1
1
openmpi/opal/mca/pmix/s2/pmix_s2.c

622 строки
18 KiB
C
Исходник Обычный вид История

Per the PMIx RFC: WHAT: Merge the PMIx branch into the devel repo, creating a new OPAL “lmix” framework to abstract PMI support for all RTEs. Replace the ORTE daemon-level collectives with a new PMIx server and update the ORTE grpcomm framework to support server-to-server collectives WHY: We’ve had problems dealing with variations in PMI implementations, and need to extend the existing PMI definitions to meet exascale requirements. WHEN: Mon, Aug 25 WHERE: https://github.com/rhc54/ompi-svn-mirror.git Several community members have been working on a refactoring of the current PMI support within OMPI. Although the APIs are common, Slurm and Cray implement a different range of capabilities, and package them differently. For example, Cray provides an integrated PMI-1/2 library, while Slurm separates the two and requires the user to specify the one to be used at runtime. In addition, several bugs in the Slurm implementations have caused problems requiring extra coding. All this has led to a slew of #if’s in the PMI code and bugs when the corner-case logic for one implementation accidentally traps the other. Extending this support to other implementations would have increased this complexity to an unacceptable level. Accordingly, we have: * created a new OPAL “pmix” framework to abstract the PMI support, with separate components for Cray, Slurm PMI-1, and Slurm PMI-2 implementations. * Replaced the current ORTE grpcomm daemon-based collective operation with an integrated PMIx server, and updated the grpcomm APIs to provide more flexible, multi-algorithm support for collective operations. At this time, only the xcast and allgather operations are supported. * Replaced the current global collective id with a signature based on the names of the participating procs. The allows an unlimited number of collectives to be executed by any group of processes, subject to the requirement that only one collective can be active at a time for a unique combination of procs. Note that a proc can be involved in any number of simultaneous collectives - it is the specific combination of procs that is subject to the constraint * removed the prior OMPI/OPAL modex code * added new macros for executing modex send/recv to simplify use of the new APIs. The send macros allow the caller to specify whether or not the BTL supports async modex operations - if so, then the non-blocking “fence” operation is used, if the active PMIx component supports it. Otherwise, the default is a full blocking modex exchange as we currently perform. * retained the current flag that directs us to use a blocking fence operation, but only to retrieve data upon demand This commit was SVN r32570.
2014-08-21 18:56:47 +00:00
/* -*- Mode: C; c-basic-offset:4 ; indent-tabs-mode:nil -*- */
/*
* Copyright (c) 2007 The Trustees of Indiana University.
* All rights reserved.
* Copyright (c) 2011 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2011-2013 Los Alamos National Security, LLC. All
* rights reserved.
* Copyright (c) 2013-2014 Intel, Inc. All rights reserved.
