ext4: avoid declaring fs inconsistent due to invalid file handles
authorTheodore Ts'o <tytso@mit.edu>
Wed, 19 Dec 2018 17:29:13 +0000 (12:29 -0500)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 9 Jan 2019 16:38:43 +0000 (17:38 +0100)
commit263663888d2f00421887e251d93ebc52650a37ab
tree0a7c6b51e2ba29cf3cc0f71f2ed38b3146b4b4c2
parent6633fcb231a0864ca5c961a7664b137474ca3779
ext4: avoid declaring fs inconsistent due to invalid file handles

commit 8a363970d1dc38c4ec4ad575c862f776f468d057 upstream.

If we receive a file handle, either from NFS or open_by_handle_at(2),
and it points at an inode which has not been initialized, and the file
system has metadata checksums enabled, we shouldn't try to get the
inode, discover the checksum is invalid, and then declare the file
system as being inconsistent.

This can be reproduced by creating a test file system via "mke2fs -t
ext4 -O metadata_csum /tmp/foo.img 8M", mounting it, cd'ing into that
directory, and then running the following program.

#define _GNU_SOURCE
#include <fcntl.h>

struct handle {
struct file_handle fh;
unsigned char fid[MAX_HANDLE_SZ];
};

int main(int argc, char **argv)
{
struct handle h = {{8, 1 }, { 12, }};

open_by_handle_at(AT_FDCWD, &h.fh, O_RDONLY);
return 0;
}

Google-Bug-Id: 120690101
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/ext4/ext4.h
fs/ext4/ialloc.c
fs/ext4/inode.c
fs/ext4/ioctl.c
fs/ext4/namei.c
fs/ext4/resize.c
fs/ext4/super.c
fs/ext4/xattr.c