1
1
openmpi/orte/mca/rmgr/base/rmgr_base_unpack.c

284 строки
7.0 KiB
C
Исходник Обычный вид История

/*
* 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.
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
* Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
* University of Stuttgart. All rights reserved.
* $COPYRIGHT$
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
*
* Additional copyrights may follow
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
*
* $HEADER$
*/
#include "orte_config.h"
#include <errno.h>
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif
#include <string.h>
#include "orte/orte_constants.h"
#include "opal/util/output.h"
#include "opal/util/trace.h"
#include "orte/dss/dss.h"
#include "orte/mca/rmgr/base/base.h"
#include "orte/mca/errmgr/errmgr.h"
/*
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
*
*/
static int orte_rmgr_base_cmd_query(orte_buffer_t* req, orte_buffer_t* rsp)
{
int32_t rc = orte_rmgr.query();
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
static int orte_rmgr_base_cmd_create(orte_buffer_t* req, orte_buffer_t* rsp)
{
int rc;
int32_t ret;
orte_app_context_t** context;
orte_jobid_t jobid;
orte_std_cntr_t i, cnt, num_context;
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
cnt = 1;
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &num_context, &cnt, ORTE_STD_CNTR))) {
ORTE_ERROR_LOG(rc);
return rc;
}
if(NULL == (context = malloc(sizeof(orte_app_context_t*)*num_context))) {
ORTE_ERROR_LOG(ORTE_ERR_OUT_OF_RESOURCE);
return ORTE_ERR_OUT_OF_RESOURCE;
}
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
cnt = num_context;
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, context, &cnt, ORTE_APP_CONTEXT))) {
ORTE_ERROR_LOG(rc);
free(context);
return rc;
}
ret = orte_rmgr.create(context, num_context, &jobid);
ret = orte_rmgr_base_proc_stage_gate_init(jobid);
if(ORTE_SUCCESS != (rc = orte_dss.pack(rsp, &jobid, 1, ORTE_JOBID))) {
ORTE_ERROR_LOG(rc);
goto cleanup;
}
if(ORTE_SUCCESS != (rc = orte_dss.pack(rsp, &ret, 1, ORTE_INT32))) {
ORTE_ERROR_LOG(rc);
goto cleanup;
}
cleanup:
for(i=0; i<num_context; i++) {
OBJ_RELEASE(context[i]);
}
free(context);
return ret;
}
static int orte_rmgr_base_cmd_allocate(orte_buffer_t* req, orte_buffer_t* rsp)
{
int32_t rc;
orte_jobid_t jobid;
orte_std_cntr_t cnt = 1;
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &jobid, &cnt, ORTE_JOBID))) {
ORTE_ERROR_LOG(rc);
} else {
rc = orte_rmgr.allocate(jobid);
}
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
static int orte_rmgr_base_cmd_deallocate(orte_buffer_t* req, orte_buffer_t* rsp)
{
int32_t rc;
orte_jobid_t jobid;
orte_std_cntr_t cnt = 1;
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &jobid, &cnt, ORTE_JOBID))) {
ORTE_ERROR_LOG(rc);
} else {
rc = orte_rmgr.deallocate(jobid);
}
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
static int orte_rmgr_base_cmd_map(orte_buffer_t* req, orte_buffer_t* rsp)
{
int rc;
orte_jobid_t jobid;
orte_std_cntr_t cnt = 1;
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &jobid, &cnt, ORTE_JOBID))) {
ORTE_ERROR_LOG(rc);
} else {
rc = orte_rmgr.map(jobid);
}
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
static int orte_rmgr_base_cmd_launch(orte_buffer_t* req, orte_buffer_t* rsp)
{
int rc;
orte_jobid_t jobid;
orte_std_cntr_t cnt = 1;
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &jobid, &cnt, ORTE_JOBID))) {
ORTE_ERROR_LOG(rc);
} else {
rc = orte_rmgr.launch(jobid);
