
This is a meta commit, that encapsulate all the ADAPT commits in the master into a single PR for 4.1. The master commits included here are: fe73586, a4be3bb, d712645, c2970a3, e59bde9, ee592f3 and c98e387. Here is a detailed list of added capabilities: * coll/adapt: Fix naming conventions and C11 atomic use * coll/adapt: Remove unused component field in module * Consistent handling of zero counts in the MPI API. * Correctly handle non-blocking collectives tags * As it is possible to have multiple outstanding non-blocking collectives provided by different collective modules, we need a consistent mechanism to allow them to select unique tags for each instance of a collective. * Add support for fallback to previous coll module on non-commutative operations (#30) * Replace mutexes by atomic operations. * Use the correct nbc request type (for both ibcast and ireduce) * coll/base: document type casts in ompi_coll_base_retain_* * add module-wide topology cache * use standard instead of synchronous send and add mca parameter to control mode of initial send in ireduce/ibcast * reduce number of memory allocations * call the default request completion. * Remove the requests from the Fortran lookup conversion tables before completing and free it. * piggybacking Bull functionalities Signed-off-by: Xi Luo <xluo12@vols.utk.edu> Signed-off-by: George Bosilca <bosilca@icl.utk.edu> Signed-off-by: Marc Sergent <marc.sergent@atos.net> Co-authored-by: Joseph Schuchart <schuchart@hlrs.de> Co-authored-by: Lemarinier, Pierre <pierre.lemarinier@atos.net> Co-authored-by: pierrele <31764860+pierrele@users.noreply.github.com>
156 строки
5.3 KiB
C
156 строки
5.3 KiB
C
/*
|
|
* Copyright (c) 2014-2020 The University of Tennessee and The University
|
|
* of Tennessee Research Foundation. All rights
|
|
* reserved.
|
|
* $COPYRIGHT$
|
|
*
|
|
* Additional copyrights may follow
|
|
*
|
|
* $HEADER$
|
|
*/
|
|
|
|
#include "ompi_config.h"
|
|
|
|
#include "opal/util/show_help.h"
|
|
#include "ompi/constants.h"
|
|
#include "ompi/mca/coll/coll.h"
|
|
#include "coll_adapt.h"
|
|
#include "coll_adapt_algorithms.h"
|
|
|
|
/*
|
|
* Public string showing the coll ompi_adapt component version number
|
|
*/
|
|
const char *mca_coll_adapt_component_version_string =
|
|
"Open MPI ADAPT collective MCA component version " OMPI_VERSION;
|
|
|
|
/*
|
|
* Local functions
|
|
*/
|
|
static int adapt_open(void);
|
|
static int adapt_close(void);
|
|
static int adapt_register(void);
|
|
|
|
/*
|
|
* Instantiate the public struct with all of our public information
|
|
* and pointers to our public functions in it
|
|
*/
|
|
|
|
mca_coll_adapt_component_t mca_coll_adapt_component = {
|
|
/* First, fill in the super */
|
|
{
|
|
/* First, the mca_component_t struct containing meta
|
|
information about the component itself */
|
|
.collm_version = {
|
|
MCA_COLL_BASE_VERSION_2_0_0,
|
|
|
|
/* Component name and version */
|
|
.mca_component_name = "adapt",
|
|
MCA_BASE_MAKE_VERSION(component, OMPI_MAJOR_VERSION, OMPI_MINOR_VERSION,
|
|
OMPI_RELEASE_VERSION),
|
|
|
|
/* Component functions */
|
|
.mca_open_component = adapt_open,
|
|
.mca_close_component = adapt_close,
