Skip to content

v3.1.x: Allow individual jobs to set map policy, override LSF bindfile #6873

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Aug 20, 2019
Merged

v3.1.x: Allow individual jobs to set map policy, override LSF bindfile #6873

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9047d68
Allow individual jobs to set their map/rank/bind policies
rhc54 Aug 6, 2019
7ebc42c
Fix typos
rhc54 Aug 7, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 15 additions & 2 deletions orte/mca/ras/lsf/ras_lsf_module.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@
* Copyright (c) 2004-2005 The Regents of the University of California.
* All rights reserved.
* Copyright (c) 2007-2017 Cisco Systems, Inc. All rights reserved
* Copyright (c) 2014 Intel, Inc. All rights reserved
* Copyright (c) 2014-2019 Intel, Inc. All rights reserved.
* Copyright (c) 2016 IBM Corporation. All rights reserved.
* $COPYRIGHT$
*
Expand All @@ -36,6 +36,7 @@

#include "orte/mca/rmaps/rmaps_types.h"
#include "orte/mca/errmgr/errmgr.h"
#include "orte/mca/rmaps/base/base.h"
#include "orte/runtime/orte_globals.h"
#include "orte/util/show_help.h"

Expand Down Expand Up @@ -70,6 +71,7 @@ static int allocate(orte_job_t *jdata, opal_list_t *nodes)
char *affinity_file;
struct stat buf;
char *ptr;
bool directives_given = false;

/* get the list of allocated nodes */
if ((num_nodes = lsb_getalloc(&nodelist)) < 0) {
Expand Down Expand Up @@ -112,8 +114,19 @@ static int allocate(orte_job_t *jdata, opal_list_t *nodes)
/* release the nodelist from lsf */
opal_argv_free(nodelist);

/* check to see if any mapping or binding directives were given */
if (NULL != jdata && NULL != jdata->map) {
if ((ORTE_MAPPING_GIVEN & ORTE_GET_MAPPING_DIRECTIVE(jdata->map->mapping)) ||
OPAL_BINDING_POLICY_IS_SET(jdata->map->binding)) {
directives_given = true;
}
} else if ((ORTE_MAPPING_GIVEN & ORTE_GET_MAPPING_DIRECTIVE(orte_rmaps_base.mapping)) ||
OPAL_BINDING_POLICY_IS_SET(opal_hwloc_binding_policy)) {
directives_given = true;
}

/* check for an affinity file */
if (NULL != (affinity_file = getenv("LSB_AFFINITY_HOSTFILE"))) {
if (!directives_given && NULL != (affinity_file = getenv("LSB_AFFINITY_HOSTFILE"))) {
/* check to see if the file is empty - if it is,
* then affinity wasn't actually set for this job */
if (0 != stat(affinity_file, &buf)) {
Expand Down
1 change: 1 addition & 0 deletions orte/mca/rmaps/base/rmaps_base_frame.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -562,6 +562,7 @@ static int check_modifiers(char *ck, orte_mapping_policy_t *tmp)
for (i=0; NULL != ck2[i]; i++) {
if (0 == strncasecmp(ck2[i], "span", strlen(ck2[i]))) {
ORTE_SET_MAPPING_DIRECTIVE(*tmp, ORTE_MAPPING_SPAN);
ORTE_SET_MAPPING_DIRECTIVE(*tmp, ORTE_MAPPING_GIVEN);
found = true;
} else if (0 == strncasecmp(ck2[i], "pe", strlen("pe"))) {
/* break this at the = sign to get the number */
Expand Down
21 changes: 0 additions & 21 deletions orte/orted/pmix/pmix_server_dyn.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -282,13 +282,6 @@ int pmix_server_spawn_fn(opal_process_name_t *requestor,

/*** MAP-BY ***/
} else if (0 == strcmp(info->key, OPAL_PMIX_MAPBY)) {
if (ORTE_MAPPING_POLICY_IS_SET(jdata->map->mapping)) {
/* not allowed to provide multiple mapping policies */
orte_show_help("help-orte-rmaps-base.txt", "redefining-policy",
true, "mapping", info->data.string,
orte_rmaps_base_print_mapping(orte_rmaps_base.mapping));
return ORTE_ERR_BAD_PARAM;
}
rc = orte_rmaps_base_set_mapping_policy(jdata, &jdata->map->mapping,
NULL, info->data.string);
if (ORTE_SUCCESS != rc) {
Expand All @@ -297,13 +290,6 @@ int pmix_server_spawn_fn(opal_process_name_t *requestor,

/*** RANK-BY ***/
} else if (0 == strcmp(info->key, OPAL_PMIX_RANKBY)) {
if (ORTE_RANKING_POLICY_IS_SET(jdata->map->ranking)) {
/* not allowed to provide multiple ranking policies */
orte_show_help("help-orte-rmaps-base.txt", "redefining-policy",
true, "ranking", info->data.string,
orte_rmaps_base_print_ranking(orte_rmaps_base.ranking));
return ORTE_ERR_BAD_PARAM;
}
rc = orte_rmaps_base_set_ranking_policy(&jdata->map->ranking,
jdata->map->mapping,
info->data.string);
Expand All @@ -313,13 +299,6 @@ int pmix_server_spawn_fn(opal_process_name_t *requestor,

/*** BIND-TO ***/
} else if (0 == strcmp(info->key, OPAL_PMIX_BINDTO)) {
if (OPAL_BINDING_POLICY_IS_SET(jdata->map->binding)) {
/* not allowed to provide multiple mapping policies */
orte_show_help("help-opal-hwloc-base.txt", "redefining-policy", true,
info->data.string,
opal_hwloc_base_print_binding(opal_hwloc_binding_policy));
return ORTE_ERR_BAD_PARAM;
}
rc = opal_hwloc_base_set_binding_policy(&jdata->map->binding,
info->data.string);
if (ORTE_SUCCESS != rc) {
Expand Down