Re: [dm-crypt] DM-Crypt resistance against Cold Boot Attacks

From: Arno Wagner <arno@wagner.name>
To: dm-crypt@saout.de
Subject: Re: [dm-crypt] DM-Crypt resistance against Cold Boot Attacks
Date: Thu, 19 May 2011 08:01:05 +0200	[thread overview]
Message-ID: <20110519060105.GA15966@tansi.org> (raw)
In-Reply-To: <BANLkTimC2YqOwWV2-f1dN96YYG-xz5RHMg@mail.gmail.com>

I have severe doubts as to its practical usefullness and 
security. It blocks CPU registers. That means if you 
want to do two ciphers or even only two keys, you may
be out of luck or (worse) insecure as the registers
get dumped somewhere. I also have my doubts whether these
registers really do not get read and stored to memory.

Without reading the paper in detail, I think this is a 
specialized tool for a scenario where you need to prevent
exactly one cipher with exactly one key from being written 
to RAM. That does not scale at all and may cause problems 
in other places. 

It looks like a neat bit of research though.

Arno

On Thu, May 19, 2011 at 01:36:02AM +0000, Kraktus wrote:
> Looks like, in theory at least, it could be use be used for other
> encryption algorithms too, like Twofish or Serpent.
> 
> I hope the hardware specifics would still be practical with Twofish.
> From what I've read, Serpent and Ridjindael are already broken.
> http://www.schneier.com/crypto-gram-0209.html#1
> http://www.schneier.com/blog/archives/2009/07/new_attack_on_a.html
> http://www.schneier.com/blog/archives/2009/07/another_new_aes.html
> 
> On 18 May 2011 22:03, Arno Wagner <arno@wagner.name> wrote:
> > From the paper, TRESOR is a Linux kernel patch, hence not something
> > that would be added to cryptsetup, but the kernel.
> >
> > Arno
> >
> > On Wed, May 18, 2011 at 03:24:17PM +0200, Philipp Deppenwiese wrote:
> >> Hi,
> >>
> >> a guy from a germany university impelemented an AES-NI with resistance
> >> against Cold boot attacks and timing attacks. Maybe it's usefule to
> >> modify the dm-crypt code.
> >>
> >> http://www1.informatik.uni-erlangen.de/tresor/
> >>
> >> Greetz Philipp
> >>
> >> --
> >>
> >>
> >> ?? ?? ?? ?? ??Philipp Deppenwiese<Scientific
> >> ?? ?? ?? ?? ??Assistant> Institute for Internet Security -
> >> ?? ?? ?? ?? ??if(is) University of Applied Sciences
> >> ?? ?? ?? ?? ??Gelsenkirchen
> >> ?? ?? ?? ?? ??https://www.internet-sicherheit.de
> >>
> >>
> >> GPG Fingerprint:
> >>
> >> 64E4 0F3D E2EC B70D 7468 ??6356 3FBA F0E7 D814 27AB
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >
> >
> >
> >> _______________________________________________
> >> dm-crypt mailing list
> >> dm-crypt@saout.de
> >> http://www.saout.de/mailman/listinfo/dm-crypt
> >
> >
> > --
> > Arno Wagner, Dr. sc. techn., Dipl. Inform., CISSP -- Email: arno@wagner.name
> > GnuPG: ??ID: 1E25338F ??FP: 0C30 5782 9D93 F785 E79C ??0296 797F 6B50 1E25 338F
> > ----
> > Cuddly UI's are the manifestation of wishful thinking. -- Dylan Evans
> >
> > If it's in the news, don't worry about it. ??The very definition of
> > "news" is "something that hardly ever happens." -- Bruce Schneier
> > _______________________________________________
> > dm-crypt mailing list
> > dm-crypt@saout.de
> > http://www.saout.de/mailman/listinfo/dm-crypt
> >
> _______________________________________________
> dm-crypt mailing list
> dm-crypt@saout.de
> http://www.saout.de/mailman/listinfo/dm-crypt

-- 
Arno Wagner, Dr. sc. techn., Dipl. Inform., CISSP -- Email: arno@wagner.name 
GnuPG:  ID: 1E25338F  FP: 0C30 5782 9D93 F785 E79C  0296 797F 6B50 1E25 338F
----
Cuddly UI's are the manifestation of wishful thinking. -- Dylan Evans

If it's in the news, don't worry about it.  The very definition of 
"news" is "something that hardly ever happens." -- Bruce Schneier 