_coll_tuned_sendrecv_actual( void* sendbuf, size_t scount,
     return (err);
 }
 
+int ompi_coll_tuned_sendrecv_nonzero_actual( void* sendbuf, size_t scount, 
+                                             ompi_datatype_t* sdatatype,
+                                             int dest, int stag,
+                                             void* recvbuf, size_t rcount, 
+                                             ompi_datatype_t* rdatatype,
+                                             int source, int rtag,
+                                             struct ompi_communicator_t* comm,
+                                             ompi_status_public_t* status )
+
+{ /* post receive first, then send, then waitall... should be fast (I hope) */
+    int err, line = 0, nreqs = 0;
+    size_t typesize;
+    ompi_request_t* reqs[2], **req = reqs;
+    ompi_status_public_t statuses[2];
+
+    /* post new irecv */
+    ompi_datatype_type_size(rdatatype, &typesize);
+    if (0 != rcount && 0 != typesize) {
+        err = MCA_PML_CALL(irecv( recvbuf, rcount, rdatatype, source, rtag, 
+                                  comm, req++));
+        ++nreqs;
+        if (err != MPI_SUCCESS) { line = __LINE__; goto error_handler; }
+    }
+
+    /* send data to children */
+    ompi_datatype_type_size(sdatatype, &typesize);
+    if (0 != scount && 0 != typesize) {
+        err = MCA_PML_CALL(isend( sendbuf, scount, sdatatype, dest, stag, 
+                                  MCA_PML_BASE_SEND_STANDARD, comm, req++));
+        ++nreqs;
+        if (err != MPI_SUCCESS) { line = __LINE__; goto error_handler; }
+    }
+
+    if (0 != nreqs) {
+        err = ompi_request_wait_all( nreqs, reqs, statuses );
+        if (err != MPI_SUCCESS) { line = __LINE__; goto error_handler; }
+
+        if (MPI_STATUS_IGNORE != status) {
+            *status = statuses[0];
+        }
+    } else {
+        /* FIXME this is currently unsupported but unused */
+        assert (MPI_STATUS_IGNORE == status);
+    }
+    
+    return (MPI_SUCCESS);
+
+ error_handler:
+    /* As we use wait_all we will get MPI_ERR_IN_STATUS which is not an error
+     * code that we can propagate up the stack. Instead, look for the real
+     * error code from the MPI_ERROR in the status.
+     */
+    if( MPI_ERR_IN_STATUS == err ) {
+        /* At least we know the error was detected during the wait_all */
+        int err_index = 0;
+        if( MPI_SUCCESS == statuses[0].MPI_ERROR ) {
+            err_index = 1;
+        }
+        if (MPI_STATUS_IGNORE != status) {
+            *status = statuses[err_index];
+        }
+        err = statuses[err_index].MPI_ERROR;
+        OPAL_OUTPUT ((ompi_coll_tuned_stream, "%s:%d: Error %d occurred (req index %d)\n",
+                      __FILE__, line, err, err_index));
+    } else {
+        /* Error discovered during the posting of the irecv or isend,
+         * and no status is available.
+         */
+        OPAL_OUTPUT ((ompi_coll_tuned_stream, "%s:%d: Error %d occurred\n",
+                      __FILE__, line, err));
+        if (MPI_STATUS_IGNORE != status) {
+            status->MPI_ERROR = err;
+        }
+    }
+    return (err);
+}
 /*
  * localcompleted version that makes sure the send has completed locally 
  * Currently this is a sync call, but will change to locally completed

ompi/mca/coll/tuned/coll_tuned_util.h

@@ -9,6 +9,8 @@
  *                         University of Stuttgart.  All rights reserved.
  * Copyright (c) 2004-2005 The Regents of the University of California.
  *                         All rights reserved.
+ * Copyright (c) 2014      Research Organization for Information Science
+ *                         and Technology (RIST). All rights reserved.
  * $COPYRIGHT$
  * 
  * Additional copyrights may follow
@@ -39,6 +41,15 @@ int ompi_coll_tuned_sendrecv_actual( void* sendbuf, size_t scount,
                                      struct ompi_communicator_t* comm,
                                      ompi_status_public_t* status );
 
+int ompi_coll_tuned_sendrecv_nonzero_actual( void* sendbuf, size_t scount, 
+                                             ompi_datatype_t* sdatatype,
+                                             int dest, int stag,
+                                             void* recvbuf, size_t rcount, 
+                                             ompi_datatype_t* rdatatype,
+                                             int source, int rtag,
+                                             struct ompi_communicator_t* comm,
+                                             ompi_status_public_t* status );
+
 
 /* inline functions */
 
@@ -54,7 +65,7 @@ ompi_coll_tuned_sendrecv( void* sendbuf, size_t scount, ompi_datatype_t* sdataty
         return (int) ompi_datatype_sndrcv(sendbuf, (int32_t) scount, sdatatype, 
                                           recvbuf, (int32_t) rcount, rdatatype);
     }
-    return ompi_coll_tuned_sendrecv_actual (sendbuf, scount, sdatatype, 
+    return ompi_coll_tuned_sendrecv_nonzero_actual (sendbuf, scount, sdatatype, 
                                             dest, stag, 
                                             recvbuf, rcount, rdatatype,
                                             source, rtag, comm, status);