Fix rank_file operations. We changed the syntax to use semi-colons between multiple slot assignments so that we could use the comma to separate specific cores, but somehow the flex definitions didn't get updated to accept that character. We also incorrectly zero'd the bitmap between slot assignment sections, and so multiple slot assignments only wound up making the last one in the list.

This commit was SVN r27908.
2013-01-25 18:33:25 +00:00 · 2013-01-25 18:33:25 +00:00 · f6b4db0b79
--- a/opal/mca/hwloc/base/hwloc_base_util.c
+++ b/opal/mca/hwloc/base/hwloc_base_util.c
@ -10,7 +10,7 @@
 * Copyright (c) 2004-2005 The Regents of the University of California.
 *                         All rights reserved.
 * Copyright (c) 2011-2012 Cisco Systems, Inc.  All rights reserved.
- * Copyright (c) 2012      Los Alamos National Security, LLC.
+ * Copyright (c) 2012-2013 Los Alamos National Security, LLC.
 *                         All rights reserved.
 * $COPYRIGHT$
 * 
@ -941,7 +941,7 @@ static int socket_to_cpu_set(char *cpus,
    int lower_range, upper_range;
    int socket_id;
    hwloc_obj_t obj;
-    hwloc_bitmap_t avail, res;
+    hwloc_bitmap_t res;

    if ('*' == cpus[0]) {
        /* requesting cpumask for ALL sockets */
@ -950,7 +950,7 @@ static int socket_to_cpu_set(char *cpus,
         * this specification equates to unbound
         */
        res = opal_hwloc_base_get_available_cpus(topo, obj);
-        hwloc_bitmap_copy(cpumask, res);
+        hwloc_bitmap_or(cpumask, cpumask, res);
        return OPAL_SUCCESS;
    }

@ -962,25 +962,20 @@ static int socket_to_cpu_set(char *cpus,
        obj = opal_hwloc_base_get_obj_by_type(topo, HWLOC_OBJ_SOCKET, 0, socket_id, OPAL_HWLOC_LOGICAL);
        /* get the available logical cpus for this socket */
        res = opal_hwloc_base_get_available_cpus(topo, obj);
-        hwloc_bitmap_copy(cpumask, res);
+        hwloc_bitmap_or(cpumask, cpumask, res);
        break;
                
    case 2:  /* range of sockets was given */
        lower_range = atoi(range[0]);
        upper_range = atoi(range[1]);
-        /* zero the bitmask */
-        hwloc_bitmap_zero(cpumask);
-        avail = hwloc_bitmap_alloc();
        /* cycle across the range of sockets */
        for (socket_id=lower_range; socket_id<=upper_range; socket_id++) {
            obj = opal_hwloc_base_get_obj_by_type(topo, HWLOC_OBJ_SOCKET, 0, socket_id, OPAL_HWLOC_LOGICAL);
            /* get the available logical cpus for this socket */
            res = opal_hwloc_base_get_available_cpus(topo, obj);
            /* set the corresponding bits in the bitmask */
-            hwloc_bitmap_or(avail, cpumask, res);
-            hwloc_bitmap_copy(cpumask, avail);
+            hwloc_bitmap_or(cpumask, cpumask, res);
        }
-        hwloc_bitmap_free(avail);
        break;
    default:
        opal_argv_free(range);
@ -1002,7 +997,7 @@ static int socket_core_to_cpu_set(char *socket_core_list,
    int lower_range, upper_range;
    int socket_id, core_id;
    hwloc_obj_t socket, core;
-    hwloc_cpuset_t res, avail;
+    hwloc_cpuset_t res;
    unsigned int idx;
    hwloc_obj_type_t obj_type = HWLOC_OBJ_CORE;

