2004-08-02 04:24:22 +04:00
|
|
|
/*
|
2011-07-11 03:32:23 +04:00
|
|
|
* Copyright (c) 2004-2011 The Trustees of Indiana University and Indiana
|
2005-11-05 22:57:48 +03:00
|
|
|
* 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.
|
2004-11-28 23:09:25 +03:00
|
|
|
* Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
|
|
|
|
* University of Stuttgart. All rights reserved.
|
2005-03-24 15:43:37 +03:00
|
|
|
* Copyright (c) 2004-2005 The Regents of the University of California.
|
|
|
|
* All rights reserved.
|
Per RFC, bring in the following changes:
* Remove paffinity, maffinity, and carto frameworks -- they've been
wholly replaced by hwloc.
* Move ompi_mpi_init() affinity-setting/checking code down to ORTE.
* Update sm, smcuda, wv, and openib components to no longer use carto.
Instead, use hwloc data. There are still optimizations possible in
the sm/smcuda BTLs (i.e., making multiple mpools). Also, the old
carto-based code found out how many NUMA nodes were ''available''
-- not how many were used ''in this job''. The new hwloc-using
code computes the same value -- it was not updated to calculate how
many NUMA nodes are used ''by this job.''
* Note that I cannot compile the smcuda and wv BTLs -- I ''think''
they're right, but they need to be verified by their owners.
* The openib component now does a bunch of stuff to figure out where
"near" OpenFabrics devices are. '''THIS IS A CHANGE IN DEFAULT
BEHAVIOR!!''' and still needs to be verified by OpenFabrics vendors
(I do not have a NUMA machine with an OpenFabrics device that is a
non-uniform distance from multiple different NUMA nodes).
* Completely rewrite the OMPI_Affinity_str() routine from the
"affinity" mpiext extension. This extension now understands
hyperthreads; the output format of it has changed a bit to reflect
this new information.
* Bunches of minor changes around the code base to update names/types
from maffinity/paffinity-based names to hwloc-based names.
* Add some helper functions into the hwloc base, mainly having to do
with the fact that we have the hwloc data reporting ''all''
topology information, but sometimes you really only want the
(online | available) data.
This commit was SVN r26391.
2012-05-07 18:52:54 +04:00
|
|
|
* Copyright (c) 2009-2012 Cisco Systems, Inc. All rights reserved.
|
|
|
|
* Copyright (c) 2011-2012 Los Alamos National Security, LLC.
|
2010-06-09 20:58:52 +04:00
|
|
|
* All rights reserved.
|
2011-08-04 14:15:45 +04:00
|
|
|
* Copyright (c) 2011 NVIDIA Corporation. All rights reserved.
|
2004-11-22 04:38:40 +03:00
|
|
|
* $COPYRIGHT$
|
|
|
|
*
|
|
|
|
* Additional copyrights may follow
|
|
|
|
*
|
2004-08-02 04:24:22 +04:00
|
|
|
* $HEADER$
|
|
|
|
*/
|
|
|
|
|
2004-10-20 05:03:09 +04:00
|
|
|
#include "ompi_config.h"
|
2004-08-02 04:24:22 +04:00
|
|
|
#include <string.h>
|
2006-02-12 04:33:29 +03:00
|
|
|
#include "ompi/mca/mpool/sm/mpool_sm.h"
|
2010-06-09 20:58:52 +04:00
|
|
|
#include "ompi/mca/common/sm/common_sm.h"
|
2011-08-04 14:15:45 +04:00
|
|
|
#include "ompi/mca/common/cuda/common_cuda.h"
|
2008-06-27 20:25:31 +04:00
|
|
|
#ifdef HAVE_UNISTD_H
|
2008-06-15 17:43:28 +04:00
|
|
|
#include <unistd.h>
|
2008-06-27 20:25:31 +04:00
|
|
|
#endif
|
Per RFC, bring in the following changes:
* Remove paffinity, maffinity, and carto frameworks -- they've been
wholly replaced by hwloc.
* Move ompi_mpi_init() affinity-setting/checking code down to ORTE.
* Update sm, smcuda, wv, and openib components to no longer use carto.
Instead, use hwloc data. There are still optimizations possible in
the sm/smcuda BTLs (i.e., making multiple mpools). Also, the old
carto-based code found out how many NUMA nodes were ''available''
-- not how many were used ''in this job''. The new hwloc-using
code computes the same value -- it was not updated to calculate how
many NUMA nodes are used ''by this job.''
* Note that I cannot compile the smcuda and wv BTLs -- I ''think''
they're right, but they need to be verified by their owners.
* The openib component now does a bunch of stuff to figure out where
"near" OpenFabrics devices are. '''THIS IS A CHANGE IN DEFAULT
BEHAVIOR!!''' and still needs to be verified by OpenFabrics vendors
(I do not have a NUMA machine with an OpenFabrics device that is a
non-uniform distance from multiple different NUMA nodes).
* Completely rewrite the OMPI_Affinity_str() routine from the
"affinity" mpiext extension. This extension now understands
hyperthreads; the output format of it has changed a bit to reflect
this new information.
* Bunches of minor changes around the code base to update names/types
from maffinity/paffinity-based names to hwloc-based names.
