a200e4f865
*** THIS RFC INCLUDES A MINOR CHANGE TO THE MPI-RTE INTERFACE *** Note: during the course of this work, it was necessary to completely separate the MPI and RTE progress engines. There were multiple places in the MPI layer where ORTE_WAIT_FOR_COMPLETION was being used. A new OMPI_WAIT_FOR_COMPLETION macro was created (defined in ompi/mca/rte/rte.h) that simply cycles across opal_progress until the provided flag becomes false. Places where the MPI layer blocked waiting for RTE to complete an event have been modified to use this macro. *************************************************************************************** I am reissuing this RFC because of the time that has passed since its original release. Since its initial release and review, I have debugged it further to ensure it fully supports tests like loop_spawn. It therefore seems ready for merge back to the trunk. Given its prior review, I have set the timeout for one week. The code is in https://bitbucket.org/rhc/ompi-oob2 WHAT: Rewrite of ORTE OOB WHY: Support asynchronous progress and a host of other features WHEN: Wed, August 21 SYNOPSIS: The current OOB has served us well, but a number of limitations have been identified over the years. Specifically: * it is only progressed when called via opal_progress, which can lead to hangs or recursive calls into libevent (which is not supported by that code) * we've had issues when multiple NICs are available as the code doesn't "shift" messages between transports - thus, all nodes had to be available via the same TCP interface. * the OOB "unloads" incoming opal_buffer_t objects during the transmission, thus preventing use of OBJ_RETAIN in the code when repeatedly sending the same message to multiple recipients * there is no failover mechanism across NICs - if the selected NIC (or its attached switch) fails, we are forced to abort * only one transport (i.e., component) can be "active" The revised OOB resolves these problems: * async progress is used for all application processes, with the progress thread blocking in the event library * each available TCP NIC is supported by its own TCP module. The ability to asynchronously progress each module independently is provided, but not enabled by default (a runtime MCA parameter turns it "on") * multi-address TCP NICs (e.g., a NIC with both an IPv4 and IPv6 address, or with virtual interfaces) are supported - reachability is determined by comparing the contact info for a peer against all addresses within the range covered by the address/mask pairs for the NIC. * a message that arrives on one TCP NIC is automatically shifted to whatever NIC that is connected to the next "hop" if that peer cannot be reached by the incoming NIC. If no TCP module will reach the peer, then the OOB attempts to send the message via all other available components - if none can reach the peer, then an "error" is reported back to the RML, which then calls the errmgr for instructions. * opal_buffer_t now conforms to standard object rules re OBJ_RETAIN as we no longer "unload" the incoming object * NIC failure is reported to the TCP component, which then tries to resend the message across any other available TCP NIC. If that doesn't work, then the message is given back to the OOB base to try using other components. If all that fails, then the error is reported to the RML, which reports to the errmgr for instructions * obviously from the above, multiple OOB components (e.g., TCP and UD) can be active in parallel * the matching code has been moved to the RML (and out of the OOB/TCP component) so it is independent of transport * routing is done by the individual OOB modules (as opposed to the RML). Thus, both routed and non-routed transports can simultaneously be active * all blocking send/recv APIs have been removed. Everything operates asynchronously. KNOWN LIMITATIONS: * although provision is made for component failover as described above, the code for doing so has not been fully implemented yet. At the moment, if all connections for a given peer fail, the errmgr is notified of a "lost connection", which by default results in termination of the job if it was a lifeline * the IPv6 code is present and compiles, but is not complete. Since the current IPv6 support in the OOB doesn't work anyway, I don't consider this a blocker * routing is performed at the individual module level, yet the active routed component is selected on a global basis. We probably should update that to reflect that different transports may need/choose to route in different ways * obviously, not every error path has been tested nor necessarily covered * determining abnormal termination is more challenging than in the old code as we now potentially have multiple ways of connecting to a process. Ideally, we would declare "connection failed" when *all* transports can no longer reach the process, but that requires some additional (possibly complex) code. For now, the code replicates the old behavior only somewhat modified - i.e., if a module sees its connection fail, it checks to see if it is a lifeline. If so, it notifies the errmgr that the lifeline is lost - otherwise, it notifies the errmgr that a non-lifeline connection was lost. * reachability is determined solely on the basis of a shared subnet address/mask - more sophisticated algorithms (e.g., the one used in the tcp btl) are required to handle routing via gateways * the RML needs to assign sequence numbers to each message on a per-peer basis. The receiving RML will then deliver messages in order, thus preventing out-of-order messaging in the case where messages travel across different transports or a message needs to be redirected/resent due to failure of a NIC This commit was SVN r29058.
