diff options
Diffstat (limited to 'fs/btrfs/disk-io.c')
-rw-r--r-- | fs/btrfs/disk-io.c | 194 |
1 files changed, 102 insertions, 92 deletions
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 6628fca9f4ed..87dad552e39a 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -101,7 +101,7 @@ int __init btrfs_end_io_wq_init(void) btrfs_end_io_wq_cache = kmem_cache_create("btrfs_end_io_wq", sizeof(struct btrfs_end_io_wq), 0, - SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, + SLAB_MEM_SPREAD, NULL); if (!btrfs_end_io_wq_cache) return -ENOMEM; @@ -124,7 +124,6 @@ struct async_submit_bio { struct list_head list; extent_submit_bio_hook_t *submit_bio_start; extent_submit_bio_hook_t *submit_bio_done; - int rw; int mirror_num; unsigned long bio_flags; /* @@ -727,7 +726,7 @@ static void end_workqueue_bio(struct bio *bio) fs_info = end_io_wq->info; end_io_wq->error = bio->bi_error; - if (bio->bi_rw & REQ_WRITE) { + if (bio_op(bio) == REQ_OP_WRITE) { if (end_io_wq->metadata == BTRFS_WQ_ENDIO_METADATA) { wq = fs_info->endio_meta_write_workers; func = btrfs_endio_meta_write_helper; @@ -797,7 +796,7 @@ static void run_one_async_start(struct btrfs_work *work) int ret; async = container_of(work, struct async_submit_bio, work); - ret = async->submit_bio_start(async->inode, async->rw, async->bio, + ret = async->submit_bio_start(async->inode, async->bio, async->mirror_num, async->bio_flags, async->bio_offset); if (ret) @@ -830,9 +829,8 @@ static void run_one_async_done(struct btrfs_work *work) return; } - async->submit_bio_done(async->inode, async->rw, async->bio, - async->mirror_num, async->bio_flags, - async->bio_offset); + async->submit_bio_done(async->inode, async->bio, async->mirror_num, + async->bio_flags, async->bio_offset); } static void run_one_async_free(struct btrfs_work *work) @@ -844,7 +842,7 @@ static void run_one_async_free(struct btrfs_work *work) } int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode, - int rw, struct bio *bio, int mirror_num, + struct bio *bio, int mirror_num, unsigned long bio_flags, u64 bio_offset, extent_submit_bio_hook_t *submit_bio_start, @@ -857,7 +855,6 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode, return -ENOMEM; async->inode = inode; - async->rw = rw; async->bio = bio; async->mirror_num = mirror_num; async->submit_bio_start = submit_bio_start; @@ -873,7 +870,7 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode, atomic_inc(&fs_info->nr_async_submits); - if (rw & REQ_SYNC) + if (bio->bi_rw & REQ_SYNC) btrfs_set_work_high_priority(&async->work); btrfs_queue_work(fs_info->workers, &async->work); @@ -903,9 +900,8 @@ static int btree_csum_one_bio(struct bio *bio) return ret; } -static int __btree_submit_bio_start(struct inode *inode, int rw, - struct bio *bio, int mirror_num, - unsigned long bio_flags, +static int __btree_submit_bio_start(struct inode *inode, struct bio *bio, + int mirror_num, unsigned long bio_flags, u64 bio_offset) { /* @@ -915,7 +911,7 @@ static int __btree_submit_bio_start(struct inode *inode, int rw, return btree_csum_one_bio(bio); } -static int __btree_submit_bio_done(struct inode *inode, int rw, struct bio *bio, +static int __btree_submit_bio_done(struct inode *inode, struct bio *bio, int mirror_num, unsigned long bio_flags, u64 bio_offset) { @@ -925,7 +921,7 @@ static int __btree_submit_bio_done(struct inode *inode, int rw, struct bio *bio, * when we're called for a write, we're already in the async * submission context. Just jump into btrfs_map_bio */ - ret = btrfs_map_bio(BTRFS_I(inode)->root, rw, bio, mirror_num, 1); + ret = btrfs_map_bio(BTRFS_I(inode)->root, bio, mirror_num, 1); if (ret) { bio->bi_error = ret; bio_endio(bio); @@ -944,14 +940,14 @@ static int check_async_write(struct inode *inode, unsigned long bio_flags) return 1; } -static int btree_submit_bio_hook(struct inode *inode, int rw, struct bio *bio, +static int btree_submit_bio_hook(struct inode *inode, struct bio *bio, int mirror_num, unsigned long bio_flags, u64 bio_offset) { int async = check_async_write(inode, bio_flags); int ret; - if (!(rw & REQ_WRITE)) { + if (bio_op(bio) != REQ_OP_WRITE) { /* * called for a read, do the setup so that checksum validation * can happen in the async kernel threads @@ -960,21 +956,19 @@ static int btree_submit_bio_hook(struct inode *inode, int rw, struct bio *bio, bio, BTRFS_WQ_ENDIO_METADATA); if (ret) goto out_w_error; - ret = btrfs_map_bio(BTRFS_I(inode)->root, rw, bio, - mirror_num, 0); + ret = btrfs_map_bio(BTRFS_I(inode)->root, bio, mirror_num, 0); } else if (!async) { ret = btree_csum_one_bio(bio); if (ret) goto out_w_error; - ret = btrfs_map_bio(BTRFS_I(inode)->root, rw, bio, - mirror_num, 0); + ret = btrfs_map_bio(BTRFS_I(inode)->root, bio, mirror_num, 0); } else { /* * kthread helpers are used to submit writes so that * checksumming can happen in parallel across all CPUs */ ret = btrfs_wq_submit_bio(BTRFS_I(inode)->root->fs_info, - inode, rw, bio, mirror_num, 0, + inode, bio, mirror_num, 0, bio_offset, __btree_submit_bio_start, __btree_submit_bio_done); @@ -1098,7 +1092,7 @@ void readahead_tree_block(struct btrfs_root *root, u64 bytenr) struct inode *btree_inode = root->fs_info->btree_inode; buf = btrfs_find_create_tree_block(root, bytenr); - if (!buf) + if (IS_ERR(buf)) return; read_extent_buffer_pages(&BTRFS_I(btree_inode)->io_tree, buf, 0, WAIT_NONE, btree_get_extent, 0); @@ -1114,7 +1108,7 @@ int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, int ret; buf = btrfs_find_create_tree_block(root, bytenr); - if (!buf) + if (IS_ERR(buf)) return 0; set_bit(EXTENT_BUFFER_READAHEAD, &buf->bflags); @@ -1146,8 +1140,9 @@ struct extent_buffer *btrfs_find_tree_block(struct btrfs_fs_info *fs_info, struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root, u64 bytenr) { - if (btrfs_test_is_dummy_root(root)) - return alloc_test_extent_buffer(root->fs_info, bytenr); + if (btrfs_is_testing(root->fs_info)) + return alloc_test_extent_buffer(root->fs_info, bytenr, + root->nodesize); return alloc_extent_buffer(root->fs_info, bytenr); } @@ -1171,8 +1166,8 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr, int ret; buf = btrfs_find_create_tree_block(root, bytenr); - if (!buf) - return ERR_PTR(-ENOMEM); + if (IS_ERR(buf)) + return buf; ret = btree_read_extent_buffer_pages(root, buf, 