* Add some helper functions into the hwloc base, mainly having to do
with the fact that we have the hwloc data reporting ''all''
topology information, but sometimes you really only want the
(online | available) data.
This commit was SVN r26391.
2012-05-07 18:52:54 +04:00
|
|
|
#include "opal/mca/hwloc/base/base.h"
|
2004-08-02 04:24:22 +04:00
|
|
|
|
2010-03-13 02:57:50 +03:00
|
|
|
#if OPAL_ENABLE_FT_CR == 1
|
A number of C/R enhancements per RFC below:
http://www.open-mpi.org/community/lists/devel/2010/07/8240.php
Documentation:
http://osl.iu.edu/research/ft/
Major Changes:
--------------
* Added C/R-enabled Debugging support.
Enabled with the --enable-crdebug flag. See the following website for more information:
http://osl.iu.edu/research/ft/crdebug/
* Added Stable Storage (SStore) framework for checkpoint storage
* 'central' component does a direct to central storage save
* 'stage' component stages checkpoints to central storage while the application continues execution.
* 'stage' supports offline compression of checkpoints before moving (sstore_stage_compress)
* 'stage' supports local caching of checkpoints to improve automatic recovery (sstore_stage_caching)
* Added Compression (compress) framework to support
* Add two new ErrMgr recovery policies
* {{{crmig}}} C/R Process Migration
* {{{autor}}} C/R Automatic Recovery
* Added the {{{ompi-migrate}}} command line tool to support the {{{crmig}}} ErrMgr component
* Added CR MPI Ext functions (enable them with {{{--enable-mpi-ext=cr}}} configure option)
* {{{OMPI_CR_Checkpoint}}} (Fixes trac:2342)
* {{{OMPI_CR_Restart}}}
* {{{OMPI_CR_Migrate}}} (may need some more work for mapping rules)
* {{{OMPI_CR_INC_register_callback}}} (Fixes trac:2192)
* {{{OMPI_CR_Quiesce_start}}}
* {{{OMPI_CR_Quiesce_checkpoint}}}
* {{{OMPI_CR_Quiesce_end}}}
* {{{OMPI_CR_self_register_checkpoint_callback}}}
* {{{OMPI_CR_self_register_restart_callback}}}
* {{{OMPI_CR_self_register_continue_callback}}}
* The ErrMgr predicted_fault() interface has been changed to take an opal_list_t of ErrMgr defined types. This will allow us to better support a wider range of fault prediction services in the future.
* Add a progress meter to:
* FileM rsh (filem_rsh_process_meter)
* SnapC full (snapc_full_progress_meter)
* SStore stage (sstore_stage_progress_meter)
* Added 2 new command line options to ompi-restart
* --showme : Display the full command line that would have been exec'ed.
* --mpirun_opts : Command line options to pass directly to mpirun. (Fixes trac:2413)
* Deprecated some MCA params:
* crs_base_snapshot_dir deprecated, use sstore_stage_local_snapshot_dir
* snapc_base_global_snapshot_dir deprecated, use sstore_base_global_snapshot_dir
* snapc_base_global_shared deprecated, use sstore_stage_global_is_shared
* snapc_base_store_in_place deprecated, replaced with different components of SStore
* snapc_base_global_snapshot_ref deprecated, use sstore_base_global_snapshot_ref
* snapc_base_establish_global_snapshot_dir deprecated, never well supported
* snapc_full_skip_filem deprecated, use sstore_stage_skip_filem
Minor Changes:
--------------
* Fixes trac:1924 : {{{ompi-restart}}} now recognizes path prefixed checkpoint handles and does the right thing.
* Fixes trac:2097 : {{{ompi-info}}} should now report all available CRS components
* Fixes trac:2161 : Manual checkpoint movement. A user can 'mv' a checkpoint directory from the original location to another and still restart from it.
* Fixes trac:2208 : Honor various TMPDIR varaibles instead of forcing {{{/tmp}}}
* Move {{{ompi_cr_continue_like_restart}}} to {{{orte_cr_continue_like_restart}}} to be more flexible in where this should be set.
* opal_crs_base_metadata_write* functions have been moved to SStore to support a wider range of metadata handling functionality.
* Cleanup the CRS framework and components to work with the SStore framework.
* Cleanup the SnapC framework and components to work with the SStore framework (cleans up these code paths considerably).
* Add 'quiesce' hook to CRCP for a future enhancement.
* We now require a BLCR version that supports {{{cr_request_file()}}} or {{{cr_request_checkpoint()}}} in order to make the code more maintainable. Note that {{{cr_request_file}}} has been deprecated since 0.7.0, so we prefer to use {{{cr_request_checkpoint()}}}.
* Add optional application level INC callbacks (registered through the CR MPI Ext interface).
* Increase the {{{opal_cr_thread_sleep_wait}}} parameter to 1000 microseconds to make the C/R thread less aggressive.
* {{{opal-restart}}} now looks for cache directories before falling back on stable storage when asked.
* {{{opal-restart}}} also support local decompression before restarting
* {{{orte-checkpoint}}} now uses the SStore framework to work with the metadata
* {{{orte-restart}}} now uses the SStore framework to work with the metadata
* Remove the {{{orte-restart}}} preload option. This was removed since the user only needs to select the 'stage' component in order to support this functionality.
