ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
openmpi/ompi/op/op.c
Nathan Hjelm 1a021b8f2d coll/ml: add support for blocking and non-blocking allreduce, reduce, and 
allgather.

The new collectives provide a signifigant performance increase over tuned for
small and medium messages. We are initially setting the priority lower than
tuned until this has had some time to soak in the trunk. Please set
coll_ml_priority to 90 for MTT runs.

Credit for this work goes to Manjunath Gorentla Venkata (ORNL), Pavel Shamis (ORNL),
and Nathan Hjelm (LANL).

Commit details (for reference):

Import ORNL's collectives for MPI_Allreduce, MPI_Reduce, and MPI_Allgather.

We need to take the basesmuma header into account when calculating the
ptpcoll small message thresholds. Add a define to bcol.h indicating the
maximum header size so we can take the header into account while not
making ptpcoll dependent on information from basesmuma.

This resolves an issue with allreduce where ptpcoll overwrites the
header of the next buffer in the basesmuma bank.

Fix reduce and make a sequential collective launcher in coll_ml_inlines.h

The root calculation for reduce was wrong for any root != 0. There are
four possibilities for the root:

 - The root is not the current process but is in the current hierarchy. In
   this case the root is the index of the global root as specified in the
   root vector.

 - The root is not the current process and is not in the next level of the
   hierarchy. In this case 0 must be the local root since this process will
   never communicate with the real root.

 - The root is not the current process but will be in next level of the
   hierarchy. In this case the current process must be the root.

 - I am the root. The root is my index.

Tested with IMB which rotates the root on every call to MPI_Reduce. Consider
IMB the reproducer for the issue this commit solves.

Make the bcast algorithm decision an enumerated variable

Resolve various asset failures when destructing coll ml requests.

Two issues:

 - Always reset the request to be invalid before returning it to the
   free list. This will avoid an asset in ompi_request_t's destructor.
   OMPI_REQUEST_FINI does this (and also releases the fortran handle
   index).

 - Never explicitly construct or destruct the superclass of an opal
   object. This screws up the class function tables and will cause
   either an assert failure or a segmentation fault when destructing
   coll ml requests.

Cleanup allgather.

I removed the duplicate non-blocking and blocking functions and modeled
the cleanup after what I found in allreduce. Also cleaned up the code
somewhat.

Don't bother copying from the send to the recieve buffer in
bcol_basesmuma_allreduce_intra_fanin_fanout if the pointers are the
same.

The eliminates a warning about memcpy and aliasing and avoids an
unnecessary call to memcpy.

Alwasy call CHECK_AND_RELEASE on memsync collectives.

There was a call to OBJ_RELEASE on the collective communicator but
because CHECK_AND_RECYLCE was never called there was not matching call
to OBJ_RELEASE. This caused coll ml to leak communicators.

Make allreduce use the sequential collective launcher in coll_ml_inlines.h

Just launch the next collective in the component progress.

I am a little unsure about this patch. There appears to be some sort
of race between collectives that causes buffer exhaustion in some cases
(IMB Allreduce is a reproducer). Changing progress to only launch the
next bcol seems to resolve the issue but might not be the best fix.

Note that I see little-no performance penalty for this change.

Fix allreduce when there are extra sources.

There was an issue with the buffer offset calculation when there are
extra sources. In the case of extra sources == 1 the offset was set
to buffer_size (just past the header of the next buffer). I adjusted
the buffer size to take into accoun the maximum header size (see the
earlier commit that added this) and simplified the offset calculation.

Make reduce/allreduce non-blocking. This is required for MPI_Comm_idup
to work correctly.

This has been tested with various layouts using the ibm testsuite and
imb and appears to have the same performance as the old blocking version.

Fix allgather for non-contiguous layouts and simplify parsing the
topology.

Some things in this patch:

 - There were several comments to the effect that level 0 of the
   hierarchy MUST contain all of the ranks. At least one function
   made this assumption but it was not true. I changed the sbgp
   components and the coll ml initization code to enforce this
   requirement.

