orte: Expand the application of !orte_keep_fqdn_hostnames

jjhursey · jjhursey · commit d26dd2c20e19 · 2016-08-26T16:09:04.000-05:00
* Expand the use of the `orte_keep_fqdn_hostnames` MCA parameter when
   it is set to false.
 * If that parameter is set to false (default) then short hostnames
   (e.g., `node01`) will match with the long hostnames (e.g.,
   `node01.mycluster.org`). This allows a user (or resource manager)
    to mix the use of short and long hostnames.
  - Note that this mechanism does _not_ perform a DNS lookup, but
    instead strips off the FQDN by truncating the hostname string at
    the first `.` character (when not an IP address).
     - By default (`false`) the following is true:
       `node01 == node01.mycluster.org == node01.bogus.com`
       since we use `node01` as the hostname.
diff --git a/orte/mca/plm/base/plm_base_launch_support.c b/orte/mca/plm/base/plm_base_launch_support.c
@@ -16,6 +16,7 @@
  * Copyright (c) 2013-2016 Intel, Inc. All rights reserved.
  * Copyright (c) 2014-2015 Research Organization for Information Science
  *                         and Technology (RIST). All rights reserved.
+ * Copyright (c) 2016      IBM Corporation.  All rights reserved.
  * $COPYRIGHT$
  *
  * Additional copyrights may follow
@@ -1528,7 +1529,7 @@ int orte_plm_base_setup_virtual_machine(orte_job_t *jdata)
     bool one_filter = false;
     int num_nodes;
     bool default_hostfile_used;
-    char *hosts;
+    char *hosts = NULL;
     bool singleton=false;
     bool multi_sim = false;
 
diff --git a/orte/mca/rmaps/base/rmaps_base_support_fns.c b/orte/mca/rmaps/base/rmaps_base_support_fns.c
@@ -13,6 +13,7 @@
  * Copyright (c) 2011-2012 Los Alamos National Security, LLC.
  *                         All rights reserved.
  * Copyright (c) 2014-2016 Intel, Inc. All rights reserved.
+ * Copyright (c) 2016      IBM Corporation.  All rights reserved.
  * $COPYRIGHT$
  *
  * Additional copyrights may follow
@@ -147,7 +148,7 @@ int orte_rmaps_base_get_target_nodes(opal_list_t *allocated_nodes, orte_std_cntr
     orte_job_t *daemons;
     bool novm;
     opal_list_t nodes;
-    char *hosts;
+    char *hosts = NULL;
 
     /** set default answer */
     *total_num_slots = 0;
diff --git a/orte/mca/rmaps/rank_file/rmaps_rank_file.c b/orte/mca/rmaps/rank_file/rmaps_rank_file.c
@@ -17,6 +17,7 @@
  * Copyright (c) 2014-2016 Intel, Inc. All rights reserved.
  * Copyright (c) 2015      Research Organization for Information Science
  *                         and Technology (RIST). All rights reserved.
+ * Copyright (c) 2016      IBM Corporation.  All rights reserved.
  *
  * $COPYRIGHT$
  *
@@ -33,6 +34,9 @@
 #ifdef HAVE_UNISTD_H
 #include <unistd.h>
 #endif  /* HAVE_UNISTD_H */
+#if HAVE_ARPA_INET_H
+#include <arpa/inet.h>
+#endif
 #include <string.h>
 
