1
1
openmpi/orte/mca/rmaps/ppr/rmaps_ppr.c
Ralph Castain bc7cc09749 After a lot of pain, I've managed to resolve the problem of conflicting mapping directives caused by mismatched MCA params - i.e., where someone has one variant of an MCA param (e.g., rmaps_base_mapping_policy) in their default MCA param file, and then specifies another variant (e.g., --npernode) on the command line. I can't fully resolve the problem as there is no way to know precisely what the user meant - we can only guess which param was really intended since the MCA param system
can't apply its normal precedence rules.

So...print a big "deprecated" warning for the old params and error out if a conflict is detected. I know that isn't what people really wanted, but it's the best we
 can do. If only the old style param is given, then process it after the warning.

Extend the current map-by param to add support for ppr and cpus-per-proc, adding the latter to the list of allowed modifiers using "pe=n" for processing elements/proc. Thus, you can map-by socket:pe=2,oversubscribe to map by socket, binding 2 processing elements/process, with oversubscription allowed. Or you can map-by ppr:2:socket:pe=4 to map two processes to every socket in the allocation, binding each process to 4 processing elements.

For those wondering, a processing element is defined as a hwthread if --use-hwthreads-as-cpus is given, or else as a core.

Refs trac:4117

This commit was SVN r30620.

The following Trac tickets were found above:
  Ticket 4117 --> https://svn.open-mpi.org/trac/ompi/ticket/4117
2014-02-07 21:25:40 +00:00

610 строки
22 KiB
C

/*
* Copyright (c) 2011 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2011 Los Alamos National Security, LLC.
* All rights reserved.
* $COPYRIGHT$
*
* Additional copyrights may follow
*
* $HEADER$
*/
#include "orte_config.h"
#include "orte/constants.h"
#include "orte/types.h"
#include <errno.h>
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif /* HAVE_UNISTD_H */
#ifdef HAVE_STRING_H
#include <string.h>
#endif /* HAVE_STRING_H */
#include "opal/mca/hwloc/base/base.h"
#include "opal/util/argv.h"
#include "orte/util/show_help.h"
#include "orte/mca/errmgr/errmgr.h"
#include "orte/mca/rmaps/base/rmaps_private.h"
#include "orte/mca/rmaps/base/base.h"
#include "rmaps_ppr.h"
static int ppr_mapper(orte_job_t *jdata);
orte_rmaps_base_module_t orte_rmaps_ppr_module = {
ppr_mapper
};
/* RHC: will eventually remove this
* definition as it is no longer reqd
* in the rest of OMPI system.
*
* Define a hierarchical level value that
* helps resolve the hwloc behavior of
* treating caches as a single type of
* entity - must always be available
*/
typedef enum {
OPAL_HWLOC_NODE_LEVEL=0,
OPAL_HWLOC_NUMA_LEVEL,
OPAL_HWLOC_SOCKET_LEVEL,
OPAL_HWLOC_L3CACHE_LEVEL,
OPAL_HWLOC_L2CACHE_LEVEL,
OPAL_HWLOC_L1CACHE_LEVEL,
OPAL_HWLOC_CORE_LEVEL,
OPAL_HWLOC_HWTHREAD_LEVEL
} opal_hwloc_level_t;
#if OPAL_HAVE_HWLOC
static void prune(orte_jobid_t jobid,
orte_app_idx_t app_idx,
orte_node_t *node,
opal_hwloc_level_t *level,
orte_vpid_t *nmapped);
#endif
static int ppr[OPAL_HWLOC_HWTHREAD_LEVEL+1];
static int ppr_mapper(orte_job_t *jdata)
{
int rc = ORTE_SUCCESS, j, n;
mca_base_component_t *c=&mca_rmaps_ppr_component.base_version;
orte_node_t *node;
orte_proc_t *proc;
orte_app_context_t *app;
orte_vpid_t total_procs, nprocs_mapped;
opal_hwloc_level_t start=OPAL_HWLOC_NODE_LEVEL;
#if OPAL_HAVE_HWLOC
hwloc_obj_t obj;
hwloc_obj_type_t lowest;
unsigned cache_level=0;
unsigned int nobjs, i;
bool pruning_reqd = false;
