diff options
Diffstat (limited to 'drivers/md')
34 files changed, 916 insertions, 424 deletions
diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c index 00cde40db572..43829d9493f7 100644 --- a/drivers/md/bcache/btree.c +++ b/drivers/md/bcache/btree.c @@ -1741,6 +1741,7 @@ static void bch_btree_gc(struct cache_set *c) do { ret = btree_root(gc_root, c, &op, &writes, &stats); closure_sync(&writes); + cond_resched(); if (ret && ret != -EAGAIN) pr_warn("gc failed!"); @@ -2162,8 +2163,10 @@ int bch_btree_insert_check_key(struct btree *b, struct btree_op *op, rw_lock(true, b, b->level); if (b->key.ptr[0] != btree_ptr || - b->seq != seq + 1) + b->seq != seq + 1) { + op->lock = b->level; goto out; + } } SET_KEY_PTRS(check_key, 1); diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c index 4dd2bb7167f0..a7a03a21d78a 100644 --- a/drivers/md/bcache/super.c +++ b/drivers/md/bcache/super.c @@ -708,6 +708,8 @@ static void bcache_device_link(struct bcache_device *d, struct cache_set *c, WARN(sysfs_create_link(&d->kobj, &c->kobj, "cache") || sysfs_create_link(&c->kobj, &d->kobj, d->name), "Couldn't create device <-> cache set symlinks"); + + clear_bit(BCACHE_DEV_UNLINK_DONE, &d->flags); } static void bcache_device_detach(struct bcache_device *d) @@ -878,8 +880,11 @@ void bch_cached_dev_run(struct cached_dev *dc) buf[SB_LABEL_SIZE] = '\0'; env[2] = kasprintf(GFP_KERNEL, "CACHED_LABEL=%s", buf); - if (atomic_xchg(&dc->running, 1)) + if (atomic_xchg(&dc->running, 1)) { + kfree(env[1]); + kfree(env[2]); return; + } if (!d->c && BDEV_STATE(&dc->sb) != BDEV_STATE_NONE) { @@ -1041,8 +1046,12 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c) */ atomic_set(&dc->count, 1); - if (bch_cached_dev_writeback_start(dc)) + /* Block writeback thread, but spawn it */ + down_write(&dc->writeback_lock); + if (bch_cached_dev_writeback_start(dc)) { + up_write(&dc->writeback_lock); return -ENOMEM; + } if (BDEV_STATE(&dc->sb) == BDEV_STATE_DIRTY) { bch_sectors_dirty_init(dc); @@ -1054,6 +1063,9 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c) bch_cached_dev_run(dc); bcache_device_link(&dc->disk, c, "bdev"); + /* Allow the writeback thread to proceed */ + up_write(&dc->writeback_lock); + pr_info("Caching %s as %s on set %pU", bdevname(dc->bdev, buf), dc->disk.disk->disk_name, dc->disk.c->sb.set_uuid); @@ -1392,6 +1404,9 @@ static void cache_set_flush(struct closure *cl) struct btree *b; unsigned i; + if (!c) + closure_return(cl); + bch_cache_accounting_destroy(&c->accounting); kobject_put(&c->internal); @@ -1836,7 +1851,7 @@ static int cache_alloc(struct cache_sb *sb, struct cache *ca) free = roundup_pow_of_two(ca->sb.nbuckets) >> 10; if (!init_fifo(&ca->free[RESERVE_BTREE], 8, GFP_KERNEL) || - !init_fifo(&ca->free[RESERVE_PRIO], prio_buckets(ca), GFP_KERNEL) || + !init_fifo_exact(&ca->free[RESERVE_PRIO], prio_buckets(ca), GFP_KERNEL) || !init_fifo(&ca->free[RESERVE_MOVINGGC], free, GFP_KERNEL) || !init_fifo(&ca->free[RESERVE_NONE], free, GFP_KERNEL) || !init_fifo(&ca->free_inc, free << 2, GFP_KERNEL) || @@ -1857,11 +1872,12 @@ static int cache_alloc(struct cache_sb *sb, struct cache *ca) return 0; } -static void register_cache(struct cache_sb *sb, struct page *sb_page, +static int register_cache(struct cache_sb *sb, struct page *sb_page, struct block_device *bdev, struct cache *ca) { char name[BDEVNAME_SIZE]; - const char *err = "cannot allocate memory"; + const char *err = NULL; /* must be set for any error case */ + int ret = 0; memcpy(&ca->sb, sb, sizeof(struct cache_sb)); ca->bdev = bdev; @@ -1876,27 +1892,40 @@ static void register_cache(struct cache_sb *sb, struct page *sb_page, if (blk_queue_discard(bdev_get_queue(ca->bdev))) ca->discard = CACHE_DISCARD(&ca->sb); - if (cache_alloc(sb, ca) != 0) + ret = cache_alloc(sb, ca); + if (ret != 0) { + if (ret == -ENOMEM) + err = "cache_alloc(): -ENOMEM"; + else + err = "cache_alloc(): unknown error"; goto err; + } - err = "error creating kobject"; - if (kobject_add(&ca->kobj, &part_to_dev(bdev->bd_part)->kobj, "bcache")) - goto err; + if (kobject_add(&ca->kobj, &part_to_dev(bdev->bd_part)->kobj, "bcache")) { + err = "error calling kobject_add"; + ret = -ENOMEM; + goto out; + } mutex_lock(&bch_register_lock); err = register_cache_set(ca); mutex_unlock(&bch_register_lock); - if (err) - goto err; + if (err) { + ret = -ENODEV; + goto out; + } pr_info("registered cache device %s", bdevname(bdev, name)); + out: kobject_put(&ca->kobj); - return; + err: - pr_notice("error opening %s: %s", bdevname(bdev, name), err); - goto out; + if (err) + pr_notice("error opening %s: %s", bdevname(bdev, name), err); + + return ret; } /* Global interfaces/init */ @@ -1967,6 +1996,8 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr, else err = "device busy"; mutex_unlock(&bch_register_lock); + if (attr == &ksysfs_register_quiet) + goto out; } goto err; } @@ -1992,7 +2023,8 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr, if (!ca) goto err_close; - register_cache(sb, sb_page, bdev, ca); + if (register_cache(sb, sb_page, bdev, ca) != 0) + goto err_close; } out: if (sb_page) @@ -2005,8 +2037,7 @@ out: err_close: blkdev_put(bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL); err: - if (attr != &ksysfs_register_quiet) - pr_info("error opening %s: %s", path, err); + pr_info("error opening %s: %s", path, err); ret = -EINVAL; goto out; } @@ -2100,8 +2131,10 @@ static int __init bcache_init(void) closure_debug_init(); bcache_major = register_blkdev(0, "bcache"); - if (bcache_major < 0) + if (bcache_major < 0) { + unregister_reboot_notifier(&reboot); return bcache_major; + } if (!(bcache_wq = create_workqueue("bcache")) || !(bcache_kobj = kobject_create_and_add("bcache", fs_kobj)) || diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c index f1986bcd1bf0..540256a0df4f 100644 --- a/drivers/md/bcache/writeback.c +++ b/drivers/md/bcache/writeback.c @@ -323,6 +323,10 @@ void bcache_dev_sectors_dirty_add(struct cache_set *c, unsigned inode, static bool dirty_pred(struct keybuf *buf, struct bkey *k) { + struct cached_dev *dc = container_of(buf, struct cached_dev, writeback_keys); + + BUG_ON(KEY_INODE(k) != dc->disk.id); + return KEY_DIRTY(k); } @@ -372,11 +376,24 @@ next: } } +/* + * Returns true if we scanned the entire disk + */ static bool refill_dirty(struct cached_dev *dc) { struct keybuf *buf = &dc->writeback_keys; + struct bkey start = KEY(dc->disk.id, 0, 0); struct bkey end = KEY(dc->disk.id, MAX_KEY_OFFSET, 0); - bool searched_from_start = false; + struct bkey start_pos; + + /* + * make sure keybuf pos is inside the range for this disk - at bringup + * we might not be attached yet so this disk's inode nr isn't + * initialized then + */ + if (bkey_cmp(&buf->last_scanned, &start) < 0 || + bkey_cmp(&buf->last_scanned, &end) > 0) + buf->last_scanned = start; if (dc->partial_stripes_expensive) { refill_full_stripes(dc); @@ -384,14 +401,20 @@ static bool refill_dirty(struct cached_dev *dc) return false; } - if (bkey_cmp(&buf->last_scanned, &end) >= 0) { - buf->last_scanned = KEY(dc->disk.id, 0, 0); - searched_from_start = true; - } - + start_pos = buf->last_scanned; bch_refill_keybuf(dc->disk.c, buf, &end, dirty_pred); - return bkey_cmp(&buf->last_scanned, &end) >= 0 && searched_from_start; + if (bkey_cmp(&buf->last_scanned, &end) < 0) + return false; + + /* + * If we get to the end start scanning again from the beginning, and + * only scan up to where we initially started scanning from: + */ + buf->last_scanned = start; + bch_refill_keybuf(dc->disk.c, buf, &start_pos, dirty_pred); + + return bkey_cmp(&buf->last_scanned, &start_pos) >= 0; } static int bch_writeback_thread(void *arg) diff --git a/drivers/md/bcache/writeback.h b/drivers/md/bcache/writeback.h index 0a9dab187b79..073a042aed24 100644 --- a/drivers/md/bcache/writeback.h +++ b/drivers/md/bcache/writeback.h @@ -63,7 +63,8 @@ static inline bool should_writeback(struct cached_dev *dc, struct bio *bio, static inline void bch_writeback_queue(struct cached_dev *dc) { - wake_up_process(dc->writeback_thread); + if (!IS_ERR_OR_NULL(dc->writeback_thread)) + wake_up_process(dc->writeback_thread); } static inline void bch_writeback_add(struct cached_dev *dc) diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c index da3604e73e8a..cced062cab4b 100644 --- a/drivers/md/bitmap.c +++ b/drivers/md/bitmap.c @@ -564,6 +564,8 @@ static int bitmap_read_sb(struct bitmap *bitmap) if (err) return err; + err = -EINVAL; + sb = kmap_atomic(sb_page); chunksize = le32_to_cpu(sb->chunksize); diff --git a/drivers/md/dm-bufio.c b/drivers/md/dm-bufio.c index afe79719ea32..ecbd3ffcf359 100644 --- a/drivers/md/dm-bufio.c +++ b/drivers/md/dm-bufio.c @@ -532,6 +532,19 @@ static void use_dmio(struct dm_buffer *b, int rw, sector_t block, end_io(&b->bio, r); } +static void inline_endio(struct bio *bio, int error) +{ + bio_end_io_t *end_fn = bio->bi_private; + + /* + * Reset the bio to free any attached resources + * (e.g. bio integrity profiles). + */ + bio_reset(bio); + + end_fn(bio, error); +} + static void use_inline_bio(struct dm_buffer *b, int rw, sector_t block, bio_end_io_t *end_io) { @@ -543,7 +556,12 @@ static void use_inline_bio(struct dm_buffer *b, int rw, sector_t block, b->bio.bi_max_vecs = DM_BUFIO_INLINE_VECS; b->bio.bi_iter.bi_sector = block << b->c->sectors_per_block_bits; b->bio.bi_bdev = b->c->bdev; - b->bio.bi_end_io = end_io; + b->bio.bi_end_io = inline_endio; + /* + * Use of .bi_private isn't a problem here because + * the dm_buffer's inline bio is local to bufio. + */ + b->bio.bi_private = end_io; /* * We assume that if len >= PAGE_SIZE ptr is page-aligned. diff --git a/drivers/md/dm-cache-metadata.c b/drivers/md/dm-cache-metadata.c index 06709257adde..1e64e9c50d85 100644 --- a/drivers/md/dm-cache-metadata.c +++ b/drivers/md/dm-cache-metadata.c @@ -94,6 +94,9 @@ struct cache_disk_superblock { } __packed; struct dm_cache_metadata { + atomic_t ref_count; + struct list_head list; + struct block_device *bdev; struct dm_block_manager *bm; struct dm_space_map *metadata_sm; @@ -669,10 +672,10 @@ static void unpack_value(__le64 value_le, dm_oblock_t *block, unsigned *flags) /*----------------------------------------------------------------*/ -struct dm_cache_metadata *dm_cache_metadata_open(struct block_device *bdev, - sector_t data_block_size, - bool may_format_device, - size_t policy_hint_size) +static struct dm_cache_metadata *metadata_open(struct block_device *bdev, + sector_t data_block_size, + bool may_format_device, + size_t policy_hint_size) { int r; struct dm_cache_metadata *cmd; @@ -680,9 +683,10 @@ struct dm_cache_metadata *dm_cache_metadata_open(struct block_device *bdev, cmd = kzalloc(sizeof(*cmd), GFP_KERNEL); if (!cmd) { DMERR("could not allocate metadata struct"); - return NULL; + return ERR_PTR(-ENOMEM); } + atomic_set(&cmd->ref_count, 1); init_rwsem(&cmd->root_lock); cmd->bdev = bdev; cmd->data_block_size = data_block_size; @@ -705,10 +709,96 @@ struct dm_cache_metadata *dm_cache_metadata_open(struct block_device *bdev, return cmd; } +/* + * We keep a little list of ref counted metadata objects to prevent two + * different target instances creating separate bufio instances. This is + * an issue if a table is reloaded before the suspend. + */ +static DEFINE_MUTEX(table_lock); +static LIST_HEAD(table); + +static struct dm_cache_metadata *lookup(struct block_device *bdev) +{ + struct dm_cache_metadata *cmd; + + list_for_each_entry(cmd, &table, list) + if (cmd->bdev == bdev) { + atomic_inc(&cmd->ref_count); + return cmd; + } + + return NULL; +} + +static struct dm_cache_metadata *lookup_or_open(struct block_device *bdev, + sector_t data_block_size, + bool may_format_device, + size_t policy_hint_size) +{ + struct dm_cache_metadata *cmd, *cmd2; + + mutex_lock(&table_lock); + cmd = lookup(bdev); + mutex_unlock(&table_lock); + + if (cmd) + return cmd; + + cmd = metadata_open(bdev, data_block_size, may_format_device, policy_hint_size); + if (!IS_ERR(cmd)) { + mutex_lock(&table_lock); + cmd2 = lookup(bdev); + if (cmd2) { + mutex_unlock(&table_lock); + __destroy_persistent_data_objects(cmd); + kfree(cmd); + return cmd2; + } + list_add(&cmd->list, &table); + mutex_unlock(&table_lock); + } + + return cmd; +} + +static bool same_params(struct dm_cache_metadata *cmd, sector_t data_block_size) +{ + if (cmd->data_block_size != data_block_size) { + DMERR("data_block_size (%llu) different from that in metadata (%llu)\n", + (unsigned long long) data_block_size, + (unsigned long long) cmd->data_block_size); + return false; + } + + return true; +} + +struct dm_cache_metadata *dm_cache_metadata_open(struct block_device *bdev, + sector_t data_block_size, + bool may_format_device, + size_t policy_hint_size) +{ + struct dm_cache_metadata *cmd = lookup_or_open(bdev, data_block_size, + may_format_device, policy_hint_size); + + if (!IS_ERR(cmd) && !same_params(cmd, data_block_size)) { + dm_cache_metadata_close(cmd); + return ERR_PTR(-EINVAL); + } + + return cmd; +} + void dm_cache_metadata_close(struct dm_cache_metadata *cmd) { - __destroy_persistent_data_objects(cmd); - kfree(cmd); + if (atomic_dec_and_test(&cmd->ref_count)) { + mutex_lock(&table_lock); + list_del(&cmd->list); + mutex_unlock(&table_lock); + + __destroy_persistent_data_objects(cmd); + kfree(cmd); + } } /* diff --git a/drivers/md/dm-cache-policy-cleaner.c b/drivers/md/dm-cache-policy-cleaner.c index b04d1f904d07..2eca9084defe 100644 --- a/drivers/md/dm-cache-policy-cleaner.c +++ b/drivers/md/dm-cache-policy-cleaner.c @@ -434,7 +434,7 @@ static struct dm_cache_policy *wb_create(dm_cblock_t cache_size, static struct dm_cache_policy_type wb_policy_type = { .name = "cleaner", .version = {1, 0, 0}, - .hint_size = 0, + .hint_size = 4, .owner = THIS_MODULE, .create = wb_create }; diff --git a/drivers/md/dm-cache-target.c b/drivers/md/dm-cache-target.c index 7130505c2425..3baed67cf26f 100644 --- a/drivers/md/dm-cache-target.c +++ b/drivers/md/dm-cache-target.c @@ -222,7 +222,13 @@ struct cache { struct list_head need_commit_migrations; sector_t migration_threshold; wait_queue_head_t migration_wait; - atomic_t nr_migrations; + atomic_t nr_allocated_migrations; + + /* + * The number of in flight migrations that are performing + * background io. eg, promotion, writeback. + */ + atomic_t nr_io_migrations; wait_queue_head_t quiescing_wait; atomic_t quiescing; @@ -258,7 +264,6 @@ struct cache { struct dm_deferred_set *all_io_ds; mempool_t *migration_pool; - struct dm_cache_migration *next_migration; struct dm_cache_policy *policy; unsigned policy_nr_args; @@ -349,10 +354,31 @@ static void free_prison_cell(struct cache *cache, struct dm_bio_prison_cell *cel dm_bio_prison_free_cell(cache->prison, cell); } +static struct dm_cache_migration *alloc_migration(struct cache *cache) +{ + struct dm_cache_migration *mg; + + mg = mempool_alloc(cache->migration_pool, GFP_NOWAIT); + if (mg) { + mg->cache = cache; + atomic_inc(&mg->cache->nr_allocated_migrations); + } + + return mg; +} + +static void free_migration(struct dm_cache_migration *mg) +{ + if (atomic_dec_and_test(&mg->cache->nr_allocated_migrations)) + wake_up(&mg->cache->migration_wait); + + mempool_free(mg, mg->cache->migration_pool); +} + static int prealloc_data_structs(struct cache *cache, struct prealloc *p) { if (!p->mg) { - p->mg = mempool_alloc(cache->migration_pool, GFP_NOWAIT); + p->mg = alloc_migration(cache); if (!p->mg) return -ENOMEM; } @@ -381,7 +407,7 @@ static void prealloc_free_structs(struct cache *cache, struct prealloc *p) free_prison_cell(cache, p->cell1); if (p->mg) - mempool_free(p->mg, cache->migration_pool); + free_migration(p->mg); } static struct dm_cache_migration *prealloc_get_migration(struct prealloc *p) @@ -817,24 +843,14 @@ static void remap_to_origin_then_cache(struct cache *cache, struct bio *bio, * Migration covers moving data from the origin device to the cache, or * vice versa. *--------------------------------------------------------------*/ -static void free_migration(struct dm_cache_migration *mg) -{ - mempool_free(mg, mg->cache->migration_pool); -} - -static void inc_nr_migrations(struct cache *cache) +static void inc_io_migrations(struct cache *cache) { - atomic_inc(&cache->nr_migrations); + atomic_inc(&cache->nr_io_migrations); } -static void dec_nr_migrations(struct cache *cache) +static void dec_io_migrations(struct cache *cache) { - atomic_dec(&cache->nr_migrations); - - /* - * Wake the worker in case we're suspending the target. - */ - wake_up(&cache->migration_wait); + atomic_dec(&cache->nr_io_migrations); } static void __cell_defer(struct cache *cache, struct dm_bio_prison_cell *cell, @@ -857,11 +873,10 @@ static void cell_defer(struct cache *cache, struct dm_bio_prison_cell *cell, wake_worker(cache); } -static void cleanup_migration(struct dm_cache_migration *mg) +static void free_io_migration(struct dm_cache_migration *mg) { - struct cache *cache = mg->cache; + dec_io_migrations(mg->cache); free_migration(mg); - dec_nr_migrations(cache); } static void migration_failure(struct dm_cache_migration *mg) @@ -886,7 +901,7 @@ static void migration_failure(struct dm_cache_migration *mg) cell_defer(cache, mg->new_ocell, true); } - cleanup_migration(mg); + free_io_migration(mg); } static void migration_success_pre_commit(struct dm_cache_migration *mg) @@ -897,7 +912,7 @@ static void migration_success_pre_commit(struct dm_cache_migration *mg) if (mg->writeback) { clear_dirty(cache, mg->old_oblock, mg->cblock); cell_defer(cache, mg->old_ocell, false); - cleanup_migration(mg); + free_io_migration(mg); return; } else if (mg->demote) { @@ -907,14 +922,14 @@ static void migration_success_pre_commit(struct dm_cache_migration *mg) mg->old_oblock); if (mg->promote) cell_defer(cache, mg->new_ocell, true); - cleanup_migration(mg); + free_io_migration(mg); return; } } else { if (dm_cache_insert_mapping(cache->cmd, mg->cblock, mg->new_oblock)) { DMWARN_LIMIT("promotion failed; couldn't update on disk metadata"); policy_remove_mapping(cache->policy, mg->new_oblock); - cleanup_migration(mg); + free_io_migration(mg); return; } } @@ -947,18 +962,22 @@ static void migration_success_post_commit(struct dm_cache_migration *mg) } else { if (mg->invalidate) policy_remove_mapping(cache->policy, mg->old_oblock); - cleanup_migration(mg); + free_io_migration(mg); } } else { - clear_dirty(cache, mg->new_oblock, mg->cblock); - if (mg->requeue_holder) + if (mg->requeue_holder) { + clear_dirty(cache, mg->new_oblock, mg->cblock); cell_defer(cache, mg->new_ocell, true); - else { + } else { + /* + * The block was promoted via an overwrite, so it's dirty. + */ + set_dirty(cache, mg->new_oblock, mg->cblock); bio_endio(mg->new_ocell->holder, 0); cell_defer(cache, mg->new_ocell, false); } - cleanup_migration(mg); + free_io_migration(mg); } } @@ -1070,7 +1089,8 @@ static void issue_copy(struct dm_cache_migration *mg) avoid = is_discarded_oblock(cache, mg->new_oblock); - if (!avoid && bio_writes_complete_block(cache, bio)) { + if (writeback_mode(&cache->features) && + !avoid && bio_writes_complete_block(cache, bio)) { issue_overwrite(mg, bio); return; } @@ -1173,7 +1193,7 @@ static void promote(struct cache *cache, struct prealloc *structs, mg->new_ocell = cell; mg->start_jiffies = jiffies; - inc_nr_migrations(cache); + inc_io_migrations(cache); quiesce_migration(mg); } @@ -1196,7 +1216,7 @@ static void writeback(struct cache *cache, struct prealloc *structs, mg->new_ocell = NULL; mg->start_jiffies = jiffies; - inc_nr_migrations(cache); + inc_io_migrations(cache); quiesce_migration(mg); } @@ -1222,7 +1242,7 @@ static void demote_then_promote(struct cache *cache, struct prealloc *structs, mg->new_ocell = new_ocell; mg->start_jiffies = jiffies; - inc_nr_migrations(cache); + inc_io_migrations(cache); quiesce_migration(mg); } @@ -1249,7 +1269,7 @@ static void invalidate(struct cache *cache, struct prealloc *structs, mg->new_ocell = NULL; mg->start_jiffies = jiffies; - inc_nr_migrations(cache); + inc_io_migrations(cache); quiesce_migration(mg); } @@ -1315,7 +1335,7 @@ static void process_discard_bio(struct cache *cache, struct bio *bio) static bool spare_migration_bandwidth(struct cache *cache) { - sector_t current_volume = (atomic_read(&cache->nr_migrations) + 1) * + sector_t current_volume = (atomic_read(&cache->nr_io_migrations) + 1) * cache->sectors_per_block; return current_volume < cache->migration_threshold; } @@ -1665,7 +1685,7 @@ static void stop_quiescing(struct cache *cache) static void wait_for_migrations(struct cache *cache) { - wait_event(cache->migration_wait, !atomic_read(&cache->nr_migrations)); + wait_event(cache->migration_wait, !atomic_read(&cache->nr_allocated_migrations)); } static void stop_worker(struct cache *cache) @@ -1777,9 +1797,6 @@ static void destroy(struct cache *cache) { unsigned i; - if (cache->next_migration) - mempool_free(cache->next_migration, cache->migration_pool); - if (cache->migration_pool) mempool_destroy(cache->migration_pool); @@ -2287,7 +2304,8 @@ static int cache_create(struct cache_args *ca, struct cache **result) INIT_LIST_HEAD(&cache->quiesced_migrations); INIT_LIST_HEAD(&cache->completed_migrations); INIT_LIST_HEAD(&cache->need_commit_migrations); - atomic_set(&cache->nr_migrations, 0); + atomic_set(&cache->nr_allocated_migrations, 0); + atomic_set(&cache->nr_io_migrations, 