 - Ensure that hierarchy level 0 has the ranks in the correct
   scatter gather order. This removes the need for a separate
   sort list and fixes the offset calculation for allgather.

 - There were several passes over the hierarchy to determine
   properties of the hierarchy. I eliminated these extra passes
   and the memory allocation associated with them and calculate the
   tree properties on the fly. The same DFS recursion also handles
   the re-order of level 0.

All these changes have been verified with MPI_Allreduce, MPI_Reduce, and
MPI_Allgather. All functions now pass all IBM/Open MPI, and IMB tests.

coll/ml: correct pointer usage for MPI_BOTTOM

Since contiguous datatypes are copied via memcpy (bypassing the convertor) we
need to adjust for the lb of the datatype. This corrects problems found testing
code that uses MPI_BOTTOM (NULL) as the send pointer.

Add fallback collectives for allreduce and reduce.

cmr=v1.7.5:reviewer=pasha

This commit was SVN r30363.
2014-01-22 15:39:19 +00:00

482 строки
17 KiB
C
Исходник Ответственный История

/* -*- Mode: C; c-basic-offset:4 ; indent-tabs-mode:nil -*- */
/*
* Copyright (c) 2004-2006 The Trustees of Indiana University and Indiana
* University Research and Technology
* Corporation. All rights reserved.
* Copyright (c) 2004-2010 The University of Tennessee and The University
* of Tennessee Research Foundation. All rights
* reserved.
* Copyright (c) 2004-2007 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) 2007-2012 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2009 Sun Microsystems, Inc. All rights reserved.
* Copyright (c) 2013 Los Alamos National Security, LLC. All rights
* reserved.
* $COPYRIGHT$
*
* Additional copyrights may follow
*
* $HEADER$
*/
#include "ompi_config.h"
#include "opal/class/opal_pointer_array.h"
#include "ompi/constants.h"
#include "ompi/op/op.h"
#include "ompi/mca/op/base/base.h"
#include "ompi/datatype/ompi_datatype_internal.h"
/*
* Table for Fortran <-> C op handle conversion
*/
opal_pointer_array_t *ompi_op_f_to_c_table;
/*
* Create intrinsic op
*/
static int add_intrinsic(ompi_op_t *op, int fort_handle, int flags,
const char *name);
/*
* Class information
*/
static void ompi_op_construct(ompi_op_t *eh);
static void ompi_op_destruct(ompi_op_t *eh);
/*
* Class instance
*/
OBJ_CLASS_INSTANCE(ompi_op_t, opal_object_t,
ompi_op_construct, ompi_op_destruct);
/*
* Intrinsic MPI_Op objects (_addr flavors are for F03 bindings)
*/
ompi_predefined_op_t ompi_mpi_op_null;
ompi_predefined_op_t *ompi_mpi_op_null_addr = &ompi_mpi_op_null;
ompi_predefined_op_t ompi_mpi_op_max;
ompi_predefined_op_t *ompi_mpi_op_max_addr = &ompi_mpi_op_max;
ompi_predefined_op_t ompi_mpi_op_min;
ompi_predefined_op_t *ompi_mpi_op_min_addr = &ompi_mpi_op_min;
ompi_predefined_op_t ompi_mpi_op_sum;
ompi_predefined_op_t *ompi_mpi_op_sum_addr = &ompi_mpi_op_sum;
