diff options
Diffstat (limited to 'drivers/md/bcache')
-rw-r--r-- | drivers/md/bcache/alloc.c | 5 | ||||
-rw-r--r-- | drivers/md/bcache/bcache.h | 7 | ||||
-rw-r--r-- | drivers/md/bcache/bset.c | 2 | ||||
-rw-r--r-- | drivers/md/bcache/bset.h | 3 | ||||
-rw-r--r-- | drivers/md/bcache/btree.c | 12 | ||||
-rw-r--r-- | drivers/md/bcache/journal.c | 4 | ||||
-rw-r--r-- | drivers/md/bcache/request.c | 17 | ||||
-rw-r--r-- | drivers/md/bcache/stats.c | 10 | ||||
-rw-r--r-- | drivers/md/bcache/super.c | 144 | ||||
-rw-r--r-- | drivers/md/bcache/sysfs.c | 22 | ||||
-rw-r--r-- | drivers/md/bcache/writeback.c | 14 | ||||
-rw-r--r-- | drivers/md/bcache/writeback.h | 19 |
12 files changed, 189 insertions, 70 deletions
diff --git a/drivers/md/bcache/alloc.c b/drivers/md/bcache/alloc.c index 6f776823b9ba..a1df0d95151c 100644 --- a/drivers/md/bcache/alloc.c +++ b/drivers/md/bcache/alloc.c @@ -377,7 +377,10 @@ retry_invalidate: if (!fifo_full(&ca->free_inc)) goto retry_invalidate; - bch_prio_write(ca); + if (bch_prio_write(ca, false) < 0) { + ca->invalidate_needs_gc = 1; + wake_up_gc(ca->set); + } } } out: diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h index e30a983a68cd..2e396954b39d 100644 --- a/drivers/md/bcache/bcache.h +++ b/drivers/md/bcache/bcache.h @@ -264,7 +264,7 @@ struct bcache_device { #define BCACHE_DEV_UNLINK_DONE 2 #define BCACHE_DEV_WB_RUNNING 3 #define BCACHE_DEV_RATE_DW_RUNNING 4 - unsigned int nr_stripes; + int nr_stripes; unsigned int stripe_size; atomic_t *stripe_sectors_dirty; unsigned long *full_dirty_stripes; @@ -329,6 +329,9 @@ struct cached_dev { */ atomic_t has_dirty; +#define BCH_CACHE_READA_ALL 0 +#define BCH_CACHE_READA_META_ONLY 1 + unsigned int cache_readahead_policy; struct bch_ratelimit writeback_rate; struct delayed_work writeback_rate_update; @@ -977,7 +980,7 @@ bool bch_cached_dev_error(struct cached_dev *dc); __printf(2, 3) bool bch_cache_set_error(struct cache_set *c, const char *fmt, ...); -void bch_prio_write(struct cache *ca); +int bch_prio_write(struct cache *ca, bool wait); void bch_write_bdev_super(struct cached_dev *dc, struct closure *parent); extern struct workqueue_struct *bcache_wq; diff --git a/drivers/md/bcache/bset.c b/drivers/md/bcache/bset.c index 268f1b685084..ec48cf86cab6 100644 --- a/drivers/md/bcache/bset.c +++ b/drivers/md/bcache/bset.c @@ -321,7 +321,7 @@ int bch_btree_keys_alloc(struct btree_keys *b, b->page_order = page_order; - t->data = (void *) __get_free_pages(gfp, b->page_order); + t->data = (void *) __get_free_pages(__GFP_COMP|gfp, b->page_order); if (!t->data) goto err; diff --git a/drivers/md/bcache/bset.h b/drivers/md/bcache/bset.h index c71365e7c1fa..a50dcfda656f 100644 --- a/drivers/md/bcache/bset.h +++ b/drivers/md/bcache/bset.h @@ -397,7 +397,8 @@ void bch_btree_keys_stats(struct btree_keys *b, struct bset_stats *state); /* Bkey utility code */ -#define bset_bkey_last(i) bkey_idx((struct bkey *) (i)->d, (i)->keys) +#define bset_bkey_last(i) bkey_idx((struct bkey *) (i)->d, \ + (unsigned int)(i)->keys) static inline struct bkey *bset_bkey_idx(struct