Merging upstream version 4.19.304.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

11 files changed, 64 insertions, 27 deletions

diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 34d56f0fa..e48c6d7a8 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -3590,6 +3590,8 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 
 	ret = tree_mod_log_eb_copy(fs_info, split, c, 0, mid, c_nritems - mid);
 	if (ret) {
+		btrfs_tree_unlock(split);
+		free_extent_buffer(split);
 		btrfs_abort_transaction(trans, ret);
 		return ret;
 	}
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index 1fbe2dee1..469a90b07 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -1171,20 +1171,33 @@ static int __btrfs_run_delayed_items(struct btrfs_trans_handle *trans, int nr)
 		ret = __btrfs_commit_inode_delayed_items(trans, path,
 							 curr_node);
 		if (ret) {
-			btrfs_release_delayed_node(curr_node);
-			curr_node = NULL;
 			btrfs_abort_transaction(trans, ret);
 			break;
 		}
 
 		prev_node = curr_node;
 		curr_node = btrfs_next_delayed_node(curr_node);
+		/*
+		 * See the comment below about releasing path before releasing
+		 * node. If the commit of delayed items was successful the path
+		 * should always be released, but in case of an error, it may
+		 * point to locked extent buffers (a leaf at the very least).
+		 */
+		ASSERT(path->nodes[0] == NULL);
 		btrfs_release_delayed_node(prev_node);
 	}
 
+	/*
+	 * Release the path to avoid a potential deadlock and lockdep splat when
+	 * releasing the delayed node, as that requires taking the delayed node's
+	 * mutex. If another task starts running delayed items before we take
+	 * the mutex, it will first lock the mutex and then it may try to lock
+	 * the same btree path (leaf).
+	 */
+	btrfs_free_path(path);
+
 	if (curr_node)
 		btrfs_release_delayed_node(curr_node);
-	btrfs_free_path(path);
 	trans->block_rsv = block_rsv;
 
 	return ret;
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index d71f800e8..902ab00bf 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -1549,6 +1549,11 @@ again:
 		err = -ENOENT;
 		goto out;
 	} else if (WARN_ON(ret)) {
+		btrfs_print_leaf(path->nodes[0]);
+		btrfs_err(fs_info,
+"extent item not found for insert, bytenr %llu num_bytes %llu parent %llu root_objectid %llu owner %llu offset %llu",
+			  bytenr, num_bytes, parent, root_objectid, owner,
+			  offset);
 		err = -EIO;
 		goto out;
 	}
@@ -2322,12 +2327,12 @@ static int run_delayed_tree_ref(struct btrfs_trans_handle *trans,
 		parent = ref->parent;
 	ref_root = ref->root;
 
-	if (node->ref_mod != 1) {
+	if (unlikely(node->ref_mod != 1)) {
 		btrfs_err(trans->fs_info,
-	"btree block(%llu) has %d references rather than 1: action %d ref_root %llu parent %llu",
+	"btree block %llu has %d references rather than 1: action %d ref_root %llu parent %llu",
 			  node->bytenr, node->ref_mod, node->action, ref_root,
 			  parent);
-		return -EIO;
+		return -EUCLEAN;
 	}
 	if (node->action == BTRFS_ADD_DELAYED_REF && insert_reserved) {
 		BUG_ON(!extent_op || !extent_op->update_flags);
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index dabf15384..504d63fb8 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -3928,11 +3928,12 @@ retry:
 			free_extent_buffer(eb);
 
 			/*
-			 * the filesystem may choose to bump up nr_to_write.
+			 * The filesystem may choose to bump up nr_to_write.
 			 * We have to make sure to honor the new nr_to_write
-			 * at any time
+			 * at any time.
 			 */
-			nr_to_write_done = wbc->nr_to_write <= 0;
+			nr_to_write_done = (wbc->sync_mode == WB_SYNC_NONE &&
+					    wbc->nr_to_write <= 0);
 		}
 		pagevec_release(&pvec);
 		cond_resched();
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index f009d585e..c76277ccf 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -1789,6 +1789,15 @@ static noinline int btrfs_ioctl_snap_create_transid(struct file *file,
 			 * are limited to own subvolumes only
 			 */
 			ret = -EPERM;
+		} else if (btrfs_ino(BTRFS_I(src_inode)) != BTRFS_FIRST_FREE_OBJECTID) {
+			/*
+			 * Snapshots must be made with the src_inode referring
+			 * to the subvolume inode, otherwise the permission
+			 * checking above is useless because we may have
+			 * permission on a lower directory but not the subvol
+			 * itself.
+			 */
+			ret = -EINVAL;
 		} else {
 			ret = btrfs_mksubvol(&file->f_path, name, namelen,
 					     BTRFS_I(src_inode)->root,
@@ -2045,7 +2054,7 @@ static noinline int key_in_sk(struct btrfs_key *key,
 static noinline int copy_to_sk(struct btrfs_path *path,
 			       struct btrfs_key *key,
 			       struct btrfs_ioctl_search_key *sk,
-			       size_t *buf_size,
+			       u64 *buf_size,
 			       char __user *ubuf,
 			       unsigned long *sk_offset,
 			       int *num_found)
@@ -2177,7 +2186,7 @@ out:
 
