On Thu, 12 Jun 2025, Miklos Szeredi wrote: > On Thu, 12 Jun 2025 at 01:38, Al Viro <viro@xxxxxxxxxxxxxxxxxx> wrote: > > > Umm... No objections, as long as overlayfs part is correct; it seems > > to be, but I hadn't checked every chunk there... > > Overlayfs parts looks okay too. > > A followup would be nice (e.g. make ovl_cleanup() take a dentry for > the directory as well, etc) so that there's no need to have local > variables for both the inode and dentry of the directory. I am planning some followups and will include that in them. I'll also be sure to test with fs-tests after consulting README.overlay as you suggest elsewhere - thanks. NeilBrown