982149d8c8
1. fix in oshmem scoll component: basic algorithms should call basic collectives since their implementation incompatible with others (fca, hcoll). 2. Set OPAL_EVLOOP_ONCE flag ON for libevent in the case of yoda smpl. Otherwise there is possible deadlock in atomic_basic_lock call fixed by Val, Igor, reviewed by Miked cmr=v1.7.5:reviewer=ompi-rm1.7 This commit was SVN r30762.
251 строка
8.7 KiB
C
251 строка
8.7 KiB
C
/*
|
||
* Copyright (c) 2013 Mellanox Technologies, Inc.
|
||
* All rights reserved.
|
||
* $COPYRIGHT$
|
||
*
|
||
* Additional copyrights may follow
|
||
*
|
||
* $HEADER$
|
||
*/
|
||
|
||
#include "oshmem_config.h"
|
||
#include <stdio.h>
|
||
#include <stdlib.h>
|
||
|
||
#include "orte/mca/grpcomm/grpcomm.h"
|
||
|
||
#include "opal/util/bit_ops.h"
|
||
|
||
#include "oshmem/constants.h"
|
||
#include "oshmem/mca/spml/spml.h"
|
||
#include "oshmem/mca/scoll/scoll.h"
|
||
#include "oshmem/mca/scoll/base/base.h"
|
||
#include "scoll_basic.h"
|
||
|
||
static int _algorithm_central_counter(struct oshmem_group_t *group,
|
||
int PE_root,
|
||
void *target,
|
||
const void *source,
|
||
size_t nlong,
|
||
long *pSync);
|
||
static int _algorithm_binomial_tree(struct oshmem_group_t *group,
|
||
int PE_root,
|
||
void *target,
|
||
const void *source,
|
||
size_t nlong,
|
||
long *pSync);
|
||
|
||
int mca_scoll_basic_broadcast(struct oshmem_group_t *group,
|
||
int PE_root,
|
||
void *target,
|
||
const void *source,
|
||
size_t nlong,
|
||
long *pSync,
|
||
int alg)
|
||
{
|
||
int rc = OSHMEM_SUCCESS;
|
||
|
||
/* Arguments validation */
|
||
if (!group) {
|
||
SCOLL_ERROR("Active set (group) of PE is not defined");
|
||
rc = OSHMEM_ERR_BAD_PARAM;
|
||
}
|
||
|
||
/* Check if this PE is part of the group */
|
||
if ((rc == OSHMEM_SUCCESS) && oshmem_proc_group_is_member(group)) {
|
||
int i = 0;
|
||
|
||
if (pSync) {
|
||
alg = (alg == SCOLL_DEFAULT_ALG ?
|
||
mca_scoll_basic_param_broadcast_algorithm : alg);
|
||
switch (alg) {
|
||
case SCOLL_ALG_BROADCAST_CENTRAL_COUNTER:
|
||
{
|
||
rc = _algorithm_central_counter(group,
|
||
PE_root,
|
||
target,
|
||
source,
|
||
nlong,
|
||
pSync);
|
||
break;
|
||
}
|
||
case SCOLL_ALG_BROADCAST_BINOMIAL:
|
||
{
|
||
rc = _algorithm_binomial_tree(group,
|
||
PE_root,
|
||
target,
|
||
source,
|
||
nlong,
|
||
pSync);
|
||
break;
|
||
}
|
||
default:
|
||
{
|
||
rc = _algorithm_binomial_tree(group,
|
||
PE_root,
|
||
target,
|
||
source,
|
||
nlong,
|
||
pSync);
|
||
}
|
||
}
|
||
} else {
|
||
SCOLL_ERROR("Incorrect argument pSync");
|
||
rc = OSHMEM_ERR_BAD_PARAM;
|
||
}
|
||
|
||
/* Restore initial values */
|
||
SCOLL_VERBOSE(12,
|
||
"[#%d] Restore special synchronization array",
|
||
group->my_pe);
|
||
for (i = 0; pSync && (i < _SHMEM_BCAST_SYNC_SIZE); i++) {
|
||
pSync[i] = _SHMEM_SYNC_VALUE;
|
||
}
|
||
}
|
||
|
||
return rc;
|
||
}
|
||
|
||
/*
|
||
This algorithm is quite simple and straightforward. But because of it<69>s obvious simplicity and
|
||
the naive prove for correctness it is implemented quite often. The root send data to all.
|
||
Outlay:
|
||
NP-1 competing network transfers are needed to implement the counter
|
||
The memory usage is constant (1 byte) per node.