opal_hwloc_level_t level;
#endif
opal_list_t node_list;
opal_list_item_t *item;
orte_std_cntr_t num_slots;
orte_app_idx_t idx;
char **ppr_req, **ck;
size_t len;
bool initial_map=true;
/* only handle initial launch of loadbalanced
* or NPERxxx jobs - allow restarting of failed apps
*/
if (ORTE_JOB_CONTROL_RESTART & jdata->controls) {
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: job %s being restarted - ppr cannot map",
ORTE_JOBID_PRINT(jdata->jobid));
return ORTE_ERR_TAKE_NEXT_OPTION;
}
if (NULL != jdata->map->req_mapper &&
0 != strcasecmp(jdata->map->req_mapper, c->mca_component_name)) {
/* a mapper has been specified, and it isn't me */
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: job %s not using ppr mapper",
ORTE_JOBID_PRINT(jdata->jobid));
return ORTE_ERR_TAKE_NEXT_OPTION;
}
if (NULL == jdata->map->ppr ||
ORTE_MAPPING_PPR != ORTE_GET_MAPPING_POLICY(jdata->map->mapping)) {
/* not for us */
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: job %s not using ppr mapper",
ORTE_JOBID_PRINT(jdata->jobid));
return ORTE_ERR_TAKE_NEXT_OPTION;
}
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: mapping job %s with ppr %s",
ORTE_JOBID_PRINT(jdata->jobid), jdata->map->ppr);
/* flag that I did the mapping */
if (NULL != jdata->map->last_mapper) {
free(jdata->map->last_mapper);
}
jdata->map->last_mapper = strdup(c->mca_component_name);
/* initialize */
memset(ppr, 0, OPAL_HWLOC_HWTHREAD_LEVEL * sizeof(opal_hwloc_level_t));
/* parse option */
n=0;
ppr_req = opal_argv_split(jdata->map->ppr, ',');
for (j=0; NULL != ppr_req[j]; j++) {
/* split on the colon */
ck = opal_argv_split(ppr_req[j], ':');
if (2 != opal_argv_count(ck)) {
/* must provide a specification */
orte_show_help("help-orte-rmaps-ppr.txt", "invalid-ppr", true, jdata->map->ppr);
opal_argv_free(ppr_req);
opal_argv_free(ck);
return ORTE_ERR_SILENT;
}
len = strlen(ck[1]);
if (0 == strncasecmp(ck[1], "node", len)) {
ppr[OPAL_HWLOC_NODE_LEVEL] = strtol(ck[0], NULL, 10);
ORTE_SET_MAPPING_POLICY(jdata->map->mapping, ORTE_MAPPING_BYNODE);
start = OPAL_HWLOC_NODE_LEVEL;
n++;
#if OPAL_HAVE_HWLOC
} else if (0 == strncasecmp(ck[1], "hwthread", len) ||
0 == strncasecmp(ck[1], "thread", len)) {
ppr[OPAL_HWLOC_HWTHREAD_LEVEL] = strtol(ck[0], NULL, 10);
start = OPAL_HWLOC_HWTHREAD_LEVEL;
ORTE_SET_MAPPING_POLICY(jdata->map->mapping, ORTE_MAPPING_BYHWTHREAD);
n++;
} else if (0 == strncasecmp(ck[1], "core", len)) {
ppr[OPAL_HWLOC_CORE_LEVEL] = strtol(ck[0], NULL, 10);
if (start < OPAL_HWLOC_CORE_LEVEL) {
start = OPAL_HWLOC_CORE_LEVEL;
ORTE_SET_MAPPING_POLICY(jdata->map->mapping, ORTE_MAPPING_BYCORE);
}
n++;
} else if (0 == strncasecmp(ck[1], "socket", len) ||
0 == strncasecmp(ck[1], "skt", len)) {
ppr[OPAL_HWLOC_SOCKET_LEVEL] = strtol(ck[0], NULL, 10);
if (start < OPAL_HWLOC_SOCKET_LEVEL) {
start = OPAL_HWLOC_SOCKET_LEVEL;
ORTE_SET_MAPPING_POLICY(jdata->map->mapping, ORTE_MAPPING_BYSOCKET);
}
n++;
} else if (0 == strncasecmp(ck[1], "l1cache", len)) {
ppr[OPAL_HWLOC_L1CACHE_LEVEL] = strtol(ck[0], NULL, 10);
if (start < OPAL_HWLOC_L1CACHE_LEVEL) {
start = OPAL_HWLOC_L1CACHE_LEVEL;
cache_level = 1;
ORTE_SET_MAPPING_POLICY(jdata->map->mapping, ORTE_MAPPING_BYL1CACHE);
}
n++;
} else if (0 == strncasecmp(ck[1], "l2cache", len)) {
ppr[OPAL_HWLOC_L2CACHE_LEVEL] = strtol(ck[0], NULL, 10);
if (start < OPAL_HWLOC_L2CACHE_LEVEL) {
start = OPAL_HWLOC_L2CACHE_LEVEL;
cache_level = 2;
ORTE_SET_MAPPING_POLICY(jdata->map->mapping, ORTE_MAPPING_BYL2CACHE);
}
n++;
} else if (0 == strncasecmp(ck[1], "l3cache", len)) {
ppr[OPAL_HWLOC_L3CACHE_LEVEL] = strtol(ck[0], NULL, 10);