* $COPYRIGHT$
*
* Additional copyrights may follow
*
* $HEADER$
*/
#include "opal_config.h"
#include "opal/constants.h"
#include "opal/types.h"
#include "opal_stdint.h"
#include "opal/mca/hwloc/base/base.h"
#include "opal/util/output.h"
#include "opal/util/proc.h"
#include "opal/util/show_help.h"
#include "pmi2_pmap_parser.h"
#include <string.h>
#include <pmi.h>
#include <pmi2.h>
#include "opal/mca/pmix/base/base.h"
#include "pmix_s2.h"
static int s2_init(void);
static int s2_fini(void);
static bool s2_initialized(void);
static int s2_abort(int flag, const char msg[]);
static int s2_spawn(int count, const char * cmds[],
int argcs[], const char ** argvs[],
const int maxprocs[],
opal_list_t *info_keyval_vector,
opal_list_t *preput_keyval_vector,
char jobId[], int jobIdSize,
int errors[]);
static int s2_put(opal_pmix_scope_t scope,
opal_value_t *kv);
static int s2_fence(opal_process_name_t *procs, size_t nprocs);
static int s2_get(const opal_identifier_t *id,
const char *key,
opal_value_t **kv);
static int s2_publish(const char service_name[],
opal_list_t *info,
const char port[]);
static int s2_lookup(const char service_name[],
opal_list_t *info,
char port[], int portLen);
static int s2_unpublish(const char service_name[],
opal_list_t *info);
static bool s2_get_attr(const char *attr, opal_value_t **kv);
static int s2_job_connect(const char jobId[]);
static int s2_job_disconnect(const char jobId[]);
const opal_pmix_base_module_t opal_pmix_s2_module = {
s2_init,
s2_fini,
s2_initialized,
s2_abort,
s2_fence,
NULL,
s2_put,
s2_get,
NULL,
s2_publish,
s2_lookup,
s2_unpublish,
s2_get_attr,
NULL,
s2_spawn,
s2_job_connect,
s2_job_disconnect,
NULL,
NULL
};
// usage accounting
static int pmix_init_count = 0;
// PMI constant values:
static int pmix_kvslen_max = 0;
static int pmix_keylen_max = 0;
static int pmix_vallen_max = 0;
// Job environment description
static char *pmix_kvs_name = NULL;
static char* pmix_packed_data = NULL;
static int pmix_packed_data_offset = 0;
static int pmix_pack_key = 0;
static uint32_t s2_jobid;
static int s2_rank;
static uint16_t s2_lrank;
static uint16_t s2_nrank;
static int s2_usize;
static int s2_jsize;
static int s2_appnum;
static int s2_nlranks;
static int *s2_lranks=NULL;
static struct {
uint32_t jid;
uint32_t vid;
} s2_pname;
static bool got_modex_data = false;
static char* pmix_error(int pmix_err);
#define OPAL_PMI_ERROR(pmi_err, pmi_func) \
do { \
opal_output(0, "%s [%s:%d:%s]: %s\n", \
pmi_func, __FILE__, __LINE__, __func__, \
pmix_error(pmi_err)); \
} while(0);
static int kvs_put(const char key[], const char value[])
{
int rc;
rc = PMI2_KVS_Put(key, value);
if( PMI2_SUCCESS != rc ){
OPAL_PMI_ERROR(rc, "PMI2_KVS_Put");
return OPAL_ERROR;
}
return OPAL_SUCCESS;
}
static int kvs_get(const char key[], char value [], int maxvalue)
{
int rc;
int len;
rc = PMI2_KVS_Get(pmix_kvs_name, PMI2_ID_NULL, key, value, maxvalue, &len);
if( PMI2_SUCCESS != rc || len < 0){
OPAL_PMI_ERROR(rc, "PMI2_KVS_Get");
return OPAL_ERROR;
}
return OPAL_SUCCESS;
}
static int s2_init(void)
{
int spawned, size, rank, appnum;
int rc, ret = OPAL_ERROR;
char buf[16];
int found;
int my_node;
char *tmp;
uint32_t jobfam, stepid;
int i;
/* if we can't startup PMI, we can't be used */
if ( PMI2_Initialized () ) {
return OPAL_SUCCESS;
}
size = -1;
rank = -1;
appnum = -1;
if (PMI2_SUCCESS != (rc = PMI2_Init(&spawned, &size, &rank, &appnum))) {
opal_show_help("help-pmix-base.txt", "pmix2-init-failed", true, rc);
return OPAL_ERROR;
}
if( size < 0 || rank < 0 ){
opal_show_help("help-pmix-base.txt", "pmix2-init-returned-bad-values", true);
goto err_exit;
}
s2_jsize = size;
s2_rank = rank;
s2_appnum = appnum;
pmix_vallen_max = PMI2_MAX_VALLEN;
pmix_kvslen_max = PMI2_MAX_VALLEN; // FIX ME: What to put here for versatility?