* Since the '-am' parameter is saved in the metadata, {{{ompi-restart}}} no longer hard codes {{{-am ft-enable-cr}}}.
* Fix {{{hnp}}} ErrMgr so that if a previous component in the stack has 'fixed' the problem, then it should be skipped.
* Make sure to decrement the number of 'num_local_procs' in the orted when one goes away.
* odls now checks the SStore framework to see if it needs to load any checkpoint files before launching (to support 'stage'). This separates the SStore logic from the --preload-[binary|files] options.
* Add unique IDs to the named pipes established between the orted and the app in SnapC. This is to better support migration and automatic recovery activities.
* Improve the checks for 'already checkpointing' error path.
* A a recovery output timer, to show how long it takes to restart a job
* Do a better job of cleaning up the old session directory on restart.
* Add a local module to the autor and crmig ErrMgr components. These small modules prevent the 'orted' component from attempting a local recovery (Which does not work for MPI apps at the moment)
* Add a fix for bounding the checkpointable region between MPI_Init and MPI_Finalize.
This commit was SVN r23587.
The following Trac tickets were found above:
Ticket 1924 --> https://svn.open-mpi.org/trac/ompi/ticket/1924
Ticket 2097 --> https://svn.open-mpi.org/trac/ompi/ticket/2097
Ticket 2161 --> https://svn.open-mpi.org/trac/ompi/ticket/2161
Ticket 2192 --> https://svn.open-mpi.org/trac/ompi/ticket/2192
Ticket 2208 --> https://svn.open-mpi.org/trac/ompi/ticket/2208
Ticket 2342 --> https://svn.open-mpi.org/trac/ompi/ticket/2342
Ticket 2413 --> https://svn.open-mpi.org/trac/ompi/ticket/2413
2010-08-11 00:51:11 +04:00
|
|
|
#include "orte/mca/sstore/sstore.h"
|
2008-10-16 19:09:00 +04:00
|
|
|
#include "ompi/mca/mpool/base/base.h"
|
|
|
|
#include "ompi/runtime/ompi_cr.h"
|
|
|
|
#endif
|
2004-08-02 04:24:22 +04:00
|
|
|
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
static void sm_module_finalize(mca_mpool_base_module_t* module);
|
|
|
|
|
2005-05-31 21:06:55 +04:00
|
|
|
/*
|
|
|
|
* Initializes the mpool module.
|
|
|
|
*/
|
|
|
|
void mca_mpool_sm_module_init(mca_mpool_sm_module_t* mpool)
|
|
|
|
{
|
2008-08-12 20:58:32 +04:00
|
|
|
mpool->super.mpool_component = &mca_mpool_sm_component.super;
|
|
|
|
mpool->super.mpool_base = mca_mpool_sm_base;
|
|
|
|
mpool->super.mpool_alloc = mca_mpool_sm_alloc;
|
|
|
|
mpool->super.mpool_realloc = mca_mpool_sm_realloc;
|
|
|
|
mpool->super.mpool_free = mca_mpool_sm_free;
|
|
|
|
mpool->super.mpool_find = NULL;
|
|
|
|
mpool->super.mpool_register = NULL;
|
|
|
|
mpool->super.mpool_deregister = NULL;
|
|
|
|
mpool->super.mpool_release_memory = NULL;
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
mpool->super.mpool_finalize = sm_module_finalize;
|
2008-08-12 20:58:32 +04:00
|
|
|
mpool->super.mpool_ft_event = mca_mpool_sm_ft_event;
|
|
|
|
mpool->super.flags = 0;
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
|
|
|
|
mpool->sm_size = 0;
|
|
|
|
mpool->sm_allocator = NULL;
|
|
|
|
mpool->sm_mmap = NULL;
|
2010-06-09 20:58:52 +04:00
|
|
|
mpool->sm_common_module = NULL;
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
mpool->mem_node = -1;
|
2005-05-31 21:06:55 +04:00
|
|
|
}
|
2004-08-02 04:24:22 +04:00
|
|
|
|
|
|
|
/*
|
|
|
|
* base address of shared memory mapping
|
|
|
|
*/
|
2005-05-31 21:06:55 +04:00
|
|
|
void* mca_mpool_sm_base(mca_mpool_base_module_t* mpool)
|
2004-08-02 04:24:22 +04:00
|
|
|
{
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
mca_mpool_sm_module_t *sm_mpool = (mca_mpool_sm_module_t*) mpool;
|
2010-06-09 20:58:52 +04:00
|
|
|
return (NULL != sm_mpool->sm_common_module) ?
|
|
|
|
sm_mpool->sm_common_module->module_seg_addr : NULL;
|
2004-08-02 04:24:22 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2008-06-15 17:43:28 +04:00
|
|
|
* allocate function
|
2004-08-02 04:24:22 +04:00
|
|
|
*/
|
2005-06-21 21:10:28 +04:00
|
|
|
void* mca_mpool_sm_alloc(
|
2008-06-15 17:43:28 +04:00
|
|
|
mca_mpool_base_module_t* mpool,
|
|
|
|
size_t size,
|
|
|
|
size_t align,
|
2005-09-13 02:28:23 +04:00
|
|
|
uint32_t flags,
|
2005-06-25 01:12:38 +04:00
|
|
|
mca_mpool_base_registration_t** registration)
|
2004-08-02 04:24:22 +04:00
|
|
|
{
|
2008-06-15 17:43:28 +04:00
|
|
|
mca_mpool_sm_module_t* mpool_sm = (mca_mpool_sm_module_t*)mpool;
|
Per RFC, bring in the following changes:
* Remove paffinity, maffinity, and carto frameworks -- they've been
wholly replaced by hwloc.
