numam-spdk/lib/util/bit_array.c
GangCao 621f96f7aa util/bit_array: use spdk_realloc for process sharing
Change-Id: I8fe49388e7bec9306474f27de7c17e767dfa19e8
Signed-off-by: GangCao <gang.cao@intel.com>
2016-11-09 15:04:20 -07:00

278 lines
7.8 KiB
C

/*-
* BSD LICENSE
*
* Copyright (c) Intel Corporation.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
* * Neither the name of Intel Corporation nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include "spdk/bit_array.h"
#include "spdk/env.h"
#include <assert.h>
#include <errno.h>
#include <limits.h>
#include <stdbool.h>
#include <stddef.h>
#include <stdlib.h>
#include <string.h>
#include "spdk/likely.h"
typedef uint64_t spdk_bit_array_word;
#define SPDK_BIT_ARRAY_WORD_TZCNT(x) (__builtin_ctzll(x))
#define SPDK_BIT_ARRAY_WORD_C(x) ((spdk_bit_array_word)(x))
#define SPDK_BIT_ARRAY_WORD_BYTES sizeof(spdk_bit_array_word)
#define SPDK_BIT_ARRAY_WORD_BITS (SPDK_BIT_ARRAY_WORD_BYTES * 8)
#define SPDK_BIT_ARRAY_WORD_INDEX_SHIFT (31u - __builtin_clz(SPDK_BIT_ARRAY_WORD_BITS))
#define SPDK_BIT_ARRAY_WORD_INDEX_MASK ((1u << SPDK_BIT_ARRAY_WORD_INDEX_SHIFT) - 1)
struct spdk_bit_array {
uint32_t bit_count;
spdk_bit_array_word words[];
};
struct spdk_bit_array *
spdk_bit_array_create(uint32_t num_bits)
{
struct spdk_bit_array *ba = NULL;
spdk_bit_array_resize(&ba, num_bits);
return ba;
}
void
spdk_bit_array_free(struct spdk_bit_array **bap)
{
struct spdk_bit_array *ba;
if (!bap) {
return;
}
ba = *bap;
*bap = NULL;
spdk_free(ba);
}
static inline uint32_t
spdk_bit_array_word_count(uint32_t num_bits)
{
return (num_bits + SPDK_BIT_ARRAY_WORD_BITS - 1) >> SPDK_BIT_ARRAY_WORD_INDEX_SHIFT;
}
static inline spdk_bit_array_word
spdk_bit_array_word_mask(uint32_t num_bits)
{
assert(num_bits < SPDK_BIT_ARRAY_WORD_BITS);
return (SPDK_BIT_ARRAY_WORD_C(1) << num_bits) - 1;
}
int
spdk_bit_array_resize(struct spdk_bit_array **bap, uint32_t num_bits)
{
struct spdk_bit_array *new_ba;
uint32_t old_word_count, new_word_count;
size_t new_size;
if (!bap) {
return -EINVAL;
}
new_word_count = spdk_bit_array_word_count(num_bits);
new_size = offsetof(struct spdk_bit_array, words) + new_word_count * SPDK_BIT_ARRAY_WORD_BYTES;
/*
* Always keep one extra word with a 0 and a 1 past the actual required size so that the
* find_first functions can just keep going until they match.
*/
new_size += SPDK_BIT_ARRAY_WORD_BYTES;
new_ba = (struct spdk_bit_array *)spdk_realloc(*bap, new_size, 64, NULL);
if (!new_ba) {
return -ENOMEM;
}
/*
* Set up special extra word (see above comment about find_first_clear).
*
* This is set to 0b10 so that find_first_clear will find a 0 at the very first
* bit past the end of the buffer, and find_first_set will find a 1 at the next bit
* past that.
