552c9ca5a0
WHAT: Open our low-level communication infrastructure by moving all necessary components (btl/rcache/allocator/mpool) down in OPAL All the components required for inter-process communications are currently deeply integrated in the OMPI layer. Several groups/institutions have express interest in having a more generic communication infrastructure, without all the OMPI layer dependencies. This communication layer should be made available at a different software level, available to all layers in the Open MPI software stack. As an example, our ORTE layer could replace the current OOB and instead use the BTL directly, gaining access to more reactive network interfaces than TCP. Similarly, external software libraries could take advantage of our highly optimized AM (active message) communication layer for their own purpose. UTK with support from Sandia, developped a version of Open MPI where the entire communication infrastucture has been moved down to OPAL (btl/rcache/allocator/mpool). Most of the moved components have been updated to match the new schema, with few exceptions (mainly BTLs where I have no way of compiling/testing them). Thus, the completion of this RFC is tied to being able to completing this move for all BTLs. For this we need help from the rest of the Open MPI community, especially those supporting some of the BTLs. A non-exhaustive list of BTLs that qualify here is: mx, portals4, scif, udapl, ugni, usnic. This commit was SVN r32317.
102 строки
3.7 KiB
C
102 строки
3.7 KiB
C
/* -*- Mode: C; c-basic-offset:4 ; -*- */
|
|
/*
|
|
* Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
|
|
* University Research and Technology
|
|
* Corporation. All rights reserved.
|
|
* Copyright (c) 2004-2007 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) 2009 Cisco Systems, Inc. All rights reserved.
|
|
* Copyright (c) 2012 Los Alamos National Security, LLC.
|
|
* All rights reserved.
|
|
* $COPYRIGHT$
|
|
*
|
|
* Additional copyrights may follow
|
|
*
|
|
* $HEADER$
|
|
*/
|
|
/**
|
|
* @file
|
|
*/
|
|
#include "opal_config.h"
|
|
|
|
#ifdef HAVE_UNISTD_H
|
|
#include <unistd.h>
|
|
#endif
|
|
|
|
#include "opal/util/show_help.h"
|
|
#include "opal/util/proc.h"
|
|
#include "opal/runtime/opal_params.h"
|
|
|
|
#include "opal/mca/mpool/base/mpool_base_mem_cb.h"
|
|
#include "opal/mca/mpool/base/base.h"
|
|
|
|
#include "opal/mca/mca.h"
|
|
#include "opal/memoryhooks/memory.h"
|
|
|
|
static char msg[512];
|
|
|
|
|
|
/*
|
|
* memory hook callback, called when memory is free'd out from under
|
|
* us. Be wary of the from_alloc flag -- if you're called with
|
|
* from_alloc==true, then you cannot call malloc (or any of its
|
|
* friends)!
|
|
*/
|
|
void mca_mpool_base_mem_cb(void* base, size_t size, void* cbdata,
|
|
bool from_alloc)
|
|
{
|
|
mca_mpool_base_selected_module_t* current;
|
|
int rc;
|
|
opal_list_item_t* item;
|
|
|
|
/* Only do anything meaningful if the OPAL layer is up and running
|
|
and size != 0 */
|
|
if ((from_alloc && (!opal_initialized)) ||
|
|
size == 0) {
|
|
return;
|
|
}
|
|
|
|
for(item = opal_list_get_first(&mca_mpool_base_modules);
|
|
item != opal_list_get_end(&mca_mpool_base_modules);
|
|
item = opal_list_get_next(item)) {
|
|
|
|
current = (mca_mpool_base_selected_module_t*) item;
|
|
|
|
if(current->mpool_module->mpool_release_memory != NULL) {
|
|
rc = current->mpool_module->mpool_release_memory(current->mpool_module,
|
|
base, size);
|
|
|
|
if (rc != OPAL_SUCCESS) {
|
|
if (from_alloc) {
|
|
int len;
|
|
len = snprintf(msg, sizeof(msg), "[%s:%d] Attempt to free memory that is still in use by an ongoing MPI communication (buffer %p, size %lu). MPI job will now abort.\n",
|
|
opal_proc_local_get()->proc_hostname,
|
|
getpid(),
|
|
base, (unsigned long) size);
|
|
msg[sizeof(msg) - 1] = '\0';
|
|
write(2, msg, len);
|
|
} else {
|
|
opal_show_help("help-mpool-base.txt",
|
|
"cannot deregister in-use memory", true,
|
|
current->mpool_component->mpool_version.mca_component_name,
|
|
opal_proc_local_get()->proc_hostname,
|
|
base, (unsigned long) size);
|
|
}
|
|
|
|
/* We're in a callback from somewhere; we can't do
|
|
anything meaningful to pass an error back up. :-(
|
|
So just exit. Call _exit() so that we don't try to
|
|
call anything on the way out -- just exit!
|
|
(remember that we're in a callback, and state may
|
|
be very undefined at this point...) */
|
|
_exit(1);
|
|
}
|
|
}
|
|
}
|
|
}
|