Ensure that the orteds do not directly respond to USR1/2 signals. Those signals are trapped by mpirun and propagated from there - at most, the orteds are involved in the propagation process, but should never do anything on their own.

This commit was SVN r16098.

orte/mca/odls/default/odls_default_module.c

@@ -1613,6 +1613,11 @@ static int send_signal(pid_t pid, int signal)
 {
     int rc = ORTE_SUCCESS;
     
+    OPAL_OUTPUT_VERBOSE((1, orte_odls_globals.output,
+                         "%s sending signal %d to pid %ld",
+                         ORTE_NAME_PRINT(ORTE_PROC_MY_NAME),
+                         signal, (long)pid));
+    
     if (kill(pid, signal) != 0) {
         switch(errno) {
             case EINVAL:
@@ -1641,6 +1646,11 @@ int orte_odls_default_signal_local_procs(const orte_process_name_t *proc, int32_
     opal_list_item_t *item;
     orte_odls_child_t *child;
     
+    OPAL_OUTPUT_VERBOSE((1, orte_odls_globals.output,
+                         "%s signaling proc %s",
+                         ORTE_NAME_PRINT(ORTE_PROC_MY_NAME),
+                         (NULL == proc) ? "NULL" : ORTE_NAME_PRINT(proc)));
+
     /* protect operations involving the global list of children */
     OPAL_THREAD_LOCK(&orte_odls_default.mutex);
 

orte/orted/orted_main.c

@@ -91,6 +91,10 @@
 static struct opal_event term_handler;
 static struct opal_event int_handler;
 static struct opal_event pipe_handler;
+#ifndef __WINDOWS__
+static struct opal_event sigusr1_handler;
+static struct opal_event sigusr2_handler;
+#endif  /* __WINDOWS__ */
 
 static void shutdown_callback(int fd, short flags, void *arg);
 
@@ -200,6 +204,8 @@ opal_cmd_line_init_t orte_cmd_line_opts[] = {
       NULL, OPAL_CMD_LINE_TYPE_NULL, NULL }
 };
 
+static void signal_callback(int fd, short event, void *arg);
+
 int orte_daemon(int argc, char *argv[])
 {
     int ret = 0;
@@ -446,6 +452,16 @@ int orte_daemon(int argc, char *argv[])
                    shutdown_callback, NULL);
     opal_event_add(&int_handler, NULL);
 
+#ifndef __WINDOWS__
+    /** setup callbacks for signals we should ignore */
+    opal_signal_set(&sigusr1_handler, SIGUSR1,
+                    signal_callback, &sigusr1_handler);
+    opal_signal_add(&sigusr1_handler, NULL);
+    opal_signal_set(&sigusr2_handler, SIGUSR2,
+                    signal_callback, &sigusr2_handler);
+    opal_signal_add(&sigusr2_handler, NULL);
+#endif  /* __WINDOWS__ */
+
     /* if requested, report my uri to the indicated pipe */
     if (orted_globals.uri_pipe > 0) {
         write(orted_globals.uri_pipe, orte_universe_info.seed_uri,
@@ -601,3 +617,7 @@ static void shutdown_callback(int fd, short flags, void *arg)
     opal_condition_signal(&orted_comm_cond);
 }
 
+static void signal_callback(int fd, short event, void *arg)
+{
+    /* just ignore these signals */
+}