v3.0.x: pml/ob1: fixed OOS message handling.

ompi/mca/pml/ob1/pml_ob1.c

@@ -3,7 +3,7 @@
  * Copyright (c) 2004-2010 The Trustees of Indiana University and Indiana
  *                         University Research and Technology
  *                         Corporation.  All rights reserved.
- * Copyright (c) 2004-2012 The University of Tennessee and The University
+ * Copyright (c) 2004-2018 The University of Tennessee and The University
  *                         of Tennessee Research Foundation.  All rights
  *                         reserved.
  * Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
@@ -223,8 +223,6 @@ int mca_pml_ob1_add_comm(ompi_communicator_t* comm)
         opal_list_remove_item (&mca_pml_ob1.non_existing_communicator_pending,
                                (opal_list_item_t *) frag);
 
-      add_fragment_to_unexpected:
-
         /* We generate the MSG_ARRIVED event as soon as the PML is aware
          * of a matching fragment arrival. Independing if it is received
          * on the correct order or not. This will allow the tools to
@@ -242,7 +240,9 @@ int mca_pml_ob1_add_comm(ompi_communicator_t* comm)
          */
         pml_proc = mca_pml_ob1_peer_lookup(comm, hdr->hdr_src);
 
-        if( ((uint16_t)hdr->hdr_seq) == ((uint16_t)pml_proc->expected_sequence) ) {
+        if (((uint16_t)hdr->hdr_seq) == ((uint16_t)pml_proc->expected_sequence) ) {
+
+        add_fragment_to_unexpected:
             /* We're now expecting the next sequence number. */
             pml_proc->expected_sequence++;
             opal_list_append( &pml_proc->unexpected_frags, (opal_list_item_t*)frag );
@@ -254,19 +254,16 @@ int mca_pml_ob1_add_comm(ompi_communicator_t* comm)
              * situation as the cant_match is only checked when a new fragment is received from
              * the network.
              */
-           for(frag = (mca_pml_ob1_recv_frag_t *)opal_list_get_first(&pml_proc->frags_cant_match);
-               frag != (mca_pml_ob1_recv_frag_t *)opal_list_get_end(&pml_proc->frags_cant_match);
-               frag = (mca_pml_ob1_recv_frag_t *)opal_list_get_next(frag)) {
-               hdr = &frag->hdr.hdr_match;
-               /* If the message has the next expected seq from that proc...  */
-               if(hdr->hdr_seq != pml_proc->expected_sequence)
-                   continue;
-
-               opal_list_remove_item(&pml_proc->frags_cant_match, (opal_list_item_t*)frag);
-               goto add_fragment_to_unexpected;
-           }
+            if( NULL != pml_proc->frags_cant_match ) {
+                frag = check_cantmatch_for_match(pml_proc);
+                if( NULL != frag ) {
+                    hdr = &frag->hdr.hdr_match;
+                    goto add_fragment_to_unexpected;
+                }
+            }
         } else {
-            opal_list_append( &pml_proc->frags_cant_match, (opal_list_item_t*)frag );
+            append_frag_to_ordered_list(&pml_proc->frags_cant_match, frag,
+                                        pml_proc->expected_sequence);
         }
     }
     return OMPI_SUCCESS;
@@ -553,6 +550,23 @@ static void mca_pml_ob1_dump_frag_list(opal_list_t* queue, bool is_req)
     }
 }
 
