ba5498cdc6
1. repair of the linear and direct routed modules 2. repair of the ompi/pubsub/orte module to correctly init routes to the ompi-server, and correctly handle failure to correctly parse the provided ompi-server URI 3. modification of orterun to accept both "file" and "FILE" for designating where the ompi-server URI is to be found - purely a convenience feature 4. resolution of a message ordering problem during the connect/accept handshake that allowed the "send-first" proc to attempt to send to the "recv-first" proc before the HNP had actually updated its routes. Let this be a further reminder to all - message ordering is NOT guaranteed in the OOB 5. Repair the ompi/dpm/orte module to correctly init routes during connect/accept. Reminder to all: messages sent to procs in another job family (i.e., started by a different mpirun) are ALWAYS routed through the respective HNPs. As per the comments in orte/routed, this is REQUIRED to maintain connect/accept (where only the root proc on each side is capable of init'ing the routes), allow communication between mpirun's using different routing modules, and to minimize connections on tools such as ompi-server. It is all taken care of "under the covers" by the OOB to ensure that a route back to the sender is maintained, even when the different mpirun's are using different routed modules. 6. corrections in the orte/odls to ensure proper identification of daemons participating in a dynamic launch 7. corrections in build/nidmap to support update of an existing nidmap during dynamic launch 8. corrected implementation of the update_arch function in the ESS, along with consolidation of a number of ESS operations into base functions for easier maintenance. The ability to support info from multiple jobs was added, although we don't currently do so - this will come later to support further fault recovery strategies 9. minor updates to several functions to remove unnecessary and/or no longer used variables and envar's, add some debugging output, etc. 10. addition of a new macro ORTE_PROC_IS_DAEMON that resolves to true if the provided proc is a daemon There is still more cleanup to be done for efficiency, but this at least works. Tested on single-node Mac, multi-node SLURM via odin. Tests included connect/accept, publish/lookup/unpublish, comm_spawn, comm_spawn_multiple, and singleton comm_spawn. Fixes ticket #1256 This commit was SVN r18804.
120 строки
4.5 KiB
C
120 строки
4.5 KiB
C
/*
|
|
* Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
|
|
* University Research and Technology
|
|
* Corporation. All rights reserved.
|
|
* Copyright (c) 2004-2005 The University of Tennessee and The University
|
|
* of Tennessee Research Foundation. All rights
|
|
* reserved.
|
|
* Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
|
|
* University of Stuttgart. All rights reserved.
|
|
* Copyright (c) 2004-2005 The Regents of the University of California.
|
|
* All rights reserved.
|
|
* $COPYRIGHT$
|
|
*
|
|
* Additional copyrights may follow
|
|
*
|
|
* $HEADER$
|
|
*/
|
|
|
|
/** @file:
|
|
*
|
|
* Populates global structure with process-specific information.
|
|
*
|
|
*
|
|
*/
|
|
|
|
#ifndef _ORTE_PROC_INFO_H_
|
|
#define _ORTE_PROC_INFO_H_
|
|
|
|
#include "orte_config.h"
|
|
#include "orte/types.h"
|
|
|
|
#ifdef HAVE_SYS_TYPES_H
|
|
#include <sys/types.h>
|
|
#endif
|
|
|
|
#include "opal/dss/dss_types.h"
|
|
|
|
BEGIN_C_DECLS
|
|
|
|
/**
|
|
* Process information structure
|
|
*
|
|
* The orte_proc_info() function fills the pid field and obtains the
|
|
* process name, storing that information in the global structure. The
|
|
* structure also holds path names to the universe, job, and process
|
|
* session directories, and to the stdin, stdout, and stderr temp
|
|
* files - however, these are all initialized elsewhere.
|
|
*/
|
|
struct orte_proc_info_t {
|
|
orte_process_name_t my_name; /**< My official process name */
|
|
orte_process_name_t my_daemon; /**< Name of my local daemon */
|
|
char *my_daemon_uri; /**< Contact info to local daemon */
|
|
orte_process_name_t my_hnp; /**< Name of my hnp */
|
|
char *my_hnp_uri; /**< Contact info for my hnp */
|
|
pid_t hnp_pid; /**< hnp pid - used if singleton */
|
|
orte_std_cntr_t app_num; /**< our index into the app_context array */
|
|
orte_vpid_t num_procs; /**< number of processes in this job */
|
|
char *nodename; /**< string name for this node */
|
|
uint32_t arch; /**< arch for this node */
|
|
pid_t pid; /**< Local process ID for this process */
|
|
bool singleton; /**< I am a singleton */
|
|
bool daemon; /**< Indicate whether or not I am a daemon */
|
|
bool hnp; /**< Indicate whether or not I am the HNP (orterun) */
|
|
bool tool; /**< I am a tool or not */
|
|
bool mpi_proc; /**< I am an MPI process */
|
|
opal_buffer_t *sync_buf; /**< buffer to store sync response */
|
|
/* The session directory has the form
|
|
* <prefix>/<openmpi-sessions-user>/<jobid>/<procid>, where the prefix
|
|
* can either be provided by the user via the
|
|
* --tmpdir command-line flag, the use of one of several
|
|
* environmental variables, or else a default location.
|
|
*/
|
|
char *tmpdir_base; /**< Base directory of the session dir tree */
|
|
char *top_session_dir; /**< Top-most directory of the session tree */
|
|
char *job_session_dir; /**< Session directory for job */
|
|
char *proc_session_dir; /**< Session directory for the process */
|
|
|
|
char *sock_stdin; /**< Path name to temp file for stdin. */
|
|
char *sock_stdout; /**< Path name to temp file for stdout. */
|
|
char *sock_stderr; /**< Path name to temp file for stderr. */
|
|
};
|
|
typedef struct orte_proc_info_t orte_proc_info_t;
|
|
|
|
|
|
/**
|
|
*
|
|
* Global process info descriptor. Initialized to almost no
|
|
* meaningful information - data is provided by calling \c
|
|
* orte_rte_init() (which calls \c orte_proc_info() to fill in the
|
|
* structure).
|
|
*
|
|
* The exception to this rule is the \c orte_process_info.seed field,
|
|
* which will be initialized to \c false, but should be set to \c true
|
|
* before calling \c orte_rte_info() if the caller is a seed daemon.
|
|
*/
|
|
ORTE_DECLSPEC extern orte_proc_info_t orte_process_info;
|
|
|
|
|
|
/**
|
|
* \internal
|
|
*
|
|
* Global structure to store a wide range of information about the
|
|
* process. orte_proc_info populates a global variable with
|
|
* information about the process being executing. This function should
|
|
* be called only once, from orte_rte_init().
|
|
*
|
|
* @param None.
|
|
*
|
|
* @retval ORTE_SUCCESS Successfully initialized the various fields.
|
|
* @retval OMPI_ERROR Failed to initialize one or more fields.
|
|
*/
|
|
|
|
ORTE_DECLSPEC int orte_proc_info(void);
|
|
|
|
ORTE_DECLSPEC int orte_proc_info_finalize(void);
|
|
|
|
END_C_DECLS
|
|
|
|
#endif
|