Hi all,
Today's linux-next merge of the vfs tree got a conflict in:
fs/f2fs/namei.c
between commit:
53edb549565f ("f2fs: fix to avoid dirent corruption")
from the f2fs tree and commit:
7deee77b993a ("f2fs: Avoid reading renamed directory if parent does not change")
from the vfs tree.
I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging. You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.
--
Cheers,
Stephen Rothwell
diff --cc fs/f2fs/namei.c
index ede6afb81762,fdc97df6bb85..000000000000
--- a/fs/f2fs/namei.c
+++ b/fs/f2fs/namei.c
@@@ -1104,8 -1106,8 +1105,8 @@@ static int f2fs_rename(struct mnt_idma
iput(whiteout);
}
- if (old_dir_entry) {
- if (old_dir != new_dir)
+ if (old_is_dir) {
- if (old_dir_entry && !whiteout)
++ if (old_dir_entry)
f2fs_set_link(old_inode, old_dir_entry,
old_dir_page, new_dir);
else
Hi Al,
On Wed, 20 Dec 2023 10:47:34 +1100 Stephen Rothwell <[email protected]> wrote:
>
> Today's linux-next merge of the vfs tree got a conflict in:
>
> fs/f2fs/namei.c
>
> between commit:
>
> 53edb549565f ("f2fs: fix to avoid dirent corruption")
>
> from the f2fs tree and commit:
>
> 7deee77b993a ("f2fs: Avoid reading renamed directory if parent does not change")
>
> from the vfs tree.
>
> I fixed it up (see below) and can carry the fix as necessary. This
> is now fixed as far as linux-next is concerned, but any non trivial
> conflicts should be mentioned to your upstream maintainer when your tree
> is submitted for merging. You may also want to consider cooperating
> with the maintainer of the conflicting tree to minimise any particularly
> complex conflicts.
Then I remembered to look at your suggested resolution and redid it
like you did (see below).
--
Cheers,
Stephen Rothwell
--- a/fs/f2fs/namei.c
+++ b/fs/f2fs/namei.c
@@@ -1104,12 -1106,12 +1105,9 @@@ static int f2fs_rename(struct mnt_idma
iput(whiteout);
}
- if (old_is_dir) {
- if (old_dir_entry && !whiteout)
- f2fs_set_link(old_inode, old_dir_entry,
+ if (old_dir_entry) {
- if (old_dir != new_dir)
- f2fs_set_link(old_inode, old_dir_entry,
++ f2fs_set_link(old_inode, old_dir_entry,
old_dir_page, new_dir);
-- else
-- f2fs_put_page(old_dir_page, 0);
f2fs_i_links_write(old_dir, false);
}
if (F2FS_OPTION(sbi).fsync_mode == FSYNC_MODE_STRICT) {
On Wed, Dec 20, 2023 at 11:13:25AM +1100, Stephen Rothwell wrote:
> Hi Al,
>
> On Wed, 20 Dec 2023 10:47:34 +1100 Stephen Rothwell <[email protected]> wrote:
> >
> > Today's linux-next merge of the vfs tree got a conflict in:
> >
> > fs/f2fs/namei.c
> >
> > between commit:
> >
> > 53edb549565f ("f2fs: fix to avoid dirent corruption")
> >
> > from the f2fs tree and commit:
> >
> > 7deee77b993a ("f2fs: Avoid reading renamed directory if parent does not change")
> >
> > from the vfs tree.
> >
> > I fixed it up (see below) and can carry the fix as necessary. This
> > is now fixed as far as linux-next is concerned, but any non trivial
> > conflicts should be mentioned to your upstream maintainer when your tree
> > is submitted for merging. You may also want to consider cooperating
> > with the maintainer of the conflicting tree to minimise any particularly
> > complex conflicts.
>
> Then I remembered to look at your suggested resolution and redid it
> like you did (see below).
My suggested resolution had been wrong, actually - the way it's written,
link count drop should be conditional on old_is_dir, cross-directory or
not.
I think the right solution is
if (old_dir_entry)
f2fs_set_link(old_inode, old_dir_entry, old_dir_page, new_dir);
if (old_is_dir)
f2fs_i_links_write(old_dir, false);
Hi Al,
On Wed, 20 Dec 2023 01:34:02 +0000 Al Viro <[email protected]> wrote:
>
> My suggested resolution had been wrong, actually - the way it's written,
> link count drop should be conditional on old_is_dir, cross-directory or
> not.
>
> I think the right solution is
> if (old_dir_entry)
> f2fs_set_link(old_inode, old_dir_entry, old_dir_page, new_dir);
> if (old_is_dir)
> f2fs_i_links_write(old_dir, false);
Ok, I will fix it up.
--
Cheers,
Stephen Rothwell
On Wed, Dec 20, 2023 at 02:05:29PM +1100, Stephen Rothwell wrote:
> Hi Al,
>
> On Wed, 20 Dec 2023 01:34:02 +0000 Al Viro <[email protected]> wrote:
> >
> > My suggested resolution had been wrong, actually - the way it's written,
> > link count drop should be conditional on old_is_dir, cross-directory or
> > not.
> >
> > I think the right solution is
> > if (old_dir_entry)
> > f2fs_set_link(old_inode, old_dir_entry, old_dir_page, new_dir);
> > if (old_is_dir)
> > f2fs_i_links_write(old_dir, false);
>
> Ok, I will fix it up.
Backmerge with conflict resolution added to #for-next...