bset *i, unsigned int idx) { diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c index 5cf3247e8afb..b60cb7a99b18 100644 --- a/drivers/md/bcache/btree.c +++ b/drivers/md/bcache/btree.c @@ -719,6 +719,8 @@ static unsigned long bch_mca_scan(struct shrinker *shrink, * IO can always make forward progress: */ nr /= c->btree_pages; + if (nr == 0) + nr = 1; nr = min_t(unsigned long, nr, mca_can_free(c)); i = 0; @@ -834,7 +836,7 @@ int bch_btree_cache_alloc(struct cache_set *c) mutex_init(&c->verify_lock); c->verify_ondisk = (void *) - __get_free_pages(GFP_KERNEL, ilog2(bucket_pages(c))); + __get_free_pages(GFP_KERNEL|__GFP_COMP, ilog2(bucket_pages(c))); c->verify_data = mca_bucket_alloc(c, &ZERO_KEY, GFP_KERNEL); @@ -1436,7 +1438,7 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op, if (__set_blocks(n1, n1->keys + n2->keys, block_bytes(b->c)) > btree_blocks(new_nodes[i])) - goto out_nocoalesce; + goto out_unlock_nocoalesce; keys = n2->keys; /* Take the key of the node we're getting rid of */ @@ -1465,7 +1467,7 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op, if (__bch_keylist_realloc(&keylist, bkey_u64s(&new_nodes[i]->key))) - goto out_nocoalesce; + goto out_unlock_nocoalesce; bch_btree_node_write(new_nodes[i], &cl); bch_keylist_add(&keylist, &new_nodes[i]->key); @@ -1511,6 +1513,10 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op, /* Invalidated our iterator */ return -EINTR; +out_unlock_nocoalesce: + for (i = 0; i < nodes; i++) + mutex_unlock(&new_nodes[i]->write_lock); + out_nocoalesce: closure_sync(&cl); diff --git a/drivers/md/bcache/journal.c b/drivers/md/bcache/journal.c index 33556acdcf9c..e145c2b1e2c9 100644 --- a/drivers/md/bcache/journal.c +++ b/drivers/md/bcache/journal.c @@ -865,8 +865,8 @@ int bch_journal_alloc(struct cache_set *c) j->w[1].c = c; if (!(init_fifo(&j->pin, JOURNAL_PIN, GFP_KERNEL)) || - !(j->w[0].data = (void *) __get_free_pages(GFP_KERNEL, JSET_BITS)) || - !(j->w[1].data = (void *) __get_free_pages(GFP_KERNEL, JSET_BITS))) + !(j->w[0].data = (void *) __get_free_pages(GFP_KERNEL|__GFP_COMP, JSET_BITS)) || + !(j->w[1].data = (void *) __get_free_pages(GFP_KERNEL|__GFP_COMP, JSET_BITS))) return -ENOMEM; return 0; diff --git a/drivers/md/bcache/request.c b/drivers/md/bcache/request.c index 41adcd1546f1..4045ae748f17 100644 --- a/drivers/md/bcache/request.c +++ b/drivers/md/bcache/request.c @@ -391,13 +391,20 @@ static bool check_should_bypass(struct cached_dev *dc, struct bio *bio) goto skip; /* - * Flag for bypass if the IO is for read-ahead or background, - * unless the read-ahead request is for metadata + * If the bio is for read-ahead or background IO, bypass it or + * not depends on the following situations, + * - If the IO is for meta data, always cache it and no bypass + * - If the IO is not meta data, check dc->cache_reada_policy, + * BCH_CACHE_READA_ALL: cache it and not bypass + * BCH_CACHE_READA_META_ONLY: not cache it and bypass + * That is, read-ahead request for metadata always get cached * (eg, for gfs2 or xfs). */ - if (bio->bi_opf & (REQ_RAHEAD|REQ_BACKGROUND) && - !