 #include "opal/util/argv.h"
@@ -488,6 +492,20 @@ static int orte_rmaps_rank_file_parse(const char *rankfile)
                             goto unlock;
                         }
                         opal_argv_free (argv);
+
+                        // Strip off the FQDN if present
+                        if( !orte_keep_fqdn_hostnames ) {
+                            char *ptr;
+                            struct in_addr buf;
+                            /* if the nodename is an IP address, do not mess with it! */
+                            if (0 == inet_pton(AF_INET, node_name, &buf) &&
+                                0 == inet_pton(AF_INET6, node_name, &buf)) {
+                                if (NULL != (ptr = strchr(node_name, '.'))) {
+                                    *ptr = '\0';
+                                }
+                            }
+                        }
+
                         /* check the rank item */
                         if (NULL == rfmap) {
                             orte_show_help("help-rmaps_rank_file.txt", "bad-syntax", true, rankfile);
diff --git a/orte/mca/rmaps/seq/rmaps_seq.c b/orte/mca/rmaps/seq/rmaps_seq.c
@@ -15,6 +15,7 @@
  * Copyright (c) 2014-2016 Intel, Inc. All rights reserved.
  * Copyright (c) 2015      Research Organization for Information Science
  *                         and Technology (RIST). All rights reserved.
+ * Copyright (c) 2016      IBM Corporation.  All rights reserved.
  * $COPYRIGHT$
  *
  * Additional copyrights may follow
@@ -70,9 +71,11 @@ static void sn_des(seq_node_t *p)
 {
     if (NULL != p->hostname) {
         free(p->hostname);
+        p->hostname = NULL;
     }
     if (NULL != p->cpuset) {
         free(p->cpuset);
+        p->cpuset = NULL;
     }
 }
 OBJ_CLASS_INSTANCE(seq_node_t,
@@ -101,7 +104,7 @@ static int orte_rmaps_seq_map(orte_job_t *jdata)
     opal_list_t node_list, *seq_list, sq_list;
     orte_proc_t *proc;
     mca_base_component_t *c = &mca_rmaps_seq_component.base_version;
-    char *hosts, *sep, *eptr;
+    char *hosts = NULL, *sep, *eptr;
     FILE *fp;
     opal_hwloc_resource_type_t rtype;
 
@@ -156,7 +159,7 @@ static int orte_rmaps_seq_map(orte_job_t *jdata)
     /* if there is a default hostfile, go and get its ordered list of nodes */
     OBJ_CONSTRUCT(&default_seq_list, opal_list_t);
     if (NULL != orte_default_hostfile) {
-        char *hstname;
+        char *hstname = NULL;
         /* open the file */
         fp = fopen(orte_default_hostfile, "r");
         if (NULL == fp) {
@@ -170,6 +173,11 @@ static int orte_rmaps_seq_map(orte_job_t *jdata)
                 /* blank line - ignore */
                 continue;
             }
+            if( '#' == hstname[0] ) {
+                free(hstname);
+                /* Comment line - ignore */
+                continue;
+            }
             sq = OBJ_NEW(seq_node_t);
             if (NULL != (sep = strchr(hstname, ' '))) {
                 *sep = '\0';
@@ -182,6 +190,21 @@ static int orte_rmaps_seq_map(orte_job_t *jdata)
                 *(eptr+1) = 0;
                 sq->cpuset = strdup(sep);
             }
+
+            // Strip off the FQDN if present
+            if( !orte_keep_fqdn_hostnames ) {
+                char *ptr;
+                struct in_addr buf;
+
+                /* if the nodename is an IP address, do not mess with it! */
+                if (0 == inet_pton(AF_INET, hstname, &buf) &&
+                    0 == inet_pton(AF_INET6, hstname, &buf)) {
+                    if (NULL != (ptr = strchr(hstname, '.'))) {
+                        *ptr = '\0';
+                    }
+                }
+            }
+
             sq->hostname = hstname;
             opal_list_append(&default_seq_list, &sq->super);
         }
@@ -255,6 +278,16 @@ static int orte_rmaps_seq_map(orte_job_t *jdata)
                 goto error;
             }
             while (NULL != (hstname = orte_getline(fp))) {
+                if (0 == strlen(hstname)) {
+                    free(hstname);
+                    /* blank line - ignore */
+                    continue;
+                }
+                if( '#' == hstname[0] ) {
+                    free(hstname);
+                    /* Comment line - ignore */
+                    continue;
+                }
                 sq = OBJ_NEW(seq_node_t);
                 if (NULL != (sep = strchr(hstname, ' '))) {
                     *sep = '\0';
@@ -267,6 +300,20 @@ static int orte_rmaps_seq_map(orte_job_t *jdata)
                     *(eptr+1) = 0;
                     sq->cpuset = strdup(sep);
                 }
+
+                // Strip off the FQDN if present
+                if( !orte_keep_fqdn_hostnames ) {
+                    char *ptr;
+                    struct in_addr buf;
+                    /* if the nodename is an IP address, do not mess with it! */
+                    if (0 == inet_pton(AF_INET, hstname, &buf) &&
+                        0 == inet_pton(AF_INET6, hstname, &buf)) {
+                        if (NULL != (ptr = strchr(hstname, '.'))) {
+                            (*ptr) = '\0';
+                        }
+                    }
+                }
+
                 sq->hostname = hstname;
                 opal_list_append(&sq_list, &sq->super);
             }
diff --git a/orte/util/dash_host/dash_host.c b/orte/util/dash_host/dash_host.c
@@ -13,6 +13,7 @@
  * Copyright (c) 2014-2016 Intel, Inc.  All rights reserved.
  * Copyright (c) 2015      Research Organization for Information Science
  *                         and Technology (RIST). All rights reserved.
+ * Copyright (c) 2016      IBM Corporation.  All rights reserved.
  * $COPYRIGHT$
  *
  * Additional copyrights may follow
@@ -23,6 +24,9 @@
 #include "orte_config.h"
 