 static noinline int search_ioctl(struct inode *inode,
 				 struct btrfs_ioctl_search_key *sk,
-				 size_t *buf_size,
+				 u64 *buf_size,
 				 char __user *ubuf)
 {
 	struct btrfs_fs_info *info = btrfs_sb(inode->i_sb);
@@ -2249,7 +2258,7 @@ static noinline int btrfs_ioctl_tree_search(struct file *file,
 	struct btrfs_ioctl_search_key sk;
 	struct inode *inode;
 	int ret;
-	size_t buf_size;
+	u64 buf_size;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
@@ -2283,8 +2292,8 @@ static noinline int btrfs_ioctl_tree_search_v2(struct file *file,
 	struct btrfs_ioctl_search_args_v2 args;
 	struct inode *inode;
 	int ret;
-	size_t buf_size;
-	const size_t buf_limit = SZ_16M;
+	u64 buf_size;
+	const u64 buf_limit = SZ_16M;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
@@ -4526,7 +4535,7 @@ static void get_block_group_info(struct list_head *groups_list,
 static long btrfs_ioctl_space_info(struct btrfs_fs_info *fs_info,
 				   void __user *arg)
 {
-	struct btrfs_ioctl_space_args space_args;
+	struct btrfs_ioctl_space_args space_args = { 0 };
 	struct btrfs_ioctl_space_info space;
 	struct btrfs_ioctl_space_info *dest;
 	struct btrfs_ioctl_space_info *dest_orig;
@@ -5884,7 +5893,7 @@ static int _btrfs_ioctl_send(struct file *file, void __user *argp, bool compat)
 
 	if (compat) {
 #if defined(CONFIG_64BIT) && defined(CONFIG_COMPAT)
-		struct btrfs_ioctl_send_args_32 args32;
+		struct btrfs_ioctl_send_args_32 args32 = { 0 };
 
 		ret = copy_from_user(&args32, argp, sizeof(args32));
 		if (ret)
diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
index 69b43df18..ef95525fa 100644
--- a/fs/btrfs/qgroup.c
+++ b/fs/btrfs/qgroup.c
@@ -1115,7 +1115,9 @@ int btrfs_quota_disable(struct btrfs_fs_info *fs_info)
 		goto end_trans;
 	}
 
+	spin_lock(&fs_info->trans_lock);
 	list_del(&quota_root->dirty_list);
+	spin_unlock(&fs_info->trans_lock);
 
 	btrfs_tree_lock(quota_root->node);
 	clean_tree_block(fs_info, quota_root->node);
diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
index 1f535cd99..cb584c044 100644
--- a/fs/btrfs/send.c
+++ b/fs/btrfs/send.c
@@ -6860,7 +6860,7 @@ long btrfs_ioctl_send(struct file *mnt_file, struct btrfs_ioctl_send_args *arg)
 	sctx->flags = arg->flags;
 
 	sctx->send_filp = fget(arg->send_fd);
-	if (!sctx->send_filp) {
+	if (!sctx->send_filp || !(sctx->send_filp->f_mode & FMODE_WRITE)) {
 		ret = -EBADF;
 		goto out;
 	}
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 521f6c209..a59543951 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -2196,7 +2196,7 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf)
 	 * calculated f_bavail.
 	 */
 	if (!mixed && block_rsv->space_info->full &&
-	    total_free_meta - thresh < block_rsv->size)
+	    (total_free_meta < thresh || total_free_meta - thresh < block_rsv->size))
 		buf->f_bavail = 0;
 
 	buf->f_type = BTRFS_SUPER_MAGIC;
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 049535115..a34c0436e 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -200,10 +200,11 @@ loop:
 	spin_unlock(&fs_info->trans_lock);
 
 	/*
-	 * If we are ATTACH, we just want to catch the current transaction,
-	 * and commit it. If there is no transaction, just return ENOENT.
+	 * If we are ATTACH or TRANS_JOIN_NOSTART, we just want to catch the
+	 * current transaction, and commit it. If there is no transaction, just
+	 * return ENOENT.
 	 */
-	if (type == TRANS_ATTACH)
+	if (type == TRANS_ATTACH || type == TRANS_JOIN_NOSTART)
 		return -ENOENT;
 
 	/*
@@ -703,8 +704,13 @@ btrfs_attach_transaction_barrier(struct btrfs_root *root)
 
 	trans = start_transaction(root, 0, TRANS_ATTACH,
 				  BTRFS_RESERVE_NO_FLUSH, true);
-	if (trans == ERR_PTR(-ENOENT))
-		btrfs_wait_for_commit(root->fs_info, 0);
+	if (trans == ERR_PTR(-ENOENT)) {
+		int ret;
+
+		ret = btrfs_wait_for_commit(root->fs_info, 0);
+		if (ret)
+			return ERR_PTR(ret);
+	}
 
 	return trans;
 }
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 0fe32c567..23ec766ee 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -4236,7 +4236,7 @@ static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans,
 	struct extent_buffer *leaf;
 	int slot;
 	int ins_nr = 0;
-	int start_slot;
+	int start_slot = 0;
 	int ret;
 
 	if (!(inode->flags & BTRFS_INODE_PREALLOC))
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 0294f519c..ceced5e56 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2760,7 +2760,7 @@ static struct extent_map *get_chunk_map(struct btrfs_fs_info *fs_info,
 		return ERR_PTR(-EINVAL);
 	}
 
-	if (em->start > logical || em->start + em->len < logical) {
+	if (em->start > logical || em->start + em->len <= logical) {
 		btrfs_crit(fs_info,
 			   "found a bad mapping, wanted %llu-%llu, found %llu-%llu",
 			   logical, length, em->start, em->start + em->len);
@@ -4106,8 +4106,7 @@ int btrfs_cancel_balance(struct btrfs_fs_info *fs_info)
 		}
 	}
 
-	BUG_ON(fs_info->balance_ctl ||
-		test_bit(BTRFS_FS_BALANCE_RUNNING, &fs_info->flags));
+	ASSERT(!test_bit(BTRFS_FS_BALANCE_RUNNING, &fs_info->flags));
 	atomic_dec(&fs_info->balance_cancel_req);
 	mutex_unlock(&fs_info->balance_mutex);
 	return 0;