Merge tag 'for-5.9-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:
 "Two small fixes and a bunch of lockdep fixes for warnings that show up
  with an upcoming tree locking update but are valid with current locks
  as well"

* tag 'for-5.9-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: tree-checker: fix the error message for transid error
  btrfs: set the lockdep class for log tree extent buffers
  btrfs: set the correct lockdep class for new nodes
  btrfs: allocate scrub workqueues outside of locks
  btrfs: fix potential deadlock in the search ioctl
  btrfs: drop path before adding new uuid tree entry
  btrfs: block-group: fix free-space bitmap threshold

Author: torvalds

fs/btrfs/block-group.c

@@ -1798,7 +1798,6 @@ static struct btrfs_block_group *btrfs_create_block_group_cache(
 
 	cache->fs_info = fs_info;
 	cache->full_stripe_len = btrfs_full_stripe_len(fs_info, start);
-	set_free_space_tree_thresholds(cache);
 
 	cache->discard_index = BTRFS_DISCARD_INDEX_UNUSED;
 
@@ -1912,6 +1911,8 @@ static int read_one_block_group(struct btrfs_fs_info *info,
 	if (ret < 0)
 		goto error;
 
+	set_free_space_tree_thresholds(cache);
+
 	if (need_clear) {
 		/*
 		 * When we mount with old space cache, we need to
@@ -2132,6 +2133,7 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, u64 bytes_used,
 		return -ENOMEM;
 
 	cache->length = size;
+	set_free_space_tree_thresholds(cache);
 	cache->used = bytes_used;
 	cache->flags = type;
 	cache->last_byte_to_unpin = (u64)-1;

fs/btrfs/ctree.c

@@ -1297,6 +1297,8 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 	btrfs_tree_read_unlock_blocking(eb);
 	free_extent_buffer(eb);
 
+	btrfs_set_buffer_lockdep_class(btrfs_header_owner(eb_rewin),
+				       eb_rewin, btrfs_header_level(eb_rewin));
 	btrfs_tree_read_lock(eb_rewin);
 	__tree_mod_log_rewind(fs_info, eb_rewin, time_seq, tm);
 	WARN_ON(btrfs_header_nritems(eb_rewin) >
@@ -1370,14 +1372,16 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 
 	if (!eb)
 		return NULL;
-	btrfs_tree_read_lock(eb);
 	if (old_root) {
 		btrfs_set_header_bytenr(eb, eb->start);
 		btrfs_set_header_backref_rev(eb, BTRFS_MIXED_BACKREF_REV);
 		btrfs_set_header_owner(eb, eb_root_owner);
 		btrfs_set_header_level(eb, old_root->level);
 		btrfs_set_header_generation(eb, old_generation);
 	}
+	btrfs_set_buffer_lockdep_class(btrfs_header_owner(eb), eb,
+				       btrfs_header_level(eb));
+	btrfs_tree_read_lock(eb);
 	if (tm)
 		__tree_mod_log_rewind(fs_info, eb, time_seq, tm);
 	else

fs/btrfs/extent-tree.c

@@ -4527,7 +4527,7 @@ btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		return ERR_PTR(-EUCLEAN);
 	}
 
-	btrfs_set_buffer_lockdep_class(root->root_key.objectid, buf, level);
+	btrfs_set_buffer_lockdep_class(owner, buf, level);
 	btrfs_tree_lock(buf);
 	btrfs_clean_tree_block(buf);
 	clear_bit(EXTENT_BUFFER_STALE, &buf->bflags);

fs/btrfs/extent_io.c

@@ -5655,9 +5655,9 @@ void read_extent_buffer(const struct extent_buffer *eb, void *dstv,
 	}
 }
 
-int read_extent_buffer_to_user(const struct extent_buffer *eb,
-			       void __user *dstv,
-			       unsigned long start, unsigned long len)
+int read_extent_buffer_to_user_nofault(const struct extent_buffer *eb,
+				       void __user *dstv,
+				       unsigned long start, unsigned long len)
 {
 	size_t cur;
 	size_t offset;
@@ -5677,7 +5677,7 @@ int read_extent_buffer_to_user(const struct extent_buffer *eb,
 
 		cur = min(len, (PAGE_SIZE - offset));
 		kaddr = page_address(page);
-		if (copy_to_user(dst, kaddr + offset, cur)) {
+		if (copy_to_user_nofault(dst, kaddr + offset, cur)) {
 			ret = -EFAULT;
 			break;
 		}