if (start < OPAL_HWLOC_L3CACHE_LEVEL) {
start = OPAL_HWLOC_L3CACHE_LEVEL;
cache_level = 3;
ORTE_SET_MAPPING_POLICY(jdata->map->mapping, ORTE_MAPPING_BYL3CACHE);
}
n++;
} else if (0 == strncasecmp(ck[1], "numa", len)) {
ppr[OPAL_HWLOC_NUMA_LEVEL] = strtol(ck[0], NULL, 10);
if (start < OPAL_HWLOC_NUMA_LEVEL) {
start = OPAL_HWLOC_NUMA_LEVEL;
ORTE_SET_MAPPING_POLICY(jdata->map->mapping, ORTE_MAPPING_BYNUMA);
}
n++;
#endif
} else {
/* unknown spec */
orte_show_help("help-orte-rmaps-ppr.txt", "unrecognized-ppr-option", true, ck[1], jdata->map->ppr);
opal_argv_free(ppr_req);
opal_argv_free(ck);
return ORTE_ERR_SILENT;
}
opal_argv_free(ck);
}
opal_argv_free(ppr_req);
/* if nothing was given, that's an error */
if (0 == n) {
opal_output(0, "NOTHING GIVEN");
return ORTE_ERR_SILENT;
}
#if OPAL_HAVE_HWLOC
/* if more than one level was specified, then pruning will be reqd */
if (1 < n) {
pruning_reqd = true;
}
#endif
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: job %s assigned policy %s",
ORTE_JOBID_PRINT(jdata->jobid),
orte_rmaps_base_print_mapping(jdata->map->mapping));
#if OPAL_HAVE_HWLOC
/* convenience */
level = start;
lowest = opal_hwloc_levels[start];
#endif
for (idx=0; idx < (orte_app_idx_t)jdata->apps->size; idx++) {
if (NULL == (app = (orte_app_context_t*)opal_pointer_array_get_item(jdata->apps, idx))) {
continue;
}
/* if the number of total procs was given, set that
* limit - otherwise, set to max so we simply fill
* all the nodes with the pattern
*/
if (0 < app->num_procs) {
total_procs = app->num_procs;
} else {
total_procs = ORTE_VPID_MAX;
}
/* get the available nodes */
OBJ_CONSTRUCT(&node_list, opal_list_t);
if(ORTE_SUCCESS != (rc = orte_rmaps_base_get_target_nodes(&node_list, &num_slots, app,
jdata->map->mapping, initial_map, false))) {
ORTE_ERROR_LOG(rc);
goto error;
}
/* flag that all subsequent requests should not reset the node->mapped flag */
initial_map = false;
/* if a bookmark exists from some prior mapping, set us to start there */
jdata->bookmark = orte_rmaps_base_get_starting_point(&node_list, jdata);
/* cycle across the nodes */
nprocs_mapped = 0;
for (item = opal_list_get_first(&node_list);
item != opal_list_get_end(&node_list);
item = opal_list_get_next(item)) {
node = (orte_node_t*)item;
#if OPAL_HAVE_HWLOC
/* bozo check */
if (NULL == node->topology) {
orte_show_help("help-orte-rmaps-ppr.txt", "ppr-topo-missing",
true, node->name);
rc = ORTE_ERR_SILENT;
goto error;
}
#endif
/* add the node to the map, if needed */
if (!node->mapped) {
if (ORTE_SUCCESS > (rc = opal_pointer_array_add(jdata->map->nodes, (void*)node))) {
ORTE_ERROR_LOG(rc);
goto error;
}
node->mapped = true;
OBJ_RETAIN(node); /* maintain accounting on object */
jdata->map->num_nodes++;
}
/* if we are mapping solely at the node level, just put
* that many procs on this node
*/
if (OPAL_HWLOC_NODE_LEVEL == start) {
#if OPAL_HAVE_HWLOC
obj = hwloc_get_root_obj(node->topology);
#endif
for (j=0; j < ppr[start] && nprocs_mapped < total_procs; j++) {
if (NULL == (proc = orte_rmaps_base_setup_proc(jdata, node, idx))) {
rc = ORTE_ERR_OUT_OF_RESOURCE;
goto error;
}
nprocs_mapped++;
#if OPAL_HAVE_HWLOC
proc->locale = obj;
#endif
}
#if OPAL_HAVE_HWLOC
} else {
/* get the number of lowest resources on this node */
nobjs = opal_hwloc_base_get_nbobjs_by_type(node->topology,
lowest, cache_level,
OPAL_HWLOC_AVAILABLE);
/* map the specified number of procs to each such resource on this node,
* recording the locale of each proc so we know its cpuset
*/
for (i=0; i < nobjs; i++) {
