Re: Encrypted File systems implementation into the kernel?

From: Michael H. Warfield (mhw@wittsend.com)
Date: Sun Feb 06 2000 - 13:34:44 EST


On Sun, Feb 06, 2000 at 07:24:42PM +0100, Marc Mutz wrote:
> "Michael H. Warfield" wrote:

        [...]

> > What algorithms do we need?
> > DES, 3DES, Idea, BlowFish...

> Renove DES! Add Serpent and the (five?) finalists of the AES
> competition.

        Note for the unwashed... You technically can not remove DES without
removing 3DES. 3DES is three passes through the DES algorithm with either
two keys (112 bit EDE mode) or three keys. You still have the DES algo
at it's heart. We may removed external ACCESS to the DES algo, but it
still has to be there.

> > Do we need public key algorithms?

> I think that is application-specific and uncritical w.r.t speed and
> therefore belongs to userspace.

        Unless we have PK signed modules. Then it gets real iffy.

> > What are the initial cryptograhic service consumers?
> > CIPE, IPSec, loopback crypto...

> Crypted swap with hidden key generated on boot-up from /dev/random? IIRC
> HPA was talking about such.

> > Do we provide access from user level programs to kernel level
> > algorithms?

> Wouldn't the userspace-to-kernel-to-userspace transitions be too great a
> penalty w.r.t speed? Maybe making the CryptoAPI a library would help.

        Perhaps. Perhaps not. It would provide transparent access to
crypto cards and hardware accelerators.

> > How to we coordinate with the different projects to get them
> > to use the integrated crypto?

> Let the implementations of ciphers for the CryptoAPI get a _lot_ of
> peer-reviewing by crypto-professors and they'll run to use these
> routines.

> <snip>
> > If there is sufficient interest and discussion, we may want to set
> > up a separate mailing list just for common crypto issues and development.
> > I would be willing to host such a list, but I want to be sure there is
> > sufficient interest and agreement before going to that trouble. If the
> > threads remain managable and well identified, we may keep it here on
> > linux-kernel unless we get too many complaints.
> >
> <snip>

> As it seems that a linux-crypto ml does not exist yet:

> On Tue, 16 Nov 1999 Rik van Riel wrote:
> >
> > On Tue, 16 Nov 1999, Marc Mutz wrote:
> >
> > > As this is the official mailing list for discussion concerning the
> > > kerneli patches (maybe there should be a linux-crypto ml, please
> > > tell me if it exists):
> >
> > If it doesn't exist yet, you're welcome to host it at
> > nl.linux.org...
> >
> <snip>

> I had not the time to let this happen. Michael, maybe you can talk to
> Rik, if you are interested in doing the hosting.

        I will follow up on it. Thanks for that reference!

> Marc

> --
> Marc Mutz <Marc@Mutz.com> http://marc.mutz.com/Encryption-HOWTO/
> University of Bielefeld, Dep. of Mathematics / Dep. of Physics

> PGP-keyID's: 0xd46ce9ab (RSA), 0x7ae55b9e (DSS/DH)

        Mike

-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 331-2437   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon Feb 07 2000 - 21:00:13 EST