* Move ompi_mpi_init() affinity-setting/checking code down to ORTE.
* Update sm, smcuda, wv, and openib components to no longer use carto.
Instead, use hwloc data. There are still optimizations possible in
the sm/smcuda BTLs (i.e., making multiple mpools). Also, the old
carto-based code found out how many NUMA nodes were ''available''
-- not how many were used ''in this job''. The new hwloc-using
code computes the same value -- it was not updated to calculate how
many NUMA nodes are used ''by this job.''
* Note that I cannot compile the smcuda and wv BTLs -- I ''think''
they're right, but they need to be verified by their owners.
* The openib component now does a bunch of stuff to figure out where
"near" OpenFabrics devices are. '''THIS IS A CHANGE IN DEFAULT
BEHAVIOR!!''' and still needs to be verified by OpenFabrics vendors
(I do not have a NUMA machine with an OpenFabrics device that is a
non-uniform distance from multiple different NUMA nodes).
* Completely rewrite the OMPI_Affinity_str() routine from the
"affinity" mpiext extension. This extension now understands
hyperthreads; the output format of it has changed a bit to reflect
this new information.
* Bunches of minor changes around the code base to update names/types
from maffinity/paffinity-based names to hwloc-based names.
* Add some helper functions into the hwloc base, mainly having to do
with the fact that we have the hwloc data reporting ''all''
topology information, but sometimes you really only want the
(online | available) data.
This commit was SVN r26391.
2012-05-07 18:52:54 +04:00
|
|
|
opal_hwloc_base_memory_segment_t mseg;
|
2008-06-15 17:43:28 +04:00
|
|
|
|
|
|
|
mseg.mbs_start_addr =
|
2009-02-20 22:51:57 +03:00
|
|
|
mpool_sm->sm_allocator->alc_alloc(mpool_sm->sm_allocator, size, align, registration);
|
2008-06-15 17:43:28 +04:00
|
|
|
|
|
|
|
if(mpool_sm->mem_node >= 0) {
|
|
|
|
mseg.mbs_len = size;
|
2012-09-27 03:24:27 +04:00
|
|
|
#if OPAL_HAVE_HWLOC
|
Per RFC, bring in the following changes:
* Remove paffinity, maffinity, and carto frameworks -- they've been
wholly replaced by hwloc.
* Move ompi_mpi_init() affinity-setting/checking code down to ORTE.
* Update sm, smcuda, wv, and openib components to no longer use carto.
Instead, use hwloc data. There are still optimizations possible in
the sm/smcuda BTLs (i.e., making multiple mpools). Also, the old
carto-based code found out how many NUMA nodes were ''available''
-- not how many were used ''in this job''. The new hwloc-using
code computes the same value -- it was not updated to calculate how
many NUMA nodes are used ''by this job.''
* Note that I cannot compile the smcuda and wv BTLs -- I ''think''
they're right, but they need to be verified by their owners.
* The openib component now does a bunch of stuff to figure out where
"near" OpenFabrics devices are. '''THIS IS A CHANGE IN DEFAULT
BEHAVIOR!!''' and still needs to be verified by OpenFabrics vendors
(I do not have a NUMA machine with an OpenFabrics device that is a
non-uniform distance from multiple different NUMA nodes).
* Completely rewrite the OMPI_Affinity_str() routine from the
"affinity" mpiext extension. This extension now understands
hyperthreads; the output format of it has changed a bit to reflect
this new information.
* Bunches of minor changes around the code base to update names/types
from maffinity/paffinity-based names to hwloc-based names.
* Add some helper functions into the hwloc base, mainly having to do
with the fact that we have the hwloc data reporting ''all''
topology information, but sometimes you really only want the
(online | available) data.
This commit was SVN r26391.
