btrfs: always log symlinks in full mode

commit d0e64a9 upstream.

On Linux, empty symlinks are invalid, and attempting to create one with
the system call symlink(2) results in an -ENOENT error and this is
explicitly documented in the man page.

If we rename a symlink that was created in the current transaction and its
parent directory was logged before, we actually end up logging the symlink
without logging its content, which is stored in an inline extent. That
means that after a power failure we can end up with an empty symlink,
having no content and an i_size of 0 bytes.

It can be easily reproduced like this:

  $ mkfs.btrfs -f /dev/sdc
  $ mount /dev/sdc /mnt

  $ mkdir /mnt/testdir
  $ sync

  # Create a file inside the directory and fsync the directory.
  $ touch /mnt/testdir/foo
  $ xfs_io -c "fsync" /mnt/testdir

  # Create a symlink inside the directory and then rename the symlink.
  $ ln -s /mnt/testdir/foo /mnt/testdir/bar
  $ mv /mnt/testdir/bar /mnt/testdir/baz

  # Now fsync again the directory, this persist the log tree.
  $ xfs_io -c "fsync" /mnt/testdir

  <power failure>

  $ mount /dev/sdc /mnt
  $ stat -c %s /mnt/testdir/baz
  0
  $ readlink /mnt/testdir/baz
  $

Fix this by always logging symlinks in full mode (LOG_INODE_ALL), so that
their content is also logged.

A test case for fstests will follow.

CC: stable@vger.kernel.org # 4.9+
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

fs/btrfs/tree-log.c

@@ -5483,6 +5483,18 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 		mutex_lock(&inode->log_mutex);
 	}
 
+	/*
+	 * For symlinks, we must always log their content, which is stored in an
+	 * inline extent, otherwise we could end up with an empty symlink after
+	 * log replay, which is invalid on linux (symlink(2) returns -ENOENT if
+	 * one attempts to create an empty symlink).
+	 * We don't need to worry about flushing delalloc, because when we create
+	 * the inline extent when the symlink is created (we never have delalloc
+	 * for symlinks).
+	 */
+	if (S_ISLNK(inode->vfs_inode.i_mode))
+		inode_only = LOG_INODE_ALL;
+
 	/*
 	 * This is for cases where logging a directory could result in losing a
 	 * a file after replaying the log. For example, if we move a file from a
@@ -5853,7 +5865,7 @@ static int log_new_dir_dentries(struct btrfs_trans_handle *trans,
 			}
 
 			ctx->log_new_dentries = false;
-			if (type == BTRFS_FT_DIR || type == BTRFS_FT_SYMLINK)
+			if (type == BTRFS_FT_DIR)
 				log_mode = LOG_INODE_ALL;
 			ret = btrfs_log_inode(trans, root, BTRFS_I(di_inode),
 					      log_mode, ctx);