mirror of https://github.com/borgbackup/borg.git
Various hashindex improvements
This commit is contained in:
parent
7e06d30ed4
commit
07270cfd51
|
@ -1,4 +1,5 @@
|
|||
#include <assert.h>
|
||||
#include <endian.h>
|
||||
#include <stdio.h>
|
||||
#include <stdlib.h>
|
||||
#include <stdint.h>
|
||||
|
@ -9,7 +10,7 @@
|
|||
#include <unistd.h>
|
||||
#include <sys/mman.h>
|
||||
|
||||
#if defined(__BYTE_ORDER__)&&(__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)
|
||||
#if defined(__BYTE_ORDER)&&(__BYTE_ORDER == __BIG_ENDIAN)
|
||||
#error This code is not big endian safe yet
|
||||
#endif
|
||||
|
||||
|
@ -31,12 +32,18 @@ typedef struct {
|
|||
int key_size;
|
||||
int value_size;
|
||||
int bucket_size;
|
||||
int limit;
|
||||
int lower_limit;
|
||||
int upper_limit;
|
||||
} HashIndex;
|
||||
|
||||
#define MAGIC "DARCHASH"
|
||||
#define EMPTY ((int32_t)-1)
|
||||
#define DELETED ((int32_t)-2)
|
||||
#define MAX_BUCKET_SIZE 512
|
||||
#define BUCKET_LOWER_LIMIT .25
|
||||
#define BUCKET_UPPER_LIMIT .90
|
||||
#define MIN_BUCKETS 1024
|
||||
#define MAX(x, y) ((x) > (y) ? (x): (y))
|
||||
#define BUCKET_ADDR_READ(index, idx) (index->buckets + (idx * index->bucket_size))
|
||||
#define BUCKET_ADDR_WRITE(index, idx) (index->buckets + (idx * index->bucket_size))
|
||||
|
||||
|
@ -47,14 +54,17 @@ typedef struct {
|
|||
|
||||
#define BUCKET_MARK_DELETED(index, idx) (*((int32_t *)(BUCKET_ADDR_WRITE(index, idx) + index->key_size)) = DELETED)
|
||||
|
||||
#define EPRINTF(msg, ...) EPRINTF_PATH(index->path, msg, ##__VA_ARGS__)
|
||||
#define EPRINTF_PATH(path, msg, ...) fprintf(stderr, "hashindex: %s: " msg "\n", path, ##__VA_ARGS__)
|
||||
|
||||
static HashIndex *hashindex_open(const char *path);
|
||||
static void hashindex_close(HashIndex *index);
|
||||
static void hashindex_clear(HashIndex *index);
|
||||
static void hashindex_flush(HashIndex *index);
|
||||
static int hashindex_close(HashIndex *index);
|
||||
static int hashindex_clear(HashIndex *index);
|
||||
static int hashindex_flush(HashIndex *index);
|
||||
static HashIndex *hashindex_create(const char *path, int capacity, int key_size, int value_size);
|
||||
static const void *hashindex_get(HashIndex *index, const void *key);
|
||||
static void hashindex_set(HashIndex *index, const void *key, const void *value);
|
||||
static void hashindex_delete(HashIndex *index, const void *key);
|
||||
static int hashindex_set(HashIndex *index, const void *key, const void *value);
|
||||
static int hashindex_delete(HashIndex *index, const void *key);
|
||||
static void *hashindex_next_key(HashIndex *index, const void *key);
|
||||
|
||||
|
||||
|
@ -96,13 +106,18 @@ hashindex_lookup(HashIndex *index, const void *key)
|
|||
}
|
||||
}
|
||||
|
||||
static void
|
||||
static int
|
||||
hashindex_resize(HashIndex *index, int capacity)
|
||||
{
|
||||
char *new_path = malloc(strlen(index->path) + 5);
|
||||
strcpy(new_path, index->path);
|
||||
strcat(new_path, ".tmp");
|
||||
HashIndex *new = hashindex_create(new_path, capacity, index->key_size, index->value_size);
|
||||
HashIndex *new;
|
||||
|
||||
if(!(new = hashindex_create(new_path, capacity, index->key_size, index->value_size))) {
|
||||
free(new_path);
|
||||
return 0;
|
||||
}
|
||||
void *key = NULL;
|
||||
while((key = hashindex_next_key(index, key))) {