0, parent_transid); if (ret) { @@ -1232,6 +1227,7 @@ static void __setup_root(u32 nodesize, u32 sectorsize, u32 stripesize, struct btrfs_root *root, struct btrfs_fs_info *fs_info, u64 objectid) { + bool dummy = test_bit(BTRFS_FS_STATE_DUMMY_FS_INFO, &fs_info->fs_state); root->node = NULL; root->commit_root = NULL; root->sectorsize = sectorsize; @@ -1286,14 +1282,14 @@ static void __setup_root(u32 nodesize, u32 sectorsize, u32 stripesize, root->log_transid = 0; root->log_transid_committed = -1; root->last_log_commit = 0; - if (fs_info) + if (!dummy) extent_io_tree_init(&root->dirty_log_pages, fs_info->btree_inode->i_mapping); memset(&root->root_key, 0, sizeof(root->root_key)); memset(&root->root_item, 0, sizeof(root->root_item)); memset(&root->defrag_progress, 0, sizeof(root->defrag_progress)); - if (fs_info) + if (!dummy) root->defrag_trans_start = fs_info->generation; else root->defrag_trans_start = 0; @@ -1314,15 +1310,20 @@ static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info, #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS /* Should only be used by the testing infrastructure */ -struct btrfs_root *btrfs_alloc_dummy_root(void) +struct btrfs_root *btrfs_alloc_dummy_root(struct btrfs_fs_info *fs_info, + u32 sectorsize, u32 nodesize) { struct btrfs_root *root; - root = btrfs_alloc_root(NULL, GFP_KERNEL); + if (!fs_info) + return ERR_PTR(-EINVAL); + + root = btrfs_alloc_root(fs_info, GFP_KERNEL); if (!root) return ERR_PTR(-ENOMEM); - __setup_root(4096, 4096, 4096, root, NULL, 1); - set_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state); + /* We don't use the stripesize in selftest, set it as sectorsize */ + __setup_root(nodesize, sectorsize, sectorsize, root, fs_info, + BTRFS_ROOT_TREE_OBJECTID); root->alloc_bytenr = 0; return root; @@ -1597,14 +1598,14 @@ int btrfs_init_fs_root(struct btrfs_root *root) ret = get_anon_bdev(&root->anon_dev); if (ret) - goto free_writers; + goto fail; mutex_lock(&root->objectid_mutex); ret = btrfs_find_highest_objectid(root, &root->highest_objectid); if (ret) { mutex_unlock(&root->objectid_mutex); - goto free_root_dev; + goto fail; } ASSERT(root->highest_objectid <= BTRFS_LAST_FREE_OBJECTID); @@ -1612,14 +1613,8 @@ int btrfs_init_fs_root(struct btrfs_root *root) mutex_unlock(&root->objectid_mutex); return 0; - -free_root_dev: - free_anon_bdev(root->anon_dev); -free_writers: - btrfs_free_subvolume_writers(root->subv_writers); fail: - kfree(root->free_ino_ctl); - kfree(root->free_ino_pinned); + /* the caller is responsible to call free_fs_root */ return ret; } @@ -1803,6 +1798,13 @@ static int cleaner_kthread(void *arg) if (btrfs_need_cleaner_sleep(root)) goto sleep; + /* + * Do not do anything if we might cause open_ctree() to block + * before we have finished mounting the filesystem. + */ + if (!root->fs_info->open) + goto sleep; + if (!mutex_trylock(&root->fs_info->cleaner_mutex)) goto sleep; @@ -2306,17 +2308,19 @@ static int btrfs_init_workqueues(struct btrfs_fs_info *fs_info, unsigned int flags = WQ_MEM_RECLAIM | WQ_FREEZABLE | WQ_UNBOUND; fs_info->workers = - btrfs_alloc_workqueue("worker", flags | WQ_HIGHPRI, - max_active, 