1
1
openmpi/opal/class/opal_free_list.c
Nathan Hjelm 5c770a7bec opal/free_list: fix race condition
There was a race condition in opal_free_list_get. Code throughout the
Open MPI codebase was assuming that a NULL return from this function
was due to an out-of-memory condition. In some cases this can lead to
a fatal condition (MPI_Irecv and MPI_Isend in pml/ob1 for
example). Before this commit opal_free_list_get_mt looked like this:

```c
static inline opal_free_list_item_t *opal_free_list_get_mt (opal_free_list_t *flist)
{
    opal_free_list_item_t *item =
        (opal_free_list_item_t*) opal_lifo_pop_atomic (&flist->super);

    if (OPAL_UNLIKELY(NULL == item)) {
        opal_mutex_lock (&flist->fl_lock);
        opal_free_list_grow_st (flist, flist->fl_num_per_alloc);
        opal_mutex_unlock (&flist->fl_lock);
        item = (opal_free_list_item_t *) opal_lifo_pop_atomic (&flist->super);
    }

    return item;
}
```

The problem is in a multithreaded environment is *is* possible for the
free list to be grown successfully but the thread calling
opal_free_list_get_mt to be left without an item. The happens if
between the calls to opal_lifo_push_atomic in opal_free_list_grow_st
and the call to opal_lifo_pop_atomic other threads pop all the items
added to the free list.

This commit fixes the issue by ensuring the thread that successfully
grew the free list **always** gets a free list item.