260 строки
10 KiB
C
260 строки
10 KiB
C
/*
|
|
* Copyright (c) 2004-2008 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 (c) 2007-2013 Los Alamos National Security, LLC. All rights
|
|
* reserved.
|
|
* $COPYRIGHT$
|
|
*
|
|
* Additional copyrights may follow
|
|
*
|
|
* $HEADER$
|
|
*/
|
|
|
|
/**
|
|
* @file
|
|
*
|
|
* RML Framework maintenence interface
|
|
*
|
|
* Interface for starting / stopping / controlling the RML framework,
|
|
* as well as support for modifying RML datatypes.
|
|
*
|
|
* @note The only RML datatype exposed to the user is the RML tag.
|
|
* This will always be an integral value, so the only datatype support
|
|
* really required is the internal DSS functions for packing /
|
|
* unpacking / comparing tags. The user should never need to deal
|
|
* with these.
|
|
*/
|
|
|
|
#ifndef MCA_RML_BASE_H
|
|
#define MCA_RML_BASE_H
|
|
|
|
#include "orte_config.h"
|
|
|
|
#include "opal/dss/dss_types.h"
|
|
#include "opal/mca/mca.h"
|
|
|
|
#include "orte/runtime/orte_globals.h"
|
|
|
|
#include "orte/mca/rml/rml.h"
|
|
|
|
BEGIN_C_DECLS
|
|
|
|
|
|
/*
|
|
* MCA Framework
|
|
*/
|
|
ORTE_DECLSPEC extern mca_base_framework_t orte_rml_base_framework;
|
|
/* select a component */
|
|
ORTE_DECLSPEC int orte_rml_base_select(void);
|
|
|
|
/**
|
|
* Post receive to get updates regarding contact information
|
|
*
|
|
* Post a non-blocking receive (likely during orte_init()) to receive
|
|
* updated contact information from the HNP when it becomes available.
|
|
* This should be called in any process that needs such updates, and
|
|
* the receive will continue to get update callbacks until
|
|
* orte_rml_base_comm_stop() is called.
|
|
*/
|
|
ORTE_DECLSPEC void orte_rml_base_comm_start(void);
|
|
|
|
|
|
/**
|
|
* Stop receiving contact information updates
|
|
*
|
|
* Shut down the receive posted during orte_rml_base_comm_start(),
|
|
* likely during orte_finalize().
|
|
*/
|
|
ORTE_DECLSPEC void orte_rml_base_comm_stop(void);
|
|
|
|
|
|
/* a global struct containing framework-level values */
|
|
typedef struct {
|
|
opal_list_t posted_recvs;
|
|
opal_list_t unmatched_msgs;
|
|
} orte_rml_base_t;
|
|
ORTE_DECLSPEC extern orte_rml_base_t orte_rml_base;
|
|
|
|
|
|
/**
|
|
* List of components that are available to the RML
|
|
*
|
|
* List of components that are currently available to the RML
|
|
* framework. Useable between calls to orte_rml_base_open() and
|
|
* orte_rml_base_close().
|
|
*
|
|
* @note This list should not be used by code outside the RML base.