fs/btrfs/extent_io.h

@@ -241,9 +241,9 @@ int memcmp_extent_buffer(const struct extent_buffer *eb, const void *ptrv,
 void read_extent_buffer(const struct extent_buffer *eb, void *dst,
 			unsigned long start,
 			unsigned long len);
-int read_extent_buffer_to_user(const struct extent_buffer *eb,
-			       void __user *dst, unsigned long start,
-			       unsigned long len);
+int read_extent_buffer_to_user_nofault(const struct extent_buffer *eb,
+				       void __user *dst, unsigned long start,
+				       unsigned long len);
 void write_extent_buffer_fsid(const struct extent_buffer *eb, const void *src);
 void write_extent_buffer_chunk_tree_uuid(const struct extent_buffer *eb,
 		const void *src);

fs/btrfs/free-space-tree.c

@@ -22,6 +22,10 @@ void set_free_space_tree_thresholds(struct btrfs_block_group *cache)
 	size_t bitmap_size;
 	u64 num_bitmaps, total_bitmap_size;
 
+	if (WARN_ON(cache->length == 0))
+		btrfs_warn(cache->fs_info, "block group %llu length is zero",
+			   cache->start);
+
 	/*
 	 * We convert to bitmaps when the disk space required for using extents
 	 * exceeds that required for using bitmaps.

fs/btrfs/ioctl.c

@@ -2086,20 +2086,29 @@ static noinline int copy_to_sk(struct btrfs_path *path,
 		sh.len = item_len;
 		sh.transid = found_transid;
 
-		/* copy search result header */
-		if (copy_to_user(ubuf + *sk_offset, &sh, sizeof(sh))) {
-			ret = -EFAULT;
+		/*
+		 * Copy search result header. If we fault then loop again so we
+		 * can fault in the pages and -EFAULT there if there's a
+		 * problem. Otherwise we'll fault and then copy the buffer in
+		 * properly this next time through
+		 */
+		if (copy_to_user_nofault(ubuf + *sk_offset, &sh, sizeof(sh))) {
+			ret = 0;
 			goto out;
 		}
 
 		*sk_offset += sizeof(sh);
 
 		if (item_len) {
 			char __user *up = ubuf + *sk_offset;
-			/* copy the item */
-			if (read_extent_buffer_to_user(leaf, up,
-						       item_off, item_len)) {
-				ret = -EFAULT;
+			/*
+			 * Copy the item, same behavior as above, but reset the
+			 * * sk_offset so we copy the full thing again.
+			 */
+			if (read_extent_buffer_to_user_nofault(leaf, up,
+						item_off, item_len)) {
+				ret = 0;
+				*sk_offset -= sizeof(sh);
 				goto out;
 			}
 
@@ -2184,6 +2193,10 @@ static noinline int search_ioctl(struct inode *inode,
 	key.offset = sk->min_offset;
 
 	while (1) {
+		ret = fault_in_pages_writeable(ubuf, *buf_size - sk_offset);
+		if (ret)
+			break;
+
 		ret = btrfs_search_forward(root, &key, path, sk->min_transid);
 		if (ret != 0) {
 			if (ret > 0)

fs/btrfs/scrub.c

@@ -3716,50 +3716,84 @@ static noinline_for_stack int scrub_supers(struct scrub_ctx *sctx,
 	return 0;
 }
 
+static void scrub_workers_put(struct btrfs_fs_info *fs_info)
+{
+	if (refcount_dec_and_mutex_lock(&fs_info->scrub_workers_refcnt,
+					&fs_info->scrub_lock)) {
+		struct btrfs_workqueue *scrub_workers = NULL;
+		struct btrfs_workqueue *scrub_wr_comp = NULL;
+		struct btrfs_workqueue *scrub_parity = NULL;
+
+		scrub_workers = fs_info->scrub_workers;
+		scrub_wr_comp = fs_info->scrub_wr_completion_workers;
+		scrub_parity = fs_info->scrub_parity_workers;
+
+		fs_info->scrub_workers = NULL;
+		fs_info->scrub_wr_completion_workers = NULL;
+		fs_info->scrub_parity_workers = NULL;
+		mutex_unlock(&fs_info->scrub_lock);
+
+		btrfs_destroy_workqueue(scrub_workers);
+		btrfs_destroy_workqueue(scrub_wr_comp);
+		btrfs_destroy_workqueue(scrub_parity);
+	}
+}
+
 /*
  * get a reference count on fs_info->scrub_workers. start worker if necessary
  */
 static noinline_for_stack int scrub_workers_get(struct btrfs_fs_info *fs_info,
 						int is_dev_replace)
 {
+	struct btrfs_workqueue *scrub_workers = NULL;
+	struct btrfs_workqueue *scrub_wr_comp = NULL;
+	struct btrfs_workqueue *scrub_parity = NULL;
 	unsigned int flags = WQ_FREEZABLE | WQ_UNBOUND;
 	int max_active = fs_info->thread_pool_size;
+	int ret = -ENOMEM;
 