2012-05-07 18:52:54 +04:00
|
|
|
opal_hwloc_base_membind(&mseg, 1, mpool_sm->mem_node);
|
2012-09-27 03:24:27 +04:00
|
|
|
#endif
|
2008-06-15 17:43:28 +04:00
|
|
|
}
|
|
|
|
|
2011-08-04 14:15:45 +04:00
|
|
|
#if OPAL_CUDA_SUPPORT
|
|
|
|
if (flags & MCA_MPOOL_FLAGS_CUDA_REGISTER_MEM) {
|
|
|
|
mca_common_cuda_register(mseg.mbs_start_addr, size,
|
|
|
|
mpool->mpool_component->mpool_version.mca_component_name);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2008-06-15 17:43:28 +04:00
|
|
|
return mseg.mbs_start_addr;
|
2004-08-02 04:24:22 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2008-06-15 17:43:28 +04:00
|
|
|
* realloc function
|
2004-08-02 04:24:22 +04:00
|
|
|
*/
|
2005-06-21 21:10:28 +04:00
|
|
|
void* mca_mpool_sm_realloc(
|
2008-06-15 17:43:28 +04:00
|
|
|
mca_mpool_base_module_t* mpool,
|
|
|
|
void* addr,
|
|
|
|
size_t size,
|
2005-06-25 01:12:38 +04:00
|
|
|
mca_mpool_base_registration_t** registration)
|
2004-08-02 04:24:22 +04:00
|
|
|
{
|
2008-06-15 17:43:28 +04:00
|
|
|
mca_mpool_sm_module_t* mpool_sm = (mca_mpool_sm_module_t*)mpool;
|
Per RFC, bring in the following changes:
* Remove paffinity, maffinity, and carto frameworks -- they've been
wholly replaced by hwloc.
* Move ompi_mpi_init() affinity-setting/checking code down to ORTE.
* Update sm, smcuda, wv, and openib components to no longer use carto.
Instead, use hwloc data. There are still optimizations possible in
the sm/smcuda BTLs (i.e., making multiple mpools). Also, the old
carto-based code found out how many NUMA nodes were ''available''
-- not how many were used ''in this job''. The new hwloc-using
code computes the same value -- it was not updated to calculate how
many NUMA nodes are used ''by this job.''
* Note that I cannot compile the smcuda and wv BTLs -- I ''think''
they're right, but they need to be verified by their owners.
* The openib component now does a bunch of stuff to figure out where
"near" OpenFabrics devices are. '''THIS IS A CHANGE IN DEFAULT
BEHAVIOR!!''' and still needs to be verified by OpenFabrics vendors
(I do not have a NUMA machine with an OpenFabrics device that is a
non-uniform distance from multiple different NUMA nodes).
* Completely rewrite the OMPI_Affinity_str() routine from the
"affinity" mpiext extension. This extension now understands
hyperthreads; the output format of it has changed a bit to reflect
this new information.
* Bunches of minor changes around the code base to update names/types
from maffinity/paffinity-based names to hwloc-based names.
* Add some helper functions into the hwloc base, mainly having to do
with the fact that we have the hwloc data reporting ''all''
topology information, but sometimes you really only want the
(online | available) data.
This commit was SVN r26391.
2012-05-07 18:52:54 +04:00
|
|
|
opal_hwloc_base_memory_segment_t mseg;
|
2008-06-15 17:43:28 +04:00
|
|
|
|
|
|
|
mseg.mbs_start_addr =
|
|
|
|
mpool_sm->sm_allocator->alc_realloc(mpool_sm->sm_allocator, addr, size,
|
|
|
|
registration);
|
|
|
|
if(mpool_sm->mem_node >= 0) {
|
|
|
|
mseg.mbs_len = size;
|
2012-09-27 03:24:27 +04:00
|
|
|
#if OPAL_HAVE_HWLOC
|
Per RFC, bring in the following changes:
* Remove paffinity, maffinity, and carto frameworks -- they've been
wholly replaced by hwloc.
* Move ompi_mpi_init() affinity-setting/checking code down to ORTE.
* Update sm, smcuda, wv, and openib components to no longer use carto.
Instead, use hwloc data. There are still optimizations possible in
the sm/smcuda BTLs (i.e., making multiple mpools). Also, the old
carto-based code found out how many NUMA nodes were ''available''
-- not how many were used ''in this job''. The new hwloc-using
code computes the same value -- it was not updated to calculate how
many NUMA nodes are used ''by this job.''
* Note that I cannot compile the smcuda and wv BTLs -- I ''think''
they're right, but they need to be verified by their owners.
* The openib component now does a bunch of stuff to figure out where
"near" OpenFabrics devices are. '''THIS IS A CHANGE IN DEFAULT
BEHAVIOR!!''' and still needs to be verified by OpenFabrics vendors
(I do not have a NUMA machine with an OpenFabrics device that is a
non-uniform distance from multiple different NUMA nodes).
* Completely rewrite the OMPI_Affinity_str() routine from the
"affinity" mpiext extension. This extension now understands
hyperthreads; the output format of it has changed a bit to reflect
this new information.
* Bunches of minor changes around the code base to update names/types
from maffinity/paffinity-based names to hwloc-based names.
* Add some helper functions into the hwloc base, mainly having to do
with the fact that we have the hwloc data reporting ''all''
topology information, but sometimes you really only want the
(online | available) data.
This commit was SVN r26391.