|
|
*/
|
|
ORTE_DECLSPEC extern opal_list_t orte_rml_base_components;
|
|
|
|
|
|
/**
|
|
* Component structure for the selected RML component
|
|
*
|
|
* Component structure pointer for the currently selected RML
|
|
* component. Useable between calls to orte_rml_base_select() and
|
|
* orte_rml_base_close().
|
|
*
|
|
* @note This pointer should not be used outside the RML base. It is
|
|
* available outside the RML base only for the F/T component.
|
|
*/
|
|
ORTE_DECLSPEC extern orte_rml_component_t *orte_rml_component;
|
|
|
|
|
|
/* structure to send RML messages - used internally */
|
|
typedef struct {
|
|
opal_list_item_t super;
|
|
orte_process_name_t peer; // targeted recipient
|
|
int status; // returned status on send
|
|
orte_rml_tag_t tag; // targeted tag
|
|
|
|
/* user's send callback functions and data */
|
|
union {
|
|
orte_rml_callback_fn_t iov;
|
|
orte_rml_buffer_callback_fn_t buffer;
|
|
} cbfunc;
|
|
void *cbdata;
|
|
|
|
/* pointer to the user's iovec array */
|
|
struct iovec *iov;
|
|
int count;
|
|
/* pointer to the user's buffer */
|
|
opal_buffer_t *buffer;
|
|
} orte_rml_send_t;
|
|
OBJ_CLASS_DECLARATION(orte_rml_send_t);
|
|
|
|
/* structure to recv RML messages - used internally */
|
|
typedef struct {
|
|
opal_list_item_t super;
|
|
opal_event_t ev;
|
|
orte_process_name_t sender; // sender
|
|
orte_rml_tag_t tag; // targeted tag
|
|
struct iovec iov; // the recvd data
|
|
} orte_rml_recv_t;
|
|
OBJ_CLASS_DECLARATION(orte_rml_recv_t);
|
|
|
|
typedef struct {
|
|
opal_list_item_t super;
|
|
bool buffer_data;
|
|
orte_process_name_t peer;
|
|
orte_rml_tag_t tag;
|
|
bool persistent;
|
|
union {
|
|
orte_rml_callback_fn_t iov;
|
|
orte_rml_buffer_callback_fn_t buffer;
|
|
} cbfunc;
|
|
void *cbdata;
|
|
} orte_rml_posted_recv_t;
|
|
OBJ_CLASS_DECLARATION(orte_rml_posted_recv_t);
|
|
|
|
/* define an object for transferring recv requests to the list of posted recvs */
|
|
typedef struct {
|
|
opal_object_t super;
|
|
opal_event_t ev;
|
|
bool cancel;
|
|
orte_rml_posted_recv_t *post;
|
|
} orte_rml_recv_request_t;
|
|
OBJ_CLASS_DECLARATION(orte_rml_recv_request_t);
|
|
|
|
#define ORTE_RML_POST_MESSAGE(p, t, b, l) \
|
|
do { \
|
|
orte_rml_recv_t *msg; \
|
|
opal_output_verbose(5, orte_rml_base_framework.framework_output, \
|
|
"%s Message posted at %s:%d", \
|
|
ORTE_NAME_PRINT(ORTE_PROC_MY_NAME), \
|
|
__FILE__, __LINE__); \
|
|
msg = OBJ_NEW(orte_rml_recv_t); \
|
|
msg->sender.jobid = (p)->jobid; \
|
|
msg->sender.vpid = (p)->vpid; \
|
|
msg->tag = (t); \
|
|
msg->iov.iov_base = (IOVBASE_TYPE*)(b); \
|
|
msg->iov.iov_len = (l); \
|
|
/* setup the event */ \