ompi_predefined_op_t ompi_mpi_op_prod;
ompi_predefined_op_t *ompi_mpi_op_prod_addr = &ompi_mpi_op_prod;
ompi_predefined_op_t ompi_mpi_op_land;
ompi_predefined_op_t *ompi_mpi_op_land_addr = &ompi_mpi_op_land;
ompi_predefined_op_t ompi_mpi_op_band;
ompi_predefined_op_t *ompi_mpi_op_band_addr = &ompi_mpi_op_band;
ompi_predefined_op_t ompi_mpi_op_lor;
ompi_predefined_op_t *ompi_mpi_op_lor_addr = &ompi_mpi_op_lor;
ompi_predefined_op_t ompi_mpi_op_bor;
ompi_predefined_op_t *ompi_mpi_op_bor_addr = &ompi_mpi_op_bor;
ompi_predefined_op_t ompi_mpi_op_lxor;
ompi_predefined_op_t *ompi_mpi_op_lxor_addr = &ompi_mpi_op_lxor;
ompi_predefined_op_t ompi_mpi_op_bxor;
ompi_predefined_op_t *ompi_mpi_op_bxor_addr = &ompi_mpi_op_bxor;
ompi_predefined_op_t ompi_mpi_op_maxloc;
ompi_predefined_op_t *ompi_mpi_op_maxloc_addr = &ompi_mpi_op_maxloc;
ompi_predefined_op_t ompi_mpi_op_minloc;
ompi_predefined_op_t *ompi_mpi_op_minloc_addr = &ompi_mpi_op_minloc;
ompi_predefined_op_t ompi_mpi_op_replace;
ompi_predefined_op_t *ompi_mpi_op_replace_addr = &ompi_mpi_op_replace;
/*
* Map from ddt->id to position in op function pointer array
*/
int ompi_op_ddt_map[OMPI_DATATYPE_MAX_PREDEFINED];
/* Get the c complex operator associated with a fortran complex type */
#define FORTRAN_COMPLEX_OP_TYPE_X(type) OMPI_OP_BASE_TYPE_ ## type
/* Preprocessor hack to ensure type gets expanded correctly */
#define FORTRAN_COMPLEX_OP_TYPE(type) FORTRAN_COMPLEX_OP_TYPE_X(type)
#define FLAGS_NO_FLOAT \
(OMPI_OP_FLAGS_INTRINSIC | OMPI_OP_FLAGS_ASSOC | OMPI_OP_FLAGS_COMMUTE)
#define FLAGS \
(OMPI_OP_FLAGS_INTRINSIC | OMPI_OP_FLAGS_ASSOC | \
OMPI_OP_FLAGS_FLOAT_ASSOC | OMPI_OP_FLAGS_COMMUTE)
/*
* Initialize OMPI op infrastructure
*/
int ompi_op_init(void)
{
int i;
/* initialize ompi_op_f_to_c_table */
ompi_op_f_to_c_table = OBJ_NEW(opal_pointer_array_t);
if (NULL == ompi_op_f_to_c_table){
return OMPI_ERROR;
}
/* Fill in the ddt.id->op_position map */
for (i = 0; i < OMPI_DATATYPE_MAX_PREDEFINED; ++i) {
ompi_op_ddt_map[i] = -1;
}
ompi_op_ddt_map[OMPI_DATATYPE_MPI_INT8_T] = OMPI_OP_BASE_TYPE_INT8_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_UINT8_T] = OMPI_OP_BASE_TYPE_UINT8_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_INT16_T] = OMPI_OP_BASE_TYPE_INT16_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_UINT16_T] = OMPI_OP_BASE_TYPE_UINT16_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_INT32_T] = OMPI_OP_BASE_TYPE_INT32_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_UINT32_T] = OMPI_OP_BASE_TYPE_UINT32_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_INT64_T] = OMPI_OP_BASE_TYPE_INT64_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_UINT64_T] = OMPI_OP_BASE_TYPE_UINT64_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_FLOAT] = OMPI_OP_BASE_TYPE_FLOAT;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_DOUBLE] = OMPI_OP_BASE_TYPE_DOUBLE;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_LONG_DOUBLE] = OMPI_OP_BASE_TYPE_LONG_DOUBLE;
