6c5532072a
OMPI and a language agnostic part in OPAL. The convertor is completely moved into OPAL. This offers several benefits as described in RFC http://www.open-mpi.org/community/lists/devel/2009/07/6387.php namely: - Fewer basic types (int* and float* types, boolean and wchar - Fixing naming scheme to ompi-nomenclature. - Usability outside of the ompi-layer. - Due to the fixed nature of simple opal types, their information is completely known at compile time and therefore constified - With fewer datatypes (22), the actual sizes of bit-field types may be reduced from 64 to 32 bits, allowing reorganizing the opal_datatype structure, eliminating holes and keeping data required in convertor (upon send/recv) in one cacheline... This has implications to the convertor-datastructure and other parts of the code. - Several performance tests have been run, the netpipe latency does not change with this patch on Linux/x86-64 on the smoky cluster. - Extensive tests have been done to verify correctness (no new regressions) using: 1. mpi_test_suite on linux/x86-64 using clean ompi-trunk and ompi-ddt: a. running both trunk and ompi-ddt resulted in no differences (except for MPI_SHORT_INT and MPI_TYPE_MIX_LB_UB do now run correctly). b. with --enable-memchecker and running under valgrind (one buglet when run with static found in test-suite, commited) 2. ibm testsuite on linux/x86-64 using clean ompi-trunk and ompi-ddt: all passed (except for the dynamic/ tests failed!! as trunk/MTT) 3. compilation and usage of HDF5 tests on Jaguar using PGI and PathScale compilers. 4. compilation and usage on Scicortex. - Please note, that for the heterogeneous case, (-m32 compiled binaries/ompi), neither ompi-trunk, nor ompi-ddt branch would successfully launch. This commit was SVN r21641.
99 строки
5.0 KiB
C
99 строки
5.0 KiB
C
/*
|
|
* Copyright (c) 2004-2007 The Trustees of Indiana University and Indiana
|
|
* University Research and Technology
|
|
* Corporation. All rights reserved.
|
|
* Copyright (c) 2004-2006 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
|
|
*
|
|
* $HEADER$
|
|
*/
|
|
|
|
#ifndef OMPI_C_BINDINGS_H
|
|
#define OMPI_C_BINDINGS_H
|
|
|
|
#include "ompi_config.h"
|
|
#include "mpi.h"
|
|
#include "ompi/datatype/ompi_datatype.h"
|
|
|
|
/* This library needs to be here so that we can define
|
|
* the OPAL_CR_* checks
|
|
*/
|
|
#include "opal/runtime/opal_cr.h"
|
|
|
|
BEGIN_C_DECLS
|
|
|
|
/* If compiling in the profile directory, then we don't have weak
|
|
symbols and therefore we need the defines to map from MPI->PMPI.
|
|
NOTE: pragma weak stuff is handled on a file-by-file basis; it
|
|
doesn't work to simply list all of the pragmas in a top-level
|
|
header file. */
|
|
|
|
/* These macros have to be used to check the corectness of the datatype depending on the
|
|
* operations that we have to do with them. They can be used on all functions, not only
|
|
* on the top level MPI functions, as they does not trigger the error handler. Is the user
|
|
* responsability to do it.
|
|
*/
|
|
#define OMPI_CHECK_DATATYPE_FOR_SEND( RC, DDT, COUNT ) \
|
|
do { \
|
|
/* (RC) = MPI_SUCCESS; */ \
|
|
if( NULL == (DDT) || MPI_DATATYPE_NULL == (DDT) ) (RC) = MPI_ERR_TYPE; \
|
|
else if( (COUNT) < 0 ) (RC) = MPI_ERR_COUNT; \
|
|
else if( !opal_datatype_is_committed(&((DDT)->super)) ) (RC) = MPI_ERR_TYPE; \
|
|
else if( !opal_datatype_is_valid(&((DDT)->super)) ) (RC) = MPI_ERR_TYPE; \
|
|
} while (0)
|
|
|
|
#define OMPI_CHECK_DATATYPE_FOR_RECV( RC, DDT, COUNT ) \
|
|
do { \
|
|
/* (RC) = MPI_SUCCESS; */ \
|
|
if( NULL == (DDT) || MPI_DATATYPE_NULL == (DDT) ) (RC) = MPI_ERR_TYPE; \
|
|
else if( (COUNT) < 0 ) (RC) = MPI_ERR_COUNT; \
|
|
else if( !opal_datatype_is_committed(&((DDT)->super)) ) (RC) = MPI_ERR_TYPE; \
|
|
/* XXX Fix flags else if( ompi_datatype_is_overlapped((DDT)) ) (RC) = MPI_ERR_TYPE; */ \
|
|
else if( !opal_datatype_is_valid(&((DDT)->super)) ) (RC) = MPI_ERR_TYPE; \
|
|
} while (0)
|
|
|
|
#define OMPI_CHECK_DATATYPE_FOR_ONE_SIDED( RC, DDT, COUNT ) \
|
|
do { \
|
|
/*(RC) = MPI_SUCCESS; */ \
|
|
if( NULL == (DDT) || MPI_DATATYPE_NULL == (DDT) ) (RC) = MPI_ERR_TYPE; \
|
|
else if( (COUNT) < 0 ) (RC) = MPI_ERR_COUNT; \
|
|
else if( !opal_datatype_is_committed(&((DDT)->super)) ) (RC) = MPI_ERR_TYPE; \
|
|
else if( opal_datatype_is_overlapped(&((DDT)->super)) ) (RC) = MPI_ERR_TYPE; \
|
|
else if( !ompi_datatype_is_acceptable_for_one_sided(DDT)) (RC) = MPI_ERR_TYPE; \
|
|
else if( !opal_datatype_is_valid(&((DDT)->super)) ) (RC) = MPI_ERR_TYPE; \
|
|
} while(0)
|
|
|
|
|
|
/* This macro has to be used to check the correctness of the user buffer depending on the datatype.
|
|
* This macro expects that the DDT parameter is a valid pointer to an ompi datatype object.
|
|
*/
|
|
#define OMPI_CHECK_USER_BUFFER(RC, BUFFER, DDT, COUNT) \
|
|
do { \
|
|
if ( NULL == (BUFFER) && 0 < (COUNT) && MPI_SUCCESS == (RC) ) { \
|
|
if ( (DDT)->super.flags & OPAL_DATATYPE_FLAG_PREDEFINED ) { \
|
|
(RC) = MPI_ERR_BUFFER; \
|
|
} else { \
|
|
size_t size = 0; \
|
|
ptrdiff_t true_lb = 0; \
|
|
ptrdiff_t true_extended = 0; \
|
|
ompi_datatype_type_size((DDT), &size); \
|
|
ompi_datatype_get_true_extent((DDT), &true_lb, &true_extended); \
|
|
if ( 0 < size && 0 == true_lb ) { \
|
|
(RC) = MPI_ERR_BUFFER; \
|
|
} \
|
|
} \
|
|
} \
|
|
} while (0)
|
|
|
|
END_C_DECLS
|
|
|
|
#endif /* OMPI_C_BINDINGS_H */
|