2012-05-07 18:52:54 +04:00
|
|
|
opal_hwloc_base_membind(&mseg, 1, mpool_sm->mem_node);
|
2012-09-27 03:24:27 +04:00
|
|
|
#endif
|
2008-06-15 17:43:28 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
return mseg.mbs_start_addr;
|
2004-08-02 04:24:22 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2008-06-15 17:43:28 +04:00
|
|
|
* free function
|
2004-08-02 04:24:22 +04:00
|
|
|
*/
|
2008-06-15 17:43:28 +04:00
|
|
|
void mca_mpool_sm_free(mca_mpool_base_module_t* mpool, void * addr,
|
2005-06-25 01:12:38 +04:00
|
|
|
mca_mpool_base_registration_t* registration)
|
2004-08-02 04:24:22 +04:00
|
|
|
{
|
2008-06-15 17:43:28 +04:00
|
|
|
mca_mpool_sm_module_t* mpool_sm = (mca_mpool_sm_module_t*)mpool;
|
2005-06-21 21:10:28 +04:00
|
|
|
mpool_sm->sm_allocator->alc_free(mpool_sm->sm_allocator, addr);
|
2004-08-02 04:24:22 +04:00
|
|
|
}
|
2007-03-17 02:11:45 +03:00
|
|
|
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
static void sm_module_finalize(mca_mpool_base_module_t* module)
|
|
|
|
{
|
|
|
|
mca_mpool_sm_module_t *sm_module = (mca_mpool_sm_module_t*) module;
|
|
|
|
|
2010-06-09 20:58:52 +04:00
|
|
|
if (NULL != sm_module->sm_common_module) {
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
if (OMPI_SUCCESS ==
|
2010-06-09 20:58:52 +04:00
|
|
|
mca_common_sm_fini(sm_module->sm_common_module)) {
|
2010-03-13 02:57:50 +03:00
|
|
|
#if OPAL_ENABLE_FT_CR == 1
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
/* Only unlink the file if we are *not* restarting. If we
|
|
|
|
are restarting the file will be unlinked at a later
|
|
|
|
time. */
|
|
|
|
if (OPAL_CR_STATUS_RESTART_PRE != opal_cr_checkpointing_state &&
|
|
|
|
OPAL_CR_STATUS_RESTART_POST != opal_cr_checkpointing_state ) {
|
2011-07-11 03:32:23 +04:00
|
|
|
unlink(sm_module->sm_common_module->shmem_ds.seg_name);
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
}
|
|
|
|
#else
|
2011-06-21 19:41:57 +04:00
|
|
|
unlink(sm_module->sm_common_module->shmem_ds.seg_name);
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
#endif
|
|
|
|
}
|
2010-06-09 20:58:52 +04:00
|
|
|
OBJ_RELEASE(sm_module->sm_common_module);
|
|
|
|
sm_module->sm_common_module = NULL;
|
Fixes trac:1988. The little bug that turned out to be huge. Yoinks.
* Various cosmetic/style updates in the btl sm
* Clean up concept of mpool module (I think that code was written way
back when the concept of "modules" was fuzzy)
* Bring over some old fixes from the /tmp/timattox-sm-coll/ tree to
fix potential segv's when mmap'ed regions were at different
addresses in different processes (thanks Tim!).
* Change sm coll to no longer use mpool as its main source of shmem;
rather, just mmap its own segment (because it's fixed size --
there was nothing to be gained by using mpool; shedding the use of
mpool saved a lot of complexity in the sm coll setup). This
effectively made Tim's fixes moot (because now everything is an
offset into the mmap that is computed locally; there are no global
pointers). :-)
* Slightly updated common/sm to allow making mmap's for a specific
set of procs (vs. ''all'' procs in the process). This potentially
allows for same-host-inter-proc mmaps -- yay!
* Fixed many, many things in the coll sm (particularly in reduce):
* Fixed handling of MPI_IN_PLACE in reduce and allreduce
* Fixed handling of non-contiguous datatypes in reduce
* Changed the order of reductions to go from process (n-1)'s data
to process 0's data, because that's how all other OMPI coll
components work
* Fixed lots of usage of ddt functions
* When using a non-contiguous datatype, if the root process is not
(n-1), now we used a 2nd convertor to copy from shmem to the rbuf
(saves a memory copy vs. what was done before)
* Lots and lots of little cleanups, clarifications, and minor
optimizations (although still more could be done -- e.g., I think
the use of write memory barriers is fairly sub-optimal; they
could be ganged together at the root, for example)
I'm marking this as "fixes trac:1988" and closing the ticket; if something
is still broken, we can re-open the ticket.
This commit was SVN r21967.
The following Trac tickets were found above:
Ticket 1988 --> https://svn.open-mpi.org/trac/ompi/ticket/1988
2009-09-15 04:25:21 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-03-13 02:57:50 +03:00
|
|
|
#if OPAL_ENABLE_FT_CR == 0
|
2007-03-17 02:11:45 +03:00
|
|
|
int mca_mpool_sm_ft_event(int state) {
|
2008-10-16 19:09:00 +04:00
|
|
|
return OMPI_SUCCESS;
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
int mca_mpool_sm_ft_event(int state) {
|
|
|
|
mca_mpool_base_module_t *self_module = NULL;
|
2009-09-17 18:45:17 +04:00
|
|
|
mca_mpool_sm_module_t *self_sm_module = NULL;
|
2008-10-16 19:09:00 +04:00
|
|
|
char * file_name = NULL;
|
|
|
|
|
2007-03-17 02:11:45 +03:00
|
|
|
if(OPAL_CRS_CHECKPOINT == state) {
|
2008-10-16 19:09:00 +04:00
|
|
|
/* Record the shared memory filename */
|
|
|
|
asprintf( &file_name, "%s"OPAL_PATH_SEP"shared_mem_pool.%s",
|
2013-01-28 03:25:10 +04:00
|
|
|
ompi_process_info.job_session_dir,
|
|
|
|
ompi_process_info.nodename );
|
A number of C/R enhancements per RFC below:
http://www.open-mpi.org/community/lists/devel/2010/07/8240.php
Documentation:
http://osl.iu.edu/research/ft/
Major Changes:
--------------
* Added C/R-enabled Debugging support.