-	lockdep_assert_held(&fs_info->scrub_lock);
+	if (refcount_inc_not_zero(&fs_info->scrub_workers_refcnt))
+		return 0;
 
-	if (refcount_read(&fs_info->scrub_workers_refcnt) == 0) {
-		ASSERT(fs_info->scrub_workers == NULL);
-		fs_info->scrub_workers = btrfs_alloc_workqueue(fs_info, "scrub",
-				flags, is_dev_replace ? 1 : max_active, 4);
-		if (!fs_info->scrub_workers)
-			goto fail_scrub_workers;
-
-		ASSERT(fs_info->scrub_wr_completion_workers == NULL);
-		fs_info->scrub_wr_completion_workers =
-			btrfs_alloc_workqueue(fs_info, "scrubwrc", flags,
-					      max_active, 2);
-		if (!fs_info->scrub_wr_completion_workers)
-			goto fail_scrub_wr_completion_workers;
+	scrub_workers = btrfs_alloc_workqueue(fs_info, "scrub", flags,
+					      is_dev_replace ? 1 : max_active, 4);
+	if (!scrub_workers)
+		goto fail_scrub_workers;
 
-		ASSERT(fs_info->scrub_parity_workers == NULL);
-		fs_info->scrub_parity_workers =
-			btrfs_alloc_workqueue(fs_info, "scrubparity", flags,
+	scrub_wr_comp = btrfs_alloc_workqueue(fs_info, "scrubwrc", flags,
 					      max_active, 2);
-		if (!fs_info->scrub_parity_workers)
-			goto fail_scrub_parity_workers;
+	if (!scrub_wr_comp)
+		goto fail_scrub_wr_completion_workers;
 
+	scrub_parity = btrfs_alloc_workqueue(fs_info, "scrubparity", flags,
+					     max_active, 2);
+	if (!scrub_parity)
+		goto fail_scrub_parity_workers;
+
+	mutex_lock(&fs_info->scrub_lock);
+	if (refcount_read(&fs_info->scrub_workers_refcnt) == 0) {
+		ASSERT(fs_info->scrub_workers == NULL &&
+		       fs_info->scrub_wr_completion_workers == NULL &&
+		       fs_info->scrub_parity_workers == NULL);
+		fs_info->scrub_workers = scrub_workers;
+		fs_info->scrub_wr_completion_workers = scrub_wr_comp;
+		fs_info->scrub_parity_workers = scrub_parity;
 		refcount_set(&fs_info->scrub_workers_refcnt, 1);
-	} else {
-		refcount_inc(&fs_info->scrub_workers_refcnt);
+		mutex_unlock(&fs_info->scrub_lock);
+		return 0;
 	}
-	return 0;
+	/* Other thread raced in and created the workers for us */
+	refcount_inc(&fs_info->scrub_workers_refcnt);
+	mutex_unlock(&fs_info->scrub_lock);
 
+	ret = 0;
+	btrfs_destroy_workqueue(scrub_parity);
 fail_scrub_parity_workers:
-	btrfs_destroy_workqueue(fs_info->scrub_wr_completion_workers);
+	btrfs_destroy_workqueue(scrub_wr_comp);
 fail_scrub_wr_completion_workers:
-	btrfs_destroy_workqueue(fs_info->scrub_workers);
+	btrfs_destroy_workqueue(scrub_workers);
 fail_scrub_workers:
-	return -ENOMEM;
+	return ret;
 }
 
 int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
@@ -3770,9 +3804,6 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	int ret;
 	struct btrfs_device *dev;
 	unsigned int nofs_flag;
-	struct btrfs_workqueue *scrub_workers = NULL;
-	struct btrfs_workqueue *scrub_wr_comp = NULL;
-	struct btrfs_workqueue *scrub_parity = NULL;
 
 	if (btrfs_fs_closing(fs_info))
 		return -EAGAIN;
@@ -3819,13 +3850,17 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	if (IS_ERR(sctx))
 		return PTR_ERR(sctx);
 
