1
1
openmpi/ompi/mpi/c/intercomm_merge.c

162 строки
5.5 KiB
C
Исходник Обычный вид История

/*
* Copyright (c) 2004-2007 The Trustees of Indiana University and Indiana
* University Research and Technology
* Corporation. All rights reserved.
* Copyright (c) 2004-2008 The University of Tennessee and The University
* of Tennessee Research Foundation. All rights
* reserved.
* Copyright (c) 2004-2008 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) 2006-2012 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2006-2009 University of Houston. All rights reserved.
* $COPYRIGHT$
*
* Additional copyrights may follow
*
* $HEADER$
*/
#include "ompi_config.h"
#include <string.h>
#include "ompi/mpi/c/bindings.h"
#include "ompi/runtime/params.h"
#include "ompi/errhandler/errhandler.h"
#include "ompi/communicator/communicator.h"
#include "ompi/proc/proc.h"
#include "ompi/memchecker.h"
#if OPAL_HAVE_WEAK_SYMBOLS && OMPI_PROFILING_DEFINES
#pragma weak MPI_Intercomm_merge = PMPI_Intercomm_merge
#endif
#if OMPI_PROFILING_DEFINES
#include "ompi/mpi/c/profile/defines.h"
#endif
static const char FUNC_NAME[] = "MPI_Intercomm_merge";
int MPI_Intercomm_merge(MPI_Comm intercomm, int high,
MPI_Comm *newcomm)
{
ompi_communicator_t *newcomp=MPI_COMM_NULL;
ompi_proc_t **procs=NULL;
int local_size, remote_size;
int first;
int total_size;
int rc=MPI_SUCCESS;
int thigh = high;
ompi_group_t *new_group_pointer;
MEMCHECKER(
memchecker_comm(intercomm);
);
if ( MPI_PARAM_CHECK ) {
OMPI_ERR_INIT_FINALIZE(FUNC_NAME);
if (ompi_comm_invalid ( intercomm ) ||
!( intercomm->c_flags & OMPI_COMM_INTER ) )
return OMPI_ERRHANDLER_INVOKE ( MPI_COMM_WORLD, MPI_ERR_COMM,
FUNC_NAME);
if ( NULL == newcomm )
return OMPI_ERRHANDLER_INVOKE ( intercomm, MPI_ERR_ARG,
FUNC_NAME);
}
OPAL_CR_ENTER_LIBRARY();
local_size = ompi_comm_size ( intercomm );
remote_size = ompi_comm_remote_size ( intercomm );
total_size = local_size + remote_size;
procs = (ompi_proc_t **) malloc ( total_size * sizeof(ompi_proc_t *));
if ( NULL == procs ) {
rc = MPI_ERR_INTERN;
goto exit;
}
first = ompi_comm_determine_first ( intercomm, thigh );
if ( MPI_UNDEFINED == first ) {
rc = MPI_ERR_INTERN;
goto exit;
}
if ( first ) {
ompi_group_union ( intercomm->c_local_group, intercomm->c_remote_group, &new_group_pointer );
}
else {
ompi_group_union ( intercomm->c_remote_group, intercomm->c_local_group, &new_group_pointer );
}
rc = ompi_comm_set ( &newcomp, /* new comm */
intercomm, /* old comm */
total_size, /* local_size */
NULL, /* local_procs*/
0, /* remote_size */
NULL, /* remote_procs */
NULL, /* attrs */
intercomm->error_handler, /* error handler*/
false, /* don't copy the topo */
new_group_pointer, /* local group */
NULL /* remote group */
);
if ( NULL == newcomp ) {
rc = MPI_ERR_INTERN;
goto exit;
}
if ( MPI_SUCCESS != rc ) {
goto exit;
}
ompi_group_decrement_proc_count(new_group_pointer);
OBJ_RELEASE(new_group_pointer);
new_group_pointer = MPI_GROUP_NULL;
/* Determine context id. It is identical to f_2_c_handle */
rc = ompi_comm_nextcid ( newcomp, /* new comm */
intercomm, /* old comm */
NULL, /* bridge comm */
NULL, /* local leader */
NULL, /* remote_leader */
OMPI_COMM_CID_INTER, /* mode */
-1 ); /* send_first */
if ( OMPI_SUCCESS != rc ) {
goto exit;
}
/* activate communicator and init coll-module */
so here is what happens: in the v1.2 series the cid's could never go above the max. allowed for a particular pml. Because of that, pml_add_comm never checked for the cid, and in fact pml_add_comm was called in comm_set, which is *before* we knew the cid. in the v1.3 series (and trunk) we check now the cid to detect overflow, and because of that pml_add_comm has been moved *after* the cid allocation routine, namely into the comm_activate routine. in the v1.2 series, the comm_activate contained a synchronization step of the old communicator in order to prevent incoming fragments on the new communicator, with the main problem being that the allreduce in the communicator allocation finished at different times on different processes, and thus, this scenario could and did really occur. in the v1.3 series, the comm_activate does not contain the synchronization step anymore, since we introduced the new queue for fragments with unknown cid. The problem is however, that whether a fragment is known or not is decided by using ompi_comm_lookup(), which will return something useful as soon as the cid allocation finished, even before pml_add_comm has been called. So there is a small time gap where we will not post a message into queue for unknown cid's, but we can also not look up the process structure belonging to the rank in that comm ( that is in pml_ob1_match_recv_frag or something like that). The current fix reintroduces the synchronization step in comm_activate, and ensures that no fragment can be received for a new communicator before the synchronization occurs , and thus comm_nextcid() and pml_add_comm has been called. It seems to be the safest and easiest way for now. Welcome back, v1.2. This commit was SVN r21970.
2009-09-17 18:37:02 +04:00
rc = ompi_comm_activate( &newcomp, /* new comm */
intercomm, /* old comm */
NULL, /* bridge comm */
NULL, /* local leader */
NULL, /* remote_leader */
OMPI_COMM_CID_INTER, /* mode */
-1 ); /* send_first */
if ( OMPI_SUCCESS != rc ) {
goto exit;
}
exit:
OPAL_CR_EXIT_LIBRARY();
if ( NULL != procs ) {
free ( procs );
}
if ( MPI_SUCCESS != rc ) {
if ( MPI_COMM_NULL != newcomp && NULL != newcomp ) {
OBJ_RELEASE(newcomp);
}
*newcomm = MPI_COMM_NULL;
return OMPI_ERRHANDLER_INVOKE(intercomm, rc, FUNC_NAME);
}
*newcomm = newcomp;
return MPI_SUCCESS;
}