#if OMPI_HAVE_FORTRAN_COMPLEX8
ompi_op_ddt_map[OMPI_DATATYPE_MPI_COMPLEX8] = FORTRAN_COMPLEX_OP_TYPE(OMPI_KIND_FORTRAN_COMPLEX8);
#endif
#if OMPI_HAVE_FORTRAN_COMPLEX16
ompi_op_ddt_map[OMPI_DATATYPE_MPI_COMPLEX16] = FORTRAN_COMPLEX_OP_TYPE(OMPI_KIND_FORTRAN_COMPLEX16);
#endif
/* Only enable reductions on COMPLEX32 if the REAL*16 matches the equivalent C type */
#if OMPI_REAL16_MATCHES_C && OMPI_HAVE_FORTRAN_COMPLEX32
ompi_op_ddt_map[OMPI_DATATYPE_MPI_COMPLEX32] = FORTRAN_COMPLEX_OP_TYPE(OMPI_KIND_FORTRAN_COMPLEX32);
#endif
ompi_op_ddt_map[OMPI_DATATYPE_MPI_WCHAR] = OMPI_OP_BASE_TYPE_WCHAR;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_BOOL] = OMPI_OP_BASE_TYPE_BOOL;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_LOGICAL] = OMPI_OP_BASE_TYPE_LOGICAL;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_CHARACTER] = OMPI_OP_BASE_TYPE_UINT8_T;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_INTEGER] = OMPI_OP_BASE_TYPE_INTEGER;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_REAL] = OMPI_OP_BASE_TYPE_REAL;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_DOUBLE_PRECISION] = OMPI_OP_BASE_TYPE_DOUBLE_PRECISION;
#if OMPI_HAVE_FORTRAN_COMPLEX
ompi_op_ddt_map[OMPI_DATATYPE_MPI_COMPLEX] = FORTRAN_COMPLEX_OP_TYPE(OMPI_KIND_FORTRAN_COMPLEX);
#endif
#if OMPI_HAVE_FORTRAN_DOUBLE_COMPLEX
ompi_op_ddt_map[OMPI_DATATYPE_MPI_DOUBLE_COMPLEX] = FORTRAN_COMPLEX_OP_TYPE(OMPI_KIND_FORTRAN_DOUBLE_COMPLEX);
#endif
ompi_op_ddt_map[OMPI_DATATYPE_MPI_LONG_DOUBLE_COMPLEX] = OMPI_OP_BASE_TYPE_C_LONG_DOUBLE_COMPLEX;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_2INT] = OMPI_OP_BASE_TYPE_2INT;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_2INTEGER] = OMPI_OP_BASE_TYPE_2INTEGER;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_2REAL] = OMPI_OP_BASE_TYPE_2REAL;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_2DBLPREC] = OMPI_OP_BASE_TYPE_2DOUBLE_PRECISION;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_2COMPLEX] = -1; /* Not defined */
ompi_op_ddt_map[OMPI_DATATYPE_MPI_2DOUBLE_COMPLEX] = -1; /* Not defined */
ompi_op_ddt_map[OMPI_DATATYPE_MPI_FLOAT_INT] = OMPI_OP_BASE_TYPE_FLOAT_INT;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_DOUBLE_INT] = OMPI_OP_BASE_TYPE_DOUBLE_INT;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_LONG_DOUBLE_INT] = OMPI_OP_BASE_TYPE_LONG_DOUBLE_INT;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_LONG_INT] = OMPI_OP_BASE_TYPE_LONG_INT;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_SHORT_INT] = OMPI_OP_BASE_TYPE_SHORT_INT;
#if SIZEOF_PTRDIFF_T == 4
ompi_op_ddt_map[OMPI_DATATYPE_MPI_AINT] = OMPI_OP_BASE_TYPE_INT32_T;
#elif SIZEOF_PTRDIFF_T == 8
ompi_op_ddt_map[OMPI_DATATYPE_MPI_AINT] = OMPI_OP_BASE_TYPE_INT64_T;
