kernel: backport fix for non-regular inodes on f2fs

Upstream commit dda9f4b9ca ("f2fs: fix to skip verifying block address
for non-regular inode").

On 4.14, attempting to perform operations on a non-regular inode
residing on an f2fs filesystem, such rm-ing a device node, would fail
and lead to a warning / call trace in dmesg. This fix was already
applied to other kernels upstream - including 4.19, from which the patch
was taken.

More info at https://bugzilla.kernel.org/show_bug.cgi?id=202495.

Signed-off-by: Matt Merhar <mattmerhar@protonmail.com>
(cherry picked from commit ee500186a5)
openwrt-19.07
Matt Merhar 2020-04-19 17:12:03 -04:00 committed by Hauke Mehrtens
parent f40947a8c0
commit 0974d59b5f
1 changed files with 69 additions and 0 deletions

View File

@ -0,0 +1,69 @@
From dda9f4b9cac6bdd2a96253b4444d7a6ce5132edb Mon Sep 17 00:00:00 2001
From: Chao Yu <yuchao0@huawei.com>
Date: Sat, 11 Aug 2018 23:42:09 +0800
Subject: f2fs: fix to skip verifying block address for non-regular inode
generic/184 1s ... [failed, exit status 1]- output mismatch
--- tests/generic/184.out 2015-01-11 16:52:27.643681072 +0800
QA output created by 184 - silence is golden
+rm: cannot remove '/mnt/f2fs/null': Bad address
+mknod: '/mnt/f2fs/null': Bad address
+chmod: cannot access '/mnt/f2fs/null': Bad address
+./tests/generic/184: line 36: /mnt/f2fs/null: Bad address
...
F2FS-fs (zram0): access invalid blkaddr:259
EIP: f2fs_is_valid_blkaddr+0x14b/0x1b0 [f2fs]
f2fs_iget+0x927/0x1010 [f2fs]
f2fs_lookup+0x26e/0x630 [f2fs]
__lookup_slow+0xb3/0x140
lookup_slow+0x31/0x50
walk_component+0x185/0x1f0
path_lookupat+0x51/0x190
filename_lookup+0x7f/0x140
user_path_at_empty+0x36/0x40
vfs_statx+0x61/0xc0
__do_sys_stat64+0x29/0x40
sys_stat64+0x13/0x20
do_fast_syscall_32+0xaa/0x22c
entry_SYSENTER_32+0x53/0x86
In f2fs_iget(), we will check inode's first block address, if it is valid,
we will set FI_FIRST_BLOCK_WRITTEN flag in inode.
But we should only do this for regular inode, otherwise, like special
inode, i_addr[0] is used for storing device info instead of block address,
it will fail checking flow obviously.
So for non-regular inode, let's skip verifying address and setting flag.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
fs/f2fs/inode.c | 14 ++++++++------
1 file changed, 8 insertions(+), 6 deletions(-)
--- a/fs/f2fs/inode.c
+++ b/fs/f2fs/inode.c
@@ -310,13 +310,15 @@ static int do_read_inode(struct inode *i
/* get rdev by using inline_info */
__get_inode_rdev(inode, ri);
- err = __written_first_block(sbi, ri);
- if (err < 0) {
- f2fs_put_page(node_page, 1);
- return err;
+ if (S_ISREG(inode->i_mode)) {
+ err = __written_first_block(sbi, ri);
+ if (err < 0) {
+ f2fs_put_page(node_page, 1);
+ return err;
+ }
+ if (!err)
+ set_inode_flag(inode, FI_FIRST_BLOCK_WRITTEN);
}
- if (!err)
- set_inode_flag(inode, FI_FIRST_BLOCK_WRITTEN);
if (!need_inode_block_update(sbi, inode->i_ino))
fi->last_disk_size = inode->i_size;