Add a new test that checks behavior when we call exit with a non-zero return code after calling finalize - don't ask why.

Modify the check_complete code so it finds the first non-zero exit status (i.e., the one from the lowest rank) in a job that terminates normally, and sets the mpirun exit code to that status.

This commit was SVN r23071.

orte/mca/errmgr/hnp/errmgr_hnp.c

@@ -580,6 +580,7 @@ static void check_job_complete(orte_job_t *jdata)
     orte_job_map_t *map;
     orte_std_cntr_t index;
     bool one_still_alive;
+    orte_exit_code_t first_non_zero=0;
     
 #if 0
     /* Check if FileM is active. If so then keep processing. */
@@ -594,6 +595,10 @@ static void check_job_complete(orte_job_t *jdata)
             continue;
         }
         
+        if (0 == first_non_zero && 0 != proc->exit_code) {
+            first_non_zero = proc->exit_code;
+        }
+        
         /*
          * Determine how the process state affects the job state
          */
@@ -727,6 +732,8 @@ static void check_job_complete(orte_job_t *jdata)
         /* turn off any sensor monitors on this job */
         orte_sensor.stop(jdata->jobid);
 #endif
+        /* update our exit code */
+        ORTE_UPDATE_EXIT_STATUS(first_non_zero);
         OPAL_OUTPUT_VERBOSE((5, orte_errmgr_base.output,
                              "%s errmgr:hnp:check_job_completed declared job %s normally terminated - checking all jobs",
                              ORTE_NAME_PRINT(ORTE_PROC_MY_NAME),

orte/test/system/Makefile

@@ -1,4 +1,4 @@
-PROGS = no_op sigusr_trap spin orte_nodename orte_spawn orte_loop_spawn orte_loop_child orte_abort get_limits orte_ring spawn_child orte_tool orte_no_op binom oob_stress iof_stress iof_delay radix orte_barrier orte_mcast opal_interface mcast mcast_recv orte_spin segfault sysinfo
+PROGS = no_op sigusr_trap spin orte_nodename orte_spawn orte_loop_spawn orte_loop_child orte_abort get_limits orte_ring spawn_child orte_tool orte_no_op binom oob_stress iof_stress iof_delay radix orte_barrier orte_mcast opal_interface mcast mcast_recv orte_spin segfault sysinfo orte_exit
 
 all: $(PROGS)
 

orte/test/system/orte_exit.c

@@ -0,0 +1,47 @@
+/* -*- C -*-
+ *
+ * $HEADER$
+ *
+ * A program that just spins, with vpid 3 aborting - provides mechanism for testing
+ * abnormal program termination
+ */
+
+#include <stdio.h>
+#include <unistd.h>
+
+#include "orte/runtime/runtime.h"
+#include "orte/util/proc_info.h"
+#include "orte/util/name_fns.h"
+#include "orte/runtime/orte_globals.h"
+#include "orte/mca/errmgr/errmgr.h"
+
+int main(int argc, char* argv[])
+{
+    
+    int i, rc;
+    double pi;
+    pid_t pid;
+    char hostname[500];
+    
+    if (0 > (rc = orte_init(&argc, &argv, ORTE_PROC_NON_MPI))) {
+        fprintf(stderr, "orte_abort: couldn't init orte - error code %d\n", rc);
+        return rc;
+    }
+    pid = getpid();
+    gethostname(hostname, 500);
+    
+    printf("orte_abort: Name %s Host: %s Pid %ld\n", ORTE_NAME_PRINT(ORTE_PROC_MY_NAME),
+           hostname, (long)pid);
+    
+    i = 0;
+    while (1) {
+        i++;
+        pi = i / 3.14159256;
+        if (i == 9995) {
+            orte_finalize();
+            exit(ORTE_PROC_MY_NAME->vpid);
+        }
+    }
+    
+    return 0;
+}