Fixes #2921

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2018-10-16 13:17:09 -06:00

318 строки
11 KiB
C

/* -*- Mode: C; c-basic-offset:4 ; indent-tabs-mode:nil -*- */
/*
* Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
* University Research and Technology
* Corporation. All rights reserved.
* Copyright (c) 2004-2009 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) 2006-2007 Mellanox Technologies. All rights reserved.
* Copyright (c) 2010-2013 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2011 NVIDIA Corporation. All rights reserved.
* Copyright (c) 2012-2018 Los Alamos National Security, LLC. All rights
* reserved.
* $COPYRIGHT$
*
* Additional copyrights may follow
*
* $HEADER$
*/
#include "opal_config.h"
#include "opal/class/opal_free_list.h"
#include "opal/align.h"
#include "opal/util/output.h"
#include "opal/mca/mpool/mpool.h"
#include "opal/mca/mpool/base/base.h"
#include "opal/mca/rcache/rcache.h"
#include "opal/util/sys_limits.h"
typedef struct opal_free_list_item_t opal_free_list_memory_t;
OBJ_CLASS_INSTANCE(opal_free_list_item_t,
opal_list_item_t,
NULL, NULL);
static void opal_free_list_construct(opal_free_list_t* fl)
{
OBJ_CONSTRUCT(&fl->fl_lock, opal_mutex_t);
OBJ_CONSTRUCT(&fl->fl_condition, opal_condition_t);
fl->fl_max_to_alloc = 0;
fl->fl_num_allocated = 0;
fl->fl_num_per_alloc = 0;
fl->fl_num_waiting = 0;
fl->fl_frag_size = sizeof(opal_free_list_item_t);
fl->fl_frag_alignment = 0;
fl->fl_payload_buffer_size = 0;
fl->fl_payload_buffer_alignment = 0;
fl->fl_frag_class = OBJ_CLASS(opal_free_list_item_t);
fl->fl_mpool = NULL;
fl->fl_rcache = NULL;
/* default flags */
fl->fl_rcache_reg_flags = MCA_RCACHE_FLAGS_CACHE_BYPASS |
MCA_RCACHE_FLAGS_CUDA_REGISTER_MEM;
fl->ctx = NULL;
OBJ_CONSTRUCT(&(fl->fl_allocations), opal_list_t);
}
static void opal_free_list_allocation_release (opal_free_list_t *fl, opal_free_list_memory_t *fl_mem)
{
if (NULL != fl->fl_rcache) {
fl->fl_rcache->rcache_deregister (fl->fl_rcache, fl_mem->registration);
}
if (NULL != fl->fl_mpool) {
fl->fl_mpool->mpool_free (fl->fl_mpool, fl_mem->ptr);
} else if (fl_mem->ptr) {
free (fl_mem->ptr);
}
/* destruct the item (we constructed it), then free the memory chunk */
OBJ_DESTRUCT(fl_mem);
free(fl_mem);
}
static void opal_free_list_destruct(opal_free_list_t *fl)
{
opal_list_item_t *item;
opal_free_list_item_t *fl_item;
#if 0 && OPAL_ENABLE_DEBUG
if(opal_list_get_size(&fl->super) != fl->fl_num_allocated) {
opal_output(0, "opal_free_list: %d allocated %d returned: %s:%d\n",
fl->fl_num_allocated, opal_list_get_size(&fl->super),
fl->super.super.cls_init_file_name, fl->super.super.cls_init_lineno);
}
#endif
while(NULL != (item = opal_lifo_pop(&(fl->super)))) {
fl_item = (opal_free_list_item_t*)item;
/* destruct the item (we constructed it), the underlying memory will be
* reclaimed when we free the slab (opal_free_list_memory_t ptr)
* containing it */
OBJ_DESTRUCT(fl_item);
}
while(NULL != (item = opal_list_remove_first(&fl->fl_allocations))) {
opal_free_list_allocation_release (fl, (opal_free_list_memory_t *) item);
}
OBJ_DESTRUCT(&fl->fl_allocations);
OBJ_DESTRUCT(&fl->fl_condition);
OBJ_DESTRUCT(&fl->fl_lock);
}
OBJ_CLASS_INSTANCE(opal_free_list_t, opal_lifo_t, opal_free_list_construct,
opal_free_list_destruct);
int opal_free_list_init (opal_free_list_t *flist, size_t frag_size, size_t frag_alignment,
opal_class_t *frag_class, size_t payload_buffer_size,
size_t payload_buffer_alignment, int num_elements_to_alloc,
int max_elements_to_alloc, int num_elements_per_alloc,
mca_mpool_base_module_t *mpool, int rcache_reg_flags,
mca_rcache_base_module_t *rcache, opal_free_list_item_init_fn_t item_init,
void *ctx)
{
/* alignment must be more than zero and power of two */
if (frag_alignment <= 1 || (frag_alignment & (frag_alignment - 1))) {
return OPAL_ERROR;
}
if (0 < payload_buffer_size) {
if (payload_buffer_alignment <= 1 || (payload_buffer_alignment & (payload_buffer_alignment - 1)))
return OPAL_ERROR;
}
if (frag_class && frag_size < frag_class->cls_sizeof) {
frag_size = frag_class->cls_sizeof;
}
if (frag_size > flist->fl_frag_size) {
flist->fl_frag_size = frag_size;
}
if (frag_class) {
flist->fl_frag_class = frag_class;
}
flist->fl_payload_buffer_size = payload_buffer_size;
flist->fl_max_to_alloc = max_elements_to_alloc;
flist->fl_num_allocated = 0;
flist->fl_num_per_alloc = num_elements_per_alloc;
flist->fl_mpool = mpool ? mpool : mca_mpool_base_default_module;
flist->fl_rcache = rcache;
flist->fl_frag_alignment = frag_alignment;
flist->fl_payload_buffer_alignment = payload_buffer_alignment;
flist->item_init = item_init;
flist->fl_rcache_reg_flags |= rcache_reg_flags;
flist->ctx = ctx;
if (num_elements_to_alloc) {
return opal_free_list_grow_st (flist, num_elements_to_alloc, NULL);