pmix_keylen_max = PMI2_MAX_KEYLEN;
rc = PMI2_Info_GetJobAttr("universeSize", buf, 16, &found);
if( PMI2_SUCCESS != rc ) {
OPAL_PMI_ERROR(rc, "PMI_Get_universe_size");
goto err_exit;
}
s2_usize = atoi(buf);
pmix_kvs_name = (char*)malloc(pmix_kvslen_max);
if( pmix_kvs_name == NULL ){
PMI2_Finalize();
ret = OPAL_ERR_OUT_OF_RESOURCE;
goto err_exit;
}
rc = PMI2_Job_GetId(pmix_kvs_name, pmix_kvslen_max);
if( PMI2_SUCCESS != rc ) {
OPAL_PMI_ERROR(rc, "PMI2_Job_GetId");
goto err_exit;
}
/* Slurm PMI provides the job id as an integer followed
* by a '.', followed by essentially a stepid. The first integer
* defines an overall job number. The second integer is the number of
* individual jobs we have run within that allocation. So we translate
* this as the overall job number equating to our job family, and
* the individual number equating to our local jobid
*/
jobfam = strtoul(pmix_kvs_name, &tmp, 10);
if (NULL == tmp) {
/* hmmm - no '.', so let's just use zero */
stepid = 0;
} else {
tmp++; /* step over the '.' */
stepid = strtoul(tmp, NULL, 10);
}
/* now build the jobid */
s2_jobid = (jobfam << 16) | stepid;
/* store our name in the opal_proc_t so that
* debug messages will make sense - an upper
* layer will eventually overwrite it, but that
* won't do any harm */
s2_pname.jid = s2_jobid;
s2_pname.vid = s2_rank;
opal_proc_set_name((opal_process_name_t*)&s2_pname);
opal_output_verbose(2, opal_pmix_base_framework.framework_output,
"%s pmix:s2: assigned tmp name",
OPAL_NAME_PRINT(*(opal_process_name_t*)&s2_pname));
char *pmapping = (char*)malloc(PMI2_MAX_VALLEN);
if( pmapping == NULL ){
rc = OPAL_ERR_OUT_OF_RESOURCE;
OPAL_ERROR_LOG(rc);
return rc;
}
rc = PMI2_Info_GetJobAttr("PMI_process_mapping", pmapping, PMI2_MAX_VALLEN, &found);
if( !found || PMI2_SUCCESS != rc ) {
OPAL_PMI_ERROR(rc,"PMI2_Info_GetJobAttr");
return OPAL_ERROR;
}
s2_lranks = mca_common_pmi2_parse_pmap(pmapping, s2_pname.vid, &my_node, &s2_nlranks);
if (NULL == s2_lranks) {
rc = OPAL_ERR_OUT_OF_RESOURCE;
OPAL_ERROR_LOG(rc);
return rc;
}
free(pmapping);
/* find ourselves */
for (i=0; i < s2_nlranks; i++) {
if (s2_rank == s2_lranks[i]) {
s2_lrank = i;
s2_nrank = my_node;
break;
}
}
return OPAL_SUCCESS;
err_exit:
PMI2_Finalize();
return ret;
}
static int s2_fini(void) {
if (0 == pmix_init_count) {
return OPAL_SUCCESS;
}
if (0 == --pmix_init_count) {
PMI2_Finalize();
}
if (NULL != pmix_kvs_name) {
free(pmix_kvs_name);
pmix_kvs_name = NULL;
}
if (NULL != s2_lranks) {
free(s2_lranks);
}
return OPAL_SUCCESS;
}
static bool s2_initialized(void)
{
if (0 < pmix_init_count) {
return true;
}
return false;
}
static int s2_abort(int flag, const char msg[])
{
PMI2_Abort(flag, msg);
return OPAL_SUCCESS;
}
static int s2_spawn(int count, const char * cmds[],
int argcs[], const char ** argvs[],
const int maxprocs[],
opal_list_t *info_keyval_vector,
opal_list_t *preput_keyval_vector,
char jobId[], int jobIdSize,
int errors[])
{
/*
int rc;
size_t preput_vector_size;
const int info_keyval_sizes[1];
info_keyval_sizes[0] = (int)opal_list_get_size(info_keyval_vector);
//FIXME what's the size of array of lists?