|
||||
hashindex_set(new, key, hashindex_get(index, key));
|
||||
|
@ -111,34 +126,55 @@ hashindex_resize(HashIndex *index, int capacity)
|
|||
index->map_addr = new->map_addr;
|
||||
index->map_length = new->map_length;
|
||||
index->num_buckets = new->num_buckets;
|
||||
index->limit = new->limit;
|
||||
index->lower_limit = new->lower_limit;
|
||||
index->upper_limit = new->upper_limit;
|
||||
index->buckets = new->buckets;
|
||||
unlink(index->path);
|
||||
rename(new_path, index->path);
|
||||
free(new_path);
|
||||
free(new->path);
|
||||
free(new);
|
||||
return 1;
|
||||
}
|
||||
|
||||
/* Public API */
|
||||
static HashIndex *
|
||||
hashindex_open(const char *path)
|
||||
{
|
||||
int fd = open(path, O_RDWR);
|
||||
if(fd < 0) {
|
||||
void *addr;
|
||||
int fd;
|
||||
off_t length;
|
||||
HashHeader *header;
|
||||
HashIndex *index;
|
||||
|
||||
if((fd = open(path, O_RDWR)) < 0) {
|
||||
EPRINTF_PATH(path, "open failed");
|
||||
fprintf(stderr, "Failed to open %s\n", path);
|
||||
return NULL;
|
||||
}
|
||||
off_t length = lseek(fd, 0, SEEK_END);
|
||||
void *addr = mmap(0, length, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0);
|
||||
close(fd);
|
||||
if(addr == MAP_FAILED) {
|
||||
fprintf(stderr, "Failed to mmap %s", path);
|
||||
if((length = lseek(fd, 0, SEEK_END)) < 0) {
|
||||
EPRINTF_PATH(path, "lseek failed");
|
||||
close(fd);
|
||||
return NULL;
|
||||
}
|
||||
if((addr = mmap(0, length, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0)) == MAP_FAILED) {
|
||||
EPRINTF_PATH(path, "mmap failed");
|
||||
close(fd);
|
||||
return NULL;
|
||||
}
|
||||
header = (HashHeader *)addr;
|
||||
if(memcmp(header->magic, MAGIC, 8)) {
|
||||
EPRINTF_PATH(path, "Unknown file header");
|
||||
return NULL;
|
||||
}
|
||||
if(length != sizeof(HashHeader) + header->num_buckets * (header->key_size + header->value_size)) {
|
||||
EPRINTF_PATH(path, "Incorrect file length");
|
||||
return NULL;
|
||||
}
|
||||
if(!(index = malloc(sizeof(HashIndex)))) {
|
||||
EPRINTF_PATH(path, "malloc failed");
|
||||
return NULL;
|
||||
}
|
||||
HashHeader *header = (HashHeader *)addr;
|
||||
HashIndex *index = malloc(sizeof(HashIndex));
|
||||
index->path = malloc(strlen(path) + 1);
|
||||
strcpy(index->path, path);
|
||||
index->map_addr = addr;
|
||||
index->map_length = length;
|
||||
index->num_entries = header->num_entries;
|
||||
|
@ -147,7 +183,13 @@ hashindex_open(const char *path)
|
|||
index->value_size = header->value_size;
|
||||
index->bucket_size = index->key_size + index->value_size;
|
||||
index->buckets = (addr + sizeof(HashHeader));
|
||||
index->limit = (int)(index->num_buckets * .75);
|
||||
index->lower_limit = index->num_buckets > MIN_BUCKETS ? ((int)(index->num_buckets * BUCKET_LOWER_LIMIT)) : 0;
|
||||
index->upper_limit = (int)(index->num_buckets * BUCKET_UPPER_LIMIT);
|
||||
if(!(index->path = strdup(path))) {
|
||||
EPRINTF_PATH(path, "strdup failed");
|
||||
free(index);
|
||||
return NULL;
|
||||
}
|
||||
return index;
|
||||
}
|
||||
|
||||
|
@ -155,36 +197,42 @@ static HashIndex *
|
|||
hashindex_create(const char *path, int capacity, int key_size, int value_size)
|
||||
{
|
||||
FILE *fd;
|
||||
int i;
|
||||
char bucket[MAX_BUCKET_SIZE] = {};
|
||||
int i, bucket_size;
|
||||
HashHeader header = {
|
||||