obj = opal_hwloc_base_get_obj_by_type(node->topology,
lowest, cache_level,
i, OPAL_HWLOC_AVAILABLE);
for (j=0; j < ppr[start] && nprocs_mapped < total_procs; j++) {
if (NULL == (proc = orte_rmaps_base_setup_proc(jdata, node, idx))) {
rc = ORTE_ERR_OUT_OF_RESOURCE;
goto error;
}
nprocs_mapped++;
proc->locale = obj;
}
}
if (pruning_reqd) {
/* go up the ladder and prune the procs according to
* the specification, adjusting the count of procs on the
* node as we go
*/
level--;
prune(jdata->jobid, idx, node, &level, &nprocs_mapped);
}
#endif
}
/* set the total slots used */
if ((int)node->num_procs <= node->slots) {
node->slots_inuse = (int)node->num_procs;
} else {
node->slots_inuse = node->slots;
}
/* if no-oversubscribe was specified, check to see if
* we have violated the total slot specification - regardless,
* if slots_max was given, we are not allowed to violate it!
*/
if ((node->slots < (int)node->num_procs) ||
(0 < node->slots_max && node->slots_max < (int)node->num_procs)) {
if (ORTE_MAPPING_NO_OVERSUBSCRIBE & ORTE_GET_MAPPING_DIRECTIVE(jdata->map->mapping)) {
orte_show_help("help-orte-rmaps-base.txt", "orte-rmaps-base:alloc-error",
true, node->num_procs, app->app);
rc = ORTE_ERR_SILENT;
goto error;
}
/* flag the node as oversubscribed so that sched-yield gets
* properly set
*/
node->oversubscribed = true;
}
/* if we haven't mapped all the procs, continue on to the
* next node
*/
if (total_procs == nprocs_mapped) {
break;
}
}
if (0 == app->num_procs) {
app->num_procs = nprocs_mapped;
}
if (ORTE_VPID_MAX != total_procs && nprocs_mapped < total_procs) {
/* couldn't map them all */
orte_show_help("help-orte-rmaps-ppr.txt", "ppr-too-many-procs",
true, app->app, app->num_procs, jdata->map->ppr);
rc = ORTE_ERR_SILENT;
goto error;
}
/* compute vpids and add proc objects to the job */
if (ORTE_SUCCESS != (rc = orte_rmaps_base_compute_vpids(jdata, app, &node_list))) {
ORTE_ERROR_LOG(rc);
goto error;
}
/* track the total number of processes we mapped - must update
* this AFTER we compute vpids so that computation is done
* correctly
*/
jdata->num_procs += app->num_procs;
while (NULL != (item = opal_list_remove_first(&node_list))) {
OBJ_RELEASE(item);
}
OBJ_DESTRUCT(&node_list);
}
return ORTE_SUCCESS;
error:
while (NULL != (item = opal_list_remove_first(&node_list))) {
OBJ_RELEASE(item);
}
OBJ_DESTRUCT(&node_list);
return rc;
}
#if OPAL_HAVE_HWLOC
static hwloc_obj_t find_split(hwloc_topology_t topo, hwloc_obj_t obj)
{
unsigned k;
hwloc_obj_t nxt;
if (1 < obj->arity) {
return obj;
}
for (k=0; k < obj->arity; k++) {
nxt = find_split(topo, obj->children[k]);
if (NULL != nxt) {
return nxt;
}
}
return NULL;
}
/* recursively climb the topology, pruning procs beyond that allowed
* by the given ppr
*/
static void prune(orte_jobid_t jobid,
orte_app_idx_t app_idx,
orte_node_t *node,
opal_hwloc_level_t *level,
orte_vpid_t *nmapped)
{
hwloc_obj_t obj, top;
unsigned int i, nobjs;
hwloc_obj_type_t lvl;
unsigned cache_level = 0, k;
int nprocs;
hwloc_cpuset_t avail, cpus, childcpus;
int n, limit, nmax, nunder, idx, idxmax = 0;
orte_proc_t *proc, *pptr, *procmax;
opal_hwloc_level_t ll;
char dang[64];
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: pruning level %d",
*level);
/* convenience */
ll = *level;
/* convenience */
lvl = opal_hwloc_levels[ll];
limit = ppr[ll];
if (0 == limit) {
/* no limit at this level, so move up if necessary */
if (0 == ll) {
/* done */
return;
}
--(*level);
prune(jobid, app_idx, node, level, nmapped);
return;
}
/* handle the darn cache thing again */
if (OPAL_HWLOC_L3CACHE_LEVEL == ll) {