Enabled with the --enable-crdebug flag. See the following website for more information:
http://osl.iu.edu/research/ft/crdebug/
* Added Stable Storage (SStore) framework for checkpoint storage
* 'central' component does a direct to central storage save
* 'stage' component stages checkpoints to central storage while the application continues execution.
* 'stage' supports offline compression of checkpoints before moving (sstore_stage_compress)
* 'stage' supports local caching of checkpoints to improve automatic recovery (sstore_stage_caching)
* Added Compression (compress) framework to support
* Add two new ErrMgr recovery policies
* {{{crmig}}} C/R Process Migration
* {{{autor}}} C/R Automatic Recovery
* Added the {{{ompi-migrate}}} command line tool to support the {{{crmig}}} ErrMgr component
* Added CR MPI Ext functions (enable them with {{{--enable-mpi-ext=cr}}} configure option)
* {{{OMPI_CR_Checkpoint}}} (Fixes trac:2342)
* {{{OMPI_CR_Restart}}}
* {{{OMPI_CR_Migrate}}} (may need some more work for mapping rules)
* {{{OMPI_CR_INC_register_callback}}} (Fixes trac:2192)
* {{{OMPI_CR_Quiesce_start}}}
* {{{OMPI_CR_Quiesce_checkpoint}}}
* {{{OMPI_CR_Quiesce_end}}}
* {{{OMPI_CR_self_register_checkpoint_callback}}}
* {{{OMPI_CR_self_register_restart_callback}}}
* {{{OMPI_CR_self_register_continue_callback}}}
* The ErrMgr predicted_fault() interface has been changed to take an opal_list_t of ErrMgr defined types. This will allow us to better support a wider range of fault prediction services in the future.
* Add a progress meter to:
* FileM rsh (filem_rsh_process_meter)
* SnapC full (snapc_full_progress_meter)
* SStore stage (sstore_stage_progress_meter)
* Added 2 new command line options to ompi-restart
* --showme : Display the full command line that would have been exec'ed.
* --mpirun_opts : Command line options to pass directly to mpirun. (Fixes trac:2413)
* Deprecated some MCA params:
* crs_base_snapshot_dir deprecated, use sstore_stage_local_snapshot_dir
* snapc_base_global_snapshot_dir deprecated, use sstore_base_global_snapshot_dir
* snapc_base_global_shared deprecated, use sstore_stage_global_is_shared
* snapc_base_store_in_place deprecated, replaced with different components of SStore
* snapc_base_global_snapshot_ref deprecated, use sstore_base_global_snapshot_ref
* snapc_base_establish_global_snapshot_dir deprecated, never well supported
* snapc_full_skip_filem deprecated, use sstore_stage_skip_filem
Minor Changes:
--------------
* Fixes trac:1924 : {{{ompi-restart}}} now recognizes path prefixed checkpoint handles and does the right thing.
* Fixes trac:2097 : {{{ompi-info}}} should now report all available CRS components
* Fixes trac:2161 : Manual checkpoint movement. A user can 'mv' a checkpoint directory from the original location to another and still restart from it.
* Fixes trac:2208 : Honor various TMPDIR varaibles instead of forcing {{{/tmp}}}
* Move {{{ompi_cr_continue_like_restart}}} to {{{orte_cr_continue_like_restart}}} to be more flexible in where this should be set.
* opal_crs_base_metadata_write* functions have been moved to SStore to support a wider range of metadata handling functionality.
* Cleanup the CRS framework and components to work with the SStore framework.
* Cleanup the SnapC framework and components to work with the SStore framework (cleans up these code paths considerably).
* Add 'quiesce' hook to CRCP for a future enhancement.
* We now require a BLCR version that supports {{{cr_request_file()}}} or {{{cr_request_checkpoint()}}} in order to make the code more maintainable. Note that {{{cr_request_file}}} has been deprecated since 0.7.0, so we prefer to use {{{cr_request_checkpoint()}}}.
* Add optional application level INC callbacks (registered through the CR MPI Ext interface).
* Increase the {{{opal_cr_thread_sleep_wait}}} parameter to 1000 microseconds to make the C/R thread less aggressive.
* {{{opal-restart}}} now looks for cache directories before falling back on stable storage when asked.
* {{{opal-restart}}} also support local decompression before restarting
* {{{orte-checkpoint}}} now uses the SStore framework to work with the metadata
* {{{orte-restart}}} now uses the SStore framework to work with the metadata
* Remove the {{{orte-restart}}} preload option. This was removed since the user only needs to select the 'stage' component in order to support this functionality.
* Since the '-am' parameter is saved in the metadata, {{{ompi-restart}}} no longer hard codes {{{-am ft-enable-cr}}}.
