1
1

Merge pull request #6493 from rhc54/topic/order

Ensure that nodes are always used in order provided
Этот коммит содержится в:
Jeff Squyres 2019-03-19 09:40:21 -04:00 коммит произвёл GitHub
родитель ad8c842e7d 5aa775c02e
Коммит 5111dbd480
Не найден ключ, соответствующий данной подписи
Идентификатор ключа GPG: 4AEE18F83AFDEB23
3 изменённых файлов: 124 добавлений и 92 удалений

Просмотреть файл

@ -12,7 +12,7 @@
* Copyright (c) 2011 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2011-2012 Los Alamos National Security, LLC.
* All rights reserved.
* Copyright (c) 2014-2018 Intel, Inc. All rights reserved.
* Copyright (c) 2014-2019 Intel, Inc. All rights reserved.
* Copyright (c) 2016 IBM Corporation. All rights reserved.
* $COPYRIGHT$
*
@ -210,18 +210,17 @@ int orte_rmaps_base_get_target_nodes(opal_list_t *allocated_nodes, orte_std_cntr
return ORTE_ERR_SILENT;
}
/* find the nodes in our node array and assemble them
* in daemon order if the vm was launched
* in list order as that is what the user specified
*/
for (i=0; i < orte_node_pool->size; i++) {
nd = NULL;
if (NULL == (node = (orte_node_t*)opal_pointer_array_get_item(orte_node_pool, i))) {
continue;
}
/* ignore nodes that are non-usable */
if (ORTE_FLAG_TEST(node, ORTE_NODE_NON_USABLE)) {
continue;
}
OPAL_LIST_FOREACH_SAFE(nptr, next, &nodes, orte_node_t) {
OPAL_LIST_FOREACH_SAFE(nptr, next, &nodes, orte_node_t) {
for (i=0; i < orte_node_pool->size; i++) {
if (NULL == (node = (orte_node_t*)opal_pointer_array_get_item(orte_node_pool, i))) {
continue;
}
/* ignore nodes that are non-usable */
if (ORTE_FLAG_TEST(node, ORTE_NODE_NON_USABLE)) {
continue;
}
if (0 != strcmp(node->name, nptr->name)) {
OPAL_OUTPUT_VERBOSE((10, orte_rmaps_base_framework.framework_output,
"NODE %s DOESNT MATCH NODE %s",
@ -266,37 +265,14 @@ int orte_rmaps_base_get_target_nodes(opal_list_t *allocated_nodes, orte_std_cntr
*/
ORTE_FLAG_UNSET(node, ORTE_NODE_FLAG_MAPPED);
}
if (NULL == nd || NULL == nd->daemon ||
NULL == node->daemon ||
nd->daemon->name.vpid < node->daemon->name.vpid) {
/* just append to end */
opal_list_append(allocated_nodes, &node->super);
nd = node;
} else {
/* starting from end, put this node in daemon-vpid order */
while (node->daemon->name.vpid < nd->daemon->name.vpid) {
if (opal_list_get_begin(allocated_nodes) == opal_list_get_prev(&nd->super)) {
/* insert at beginning */
opal_list_prepend(allocated_nodes, &node->super);
goto moveon1;
}
nd = (orte_node_t*)opal_list_get_prev(&nd->super);
}
item = opal_list_get_next(&nd->super);
if (item == opal_list_get_end(allocated_nodes)) {
/* we are at the end - just append */
opal_list_append(allocated_nodes, &node->super);
} else {
nd = (orte_node_t*)item;
opal_list_insert_pos(allocated_nodes, item, &node->super);
}
moveon1:
/* reset us back to the end for the next node */
nd = (orte_node_t*)opal_list_get_last(allocated_nodes);
}
opal_list_remove_item(&nodes, (opal_list_item_t*)nptr);
OBJ_RELEASE(nptr);
/* the list is ordered as per user direction using -host
* or the listing in -hostfile - preserve that ordering */
opal_list_append(allocated_nodes, &node->super);
break;
}
/* remove the item from the list as we have allocated it */
opal_list_remove_item(&nodes, (opal_list_item_t*)nptr);
OBJ_RELEASE(nptr);
}
OBJ_DESTRUCT(&nodes);
/* now prune for usage and compute total slots */

Просмотреть файл

@ -10,7 +10,7 @@
* Copyright (c) 2004-2005 The Regents of the University of California.
* All rights reserved.
* Copyright (c) 2013 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2014-2018 Intel, Inc. All rights reserved.
* Copyright (c) 2014-2019 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.
@ -52,7 +52,8 @@ int orte_util_dash_host_compute_slots(orte_node_t *node, char *hosts)
/* see if this node appears in the list */
for (n=0; NULL != specs[n]; n++) {
if (0 == strncmp(node->name, specs[n], strlen(node->name))) {
if (0 == strncmp(node->name, specs[n], strlen(node->name)) ||
(orte_ifislocal(node->name) && orte_ifislocal(specs[n]))) {
/* check if the #slots was specified */
if (NULL != (cptr = strchr(specs[n], ':'))) {
*cptr = '\0';