preput_vector_size = opal_list_get_size(preput_keyval_vector);
rc = PMI2_Job_Spawn(count, cmds, argcs, argvs, maxprocs, info_keyval_sizes, info_keyval_vector, (int)preput_vector_size, preput_keyval_vector, jobId, jobIdSize, errors);
if( PMI2_SUCCESS != rc ) {
OPAL_PMI_ERROR(rc, "PMI2_Job_Spawn");
return OPAL_ERROR;
}*/
return OPAL_ERR_NOT_IMPLEMENTED;
}
static int s2_job_connect(const char jobId[])
{
int rc;
PMI2_Connect_comm_t *conn;
/*FIXME should change function prototype to add void* conn */
rc = PMI2_Job_Connect(jobId, conn);
if( PMI2_SUCCESS != rc ){
OPAL_PMI_ERROR(rc, "PMI2_Job_Connect");
return OPAL_ERROR;
}
return OPAL_SUCCESS;
}
static int s2_job_disconnect(const char jobId[])
{
int rc;
rc = PMI2_Job_Disconnect(jobId);
if( PMI2_SUCCESS != rc ){
OPAL_PMI_ERROR(rc, "PMI2_Job_Disconnect");
return OPAL_ERROR;
}
return OPAL_SUCCESS;
}
static int s2_put(opal_pmix_scope_t scope,
opal_value_t *kv)
{
int rc;
char* buffer_to_put;
int rem_offset = 0;
int data_to_put = 0;
opal_output_verbose(2, opal_pmix_base_framework.framework_output,
"%s pmix:s2 put for key %s",
OPAL_NAME_PRINT(OPAL_PROC_MY_NAME), kv->key);
if (OPAL_SUCCESS != (rc = opal_pmix_base_store_encoded (kv->key, (void*)&kv->data, kv->type, &pmix_packed_data, &pmix_packed_data_offset))) {
OPAL_ERROR_LOG(rc);
return rc;
}
if (pmix_packed_data_offset == 0) {
/* nothing to write */
return OPAL_SUCCESS;
}
if (pmix_packed_data_offset < pmix_vallen_max) {
/* this meta-key is still being filled,
* nothing to put yet
*/
return OPAL_SUCCESS;
}
/* encode only full filled meta keys */
rem_offset = pmix_packed_data_offset % pmix_vallen_max;
data_to_put = pmix_packed_data_offset - rem_offset;
buffer_to_put = (char*)malloc(data_to_put);
memcpy(buffer_to_put, pmix_packed_data, data_to_put);
opal_pmix_base_commit_packed (buffer_to_put, data_to_put, pmix_vallen_max, &pmix_pack_key, kvs_put);
free(buffer_to_put);
pmix_packed_data_offset = rem_offset;
if (0 == pmix_packed_data_offset) {
free(pmix_packed_data);
pmix_packed_data = NULL;
} else {
memmove (pmix_packed_data, pmix_packed_data + data_to_put, pmix_packed_data_offset);
pmix_packed_data = realloc (pmix_packed_data, pmix_packed_data_offset);
}
return rc;
}
static int s2_fence(opal_process_name_t *procs, size_t nprocs)
{
int rc;
int32_t i;
opal_value_t *kp, kvn;
opal_hwloc_locality_t locality;
opal_output_verbose(2, opal_pmix_base_framework.framework_output,
"%s pmix:s2 called fence",
OPAL_NAME_PRINT(OPAL_PROC_MY_NAME));
/* check if there is partially filled meta key and put them */
if (0 != pmix_packed_data_offset && NULL != pmix_packed_data) {
opal_pmix_base_commit_packed(pmix_packed_data, pmix_packed_data_offset, pmix_vallen_max, &pmix_pack_key, kvs_put);