* Fix {{{hnp}}} ErrMgr so that if a previous component in the stack has 'fixed' the problem, then it should be skipped.
* Make sure to decrement the number of 'num_local_procs' in the orted when one goes away.
* odls now checks the SStore framework to see if it needs to load any checkpoint files before launching (to support 'stage'). This separates the SStore logic from the --preload-[binary|files] options.
* Add unique IDs to the named pipes established between the orted and the app in SnapC. This is to better support migration and automatic recovery activities.
* Improve the checks for 'already checkpointing' error path.
* A a recovery output timer, to show how long it takes to restart a job
* Do a better job of cleaning up the old session directory on restart.
* Add a local module to the autor and crmig ErrMgr components. These small modules prevent the 'orted' component from attempting a local recovery (Which does not work for MPI apps at the moment)
* Add a fix for bounding the checkpointable region between MPI_Init and MPI_Finalize.
This commit was SVN r23587.
The following Trac tickets were found above:
Ticket 1924 --> https://svn.open-mpi.org/trac/ompi/ticket/1924
Ticket 2097 --> https://svn.open-mpi.org/trac/ompi/ticket/2097
Ticket 2161 --> https://svn.open-mpi.org/trac/ompi/ticket/2161
Ticket 2192 --> https://svn.open-mpi.org/trac/ompi/ticket/2192
Ticket 2208 --> https://svn.open-mpi.org/trac/ompi/ticket/2208
Ticket 2342 --> https://svn.open-mpi.org/trac/ompi/ticket/2342
Ticket 2413 --> https://svn.open-mpi.org/trac/ompi/ticket/2413
2010-08-11 00:51:11 +04:00
|
|
|
orte_sstore.set_attr(orte_sstore_handle_current, SSTORE_METADATA_LOCAL_TOUCH, file_name);
|
2008-10-16 19:09:00 +04:00
|
|
|
free(file_name);
|
|
|
|
file_name = NULL;
|
2007-03-17 02:11:45 +03:00
|
|
|
}
|
|
|
|
else if(OPAL_CRS_CONTINUE == state) {
|
2012-06-27 05:28:28 +04:00
|
|
|
if(orte_cr_continue_like_restart) {
|
2009-09-17 18:45:17 +04:00
|
|
|
/* Find the sm module */
|
2008-10-16 19:09:00 +04:00
|
|
|
self_module = mca_mpool_base_module_lookup("sm");
|
2009-09-17 18:45:17 +04:00
|
|
|
self_sm_module = (mca_mpool_sm_module_t*) self_module;
|
2008-10-16 19:09:00 +04:00
|
|
|
|
2009-09-17 18:45:17 +04:00
|
|
|
/* Mark the old sm file for eventual removal via CRS */
|
2010-06-09 20:58:52 +04:00
|
|
|
if (NULL != self_sm_module->sm_common_module) {
|
2011-07-11 03:32:23 +04:00
|
|
|
opal_crs_base_cleanup_append(self_sm_module->sm_common_module->shmem_ds.seg_name, false);
|
2008-10-16 19:09:00 +04:00
|
|
|
}
|
2009-09-17 18:45:17 +04:00
|
|
|
|
|
|
|
/* Remove self from the list of all modules */
|
|
|
|
mca_mpool_base_module_destroy(self_module);
|
2008-10-16 19:09:00 +04:00
|
|
|
}
|
2007-03-17 02:11:45 +03:00
|
|
|
}
|
2008-10-16 19:09:00 +04:00
|
|
|
else if(OPAL_CRS_RESTART == state ||
|
|
|
|
OPAL_CRS_RESTART_PRE == state) {
|
2009-09-17 18:45:17 +04:00
|
|
|
/* Find the sm module */
|
2008-10-16 19:09:00 +04:00
|
|
|
self_module = mca_mpool_base_module_lookup("sm");
|
2009-09-17 18:45:17 +04:00
|
|
|
self_sm_module = (mca_mpool_sm_module_t*) self_module;
|
2008-10-16 19:09:00 +04:00
|
|
|
|
2009-09-17 18:45:17 +04:00
|
|
|
/* Mark the old sm file for eventual removal via CRS */
|
2010-06-09 20:58:52 +04:00
|
|
|
if (NULL != self_sm_module->sm_common_module) {
|
2011-07-11 03:32:23 +04:00
|
|
|
opal_crs_base_cleanup_append(self_sm_module->sm_common_module->shmem_ds.seg_name, false);
|
2008-10-16 19:09:00 +04:00
|
|
|
}
|
2009-09-17 18:45:17 +04:00
|
|
|
|
|
|
|
/* Remove self from the list of all modules */
|
|
|
|
mca_mpool_base_module_destroy(self_module);
|
2007-03-17 02:11:45 +03:00
|
|
|
}
|
|
|
|
else if(OPAL_CRS_TERM == state ) {
|
|
|
|
;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
;
|
|
|
|
}
|
|
|
|
|
|
|
|
return OMPI_SUCCESS;
|
|
|
|
}
|
2010-03-13 02:57:50 +03:00
|
|
|
#endif /* OPAL_ENABLE_FT_CR */
|