 #include <string.h>
+#if HAVE_ARPA_INET_H
+#include <arpa/inet.h>
+#endif
 
 #include "orte/constants.h"
 #include "orte/types.h"
@@ -207,6 +211,19 @@ int orte_util_add_dash_host_nodes(opal_list_t *nodes,
             ndname = mini_map[i];
         }
 
+        // Strip off the FQDN if present
+        if( !orte_keep_fqdn_hostnames ) {
+            char *ptr;
+            struct in_addr buf;
+            /* if the nodename is an IP address, do not mess with it! */
+            if (0 == inet_pton(AF_INET, ndname, &buf) &&
+                0 == inet_pton(AF_INET6, ndname, &buf)) {
+                if (NULL != (ptr = strchr(ndname, '.'))) {
+                    *ptr = '\0';
+                }
+            }
+        }
+
         /* see if the node is already on the list */
         found = false;
         OPAL_LIST_FOREACH(node, &adds, orte_node_t) {
diff --git a/orte/util/hostfile/hostfile.c b/orte/util/hostfile/hostfile.c
@@ -15,6 +15,7 @@
  * Copyright (c) 2013-2014 Intel, Inc. All rights reserved.
  * Copyright (c) 2015      Research Organization for Information Science
  *                         and Technology (RIST). All rights reserved.
+ * Copyright (c) 2016      IBM Corporation.  All rights reserved.
  * $COPYRIGHT$
  *
  * Additional copyrights may follow
@@ -27,6 +28,9 @@
 #ifdef HAVE_UNISTD_H
 #include <unistd.h>
 #endif
+#if HAVE_ARPA_INET_H
+#include <arpa/inet.h>
+#endif
 #include <errno.h>
 #include <string.h>
 #include <sys/stat.h>
@@ -164,6 +168,19 @@ static int hostfile_parse_line(int token, opal_list_t* updates,
         }
         opal_argv_free (argv);
 
+        // Strip off the FQDN if present
+        if( !orte_keep_fqdn_hostnames ) {
+            char *ptr;
+            struct in_addr buf;
+            /* if the nodename is an IP address, do not mess with it! */
+            if (0 == inet_pton(AF_INET, node_name, &buf) &&
+                0 == inet_pton(AF_INET6, node_name, &buf)) {
+                if (NULL != (ptr = strchr(node_name, '.'))) {
+                    *ptr = '\0';
+                }
+            }
+        }
+
         /* if the first letter of the name is '^', then this is a node
          * to be excluded. Remove the ^ character so the nodename is
          * usable, and put it on the exclude list
@@ -270,6 +287,20 @@ static int hostfile_parse_line(int token, opal_list_t* updates,
             opal_output(0, "WARNING: Unhandled user@host-combination\n"); /* XXX */
         }
         opal_argv_free (argv);
+
+        // Strip off the FQDN if present
+        if( !orte_keep_fqdn_hostnames ) {
+            char *ptr;
+            struct in_addr buf;
+            /* if the nodename is an IP address, do not mess with it! */
+            if (0 == inet_pton(AF_INET, node_name, &buf) &&
+                0 == inet_pton(AF_INET6, node_name, &buf)) {
+                if (NULL != (ptr = strchr(node_name, '.'))) {
+                    *ptr = '\0';
+                }
+            }
+        }
+
         /* Do we need to make a new node object? */
         if (NULL == (node = hostfile_lookup(updates, node_name))) {
             node = OBJ_NEW(orte_node_t);
