Re: 6man w.g. last call for <draft-ietf-6man-default-iids-11.txt>

[Alexandre Petrescu <[e-mail]>]

Le 19/05/2016 à 11:03, Lorenzo Colitti a écrit :
> On Thu, May 19, 2016 at 5:38 PM, Mark Smith <[e-mail]
> <mailto:[e-mail]>> wrote:
>
> So how would the IPv6 layer test that, and if it was tested properly,
> what happens if the test fails?
>
> The code that configures both layer 2 and layer 3 can simply call
> arc4random_buf (or your favourite random function)

I think the question is to decide when to call that random
function.

A 48bit identifier is often there already, but we need to use if it's
random, or generate one if it's not random.  So how to test whether it's
random?

> and then "ifconfig wlan0 hw ether <foo>", and then pick an EUI-64 IID
> based on that. Even if the hardware does not support the hardware
> address and does not return an error, but silently continues to use
> the hardware MAC address, the resulting IPv6 address is still random
> and free from privacy concerns.
>
> I think the testing of layer 2 address randomisation, and coming up
> with a both a user acceptable and friendly scheme to deal with a test
> failure is harder and more complex that just universally applying
> RFC7217 to all past, current and future link layers, regardless of
> how good or bad their link layer addresses are.
>
> This is really not hard to test in any hardware qualification lab.
> Really, it isn't.

On the contrary, I think it is very hard to test a number to see whether
it's random - you need a very expensive lab and out-of-band info.

Alex

>
>
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list [e-mail] Administrative
> Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
>