pmix_packed_data_offset = 0;
free(pmix_packed_data);
pmix_packed_data = NULL;
}
if (PMI2_SUCCESS != (rc = PMI2_KVS_Fence())) {
OPAL_PMI_ERROR(rc, "PMI2_KVS_Fence");
return OPAL_ERROR;
}
opal_output_verbose(2, opal_pmix_base_framework.framework_output,
"%s pmix:s2 kvs_fence complete",
OPAL_NAME_PRINT(OPAL_PROC_MY_NAME));
/* get the modex data from each local process and set the
* localities to avoid having the MPI layer fetch data
* for every process in the job */
if (!got_modex_data) {
got_modex_data = true;
/* we only need to set locality for each local rank as "not found"
* equates to "non-local" */
for (i=0; i < s2_nlranks; i++) {
s2_pname.vid = i;
rc = opal_pmix_base_cache_keys_locally((opal_identifier_t*)&s2_pname, OPAL_DSTORE_CPUSET,
&kp, pmix_kvs_name, pmix_vallen_max, kvs_get);
if (OPAL_SUCCESS != rc) {
OPAL_ERROR_LOG(rc);
return rc;
}
#if OPAL_HAVE_HWLOC
if (NULL == kp || NULL == kp->data.string) {
/* if we share a node, but we don't know anything more, then
* mark us as on the node as this is all we know
*/
locality = OPAL_PROC_ON_CLUSTER | OPAL_PROC_ON_CU | OPAL_PROC_ON_NODE;
} else {
/* determine relative location on our node */
locality = opal_hwloc_base_get_relative_locality(opal_hwloc_topology,
opal_process_info.cpuset,
kp->data.string);
}
if (NULL != kp) {
OBJ_RELEASE(kp);
}
#else
/* all we know is we share a node */
locality = OPAL_PROC_ON_CLUSTER | OPAL_PROC_ON_CU | OPAL_PROC_ON_NODE;
#endif
OPAL_OUTPUT_VERBOSE((1, opal_pmix_base_framework.framework_output,
"%s pmix:s2 proc %s locality %s",
OPAL_NAME_PRINT(OPAL_PROC_MY_NAME),
OPAL_NAME_PRINT(*(opal_identifier_t*)&s2_pname),
opal_hwloc_base_print_locality(locality)));
OBJ_CONSTRUCT(&kvn, opal_value_t);
kvn.key = strdup(OPAL_DSTORE_LOCALITY);
kvn.type = OPAL_UINT16;
kvn.data.uint16 = locality;
(void)opal_dstore.store(opal_dstore_internal, (opal_identifier_t*)&s2_pname, &kvn);
OBJ_DESTRUCT(&kvn);
}
}
return OPAL_SUCCESS;
}
static int s2_get(const opal_identifier_t *id,
const char *key,
opal_value_t **kv)
{
int rc;
rc = opal_pmix_base_cache_keys_locally(id, key, kv, pmix_kvs_name, pmix_vallen_max, kvs_get);
if (NULL == *kv) {
return OPAL_ERROR;
}
return rc;
}
static int s2_publish(const char service_name[],
opal_list_t *info,
const char port[])
{
int rc;
if (PMI2_SUCCESS != (rc = PMI2_Nameserv_publish(service_name, NULL, port))) {
OPAL_PMI_ERROR(rc, "PMI2_Nameserv_publish");
return OPAL_ERROR;
}
return OPAL_SUCCESS;
}
static int s2_lookup(const char service_name[],
opal_list_t *info,
char port[], int portLen)
{
int rc;
if (PMI2_SUCCESS != (rc = PMI2_Nameserv_lookup(service_name, NULL, port, portLen))) {
OPAL_PMI_ERROR(rc, "PMI2_Nameserv_lookup");
return OPAL_ERROR;
}
return OPAL_SUCCESS;
}