Просмотреть файл

@ -385,6 +385,7 @@ int orte_util_decode_nidmap(opal_buffer_t *buf)
/* add this name to the pool */
nd = OBJ_NEW(orte_node_t);
nd->name = strdup(names[n]);
nd->index = n;
opal_pointer_array_set_item(orte_node_pool, n, nd);
/* set the topology - always default to homogeneous
* as that is the most common scenario */
@ -409,7 +410,6 @@ int orte_util_decode_nidmap(opal_buffer_t *buf)
daemons->num_procs++;
opal_pointer_array_set_item(daemons->procs, proc->name.vpid, proc);
}
nd->index = proc->name.vpid;
OBJ_RETAIN(nd);
proc->node = nd;
OBJ_RETAIN(proc);
@ -945,8 +945,9 @@ int orte_util_parse_node_info(opal_buffer_t *buf)
int orte_util_generate_ppn(orte_job_t *jdata,
opal_buffer_t *buf)
{
uint16_t *ppn=NULL;
size_t nbytes;
uint16_t ppn;
uint8_t *bytes;
int32_t nbytes;
int rc = ORTE_SUCCESS;
orte_app_idx_t i;
int j, k;
@ -955,32 +956,39 @@ int orte_util_generate_ppn(orte_job_t *jdata,
orte_node_t *nptr;
orte_proc_t *proc;
size_t sz;
opal_buffer_t bucket;
/* make room for the number of procs on each node */
nbytes = sizeof(uint16_t) * orte_node_pool->size;
ppn = (uint16_t*)malloc(nbytes);
OBJ_CONSTRUCT(&bucket, opal_buffer_t);
for (i=0; i < jdata->num_apps; i++) {
/* reset the #procs */
memset(ppn, 0, nbytes);
/* for each app_context, compute the #procs on
* each node of the allocation */
for (j=0; j < orte_node_pool->size; j++) {
if (NULL == (nptr = (orte_node_t*)opal_pointer_array_get_item(orte_node_pool, j))) {
/* for each app_context */
for (j=0; j < jdata->map->nodes->size; j++) {
if (NULL == (nptr = (orte_node_t*)opal_pointer_array_get_item(jdata->map->nodes, j))) {
continue;
}
if (NULL == nptr->daemon) {
continue;
}
ppn = 0;
for (k=0; k < nptr->procs->size; k++) {
if (NULL != (proc = (orte_proc_t*)opal_pointer_array_get_item(nptr->procs, k))) {
if (proc->name.jobid == jdata->jobid) {
++ppn[j];
++ppn;
}
}
}
if (0 < ppn) {
if (ORTE_SUCCESS != (rc = opal_dss.pack(&bucket, &nptr->index, 1, ORTE_STD_CNTR))) {
goto cleanup;
}
if (ORTE_SUCCESS != (rc = opal_dss.pack(&bucket, &ppn, 1, OPAL_UINT16))) {
goto cleanup;
}
}
}
if (opal_compress.compress_block((uint8_t*)ppn, nbytes,
opal_dss.unload(&bucket, (void**)&bytes, &nbytes);
if (opal_compress.compress_block(bytes, (size_t)nbytes,
(uint8_t**)&bo.bytes, &sz)) {
/* mark that this was compressed */
compressed = true;
@ -988,7 +996,7 @@ int orte_util_generate_ppn(orte_job_t *jdata,
} else {
/* mark that this was not compressed */
compressed = false;
bo.bytes = (uint8_t*)ppn;
bo.bytes = bytes;
bo.size = nbytes;
}
/* indicate compression */
@ -1015,21 +1023,31 @@ int orte_util_generate_ppn(orte_job_t *jdata,
}
cleanup:
free(ppn);
OBJ_DESTRUCT(&bucket);
return rc;
}
int orte_util_decode_ppn(orte_job_t *jdata,
opal_buffer_t *buf)
{
orte_std_cntr_t index;
orte_app_idx_t n;
int m, cnt, rc;
int cnt, rc, m;
opal_byte_object_t *boptr;
bool compressed;
uint8_t *bytes;
size_t sz;
uint16_t *ppn, k;
uint16_t ppn, k;
orte_node_t *node;
orte_proc_t *proc;
opal_buffer_t bucket;
/* reset any flags */
for (m=0; m < orte_node_pool->size; m++) {
if (NULL != (node = (orte_node_t*)opal_pointer_array_get_item(orte_node_pool, m))) {
ORTE_FLAG_UNSET(node, ORTE_NODE_FLAG_MAPPED);
}
}
for (n=0; n < jdata->num_apps; n++) {
/* unpack the compression flag */
@ -1062,14 +1080,15 @@ int orte_util_decode_ppn(orte_job_t *jdata,
