[dm-crypt] LUKS and backdoors

Arno Wagner arno at wagner.name
Mon Oct 21 19:50:20 CEST 2013


On Mon, Oct 21, 2013 at 03:12:25PM +0200, Christoph Anton Mitterer wrote:
> On Mon, 2013-10-21 at 13:10 +0200, octane indice wrote:
> > But at this point, what is the quality of the random[1]?
> Well /dev/random (in Linux) should have either high quality entropy,...
> or block... at least that was my understanding (there's currently a
> discussion going on about /dev/[u]random at the well known cryptography
> mailing list)...

That understanding is unfortunately incorrect. True, /dev/random
will block if it thinks it is low on entropy, but it cannot 
reliably know. The problem is that it needs to estimate entropy 
for each event it adds, and that can go horribly wrong. In
particular virtualized anvironments can have that effect.
 
> BUT,... cryptsetup uses by default unfortunately urandom to generate the
> master key.
> I never really understood why since all arguments pro it seem weak or
> nonsense to me... anyway that's how things are.
> But you can use --use-random to change that.

Look at the mailing-list archives. It boild down to the simple
fact that if you do crypto in a low-entropy environment, you
better know what you are doing and that the splution to dealing
with such an environment is _not_ obvious and not as simple
as selecting /dev/random.

> So in principle you should be on the safe side then.
> 
> 
> Of course you can improve entropy by using stuff like haveged, or a
> TRNG[0],... but I do not really know wheter these also have a positive
> effect on the _quality_ of the entropy or just on the _quantity_.

There is no difference between quality and quantity when entropy
is concerned.

Arno

> 
> Cheers,
> Chris.
> 
> 
> [0] According to Ted Ts'o and others it's not possible to
> spoil /dev/random by seeding it with malicious entropy sources (it just
> wouldn't get better as it was already)... though I must admit that I've
> never understood why this could be like that.



> _______________________________________________
> dm-crypt mailing list
> dm-crypt at saout.de
> http://www.saout.de/mailman/listinfo/dm-crypt


-- 
Arno Wagner,     Dr. sc. techn., Dipl. Inform.,    Email: arno at wagner.name
GnuPG: ID: CB5D9718  FP: 12D6 C03B 1B30 33BB 13CF  B774 E35C 5FA1 CB5D 9718
----
There are two ways of constructing a software design: One way is to make it
so simple that there are obviously no deficiencies, and the other way is to
make it so complicated that there are no obvious deficiencies. The first
method is far more difficult.  --Tony Hoare


More information about the dm-crypt mailing list