The ib_procs list in the openib btl is accessed without the ib lock in some cases. This causes races when running multithreaded. This patch adds protection of the ib_procs list with the ib_lock.

fixes trac:2149 cmr:v1.4

This commit was SVN r22682.

The following Trac tickets were found above:
  Ticket 2149 --> https://svn.open-mpi.org/trac/ompi/ticket/2149

ompi/mca/btl/openib/connect/btl_openib_connect_ibcm.c

@@ -1,6 +1,7 @@
 /*
  * Copyright (c) 2007-2009 Cisco Systems, Inc.  All rights reserved.
  * Copyright (c) 2008-2009 Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2009      IBM Corporation.  All rights reserved.
  *
  * $COPYRIGHT$
  * 
@@ -1816,6 +1817,7 @@ static int request_received(ibcm_listen_cm_id_t *cmh,
     /* JMS: optimization target -- can we send something in private
        data to find the proc directly instead of having to search
        through *all* procs? */
+    OPAL_THREAD_LOCK(&mca_btl_openib_component.ib_lock);
     for (found = false, ib_proc = (mca_btl_openib_proc_t*)
              opal_list_get_first(&mca_btl_openib_component.ib_procs);
          !found && 
@@ -1850,6 +1852,7 @@ static int request_received(ibcm_listen_cm_id_t *cmh,
             }
         }
     }
+    OPAL_THREAD_UNLOCK(&mca_btl_openib_component.ib_lock);
     if (!found) {
         BTL_VERBOSE(("could not find match for calling endpoint!"));
         rc = OMPI_ERR_NOT_FOUND;

ompi/mca/btl/openib/connect/btl_openib_connect_oob.c

@@ -13,6 +13,7 @@
  * Copyright (c) 2006      Los Alamos National Security, LLC.  All rights
  *                         reserved. 
  * Copyright (c) 2008-2009 Mellanox Technologies.  All rights reserved.
+ * Copyright (c) 2009      IBM Corporation.  All rights reserved.
  *
  * $COPYRIGHT$
  * 
@@ -728,6 +729,9 @@ static void rml_recv_cb(int status, orte_process_name_t* process_name,
     master = orte_util_compare_name_fields(ORTE_NS_CMP_ALL, ORTE_PROC_MY_NAME,
                                     process_name) > 0 ? true : false;
     
+    /* Need to protect the ib_procs list */
+    OPAL_THREAD_LOCK(&mca_btl_openib_component.ib_lock);
+
     for (ib_proc = (mca_btl_openib_proc_t*)
             opal_list_get_first(&mca_btl_openib_component.ib_procs);
         ib_proc != (mca_btl_openib_proc_t*)
@@ -780,6 +784,7 @@ static void rml_recv_cb(int status, orte_process_name_t* process_name,
                just ignore this connection request */
             if (found && !master &&
                 MCA_BTL_IB_CLOSED != ib_endpoint->endpoint_state) {
+		OPAL_THREAD_UNLOCK(&mca_btl_openib_component.ib_lock);
                 return;
             }
         }
@@ -787,6 +792,7 @@ static void rml_recv_cb(int status, orte_process_name_t* process_name,
         if (!found) {
             BTL_ERROR(("can't find suitable endpoint for this peer\n")); 
             mca_btl_openib_endpoint_invoke_error(NULL);
+	    OPAL_THREAD_UNLOCK(&mca_btl_openib_component.ib_lock);
             return; 
         }
         
@@ -869,4 +875,5 @@ static void rml_recv_cb(int status, orte_process_name_t* process_name,
         }
         break;
     }
+    OPAL_THREAD_UNLOCK(&mca_btl_openib_component.ib_lock);
 }

ompi/mca/btl/openib/connect/btl_openib_connect_xoob.c

@@ -1,6 +1,7 @@
 /*
  * Copyright (c) 2007-2009 Mellanox Technologies.  All rights reserved.
  * Copyright (c) 2009      Cisco Systems, Inc.  All rights reserved.
+ * Copyright (c) 2009      IBM Corporation.  All rights reserved.
  *
  * $COPYRIGHT$
  *
@@ -658,6 +659,7 @@ static mca_btl_openib_endpoint_t* xoob_find_endpoint(orte_process_name_t* proces
                 "jobid %d, vpid %d, sid %" PRIx64 ", lid %d",
                 process_name->jobid, process_name->vpid, subnet_id, lid));
     /* find ibproc */
+    OPAL_THREAD_LOCK(&mca_btl_openib_component.ib_lock);
     for (ib_proc = (mca_btl_openib_proc_t*)
             opal_list_get_first(&mca_btl_openib_component.ib_procs);
             ib_proc != (mca_btl_openib_proc_t*)
@@ -696,6 +698,7 @@ static mca_btl_openib_endpoint_t* xoob_find_endpoint(orte_process_name_t* proces
     } else {
             BTL_ERROR(("can't find suitable endpoint for this peer\n"));
     }
+    OPAL_THREAD_UNLOCK(&mca_btl_openib_component.ib_lock);
     return ib_endpoint;
 }