#else
#warning Unsupported definition for MPI_AINT
#endif
#if OMPI_MPI_OFFSET_SIZE == 4
ompi_op_ddt_map[OMPI_DATATYPE_MPI_OFFSET] = OMPI_OP_BASE_TYPE_UINT32_T;
#elif OMPI_MPI_OFFSET_SIZE == 8
ompi_op_ddt_map[OMPI_DATATYPE_MPI_OFFSET] = OMPI_OP_BASE_TYPE_UINT64_T;
#else
#warning Unsupported definition for MPI_OFFSET
#endif
ompi_op_ddt_map[OMPI_DATATYPE_MPI_C_BOOL] = OMPI_OP_BASE_TYPE_BOOL;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_C_COMPLEX] = OMPI_OP_BASE_TYPE_C_FLOAT_COMPLEX;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_C_FLOAT_COMPLEX] = OMPI_OP_BASE_TYPE_C_FLOAT_COMPLEX;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_C_DOUBLE_COMPLEX] = OMPI_OP_BASE_TYPE_C_DOUBLE_COMPLEX;
ompi_op_ddt_map[OMPI_DATATYPE_MPI_C_LONG_DOUBLE_COMPLEX] = OMPI_OP_BASE_TYPE_C_LONG_DOUBLE_COMPLEX;
/* MPI 3.0 datatypes */
#if OMPI_MPI_COUNT_SIZE == 4
ompi_op_ddt_map[OMPI_DATATYPE_MPI_COUNT] = OMPI_OP_BASE_TYPE_INT32_T;
#elif OMPI_MPI_COUNT_SIZE == 8
ompi_op_ddt_map[OMPI_DATATYPE_MPI_COUNT] = OMPI_OP_BASE_TYPE_INT64_T;
#else
#warning Unsupported definition for MPI_COUNT
#endif
/* Create the intrinsic ops */
if (OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_null.op, OMPI_OP_BASE_FORTRAN_NULL,
FLAGS, "MPI_NULL") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_max.op, OMPI_OP_BASE_FORTRAN_MAX,
FLAGS, "MPI_MAX") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_min.op, OMPI_OP_BASE_FORTRAN_MIN,
FLAGS, "MPI_MIN") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_sum.op, OMPI_OP_BASE_FORTRAN_SUM,
FLAGS_NO_FLOAT, "MPI_SUM") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_prod.op, OMPI_OP_BASE_FORTRAN_PROD,
FLAGS_NO_FLOAT, "MPI_PROD") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_land.op, OMPI_OP_BASE_FORTRAN_LAND,
FLAGS, "MPI_LAND") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_band.op, OMPI_OP_BASE_FORTRAN_BAND,
FLAGS, "MPI_BAND") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_lor.op, OMPI_OP_BASE_FORTRAN_LOR,
FLAGS, "MPI_LOR") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_bor.op, OMPI_OP_BASE_FORTRAN_BOR,
FLAGS, "MPI_BOR") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_lxor.op, OMPI_OP_BASE_FORTRAN_LXOR,
FLAGS, "MPI_LXOR") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_bxor.op, OMPI_OP_BASE_FORTRAN_BXOR,
FLAGS, "MPI_BXOR") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_maxloc.op, OMPI_OP_BASE_FORTRAN_MAXLOC,
FLAGS, "MPI_MAXLOC") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_minloc.op, OMPI_OP_BASE_FORTRAN_MINLOC,
FLAGS, "MPI_MINLOC") ||
OMPI_SUCCESS !=
add_intrinsic(&ompi_mpi_op_replace.op, OMPI_OP_BASE_FORTRAN_REPLACE,
FLAGS, "MPI_REPLACE")) {
return OMPI_ERROR;
}else{
/* This code is placed back here to support
* HCOL allreduce at the moment. It is a part of bgate repository only. This conflict with OMPI v1.7