.magic = MAGIC, .num_entries = 0, .key_size = key_size, .value_size = value_size
|
||||
};
|
||||
capacity = MAX(MIN_BUCKETS, capacity);
|
||||
header.num_buckets = capacity;
|
||||
|
||||
if(!(fd = fopen(path, "w"))) {
|
||||
fprintf(stderr, "Failed to create %s\n", path);
|
||||
EPRINTF_PATH(path, "fopen failed");
|
||||
return NULL;
|
||||
}
|
||||
HashHeader header;
|
||||
memcpy(header.magic, MAGIC, sizeof(MAGIC) - 1);
|
||||
header.num_entries = 0;
|
||||
header.num_buckets = capacity;
|
||||
header.key_size = key_size;
|
||||
header.value_size = value_size;
|
||||
int bucket_size = key_size + value_size;
|
||||
char *bucket = calloc(bucket_size, 1);
|
||||
if(fwrite(&header, 1, sizeof(header), fd) != sizeof(header))
|
||||
bucket_size = key_size + value_size;
|
||||
if(fwrite(&header, 1, sizeof(header), fd) != sizeof(header)) {
|
||||
goto error;
|
||||
}
|
||||
*((int32_t *)(bucket + key_size)) = EMPTY;
|
||||
for(i = 0; i < capacity; i++) {
|
||||
if(fwrite(bucket, 1, bucket_size, fd) != bucket_size)
|
||||
if(fwrite(bucket, 1, bucket_size, fd) != bucket_size) {
|
||||
goto error;
|
||||
}
|
||||
}
|
||||
if(fclose(fd) < 0) {
|
||||
EPRINTF_PATH(path, "fclose failed");
|
||||
return NULL;
|
||||
}
|
||||
free(bucket);
|
||||
fclose(fd);
|
||||
return hashindex_open(path);
|
||||
error:
|
||||
fclose(fd);
|
||||
free(bucket);
|
||||
EPRINTF_PATH(path, "fwrite failed");
|
||||
if(fclose(fd) < 0) {
|
||||
EPRINTF_PATH(path, "fclose failed");
|
||||
}
|
||||
return NULL;
|
||||
}
|
||||
|
||||
static void
|
||||
static int
|
||||
hashindex_clear(HashIndex *index)
|
||||
{
|
||||
int i;
|
||||
|
@ -192,24 +240,36 @@ hashindex_clear(HashIndex *index)
|
|||
BUCKET_MARK_DELETED(index, i);
|
||||
}
|
||||
index->num_entries = 0;
|
||||
hashindex_resize(index, 16);
|
||||
return hashindex_resize(index, MIN_BUCKETS);
|
||||
}
|
||||
|
||||
static void
|
||||
static int
|
||||
hashindex_flush(HashIndex *index)
|
||||
{
|
||||
*((int32_t *)(index->map_addr + 8)) = index->num_entries;
|
||||
*((int32_t *)(index->map_addr + 12)) = index->num_buckets;
|
||||
msync(index->map_addr, index->map_length, MS_SYNC);
|
||||
if(msync(index->map_addr, index->map_length, MS_SYNC) < 0) {
|
||||
EPRINTF("msync failed");
|
||||
return 0;
|
||||
}
|
||||
return 1;
|
||||
}
|
||||
|
||||
static void
|
||||
static int
|
||||
hashindex_close(HashIndex *index)
|
||||
{
|
||||
hashindex_flush(index);
|
||||
munmap(index->map_addr, index->map_length);
|
||||
int rv = 1;
|
||||
|
||||
if(hashindex_flush(index) < 0) {
|
||||
rv = 0;
|
||||
}
|
||||
if(munmap(index->map_addr, index->map_length) < 0) {
|
||||
EPRINTF("munmap failed");
|
||||
rv = 0;
|
||||
}
|
||||
free(index->path);
|
||||
free(index);
|
||||
return rv;
|
||||
}
|
||||
|
||||
static const void *
|
||||
|
@ -222,15 +282,17 @@ hashindex_get(HashIndex *index, const void *key)
|
|||
return BUCKET_ADDR_READ(index, idx) + index->key_size;
|
||||
}
|
||||
|
||||
static void
|
||||
static int
|
||||
hashindex_set(HashIndex *index, const void *key, const void *value)
|
||||
{
|
||||
int idx = hashindex_lookup(index, key);
|
||||
uint8_t *ptr;
|
||||
if(idx < 0)
|
||||
{
|
||||
if(index->num_entries > index->limit) {
|
||||
hashindex_resize(index, index->num_buckets * 2);
|
||||
if(index->num_entries > index->upper_limit) {
|
||||
