btl/openib: fix locking bugs with XRC ib_addr lock

This bug fixes two issue with the ib_addr lock:

 - The ib_addr lock must always be obtained regardless of
   opal_using_threads() as the CPC is run in a seperate thread.

 - The ib_addr lock is held in mca_btl_openib_endpoint_connected when
   calling back into the CPC start_connect on any pending
   connections. This will attempt to obtain the ib_addr lock
   again. Since this is not a performance-critical part of the code
   the lock has been changed to be recursive.

Signed-off-by: Nathan Hjelm <[email protected]>

(cherry picked from commit open-mpi/ompi@371df45)

Signed-off-by: Nathan Hjelm <[email protected]>

Author: hjelmn

opal/mca/btl/openib/btl_openib_endpoint.c

@@ -11,7 +11,7 @@
  * Copyright (c) 2004-2005 The Regents of the University of California.
  *                         All rights reserved.
  * Copyright (c) 2006-2013 Cisco Systems, Inc.  All rights reserved.
- * Copyright (c) 2006-2015 Los Alamos National Security, LLC.  All rights
+ * Copyright (c) 2006-2016 Los Alamos National Security, LLC.  All rights
  *                         reserved.
  * Copyright (c) 2006-2007 Voltaire All rights reserved.
  * Copyright (c) 2006-2009 Mellanox Technologies, Inc.  All rights reserved.
@@ -579,7 +579,7 @@ void mca_btl_openib_endpoint_connected(mca_btl_openib_endpoint_t *endpoint)
 
     opal_output(-1, "Now we are CONNECTED");
     if (MCA_BTL_XRC_ENABLED) {
-        OPAL_THREAD_LOCK(&endpoint->ib_addr->addr_lock);
+        opal_mutex_lock (&endpoint->ib_addr->addr_lock);
         if (MCA_BTL_IB_ADDR_CONNECTED == endpoint->ib_addr->status) {
             /* We are not xrc master */
             /* set our qp pointer to master qp */
@@ -622,7 +622,7 @@ void mca_btl_openib_endpoint_connected(mca_btl_openib_endpoint_t *endpoint)
                 }
             }
         }
-        OPAL_THREAD_UNLOCK(&endpoint->ib_addr->addr_lock);
+        opal_mutex_unlock (&endpoint->ib_addr->addr_lock);
     }
 
 

opal/mca/btl/openib/btl_openib_xrc.c

@@ -1,10 +1,13 @@
+/* -*- Mode: C; c-basic-offset:4 ; indent-tabs-mode:nil -*- */
 /*
  * Copyright (c) 2007-2008 Mellanox Technologies. All rights reserved.
  * Copyright (c) 2009      Cisco Systems, Inc.  All rights reserved.
  * Copyright (c) 2014      NVIDIA Corporation.  All rights reserved.
  * Copyright (c) 2014-2015 Research Organization for Information Science
  *                         and Technology (RIST). All rights reserved.
  * Copyright (c) 2014      Bull SAS.  All rights reserved.
+ * Copyright (c) 2016      Los Alamos National Security, LLC. All rights
+ *                         reserved.
  * $COPYRIGHT$
  *
  * Additional copyrights may follow
@@ -122,7 +125,10 @@ static void ib_address_constructor(ib_address_t *ib_addr)
     ib_addr->lid = 0;
     ib_addr->status = MCA_BTL_IB_ADDR_CLOSED;
     ib_addr->qp = NULL;
-    OBJ_CONSTRUCT(&ib_addr->addr_lock, opal_mutex_t);
+    /* NTH: make the addr_lock recursive because mca_btl_openib_endpoint_connected can call
+     * into the CPC with the lock held. The alternative would be to drop the lock but the
+     * lock is never obtained in a critical path. */
+    OBJ_CONSTRUCT(&ib_addr->addr_lock, opal_recursive_mutex_t);
     OBJ_CONSTRUCT(&ib_addr->pending_ep, opal_list_t);
 }