* is to be resolved some other way.
* */
ompi_mpi_op_null.op.op_type = OMPI_OP_NULL;
ompi_mpi_op_max.op.op_type = OMPI_OP_MAX;
ompi_mpi_op_min.op.op_type = OMPI_OP_MIN;
ompi_mpi_op_sum.op.op_type = OMPI_OP_SUM;
ompi_mpi_op_prod.op.op_type = OMPI_OP_PROD;
ompi_mpi_op_land.op.op_type = OMPI_OP_LAND;
ompi_mpi_op_band.op.op_type = OMPI_OP_BAND;
ompi_mpi_op_lor.op.op_type = OMPI_OP_LOR;
ompi_mpi_op_bor.op.op_type = OMPI_OP_BOR;
ompi_mpi_op_lxor.op.op_type = OMPI_OP_LXOR;
ompi_mpi_op_bxor.op.op_type = OMPI_OP_BXOR;
ompi_mpi_op_maxloc.op.op_type = OMPI_OP_MAXLOC;
ompi_mpi_op_minloc.op.op_type = OMPI_OP_MINLOC;
ompi_mpi_op_replace.op.op_type = OMPI_OP_REPLACE;
}
/* All done */
return OMPI_SUCCESS;
}
/*
* Clean up the op resources
*/
int ompi_op_finalize(void)
{
/* clean up the intrinsic ops */
OBJ_DESTRUCT(&ompi_mpi_op_replace);
OBJ_DESTRUCT(&ompi_mpi_op_minloc);
OBJ_DESTRUCT(&ompi_mpi_op_maxloc);
OBJ_DESTRUCT(&ompi_mpi_op_bxor);
OBJ_DESTRUCT(&ompi_mpi_op_lxor);
OBJ_DESTRUCT(&ompi_mpi_op_bor);
OBJ_DESTRUCT(&ompi_mpi_op_lor);
OBJ_DESTRUCT(&ompi_mpi_op_band);
OBJ_DESTRUCT(&ompi_mpi_op_land);
OBJ_DESTRUCT(&ompi_mpi_op_prod);
OBJ_DESTRUCT(&ompi_mpi_op_sum);
OBJ_DESTRUCT(&ompi_mpi_op_min);
OBJ_DESTRUCT(&ompi_mpi_op_max);
OBJ_DESTRUCT(&ompi_mpi_op_null);
/* Remove op F2C table */
OBJ_RELEASE(ompi_op_f_to_c_table);
/* All done */
return OMPI_SUCCESS;
}
/*
* Create a new MPI_Op
*/
ompi_op_t *ompi_op_create_user(bool commute,
ompi_op_fortran_handler_fn_t func)
{
ompi_op_t *new_op;
/* Create a new object and ensure that it's valid */
new_op = OBJ_NEW(ompi_op_t);
if (NULL == new_op) {
goto error;
}
if (0 > new_op->o_f_to_c_index) {
OBJ_RELEASE(new_op);
new_op = NULL;
goto error;
}
/*
* The new object is valid -- initialize it. If this is being
* created from fortran, the fortran MPI API wrapper function
* will override the o_flags field directly. We cast the
* function pointer type to the fortran type arbitrarily -- it
* only has to be a function pointer in order to store properly,
* it doesn't matter what type it is (we'll cast it to the Right
* type when we *use* it).
*/
new_op->o_flags = OMPI_OP_FLAGS_ASSOC;
if (commute) {
new_op->o_flags |= OMPI_OP_FLAGS_COMMUTE;
}
/* Set the user-defined callback function. The "fort_fn" member
is part of a union, so it doesn't matter if this is a C or
Fortan callback; we'll call the right flavor (per o_flags) at
invocation time. */
new_op->o_func.fort_fn = func;
error:
/* All done */
return new_op;
}
/*
* See lengthy comment in mpi/cxx/intercepts.cc for how the C++ MPI::Op
* callbacks work.
*/
void ompi_op_set_cxx_callback(ompi_op_t *op, MPI_User_function *fn)
{
op->o_flags |= OMPI_OP_FLAGS_CXX_FUNC;
/* The OMPI C++ intercept was previously stored in
op->o_func.fort_fn by ompi_op_create_user(). So save that in
cxx.intercept_fn and put the user's fn in cxx.user_fn. */
