ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
c992e99035
openmpi/ompi/mca/coll/tuned/coll_tuned_module.c
Jeff Squyres e7ecd56bd2 This commit represents a bunch of work on a Mercurial side branch. As 
such, the commit message back to the master SVN repository is fairly
long.

= ORTE Job-Level Output Messages =

Add two new interfaces that should be used for all new code throughout
the ORTE and OMPI layers (we already make the search-and-replace on
the existing ORTE / OMPI layers):

 * orte_output(): (and corresponding friends ORTE_OUTPUT,
   orte_output_verbose, etc.)  This function sends the output directly
   to the HNP for processing as part of a job-specific output
   channel.  It supports all the same outputs as opal_output()
   (syslog, file, stdout, stderr), but for stdout/stderr, the output
   is sent to the HNP for processing and output.  More on this below.
 * orte_show_help(): This function is a drop-in-replacement for
   opal_show_help(), with two differences in functionality:
   1. the rendered text help message output is sent to the HNP for
      display (rather than outputting directly into the process' stderr
      stream)
   1. the HNP detects duplicate help messages and does not display them
      (so that you don't see the same error message N times, once from
      each of your N MPI processes); instead, it counts "new" instances
      of the help message and displays a message every ~5 seconds when
      there are new ones ("I got X new copies of the help message...")

opal_show_help and opal_output still exist, but they only output in
the current process.  The intent for the new orte_* functions is that
they can apply job-level intelligence to the output.  As such, we
recommend that all new ORTE and OMPI code use the new orte_*
functions, not thei opal_* functions.

=== New code ===

For ORTE and OMPI programmers, here's what you need to do differently
in new code:

 * Do not include opal/util/show_help.h or opal/util/output.h.
   Instead, include orte/util/output.h (this one header file has
   declarations for both the orte_output() series of functions and
   orte_show_help()).
 * Effectively s/opal_output/orte_output/gi throughout your code.
   Note that orte_output_open() takes a slightly different argument
   list (as a way to pass data to the filtering stream -- see below),
   so you if explicitly call opal_output_open(), you'll need to
   slightly adapt to the new signature of orte_output_open().
 * Literally s/opal_show_help/orte_show_help/.  The function signature
   is identical.

=== Notes ===

 * orte_output'ing to stream 0 will do similar to what
   opal_output'ing did, so leaving a hard-coded "0" as the first
   argument is safe.
 * For systems that do not use ORTE's RML or the HNP, the effect of
   orte_output_* and orte_show_help will be identical to their opal
   counterparts (the additional information passed to
   orte_output_open() will be lost!).  Indeed, the orte_* functions
   simply become trivial wrappers to their opal_* counterparts.  Note
   that we have not tested this; the code is simple but it is quite
   possible that we mucked something up.

= Filter Framework =

Messages sent view the new orte_* functions described above and
messages output via the IOF on the HNP will now optionally be passed
through a new "filter" framework before being output to
stdout/stderr.  The "filter" OPAL MCA framework is intended to allow
preprocessing to messages before they are sent to their final
destinations.  The first component that was written in the filter
framework was to create an XML stream, segregating all the messages
into different XML tags, etc.  This will allow 3rd party tools to read
the stdout/stderr from the HNP and be able to know exactly what each
text message is (e.g., a help message, another OMPI infrastructure
message, stdout from the user process, stderr from the user process,
etc.).

Filtering is not active by default.  Filter components must be
specifically requested, such as:

{{{
$ mpirun --mca filter xml ...
}}}

There can only be one filter component active.

= New MCA Parameters =

The new functionality described above introduces two new MCA
parameters:

 * '''orte_base_help_aggregate''': Defaults to 1 (true), meaning that
   help messages will be aggregated, as described above.  If set to 0,
   all help messages will be displayed, even if they are duplicates
   (i.e., the original behavior).
 * '''orte_base_show_output_recursions''': An MCA parameter to help
   debug one of the known issues, described below.  It is likely that
   this MCA parameter will disappear before v1.3 final.

= Known Issues =

 * The XML filter component is not complete.  The current output from
   this component is preliminary and not real XML.  A bit more work
   needs to be done to configure.m4 search for an appropriate XML
   library/link it in/use it at run time.
 * There are possible recursion loops in the orte_output() and
   orte_show_help() functions -- e.g., if RML send calls orte_output()
   or orte_show_help().  We have some ideas how to fix these, but
   figured that it was ok to commit before feature freeze with known
   issues.  The code currently contains sub-optimal workarounds so
   that this will not be a problem, but it would be good to actually
   solve the problem rather than have hackish workarounds before v1.3 final.