}
return OPAL_SUCCESS;
}
int opal_free_list_grow_st (opal_free_list_t* flist, size_t num_elements, opal_free_list_item_t **item_out)
{
unsigned char *ptr, *payload_ptr = NULL;
opal_free_list_memory_t *alloc_ptr;
size_t alloc_size, head_size, elem_size = 0, buffer_size = 0, align = 0;
mca_rcache_base_registration_t *reg = NULL;
int rc = OPAL_SUCCESS;
if (flist->fl_max_to_alloc && (flist->fl_num_allocated + num_elements) >
flist->fl_max_to_alloc) {
num_elements = flist->fl_max_to_alloc - flist->fl_num_allocated;
}
if (num_elements == 0) {
return OPAL_ERR_TEMP_OUT_OF_RESOURCE;
}
head_size = OPAL_ALIGN(flist->fl_frag_size, flist->fl_frag_alignment, size_t);
/* NTH: calculate allocation alignment first as it might change the number of elements */
if (0 != flist->fl_payload_buffer_size) {
elem_size = OPAL_ALIGN(flist->fl_payload_buffer_size,
flist->fl_payload_buffer_alignment, size_t);
/* elem_size should not be 0 here */
assert (elem_size > 0);
buffer_size = num_elements * elem_size;
align = flist->fl_payload_buffer_alignment;
if (MCA_RCACHE_FLAGS_CUDA_REGISTER_MEM & flist->fl_rcache_reg_flags) {
size_t pagesize = opal_getpagesize ();
/* CUDA cannot handle registering overlapping regions, so make
* sure each region is page sized and page aligned. */
align = OPAL_ALIGN(align, pagesize, size_t);
buffer_size = OPAL_ALIGN(buffer_size, pagesize, size_t);
/* avoid wasting space in the buffer */
num_elements = buffer_size / elem_size;
}
}
/* calculate head allocation size */
alloc_size = num_elements * head_size + sizeof(opal_free_list_memory_t) +
flist->fl_frag_alignment;
alloc_ptr = (opal_free_list_memory_t *) malloc(alloc_size);
if (OPAL_UNLIKELY(NULL == alloc_ptr)) {
return OPAL_ERR_TEMP_OUT_OF_RESOURCE;
}
if (0 != flist->fl_payload_buffer_size) {
/* allocate the rest from the mpool (or use memalign/malloc) */
payload_ptr = (unsigned char *) flist->fl_mpool->mpool_alloc(flist->fl_mpool, buffer_size, align, 0);
if (NULL == payload_ptr) {
free(alloc_ptr);
return OPAL_ERR_TEMP_OUT_OF_RESOURCE;
}
if (flist->fl_rcache) {
rc = flist->fl_rcache->rcache_register (flist->fl_rcache, payload_ptr, num_elements * elem_size,
flist->fl_rcache_reg_flags, MCA_RCACHE_ACCESS_ANY, &reg);
if (OPAL_UNLIKELY(OPAL_SUCCESS != rc)) {
free (alloc_ptr);
flist->fl_mpool->mpool_free (flist->fl_mpool, payload_ptr);
return rc;
}
}
}
/* make the alloc_ptr a list item, save the chunk in the allocations list,
* and have ptr point to memory right after the list item structure */
OBJ_CONSTRUCT(alloc_ptr, opal_free_list_item_t);
opal_list_append(&(flist->fl_allocations), (opal_list_item_t*)alloc_ptr);
alloc_ptr->registration = reg;
alloc_ptr->ptr = payload_ptr;
ptr = (unsigned char*)alloc_ptr + sizeof(opal_free_list_memory_t);
ptr = OPAL_ALIGN_PTR(ptr, flist->fl_frag_alignment, unsigned char*);
for(size_t i = 0; i < num_elements ; ++i) {
opal_free_list_item_t* item = (opal_free_list_item_t*)ptr;
item->registration = reg;
item->ptr = payload_ptr;
OBJ_CONSTRUCT_INTERNAL(item, flist->fl_frag_class);
item->super.item_free = 0;
/* run the initialize function if present */
if (flist->item_init) {
if (OPAL_SUCCESS != (rc = flist->item_init(item, flist->ctx))) {
num_elements = i;
OBJ_DESTRUCT (item);
break;
}
}
/* NTH: in case the free list may be accessed from multiple threads
* use the atomic lifo push. The overhead is small compared to the
* overall overhead of opal_free_list_grow(). */
if (item_out && 0 == i) {
/* ensure the thread that is growing the free list always gets an item
* if one is available */
*item_out = item;
} else {
opal_lifo_push_atomic (&flist->super, &item->super);
}
ptr += head_size;
payload_ptr += elem_size;
}
if (OPAL_SUCCESS != rc && 0 == num_elements) {
/* couldn't initialize any items */
opal_list_remove_item (&flist->fl_allocations, (opal_list_item_t *) alloc_ptr);
opal_free_list_allocation_release (flist, alloc_ptr);
return OPAL_ERR_OUT_OF_RESOURCE;
}
flist->fl_num_allocated += num_elements;
return OPAL_SUCCESS;
}
/**
* This function resize the free_list to contain at least the specified
* number of elements. We do not create all of them in the same memory
* segment. Instead we will several time the fl_num_per_alloc elements
* until we reach the required number of the maximum allowed by the
* initialization.
*/
int opal_free_list_resize_mt(opal_free_list_t *flist, size_t size)
{
ssize_t inc_num;
int ret = OPAL_SUCCESS;
if (flist->fl_num_allocated > size) {
return OPAL_SUCCESS;
}
opal_mutex_lock (&flist->fl_lock);
do {
ret = opal_free_list_grow_st (flist, flist->fl_num_per_alloc, NULL);
if (OPAL_SUCCESS != ret) {
break;
}
inc_num = (ssize_t)size - (ssize_t)flist->fl_num_allocated;
} while (inc_num > 0);
opal_mutex_unlock (&flist->fl_lock);
return ret;
}