*/
new_ba->words[new_word_count] = 0x2;
if (*bap == NULL) {
old_word_count = 0;
new_ba->bit_count = 0;
} else {
old_word_count = spdk_bit_array_word_count(new_ba->bit_count);
}
if (new_word_count > old_word_count) {
/* Zero out new entries */
memset(&new_ba->words[old_word_count], 0,
(new_word_count - old_word_count) * SPDK_BIT_ARRAY_WORD_BYTES);
} else if (new_word_count == old_word_count && num_bits < new_ba->bit_count) {
/* Make sure any existing partial last word is cleared beyond the new num_bits. */
uint32_t last_word_bits;
spdk_bit_array_word mask;
last_word_bits = num_bits & SPDK_BIT_ARRAY_WORD_INDEX_MASK;
mask = spdk_bit_array_word_mask(last_word_bits);
new_ba->words[old_word_count - 1] &= mask;
}
new_ba->bit_count = num_bits;
*bap = new_ba;
return 0;
}
uint32_t
spdk_bit_array_capacity(const struct spdk_bit_array *ba)
{
return ba->bit_count;
}
static inline int
_spdk_bit_array_get_word(const struct spdk_bit_array *ba, uint32_t bit_index,
uint32_t *word_index, uint32_t *word_bit_index)
{
if (spdk_unlikely(bit_index >= ba->bit_count)) {
return -EINVAL;
}
*word_index = bit_index >> SPDK_BIT_ARRAY_WORD_INDEX_SHIFT;
*word_bit_index = bit_index & SPDK_BIT_ARRAY_WORD_INDEX_MASK;
return 0;
}
bool
spdk_bit_array_get(const struct spdk_bit_array *ba, uint32_t bit_index)
{
uint32_t word_index, word_bit_index;
if (_spdk_bit_array_get_word(ba, bit_index, &word_index, &word_bit_index)) {
return false;
}
return (ba->words[word_index] >> word_bit_index) & 1U;
}
int
spdk_bit_array_set(struct spdk_bit_array *ba, uint32_t bit_index)
{
uint32_t word_index, word_bit_index;
if (_spdk_bit_array_get_word(ba, bit_index, &word_index, &word_bit_index)) {
return -EINVAL;
}
ba->words[word_index] |= (SPDK_BIT_ARRAY_WORD_C(1) << word_bit_index);
return 0;
}
void
spdk_bit_array_clear(struct spdk_bit_array *ba, uint32_t bit_index)
{
uint32_t word_index, word_bit_index;
if (_spdk_bit_array_get_word(ba, bit_index, &word_index, &word_bit_index)) {
/*
* Clearing past the end of the bit array is a no-op, since bit past the end
* are implicitly 0.
*/
return;
}
ba->words[word_index] &= ~(SPDK_BIT_ARRAY_WORD_C(1) << word_bit_index);
}
static inline uint32_t
_spdk_bit_array_find_first(const struct spdk_bit_array *ba, uint32_t start_bit_index,
spdk_bit_array_word xor_mask)
{
uint32_t word_index, first_word_bit_index;
spdk_bit_array_word word, first_word_mask;
const spdk_bit_array_word *words, *cur_word;
if (spdk_unlikely(start_bit_index >= ba->bit_count)) {
return ba->bit_count;
}
word_index = start_bit_index >> SPDK_BIT_ARRAY_WORD_INDEX_SHIFT;
words = ba->words;
cur_word = &words[word_index];
/*
* Special case for first word: skip start_bit_index % SPDK_BIT_ARRAY_WORD_BITS bits
* within the first word.
*/
first_word_bit_index = start_bit_index & SPDK_BIT_ARRAY_WORD_INDEX_MASK;
first_word_mask = spdk_bit_array_word_mask(first_word_bit_index);
word = (*cur_word ^ xor_mask) & ~first_word_mask;
/*
* spdk_bit_array_resize() guarantees that an extra word with a 1 and a 0 will always be
* at the end of the words[] array, so just keep going until a word matches.
*/
while (word == 0) {
word = *++cur_word ^ xor_mask;
}
return ((uintptr_t)cur_word - (uintptr_t)words) * 8 + SPDK_BIT_ARRAY_WORD_TZCNT(word);
}
uint32_t
spdk_bit_array_find_first_set(const struct spdk_bit_array *ba, uint32_t start_bit_index)
{
uint32_t bit_index;
bit_index = _spdk_bit_array_find_first(ba, start_bit_index, 0);
/*
* If we ran off the end of the array and found the 1 bit in the extra word,
* return UINT32_MAX to indicate no actual 1 bits were found.
*/
if (bit_index >= ba->bit_count) {
bit_index = UINT32_MAX;
}
return bit_index;
}
uint32_t
spdk_bit_array_find_first_clear(const struct spdk_bit_array *ba, uint32_t start_bit_index)
{
return _spdk_bit_array_find_first(ba, start_bit_index, SPDK_BIT_ARRAY_WORD_C(-1));
}