|
|
.mca_register_component_params = adapt_register,
|
|
},
|
|
.collm_data = {
|
|
/* The component is not checkpoint ready */
|
|
MCA_BASE_METADATA_PARAM_NONE
|
|
},
|
|
|
|
/* Initialization / querying functions */
|
|
.collm_init_query = ompi_coll_adapt_init_query,
|
|
.collm_comm_query = ompi_coll_adapt_comm_query,
|
|
},
|
|
|
|
/* adapt-component specific information */
|
|
|
|
0, /* (default) priority */
|
|
|
|
0, /* (default) output stream */
|
|
0, /* (default) verbose level */
|
|
|
|
/* default values for non-MCA parameters */
|
|
/* Not specifying values here gives us all 0's */
|
|
};
|
|
|
|
/* Open the component */
|
|
static int adapt_open(void)
|
|
{
|
|
mca_coll_adapt_component_t *cs = &mca_coll_adapt_component;
|
|
|
|
if (cs->adapt_verbose > 0) {
|
|
cs->adapt_output = opal_output_open(NULL);
|
|
opal_output_set_verbosity(cs->adapt_output, cs->adapt_verbose);
|
|
}
|
|
|
|
return OMPI_SUCCESS;
|
|
}
|
|
|
|
|
|
/* Shut down the component */
|
|
static int adapt_close(void)
|
|
{
|
|
ompi_coll_adapt_ibcast_fini();
|
|
ompi_coll_adapt_ireduce_fini();
|
|
|
|
return OMPI_SUCCESS;
|
|
}
|
|
|
|
static int adapt_verify_mca_variables(void)
|
|
{
|
|
return OMPI_SUCCESS;
|
|
}
|
|
|
|
/*
|
|
* Register MCA params
|
|
*/
|
|
static int adapt_register(void)
|
|
{
|
|
mca_base_component_t *c = &mca_coll_adapt_component.super.collm_version;
|
|
mca_coll_adapt_component_t *cs = &mca_coll_adapt_component;
|
|
|
|
/* If we want to be selected (i.e., all procs on one node), then
|
|
we should have a high priority */
|
|
cs->adapt_priority = 0;
|
|
(void) mca_base_component_var_register(c, "priority", "Priority of the adapt coll component",
|
|
MCA_BASE_VAR_TYPE_INT, NULL, 0, 0,
|
|
OPAL_INFO_LVL_9,
|
|
MCA_BASE_VAR_SCOPE_READONLY, &cs->adapt_priority);
|
|
|
|
cs->adapt_verbose = ompi_coll_base_framework.framework_verbose;
|
|
(void) mca_base_component_var_register(c, "verbose",
|
|
"Verbose level (default set to the collective framework verbosity)",
|
|
MCA_BASE_VAR_TYPE_INT, NULL, 0, 0,
|
|
OPAL_INFO_LVL_9,
|
|
MCA_BASE_VAR_SCOPE_READONLY, &cs->adapt_verbose);
|
|
|
|
cs->adapt_context_free_list_min = 64;
|
|
(void) mca_base_component_var_register(c, "context_free_list_min",
|
|
"Minimum number of segments in context free list",
|
|
MCA_BASE_VAR_TYPE_INT, NULL, 0, 0,
|
|
OPAL_INFO_LVL_9,
|
|
MCA_BASE_VAR_SCOPE_READONLY,
|
|
&cs->adapt_context_free_list_min);
|
|
|
|
cs->adapt_context_free_list_max = 1024;
|
|
(void) mca_base_component_var_register(c, "context_free_list_max",
|
|
"Maximum number of segments in context free list",
|
|
MCA_BASE_VAR_TYPE_INT, NULL, 0, 0,
|
|
OPAL_INFO_LVL_9,
|
|
MCA_BASE_VAR_SCOPE_READONLY,
|
|
&cs->adapt_context_free_list_max);
|
|
|
|
cs->adapt_context_free_list_inc = 32;
|
|
(void) mca_base_component_var_register(c, "context_free_list_inc",
|
|
"Increasement number of segments in context free list",
|
|
MCA_BASE_VAR_TYPE_INT, NULL, 0, 0,
|
|
OPAL_INFO_LVL_9,
|
|
MCA_BASE_VAR_SCOPE_READONLY,
|
|
&cs->adapt_context_free_list_inc);
|
|
ompi_coll_adapt_ibcast_register();
|
|
ompi_coll_adapt_ireduce_register();
|
|
|
|
return adapt_verify_mca_variables();
|
|
}
|