Re: [TLS] Inter-protocol attacks

[Ilari Liusvaara <[e-mail]>]

On Wed, Aug 13, 2014 at 04:00:28PM +0200, Martin Rex wrote:
> 
> But with your description I somehow still fail to see a problem.
> If the server certificate of such resumptions will *pass* the
> "server endpoint identification", then why would re-doing the
> server endpoint identification make a difference?

As far as I understand, the actual issue is HTTP servers issuing
false responses to HTTP reqeusts if sent to wrong server (but
that server may still have valid certificate). And clients
interpretting those false responses as proper ones then leads
to all sorts of unpleasant problems.

Usually those false responses are because of misconfiguration
(usually default virtual host is something unintended, or the
default vhost is issuing redirects to "main site" even when
such redirect must not be sent (the vhost being handled else-
where)).

HTTP/2 even defines error code specifically for responding to
such requests (421). That code would also apply to HTTP/1.1
(in pinch, 400 would do while waiting for 421).


Where session resumption comes in is that it is more vulernable
than full handshake to attacker interference, resulting in it
being easier to make user connect to wrong server.


-Ilari