+	ret = scrub_workers_get(fs_info, is_dev_replace);
+	if (ret)
+		goto out_free_ctx;
+
 	mutex_lock(&fs_info->fs_devices->device_list_mutex);
 	dev = btrfs_find_device(fs_info->fs_devices, devid, NULL, NULL, true);
 	if (!dev || (test_bit(BTRFS_DEV_STATE_MISSING, &dev->dev_state) &&
 		     !is_dev_replace)) {
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 		ret = -ENODEV;
-		goto out_free_ctx;
+		goto out;
 	}
 
 	if (!is_dev_replace && !readonly &&
@@ -3834,7 +3869,7 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		btrfs_err_in_rcu(fs_info, "scrub: device %s is not writable",
 				rcu_str_deref(dev->name));
 		ret = -EROFS;
-		goto out_free_ctx;
+		goto out;
 	}
 
 	mutex_lock(&fs_info->scrub_lock);
@@ -3843,7 +3878,7 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		mutex_unlock(&fs_info->scrub_lock);
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 		ret = -EIO;
-		goto out_free_ctx;
+		goto out;
 	}
 
 	down_read(&fs_info->dev_replace.rwsem);
@@ -3854,17 +3889,10 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		mutex_unlock(&fs_info->scrub_lock);
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 		ret = -EINPROGRESS;
-		goto out_free_ctx;
+		goto out;
 	}
 	up_read(&fs_info->dev_replace.rwsem);
 
-	ret = scrub_workers_get(fs_info, is_dev_replace);
-	if (ret) {
-		mutex_unlock(&fs_info->scrub_lock);
-		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
-		goto out_free_ctx;
-	}
-
 	sctx->readonly = readonly;
 	dev->scrub_ctx = sctx;
 	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
@@ -3917,24 +3945,14 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 
 	mutex_lock(&fs_info->scrub_lock);
 	dev->scrub_ctx = NULL;
-	if (refcount_dec_and_test(&fs_info->scrub_workers_refcnt)) {
-		scrub_workers = fs_info->scrub_workers;
-		scrub_wr_comp = fs_info->scrub_wr_completion_workers;
-		scrub_parity = fs_info->scrub_parity_workers;
-
-		fs_info->scrub_workers = NULL;
-		fs_info->scrub_wr_completion_workers = NULL;
-		fs_info->scrub_parity_workers = NULL;
-	}
 	mutex_unlock(&fs_info->scrub_lock);
 
-	btrfs_destroy_workqueue(scrub_workers);
-	btrfs_destroy_workqueue(scrub_wr_comp);
-	btrfs_destroy_workqueue(scrub_parity);
+	scrub_workers_put(fs_info);
 	scrub_put_ctx(sctx);
 
 	return ret;
-
+out:
+	scrub_workers_put(fs_info);
 out_free_ctx:
 	scrub_free_ctx(sctx);
 

fs/btrfs/tree-checker.c

@@ -984,7 +984,7 @@ static int check_inode_item(struct extent_buffer *leaf,
 	/* Note for ROOT_TREE_DIR_ITEM, mkfs could set its transid 0 */
 	if (btrfs_inode_transid(leaf, iitem) > super_gen + 1) {
 		inode_item_err(leaf, slot,
-			"invalid inode generation: has %llu expect [0, %llu]",
+			"invalid inode transid: has %llu expect [0, %llu]",
 			       btrfs_inode_transid(leaf, iitem), super_gen + 1);
 		return -EUCLEAN;
 	}

fs/btrfs/volumes.c

@@ -4462,6 +4462,7 @@ int btrfs_uuid_scan_kthread(void *data)
 			goto skip;
 		}
 update_tree:
+		btrfs_release_path(path);
 		if (!btrfs_is_empty_uuid(root_item.uuid)) {
 			ret = btrfs_uuid_tree_add(trans, root_item.uuid,
 						  BTRFS_UUID_KEY_SUBVOL,
@@ -4486,14 +4487,14 @@ int btrfs_uuid_scan_kthread(void *data)
 		}
 
 skip:
+		btrfs_release_path(path);
 		if (trans) {
 			ret = btrfs_end_transaction(trans);
 			trans = NULL;
 			if (ret)
 				break;
 		}
 
-		btrfs_release_path(path);
 		if (key.offset < (u64)-1) {
 			key.offset++;
 		} else if (key.type < BTRFS_ROOT_ITEM_KEY) {