NFSD: nfsd_unlink() clobbers non-zero status returned from fh_fill_pre_attrs()

commit d7d8e31 upstream.

If fh_fill_pre_attrs() returns a non-zero status, the error flow
takes it through out_unlock, which then overwrites the returned
status code with

	err = nfserrno(host_err);

Fixes: a332018 ("nfsd: handle failure to collect pre/post-op attrs more sanely")
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Author: chucklever

fs/nfsd/vfs.c

@@ -2011,11 +2011,9 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 		 * error status.
 		 */
 		err = nfserr_file_open;
-	} else {
-		err = nfserrno(host_err);
 	}
 out:
-	return err;
+	return err != nfs_ok ? err : nfserrno(host_err);
 out_unlock:
 	inode_unlock(dirp);
 	goto out_drop_write;