Re: [Model-t] What are we trying to protect

[Eric Rescorla <[e-mail]>]

On Sun, Aug 4, 2019 at 3:37 AM Stephen Farrell <[e-mail]>
wrote:

>
> Hiya,
>
> On 04/08/2019 06:08, Eric Rescorla wrote:
> > his seems like a reasonable problem statement for the overall problem of
> > computer security, but not really for IETF.
> >
> > To take a concrete example memory reading attacks like Spectre are a
> threat
> > to user data and something that browser vendors spend a fair amount of
> > energy working on, but they're mostly not in scope for IETF [0]. There's
> > nothing wrong with that, it's just division of labor.
>
> I tend to agree.
>
> OTOH, one might also wonder if such attacks are or ought be
> considered in NFV and similar things. Such wonderings could
> I think be useful, even if the answer is maybe that without
> the flexibility of JS a successful attack in such a scenario
> seems unlikely enough to not have to (currently) bother with.
>

That's true in the Web context, but not in other contexts, such as virtual
hosting. Indeed, cross-VM attacks go back even before Spectre, at least to,
Ristenpart et al. 2009
https://hovav.net/ucsd/papers/rtss09.html.


On 04/08/2019 10:09, Eric Rescorla wrote:
> > - Compromise of the WebPKI would obviously be very bad
> > for many protocols the IETF has designed, but the management
> > of the WebPKI is done by CABF, not IETF.
>
> That's also true, but nonetheless the IETF did document CT
> as a direct result of the WebPKI CAs doing a bad job wrt
> mis-issuance risks. So while "replace CABF, because <X>"
> would not be a sensible outcome of this discussion, it's
> reasonable to not consider the current WebPKI as a fixed
> unchangeable thing, for the purposes of this discussion.
>

i don't disagree with this. My point was merely that this is an example of
deciding that certain technical work was out of scope.

-Ekr