oob ud: better error msgs, tolerate systems without UD devices

It is perfectly ok to be on a system without UD devices.

Also, make some of the error messages better -- so that the user has a
clue about where the error messages are coming from, and what they
should do.

orte/mca/oob/ud/help-oob-ud.txt

@@ -12,32 +12,34 @@
 #                         All rights reserved.
 #               2015      Mellanox Technologies, Inc.
 #                         All rights reserved.
+# Copyright (c) 2015 Cisco Systems, Inc.  All rights reserved.
 # $COPYRIGHT$
 #
 # Additional copyrights may follow
 #
 # $HEADER$
 #
-[no-devices-available]
+[no-devices-error]
-No available RDMA devices found:
+Open MPI has detected a failure in a basic verbs function call.  This
+is unusual, and may indicate that something is malfunctioning on this
+system.
 
-Hostname: %s
+You job will continue, but Open MPI will ignore the "ud" oob component
+in this run.
+
+Verbs function: ibv_get_device_list()
+Error:          %s
+Hostname:       %s
 
 Please contact your system administrator.
 #
-[no-devices-error]
-Failed to get list of the available RDMA devices:
-
-Hostname: %s
-Error:    %s
-#
-[no-devices-usable]
-No usable devices found:
-
-Hostname: %s
-#
 [no-ports-usable]
-No usable ports found:
+Open MPI has detected that there are UD-capable Verbs devices on your
+system, but none of them were able to be setup properly.  This may
+indicate a problem on this system.
+
+You job will continue, but Open MPI will ignore the "ud" oob component
+in this run.
 
 Hostname: %s
 #

orte/mca/oob/ud/oob_ud_component.c

@@ -344,14 +344,16 @@ static int mca_oob_ud_component_startup(void)
     devices = ibv_get_device_list (&num_devices);
     if (NULL == devices) {
         orte_show_help("help-oob-ud.txt", "no-devices-error", true,
-                       orte_process_info.nodename, strerror(errno));
+                       strerror(errno),
+                       orte_process_info.nodename);
         return ORTE_ERROR;
     }
 
+    /* If there are no devices, it is not an error; we just won't use
+       this component. */
     if (0 == num_devices) {
-        orte_show_help("help-oob-ud.txt", "no-devices-available", true,
+        ibv_free_device_list(devices);
-                       orte_process_info.nodename);
+        return ORTE_ERR_NOT_FOUND;
-        return ORTE_ERROR;
     }
 
     for (i = 0 ; i < num_devices ; ++i) {
@@ -377,10 +379,10 @@ static int mca_oob_ud_component_startup(void)
 
     ibv_free_device_list (devices);
 
+    /* If no usable devices are found, then just ignore this component
+       in this run */
     if (0 == opal_list_get_size (&mca_oob_ud_component.ud_devices)) {
-        orte_show_help("help-oob-ud.txt", "no-devices-usable", true,
+        return ORTE_ERR_NOT_FOUND;
-                       orte_process_info.nodename);
-        return ORTE_ERROR;
     }
 
     opal_output_verbose(5, orte_oob_base_framework.framework_output,