static int get_wireup_info(opal_buffer_t *buf)
{
int rc;
int i;
orte_routed_jobfam_t *jfam;
if (ORTE_PROC_IS_HNP) {
/* if we are not using static ports, then we need to share the
* comm info - otherwise, just return
*/
if (orte_static_ports) {
return ORTE_SUCCESS;
}
if (ORTE_SUCCESS != (rc = orte_rml_base_get_contact_info(ORTE_PROC_MY_NAME->jobid, buf))) {
ORTE_ERROR_LOG(rc);
}
return rc;
}
/* if I am an application, this is occurring during connect_accept.
* We need to return the stored information of other HNPs we
* know about, if any
*/
if (ORTE_PROC_IS_APP) {
for (i=0; i < orte_routed_jobfams.size; i++) {
if (NULL != (jfam = (orte_routed_jobfam_t*)opal_pointer_array_get_item(&orte_routed_jobfams, i))) {
opal_dss.pack(buf, &(jfam->hnp_uri), 1, OPAL_STRING);
}
}
return ORTE_SUCCESS;
}
return ORTE_SUCCESS;
}
static int get_wireup_info(opal_buffer_t *buf)
{
int rc;
/* if I am anything other than the HNP, this
* is a meaningless command as I cannot get
* the requested info
*/
if (!ORTE_PROC_IS_HNP) {
return ORTE_ERR_NOT_SUPPORTED;
}
/* if we are not using static ports, then we need to share the
* comm info - otherwise, just return
*/
if (orte_static_ports) {
return ORTE_SUCCESS;
}
if (ORTE_SUCCESS != (rc = orte_rml_base_get_contact_info(ORTE_PROC_MY_NAME->jobid, buf))) {
ORTE_ERROR_LOG(rc);
OBJ_RELEASE(buf);
return rc;
}
return ORTE_SUCCESS;
}
static int get_wireup_info(opal_buffer_t *buf)
{
int rc;
if (ORTE_PROC_IS_HNP) {
if (ORTE_SUCCESS != (rc = orte_rml_base_get_contact_info(ORTE_PROC_MY_NAME->jobid, buf))) {
ORTE_ERROR_LOG(rc);
}
return rc;
}
return ORTE_SUCCESS;
}
static void vm_ready(int fd, short args, void *cbdata)
{
orte_state_caddy_t *caddy = (orte_state_caddy_t*)cbdata;
int rc;
opal_buffer_t *buf;
orte_daemon_cmd_flag_t command = ORTE_DAEMON_DVM_NIDMAP_CMD;
orte_grpcomm_signature_t *sig;
opal_buffer_t *wireup;
opal_byte_object_t bo, *boptr;
int8_t flag;
int32_t numbytes;
/* if this is my job, then we are done */
if (ORTE_PROC_MY_NAME->jobid == caddy->jdata->jobid) {
/* send the daemon map to every daemon in this DVM - we
* do this here so we don't have to do it for every
* job we are going to launch */
buf = OBJ_NEW(opal_buffer_t);
/* pack the "load nidmap" cmd */
if (ORTE_SUCCESS != (rc = opal_dss.pack(buf, &command, 1, ORTE_DAEMON_CMD))) {
ORTE_ERROR_LOG(rc);
OBJ_RELEASE(buf);
return;
}
/* flag that daemons were launched so we will update the nidmap */
flag = 1;
opal_dss.pack(buf, &flag, 1, OPAL_INT8);
/* construct a nodemap with everything in it */
if (ORTE_SUCCESS != (rc = orte_util_encode_nodemap(buf))) {
ORTE_ERROR_LOG(rc);
OBJ_RELEASE(buf);
return;
}
if (!orte_static_ports && !orte_fwd_mpirun_port) {
/* pack a flag indicating wiring info is provided */
flag = 1;
opal_dss.pack(buf, &flag, 1, OPAL_INT8);
/* get wireup info for daemons per the selected routing module */
wireup = OBJ_NEW(opal_buffer_t);
if (ORTE_SUCCESS != (rc = orte_rml_base_get_contact_info(ORTE_PROC_MY_NAME->jobid, wireup))) {
ORTE_ERROR_LOG(rc);
OBJ_RELEASE(wireup);
OBJ_RELEASE(buf);
return;
}
/* put it in a byte object for xmission */
opal_dss.unload(wireup, (void**)&bo.bytes, &numbytes);
/* pack the byte object - zero-byte objects are fine */
bo.size = numbytes;
boptr = &bo;
if (ORTE_SUCCESS != (rc = opal_dss.pack(buf, &boptr, 1, OPAL_BYTE_OBJECT))) {
ORTE_ERROR_LOG(rc);
OBJ_RELEASE(wireup);
OBJ_RELEASE(buf);
return;
}
/* release the data since it has now been copied into our buffer */
if (NULL != bo.bytes) {
free(bo.bytes);
}
OBJ_RELEASE(wireup);
} else {
flag = 0;
opal_dss.pack(buf, &flag, 1, OPAL_INT8);
}
/* goes to all daemons */
sig = OBJ_NEW(orte_grpcomm_signature_t);
sig->signature = (orte_process_name_t*)malloc(sizeof(orte_process_name_t));
sig->signature[0].jobid = ORTE_PROC_MY_NAME->jobid;
sig->signature[0].vpid = ORTE_VPID_WILDCARD;
if (ORTE_SUCCESS != (rc = orte_grpcomm.xcast(sig, ORTE_RML_TAG_DAEMON, buf))) {
ORTE_ERROR_LOG(rc);
OBJ_RELEASE(buf);
OBJ_RELEASE(sig);
ORTE_FORCED_TERMINATE(ORTE_ERROR_DEFAULT_EXIT_CODE);
return;
}
OBJ_RELEASE(buf);
/* notify that the vm is ready */
fprintf(stdout, "DVM ready\n");
OBJ_RELEASE(caddy);
return;
}
/* progress the job */
caddy->jdata->state = ORTE_JOB_STATE_VM_READY;
/* position any required files */
if (ORTE_SUCCESS != orte_filem.preposition_files(caddy->jdata, files_ready, caddy->jdata)) {
ORTE_FORCED_TERMINATE(ORTE_ERROR_DEFAULT_EXIT_CODE);
}
/* cleanup */
OBJ_RELEASE(caddy);
}