0); init_waitqueue_head(&cache->migration_wait); init_waitqueue_head(&cache->quiescing_wait); @@ -2346,8 +2364,6 @@ static int cache_create(struct cache_args *ca, struct cache **result) goto bad; } - cache->next_migration = NULL; - cache->need_tick_bio = true; cache->sized = false; cache->invalidate = false; @@ -2549,11 +2565,11 @@ static int __cache_map(struct cache *cache, struct bio *bio, struct dm_bio_priso static int cache_map(struct dm_target *ti, struct bio *bio) { int r; - struct dm_bio_prison_cell *cell; + struct dm_bio_prison_cell *cell = NULL; struct cache *cache = ti->private; r = __cache_map(cache, bio, &cell); - if (r == DM_MAPIO_REMAPPED) { + if (r == DM_MAPIO_REMAPPED && cell) { inc_ds(cache, bio, cell); cell_defer(cache, cell, false); } diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index fc93b9330af4..945c031cf6ee 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -18,9 +18,11 @@ #include <linux/slab.h> #include <linux/crypto.h> #include <linux/workqueue.h> +#include <linux/kthread.h> #include <linux/backing-dev.h> #include <linux/atomic.h> #include <linux/scatterlist.h> +#include <linux/rbtree.h> #include <asm/page.h> #include <asm/unaligned.h> #include <crypto/hash.h> @@ -58,7 +60,8 @@ struct dm_crypt_io { atomic_t io_pending; int error; sector_t sector; - struct dm_crypt_io *base_io; + + struct rb_node rb_node; } CRYPTO_MINALIGN_ATTR; struct dm_crypt_request { @@ -108,7 +111,8 @@ struct iv_tcw_private { * Crypt: maps a linear range of a block device * and encrypts / decrypts at the same time. */ -enum flags { DM_CRYPT_SUSPENDED, DM_CRYPT_KEY_VALID }; +enum flags { DM_CRYPT_SUSPENDED, DM_CRYPT_KEY_VALID, + DM_CRYPT_SAME_CPU, DM_CRYPT_NO_OFFLOAD }; /* * The fields in here must be read only after initialization. @@ -121,14 +125,18 @@ struct crypt_config { * pool for per bio private data, crypto requests and * encryption requeusts/buffer pages */ - mempool_t *io_pool; mempool_t *req_pool; mempool_t *page_pool; struct bio_set *bs; + struct mutex bio_alloc_lock; struct workqueue_struct *io_queue; struct workqueue_struct *crypt_queue; + struct task_struct *write_thread; + wait_queue_head_t write_thread_wait; + struct rb_root write_tree; + char *cipher; char *cipher_string; @@ -172,9 +180,6 @@ struct crypt_config { }; #define MIN_IOS 16 -#define MIN_POOL_PAGES 32 - -static struct kmem_cache *_crypt_io_pool; static void clone_init(struct dm_crypt_io *, struct bio *); static void kcryptd_queue_crypt(struct dm_crypt_io *io); @@ -223,7 +228,7 @@ static struct crypto_ablkcipher *any_tfm(struct crypt_config *cc) * * tcw: Compatible implementation of the block chaining mode used * by the TrueCrypt device encryption system (prior to version 4.1). - * For more info see: http://www.truecrypt.org + * For more info see: https://gitlab.com/cryptsetup/cryptsetup/wikis/TrueCryptOnDiskFormat * It operates on full 512 byte sectors and uses CBC * with an IV derived from initial key and the sector number. * In addition, whitening value is applied on every sector, whitening @@ -705,7 +710,7 @@ static int crypt_iv_tcw_whitening(struct crypt_config *cc, for (i = 0; i < ((1 << SECTOR_SHIFT) / 8); i++) crypto_xor(data + i * 8, buf, 8); out: - memset(buf, 0, sizeof(buf)); + memzero_explicit(buf, sizeof(buf)); return r; } @@ -946,57 +951,70 @@ static int crypt_convert(struct crypt_config *cc, return 0; } +static void crypt_free_buffer_pages(struct crypt_config *cc, struct bio *clone); + /* * Generate a new unfragmented bio with the given size * This should never violate the device limitations - * May return a smaller bio when running out of pages, indicated by - * *out_of_pages set to 1. + * + * This function may be called concurrently. If we allocate from the mempool + * concurrently, there is a possibility of deadlock. For example, if we have + * mempool of 256 pages, two processes, each wanting 256, pages allocate from + * the mempool concurrently, it may deadlock in a situation where both processes + * have allocated 128 pages and the mempool is exhausted. + * + * In order to avoid this scenario we allocate the pages under a mutex. + * + * In order to not degrade performance with excessive locking, we try + * non-blocking allocations without a mutex first but on failure we fallback + * to blocking allocations with a mutex. */ -static struct bio *crypt_alloc_buffer(struct dm_crypt_io *io, unsigned size, - unsigned *out_of_pages) +static struct bio *crypt_alloc_buffer(struct dm_crypt_io *io, unsigned size) { struct crypt_config *cc = io->cc; struct bio *clone; unsigned int nr_iovecs = (size + PAGE_SIZE - 1) >> PAGE_SHIFT; - gfp_t gfp_mask = GFP_NOIO | __GFP_HIGHMEM; - unsigned i, len; + gfp_t gfp_mask = GFP_NOWAIT | __GFP_HIGHMEM; + unsigned i, len, remaining_size; struct page *page; + struct bio_vec *bvec; + +retry: + if (unlikely(gfp_mask & __GFP_WAIT)) + mutex_lock(&cc->bio_alloc_lock); clone = bio_alloc_bioset(GFP_NOIO, nr_iovecs, cc->bs); if (!clone) - return NULL; + goto return_clone; clone_init(io, clone); - *out_of_pages = 0; + + remaining_size = size; for (i = 0; i < nr_iovecs; i++) { page = mempool_alloc(cc->page_pool, gfp_mask); if (!page) { - *out_of_pages = 1; - break; + crypt_free_buffer_pages(cc, clone); + bio_put(clone); + gfp_mask |= __GFP_WAIT; + goto retry; } - /* - * If additional pages cannot be allocated without waiting, - * return a partially-allocated bio. The caller will then try - * to allocate more bios while submitting this partial bio. - */ - gfp_mask = (gfp_mask | __GFP_NOWARN) & ~__GFP_WAIT; + len = (remaining_size > PAGE_SIZE) ? PAGE_SIZE : remaining_size; - len = (size > PAGE_SIZE) ? PAGE_SIZE : size; + bvec = &clone->bi_io_vec[clone->bi_vcnt++]; + bvec->bv_page = page; + bvec->bv_len = len; + bvec->bv_offset = 0; - if (!bio_add_page(clone, page, len, 0)) { - mempool_free(page, cc->page_pool); - break; - } + clone->bi_iter.bi_size += len; - size -= len; + remaining_size -= len; } - if (!clone->bi_iter.bi_size) { - bio_put(clone); - return NULL; - } +return_clone: + if (unlikely(gfp_mask & __GFP_WAIT)) + mutex_unlock(&cc->bio_alloc_lock); return clone; } @@ -1020,7 +1038,6 @@ static void crypt_io_init(struct dm_crypt_io *io, struct crypt_config *cc, io->base_bio = bio; io->sector = sector; io->error = 0; - io->base_io = NULL; io->ctx.req = NULL; atomic_set(&io->io_pending, 0); } @@ -1033,13 +1050,11 @@ static void crypt_inc_pending(struct dm_crypt_io *io) /* * One of the bios was finished. Check for completion of * the whole request and correctly clean up the buffer. - * If base_io is set, wait for the last fragment to complete. */ static void crypt_dec_pending(struct dm_crypt_io *io) { struct crypt_config *cc = io->cc; struct bio *base_bio = io->base_bio; - struct dm_crypt_io *base_io = io->base_io; int error = io->error; if (!atomic_dec_and_test(&io->io_pending)) @@ -1047,16 +1062,8 @@ static void crypt_dec_pending(struct dm_crypt_io *io) if (io->ctx.req) crypt_free_req(cc, io->ctx.req, base_bio); - if (io != dm_per_bio_data(base_bio, cc->per_bio_data_size)) - mempool_free(io, cc->io_pool); - - if (likely(!base_io)) - bio_endio(base_bio, error); - else { - if (error && !base_io->error) - base_io->error = error; - crypt_dec_pending(base_io); - } + + bio_endio(base_bio, error); } /* @@ -1117,15 +1124,15 @@ static void clone_init(struct dm_crypt_io *io, struct bio *clone) static int kcryptd_io_read(struct dm_crypt_io *io, gfp_t gfp) { struct crypt_config *cc = io->cc; - struct bio *base_bio = io->base_bio; struct bio *clone; /* - * The block layer might modify the bvec array, so always - * copy the required bvecs because we need the original - * one in order to decrypt the whole bio data *afterwards*. + * We need the original biovec array in order to decrypt + * the whole bio data *afterwards* -- thanks to immutable + * biovecs we don't need to worry about the block layer + * modifying the biovec array; so leverage bio_clone_fast(). */ - clone = bio_clone_bioset(base_bio, gfp, cc->bs); + clone = bio_clone_fast(io->base_bio, gfp, cc->bs); if (!clone) return 1; @@ -1138,37 +1145,97 @@ static int kcryptd_io_read(struct dm_crypt_io *io, gfp_t gfp) return 0; } +static void kcryptd_io_read_work(struct work_struct *work) +{ + struct dm_crypt_io *io = container_of(work, struct dm_crypt_io, work); + + crypt_inc_pending(io); + if (kcryptd_io_read(io, GFP_NOIO)) + io->error = -ENOMEM; + crypt_dec_pending(io); +} + +static void kcryptd_queue_read(struct dm_crypt_io *io) +{ + struct crypt_config *cc = io->cc; + + INIT_WORK(&io->work, kcryptd_io_read_work); + queue_work(cc->io_queue, &io->work); +} + static void kcryptd_io_write(struct dm_crypt_io *io) { struct bio *clone = io->ctx.bio_out; + generic_make_request(clone); } -static void kcryptd_io(struct work_struct *work) +#define crypt_io_from_node(node) rb_entry((node), struct dm_crypt_io, rb_node) + +static int dmcrypt_write(void *data) { - struct dm_crypt_io *io = container_of(work, struct dm_crypt_io, work); + struct crypt_config *cc = data; + struct dm_crypt_io *io; - if (bio_data_dir(io->base_bio) == READ) { - crypt_inc_pending(io); - if (kcryptd_io_read(io, GFP_NOIO)) - io->error = -ENOMEM; - crypt_dec_pending(io); - } else - kcryptd_io_write(io); -} + while (1) { + struct rb_root write_tree; + struct blk_plug plug; -static void kcryptd_queue_io(struct dm_crypt_io *io) -{ - struct crypt_config *cc = io->cc; + DECLARE_WAITQUEUE(wait, current); - INIT_WORK(&io->work, kcryptd_io); - queue_work(cc->io_queue, &io->work); + spin_lock_irq(&cc->write_thread_wait.lock); +continue_locked: + + if (!RB_EMPTY_ROOT(&cc->write_tree)) + goto pop_from_list; + + __set_current_state(TASK_INTERRUPTIBLE); + __add_wait_queue(&cc->write_thread_wait, &wait); + + spin_unlock_irq(&cc->write_thread_wait.lock); + + if (unlikely(kthread_should_stop())) { + set_task_state(current, TASK_RUNNING); + remove_wait_queue(&cc->write_thread_wait, &wait); + break; + } + + schedule(); + + set_task_state(current, TASK_RUNNING); + spin_lock_irq(&cc->write_thread_wait.lock); + __remove_wait_queue(&cc->write_thread_wait, &wait); + goto continue_locked; + +pop_from_list: + write_tree = cc->write_tree; + cc->write_tree = RB_ROOT; + spin_unlock_irq(&cc->write_thread_wait.lock); + + BUG_ON(rb_parent(write_tree.rb_node)); + + /* + * Note: we cannot walk the tree here with rb_next because + * the structures may be freed when kcryptd_io_write is called. + */ + blk_start_plug(&plug); + do { + io = crypt_io_from_node(rb_first(&write_tree)); + rb_erase(&io->rb_node, &write_tree); + kcryptd_io_write(io); + } while (!RB_EMPTY_ROOT(&write_tree)); + blk_finish_plug(&plug); + } + return 