Followup to: <20000601235558.A188@bug.ucw.cz>
By author: Pavel Machek <email@example.com>
In newsgroup: linux.dev.kernel
> > > That will surely confuse hell out of the backup utility. It is
> > > probably going to confuse autoclean.
> > And you allow your backups freely span the filesystem boundaries? I've
> > learnt not to do it _really_ hard way. Even harder was the autocleaner -
> > ever went "uh, /mnt is already busy, let's mount it for an hour under
> > /tmp/foo... Oh, FSCK! Let's hope these tapes are alive and there goes
> > my fscking weekend..."?
> It is bad to allow backups over fs boundaries; but it is harmless if
> you do not do "mistake". Oops. It used to be harmless. Now it is
> security hole. If you *really* want to have that mount in 2.4.0, let
> me know, and I'll make nice bugtraq announcement. Then we'll see how
> many holes this added.
It's worse than that. fs boundaries are typically detected by st_dev
changing, but that is not necessarily the case if your bind goes from
one part of one filesystem to another. Suddenly your bind is
undetectable to all existing software.
-- <firstname.lastname@example.org> at work, <email@example.com> in private! "Unix gives you enough rope to shoot yourself in the foot."
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to firstname.lastname@example.org Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Jun 07 2000 - 21:00:14 EST