This commit was SVN r18434.
2008-05-13 20:00:55 +00:00

376 строки
15 KiB
C
Исходник Ответственный История

/*
* 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 (c) 2008 Sun Microsystems, Inc. All rights reserved.
* $COPYRIGHT$
*
* Additional copyrights may follow
*
* $HEADER$
*/
#include "ompi_config.h"
#include "coll_tuned.h"
#include <stdio.h>
#include "mpi.h"
#include "ompi/communicator/communicator.h"
#include "opal/mca/base/mca_base_param.h"
#include "ompi/mca/coll/coll.h"
#include "ompi/mca/coll/base/base.h"
#include "coll_tuned.h"
#include "coll_tuned_topo.h"
#include "coll_tuned_dynamic_rules.h"
#include "coll_tuned_dynamic_file.h"
#include "coll_tuned_forced.h"
static int tuned_module_enable(struct mca_coll_base_module_1_1_0_t *module,
struct ompi_communicator_t *comm);
/*
* Initial query function that is invoked during MPI_INIT, allowing
* this component to disqualify itself if it doesn't support the
* required level of thread support.
*/
int ompi_coll_tuned_init_query(bool enable_progress_threads,
bool enable_mpi_threads)
{
/* Nothing to do */
return OMPI_SUCCESS;
}
/*
* Invoked when there's a new communicator that has been created.
* Look at the communicator and decide which set of functions and
* priority we want to return.
*/
mca_coll_base_module_1_1_0_t *
ompi_coll_tuned_comm_query(struct ompi_communicator_t *comm, int *priority)
{
mca_coll_tuned_module_t *tuned_module;
ORTE_OUTPUT((ompi_coll_tuned_stream, "coll:tuned:module_tuned query called"));
/**
* If it is inter-communicator and size is less than 2 we have specialized modules
* to handle the intra collective communications.
*/
if (OMPI_COMM_IS_INTRA(comm) && ompi_comm_size(comm) < 2) {
*priority = 0;
return NULL;
}
if (OMPI_COMM_IS_INTER(comm)) {
#if 0
if (ompi_coll_tuned_use_dynamic_rules) {
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_query using inter_dynamic"));
} else {
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_query using inter_fixed"));
}
#endif
/* tuned does not support inter-communicator yet */
*priority = 0;
return NULL;
}
tuned_module = OBJ_NEW(mca_coll_tuned_module_t);
if (NULL == tuned_module) return NULL;
*priority = ompi_coll_tuned_priority;
/*
* Choose whether to use [intra|inter] decision functions
* and if using fixed OR dynamic rule sets.
* Right now you cannot mix them, maybe later on it can be changed
* but this would probably add an extra if and funct call to the path
*/
tuned_module->super.coll_module_enable = tuned_module_enable;
tuned_module->super.ft_event = mca_coll_tuned_ft_event;
if (ompi_coll_tuned_use_dynamic_rules) {
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_query using intra_dynamic"));
tuned_module->super.coll_allgather = ompi_coll_tuned_allgather_intra_dec_dynamic;
tuned_module->super.coll_allgatherv = ompi_coll_tuned_allgatherv_intra_dec_dynamic;
tuned_module->super.coll_allreduce = ompi_coll_tuned_allreduce_intra_dec_dynamic;
tuned_module->super.coll_alltoall = ompi_coll_tuned_alltoall_intra_dec_dynamic;
tuned_module->super.coll_alltoallv = ompi_coll_tuned_alltoallv_intra_dec_dynamic;
tuned_module->super.coll_alltoallw = NULL;
tuned_module->super.coll_barrier = ompi_coll_tuned_barrier_intra_dec_dynamic;
tuned_module->super.coll_bcast = ompi_coll_tuned_bcast_intra_dec_dynamic;
tuned_module->super.coll_exscan = NULL;
tuned_module->super.coll_gather = ompi_coll_tuned_gather_intra_dec_dynamic;
tuned_module->super.coll_gatherv = NULL;
tuned_module->super.coll_reduce = ompi_coll_tuned_reduce_intra_dec_dynamic;
tuned_module->super.coll_reduce_scatter = ompi_coll_tuned_reduce_scatter_intra_dec_dynamic;
tuned_module->super.coll_scan = NULL;
tuned_module->super.coll_scatter = ompi_coll_tuned_scatter_intra_dec_dynamic;
tuned_module->super.coll_scatterv = NULL;
} else {
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_query using intra_fixed"));
tuned_module->super.coll_allgather = ompi_coll_tuned_allgather_intra_dec_fixed;
tuned_module->super.coll_allgatherv = ompi_coll_tuned_allgatherv_intra_dec_fixed;
tuned_module->super.coll_allreduce = ompi_coll_tuned_allreduce_intra_dec_fixed;
tuned_module->super.coll_alltoall = ompi_coll_tuned_alltoall_intra_dec_fixed;
tuned_module->super.coll_alltoallv = ompi_coll_tuned_alltoallv_intra_dec_fixed;
tuned_module->super.coll_alltoallw = NULL;
