nilfs2: reject filesystem with unsupported block size

This inserts sanity check that refuses to mount a filesystem with
unsupported block size.

Previously, kernel code of nilfs was looking only limitation of
devices though mkfs.nilfs2 limits the range of block sizes; there was
no check that prevents rec_len overflow with larger block sizes.

With this change, block sizes larger than 64KB or smaller than 1KB
will get rejected explicitly by kernel.

Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
diff --git a/include/linux/nilfs2_fs.h b/include/linux/nilfs2_fs.h
index 970828a..f5487b6 100644
--- a/include/linux/nilfs2_fs.h
+++ b/include/linux/nilfs2_fs.h
@@ -287,6 +287,12 @@
 #define NILFS_NAME_LEN 255
 
 /*
+ * Block size limitations
+ */
+#define NILFS_MIN_BLOCK_SIZE		1024
+#define NILFS_MAX_BLOCK_SIZE		65536
+
+/*
  * The new version of the directory entry.  Since V0 structures are
  * stored in intel byte order, and the name_len field could never be
  * bigger than 255 chars, it's safe to reclaim the extra byte for the