0; } static void kcryptd_crypt_write_io_submit(struct dm_crypt_io *io, int async) { struct bio *clone = io->ctx.bio_out; struct crypt_config *cc = io->cc; + unsigned long flags; + sector_t sector; + struct rb_node **rbp, *parent; if (unlikely(io->error < 0)) { crypt_free_buffer_pages(cc, clone); @@ -1182,20 +1249,34 @@ static void kcryptd_crypt_write_io_submit(struct dm_crypt_io *io, int async) clone->bi_iter.bi_sector = cc->start + io->sector; - if (async) - kcryptd_queue_io(io); - else + if (likely(!async) && test_bit(DM_CRYPT_NO_OFFLOAD, &cc->flags)) { generic_make_request(clone); + return; + } + + spin_lock_irqsave(&cc->write_thread_wait.lock, flags); + rbp = &cc->write_tree.rb_node; + parent = NULL; + sector = io->sector; + while (*rbp) { + parent = *rbp; + if (sector < crypt_io_from_node(parent)->sector) + rbp = &(*rbp)->rb_left; + else + rbp = &(*rbp)->rb_right; + } + rb_link_node(&io->rb_node, parent, rbp); + rb_insert_color(&io->rb_node, &cc->write_tree); + + wake_up_locked(&cc->write_thread_wait); + spin_unlock_irqrestore(&cc->write_thread_wait.lock, flags); } static void kcryptd_crypt_write_convert(struct dm_crypt_io *io) { struct crypt_config *cc = io->cc; struct bio *clone; - struct dm_crypt_io *new_io; int crypt_finished; - unsigned out_of_pages = 0; - unsigned remaining = io->base_bio->bi_iter.bi_size; sector_t sector = io->sector; int r; @@ -1205,80 +1286,30 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io) crypt_inc_pending(io); crypt_convert_init(cc, &io->ctx, NULL, io->base_bio, sector); - /* - * The allocated buffers can be smaller than the whole bio, - * so repeat the whole process until all the data can be handled. - */ - while (remaining) { - clone = crypt_alloc_buffer(io, remaining, &out_of_pages); - if (unlikely(!clone)) { - io->error = -ENOMEM; - break; - } - - io->ctx.bio_out = clone; - io->ctx.iter_out = clone->bi_iter; - - remaining -= clone->bi_iter.bi_size; - sector += bio_sectors(clone); - - crypt_inc_pending(io); - - r = crypt_convert(cc, &io->ctx); - if (r < 0) - io->error = -EIO; - - crypt_finished = atomic_dec_and_test(&io->ctx.cc_pending); - - /* Encryption was already finished, submit io now */ - if (crypt_finished) { - kcryptd_crypt_write_io_submit(io, 0); - - /* - * If there was an error, do not try next fragments. - * For async, error is processed in async handler. - */ - if (unlikely(r < 0)) - break; + clone = crypt_alloc_buffer(io, io->base_bio->bi_iter.bi_size); + if (unlikely(!clone)) { + io->error = -EIO; + goto dec; + } - io->sector = sector; - } + io->ctx.bio_out = clone; + io->ctx.iter_out = clone->bi_iter; - /* - * Out of memory -> run queues - * But don't wait if split was due to the io size restriction - */ - if (unlikely(out_of_pages)) - congestion_wait(BLK_RW_ASYNC, HZ/100); + sector += bio_sectors(clone); - /* - * With async crypto it is unsafe to share the crypto context - * between fragments, so switch to a new dm_crypt_io structure. - */ - if (unlikely(!crypt_finished && remaining)) { - new_io = mempool_alloc(cc->io_pool, GFP_NOIO); - crypt_io_init(new_io, io->cc, io->base_bio, sector); - crypt_inc_pending(new_io); - crypt_convert_init(cc, &new_io->ctx, NULL, - io->base_bio, sector); - new_io->ctx.iter_in = io->ctx.iter_in; - - /* - * Fragments after the first use the base_io - * pending count. - */ - if (!io->base_io) - new_io->base_io = io; - else { - new_io->base_io = io->base_io; - crypt_inc_pending(io->base_io); - crypt_dec_pending(io); - } + crypt_inc_pending(io); + r = crypt_convert(cc, &io->ctx); + if (r) + io->error = -EIO; + crypt_finished = atomic_dec_and_test(&io->ctx.cc_pending); - io = new_io; - } + /* Encryption was already finished, submit io now */ + if (crypt_finished) { + kcryptd_crypt_write_io_submit(io, 0); + io->sector = sector; } +dec: crypt_dec_pending(io); } @@ -1401,7 +1432,7 @@ static int crypt_alloc_tfms(struct crypt_config *cc, char *ciphermode) unsigned i; int err; - cc->tfms = kmalloc(cc->tfms_count * sizeof(struct crypto_ablkcipher *), + cc->tfms = kzalloc(cc->tfms_count * sizeof(struct crypto_ablkcipher *), GFP_KERNEL); if (!cc->tfms) return -ENOMEM; @@ -1481,6 +1512,9 @@ static void crypt_dtr(struct dm_target *ti) if (!cc) return; + if (cc->write_thread) + kthread_stop(cc->write_thread); + if (cc->io_queue) destroy_workqueue(cc->io_queue); if (cc->crypt_queue) @@ -1495,8 +1529,6 @@ static void crypt_dtr(struct dm_target *ti) mempool_destroy(cc->page_pool); if (cc->req_pool) mempool_destroy(cc->req_pool); - if (cc->io_pool) - mempool_destroy(cc->io_pool); if (cc->iv_gen_ops && cc->iv_gen_ops->dtr) cc->iv_gen_ops->dtr(cc); @@ -1688,7 +1720,7 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) char dummy; static struct dm_arg _args[] = { - {0, 1, "Invalid number of feature args"}, + {0, 3, "Invalid number of feature args"}, }; if (argc < 5) { @@ -1710,13 +1742,6 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) if (ret < 0) goto bad; - ret = -ENOMEM; - cc->io_pool = mempool_create_slab_pool(MIN_IOS, _crypt_io_pool); - if (!cc->io_pool) { - ti->error = "Cannot allocate crypt io mempool"; - goto bad; - } - cc->dmreq_start = sizeof(struct ablkcipher_request); cc->dmreq_start += crypto_ablkcipher_reqsize(any_tfm(cc)); cc->dmreq_start = ALIGN(cc->dmreq_start, __alignof__(struct dm_crypt_request)); @@ -1734,6 +1759,7 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) iv_size_padding = crypto_ablkcipher_alignmask(any_tfm(cc)); } + ret = -ENOMEM; cc->req_pool = mempool_create_kmalloc_pool(MIN_IOS, cc->dmreq_start + sizeof(struct dm_crypt_request) + iv_size_padding + cc->iv_size); if (!cc->req_pool) { @@ -1746,7 +1772,7 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) sizeof(struct dm_crypt_request) + iv_size_padding + cc->iv_size, ARCH_KMALLOC_MINALIGN); - cc->page_pool = mempool_create_page_pool(MIN_POOL_PAGES, 0); + cc->page_pool = mempool_create_page_pool(BIO_MAX_PAGES, 0); if (!cc->page_pool) { ti->error = "Cannot allocate page mempool"; goto bad; @@ -1758,6 +1784,8 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) goto bad; } + mutex_init(&cc->bio_alloc_lock); + ret = -EINVAL; if (sscanf(argv[2], "%llu%c", &tmpll, &dummy) != 1) { ti->error = "Invalid iv_offset sector"; @@ -1788,15 +1816,27 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) if (ret) goto bad; - opt_string = dm_shift_arg(&as); + ret = -EINVAL; + while (opt_params--) { + opt_string = dm_shift_arg(&as); + if (!opt_string) { + ti->error = "Not enough feature arguments"; + goto bad; + } - if (opt_params == 1 && opt_string && - !strcasecmp(opt_string, "allow_discards")) - ti->num_discard_bios = 1; - else if (opt_params) { - ret = -EINVAL; - ti->error = "Invalid feature arguments"; - goto bad; + if (!strcasecmp(opt_string, "allow_discards")) + ti->num_discard_bios = 1; + + else if (!strcasecmp(opt_string, "same_cpu_crypt")) + set_bit(DM_CRYPT_SAME_CPU, &cc->flags); + + else if (!strcasecmp(opt_string, "submit_from_crypt_cpus")) + set_bit(DM_CRYPT_NO_OFFLOAD, &cc->flags); + + else { + ti->error = "Invalid feature arguments"; + goto bad; + } } } @@ -1807,13 +1847,28 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) goto bad; } - cc->crypt_queue = alloc_workqueue("kcryptd", - WQ_CPU_INTENSIVE | WQ_MEM_RECLAIM, 1); + if (test_bit(DM_CRYPT_SAME_CPU, &cc->flags)) + cc->crypt_queue = alloc_workqueue("kcryptd", WQ_CPU_INTENSIVE | WQ_MEM_RECLAIM, 1); + else + cc->crypt_queue = alloc_workqueue("kcryptd", WQ_CPU_INTENSIVE | WQ_MEM_RECLAIM | WQ_UNBOUND, + num_online_cpus()); if (!cc->crypt_queue) { ti->error = "Couldn't create kcryptd queue"; goto bad; } + init_waitqueue_head(&cc->write_thread_wait); + cc->write_tree = RB_ROOT; + + cc->write_thread = kthread_create(dmcrypt_write, cc, "dmcrypt_write"); + if (IS_ERR(cc->write_thread)) { + ret = PTR_ERR(cc->write_thread); + cc->write_thread = NULL; + ti->error = "Couldn't spawn write thread"; + goto bad; + } + wake_up_process(cc->write_thread); + ti->num_flush_bios = 1; ti->discard_zeroes_data_unsupported = true; @@ -1842,13 +1897,20 @@ static int crypt_map(struct dm_target *ti, struct bio *bio) return DM_MAPIO_REMAPPED; } + /* + * Check if bio is too large, split as needed. + */ + if (unlikely(bio->bi_iter.bi_size > (BIO_MAX_PAGES << PAGE_SHIFT)) && + bio_data_dir(bio) == WRITE) + dm_accept_partial_bio(bio, ((BIO_MAX_PAGES << PAGE_SHIFT) >> SECTOR_SHIFT)); + io = dm_per_bio_data(bio, cc->per_bio_data_size); crypt_io_init(io, cc, bio, dm_target_offset(ti, bio->bi_iter.bi_sector)); io->ctx.req = (struct ablkcipher_request *)(io + 1); if (bio_data_dir(io->base_bio) == READ) { if (kcryptd_io_read(io, GFP_NOWAIT)) - kcryptd_queue_io(io); + kcryptd_queue_read(io); } else kcryptd_queue_crypt(io); @@ -1860,6 +1922,7 @@ static void crypt_status(struct dm_target *ti, status_type_t type, { struct crypt_config *cc = ti->private; unsigned i, sz = 0; + int num_feature_args = 0; switch (type) { case STATUSTYPE_INFO: @@ -1878,8 +1941,18 @@ static void crypt_status(struct dm_target *ti, status_type_t type, DMEMIT(" %llu %s %llu", (unsigned long long)cc->iv_offset, cc->dev->name, (unsigned long long)cc->start); - if (ti->num_discard_bios) - DMEMIT(" 1 allow_discards"); + num_feature_args += !!ti->num_discard_bios; + num_feature_args += test_bit(DM_CRYPT_SAME_CPU, &cc->flags); + num_feature_args += test_bit(DM_CRYPT_NO_OFFLOAD, &cc->flags); + if (num_feature_args) { + DMEMIT(" %d", num_feature_args); + if (ti->num_discard_bios) + DMEMIT(" allow_discards"); + if (test_bit(DM_CRYPT_SAME_CPU, &cc->flags)) + DMEMIT(" same_cpu_crypt"); + if (test_bit(DM_CRYPT_NO_OFFLOAD, &cc->flags)) + DMEMIT(" submit_from_crypt_cpus"); + } break; } @@ -1976,7 +2049,7 @@ static int crypt_iterate_devices(struct dm_target *ti, static struct target_type crypt_target = { .name = "crypt", - .version = {1, 13, 0}, + .version = {1, 14, 0}, .module = THIS_MODULE, .ctr = crypt_ctr, .dtr = crypt_dtr, @@ -1994,15 +2067,9 @@ static int __init dm_crypt_init(void) { int r; - _crypt_io_pool = KMEM_CACHE(dm_crypt_io, 0); - if (!