tuned_module->super.coll_barrier = ompi_coll_tuned_barrier_intra_dec_fixed;
tuned_module->super.coll_bcast = ompi_coll_tuned_bcast_intra_dec_fixed;
tuned_module->super.coll_exscan = NULL;
tuned_module->super.coll_gather = ompi_coll_tuned_gather_intra_dec_fixed;
tuned_module->super.coll_gatherv = NULL;
tuned_module->super.coll_reduce = ompi_coll_tuned_reduce_intra_dec_fixed;
tuned_module->super.coll_reduce_scatter = ompi_coll_tuned_reduce_scatter_intra_dec_fixed;
tuned_module->super.coll_scan = NULL;
tuned_module->super.coll_scatter = ompi_coll_tuned_scatter_intra_dec_fixed;
tuned_module->super.coll_scatterv = NULL;
}
return &(tuned_module->super);
}
/*
* Init module on the communicator
*/
static int
tuned_module_enable(struct mca_coll_base_module_1_1_0_t *module,
struct ompi_communicator_t *comm)
{
int size;
mca_coll_tuned_module_t *tuned_module = (mca_coll_tuned_module_t *) module;
mca_coll_tuned_comm_t *data = NULL;
/* fanout parameters */
int rc=0;
int i;
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_init called."));
/* This routine will become more complex and might have to be */
/* broken into more sections/function calls */
/* Order of operations:
* alloc memory for nb reqs (in case we fall through)
* add decision rules if using dynamic rules
* compact rules using communicator size info etc
* build first guess cached topologies (might depend on the rules from above)
*
* then attach all to the communicator and return base module funct ptrs
*/
/* Allocate the data that hangs off the communicator */
if (OMPI_COMM_IS_INTER(comm)) {
size = ompi_comm_remote_size(comm);
} else {
size = ompi_comm_size(comm);
}
/*
* we still malloc data as it is used by the TUNED modules
* if we don't allocate it and fall back to a BASIC module routine then confuses debuggers
* we place any special info after the default data
*
* BUT on very large systems we might not be able to allocate all this memory so
* we do check a MCA parameter to see if if we should allocate this memory
*
* The default is set very high
*
*/
/* if we within the memory/size limit, allow preallocated data */
if (size<=ompi_coll_tuned_preallocate_memory_comm_size_limit) {
data = (mca_coll_tuned_comm_t*)malloc(sizeof(struct mca_coll_tuned_comm_t) +
(sizeof(ompi_request_t *) * size * 2));
if (NULL == data) {
return OMPI_ERROR;
}
data->mcct_reqs = (ompi_request_t **) (data + 1);
data->mcct_num_reqs = size * 2;
} else {
data = (mca_coll_tuned_comm_t*)malloc(sizeof(struct mca_coll_tuned_comm_t));
if (NULL == data) {
return OMPI_ERROR;
}
data->mcct_reqs = (ompi_request_t **) NULL;
data->mcct_num_reqs = 0;
}
/*
* If using dynamic and you are MPI_COMM_WORLD and you want to use a parameter file..
* then this effects how much storage space you need
* (This is a basic version of what will go into V2)
*
*/
size = ompi_comm_size(comm); /* find size so we can (A) decide if to access the file directly */
/* (B) so we can get our very own customised ompi_coll_com_rule_t ptr */
/* which only has rules in it for our com size */
/* actually if they are below a threadhold, they all open it */
/* have to build a collective in here.. but just for MCW.. */
/* but we have to make sure we have the same rules everywhere :( */
/* if using dynamic rules make sure all overrides are NULL before we start override anything accidently */
if (ompi_coll_tuned_use_dynamic_rules) {
/* base rules */
data->all_base_rules = (ompi_coll_alg_rule_t*) NULL;
/* each collective rule for my com size */
for (i=0;i<COLLCOUNT;i++) {
data->com_rules[i] = (ompi_coll_com_rule_t*) NULL;
}
}
/* next dynamic state, recheck all forced rules as well */
/* warning, we should check to make sure this is really an INTRA comm here... */
if (ompi_coll_tuned_use_dynamic_rules) {
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[ALLREDUCE], &(data->user_forced[ALLREDUCE]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[ALLTOALL], &(data->user_forced[ALLTOALL]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[ALLGATHER], &(data->user_forced[ALLGATHER]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[ALLGATHERV], &(data->user_forced[ALLGATHERV]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[ALLTOALLV], &(data->user_forced[ALLTOALLV]));