16); + btrfs_alloc_workqueue(fs_info, "worker", + flags | WQ_HIGHPRI, max_active, 16); fs_info->delalloc_workers = - btrfs_alloc_workqueue("delalloc", flags, max_active, 2); + btrfs_alloc_workqueue(fs_info, "delalloc", + flags, max_active, 2); fs_info->flush_workers = - btrfs_alloc_workqueue("flush_delalloc", flags, max_active, 0); + btrfs_alloc_workqueue(fs_info, "flush_delalloc", + flags, max_active, 0); fs_info->caching_workers = - btrfs_alloc_workqueue("cache", flags, max_active, 0); + btrfs_alloc_workqueue(fs_info, "cache", flags, max_active, 0); /* * a higher idle thresh on the submit workers makes it much more @@ -2324,41 +2328,48 @@ static int btrfs_init_workqueues(struct btrfs_fs_info *fs_info, * devices */ fs_info->submit_workers = - btrfs_alloc_workqueue("submit", flags, + btrfs_alloc_workqueue(fs_info, "submit", flags, min_t(u64, fs_devices->num_devices, max_active), 64); fs_info->fixup_workers = - btrfs_alloc_workqueue("fixup", flags, 1, 0); + btrfs_alloc_workqueue(fs_info, "fixup", flags, 1, 0); /* * endios are largely parallel and should have a very * low idle thresh */ fs_info->endio_workers = - btrfs_alloc_workqueue("endio", flags, max_active, 4); + btrfs_alloc_workqueue(fs_info, "endio", flags, max_active, 4); fs_info->endio_meta_workers = - btrfs_alloc_workqueue("endio-meta", flags, max_active, 4); + btrfs_alloc_workqueue(fs_info, "endio-meta", flags, + max_active, 4); fs_info->endio_meta_write_workers = - btrfs_alloc_workqueue("endio-meta-write", flags, max_active, 2); + btrfs_alloc_workqueue(fs_info, "endio-meta-write", flags, + max_active, 2); fs_info->endio_raid56_workers = - btrfs_alloc_workqueue("endio-raid56", flags, max_active, 4); + btrfs_alloc_workqueue(fs_info, "endio-raid56", flags, + max_active, 4); fs_info->endio_repair_workers = - btrfs_alloc_workqueue("endio-repair", flags, 1, 0); + btrfs_alloc_workqueue(fs_info, "endio-repair", flags, 1, 0); fs_info->rmw_workers = - btrfs_alloc_workqueue("rmw", flags, max_active, 2); + btrfs_alloc_workqueue(fs_info, "rmw", flags, max_active, 2); fs_info->endio_write_workers = - btrfs_alloc_workqueue("endio-write", flags, max_active, 2); + btrfs_alloc_workqueue(fs_info, "endio-write", flags, + max_active, 2); fs_info->endio_freespace_worker = - btrfs_alloc_workqueue("freespace-write", flags, max_active, 0); + btrfs_alloc_workqueue(fs_info, "freespace-write", flags, + max_active, 0); fs_info->delayed_workers = - btrfs_alloc_workqueue("delayed-meta", flags, max_active, 0); + btrfs_alloc_workqueue(fs_info, "delayed-meta", flags, + max_active, 0); fs_info->readahead_workers = - btrfs_alloc_workqueue("readahead", flags, max_active, 2); + btrfs_alloc_workqueue(fs_info, "readahead", flags, + max_active, 2); fs_info->qgroup_rescan_workers = - btrfs_alloc_workqueue("qgroup-rescan", flags, 1, 0); + btrfs_alloc_workqueue(fs_info, "qgroup-rescan", flags, 1, 0); fs_info->extent_workers = - btrfs_alloc_workqueue("extent-refs", flags, + btrfs_alloc_workqueue(fs_info, "extent-refs", flags, min_t(u64, fs_devices->num_devices, max_active), 8); @@ -2517,7 +2528,6 @@ int open_ctree(struct super_block *sb, int num_backups_tried = 0; int