(bio->bi_opf & (REQ_META|REQ_PRIO))) - goto skip; + if ((bio->bi_opf & (REQ_RAHEAD|REQ_BACKGROUND))) { + if (!(bio->bi_opf & (REQ_META|REQ_PRIO)) && + (dc->cache_readahead_policy != BCH_CACHE_READA_ALL)) + goto skip; + } if (bio->bi_iter.bi_sector & (c->sb.block_size - 1) || bio_sectors(bio) & (c->sb.block_size - 1)) { diff --git a/drivers/md/bcache/stats.c b/drivers/md/bcache/stats.c index ba1c93791d8d..503aafe188dc 100644 --- a/drivers/md/bcache/stats.c +++ b/drivers/md/bcache/stats.c @@ -109,9 +109,13 @@ int bch_cache_accounting_add_kobjs(struct cache_accounting *acc, void bch_cache_accounting_clear(struct cache_accounting *acc) { - memset(&acc->total.cache_hits, - 0, - sizeof(struct cache_stats)); + acc->total.cache_hits = 0; + acc->total.cache_misses = 0; + acc->total.cache_bypass_hits = 0; + acc->total.cache_bypass_misses = 0; + acc->total.cache_readaheads = 0; + acc->total.cache_miss_collisions = 0; + acc->total.sectors_bypassed = 0; } void bch_cache_accounting_destroy(struct cache_accounting *acc) diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c index 6daf777105fb..5248bca5e2ed 100644 --- a/drivers/md/bcache/super.c +++ b/drivers/md/bcache/super.c @@ -527,12 +527,29 @@ static void prio_io(struct cache *ca, uint64_t bucket, int op, closure_sync(cl); } -void bch_prio_write(struct cache *ca) +int bch_prio_write(struct cache *ca, bool wait) { int i; struct bucket *b; struct closure cl; + pr_debug("free_prio=%zu, free_none=%zu, free_inc=%zu", + fifo_used(&ca->free[RESERVE_PRIO]), + fifo_used(&ca->free[RESERVE_NONE]), + fifo_used(&ca->free_inc)); + + /* + * Pre-check if there are enough free buckets. In the non-blocking + * scenario it's better to fail early rather than starting to allocate + * buckets and do a cleanup later in case of failure. + */ + if (!wait) { + size_t avail = fifo_used(&ca->free[RESERVE_PRIO]) + + fifo_used(&ca->free[RESERVE_NONE]); + if (prio_buckets(ca) > avail) + return -ENOMEM; + } + closure_init_stack(&cl); lockdep_assert_held(&ca->set->bucket_lock); @@ -542,9 +559,6 @@ void bch_prio_write(struct cache *ca) atomic_long_add(ca->sb.bucket_size * prio_buckets(ca), &ca->meta_sectors_written); - //pr_debug("free %zu, free_inc %zu, unused %zu", fifo_used(&ca->free), - // fifo_used(&ca->free_inc), fifo_used(&ca->unused)); - for (i = prio_buckets(ca) - 1; i >= 0; --i) { long bucket; struct prio_set *p = ca->disk_buckets; @@ -562,7 +576,7 @@ void bch_prio_write(struct cache *ca) p->magic = pset_magic(&ca->sb); p->csum = bch_crc64(&p->magic, bucket_bytes(ca) - 8); - bucket = bch_bucket_alloc(ca, RESERVE_PRIO, true); + bucket = bch_bucket_alloc(ca, RESERVE_PRIO, wait); BUG_ON(bucket == -1); mutex_unlock(&ca->set->bucket_lock); @@ -591,6 +605,7 @@ void bch_prio_write(struct cache *ca) ca->prio_last_buckets[i] = ca->prio_buckets[i]; } + return 0; } static void prio_read(struct cache *ca, uint64_t bucket) @@ -754,20 +769,31 @@ static inline int idx_to_first_minor(int idx) static void bcache_device_free(struct bcache_device *d) { + struct gendisk *disk = d->disk; + lockdep_assert_held(&bch_register_lock); - pr_info("%s stopped", d->disk->disk_name); + if (disk) + pr_info("%s stopped", disk->disk_name); + else + pr_err("bcache device (NULL gendisk) stopped"); if (d->c) bcache_device_detach(d); - if (d->disk && d->disk->flags & GENHD_FL_UP) - del_gendisk(d->disk); - if (d->disk && d->disk->queue) - blk_cleanup_queue(d->disk->queue); - if (d->disk) { + + if (disk) { + bool disk_added = (disk->flags & GENHD_FL_UP) != 0; + + if (disk_added) + del_gendisk(disk); + + if (disk->queue) + blk_cleanup_queue(disk->queue); + ida_simple_remove(&bcache_device_idx, - first_minor_to_idx(d->disk->first_minor)); - put_disk(d->disk); + first_minor_to_idx(disk->first_minor)); + if (disk_added) + put_disk(disk); } bioset_exit(&d->bio_split); @@ -1209,6 +1235,9 @@ static void cached_dev_free(struct closure *cl) mutex_unlock(&bch_register_lock); + if (dc->sb_bio.bi_inline_vecs[0].bv_page) + put_page(bio_first_page_all(&dc->sb_bio)); + if (!IS_ERR_OR_NULL(dc->bdev)) blkdev_put(dc->bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL); @@ -1674,7 +1703,7 @@ void bch_cache_set_unregister(struct cache_set *c) } #define alloc_bucket_pages(gfp, c) \ - ((void *) __get_free_pages(__GFP_ZERO|gfp, ilog2(bucket_pages(c)))) + ((void *) __get_free_pages(__GFP_ZERO|__GFP_COMP|gfp, ilog2(bucket_pages(c)))) struct cache_set *bch_cache_set_alloc(struct cache_sb *sb) { @@ -1886,7 +1915,7 @@ static int run_cache_set(struct cache_set *c) mutex_lock(&c->bucket_lock); for_each_cache(ca, c, i) - bch_prio_write(ca); + bch_prio_write(ca, true); mutex_unlock(&c->bucket_lock); err = "cannot allocate new UUID bucket"; @@ -1994,7 +2023,14 @@ found: sysfs_create_link(&c->kobj, &ca->kobj, buf)) goto err; - if (ca->sb.seq > c->sb.seq) { + /* + * A special case is both ca->sb.seq and c->sb.seq are 0, + * such condition happens on a new created cache device whose + * super block is never flushed yet. In this case c->sb.version + * and other members should be updated too, otherwise we will + * have a mistaken super block version in cache set. + */ + if (ca->sb.seq > c->sb.seq || c->sb.seq == 0) { c->sb.version = ca->sb.version; memcpy(c->sb.set_uuid, ca->sb.set_uuid, 16); c->sb.flags = ca->sb.flags; @@ -2274,29 +2310,35 @@ static bool bch_is_open(struct block_device *bdev) static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr, const char *buffer, size_t size) { - ssize_t ret = -EINVAL; - const char *err = "cannot allocate memory"; + const char *err; char *path = NULL; - struct cache_sb *sb = NULL; + struct cache_sb *sb; struct block_device *bdev = NULL; - struct page *sb_page = NULL; + struct page *sb_page; + ssize_t ret; + ret = -EBUSY; + err = "failed to reference bcache module"; if (!try_module_get(THIS_MODULE)) - return -EBUSY; + goto out; /* For latest state of bcache_is_reboot */ smp_mb(); + err = "bcache is in reboot"; if (bcache_is_reboot) - return -EBUSY; + goto out_module_put; + ret = -ENOMEM; + err = "cannot allocate memory"; path = kstrndup(buffer, size, GFP_KERNEL); if (!path) - goto err; + goto out_module_put; sb = kmalloc(sizeof(struct cache_sb), GFP_KERNEL); if (!sb) - goto err; + goto out_free_path; + ret = -EINVAL; err = "failed to open device"; bdev = blkdev_get_by_path(strim(path), FMODE_READ|FMODE_WRITE|FMODE_EXCL, @@ -2313,57 +2355,69 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr, if (!IS_ERR(bdev)) bdput(bdev); if (attr == &ksysfs_register_quiet) - goto quiet_out; + goto done; } - goto err; + goto out_free_sb; } err = "failed to set blocksize"; if (set_blocksize(bdev, 4096)) - goto err_close; + goto out_blkdev_put; err = read_super(sb, bdev, &sb_page); if (err) - goto err_close; + goto out_blkdev_put; err = "failed to register device"; if (SB_IS_BDEV(sb)) { struct cached_dev *dc = kzalloc(sizeof(*dc), GFP_KERNEL); if (!dc) - goto err_close; + goto out_put_sb_page; mutex_lock(&bch_register_lock); ret = register_bdev(sb, sb_page, bdev, dc); mutex_unlock(&bch_register_lock); /* blkdev_put() will be called in cached_dev_free() */ - if (ret < 0) - goto err; + if (ret < 0) { + bdev = NULL; + goto out_put_sb_page; + } } else { struct cache *ca = kzalloc(sizeof(*ca), GFP_KERNEL); if (!ca) - goto err_close; + goto out_put_sb_page; /* blkdev_put() will be called in bch_cache_release() */ - if (register_cache(sb, sb_page, bdev, ca) != 0) - goto err; + if (register_cache(sb, sb_page, bdev, ca) != 0) { + bdev = NULL; + goto out_put_sb_page; + } } -quiet_out: - ret = size; -out: - if (sb_page) - put_page(sb_page); + + put_page(sb_page); +done: + kfree(sb); + kfree(path); + module_put(THIS_MODULE); + return size; + +out_put_sb_page: + put_page(sb_page); +out_blkdev_put: + if (bdev) + blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); +out_free_sb: kfree(sb); +out_free_path: kfree(path); + path = NULL; +out_module_put: module_put(THIS_MODULE); +out: + pr_info("error %s: %s", path?path:"", err); return ret; - -err_close: - blkdev_put(bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL); -err: - pr_info("error %s: %s", path, err); - goto out; } static int bcache_reboot(struct notifier_block *n, unsigned long code, void *x) diff --git a/drivers/md/bcache/sysfs.c b/drivers/md/bcache/sysfs.c index 54cd1727d20c..9d5cc9fc8d65 100644 --- a/drivers/md/bcache/sysfs.c +++ b/drivers/md/bcache/sysfs.c @@ -27,6 +27,12 @@ static const char * const bch_cache_modes[] = { NULL }; +static const char * const bch_reada_cache_policies[] = { + "all", + "meta-only", + NULL +}; + /* Default is 0 ("auto") */ static const char * const bch_stop_on_failure_modes[] = { "auto", @@ -100,6 +106,7 @@ rw_attribute(congested_write_threshold_us); rw_attribute(sequential_cutoff); rw_attribute(data_csum); rw_attribute(cache_mode); +rw_attribute(readahead_cache_policy); rw_attribute(stop_when_cache_set_failed); rw_attribute(writeback_metadata); rw_attribute(writeback_running); @@ -167,6 +174,11 @@ SHOW(__bch_cached_dev) bch_cache_modes, BDEV_CACHE_MODE(&dc->sb)); + if (attr == &sysfs_readahead_cache_policy) + return bch_snprint_string_list(buf, PAGE_SIZE, + bch_reada_cache_policies, + dc->cache_readahead_policy); + if (attr == &sysfs_stop_when_cache_set_failed) return