2021-03-12 06:53:33

by Shijie Luo

Subject: [PATCH] ext4: fix potential error in ext4_do_update_inode

If set_large_file = 1 and errors occur in ext4_handle_dirty_metadata(),
the error code will be overridden, go to out_brelse to avoid this
situation.

Signed-off-by: Shijie Luo <[email protected]>
---
fs/ext4/inode.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 650c5acd2f2d..8074ae0e976d 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5026,7 +5026,7 @@ static int ext4_do_update_inode(handle_t *handle,
struct ext4_inode_info *ei = EXT4_I(inode);
struct buffer_head *bh = iloc->bh;
struct super_block *sb = inode->i_sb;
- int err = 0, rc, block;
+ int err = 0, block;
int need_datasync = 0, set_large_file = 0;
uid_t i_uid;
gid_t i_gid;
@@ -5138,9 +5138,9 @@ static int ext4_do_update_inode(handle_t *handle,
bh->b_data);

BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
- rc = ext4_handle_dirty_metadata(handle, NULL, bh);
- if (!err)
- err = rc;
+ err = ext4_handle_dirty_metadata(handle, NULL, bh);
+ if (err)
+ goto out_brelse;
ext4_clear_inode_state(inode, EXT4_STATE_NEW);
if (set_large_file) {
BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get write access");
--
2.19.1

2021-03-12 12:52:45

by Jan Kara

Subject: Re: [PATCH] ext4: fix potential error in ext4_do_update_inode

On Fri 12-03-21 01:50:51, Shijie Luo wrote:
> If set_large_file = 1 and errors occur in ext4_handle_dirty_metadata(),
> the error code will be overridden, go to out_brelse to avoid this
> situation.
>
> Signed-off-by: Shijie Luo <[email protected]>

Yeah, looks good. Once ext4_handle_dirty_metadata() fails, the journal is
aborted anyway so we are unlikely to do anything useful with the
filesystem. Feel free to add:

Reviewed-by: Jan Kara <[email protected]>

Honza

> ---
> fs/ext4/inode.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index 650c5acd2f2d..8074ae0e976d 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -5026,7 +5026,7 @@ static int ext4_do_update_inode(handle_t *handle,
> struct ext4_inode_info *ei = EXT4_I(inode);
> struct buffer_head *bh = iloc->bh;
> struct super_block *sb = inode->i_sb;
> - int err = 0, rc, block;
> + int err = 0, block;
> int need_datasync = 0, set_large_file = 0;
> uid_t i_uid;
> gid_t i_gid;
> @@ -5138,9 +5138,9 @@ static int ext4_do_update_inode(handle_t *handle,
> bh->b_data);
>
> BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
> - rc = ext4_handle_dirty_metadata(handle, NULL, bh);
> - if (!err)
> - err = rc;
> + err = ext4_handle_dirty_metadata(handle, NULL, bh);
> + if (err)
> + goto out_brelse;
> ext4_clear_inode_state(inode, EXT4_STATE_NEW);
> if (set_large_file) {
> BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get write access");
> --
> 2.19.1
>
--
Jan Kara <[email protected]>
SUSE Labs, CR

2021-03-21 04:27:48

by Theodore Ts'o

Subject: Re: [PATCH] ext4: fix potential error in ext4_do_update_inode

On Fri, Mar 12, 2021 at 01:50:55PM +0100, Jan Kara wrote:
> On Fri 12-03-21 01:50:51, Shijie Luo wrote:
> > If set_large_file = 1 and errors occur in ext4_handle_dirty_metadata(),
> > the error code will be overridden, go to out_brelse to avoid this
> > situation.
> >
> > Signed-off-by: Shijie Luo <[email protected]>
>
> Yeah, looks good. Once ext4_handle_dirty_metadata() fails, the journal is
> aborted anyway so we are unlikely to do anything useful with the
> filesystem. Feel free to add:
>
> Reviewed-by: Jan Kara <[email protected]>

Thanks, applied.

- Ted