cache_level = 3;
} else if (OPAL_HWLOC_L2CACHE_LEVEL == ll) {
cache_level = 2;
} else if (OPAL_HWLOC_L1CACHE_LEVEL == ll) {
cache_level = 1;
}
/* get the number of resources at this level on this node */
nobjs = opal_hwloc_base_get_nbobjs_by_type(node->topology,
lvl, cache_level,
OPAL_HWLOC_AVAILABLE);
/* for each resource, compute the number of procs sitting
* underneath it and check against the limit
*/
for (i=0; i < nobjs; i++) {
obj = opal_hwloc_base_get_obj_by_type(node->topology,
lvl, cache_level,
i, OPAL_HWLOC_AVAILABLE);
/* get the available cpuset */
avail = opal_hwloc_base_get_available_cpus(node->topology, obj);
/* look at the intersection of this object's cpuset and that
* of each proc in the job/app - if they intersect, then count this proc
* against the limit
*/
nprocs = 0;
for (n=0; n < node->procs->size; n++) {
if (NULL == (proc = (orte_proc_t*)opal_pointer_array_get_item(node->procs, n))) {
continue;
}
if (proc->name.jobid != jobid ||
proc->app_idx != app_idx) {
continue;
}
cpus = opal_hwloc_base_get_available_cpus(node->topology, proc->locale);
if (hwloc_bitmap_intersects(avail, cpus)) {
nprocs++;
}
}
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: found %d procs limit %d",
nprocs, limit);
/* check against the limit */
while (limit < nprocs) {
/* need to remove procs - do this in a semi-intelligent
* manner to provide a little load balancing by cycling
* across the objects beneath this one, removing procs
* in a round-robin fashion until the limit is satisfied
*
* NOTE: I'm sure someone more knowledgeable with hwloc
* will come up with a more efficient way to do this, so
* consider this is a starting point
*/
/* find the first level that has more than
* one child beneath it - if all levels
* have only one child, then return this
* object
*/
top = find_split(node->topology, obj);
hwloc_obj_type_snprintf(dang, 64, top, 1);
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: SPLIT AT LEVEL %s", dang);
/* cycle across the children of this object */
nmax = 0;
procmax = NULL;
idx = 0;
/* find the child with the most procs underneath it */
for (k=0; k < top->arity && limit < nprocs; k++) {
/* get this object's available cpuset */
childcpus = opal_hwloc_base_get_available_cpus(node->topology, top->children[k]);
nunder = 0;
pptr = NULL;
for (n=0; n < node->procs->size; n++) {
if (NULL == (proc = (orte_proc_t*)opal_pointer_array_get_item(node->procs, n))) {
continue;
}
if (proc->name.jobid != jobid ||
proc->app_idx != app_idx) {
continue;
}
cpus = opal_hwloc_base_get_available_cpus(node->topology, proc->locale);
if (hwloc_bitmap_intersects(childcpus, cpus)) {
nunder++;
if (NULL == pptr) {
/* save the location of the first proc under this object */
pptr = proc;
idx = n;
}
}
}
if (nmax < nunder) {
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: PROCS UNDER CHILD %d %d MAX %d",
k, nunder, nmax);
nmax = nunder;
procmax = pptr;
idxmax = idx;
}
}
if (NULL == procmax) {
/* can't find anything to remove - error out */
goto error;
}
/* remove it */
opal_output_verbose(5, orte_rmaps_base_framework.framework_output,
"mca:rmaps:ppr: removing proc at posn %d",
idxmax);
opal_pointer_array_set_item(node->procs, idxmax, NULL);
node->num_procs--;
node->slots_inuse--;
if (node->slots_inuse < 0) {
node->slots_inuse = 0;
}
nprocs--;
*nmapped -= 1;
OBJ_RELEASE(procmax);
}
}
/* finished with this level - move up if necessary */
if (0 == ll) {
return;
}
--(*level);
prune(jobid, app_idx, node, level, nmapped);
return;
error:
opal_output(0, "INFINITE LOOP");
}
#endif