if(!hashindex_resize(index, index->num_buckets * 2)) {
|
||||
return 0;
|
||||
}
|
||||
}
|
||||
idx = hashindex_index(index, key);
|
||||
while(!BUCKET_IS_EMPTY(index, idx) && !BUCKET_IS_DELETED(index, idx)) {
|
||||
|
@ -245,17 +307,24 @@ hashindex_set(HashIndex *index, const void *key, const void *value)
|
|||
{
|
||||
memcpy(BUCKET_ADDR_WRITE(index, idx) + index->key_size, value, index->value_size);
|
||||
}
|
||||
return 1;
|
||||
}
|
||||
|
||||
static void
|
||||
static int
|
||||
hashindex_delete(HashIndex *index, const void *key)
|
||||
{
|
||||
int idx = hashindex_lookup(index, key);
|
||||
if (idx < 0) {
|
||||
return;
|
||||
return 1;
|
||||
}
|
||||
BUCKET_MARK_DELETED(index, idx);
|
||||
index->num_entries -= 1;
|
||||
if(index->num_entries < index->lower_limit) {
|
||||
if(!hashindex_resize(index, index->num_buckets * 2)) {
|
||||
return 0;
|
||||
}
|
||||
}
|
||||
return 1;
|
||||
}
|
||||
|
||||
static void *
|
||||
|
@ -265,12 +334,14 @@ hashindex_next_key(HashIndex *index, const void *key)
|
|||
if(key) {
|
||||
idx = 1 + (key - index->buckets) / index->bucket_size;
|
||||
}
|
||||
if (idx == index->num_buckets)
|
||||
if (idx == index->num_buckets) {
|
||||
return NULL;
|
||||
}
|
||||
while(BUCKET_IS_EMPTY(index, idx) || BUCKET_IS_DELETED(index, idx)) {
|
||||
idx ++;
|
||||
if (idx == index->num_buckets)
|
||||
if (idx == index->num_buckets) {
|
||||
return NULL;
|
||||
}
|
||||
}
|
||||
return BUCKET_ADDR_READ(index, idx);
|
||||
}
|
||||
|
|
|
@ -9,19 +9,20 @@ cdef extern from "_hashindex.c":
|
|||
HashIndex *hashindex_open(char *path)
|
||||
HashIndex *hashindex_create(char *path, int capacity, int key_size, int value_size)
|
||||
int hashindex_get_size(HashIndex *index)
|
||||
void hashindex_clear(HashIndex *index)
|
||||
void hashindex_close(HashIndex *index)
|
||||
void hashindex_flush(HashIndex *index)
|
||||
int hashindex_clear(HashIndex *index)
|
||||
int hashindex_close(HashIndex *index)
|
||||
int hashindex_flush(HashIndex *index)
|
||||
void *hashindex_get(HashIndex *index, void *key)
|
||||
void *hashindex_next_key(HashIndex *index, void *key)
|
||||
void hashindex_delete(HashIndex *index, void *key)
|
||||
void hashindex_set(HashIndex *index, void *key, void *value)
|
||||
int hashindex_delete(HashIndex *index, void *key)
|
||||
int hashindex_set(HashIndex *index, void *key, void *value)
|
||||
|
||||
|
||||
_NoDefault = object()
|
||||
|
||||
cdef class IndexBase:
|
||||
cdef HashIndex *index
|
||||
key_size = 32
|
||||
|
||||
def __cinit__(self, path):
|
||||
self.index = hashindex_open(<bytes>os.fsencode(path))
|
||||
|
@ -30,18 +31,34 @@ cdef class IndexBase:
|
|||
|
||||
def __dealloc__(self):
|
||||
if self.index:
|
||||
hashindex_close(self.index)
|
||||
if not hashindex_close(self.index):
|
||||
raise Exception('hashindex_close failed')
|
||||
|
||||
@classmethod
|
||||
def create(cls, path):
|
||||
index = hashindex_create(<bytes>os.fsencode(path), 0, cls.key_size, cls.value_size)
|
||||
if not index:
|
||||
raise Exception('Failed to create %s' % path)
|
||||
hashindex_close(index)
|
||||
return cls(path)
|
||||
|
||||
def clear(self):
|
||||
hashindex_clear(self.index)
|
||||
if not hashindex_clear(self.index):
|
||||
raise Exception('hashindex_clear failed')
|
||||
|
||||
def flush(self):
|
||||
hashindex_flush(self.index)
|
||||