}
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
static int orte_rmgr_base_cmd_term_job(orte_buffer_t* req, orte_buffer_t* rsp)
{
int rc;
orte_jobid_t jobid;
orte_std_cntr_t cnt = 1;
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &jobid, &cnt, ORTE_JOBID))) {
ORTE_ERROR_LOG(rc);
} else {
rc = orte_rmgr.terminate_job(jobid);
}
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
static int orte_rmgr_base_cmd_term_proc(orte_buffer_t* req, orte_buffer_t* rsp)
{
int rc;
orte_process_name_t name;
orte_std_cntr_t cnt = 1;
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &name, &cnt, ORTE_NAME))) {
ORTE_ERROR_LOG(rc);
} else {
rc = orte_rmgr.terminate_proc(&name);
}
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
static int orte_rmgr_base_cmd_signal_job(orte_buffer_t* req, orte_buffer_t* rsp)
{
int rc;
orte_jobid_t jobid;
orte_std_cntr_t cnt = 1;
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
int32_t signal;
OPAL_TRACE(4);
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &jobid, &cnt, ORTE_JOBID))) {
ORTE_ERROR_LOG(rc);
return rc;
}
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &signal, &cnt, ORTE_INT32))) {
ORTE_ERROR_LOG(rc);
return rc;
}
rc = orte_rmgr.signal_job(jobid, signal);
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
static int orte_rmgr_base_cmd_signal_proc(orte_buffer_t* req, orte_buffer_t* rsp)
{
int rc;
orte_process_name_t name;
orte_std_cntr_t cnt = 1;
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
int32_t signal;
OPAL_TRACE(4);
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &name, &cnt, ORTE_NAME))) {
ORTE_ERROR_LOG(rc);
}
if(ORTE_SUCCESS != (rc = orte_dss.unpack(req, &signal, &cnt, ORTE_INT32))) {
ORTE_ERROR_LOG(rc);
return rc;
}
rc = orte_rmgr.signal_proc(&name, signal);
return orte_dss.pack(rsp, &rc, 1, ORTE_INT32);
}
int orte_rmgr_base_cmd_dispatch(orte_buffer_t* req, orte_buffer_t* rsp)
{
orte_rmgr_cmd_t cmd;
orte_std_cntr_t cnt = 1;
int rc;
OPAL_TRACE(4);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
rc = orte_dss.unpack(req, &cmd, &cnt, ORTE_RMGR_CMD);
if(ORTE_SUCCESS != rc) {
ORTE_ERROR_LOG(rc);
return rc;
}
switch(cmd) {
case ORTE_RMGR_CMD_QUERY:
return orte_rmgr_base_cmd_query(req,rsp);
case ORTE_RMGR_CMD_CREATE:
return orte_rmgr_base_cmd_create(req,rsp);
case ORTE_RMGR_CMD_ALLOCATE:
return orte_rmgr_base_cmd_allocate(req,rsp);
case ORTE_RMGR_CMD_DEALLOCATE:
return orte_rmgr_base_cmd_deallocate(req,rsp);
case ORTE_RMGR_CMD_MAP:
return orte_rmgr_base_cmd_map(req,rsp);
case ORTE_RMGR_CMD_LAUNCH:
return orte_rmgr_base_cmd_launch(req,rsp);
case ORTE_RMGR_CMD_TERM_JOB:
return orte_rmgr_base_cmd_term_job(req,rsp);
case ORTE_RMGR_CMD_TERM_PROC:
return orte_rmgr_base_cmd_term_proc(req,rsp);
Add ability to trap and propagate SIGUSR1/2 to remote processes. There are a number of small changes that hit a bunch of files: 1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc. 2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X. 3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job. 4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality. I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun. Ralph This commit was SVN r10258.
2006-06-08 18:27:17 +00:00
case ORTE_RMGR_CMD_SIGNAL_JOB:
return orte_rmgr_base_cmd_signal_job(req,rsp);
case ORTE_RMGR_CMD_SIGNAL_PROC:
return orte_rmgr_base_cmd_signal_proc(req,rsp);
default:
ORTE_ERROR_LOG(ORTE_ERR_BAD_PARAM);
return ORTE_ERR_BAD_PARAM;
}
}