backup_index = 0; int max_active; - bool cleaner_mutex_locked = false; tree_root = fs_info->tree_root = btrfs_alloc_root(fs_info, GFP_KERNEL); chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info, GFP_KERNEL); @@ -2797,7 +2807,7 @@ int open_ctree(struct super_block *sb, nodesize = btrfs_super_nodesize(disk_super); sectorsize = btrfs_super_sectorsize(disk_super); - stripesize = btrfs_super_stripesize(disk_super); + stripesize = sectorsize; fs_info->dirty_metadata_batch = nodesize * (1 + ilog2(nr_cpu_ids)); fs_info->delalloc_batch = sectorsize * 512 * (1 + ilog2(nr_cpu_ids)); @@ -2996,13 +3006,6 @@ retry_root_backup: goto fail_sysfs; } - /* - * Hold the cleaner_mutex thread here so that we don't block - * for a long time on btrfs_recover_relocation. cleaner_kthread - * will wait for us to finish mounting the filesystem. - */ - mutex_lock(&fs_info->cleaner_mutex); - cleaner_mutex_locked = true; fs_info->cleaner_kthread = kthread_run(cleaner_kthread, tree_root, "btrfs-cleaner"); if (IS_ERR(fs_info->cleaner_kthread)) @@ -3014,8 +3017,8 @@ retry_root_backup: if (IS_ERR(fs_info->transaction_kthread)) goto fail_cleaner; - if (!btrfs_test_opt(tree_root, SSD) && - !btrfs_test_opt(tree_root, NOSSD) && + if (!btrfs_test_opt(tree_root->fs_info, SSD) && + !btrfs_test_opt(tree_root->fs_info, NOSSD) && !fs_info->fs_devices->rotating) { btrfs_info(fs_info, "detected SSD devices, enabling SSD mode"); btrfs_set_opt(fs_info->mount_opt, SSD); @@ -3028,9 +3031,9 @@ retry_root_backup: btrfs_apply_pending_changes(fs_info); #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY - if (btrfs_test_opt(tree_root, CHECK_INTEGRITY)) { + if (btrfs_test_opt(tree_root->fs_info, CHECK_INTEGRITY)) { ret = btrfsic_mount(tree_root, fs_devices, - btrfs_test_opt(tree_root, + btrfs_test_opt(tree_root->fs_info, CHECK_INTEGRITY_INCLUDING_EXTENT_DATA) ? 1 : 0, fs_info->check_integrity_print_mask); @@ -3046,7 +3049,7 @@ retry_root_backup: /* do not make disk changes in broken FS or nologreplay is given */ if (btrfs_super_log_root(disk_super) != 0 && - !btrfs_test_opt(tree_root, NOLOGREPLAY)) { + !btrfs_test_opt(tree_root->fs_info, NOLOGREPLAY)) { ret = btrfs_replay_log(fs_info, fs_devices); if (ret) { err = ret; @@ -3062,8 +3065,10 @@ retry_root_backup: ret = btrfs_cleanup_fs_roots(fs_info); if (ret) goto fail_qgroup; - /* We locked cleaner_mutex before creating cleaner_kthread. */ + + mutex_lock(&fs_info->cleaner_mutex); ret = btrfs_recover_relocation(tree_root); + mutex_unlock(&fs_info->cleaner_mutex); if (ret < 0) { btrfs_warn(fs_info, "failed to recover relocation: %d", ret); @@ -3071,8 +3076,6 @@ retry_root_backup: goto fail_qgroup; } } - mutex_unlock(&fs_info->cleaner_mutex); - cleaner_mutex_locked = false; location.objectid = BTRFS_FS_TREE_OBJECTID; location.type = BTRFS_ROOT_ITEM_KEY; @@ -3087,7 +3090,7 @@ retry_root_backup: if (sb->s_flags & MS_RDONLY) return 0; - if (btrfs_test_opt(tree_root, FREE_SPACE_TREE) && + if (btrfs_test_opt(tree_root->fs_info, FREE_SPACE_TREE) && !btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE)) { btrfs_info(fs_info, "creating free space tree"); ret = btrfs_create_free_space_tree(fs_info); @@ -3124,7 +3127,7 @@ retry_root_backup: btrfs_qgroup_rescan_resume(fs_info); - if (btrfs_test_opt(tree_root, CLEAR_CACHE) && + if (btrfs_test_opt(tree_root->fs_info, CLEAR_CACHE) && btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE)) { btrfs_info(fs_info, "clearing free space tree"); ret = btrfs_clear_free_space_tree(fs_info); @@ -3145,7 +3148,7 @@ retry_root_backup: close_ctree(tree_root); return ret; } - } else if (btrfs_test_opt(tree_root, RESCAN_UUID_TREE) || + } else if (btrfs_test_opt(tree_root->fs_info, RESCAN_UUID_TREE) || fs_info->generation != btrfs_super_uuid_tree_generation(disk_super)) { btrfs_info(fs_info, "checking UUID tree"); @@ -3186,10 +3189,6 @@ fail_cleaner: filemap_write_and_wait(fs_info->btree_inode->i_mapping); fail_sysfs: - if (cleaner_mutex_locked) { - mutex_unlock(&fs_info->cleaner_mutex); - cleaner_mutex_locked = false; - } btrfs_sysfs_remove_mounted(fs_info); fail_fsdev_sysfs: @@ -3226,7 +3225,7 @@ fail: return err; recovery_tree_root: - if (!btrfs_test_opt(tree_root, USEBACKUPROOT)) + if (!btrfs_test_opt(tree_root->fs_info, USEBACKUPROOT)) goto fail_tree_roots; free_root_pointers(fs_info, 0); @@ -3420,9 +3419,9 @@ static int write_dev_supers(struct btrfs_device *device, * to go down lazy. */ if (i == 0) - ret = btrfsic_submit_bh(WRITE_FUA, bh); + ret = btrfsic_submit_bh(REQ_OP_WRITE, WRITE_FUA, bh); else - ret = btrfsic_submit_bh(WRITE_SYNC, bh); + ret = btrfsic_submit_bh(REQ_OP_WRITE, WRITE_SYNC, bh); if (ret) errors++; } @@ -3486,12 +3485,13 @@ static int write_dev_flush(struct btrfs_device *device, int wait) bio->bi_end_io = btrfs_end_empty_barrier; bio->bi_bdev = device->bdev; + bio_set_op_attrs(bio, REQ_OP_WRITE, WRITE_FLUSH); init_completion(&device->flush_wait); bio->bi_private = &device->flush_wait; device->flush_bio = bio; bio_get(bio); - btrfsic_submit_bio(WRITE_FLUSH, bio); + btrfsic_submit_bio(bio); return 0; } @@ -3641,7 +3641,7 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors) int total_errors = 0; u64 flags; - do_barriers = !btrfs_test_opt(root, NOBARRIER); + do_barriers = !btrfs_test_opt(root->fs_info, NOBARRIER); backup_super_roots(root->fs_info); sb = root->fs_info->super_for_commit; @@ -3925,7 +3925,7 @@ void close_ctree(struct btrfs_root *root) iput(fs_info->btree_inode); #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY - if (btrfs_test_opt(root, CHECK_INTEGRITY)) + if (btrfs_test_opt(root->fs_info, CHECK_INTEGRITY)) btrfsic_unmount(root, fs_info->fs_devices); #endif @@ -4130,6 +4130,16 @@ static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info, * Hint to catch really bogus numbers, bitflips or so, more exact checks are * done later */ + if (btrfs_super_bytes_used(sb) < 6 * btrfs_super_nodesize(sb)) { + btrfs_err(fs_info, "bytes_used is too small %llu", + btrfs_super_bytes_used(sb)); + ret = -EINVAL; + } + if (!is_power_of_2(btrfs_super_stripesize(sb))) { + btrfs_err(fs_info, "invalid stripesize %u", + btrfs_super_stripesize(sb)); + ret = -EINVAL; + } if (btrfs_super_num_devices(sb) > (1UL << 31)) printk(KERN_WARNING "BTRFS: suspicious number of devices: %llu\n", btrfs_super_num_devices(sb)); |