2006-09-14 20:44:02 +04:00
|
|
|
/*
|
|
|
|
* Copyright (c) 2004-2006 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) 2004-2006 The Regents of the University of California.
|
|
|
|
* All rights reserved.
|
|
|
|
* Copyright (c) 2006 QLogic Corporation. All rights reserved.
|
|
|
|
* $COPYRIGHT$
|
|
|
|
*
|
|
|
|
* Additional copyrights may follow
|
|
|
|
*
|
|
|
|
* $HEADER$
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "ompi_config.h"
|
2009-03-04 20:06:51 +03:00
|
|
|
#include "ompi/mca/pml/pml.h"
|
2006-09-14 20:44:02 +04:00
|
|
|
#include "ompi/communicator/communicator.h"
|
- Split the datatype engine into two parts: an MPI specific part in
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.
2009-07-13 08:56:31 +04:00
|
|
|
#include "opal/datatype/opal_convertor.h"
|
2006-09-14 20:44:02 +04:00
|
|
|
|
|
|
|
#include "mtl_psm.h"
|
|
|
|
#include "mtl_psm_types.h"
|
|
|
|
#include "mtl_psm_request.h"
|
|
|
|
#include "ompi/mca/mtl/base/mtl_base_datatype.h"
|
|
|
|
|
|
|
|
int
|
|
|
|
ompi_mtl_psm_send(struct mca_mtl_base_module_t* mtl,
|
|
|
|
struct ompi_communicator_t* comm,
|
|
|
|
int dest,
|
|
|
|
int tag,
|
- Split the datatype engine into two parts: an MPI specific part in
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.
2009-07-13 08:56:31 +04:00
|
|
|
struct opal_convertor_t *convertor,
|
2006-09-14 20:44:02 +04:00
|
|
|
mca_pml_base_send_mode_t mode)
|
|
|
|
{
|
|
|
|
psm_error_t err;
|
|
|
|
mca_mtl_psm_request_t mtl_psm_request;
|
|
|
|
uint64_t mqtag;
|
|
|
|
uint32_t flags = 0;
|
|
|
|
int ret;
|
|
|
|
size_t length;
|
2006-09-21 02:14:46 +04:00
|
|
|
ompi_proc_t* ompi_proc = ompi_comm_peer_lookup( comm, dest );
|
2013-08-30 20:54:55 +04:00
|
|
|
mca_mtl_psm_endpoint_t* psm_endpoint = (mca_mtl_psm_endpoint_t*) ompi_proc->proc_endpoints[OMPI_PROC_ENDPOINT_TAG_MTL];
|
2006-09-14 20:44:02 +04:00
|
|
|
|
|
|
|
assert(mtl == &ompi_mtl_psm.super);
|
|
|
|
|
|
|
|
mqtag = PSM_MAKE_MQTAG(comm->c_contextid, comm->c_my_rank, tag);
|
|
|
|
|
|
|
|
ret = ompi_mtl_datatype_pack(convertor,
|
|
|
|
&mtl_psm_request.buf,
|
|
|
|
&length,
|
|
|
|
&mtl_psm_request.free_after);
|
|
|
|
|
|
|
|
|
|
|
|
mtl_psm_request.length = length;
|
|
|
|
mtl_psm_request.convertor = convertor;
|
|
|
|
mtl_psm_request.type = OMPI_MTL_PSM_ISEND;
|
|
|
|
|
|
|
|
if (OMPI_SUCCESS != ret) return ret;
|
|
|
|
|
|
|
|
if (mode == MCA_PML_BASE_SEND_SYNCHRONOUS)
|
|
|
|
flags |= PSM_MQ_FLAG_SENDSYNC;
|
|
|
|
|
|
|
|
err = psm_mq_send(ompi_mtl_psm.mq,
|
|
|
|
psm_endpoint->peer_addr,
|
|
|
|
flags,
|
|
|
|
mqtag,
|
|
|
|
mtl_psm_request.buf,
|
|
|
|
length);
|
|
|
|
|
2009-07-30 06:55:20 +04:00
|
|
|
if (mtl_psm_request.free_after) {
|
2007-02-03 00:55:04 +03:00
|
|
|
free(mtl_psm_request.buf);
|
2009-07-30 06:55:20 +04:00
|
|
|
}
|
2007-02-03 00:55:04 +03:00
|
|
|
|
2006-09-14 20:44:02 +04:00
|
|
|
return err == PSM_OK ? OMPI_SUCCESS : OMPI_ERROR;
|
|
|
|
}
|
|
|
|
|
|
|
|
int
|
|
|
|
ompi_mtl_psm_isend(struct mca_mtl_base_module_t* mtl,
|
|
|
|
struct ompi_communicator_t* comm,
|
|
|
|
int dest,
|
|
|
|
int tag,
|
- Split the datatype engine into two parts: an MPI specific part in
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.
2009-07-13 08:56:31 +04:00
|
|
|
struct opal_convertor_t *convertor,
|
2006-09-14 20:44:02 +04:00
|
|
|
mca_pml_base_send_mode_t mode,
|
|
|
|
bool blocking,
|
|
|
|
mca_mtl_request_t * mtl_request)
|
|
|
|
{
|
|
|
|
psm_error_t psm_error;
|
|
|
|
uint64_t mqtag;
|
|
|
|
uint32_t flags = 0;
|
|
|
|
int ret;
|
|
|
|
mca_mtl_psm_request_t * mtl_psm_request = (mca_mtl_psm_request_t*) mtl_request;
|
|
|
|
size_t length;
|
2006-10-10 03:14:22 +04:00
|
|
|
ompi_proc_t* ompi_proc = ompi_comm_peer_lookup( comm, dest );
|
2013-08-30 20:54:55 +04:00
|
|
|
mca_mtl_psm_endpoint_t* psm_endpoint = (mca_mtl_psm_endpoint_t*)ompi_proc->proc_endpoints[OMPI_PROC_ENDPOINT_TAG_MTL];
|
2006-09-14 20:44:02 +04:00
|
|
|
|
|
|
|
assert(mtl == &ompi_mtl_psm.super);
|
|
|
|
|
|
|
|
mqtag = PSM_MAKE_MQTAG(comm->c_contextid, comm->c_my_rank, tag);
|
|
|
|
|
|
|
|
|
|
|
|
ret = ompi_mtl_datatype_pack(convertor,
|
|
|
|
&mtl_psm_request->buf,
|
|
|
|
&length,
|
|
|
|
&mtl_psm_request->free_after);
|
|
|
|
|
|
|
|
mtl_psm_request->length= length;
|
|
|
|
mtl_psm_request->convertor = convertor;
|
|
|
|
mtl_psm_request->type = OMPI_MTL_PSM_ISEND;
|
|
|
|
|
|
|
|
if (OMPI_SUCCESS != ret) return ret;
|
|
|
|
|
|
|
|
if (mode == MCA_PML_BASE_SEND_SYNCHRONOUS)
|
|
|
|
flags |= PSM_MQ_FLAG_SENDSYNC;
|
|
|
|
|
|
|
|
psm_error = psm_mq_isend(ompi_mtl_psm.mq,
|
|
|
|
psm_endpoint->peer_addr,
|
|
|
|
flags,
|
|
|
|
mqtag,
|
|
|
|
mtl_psm_request->buf,
|
|
|
|
length,
|
|
|
|
mtl_psm_request,
|
|
|
|
&mtl_psm_request->psm_request);
|
|
|
|
|
|
|
|
return psm_error == PSM_OK ? OMPI_SUCCESS : OMPI_ERROR;
|
|
|
|
}
|