On Wednesday, 20 October, 2010, Sage Weil wrote: > On Mon, 18 Oct 2010, Goffredo Baroncelli wrote: > > > Hi all > > > > like my previous patch, this one allow to remove a subvolume by an ordinary > > user. Instead of adding this capability to the rmdir(2) syscall, I update the > > BTRFS_IOC_SNAP_DESTROY ioctl, relaxing the rules to be execute. > > The checks are the ones performed by the rmdir(2) syscall. So a > > subvolume must be empty to be removed by a non-root user. I think that this > > increases a lot the usefulness of the snapshot/subvolume. > > > > It is possible to pull the code from the branch named "rm-subvolume-not- root" > > of the following repository: > > > > http://cassiopea.homelinux.net/git/btrfs-unstable.git > > > > Comments are welcome. > > This looks okay to me. I posted a similar patch a while back[1], but > didn't want to duplicate the check_sticky and may_delete code and > implemented a simpler set of checks instead. The full checks are probably > a better route, although it would be nice if we could avoid duplicating > the VFS checks in the process. Whether those helpers should be exported > is someone else's call, though. (The only other may_ functions that are > exported are may_umount and may_umount_tree.) I agree about the may_* function. But also there is the case of the may_create.. Anyway I want to highlight that the main differences between our patches is the fact that may patches needed the subvolume to be empty. So I skip all the problem related to removing a "not owned directory - not empty directory" > > sage > > [1] http://marc.info/?l=linux-btrfs&m=128086492512628&w=2 -- gpg key@ keyserver.linux.it: Goffredo Baroncelli (ghigo) <kreijack@xxxxxxxxx> Key fingerprint = 4769 7E51 5293 D36C 814E C054 BF04 F161 3DC5 0512 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