/* decompress if required */
if (compressed) {
if (!opal_compress.decompress_block((uint8_t**)&ppn, sz,
if (!opal_compress.decompress_block(&bytes, sz,
boptr->bytes, boptr->size)) {
ORTE_ERROR_LOG(ORTE_ERROR);
OBJ_RELEASE(boptr);
return ORTE_ERROR;
}
} else {
ppn = (uint16_t*)boptr->bytes;
bytes = boptr->bytes;
sz = boptr->size;
boptr->bytes = NULL;
boptr->size = 0;
}
@ -1078,38 +1097,74 @@ int orte_util_decode_ppn(orte_job_t *jdata,
}
free(boptr);
/* cycle thru the node pool */
for (m=0; m < orte_node_pool->size; m++) {
if (NULL == (node = (orte_node_t*)opal_pointer_array_get_item(orte_node_pool, m))) {
continue;
/* setup to unpack */
OBJ_CONSTRUCT(&bucket, opal_buffer_t);
opal_dss.load(&bucket, bytes, sz);
/* unpack each node and its ppn */
cnt = 1;
while (OPAL_SUCCESS == (rc = opal_dss.unpack(&bucket, &index, &cnt, ORTE_STD_CNTR))) {
/* get the corresponding node object */
if (NULL == (node = (orte_node_t*)opal_pointer_array_get_item(orte_node_pool, index))) {
rc = ORTE_ERR_NOT_FOUND;
ORTE_ERROR_LOG(rc);
goto error;
}
if (0 < ppn[m]) {
if (!ORTE_FLAG_TEST(node, ORTE_NODE_FLAG_MAPPED)) {
OBJ_RETAIN(node);
ORTE_FLAG_SET(node, ORTE_NODE_FLAG_MAPPED);
opal_pointer_array_add(jdata->map->nodes, node);
}
/* create a proc object for each one */
for (k=0; k < ppn[m]; k++) {
proc = OBJ_NEW(orte_proc_t);
proc->name.jobid = jdata->jobid;
/* leave the vpid undefined as this will be determined
* later when we do the overall ranking */
proc->app_idx = n;
proc->parent = node->daemon->name.vpid;
OBJ_RETAIN(node);
proc->node = node;
/* flag the proc as ready for launch */
proc->state = ORTE_PROC_STATE_INIT;
opal_pointer_array_add(node->procs, proc);
/* we will add the proc to the jdata array when we
* compute its rank */
}
node->num_procs += ppn[m];
/* add the node to the job map if not already assigned */
if (!ORTE_FLAG_TEST(node, ORTE_NODE_FLAG_MAPPED)) {
OBJ_RETAIN(node);
opal_pointer_array_add(jdata->map->nodes, node);
ORTE_FLAG_SET(node, ORTE_NODE_FLAG_MAPPED);
}
/* get the ppn */
cnt = 1;
if (OPAL_SUCCESS != (rc = opal_dss.unpack(&bucket, &ppn, &cnt, OPAL_UINT16))) {
ORTE_ERROR_LOG(rc);
goto error;
}
/* create a proc object for each one */
for (k=0; k < ppn; k++) {
proc = OBJ_NEW(orte_proc_t);
proc->name.jobid = jdata->jobid;
/* leave the vpid undefined as this will be determined
* later when we do the overall ranking */
proc->app_idx = n;
proc->parent = node->daemon->name.vpid;
OBJ_RETAIN(node);
proc->node = node;
/* flag the proc as ready for launch */
proc->state = ORTE_PROC_STATE_INIT;
opal_pointer_array_add(node->procs, proc);
node->num_procs++;
/* we will add the proc to the jdata array when we
* compute its rank */
}
node->num_procs += ppn;
cnt = 1;
}
free(ppn);
OBJ_DESTRUCT(&bucket);
}
if (OPAL_ERR_UNPACK_READ_PAST_END_OF_BUFFER != rc) {
ORTE_ERROR_LOG(rc);
}
/* reset any flags */
for (m=0; m < jdata->map->nodes->size; m++) {
node = (orte_node_t*)opal_pointer_array_get_item(jdata->map->nodes, m);
if (NULL != node) {
ORTE_FLAG_UNSET(node, ORTE_NODE_FLAG_MAPPED);
}
}
return ORTE_SUCCESS;
error:
OBJ_DESTRUCT(&bucket);
/* reset any flags */
for (m=0; m < jdata->map->nodes->size; m++) {
node = (orte_node_t*)opal_pointer_array_get_item(jdata->map->nodes, m);
if (NULL != node) {
ORTE_FLAG_UNSET(node, ORTE_NODE_FLAG_MAPPED);
}
}
return rc;
}