@ -1034,7 +1029,7 @@ static int socket_core_to_cpu_set(char *socket_core_list,
        if ('*' == corestr[0]) {
            /* set to all available logical cpus on this socket */
            res = opal_hwloc_base_get_available_cpus(topo, socket);
-            hwloc_bitmap_copy(cpumask, res);
+            hwloc_bitmap_or(cpumask, cpumask, res);
            /* we are done - already assigned all cores! */
            rc = OPAL_SUCCESS;
            break;
@ -1054,14 +1049,15 @@ static int socket_core_to_cpu_set(char *socket_core_list,
                }
                /* get the cpus */
                res = opal_hwloc_base_get_available_cpus(topo, core);
-                hwloc_bitmap_copy(cpumask, res);
+                hwloc_bitmap_or(cpumask, cpumask, res);
                break;
                
            case 2:  /* range of core id's was given */
+                opal_output_verbose(5, opal_hwloc_base_output,
+                                    "range of cores given: start %s stop %s",
+                                    range[0], range[1]);
                lower_range = atoi(range[0]);
                upper_range = atoi(range[1]);
-                hwloc_bitmap_zero(cpumask);
-                avail = hwloc_bitmap_alloc();
                for (core_id=lower_range; core_id <= upper_range; core_id++) {
                    /* get that object */
                    idx = 0;
@ -1073,10 +1069,8 @@ static int socket_core_to_cpu_set(char *socket_core_list,
                    /* get the cpus */
                    res = opal_hwloc_base_get_available_cpus(topo, core);
                    /* add them into the result */
-                    hwloc_bitmap_or(avail, cpumask, res);
-                    hwloc_bitmap_copy(cpumask, avail);
+                    hwloc_bitmap_or(cpumask, cpumask, res);
                }
-                hwloc_bitmap_free(avail);
                break;
                
            default:
@ -1123,6 +1117,9 @@ int opal_hwloc_base_slot_list_parse(const char *slot_str,
    hwloc_bitmap_zero(cpumask);
    /* loop across the items and accumulate the mask */
    for (i=0; NULL != item[i]; i++) {
+        opal_output_verbose(5, opal_hwloc_base_output,
+                            "working assignment %s",
+                            item[i]);
        /* if they specified "socket" by starting with an S/s,
         * or if they use socket:core notation, then parse the
         * socket/core info
@ -1161,7 +1158,6 @@ int opal_hwloc_base_slot_list_parse(const char *slot_str,
            /* just a core specification - see if one or a range was given */
            range = opal_argv_split(item[i], '-');
            range_cnt = opal_argv_count(range);
-            hwloc_bitmap_zero(cpumask);
            /* see if a range was set or not */
            switch (range_cnt) {
            case 1:  /* only one core specified */
@ -1175,13 +1171,12 @@ int opal_hwloc_base_slot_list_parse(const char *slot_str,
                /* get the available cpus for that object */
                pucpus = opal_hwloc_base_get_available_cpus(topo, pu);
                 /* set that in the mask */
-                hwloc_bitmap_copy(cpumask, pucpus);
+                hwloc_bitmap_or(cpumask, cpumask, pucpus);
                break;
                    
            case 2:  /* range of core id's was given */
                lower_range = atoi(range[0]);
                upper_range = atoi(range[1]);
-                hwloc_bitmap_zero(cpumask);
                for (core_id=lower_range; core_id <= upper_range; core_id++) {
                    /* find the specified logical available cpu */
                    if (NULL == (pu = get_pu(topo, core_id))) {
--- a/orte/mca/rmaps/rank_file/help-rmaps_rank_file.txt
+++ b/orte/mca/rmaps/rank_file/help-rmaps_rank_file.txt
@ -1,6 +1,8 @@
 # Copyright (c) 2004-2005 The Regents of the University of California.
 #                         All rights reserved.
 # Copyright (c) 2011      Cisco Systems, Inc.  All rights reserved.
+# Copyright (c) 2013      Los Alamos National Security, LLC.
+#                         All rights reserved.
 # $COPYRIGHT$
 #
 # Additional copyrights may follow
@ -27,7 +29,8 @@ Check to make sure the path and filename are correct.