static int s2_unpublish(const char service_name[],
opal_list_t *info)
{
int rc;
if (PMI2_SUCCESS != (rc = PMI2_Nameserv_unpublish(service_name, NULL))) {
OPAL_PMI_ERROR(rc, "PMI2_Nameserv_unpublish");
return OPAL_ERROR;
}
return OPAL_SUCCESS;;
}
static bool s2_get_attr(const char *attr, opal_value_t **kv)
{
opal_value_t *kp;
if (0 == strcmp(PMIX_JOBID, attr)) {
kp = OBJ_NEW(opal_value_t);
kp->key = strdup(attr);
kp->type = OPAL_UINT32;
kp->data.uint32 = s2_jobid;
*kv = kp;
return true;
}
if (0 == strcmp(PMIX_RANK, attr)) {
kp = OBJ_NEW(opal_value_t);
kp->key = strdup(attr);
kp->type = OPAL_UINT32;
kp->data.uint32 = s2_rank;
*kv = kp;
return true;
}
if (0 == strcmp(PMIX_UNIV_SIZE, attr)) {
kp = OBJ_NEW(opal_value_t);
kp->key = strdup(attr);
kp->type = OPAL_UINT32;
kp->data.uint32 = s2_usize;
*kv = kp;
return true;
}
if (0 == strcmp(PMIX_JOB_SIZE, attr)) {
kp = OBJ_NEW(opal_value_t);
kp->key = strdup(attr);
kp->type = OPAL_UINT32;
kp->data.uint32 = s2_jsize;
*kv = kp;
return true;
}
if (0 == strcmp(PMIX_APPNUM, attr)) {
kp = OBJ_NEW(opal_value_t);
kp->key = strdup(attr);
kp->type = OPAL_UINT32;
kp->data.uint32 = s2_appnum;
*kv = kp;
return true;
}
if (0 == strcmp(PMIX_LOCAL_RANK, attr)) {
kp = OBJ_NEW(opal_value_t);
kp->key = strdup(attr);
kp->type = OPAL_UINT32;
kp->data.uint32 = s2_lrank;
*kv = kp;
return true;
}
if (0 == strcmp(PMIX_NODE_RANK, attr)) {
kp = OBJ_NEW(opal_value_t);
kp->key = strdup(attr);
kp->type = OPAL_UINT32;
kp->data.uint32 = s2_nrank;
*kv = kp;
return true;
}
return false;
}
static char* pmix_error(int pmix_err)
{
char * err_msg;
switch(pmix_err) {
case PMI2_FAIL: err_msg = "Operation failed"; break;
case PMI2_ERR_INIT: err_msg = "PMI is not initialized"; break;
case PMI2_ERR_NOMEM: err_msg = "Input buffer not large enough"; break;
case PMI2_ERR_INVALID_ARG: err_msg = "Invalid argument"; break;
case PMI2_ERR_INVALID_KEY: err_msg = "Invalid key argument"; break;
case PMI2_ERR_INVALID_KEY_LENGTH: err_msg = "Invalid key length argument"; break;
case PMI2_ERR_INVALID_VAL: err_msg = "Invalid value argument"; break;
case PMI2_ERR_INVALID_VAL_LENGTH: err_msg = "Invalid value length argument"; break;
case PMI2_ERR_INVALID_LENGTH: err_msg = "Invalid length argument"; break;
case PMI2_ERR_INVALID_NUM_ARGS: err_msg = "Invalid number of arguments"; break;
case PMI2_ERR_INVALID_ARGS: err_msg = "Invalid args argument"; break;
case PMI2_ERR_INVALID_NUM_PARSED: err_msg = "Invalid num_parsed length argument"; break;
case PMI2_ERR_INVALID_KEYVALP: err_msg = "Invalid keyvalp argument"; break;
case PMI2_ERR_INVALID_SIZE: err_msg = "Invalid size argument"; break;
case PMI2_SUCCESS: err_msg = "Success"; break;
default: err_msg = "Unkown error";
}
return err_msg;
}