bch_snprint_string_list(buf, PAGE_SIZE, bch_stop_on_failure_modes, @@ -349,6 +361,15 @@ STORE(__cached_dev) } } + if (attr == &sysfs_readahead_cache_policy) { + v = __sysfs_match_string(bch_reada_cache_policies, -1, buf); + if (v < 0) + return v; + + if ((unsigned int) v != dc->cache_readahead_policy) + dc->cache_readahead_policy = v; + } + if (attr == &sysfs_stop_when_cache_set_failed) { v = __sysfs_match_string(bch_stop_on_failure_modes, -1, buf); if (v < 0) @@ -463,6 +484,7 @@ static struct attribute *bch_cached_dev_files[] = { &sysfs_data_csum, #endif &sysfs_cache_mode, + &sysfs_readahead_cache_policy, &sysfs_stop_when_cache_set_failed, &sysfs_writeback_metadata, &sysfs_writeback_running, diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c index e9ffcea1ca50..58bedd51e11b 100644 --- a/drivers/md/bcache/writeback.c +++ b/drivers/md/bcache/writeback.c @@ -516,15 +516,19 @@ void bcache_dev_sectors_dirty_add(struct cache_set *c, unsigned int inode, uint64_t offset, int nr_sectors) { struct bcache_device *d = c->devices[inode]; - unsigned int stripe_offset, stripe, sectors_dirty; + unsigned int stripe_offset, sectors_dirty; + int stripe; if (!d) return; + stripe = offset_to_stripe(d, offset); + if (stripe < 0) + return; + if (UUID_FLASH_ONLY(&c->uuids[inode])) atomic_long_add(nr_sectors, &c->flash_dev_dirty_sectors); - stripe = offset_to_stripe(d, offset); stripe_offset = offset & (d->stripe_size - 1); while (nr_sectors) { @@ -564,12 +568,12 @@ static bool dirty_pred(struct keybuf *buf, struct bkey *k) static void refill_full_stripes(struct cached_dev *dc) { struct keybuf *buf = &dc->writeback_keys; - unsigned int start_stripe, stripe, next_stripe; + unsigned int start_stripe, next_stripe; + int stripe; bool wrapped = false; stripe = offset_to_stripe(&dc->disk, KEY_OFFSET(&buf->last_scanned)); - - if (stripe >= dc->disk.nr_stripes) + if (stripe < 0) stripe = 0; start_stripe = stripe; diff --git a/drivers/md/bcache/writeback.h b/drivers/md/bcache/writeback.h index 4e4c6810dc3c..c4ff76037227 100644 --- a/drivers/md/bcache/writeback.h +++ b/drivers/md/bcache/writeback.h @@ -33,10 +33,22 @@ static inline uint64_t bcache_dev_sectors_dirty(struct bcache_device *d) return ret; } -static inline unsigned int offset_to_stripe(struct bcache_device *d, +static inline int offset_to_stripe(struct bcache_device *d, uint64_t offset) { do_div(offset, d->stripe_size); + + /* d->nr_stripes is in range [1, INT_MAX] */ + if (unlikely(offset >= d->nr_stripes)) { + pr_err("Invalid stripe %llu (>= nr_stripes %d).\n", + offset, d->nr_stripes); + return -EINVAL; + } + + /* + * Here offset is definitly smaller than INT_MAX, + * return it as int will never overflow. + */ return offset; } @@ -44,7 +56,10 @@ static inline bool bcache_dev_stripe_dirty(struct cached_dev *dc, uint64_t offset, unsigned int nr_sectors) { - unsigned int stripe = offset_to_stripe(&dc->disk, offset); + int stripe = offset_to_stripe(&dc->disk, offset); + + if (stripe < 0) + return false; while (1) { if (atomic_read(dc->disk.stripe_sectors_dirty + stripe)) |