X-Git-Url: https://git.kernel.dk/?a=blobdiff_plain;f=t%2Fdedupe.c;h=2ef8dc539809884cc0372597ef1a260385ee9165;hb=36d108a47d4f97927b3acccab6cb79f75af5f88b;hp=16c1748d86f19131f7985daa9b6eecda168fbbf0;hpb=1956a9e5ce9947a81a3c6de619957b6d1313b3f9;p=fio.git diff --git a/t/dedupe.c b/t/dedupe.c index 16c1748d..2ef8dc53 100644 --- a/t/dedupe.c +++ b/t/dedupe.c @@ -3,37 +3,27 @@ * just scans the filename for extents of the given size, checksums them, * and orders them up. */ +#include +#include #include -#include +#include #include -#include -#include -#include #include -#include -#include -#include -#include -#include "../lib/rbtree.h" #include "../flist.h" #include "../log.h" -#include "../mutex.h" +#include "../fio_sem.h" #include "../smalloc.h" #include "../minmax.h" #include "../crc/md5.h" -#include "../memalign.h" +#include "../lib/memalign.h" #include "../os/os.h" #include "../gettime.h" #include "../fio_time.h" +#include "../lib/rbtree.h" -FILE *f_err; -struct timeval *fio_tv = NULL; -unsigned int fio_debug = 0; - -void __dprint(int type, const char *str, ...) -{ -} +#include "../lib/bloom.h" +#include "debug.h" struct worker_thread { pthread_t thread; @@ -45,6 +35,7 @@ struct worker_thread { uint64_t size; unsigned long items; + unsigned long dupes; int err; }; @@ -54,7 +45,7 @@ struct extent { }; struct chunk { - struct rb_node rb_node; + struct fio_rb_node rb_node; uint64_t count; uint32_t hash[MD5_HASH_WORDS]; struct flist_head extent_list[0]; @@ -66,7 +57,8 @@ struct item { }; static struct rb_root rb_root; -static struct fio_mutex *rb_lock; +static struct bloom *bloom; +static struct fio_sem *rb_lock; static unsigned int blocksize = 4096; static unsigned int num_threads; @@ -75,22 +67,26 @@ static unsigned int dump_output; static unsigned int odirect; static unsigned int collision_check; static unsigned int print_progress = 1; +static unsigned int use_bloom = 1; static uint64_t total_size; static uint64_t cur_offset; -static struct fio_mutex *size_lock; +static struct fio_sem *size_lock; -static int dev_fd; +static struct fio_file file; -static uint64_t get_size(int fd, struct stat *sb) +static uint64_t get_size(struct fio_file *f, struct stat *sb) { uint64_t ret; if (S_ISBLK(sb->st_mode)) { - if (ioctl(fd, BLKGETSIZE64, &ret) < 0) { - perror("ioctl"); + unsigned long long bytes = 0; + + if (blockdev_size(f, &bytes)) { + log_err("dedupe: failed getting bdev size\n"); return 0; } + ret = bytes; } else ret = sb->st_size; @@ -102,7 +98,7 @@ static int get_work(uint64_t *offset, uint64_t *size) uint64_t this_chunk; int ret = 1; - fio_mutex_down(size_lock); + fio_sem_down(size_lock); if (cur_offset < total_size) { *offset = cur_offset; @@ -112,21 +108,21 @@ static int get_work(uint64_t *offset, uint64_t *size) ret = 0; } - fio_mutex_up(size_lock); + fio_sem_up(size_lock); return ret; } -static int read_block(int fd, void *buf, off_t offset) +static int __read_block(int fd, void *buf, off_t offset, size_t count) { ssize_t ret; - ret = pread(fd, buf, blocksize, offset); + ret = pread(fd, buf, count, offset); if (ret < 0) { perror("pread"); return 1; } else if (!ret) return 1; - else if (ret != blocksize) { + else if (ret != count) { log_err("dedupe: short read on block\n"); return 1; } @@ -134,6 +130,11 @@ static int read_block(int fd, void *buf, off_t offset) return 0; } +static int read_block(int fd, void *buf, off_t offset) +{ + return __read_block(fd, buf, offset, blocksize); +} + static void add_item(struct chunk *c, struct item *i) { /* @@ -157,20 +158,20 @@ static int col_check(struct chunk *c, struct item *i) char *cbuf, *ibuf; int ret = 1; - cbuf = fio_memalign(blocksize, blocksize); - ibuf = fio_memalign(blocksize, blocksize); + cbuf = fio_memalign(blocksize, blocksize, false); + ibuf = fio_memalign(blocksize, blocksize, false); e = flist_entry(c->extent_list[0].next, struct extent, list); - if (read_block(dev_fd, cbuf, e->offset)) + if (read_block(file.fd, cbuf, e->offset)) goto out; - if (read_block(dev_fd, ibuf, i->offset)) + if (read_block(file.fd, ibuf, i->offset)) goto out; ret = memcmp(ibuf, cbuf, blocksize); out: - fio_memfree(cbuf, blocksize); - fio_memfree(ibuf, blocksize); + fio_memfree(cbuf, blocksize, false); + fio_memfree(ibuf, blocksize, false); return ret; } @@ -189,7 +190,7 @@ static struct chunk *alloc_chunk(void) static void insert_chunk(struct item *i) { - struct rb_node **p, *parent; + struct fio_rb_node **p, *parent; struct chunk *c; int diff; @@ -210,9 +211,9 @@ static void insert_chunk(struct item *i) if (!collision_check) goto add; - fio_mutex_up(rb_lock); + fio_sem_up(rb_lock); ret = col_check(c, i); - fio_mutex_down(rb_lock); + fio_sem_down(rb_lock); if (!ret) goto add; @@ -231,16 +232,26 @@ add: add_item(c, i); } -static void insert_chunks(struct item *items, unsigned int nitems) +static void insert_chunks(struct item *items, unsigned int nitems, + uint64_t *ndupes) { int i; - fio_mutex_down(rb_lock); + fio_sem_down(rb_lock); + + for (i = 0; i < nitems; i++) { + if (bloom) { + unsigned int s; + int r; - for (i = 0; i < nitems; i++) - insert_chunk(&items[i]); + s = sizeof(items[i].hash) / sizeof(uint32_t); + r = bloom_set(bloom, items[i].hash, s); + *ndupes += r; + } else + insert_chunk(&items[i]); + } - fio_mutex_up(rb_lock); + fio_sem_up(rb_lock); } static void crc_buf(void *buf, uint32_t *hash) @@ -252,30 +263,45 @@ static void crc_buf(void *buf, uint32_t *hash) fio_md5_final(&ctx); } +static unsigned int read_blocks(int fd, void *buf, off_t offset, size_t size) +{ + if (__read_block(fd, buf, offset, size)) + return 0; + + return size / blocksize; +} + static int do_work(struct worker_thread *thread, void *buf) { unsigned int nblocks, i; off_t offset; - int err = 0, nitems = 0; + int nitems = 0; + uint64_t ndupes = 0; struct item *items; - nblocks = thread->size / blocksize; offset = thread->cur_offset; + + nblocks = read_blocks(thread->fd, buf, offset, min(thread->size, (uint64_t)chunk_size)); + if (!nblocks) + return 1; + items = malloc(sizeof(*items) * nblocks); for (i = 0; i < nblocks; i++) { - if (read_block(thread->fd, buf, offset)) - break; + void *thisptr = buf + (i * blocksize); + items[i].offset = offset; - crc_buf(buf, items[i].hash); + crc_buf(thisptr, items[i].hash); offset += blocksize; nitems++; } - insert_chunks(items, nitems); - thread->items += nitems; + insert_chunks(items, nitems, &ndupes); + free(items); - return err; + thread->items += nitems; + thread->dupes += ndupes; + return 0; } static void *thread_fn(void *data) @@ -283,7 +309,7 @@ static void *thread_fn(void *data) struct worker_thread *thread = data; void *buf; - buf = fio_memalign(blocksize, blocksize); + buf = fio_memalign(blocksize, chunk_size, false); do { if (get_work(&thread->cur_offset, &thread->size)) { @@ -297,14 +323,14 @@ static void *thread_fn(void *data) } while (1); thread->done = 1; - fio_memfree(buf, blocksize); + fio_memfree(buf, chunk_size, false); return NULL; } static void show_progress(struct worker_thread *threads, unsigned long total) { unsigned long last_nitems = 0; - struct timeval last_tv; + struct timespec last_tv; fio_gettime(&last_tv, NULL); @@ -313,7 +339,7 @@ static void show_progress(struct worker_thread *threads, unsigned long total) unsigned long nitems = 0; uint64_t tdiff; float perc; - int some_done; + int some_done = 0; int i; for (i = 0; i < num_threads; i++) { @@ -332,8 +358,8 @@ static void show_progress(struct worker_thread *threads, unsigned long total) this_items *= blocksize; tdiff = mtime_since_now(&last_tv); if (tdiff) { - this_items /= tdiff; - printf("%3.2f%% done (%luKB/sec)\r", perc, this_items); + this_items = (this_items * 1000) / (tdiff * 1024); + printf("%3.2f%% done (%luKiB/sec)\r", perc, this_items); last_nitems = nitems; fio_gettime(&last_tv, NULL); } else @@ -343,7 +369,8 @@ static void show_progress(struct worker_thread *threads, unsigned long total) }; } -static int run_dedupe_threads(int fd, uint64_t dev_size) +static int run_dedupe_threads(struct fio_file *f, uint64_t dev_size, + uint64_t *nextents, uint64_t *nchunks) { struct worker_thread *threads; unsigned long nitems, total_items; @@ -352,14 +379,12 @@ static int run_dedupe_threads(int fd, uint64_t dev_size) total_size = dev_size; total_items = dev_size / blocksize; cur_offset = 0; - size_lock = fio_mutex_init(FIO_MUTEX_UNLOCKED); + size_lock = fio_sem_init(FIO_SEM_UNLOCKED); threads = malloc(num_threads * sizeof(struct worker_thread)); for (i = 0; i < num_threads; i++) { - threads[i].fd = fd; - threads[i].items = 0; - threads[i].err = 0; - threads[i].done = 0; + memset(&threads[i], 0, sizeof(struct worker_thread)); + threads[i].fd = f->fd; err = pthread_create(&threads[i].thread, NULL, thread_fn, &threads[i]); if (err) { @@ -371,20 +396,27 @@ static int run_dedupe_threads(int fd, uint64_t dev_size) show_progress(threads, total_items); nitems = 0; + *nextents = 0; + *nchunks = 1; for (i = 0; i < num_threads; i++) { void *ret; pthread_join(threads[i].thread, &ret); nitems += threads[i].items; + *nchunks += threads[i].dupes; } printf("Threads(%u): %lu items processed\n", num_threads, nitems); - fio_mutex_remove(size_lock); + *nextents = nitems; + *nchunks = nitems - *nchunks; + + fio_sem_remove(size_lock); free(threads); return err; } -static int dedupe_check(const char *filename) +static int dedupe_check(const char *filename, uint64_t *nextents, + uint64_t *nchunks) { uint64_t dev_size; struct stat sb; @@ -392,29 +424,41 @@ static int dedupe_check(const char *filename) flags = O_RDONLY; if (odirect) - flags |= O_DIRECT; + flags |= OS_O_DIRECT; - dev_fd = open(filename, flags); - if (dev_fd == -1) { + memset(&file, 0, sizeof(file)); + file.file_name = strdup(filename); + + file.fd = open(filename, flags); + if (file.fd == -1) { perror("open"); - return 1; + goto err; } - if (fstat(dev_fd, &sb) < 0) { + if (fstat(file.fd, &sb) < 0) { perror("fstat"); - close(dev_fd); - return 1; + goto err; } - dev_size = get_size(dev_fd, &sb); - if (!dev_size) { - close(dev_fd); - return 1; + dev_size = get_size(&file, &sb); + if (!dev_size) + goto err; + + if (use_bloom) { + uint64_t bloom_entries; + + bloom_entries = 8 * (dev_size / blocksize); + bloom = bloom_new(bloom_entries); } - printf("Will check <%s>, size <%llu>\n", filename, (unsigned long long) dev_size); + printf("Will check <%s>, size <%llu>, using %u threads\n", filename, (unsigned long long) dev_size, num_threads); - return run_dedupe_threads(dev_fd, dev_size); + return run_dedupe_threads(&file, dev_size, nextents, nchunks); +err: + if (file.fd != -1) + close(file.fd); + free(file.file_name); + return 1; } static void show_chunk(struct chunk *c) @@ -429,14 +473,29 @@ static void show_chunk(struct chunk *c) } } -static void iter_rb_tree(void) +static void show_stat(uint64_t nextents, uint64_t nchunks) { - struct rb_node *n; - uint64_t nchunks; - uint64_t nextents; - double perc; + double perc, ratio; + + printf("Extents=%lu, Unique extents=%lu\n", (unsigned long) nextents, (unsigned long) nchunks); + + if (nchunks) { + ratio = (double) nextents / (double) nchunks; + printf("De-dupe ratio: 1:%3.2f\n", ratio - 1.0); + } else + printf("De-dupe ratio: 1:infinite\n"); - nchunks = nextents = 0; + perc = 1.00 - ((double) nchunks / (double) nextents); + perc *= 100.0; + printf("Fio setting: dedupe_percentage=%u\n", (int) (perc + 0.50)); + +} + +static void iter_rb_tree(uint64_t *nextents, uint64_t *nchunks) +{ + struct fio_rb_node *n; + + *nchunks = *nextents = 0; n = rb_first(&rb_root); if (!n) @@ -446,20 +505,13 @@ static void iter_rb_tree(void) struct chunk *c; c = rb_entry(n, struct chunk, rb_node); - nchunks++; - nextents += c->count; + (*nchunks)++; + *nextents += c->count; if (dump_output) show_chunk(c); } while ((n = rb_next(n)) != NULL); - - printf("Extents=%lu, Unique extents=%lu\n", (unsigned long) nextents, (unsigned long) nchunks); - printf("De-dupe factor: %3.2f\n", (double) nextents / (double) nchunks); - - perc = 1.00 - ((double) nchunks / (double) nextents); - perc *= 100.0; - printf("Fio setting: dedupe_percentage=%u\n", (int) (perc + 0.50)); } static int usage(char *argv[]) @@ -471,15 +523,20 @@ static int usage(char *argv[]) log_err("\t-d\tFull extent/chunk debug output\n"); log_err("\t-o\tUse O_DIRECT\n"); log_err("\t-c\tFull collision check\n"); + log_err("\t-B\tUse probabilistic bloom filter\n"); log_err("\t-p\tPrint progress indicator\n"); return 1; } int main(int argc, char *argv[]) { + uint64_t nextents = 0, nchunks = 0; int c, ret; - while ((c = getopt(argc, argv, "b:t:d:o:c:p:")) != -1) { + arch_init(argv); + debug_init(); + + while ((c = getopt(argc, argv, "b:t:d:o:c:p:B:")) != -1) { switch (c) { case 'b': blocksize = atoi(optarg); @@ -499,12 +556,18 @@ int main(int argc, char *argv[]) case 'p': print_progress = atoi(optarg); break; + case 'B': + use_bloom = atoi(optarg); + break; case '?': default: return usage(argv); } } + if (collision_check || dump_output) + use_bloom = 0; + if (!num_threads) num_threads = cpus_online(); @@ -514,13 +577,20 @@ int main(int argc, char *argv[]) sinit(); rb_root = RB_ROOT; - rb_lock = fio_mutex_init(FIO_MUTEX_UNLOCKED); + rb_lock = fio_sem_init(FIO_SEM_UNLOCKED); - ret = dedupe_check(argv[optind]); + ret = dedupe_check(argv[optind], &nextents, &nchunks); - iter_rb_tree(); + if (!ret) { + if (!bloom) + iter_rb_tree(&nextents, &nchunks); + + show_stat(nextents, nchunks); + } - fio_mutex_remove(rb_lock); + fio_sem_remove(rb_lock); + if (bloom) + bloom_free(bloom); scleanup(); return ret; }