ompi_coll_tuned_forced_getvalues_barrier (ompi_coll_tuned_forced_params[BARRIER], &(data->user_forced[BARRIER]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[BCAST], &(data->user_forced[BCAST]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[REDUCE], &(data->user_forced[REDUCE]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[REDUCESCATTER], &(data->user_forced[REDUCESCATTER]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[GATHER], &(data->user_forced[GATHER]));
ompi_coll_tuned_forced_getvalues (ompi_coll_tuned_forced_params[SCATTER], &(data->user_forced[SCATTER]));
}
if (&ompi_mpi_comm_world==comm) {
if (ompi_coll_tuned_use_dynamic_rules) {
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_init MCW & Dynamic"));
if (ompi_coll_tuned_dynamic_rules_filename) {
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_init Opening [%s]",
ompi_coll_tuned_dynamic_rules_filename));
rc = ompi_coll_tuned_read_rules_config_file (ompi_coll_tuned_dynamic_rules_filename,
&(data->all_base_rules), COLLCOUNT);
if (rc>=0) {
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_init Read %d valid rules\n", rc));
/* at this point we all have a base set of rules */
/* now we can get our customized communicator sized rule set, for each collective */
for (i=0;i<COLLCOUNT;i++) {
data->com_rules[i] = ompi_coll_tuned_get_com_rule_ptr (data->all_base_rules, i, size);
}
}
else { /* failed to read config file, thus make sure its a NULL... */
data->all_base_rules = (ompi_coll_alg_rule_t*) NULL;
}
} /* end if a config filename exists */
} /* end if dynamic_rules */
} /* end if MCW */
/* ok, if using dynamic rules, not MCW and we are just any rank and a base set of rules exist.. ref them */
/* order of eval is important here, if we are MCW ompi_mpi_comm_world.c_coll_selected_data is NULL still.. */
#if 0 /* FIXME: don't know how to deal with this */
if ((ompi_coll_tuned_use_dynamic_rules)&&(!(&ompi_mpi_comm_world==comm))&&
((ompi_mpi_comm_world.c_coll_selected_data)->all_base_rules)) {
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_init NOT MCW & Dynamic"));
/* this will, erm fail if MCW doesn't exist which it should! */
data->all_base_rules = (ompi_mpi_comm_world.c_coll_selected_data)->all_base_rules;
/* at this point we all have a base set of rules if they exist atall */
/* now we can get our customized communicator sized rule set, for each collective */
for (i=0;i<COLLCOUNT;i++) {
data->com_rules[i] = ompi_coll_tuned_get_com_rule_ptr (data->all_base_rules, i, size);
}
}
#endif
/*
* now for the cached topo functions
* guess the initial topologies to use rank 0 as root
*/
/* general n fan out tree */
data->cached_ntree = ompi_coll_tuned_topo_build_tree (ompi_coll_tuned_init_tree_fanout,
comm, 0);
data->cached_ntree_root = 0;
data->cached_ntree_fanout = ompi_coll_tuned_init_tree_fanout;
/* binary tree */
data->cached_bintree = ompi_coll_tuned_topo_build_tree (2, comm, 0);
data->cached_bintree_root = 0;
/* binomial tree */
data->cached_bmtree = ompi_coll_tuned_topo_build_bmtree (comm, 0);
data->cached_bmtree_root = 0;
/* binomial tree */
data->cached_in_order_bmtree = ompi_coll_tuned_topo_build_in_order_bmtree (comm, 0);
data->cached_in_order_bmtree_root = 0;
/*
* chains (fanout followed by pipelines)
* are more difficuilt as the fan out really really depends on message size [sometimes]..
* as size gets larger fan-out gets smaller [usually]
*
* will probably change how we cache this later, for now a midsize
* GEF
*/
data->cached_chain = ompi_coll_tuned_topo_build_chain (ompi_coll_tuned_init_chain_fanout,
comm, 0);
data->cached_chain_root = 0;
data->cached_chain_fanout = ompi_coll_tuned_init_chain_fanout;
/* standard pipeline */
data->cached_pipeline = ompi_coll_tuned_topo_build_chain (1, comm, 0);
data->cached_pipeline_root = 0;
/* in-order binary tree */
data->cached_in_order_bintree = ompi_coll_tuned_topo_build_in_order_bintree(comm);
/* All done */
tuned_module->tuned_data = data;
ORTE_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:module_init Tuned is in use"));
return OMPI_SUCCESS;
}
int mca_coll_tuned_ft_event(int state) {
if(OPAL_CRS_CHECKPOINT == state) {
;
}
else if(OPAL_CRS_CONTINUE == state) {
;
}
else if(OPAL_CRS_RESTART == state) {
;
}
else if(OPAL_CRS_TERM == state ) {
;
}
else {
;
}
return OMPI_SUCCESS;
}
Показать git blame Копировать постоянную ссылку