On Wed, 9 Apr 2025 at 10:25, Amir Goldstein <amir73il@xxxxxxxxx> wrote: > On second thought, if unpriv user suppresses ovl_set_redirect() > by setting some mock redirect value on index maybe that lead to some > risk. Not worth overthinking about it. > > Attached patch removed next* variables without this compromise. > > Tested it squashed to patch 1 and minor rebase conflicts fixes in patch 2. > It passed your tests. Thanks. One more change: in this patch we just want the consistency fix, not the behavior change introduced in 2/3. So move the ovl_check_follow_redirect() to before the lazy-data check here and restore the order in the next patch. Pushed to overlayfs/vfs.git#overlayfs-next Thanks, Miklos