if not hashindex_flush(self.index):
|
||||
raise Exception('hashindex_flush failed')
|
||||
|
||||
def setdefault(self, key, value):
|
||||
if not key in self:
|
||||
self[key] = value
|
||||
|
||||
def __delitem__(self, key):
|
||||
assert len(key) == 32
|
||||
if not hashindex_delete(self.index, <char *>key):
|
||||
raise Exception('hashindex_delete failed')
|
||||
|
||||
def get(self, key, default=None):
|
||||
try:
|
||||
return self[key]
|
||||
|
@ -64,13 +81,7 @@ cdef class IndexBase:
|
|||
|
||||
cdef class NSIndex(IndexBase):
|
||||
|
||||
@classmethod
|
||||
def create(cls, path, capacity=16):
|
||||
index = hashindex_create(<bytes>os.fsencode(path), capacity, 32, 8)
|
||||
if not index:
|
||||
raise Exception('Failed to create %s' % path)
|
||||
hashindex_close(index)
|
||||
return cls(path)
|
||||
value_size = 8
|
||||
|
||||
def __getitem__(self, key):
|
||||
assert len(key) == 32
|
||||
|
@ -79,16 +90,13 @@ cdef class NSIndex(IndexBase):
|
|||
raise KeyError
|
||||
return data[0], data[1]
|
||||
|
||||
def __delitem__(self, key):
|
||||
assert len(key) == 32
|
||||
hashindex_delete(self.index, <char *>key)
|
||||
|
||||
def __setitem__(self, key, value):
|
||||
assert len(key) == 32
|
||||
cdef int[2] data
|
||||
data[0] = value[0]
|
||||
data[1] = value[1]
|
||||
hashindex_set(self.index, <char *>key, data)
|
||||
if not hashindex_set(self.index, <char *>key, data):
|
||||
raise Exception('hashindex_set failed')
|
||||
|
||||
def __contains__(self, key):
|
||||
assert len(key) == 32
|
||||
|
@ -121,11 +129,7 @@ cdef class NSKeyIterator:
|
|||
|
||||
cdef class ChunkIndex(IndexBase):
|
||||
|
||||
@classmethod
|
||||
def create(cls, path, capacity=16):
|
||||
index = hashindex_create(<bytes>os.fsencode(path), capacity, 32, 12)
|
||||
hashindex_close(index)
|
||||
return cls(path)
|
||||
value_size = 12
|
||||
|
||||
def __getitem__(self, key):
|
||||
assert len(key) == 32
|
||||
|
@ -134,17 +138,14 @@ cdef class ChunkIndex(IndexBase):
|
|||
raise KeyError
|
||||
return data[0], data[1], data[2]
|
||||
|
||||
def __delitem__(self, key):
|
||||
assert len(key) == 32
|
||||
hashindex_delete(self.index, <char *>key)
|
||||
|
||||
def __setitem__(self, key, value):
|
||||
assert len(key) == 32
|
||||
cdef int[3] data
|
||||
data[0] = value[0]
|
||||
data[1] = value[1]
|
||||
data[2] = value[2]
|
||||
hashindex_set(self.index, <char *>key, data)
|
||||
if not hashindex_set(self.index, <char *>key, data):
|
||||
raise Exception('hashindex_set failed')
|
||||
|
||||
def __contains__(self, key):
|
||||
assert len(key) == 32
|
||||
|
|
|
@ -41,8 +41,8 @@ class HashIndexTestCase(DarcTestCase):
|
|||
self.assert_equal(len(cls(idx_name.name)), 0)
|
||||
|
||||
def test_nsindex(self):
|
||||
self._generic_test(NSIndex, lambda x: (x, x), '926af057de19a82903fab517e0813e6fa16e4129df5b7062e9b1f24bbe98fb24')
|
||||
self._generic_test(NSIndex, lambda x: (x, x), '9a6f9cb3c03d83ed611265eeef1f9a9d69c2f0417a35ac14d56ce573d0c8b356')
|
||||
|
||||
def test_chunkindex(self):
|
||||
self._generic_test(ChunkIndex, lambda x: (x, x, x), 'cc60ec9265d172b20c33b6c00160abda1a6e881d07088b9d1c5997a82e05f5a7')
|
||||
self._generic_test(ChunkIndex, lambda x: (x, x, x), '9c35f237e533b6d2533d2646da127052d615ab9b66de65a795cd922b337741ca')
|
||||
|
||||
|
|
Loading…
Reference in New Issue