|
|
opal_event_set(orte_event_base, &msg->ev, -1, \
|
|
OPAL_EV_WRITE, \
|
|
orte_rml_base_process_msg, msg); \
|
|
opal_event_set_priority(&msg->ev, ORTE_MSG_PRI); \
|
|
opal_event_active(&msg->ev, OPAL_EV_WRITE, 1); \
|
|
} while(0);
|
|
|
|
#define ORTE_RML_ACTIVATE_MESSAGE(m) \
|
|
do { \
|
|
/* setup the event */ \
|
|
opal_event_set(orte_event_base, &(m)->ev, -1, \
|
|
OPAL_EV_WRITE, \
|
|
orte_rml_base_process_msg, (m)); \
|
|
opal_event_set_priority(&(m)->ev, ORTE_MSG_PRI); \
|
|
opal_event_active(&(m)->ev, OPAL_EV_WRITE, 1); \
|
|
} while(0);
|
|
|
|
#define ORTE_RML_SEND_COMPLETE(m) \
|
|
do { \
|
|
opal_output_verbose(5, orte_rml_base_framework.framework_output, \
|
|
"%s-%s Send message complete at %s:%d", \
|
|
ORTE_NAME_PRINT(ORTE_PROC_MY_NAME), \
|
|
ORTE_NAME_PRINT(&((m)->peer)), \
|
|
__FILE__, __LINE__); \
|
|
if (NULL != (m)->iov) { \
|
|
if (NULL != (m)->cbfunc.iov) { \
|
|
(m)->cbfunc.iov((m)->status, \
|
|
&((m)->peer), \
|
|
(m)->iov, (m)->count, \
|
|
(m)->tag, (m)->cbdata); \
|
|
} \
|
|
} else { \
|
|
/* non-blocking buffer send */ \
|
|
(m)->cbfunc.buffer((m)->status, &((m)->peer), \
|
|
(m)->buffer, \
|
|
(m)->tag, (m)->cbdata); \
|
|
} \
|
|
}while(0);
|
|
|
|
/*
|
|
* This is the base priority for a RML wrapper component
|
|
* If there exists more than one wrapper, then the one with
|
|
* the lowest priority wins.
|
|
*/
|
|
#define RML_SELECT_WRAPPER_PRIORITY -128
|
|
|
|
/* common implementations */
|
|
ORTE_DECLSPEC void orte_rml_base_post_recv(int sd, short args, void *cbdata);
|
|
ORTE_DECLSPEC void orte_rml_base_process_msg(int fd, short flags, void *cbdata);
|
|
ORTE_DECLSPEC void orte_rml_base_process_error(int fd, short flags, void *cbdata);
|
|
|
|
/* null functions */
|
|
int orte_rml_base_null_send_nb(struct orte_process_name_t* peer,
|
|
struct iovec* msg,
|
|
int count,
|
|
orte_rml_tag_t tag,
|
|
orte_rml_callback_fn_t cbfunc,
|
|
void* cbdata);
|
|
int orte_rml_base_null_send_buffer_nb(struct orte_process_name_t* peer,
|
|
struct opal_buffer_t* buffer,
|
|
orte_rml_tag_t tag,
|
|
orte_rml_buffer_callback_fn_t cbfunc,
|
|
void* cbdata);
|
|
void orte_rml_base_null_recv_nb(struct orte_process_name_t* peer,
|
|
orte_rml_tag_t tag,
|
|
bool persistent,
|
|
orte_rml_callback_fn_t cbfunc,
|
|
void* cbdata);
|
|
void orte_rml_base_null_recv_buffer_nb(struct orte_process_name_t* peer,
|
|
orte_rml_tag_t tag,
|
|
bool persistent,
|
|
orte_rml_buffer_callback_fn_t cbfunc,
|
|
void* cbdata);
|
|
void orte_rml_base_null_recv_cancel(orte_process_name_t* peer,
|
|
orte_rml_tag_t tag);
|
|
|
|
END_C_DECLS
|
|
|
|
#endif /* MCA_RML_BASE_H */
|