 usage:  mpirun -mca rmaps_rankfile_path rankfile ./app

-example:  cat hostfile
+Examples of proper syntax include:
+  cat hostfile
    host1
    host2
    host3
@ -36,21 +39,37 @@ example:  cat hostfile
    rank 1=host1 slot=1:0,1
    rank 0=host2 slot=0:*
    rank 2=host4 slot=1-2
-    rank 3=host3 slot=0:1,1:0-2
+    rank 3=host3 slot=0:1;1:0-2
 #
 [parse_error_string]
 Open RTE detected a parse error in the rankfile (%s)
 It occured on line number %d on token %d:
    %s
+Examples of proper syntax include:
+    rank 1=host1 slot=1:0,1
+    rank 0=host2 slot=0:*
+    rank 2=host4 slot=1-2
+    rank 3=host3 slot=0:1;1:0-2
 #
 [parse_error_int]
 Open RTE detected a parse error in the rankfile (%s)
 It occured on line number %d on token %d:
    %d
+Examples of proper syntax include:
+    rank 1=host1 slot=1:0,1
+    rank 0=host2 slot=0:*
+    rank 2=host4 slot=1-2
+    rank 3=host3 slot=0:1;1:0-2
 #
 [parse_error]
 Open RTE detected a parse error in the rankfile (%s)
-It occured on line number %d on token %d.
+It occured on line number %d on token %d. Examples of
+proper syntax include:
+    rank 1=host1 slot=1:0,1
+    rank 0=host2 slot=0:*
+    rank 2=host4 slot=1-2
+    rank 3=host3 slot=0:1;1:0-2
+
 #
 [not-all-mapped-alloc]
 Some of the requested ranks are not included in the current allocation.
@ -89,7 +108,11 @@ Error, rank %d is already assigned to %s, check %s
 Error, invalid syntax in the rankfile (%s)
 syntax must be the fallowing
 rank i=host_i slot=string
-ex: rank 1=host1 slot=1:0,1
+Examples of proper syntax include:
+    rank 1=host1 slot=1:0,1
+    rank 0=host2 slot=0:*
+    rank 2=host4 slot=1-2
+    rank 3=host3 slot=0:1;1:0-2
 #
 [orte-rmaps-rf:multi-apps-and-zero-np]
 RMAPS found multiple applications to be launched, with
--- a/orte/mca/rmaps/rank_file/rmaps_rank_file.c
+++ b/orte/mca/rmaps/rank_file/rmaps_rank_file.c
@ -10,7 +10,7 @@
 * Copyright (c) 2004-2005 The Regents of the University of California.
 *                         All rights reserved.
 * Copyright (c) 2006-2011 Cisco Systems, Inc.  All rights reserved.
- * Copyright (c) 2011      Los Alamos National Security, LLC.
+ * Copyright (c) 2011-2013 Los Alamos National Security, LLC.
 *                         All rights reserved.
 * Copyright (c) 2008      Voltaire. All rights reserved
 * Copyright (c) 2010      Oracle and/or its affiliates.  All rights reserved.
@ -382,7 +382,10 @@ static int orte_rmaps_rank_file_parse(const char *rankfile)
        
        switch (token) {
            case ORTE_RANKFILE_ERROR:
-                opal_output(0, "Got an error!");
+                orte_show_help("help-rmaps_rank_file.txt", "bad-syntax", true, rankfile);
+                rc = ORTE_ERR_BAD_PARAM;
+                ORTE_ERROR_LOG(rc);
+                goto unlock;
                break;
            case ORTE_RANKFILE_QUOTED_STRING:
                orte_show_help("help-rmaps_rank_file.txt", "not-supported-rankfile", true, "QUOTED_STRING", rankfile);
--- a/orte/mca/rmaps/rank_file/rmaps_rank_file_lex.l
+++ b/orte/mca/rmaps/rank_file/rmaps_rank_file_lex.l
@ -13,6 +13,8 @@
 * Copyright (c) 2004-2005 The Regents of the University of California.
 *                         All rights reserved.
 * Copyright (c) 2008      Voltaire. All rights reserved
+ * Copyright (c) 2013      Los Alamos National Security, LLC.
+ *                         All rights reserved.
 *  
 * $COPYRIGHT$
 * 
@ -96,7 +98,7 @@ username           { orte_rmaps_rank_file_value.sval = yytext;
    */
 %}

-[A-za-z0-9_\-,:*@]*  { orte_rmaps_rank_file_value.sval = yytext;
+[A-za-z0-9_\-,\;:*@]*  { orte_rmaps_rank_file_value.sval = yytext;
                         return ORTE_RANKFILE_STRING; }

 ([A-Za-z0-9][A-Za-z0-9_\-]*"@")?([0-9]{1,3}"."){3}[0-9]{1,3} {