_crypt_io_pool) - return -ENOMEM; - r = dm_register_target(&crypt_target); - if (r < 0) { + if (r < 0) DMERR("register failed %d", r); - kmem_cache_destroy(_crypt_io_pool); - } return r; } @@ -2010,7 +2077,6 @@ static int __init dm_crypt_init(void) static void __exit dm_crypt_exit(void) { dm_unregister_target(&crypt_target); - kmem_cache_destroy(_crypt_io_pool); } module_init(dm_crypt_init); diff --git a/drivers/md/dm-exception-store.h b/drivers/md/dm-exception-store.h index 0b2536247cf5..84e27708ad97 100644 --- a/drivers/md/dm-exception-store.h +++ b/drivers/md/dm-exception-store.h @@ -70,7 +70,7 @@ struct dm_exception_store_type { * Update the metadata with this exception. */ void (*commit_exception) (struct dm_exception_store *store, - struct dm_exception *e, + struct dm_exception *e, int valid, void (*callback) (void *, int success), void *callback_context); diff --git a/drivers/md/dm-flakey.c b/drivers/md/dm-flakey.c index b257e46876d3..768c9fdf0df8 100644 --- a/drivers/md/dm-flakey.c +++ b/drivers/md/dm-flakey.c @@ -287,10 +287,14 @@ static int flakey_map(struct dm_target *ti, struct bio *bio) pb->bio_submitted = true; /* - * Map reads as normal. + * Error reads if neither corrupt_bio_byte or drop_writes are set. + * Otherwise, flakey_end_io() will decide if the reads should be modified. */ - if (bio_data_dir(bio) == READ) + if (bio_data_dir(bio) == READ) { + if (!fc->corrupt_bio_byte && !test_bit(DROP_WRITES, &fc->flags)) + return -EIO; goto map_bio; + } /* * Drop writes? @@ -326,14 +330,22 @@ static int flakey_end_io(struct dm_target *ti, struct bio *bio, int error) struct flakey_c *fc = ti->private; struct per_bio_data *pb = dm_per_bio_data(bio, sizeof(struct per_bio_data)); - /* - * Corrupt successful READs while in down state. - * If flags were specified, only corrupt those that match. - */ - if (fc->corrupt_bio_byte && !error && pb->bio_submitted && - (bio_data_dir(bio) == READ) && (fc->corrupt_bio_rw == READ) && - all_corrupt_bio_flags_match(bio, fc)) - corrupt_bio_data(bio, fc); + if (!error && pb->bio_submitted && (bio_data_dir(bio) == READ)) { + if (fc->corrupt_bio_byte && (fc->corrupt_bio_rw == READ) && + all_corrupt_bio_flags_match(bio, fc)) { + /* + * Corrupt successful matching READs while in down state. + */ + corrupt_bio_data(bio, fc); + + } else if (!test_bit(DROP_WRITES, &fc->flags)) { + /* + * Error read during the down_interval if drop_writes + * wasn't configured. + */ + return -EIO; + } + } return error; } diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c index c09359db3a90..74adcd2c967e 100644 --- a/drivers/md/dm-io.c +++ b/drivers/md/dm-io.c @@ -289,6 +289,19 @@ static void do_region(int rw, unsigned region, struct dm_io_region *where, struct request_queue *q = bdev_get_queue(where->bdev); unsigned short logical_block_size = queue_logical_block_size(q); sector_t num_sectors; + unsigned int uninitialized_var(special_cmd_max_sectors); + + /* + * Reject unsupported discard and write same requests. + */ + if (rw & REQ_DISCARD) + special_cmd_max_sectors = q->limits.max_discard_sectors; + else if (rw & REQ_WRITE_SAME) + special_cmd_max_sectors = q->limits.max_write_same_sectors; + if ((rw & (REQ_DISCARD | REQ_WRITE_SAME)) && special_cmd_max_sectors == 0) { + dec_count(io, region, -EOPNOTSUPP); + return; + } /* * where->count may be zero if rw holds a flush and we need to @@ -311,7 +324,7 @@ static void do_region(int rw, unsigned region, struct dm_io_region *where, store_io_and_region_in_bio(bio, io, region); if (rw & REQ_DISCARD) { - num_sectors = min_t(sector_t, q->limits.max_discard_sectors, remaining); + num_sectors = min_t(sector_t, special_cmd_max_sectors, remaining); bio->bi_iter.bi_size = num_sectors << SECTOR_SHIFT; remaining -= num_sectors; } else if (rw & REQ_WRITE_SAME) { @@ -320,7 +333,7 @@ static void do_region(int rw, unsigned region, struct dm_io_region *where, */ dp->get_page(dp, &page, &len, &offset); bio_add_page(bio, page, logical_block_size, offset); - num_sectors = min_t(sector_t, q->limits.max_write_same_sectors, remaining); + num_sectors = min_t(sector_t, special_cmd_max_sectors, remaining); bio->bi_iter.bi_size = num_sectors << SECTOR_SHIFT; offset = 0; diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c index 07c0fa0fa284..e5d97da5f41e 100644 --- a/drivers/md/dm-raid.c +++ b/drivers/md/dm-raid.c @@ -327,8 +327,7 @@ static int validate_region_size(struct raid_set *rs, unsigned long region_size) */ if (min_region_size > (1 << 13)) { /* If not a power of 2, make it the next power of 2 */ - if (min_region_size & (min_region_size - 1)) - region_size = 1 << fls(region_size); + region_size = roundup_pow_of_two(min_region_size); DMINFO("Choosing default region size of %lu sectors", region_size); } else { diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c index 7dfdb5c746d6..089d62751f7f 100644 --- a/drivers/md/dm-raid1.c +++ b/drivers/md/dm-raid1.c @@ -604,6 +604,15 @@ static void write_callback(unsigned long error, void *context) return; } + /* + * If the bio is discard, return an error, but do not + * degrade the array. + */ + if (bio->bi_rw & REQ_DISCARD) { + bio_endio(bio, -EOPNOTSUPP); + return; + } + for (i = 0; i < ms->nr_mirrors; i++) if (test_bit(i, &error)) fail_mirror(ms->mirror + i, DM_RAID1_WRITE_ERROR); diff --git a/drivers/md/dm-snap-persistent.c b/drivers/md/dm-snap-persistent.c index d6e88178d22c..d3272acc0f0e 100644 --- a/drivers/md/dm-snap-persistent.c +++ b/drivers/md/dm-snap-persistent.c @@ -700,7 +700,7 @@ static int persistent_prepare_exception(struct dm_exception_store *store, } static void persistent_commit_exception(struct dm_exception_store *store, - struct dm_exception *e, + struct dm_exception *e, int valid, void (*callback) (void *, int success), void *callback_context) { @@ -709,6 +709,9 @@ static void persistent_commit_exception(struct dm_exception_store *store, struct core_exception ce; struct commit_callback *cb; + if (!valid) + ps->valid = 0; + ce.old_chunk = e->old_chunk; ce.new_chunk = e->new_chunk; write_exception(ps, ps->current_committed++, &ce); diff --git a/drivers/md/dm-snap-transient.c b/drivers/md/dm-snap-transient.c index 1ce9a2586e41..31439d53cf7e 100644 --- a/drivers/md/dm-snap-transient.c +++ b/drivers/md/dm-snap-transient.c @@ -52,12 +52,12 @@ static int transient_prepare_exception(struct dm_exception_store *store, } static void transient_commit_exception(struct dm_exception_store *store, - struct dm_exception *e, + struct dm_exception *e, int valid, void (*callback) (void *, int success), void *callback_context) { /* Just succeed */ - callback(callback_context, 1); + callback(callback_context, valid); } static void transient_usage(struct dm_exception_store *store, diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c index 864b03f47727..b91b39caf7a9 100644 --- a/drivers/md/dm-snap.c +++ b/drivers/md/dm-snap.c @@ -20,6 +20,8 @@ #include <linux/log2.h> #include <linux/dm-kcopyd.h> +#include "dm.h" + #include "dm-exception-store.h" #define DM_MSG_PREFIX "snapshots" @@ -291,12 +293,23 @@ struct origin { }; /* + * This structure is allocated for each origin target + */ +struct dm_origin { + struct dm_dev *dev; + struct dm_target *ti; + unsigned split_boundary; + struct list_head hash_list; +}; + +/* * Size of the hash table for origin volumes. If we make this * the size of the minors list then it should be nearly perfect */ #define ORIGIN_HASH_SIZE 256 #define ORIGIN_MASK 0xFF static struct list_head *_origins; +static struct list_head *_dm_origins; static struct rw_semaphore _origins_lock; static DECLARE_WAIT_QUEUE_HEAD(_pending_exceptions_done); @@ -310,12 +323,22 @@ static int init_origin_hash(void) _origins = kmalloc(ORIGIN_HASH_SIZE * sizeof(struct list_head), GFP_KERNEL); if (!_origins) { - DMERR("unable to allocate memory"); + DMERR("unable to allocate memory for _origins"); return -ENOMEM; } - for (i = 0; i < ORIGIN_HASH_SIZE; i++) INIT_LIST_HEAD(_origins + i); + + _dm_origins = kmalloc(ORIGIN_HASH_SIZE * sizeof(struct list_head), + GFP_KERNEL); + if (!_dm_origins) { + DMERR("unable to allocate memory for _dm_origins"); + kfree(_origins); + return -ENOMEM; + } + for (i = 0; i < ORIGIN_HASH_SIZE; i++) + INIT_LIST_HEAD(_dm_origins + i); + init_rwsem(&_origins_lock); return 0; @@ -324,6 +347,7 @@ static int init_origin_hash(void) static void exit_origin_hash(void) { kfree(_origins); + kfree(_dm_origins); } static unsigned origin_hash(struct block_device *bdev) @@ -350,6 +374,30 @@ static void __insert_origin(struct origin *o) list_add_tail(&o->hash_list, sl); } +static struct dm_origin *__lookup_dm_origin(struct block_device *origin) +{ + struct list_head *ol; + struct dm_origin *o; + + ol = &_dm_origins[origin_hash(origin)]; + list_for_each_entry (o, ol, hash_list) + if (bdev_equal(o->dev->bdev, origin)) + return o; + + return NULL; +} + +static void __insert_dm_origin(struct dm_origin *o) +{ + struct list_head *sl = &_dm_origins[origin_hash(o->dev->bdev)]; + list_add_tail(&o->hash_list, sl); +} + +static void __remove_dm_origin(struct dm_origin *o) +{ + list_del(&o->hash_list); +} + /* * _origins_lock must be held when calling this function. * Returns number of snapshots registered using the supplied cow device, plus: @@ -1380,8 +1428,9 @@ static void __invalidate_snapshot(struct dm_snapshot *s, int err) dm_table_event(s->ti->table); } -static void pending_complete(struct dm_snap_pending_exception *pe, int success) +static void pending_complete(void *context, int success) { + struct dm_snap_pending_exception *pe = context; struct dm_exception *e; struct dm_snapshot *s = pe->snap; struct bio *origin_bios = NULL; @@ -1432,8 +1481,6 @@ out: full_bio->bi_private = pe->full_bio_private; atomic_inc(&full_bio->bi_remaining); } - free_pending_exception(pe); - increment_pending_exceptions_done_count(); up_write(&s->lock); @@ -1450,26 +1497,17 @@ out: } retry_origin_bios(s, origin_bios); -} -static void commit_callback(void *context, int success) -{ - struct dm_snap_pending_exception *pe = context; - - pending_complete(pe, success); + free_pending_exception(pe); } static void complete_exception(struct dm_snap_pending_exception *pe) { struct dm_snapshot *s = pe->snap; - if (unlikely(pe->copy_error)) - pending_complete(pe, 0); - - else - /* Update the metadata if we are persistent */ - s->store->type->commit_exception(s->store, &pe->e, - commit_callback, pe); + /* Update the metadata if we are persistent */ + s->store->type->commit_exception(s->store, &pe->e, !pe->copy_error, + pending_complete, pe); } /* @@ -1840,9 +1878,40 @@ static int snapshot_preresume(struct dm_target *ti) static void snapshot_resume(struct dm_target *ti) { struct dm_snapshot *s = ti->private; - struct dm_snapshot *snap_src = NULL, *snap_dest = NULL; + struct dm_snapshot *snap_src = NULL, *snap_dest = NULL, *snap_merging = NULL; + struct dm_origin *o; + struct mapped_device *origin_md = NULL; + bool must_restart_merging = false; down_read(&_origins_lock); + + o = __lookup_dm_origin(s->origin->bdev); + if (o) + origin_md = dm_table_get_md(o->ti->table); + if (!origin_md) { + (void) __find_snapshots_sharing_cow(s, NULL, NULL, &snap_merging); + if (snap_merging) + origin_md = dm_table_get_md(snap_merging->ti->table); + } + if (origin_md == dm_table_get_md(ti->table)) + origin_md = NULL; + if (origin_md) { + if (dm_hold(origin_md)) + origin_md = NULL; + } + + up_read(&_origins_lock); + + if (origin_md) { + dm_internal_suspend(origin_md); + if (snap_merging && test_bit(RUNNING_MERGE, &snap_merging->state_bits)) { + must_restart_merging = true; + stop_merge(snap_merging); + } + } + + down_read(&_origins_lock); + (void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL); if (snap_src && snap_dest) { down_write(&snap_src->lock); @@ -1851,8 +1920,16 @@ static void snapshot_resume(struct dm_target *ti) up_write(&snap_dest->lock); up_write(&snap_src->lock); } + up_read(&_origins_lock); + if (origin_md) { + if (must_restart_merging) + start_merge(snap_merging); + dm_internal_resume(origin_md); + dm_put(origin_md); + } + /* Now we have correct chunk size, reregister */ reregister_snapshot(s); @@ -2133,11 +2210,6 @@ static int origin_write_extent(struct dm_snapshot *merging_snap, * Origin: maps a linear range of a device, with hooks for snapshotting. */ -struct dm_origin { - struct dm_dev *dev; - unsigned split_boundary; -}; - /* * Construct an origin mapping: <dev_path> * The context for an origin is merely a 'struct dm_dev *' @@ -2166,6 +2238,7 @@ static int origin_ctr(struct dm_target *ti, unsigned int argc, char **argv) goto bad_open; } + o->ti = ti; ti->private = o; ti->num_flush_bios = 1; @@ -2180,6 +2253,7 @@ bad_alloc: static void origin_dtr(struct dm_target *ti) { struct dm_origin *o = ti->private; + dm_put_device(ti, o->dev); kfree(o); } @@ -2216,6 +2290,19 @@ static void origin_resume(struct dm_target *ti) struct dm_origin *o = ti->private; o->split_boundary = get_origin_minimum_chunksize(o->dev->bdev); + + down_write(&_origins_lock); + __insert_dm_origin(o); + up_write(&_origins_lock); +} + +static void origin_postsuspend(struct dm_target *ti) +{ + struct dm_origin *o = ti->private; + + down_write(&_origins_lock); + __remove_dm_origin(o); + up_write(&_origins_lock); } static void origin_status(struct dm_target *ti, status_type_t type, @@ -2258,12 +2345,13 @@ static int origin_iterate_devices(struct dm_target *ti, static struct target_type origin_target = { .name = "snapshot-origin", - .version = {1, 8, 1}, + .version = {1, 9, 0}, .module = THIS_MODULE, .ctr = origin_ctr, .dtr = origin_dtr, .map = origin_map, .resume = origin_resume, + .postsuspend = origin_postsuspend, .status = origin_status, .merge = origin_merge, .iterate_devices = origin_iterate_devices, @@ -2271,7 +2359,7 @@ static struct target_type origin_target = { static struct target_type snapshot_target = { .name = "snapshot", - .version = {1, 12, 0}, + .version = {1, 13, 0}, .module = THIS_MODULE, .ctr = snapshot_ctr, .dtr = snapshot_dtr, @@ -2285,7 +2373,7 @@ static struct target_type snapshot_target = { static struct target_type merge_target = { .name = dm_snapshot_merge_target_name, - .version = {1, 2, 0}, + .version = {1, 3, 0}, .module = THIS_MODULE, .ctr = snapshot_ctr, .dtr = snapshot_dtr, diff --git a/drivers/md/dm-stats.c b/drivers/md/dm-stats.c index 87f86c77b094..6f2a4ce413f7 100644 --- a/drivers/md/dm-stats.c +++ b/drivers/md/dm-stats.c @@ -795,6 +795,8 @@ static int message_stats_create(struct mapped_device *md, return -EINVAL; if (sscanf(argv[2], "/%u%c", &divisor, &dummy) == 1) { + if (!divisor) + return -EINVAL; step = end - start; if (do_div(step, divisor)) step++; diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c index b2bd1ebf4562..fc93bd80270f 100644 --- a/drivers/md/dm-table.c +++ b/drivers/md/dm-table.c @@ -691,37 +691,32 @@ int dm_table_add_target(struct dm_table *t, const char *type, tgt->type = dm_get_target_type(type); if (!tgt->type) { - DMERR("%s: %s: unknown target type", dm_device_name(t->md), - type); + DMERR("%s: %s: unknown target type", dm_device_name(t->md), type); return -EINVAL; } if (dm_target_needs_singleton(tgt->type)) { if (t->num_targets) { - DMERR("%s: target type %s must appear alone in table", - dm_device_name(t->md), type); - return -EINVAL; + tgt->error = "singleton target type must appear alone in table"; + goto bad; } t->singleton = 1; } if (dm_target_always_writeable(tgt->type) && !(t->mode & FMODE_WRITE)) { - DMERR("%s: target type %s may not be included in read-only tables", - dm_device_name(t->md), type); - return -EINVAL; + tgt->error = "target type may not be included in a read-only table"; + goto bad; } if (t->immutable_target_type) { if (t->immutable_target_type != tgt->type) { - DMERR("%s: immutable target type %s cannot be mixed with other target types", - dm_device_name(t->md), t->immutable_target_type->name); - return -EINVAL; + tgt->error = "immutable target type cannot be mixed with other target types"; + goto bad; } } else if (dm_target_is_immutable(tgt->type)) { if (t->num_targets) { - DMERR("%s: immutable target type %s cannot be mixed with other target types", - dm_device_name(t->md), tgt->type->name); - return -EINVAL; + tgt->error = "immutable target type cannot be mixed with other target types"; + goto bad; } t->immutable_target_type = tgt->type; } @@ -736,7 +731,6 @@ int dm_table_add_target(struct dm_table *t, const char *type, */ if (!adjoin(t, tgt)) { tgt->error = "Gap in table"; - r = -EINVAL; goto bad; } diff --git a/drivers/md/dm-thin-metadata.c b/drivers/md/dm-thin-metadata.c index e9d33ad59df5..3412b86e79fd 100644 --- a/drivers/md/dm-thin-metadata.c +++ b/drivers/md/dm-thin-metadata.c @@ -1295,8 +1295,8 @@ static int __release_metadata_snap(struct dm_pool_metadata *pmd) return r; disk_super = dm_block_data(copy); - dm_sm_dec_block(pmd->metadata_sm, le64_to_cpu(disk_super->data_mapping_root)); - dm_sm_dec_block(pmd->metadata_sm, le64_to_cpu(disk_super->device_details_root)); + dm_btree_del(&pmd->info, le64_to_cpu(disk_super->data_mapping_root)); + dm_btree_del(&pmd->details_info, le64_to_cpu(disk_super->device_details_root)); dm_sm_dec_block(pmd->metadata_sm, held_root); return dm_tm_unlock(pmd->tm, copy); diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c index 0f86d802b533..9dfe2bb713c5 100644 --- a/drivers/md/dm-thin.c +++ b/drivers/md/dm-thin.c @@ -990,6 +990,24 @@ static void schedule_external_copy(struct thin_c *tc, dm_block_t virt_block, schedule_zero(tc, virt_block, data_dest, cell, bio); } +static void set_pool_mode(struct pool *pool, enum pool_mode new_mode); + +static void check_for_space(struct pool *pool) +{ + int r; + dm_block_t nr_free; + + if (get_pool_mode(pool) != PM_OUT_OF_DATA_SPACE) + return; + + r = dm_pool_get_free_block_count(pool->pmd, &nr_free); + if (r) + return; + + if (nr_free) + set_pool_mode(pool, PM_WRITE); +} + /* * A non-zero return indicates read_only or fail_io mode. * Many callers don't care about the return value. @@ -1004,6 +1022,8 @@ static int commit(struct pool *pool) r = dm_pool_commit_metadata(pool->pmd); if (r) metadata_operation_failed(pool, "dm_pool_commit_metadata", r); + else + check_for_space(pool); return r; } @@ -1022,8 +1042,6 @@ static void check_low_water_mark(struct pool *pool, dm_block_t free_blocks) } } -static void set_pool_mode(struct pool *pool, enum pool_mode new_mode); - static int alloc_data_block(struct thin_c *tc, dm_block_t *result) { int r; @@ -1824,7 +1842,7 @@ static void set_pool_mode(struct pool *pool, enum pool_mode new_mode) pool->process_bio = process_bio_read_only; pool->process_discard = process_discard; pool->process_prepared_mapping = process_prepared_mapping; - pool->process_prepared_discard = process_prepared_discard_passdown; + pool->process_prepared_discard = process_prepared_discard; if (!pool->pf.error_if_no_space && no_space_timeout) queue_delayed_work(pool->wq, &pool->no_space_timeout, no_space_timeout); @@ -1985,17 +2003,6 @@ static int thin_bio_map(struct dm_target *ti, struct bio *bio) return DM_MAPIO_REMAPPED; case -ENODATA: - if (get_pool_mode(tc->pool) == PM_READ_ONLY) { - /* - * This block isn't provisioned, and we have no way - * of doing so. - */ - handle_unserviceable_bio(tc->pool, bio); - cell_defer_no_holder_no_free(tc, &cell1); - return DM_MAPIO_SUBMITTED; - } - /* fall through */ - case -EWOULDBLOCK: /* * In future, the failed dm_thin_find_block above could @@ -2589,7 +2596,7 @@ static int pool_ctr(struct dm_target *ti, unsigned argc, char **argv) metadata_low_callback, pool); if (r) - goto out_free_pt; + goto out_flags_changed; pt->callbacks.congested_fn = pool_is_congested; dm_table_add_target_callbacks(ti->table, &pt->callbacks); @@ -2783,8 +2790,8 @@ static void pool_postsuspend(struct dm_target *ti) struct pool_c *pt = ti->private; struct pool *pool = pt->pool; - cancel_delayed_work(&pool->waker); - cancel_delayed_work(&pool->no_space_timeout); + cancel_delayed_work_sync(&pool->waker); + cancel_delayed_work_sync(&pool->no_space_timeout); flush_workqueue(pool->wq); (void) commit(pool); } @@ -2960,6 +2967,12 @@ static int pool_message(struct dm_target *ti, unsigned argc, char **argv) struct pool_c *pt = ti->private; struct pool *pool = pt->pool; + if (get_pool_mode(pool) >= PM_READ_ONLY) { + DMERR("%s: unable to service pool target messages in READ_ONLY or FAIL mode", + dm_device_name(pool->pool_md)); + return -EINVAL; + } + if (!strcasecmp(argv[0], "create_thin")) r = process_create_thin_mesg(argc, argv, pool); @@ -3248,14 +3261,14 @@ static void thin_dtr(struct dm_target *ti) struct thin_c *tc = ti->private; unsigned long flags; - thin_put(tc); - wait_for_completion(&tc->can_destroy); - spin_lock_irqsave(&tc->pool->lock, flags); list_del_rcu(&tc->list); spin_unlock_irqrestore(&tc->pool->lock, flags); synchronize_rcu(); + thin_put(tc); + wait_for_completion(&tc->can_destroy); + mutex_lock(&dm_thin_pool_table.mutex); __pool_dec(tc->pool); diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 58f3927fd7cc..042114fbd200 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -899,7 +899,7 @@ static void disable_write_same(struct mapped_device *md) static void clone_endio(struct bio *bio, int error) { - int r = 0; + int r = error; struct dm_target_io *tio = container_of(bio, struct dm_target_io, clone); struct dm_io *io = tio->io; struct mapped_device *md = tio->io->md; @@ -2462,7 +2462,7 @@ int dm_setup_md_queue(struct mapped_device *md) return 0; } -static struct mapped_device *dm_find_md(dev_t dev) +struct mapped_device *dm_get_md(dev_t dev) { struct mapped_device *md; unsigned minor = MINOR(dev); @@ -2473,12 +2473,15 @@ static struct mapped_device *dm_find_md(dev_t dev) spin_lock(&_minor_lock); md = idr_find(&_minor_idr, minor); - if (md && (md == MINOR_ALLOCED || - (MINOR(disk_devt(dm_disk(md))) != minor) || - dm_deleting_md(md) || - test_bit(DMF_FREEING, &md->flags))) { - md = NULL; - goto out; + if (md) { + if ((md == MINOR_ALLOCED || + (MINOR(disk_devt(dm_disk(md))) != minor) || + dm_deleting_md(md) || + test_bit(DMF_FREEING, &md->flags))) { + md = NULL; + goto out; + } + dm_get(md); } out: @@ -2486,16 +2489,6 @@ out: return md; } - -struct mapped_device *dm_get_md(dev_t dev) -{ - struct mapped_device *md = dm_find_md(dev); - - if (md) - dm_get(md); - - return md; -} EXPORT_SYMBOL_GPL(dm_get_md); void *dm_get_mdptr(struct mapped_device *md) @@ -2514,6 +2507,19 @@ void dm_get(struct mapped_device *md) BUG_ON(test_bit(DMF_FREEING, &md->flags)); } +int dm_hold(struct mapped_device *md) +{ + spin_lock(&_minor_lock); + if (test_bit(DMF_FREEING, &md->flags)) { + spin_unlock(&_minor_lock); + return -EBUSY; + } + dm_get(md); + spin_unlock(&_minor_lock); + return 0; +} +EXPORT_SYMBOL_GPL(dm_hold); + const char *dm_device_name(struct mapped_device *md) { return md->name; @@ -2533,10 +2539,16 @@ static void __dm_destroy(struct mapped_device *md, bool wait) set_bit(DMF_FREEING, &md->flags); spin_unlock(&_minor_lock); + /* + * Take suspend_lock so that presuspend and postsuspend methods + * do not race with internal suspend. + */ + mutex_lock(&md->suspend_lock); if (!dm_suspended_md(md)) { dm_table_presuspend_targets(map); dm_table_postsuspend_targets(map); } + mutex_unlock(&md->suspend_lock); /* dm_put_live_table must be before msleep, otherwise deadlock is possible */ dm_put_live_table(md, srcu_idx); @@ -2888,6 +2900,7 @@ void dm_internal_suspend(struct mapped_device *md) flush_workqueue(md->wq); dm_wait_for_completion(md, TASK_UNINTERRUPTIBLE); } +EXPORT_SYMBOL_GPL(dm_internal_suspend); void dm_internal_resume(struct mapped_device *md) { @@ -2899,6 +2912,7 @@ void dm_internal_resume(struct mapped_device *md) done: mutex_unlock(&md->suspend_lock); } +EXPORT_SYMBOL_GPL(dm_internal_resume); /*----------------------------------------------------------------- * Event notification. diff --git a/drivers/md/md.c b/drivers/md/md.c index 9233c71138f1..a4b0b4e43b40 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -282,6 +282,8 @@ static void md_make_request(struct request_queue *q, struct bio *bio) * go away inside make_request */ sectors = bio_sectors(bio); + /* bio could be mergeable after passing to underlayer */ + bio->bi_rw &= ~REQ_NOMERGE; mddev->pers->make_request(mddev, bio); cpu = part_stat_lock(); @@ -5073,6 +5075,8 @@ EXPORT_SYMBOL_GPL(md_stop_writes); static void __md_stop(struct mddev *mddev) { mddev->ready = 0; + /* Ensure ->event_work is done */ + flush_workqueue(md_misc_wq); mddev->pers->stop(mddev); if (mddev->pers->sync_request && mddev->to_remove == NULL) mddev->to_remove = &md_redundancy_group; @@ -5432,8 +5436,7 @@ static int get_bitmap_file(struct mddev *mddev, void __user * arg) char *ptr, *buf = NULL; int err = -ENOMEM; - file = kmalloc(sizeof(*file), GFP_NOIO); - + file = kzalloc(sizeof(*file), GFP_NOIO); if (!file) goto out; @@ -6031,7 +6034,7 @@ static int update_array_info(struct mddev *mddev, mdu_array_info_t *info) mddev->ctime != info->ctime || mddev->level != info->level || /* mddev->layout != info->layout || */ - !mddev->persistent != info->not_persistent|| + mddev->persistent != !info->not_persistent || mddev->chunk_sectors != info->chunk_size >> 9 || /* ignore bottom 8 bits of state, and allow SB_BITMAP_PRESENT to change */ ((state^info->state) & 0xfffffe00) @@ -7406,6 +7409,9 @@ void md_do_sync(struct md_thread *thread) break; j += sectors; + if (j > max_sectors) + /* when skipping, extra large numbers can be returned. */ + j = max_sectors; if (j > 2) mddev->curr_resync = j; mddev->curr_mark_cnt = io_sectors; @@ -7465,11 +7471,17 @@ void md_do_sync(struct md_thread *thread) blk_finish_plug(&plug); wait_event(mddev->recovery_wait, !atomic_read(&mddev->recovery_active)); + if (!test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery) && + !test_bit(MD_RECOVERY_INTR, &mddev->recovery) && + mddev->curr_resync > 3) { + mddev->curr_resync_completed = mddev->curr_resync; + sysfs_notify(&mddev->kobj, NULL, "sync_completed"); + } /* tell personality that we are finished */ mddev->pers->sync_request(mddev, max_sectors, &skipped, 1); if (!test_bit(MD_RECOVERY_CHECK, &mddev->recovery) && - mddev->curr_resync > 2) { + mddev->curr_resync > 3) { if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) { if (test_bit(MD_RECOVERY_INTR, &mddev->recovery)) { if (mddev->curr_resync >= mddev->recovery_cp) { @@ -7559,8 +7571,7 @@ static int remove_and_add_spares(struct mddev *mddev, !test_bit(Bitmap_sync, &rdev->flags))) continue; - if (rdev->saved_raid_disk < 0) - rdev->recovery_offset = 0; + rdev->recovery_offset = 0; if (mddev->pers-> hot_add_disk(mddev, rdev) == 0) { if (sysfs_link_rdev(mddev, rdev)) diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c index 399272f9c042..2df218da1791 100644 --- a/drivers/md/multipath.c +++ b/drivers/md/multipath.c @@ -129,7 +129,9 @@ static void multipath_make_request(struct mddev *mddev, struct bio * bio) } multipath = conf->multipaths + mp_bh->path; - mp_bh->bio = *bio; + bio_init(&mp_bh->bio); + __bio_clone_fast(&mp_bh->bio, bio); + mp_bh->bio.bi_iter.bi_sector += multipath->rdev->data_offset; mp_bh->bio.bi_bdev = multipath->rdev->bdev; mp_bh->bio.bi_rw |= REQ_FAILFAST_TRANSPORT; diff --git a/drivers/md/persistent-data/dm-btree-internal.h b/drivers/md/persistent-data/dm-btree-internal.h index bf2b80d5c470..8731b6ea026b 100644 --- a/drivers/md/persistent-data/dm-btree-internal.h +++ b/drivers/md/persistent-data/dm-btree-internal.h @@ -138,4 +138,10 @@ int lower_bound(struct btree_node *n, uint64_t key); extern struct dm_block_validator btree_node_validator; +/* + * Value type for upper levels of multi-level btrees. + */ +extern void init_le64_type(struct dm_transaction_manager *tm, + struct dm_btree_value_type *vt); + #endif /* DM_BTREE_INTERNAL_H */ diff --git a/drivers/md/persistent-data/dm-btree-remove.c b/drivers/md/persistent-data/dm-btree-remove.c index b88757cd0d1d..92cd09f3c69b 100644 --- a/drivers/md/persistent-data/dm-btree-remove.c +++ b/drivers/md/persistent-data/dm-btree-remove.c @@ -301,35 +301,40 @@ static void redistribute3(struct dm_btree_info *info, struct btree_node *parent, { int s; uint32_t max_entries = le32_to_cpu(left->header.max_entries); - unsigned target = (nr_left + nr_center + nr_right) / 3; - BUG_ON(target > max_entries); + unsigned total = nr_left + nr_center + nr_right; + unsigned target_right = total / 3; + unsigned remainder = (target_right * 3) != total; + unsigned target_left = target_right + remainder; + + BUG_ON(target_left > max_entries); + BUG_ON(target_right > max_entries); if (nr_left < nr_right) { - s = nr_left - target; + s = nr_left - target_left; if (s < 0 && nr_center < -s) { /* not enough in central node */ - shift(left, center, nr_center); - s = nr_center - target; + shift(left, center, -nr_center); + s += nr_center; shift(left, right, s); nr_right += s; } else shift(left, center, s); - shift(center, right, target - nr_right); + shift(center, right, target_right - nr_right); } else { - s = target - nr_right; + s = target_right - nr_right; if (s > 0 && nr_center < s) { /* not enough in central node */ shift(center, right, nr_center); - s = target - nr_center; + s -= nr_center; shift(left, right, s); nr_left -= s; } else shift(center, right, s); - shift(left, center, nr_left - target); + shift(left, center, nr_left - target_left); } *key_ptr(parent, c->index) = center->keys[0]; @@ -544,14 +549,6 @@ static int remove_raw(struct shadow_spine *s, struct dm_btree_info *info, return r; } -static struct dm_btree_value_type le64_type = { - .context = NULL, - .size = sizeof(__le64), - .inc = NULL, - .dec = NULL, - .equal = NULL -}; - int dm_btree_remove(struct dm_btree_info *info, dm_block_t root, uint64_t *keys, dm_block_t *new_root) { @@ -559,12 +556,14 @@ int dm_btree_remove(struct dm_btree_info *info, dm_block_t root, int index = 0, r = 0; struct shadow_spine spine; struct btree_node *n; + struct dm_btree_value_type le64_vt; + init_le64_type(info->tm, &le64_vt); init_shadow_spine(&spine, info); for (level = 0; level < info->levels; level++) { r = remove_raw(&spine, info, (level == last_level ? - &info->value_type : &le64_type), + &info->value_type : &le64_vt), root, keys[level], (unsigned *)&index); if (r < 0) break; diff --git a/drivers/md/persistent-data/dm-btree-spine.c b/drivers/md/persistent-data/dm-btree-spine.c index 1b5e13ec7f96..0dee514ba4c5 100644 --- a/drivers/md/persistent-data/dm-btree-spine.c +++ b/drivers/md/persistent-data/dm-btree-spine.c @@ -249,3 +249,40 @@ int shadow_root(struct shadow_spine *s) { return s->root; } + +static void le64_inc(void *context, const void *value_le) +{ + struct dm_transaction_manager *tm = context; + __le64 v_le; + + memcpy(&v_le, value_le, sizeof(v_le)); + dm_tm_inc(tm, le64_to_cpu(v_le)); +} + +static void le64_dec(void *context, const void *value_le) +{ + struct dm_transaction_manager *tm = context; + __le64 v_le; + + memcpy(&v_le, value_le, sizeof(v_le)); + dm_tm_dec(tm, le64_to_cpu(v_le)); +} + +static int le64_equal(void *context, const void *value1_le, const void *value2_le) +{ + __le64 v1_le, v2_le; + + memcpy(&v1_le, value1_le, sizeof(v1_le)); + memcpy(&v2_le, value2_le, sizeof(v2_le)); + return v1_le == v2_le; +} + +void init_le64_type(struct dm_transaction_manager *tm, + struct dm_btree_value_type *vt) +{ + vt->context = tm; + vt->size = sizeof(__le64); + vt->inc = le64_inc; + vt->dec = le64_dec; + vt->equal = le64_equal; +} diff --git a/drivers/md/persistent-data/dm-btree.c b/drivers/md/persistent-data/dm-btree.c index 200ac12a1d40..d6e47033b5e0 100644 --- a/drivers/md/persistent-data/dm-btree.c +++ b/drivers/md/persistent-data/dm-btree.c @@ -255,7 +255,7 @@ int dm_btree_del(struct dm_btree_info *info, dm_block_t root) int r; struct del_stack *s; - s = kmalloc(sizeof(*s), GFP_KERNEL); + s = kmalloc(sizeof(*s), GFP_NOIO); if (!s) return -ENOMEM; s->info = info; @@ -523,7 +523,7 @@ static int btree_split_beneath(struct shadow_spine *s, uint64_t key) r = new_block(s->info, &right); if (r < 0) { - /* FIXME: put left */ + unlock_block(s->info, left); return r; } @@ -667,12 +667,7 @@ static int insert(struct dm_btree_info *info, dm_block_t root, struct btree_node *n; struct dm_btree_value_type le64_type; - le64_type.context = NULL; - le64_type.size = sizeof(__le64); - le64_type.inc = NULL; - le64_type.dec = NULL; - le64_type.equal = NULL; - + init_le64_type(info->tm, &le64_type); init_shadow_spine(&spine, info); for (level = 0; level < (info->levels - 1); level++) { diff --git a/drivers/md/persistent-data/dm-space-map-metadata.c b/drivers/md/persistent-data/dm-space-map-metadata.c index 786b689bdfc7..199c9ccd1f5d 100644 --- a/drivers/md/persistent-data/dm-space-map-metadata.c +++ b/drivers/md/persistent-data/dm-space-map-metadata.c @@ -204,6 +204,27 @@ static void in(struct sm_metadata *smm) smm->recursion_count++; } +static int apply_bops(struct sm_metadata *smm) +{ + int r = 0; + + while (!brb_empty(&smm->uncommitted)) { + struct block_op bop; + + r = brb_pop(&smm->uncommitted, &bop); + if (r) { + DMERR("bug in bop ring buffer"); + break; + } + + r = commit_bop(smm, &bop); + if (r) + break; + } + + return r; +} + static int out(struct sm_metadata *smm) { int r = 0; @@ -216,21 +237,8 @@ static int out(struct sm_metadata *smm) return -ENOMEM; } - if (smm->recursion_count == 1) { - while (!brb_empty(&smm->uncommitted)) { - struct block_op bop; - - r = brb_pop(&smm->uncommitted, &bop); - if (r) { - DMERR("bug in bop ring buffer"); - break; - } - - r = commit_bop(smm, &bop); - if (r) - break; - } - } + if (smm->recursion_count == 1) + apply_bops(smm); smm->recursion_count--; @@ -564,7 +572,9 @@ static int sm_bootstrap_get_nr_blocks(struct dm_space_map *sm, dm_block_t *count { struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm); - return smm->ll.nr_blocks; + *count = smm->ll.nr_blocks; + + return 0; } static int sm_bootstrap_get_nr_free(struct dm_space_map *sm, dm_block_t *count) @@ -700,6 +710,12 @@ static int sm_metadata_extend(struct dm_space_map *sm, dm_block_t extra_blocks) } old_len = smm->begin; + r = apply_bops(smm); + if (r) { + DMERR("%s: apply_bops failed", __func__); + goto out; + } + r = sm_ll_commit(&smm->ll); if (r) goto out; @@ -769,6 +785,12 @@ int dm_sm_metadata_create(struct dm_space_map *sm, if (r) return r; + r = apply_bops(smm); + if (r) { + DMERR("%s: apply_bops failed", __func__); + return r; + } + return sm_metadata_commit(sm); } diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c index ba6b85de96d2..b974ff29b55c 100644 --- a/drivers/md/raid0.c +++ b/drivers/md/raid0.c @@ -319,7 +319,7 @@ static struct strip_zone *find_zone(struct r0conf *conf, /* * remaps the bio to the target device. we separate two flows. - * power 2 flow and a general flow for the sake of perfromance + * power 2 flow and a general flow for the sake of performance */ static struct md_rdev *map_sector(struct mddev *mddev, struct strip_zone *zone, sector_t sector, sector_t *sector_offset) @@ -530,6 +530,9 @@ static void raid0_make_request(struct mddev *mddev, struct bio *bio) ? (sector & (chunk_sects-1)) : sector_div(sector, chunk_sects)); + /* Restore due to sector_div */ + sector = bio->bi_iter.bi_sector; + if (sectors < bio_sectors(bio)) { split = bio_split(bio, sectors, GFP_NOIO, fs_bio_set); bio_chain(split, bio); diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index 40b35be34f8d..3c8ada48ca76 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -336,7 +336,7 @@ static void raid1_end_read_request(struct bio *bio, int error) spin_lock_irqsave(&conf->device_lock, flags); if (r1_bio->mddev->degraded == conf->raid_disks || (r1_bio->mddev->degraded == conf->raid_disks-1 && - !test_bit(Faulty, &conf->mirrors[mirror].rdev->flags))) + test_bit(In_sync, &conf->mirrors[mirror].rdev->flags))) uptodate = 1; spin_unlock_irqrestore(&conf->device_lock, flags); } @@ -560,7 +560,7 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect if (test_bit(WriteMostly, &rdev->flags)) { /* Don't balance among write-mostly, just * use the first as a last resort */ - if (best_disk < 0) { + if (best_dist_disk < 0) { if (is_badblock(rdev, this_sector, sectors, &first_bad, &bad_sectors)) { if (first_bad < this_sector) @@ -569,7 +569,8 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect best_good_sectors = first_bad - this_sector; } else best_good_sectors = sectors; - best_disk = disk; + best_dist_disk = disk; + best_pending_disk = disk; } continue; } @@ -1473,6 +1474,7 @@ static void error(struct mddev *mddev, struct md_rdev *rdev) { char b[BDEVNAME_SIZE]; struct r1conf *conf = mddev->private; + unsigned long flags; /* * If it is not operational, then we have already marked it as dead @@ -1492,14 +1494,13 @@ static void error(struct mddev *mddev, struct md_rdev *rdev) return; } set_bit(Blocked, &rdev->flags); + spin_lock_irqsave(&conf->device_lock, flags); if (test_and_clear_bit(In_sync, &rdev->flags)) { - unsigned long flags; - spin_lock_irqsave(&conf->device_lock, flags); mddev->degraded++; set_bit(Faulty, &rdev->flags); - spin_unlock_irqrestore(&conf->device_lock, flags); } else set_bit(Faulty, &rdev->flags); + spin_unlock_irqrestore(&conf->device_lock, flags); /* * if recovery is running, make sure it aborts. */ @@ -1565,7 +1566,10 @@ static int raid1_spare_active(struct mddev *mddev) * Find all failed disks within the RAID1 configuration * and mark them readable. * Called under mddev lock, so rcu protection not needed. + * device_lock used to avoid races with raid1_end_read_request + * which expects 'In_sync' flags and ->degraded to be consistent. */ + spin_lock_irqsave(&conf->device_lock, flags); for (i = 0; i < conf->raid_disks; i++) { struct md_rdev *rdev = conf->mirrors[i].rdev; struct md_rdev *repl = conf->mirrors[conf->raid_disks + i].rdev; @@ -1595,7 +1599,6 @@ static int raid1_spare_active(struct mddev *mddev) sysfs_notify_dirent_safe(rdev->sysfs_state); } } - spin_lock_irqsave(&conf->device_lock, flags); mddev->degraded -= count; spin_unlock_irqrestore(&conf->device_lock, flags); @@ -2242,7 +2245,7 @@ static int narrow_write_error(struct r1bio *r1_bio, int i) bio_trim(wbio, sector - r1_bio->sector, sectors); wbio->bi_iter.bi_sector += rdev->data_offset; wbio->bi_bdev = rdev->bdev; - if (submit_bio_wait(WRITE, wbio) == 0) + if (submit_bio_wait(WRITE, wbio) < 0) /* failure! */ ok = rdev_set_badblocks(rdev, sector, sectors, 0) diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 32e282f4c83c..644f9e576736 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -2599,7 +2599,7 @@ static int narrow_write_error(struct r10bio *r10_bio, int i) choose_data_offset(r10_bio, rdev) + (sector - r10_bio->sector)); wbio->bi_bdev = rdev->bdev; - if (submit_bio_wait(WRITE, wbio) == 0) + if (submit_bio_wait(WRITE, wbio) < 0) /* Failure! */ ok = rdev_set_badblocks(rdev, sector, sectors, 0) @@ -3581,6 +3581,7 @@ static struct r10conf *setup_conf(struct mddev *mddev) /* far_copies must be 1 */ conf->prev.stride = conf->dev_sectors; } + conf->reshape_safe = conf->reshape_progress; spin_lock_init(&conf->device_lock); INIT_LIST_HEAD(&conf->retry_list); @@ -3788,7 +3789,6 @@ static int run(struct mddev *mddev) } conf->offset_diff = min_offset_diff; - conf->reshape_safe = conf->reshape_progress; clear_bit(MD_RECOVERY_SYNC, &mddev->recovery); clear_bit(MD_RECOVERY_CHECK, &mddev->recovery); set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery); @@ -4135,6 +4135,7 @@ static int raid10_start_reshape(struct mddev *mddev) conf->reshape_progress = size; } else conf->reshape_progress = 0; + conf->reshape_safe = conf->reshape_progress; spin_unlock_irq(&conf->device_lock); if (mddev->delta_disks && mddev->bitmap) { @@ -4201,6 +4202,7 @@ abort: rdev->new_data_offset = rdev->data_offset; smp_wmb(); conf->reshape_progress = MaxSector; + conf->reshape_safe = MaxSector; mddev->reshape_position = MaxSector; spin_unlock_irq(&conf->device_lock); return ret; @@ -4555,6 +4557,7 @@ static void end_reshape(struct r10conf *conf) md_finish_reshape(conf->mddev); smp_wmb(); conf->reshape_progress = MaxSector; + conf->reshape_safe = MaxSector; spin_unlock_irq(&conf->device_lock); /* read-ahead size must cover two whole stripes, which is diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 9c66e5997fc8..cdd8770de1c2 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -1936,7 +1936,8 @@ static int resize_stripes(struct r5conf *conf, int newsize) conf->slab_cache = sc; conf->active_name = 1-conf->active_name; - conf->pool_size = newsize; + if (!err) + conf->pool_size = newsize; return err; } @@ -2917,8 +2918,11 @@ static int fetch_block(struct stripe_head *sh, struct stripe_head_state *s, (sh->raid_conf->level <= 5 && s->failed && fdev[0]->towrite && (!test_bit(R5_Insync, &dev->flags) || test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) && !test_bit(R5_OVERWRITE, &fdev[0]->flags)) || - (sh->raid_conf->level == 6 && s->failed && s->to_write && - s->to_write - s->non_overwrite < sh->raid_conf->raid_disks - 2 && + ((sh->raid_conf->level == 6 || + sh->sector >= sh->raid_conf->mddev->recovery_cp) + && s->failed && s->to_write && + (s->to_write - s->non_overwrite < + sh->raid_conf->raid_disks - sh->raid_conf->max_degraded) && (!test_bit(R5_Insync, &dev->flags) || test_bit(STRIPE_PREREAD_ACTIVE, &sh->state))))) { /* we would like to get this block, possibly by computing it, * otherwise read it if the backing disk is insync @@ -3056,6 +3060,8 @@ static void handle_stripe_clean_event(struct r5conf *conf, } if (!discard_pending && test_bit(R5_Discard, &sh->dev[sh->pd_idx].flags)) { + int hash = sh->hash_lock_index; + clear_bit(R5_Discard, &sh->dev[sh->pd_idx].flags); clear_bit(R5_UPTODATE, &sh->dev[sh->pd_idx].flags); if (sh->qd_idx >= 0) { @@ -3069,9 +3075,9 @@ static void handle_stripe_clean_event(struct r5conf *conf, * no updated data, so remove it from hash list and the stripe * will be reinitialized */ - spin_lock_irq(&conf->device_lock); + spin_lock_irq(conf->hash_locks + hash); remove_hash(sh); - spin_unlock_irq(&conf->device_lock); + spin_unlock_irq(conf->hash_locks + hash); if (test_bit(STRIPE_SYNC_REQUESTED, &sh->state)) set_bit(STRIPE_HANDLE, &sh->state); @@ -3099,7 +3105,8 @@ static void handle_stripe_dirtying(struct r5conf *conf, * generate correct data from the parity. */ if (conf->max_degraded == 2 || - (recovery_cp < MaxSector && sh->sector >= recovery_cp)) { + (recovery_cp < MaxSector && sh->sector >= recovery_cp && + s->failed == 0)) { /* Calculate the real rcw later - for now make it * look like rcw is cheaper */ @@ -3192,6 +3199,11 @@ static void handle_stripe_dirtying(struct r5conf *conf, (unsigned long long)sh->sector, rcw, qread, test_bit(STRIPE_DELAYED, &sh->state)); } + + if (rcw > disks && rmw > disks && + !test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) + set_bit(STRIPE_DELAYED, &sh->state); + /* now if nothing is locked, and if we have enough data, * we can start a write request */ @@ -6233,8 +6245,8 @@ static int run(struct mddev *mddev) } if (discard_supported && - mddev->queue->limits.max_discard_sectors >= stripe && - mddev->queue->limits.discard_granularity >= stripe) + mddev->queue->limits.max_discard_sectors >= (stripe >> 9) && + mddev->queue->limits.discard_granularity >= stripe) queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, mddev->queue); else |