+void mca_pml_ob1_dump_cant_match(mca_pml_ob1_recv_frag_t* queue)
+{
+    mca_pml_ob1_recv_frag_t* item = queue;
+
+    do {
+        mca_pml_ob1_dump_hdr( &item->hdr );
+        if( NULL != item->range ) {
+            mca_pml_ob1_recv_frag_t* frag = item->range;
+            do {
+                mca_pml_ob1_dump_hdr( &frag->hdr );
+                frag = (mca_pml_ob1_recv_frag_t*)frag->super.super.opal_list_next;
+            } while( frag != item->range );
+        }
+        item = (mca_pml_ob1_recv_frag_t*)item->super.super.opal_list_next;
+    } while( item != queue );
+}
+
 int mca_pml_ob1_dump(struct ompi_communicator_t* comm, int verbose)
 {
     struct mca_pml_comm_t* pml_comm = comm->c_pml_comm;
@@ -588,9 +602,9 @@ int mca_pml_ob1_dump(struct ompi_communicator_t* comm, int verbose)
             opal_output(0, "expected specific receives\n");
             mca_pml_ob1_dump_frag_list(&proc->specific_receives, true);
         }
-        if( opal_list_get_size(&proc->frags_cant_match) ) {
+        if( NULL != proc->frags_cant_match ) {
             opal_output(0, "out of sequence\n");
-            mca_pml_ob1_dump_frag_list(&proc->frags_cant_match, false);
+            mca_pml_ob1_dump_cant_match(proc->frags_cant_match);
         }
         if( opal_list_get_size(&proc->unexpected_frags) ) {
             opal_output(0, "unexpected frag\n");

ompi/mca/pml/ob1/pml_ob1.h

@@ -3,7 +3,7 @@
  * Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
  *                         University Research and Technology
  *                         Corporation.  All rights reserved.
- * Copyright (c) 2004-2016 The University of Tennessee and The University
+ * Copyright (c) 2004-2018 The University of Tennessee and The University
  *                         of Tennessee Research Foundation.  All rights
  *                         reserved.
  * Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,

ompi/mca/pml/ob1/pml_ob1_comm.c

@@ -2,7 +2,7 @@
  * Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
  *                         University Research and Technology
  *                         Corporation.  All rights reserved.
- * Copyright (c) 2004-2006 The University of Tennessee and The University
+ * Copyright (c) 2004-2018 The University of Tennessee and The University
  *                         of Tennessee Research Foundation.  All rights
  *                         reserved.
  * Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
@@ -29,15 +29,15 @@ static void mca_pml_ob1_comm_proc_construct(mca_pml_ob1_comm_proc_t* proc)
     proc->ompi_proc = NULL;
     proc->expected_sequence = 1;
     proc->send_sequence = 0;
-    OBJ_CONSTRUCT(&proc->frags_cant_match, opal_list_t);
+    proc->frags_cant_match = NULL;
     OBJ_CONSTRUCT(&proc->specific_receives, opal_list_t);
     OBJ_CONSTRUCT(&proc->unexpected_frags, opal_list_t);
 }
 
 
 static void mca_pml_ob1_comm_proc_destruct(mca_pml_ob1_comm_proc_t* proc)
 {
-    OBJ_DESTRUCT(&proc->frags_cant_match);
+    assert(NULL == proc->frags_cant_match);
     OBJ_DESTRUCT(&proc->specific_receives);
     OBJ_DESTRUCT(&proc->unexpected_frags);
     if (proc->ompi_proc) {

ompi/mca/pml/ob1/pml_ob1_comm.h

@@ -3,7 +3,7 @@
  * Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
  *                         University Research and Technology
  *                         Corporation.  All rights reserved.
- * Copyright (c) 2004-2016 The University of Tennessee and The University
+ * Copyright (c) 2004-2018 The University of Tennessee and The University
  *                         of Tennessee Research Foundation.  All rights
  *                         reserved.
  * Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
@@ -40,7 +40,7 @@ struct mca_pml_ob1_comm_proc_t {
 #else
     int32_t send_sequence; /**< send side sequence number */
 #endif
-    opal_list_t frags_cant_match;  /**< out-of-order fragment queues */
+    struct mca_pml_ob1_recv_frag_t* frags_cant_match;  /**< out-of-order fragment queues */
     opal_list_t specific_receives; /**< queues of unmatched specific receives */
     opal_list_t unexpected_frags;  /**< unexpected fragment queues */
 };