op->o_func.cxx_data.intercept_fn =
(ompi_op_cxx_handler_fn_t *) op->o_func.fort_fn;
op->o_func.cxx_data.user_fn = fn;
}
/*
* See lengthy comment in mpi/cxx/intercepts.cc for how the C++ MPI::Op
* callbacks work.
*/
void ompi_op_set_java_callback(ompi_op_t *op, void *jnienv,
void *object, int baseType)
{
op->o_flags |= OMPI_OP_FLAGS_JAVA_FUNC;
/* The OMPI Java intercept was previously stored in
op->o_func.fort_fn by ompi_op_create_user(). So save that in
cxx.intercept_fn and put the user's fn in cxx.user_fn. */
op->o_func.java_data.intercept_fn =
(ompi_op_java_handler_fn_t *) op->o_func.fort_fn;
op->o_func.java_data.jnienv = jnienv;
op->o_func.java_data.object = object;
op->o_func.java_data.baseType = baseType;
}
/**************************************************************************
*
* Static functions
*
**************************************************************************/
static int add_intrinsic(ompi_op_t *op, int fort_handle, int flags,
const char *name)
{
/* Add the op to the table */
OBJ_CONSTRUCT(op, ompi_op_t);
if (op->o_f_to_c_index != fort_handle) {
return OMPI_ERROR;
}
/* Set the members */
op->o_flags = flags;
strncpy(op->o_name, name, sizeof(op->o_name) - 1);
op->o_name[sizeof(op->o_name) - 1] = '\0';
/* Perform the selection on this op to fill in the function
pointers (except for NULL and REPLACE, which don't get
components) */
if (OMPI_OP_BASE_FORTRAN_NULL != op->o_f_to_c_index &&
OMPI_OP_BASE_FORTRAN_REPLACE != op->o_f_to_c_index) {
return ompi_op_base_op_select(op);
} else {
return OMPI_SUCCESS;
}
}
/*
* Op constructor
*/
static void ompi_op_construct(ompi_op_t *new_op)
{
int i;
/* assign entry in fortran <-> c translation array */
new_op->o_f_to_c_index =
opal_pointer_array_add(ompi_op_f_to_c_table, new_op);
/* Set everything to NULL so that we can intelligently free
non-NULL's in the destructor */
for (i = 0; i < OMPI_OP_BASE_TYPE_MAX; ++i) {
new_op->o_func.intrinsic.fns[i] = NULL;
new_op->o_func.intrinsic.modules[i] = NULL;
new_op->o_3buff_intrinsic.fns[i] = NULL;
new_op->o_3buff_intrinsic.modules[i] = NULL;
}
}
/*
* Op destructor
*/
static void ompi_op_destruct(ompi_op_t *op)
{
int i;
/* reset the ompi_op_f_to_c_table entry - make sure that the
entry is in the table */
if (NULL != opal_pointer_array_get_item(ompi_op_f_to_c_table,
op->o_f_to_c_index)) {
opal_pointer_array_set_item(ompi_op_f_to_c_table,
op->o_f_to_c_index, NULL);
}
for (i = 0; i < OMPI_OP_BASE_TYPE_MAX; ++i) {
op->o_func.intrinsic.fns[i] = NULL;
if( NULL != op->o_func.intrinsic.modules[i] ) {
OBJ_RELEASE(op->o_func.intrinsic.modules[i]);
op->o_func.intrinsic.modules[i] = NULL;
}
op->o_3buff_intrinsic.fns[i] = NULL;
if( NULL != op->o_3buff_intrinsic.modules[i] ) {
OBJ_RELEASE(op->o_3buff_intrinsic.modules[i]);
op->o_3buff_intrinsic.modules[i] = NULL;
}
}
}
Показать git blame Копировать постоянную ссылку