|
||
*/
|
||
static int _algorithm_central_counter(struct oshmem_group_t *group,
|
||
int PE_root,
|
||
void *target,
|
||
const void *source,
|
||
size_t nlong,
|
||
long *pSync)
|
||
{
|
||
int rc = OSHMEM_SUCCESS;
|
||
int i = 0;
|
||
|
||
SCOLL_VERBOSE(12,
|
||
"[#%d] Broadcast algorithm: Central Counter",
|
||
group->my_pe);
|
||
SCOLL_VERBOSE(15,
|
||
"[#%d] pSync[0] = %ld root = #%d",
|
||
group->my_pe, pSync[0], PE_root);
|
||
|
||
/* Check if this PE is the root */
|
||
if (PE_root == group->my_pe) {
|
||
int pe_cur = 0;
|
||
|
||
SCOLL_VERBOSE(14,
|
||
"[#%d] send data to all PE in the group",
|
||
group->my_pe);
|
||
for (i = 0; (i < group->proc_count) && (rc == OSHMEM_SUCCESS); i++) {
|
||
pe_cur = oshmem_proc_pe(group->proc_array[i]);
|
||
if (pe_cur != PE_root) {
|
||
SCOLL_VERBOSE(15,
|
||
"[#%d] send data to #%d",
|
||
group->my_pe, pe_cur);
|
||
rc = MCA_SPML_CALL(put(target, nlong, (void *)source, pe_cur));
|
||
}
|
||
}
|
||
}
|
||
|
||
/* Wait for operation completion to set needed size */
|
||
if (rc == OSHMEM_SUCCESS) {
|
||
SCOLL_VERBOSE(14, "[#%d] Wait for operation completion", group->my_pe);
|
||
rc = BARRIER_FUNC(group,
|
||
(pSync + 1),
|
||
SCOLL_DEFAULT_ALG);
|
||
}
|
||
|
||
return rc;
|
||
}
|
||
|
||
/*
|
||
The Binomial Spanning Tree algorithm.
|
||
Outlay:
|
||
The game scales with log2(NP) and uses 1 byte of memory.
|
||
*/
|
||
static int _algorithm_binomial_tree(struct oshmem_group_t *group,
|
||
int PE_root,
|
||
void *target,
|
||
const void *source,
|
||
size_t nlong,
|
||
long *pSync)
|
||
{
|
||
int rc = OSHMEM_SUCCESS;
|
||
long value = SHMEM_SYNC_INIT;
|
||
int root_id = oshmem_proc_group_find_id(group, PE_root);
|
||
int my_id = oshmem_proc_group_find_id(group, group->my_pe);
|
||
int peer_id = 0;
|
||
int peer_pe = 0;
|
||
int vrank;
|
||
int dim = opal_cube_dim(group->proc_count);
|
||
int hibit;
|
||
int mask;
|
||
int i = 0;
|
||
|
||
SCOLL_VERBOSE(12, "[#%d] Broadcast algorithm: Tree", group->my_pe);
|
||
SCOLL_VERBOSE(15,
|
||
"[#%d] pSync[0] = %ld root = #%d",
|
||
group->my_pe, pSync[0], PE_root);
|
||
|
||
vrank = (my_id + group->proc_count - root_id) % group->proc_count;
|
||
hibit = opal_hibit(vrank, dim);
|
||
|
||
SCOLL_VERBOSE(15,
|
||
"[#%d] dim = %d vrank = %d hibit = %d",
|
||
group->my_pe, dim, vrank, hibit);
|
||
|
||
dim--;
|
||
|
||
pSync[0] = SHMEM_SYNC_READY;
|
||
/* Receive data from parent in the tree. */
|
||
if (vrank > 0) {
|
||
value = SHMEM_SYNC_READY;
|
||
|
||
SCOLL_VERBOSE(14, "[#%d] wait", group->my_pe);
|
||
rc = MCA_SPML_CALL(wait((void*)pSync, SHMEM_CMP_NE, (void*)&value, SHMEM_LONG));
|
||
while ((value = pSync[0]) < 0) {
|
||
SCOLL_VERBOSE(14,
|
||
"[#%d] Broadcast size is a negative value (%li)\n",
|
||
group->my_pe, pSync[0]);
|
||
MCA_SPML_CALL(wait((void*)pSync, SHMEM_CMP_NE, (void*)&value, SHMEM_LONG));
|
||
}
|
||
if (OSHMEM_SUCCESS != rc) {
|
||
return rc;
|
||
}
|
||
nlong = (size_t) pSync[0];
|
||
}
|
||
|
||
/* Send data to the children. */
|
||
for (i = hibit + 1, mask = 1 << i; i <= dim; ++i, mask <<= 1) {
|
||
peer_id = vrank | mask;
|
||
|
||
if (peer_id < group->proc_count) {
|
||
/* Wait for the child to be ready to receive (pSync must have the initial value) */
|
||
peer_id = (peer_id + root_id) % group->proc_count;
|
||
peer_pe = oshmem_proc_pe(group->proc_array[peer_id]);
|
||
|
||
SCOLL_VERBOSE(14,
|
||
"[#%d] check remote pe is ready to receive #%d",
|
||
group->my_pe, peer_pe);
|
||
do {
|
||
rc = MCA_SPML_CALL(get((void*)pSync, sizeof(long), (void*)pSync, peer_pe));
|
||
} while ((OSHMEM_SUCCESS == rc) && (pSync[0] != SHMEM_SYNC_READY));
|
||
|
||
SCOLL_VERBOSE(14, "[#%d] send data to #%d", group->my_pe, peer_pe);
|
||
rc = MCA_SPML_CALL(put(target, nlong, (my_id == root_id ? (void *)source : target), peer_pe));
|
||
|
||
MCA_SPML_CALL(fence());
|
||
|
||
SCOLL_VERBOSE(14, "[#%d] signals to #%d", group->my_pe, peer_pe);
|
||
value = nlong;
|
||
rc = MCA_SPML_CALL(put((void*)pSync, sizeof(value), (void*)&value, peer_pe));
|
||
if (OSHMEM_SUCCESS != rc) {
|
||
break;
|
||
}
|
||
}
|
||
}
|
||
|
||
return rc;
|
||
}
|