[PATCH] nilfs2: fix kernel bug on rename operation of broken directory

[Ryusuke Konishi <konishi.ryusuke@gmail.com>]

Syzbot reported that in rename directory operation on broken directory
on nilfs2, __block_write_begin_int() called to prepare block write may
fail BUG_ON check for access exceeding the folio/page size.

This is because nilfs_dotdot(), which gets parent directory reference
entry ("..") of the directory to be moved or renamed, does not check
consistency enough, and may return location exceeding folio/page size
for broken directories.

Fix this issue by checking required directory entries ("." and "..")
in the first chunk of the directory in nilfs_dotdot().

Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Reported-by: syzbot+d3abed1ad3d367fa2627@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=d3abed1ad3d367fa2627
Fixes: 2ba466d74ed7 ("nilfs2: directory entry operations")
Tested-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Cc: stable@vger.kernel.org
---
Hi Andrew, please apply this as a bug fix.

This fixes a potential kernel bug reported by syzbot regarding broken
directory rename operations.

Thanks,
Ryusuke Konishi

 fs/nilfs2/dir.c | 32 ++++++++++++++++++++++++++++++--
 1 file changed, 30 insertions(+), 2 deletions(-)

diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c
index dddfa604491a..4a29b0138d75 100644
--- a/fs/nilfs2/dir.c
+++ b/fs/nilfs2/dir.c
@@ -383,11 +383,39 @@ struct nilfs_dir_entry *nilfs_find_entry(struct inode *dir,
 
 struct nilfs_dir_entry *nilfs_dotdot(struct inode *dir, struct folio **foliop)
 {
-	struct nilfs_dir_entry *de = nilfs_get_folio(dir, 0, foliop);
+	struct folio *folio;
+	struct nilfs_dir_entry *de, *next_de;
+	size_t limit;
+	char *msg;
 
+	de = nilfs_get_folio(dir, 0, &folio);
 	if (IS_ERR(de))
 		return NULL;
-	return nilfs_next_entry(de);
+
+	limit = nilfs_last_byte(dir, 0);  /* is a multiple of chunk size */
+	if (unlikely(!limit || le64_to_cpu(de->inode) != dir->i_ino ||
+		     !nilfs_match(1, ".", de))) {
+		msg = "missing '.'";
+		goto fail;
+	}
+
+	next_de = nilfs_next_entry(de);
+	/*
+	 * If "next_de" has not reached the end of the chunk, there is
+	 * at least one more record.  Check whether it matches "..".
+	 */
+	if (unlikely((char *)next_de == (char *)de + nilfs_chunk_size(dir) ||
+		     !nilfs_match(2, "..", next_de))) {
+		msg = "missing '..'";
+		goto fail;
+	}
+	*foliop = folio;
+	return next_de;
+
+fail:
+	nilfs_error(dir->i_sb, "directory #%lu %s", dir->i_ino, msg);
+	folio_release_kmap(folio, de);
+	return NULL;
 }
 
 ino_t nilfs_inode_by_name(struct inode *dir, const struct qstr *qstr)
-- 
2.34.1