Why aren't these IKEv1 implementation defending against Bleichenbacher by
hiding the nature of error (decryption error or authentication error) like
what TLS 1.2 does [1]? Is this option documented in any of the IPsec/IKE
documents? Is there a reason why it would not work in IKEv1 as it worked
for TLS?

[1] RFC 5246 Sec 7.4.7.1

Hugo



On Tue, Aug 14, 2018 at 2:16 PM, Tero Kivinen <[email protected]> wrote:

> Valery Smyslov writes:
> > after reading the paper I still don’t understand why authors
> > mentioned IKEv2 there.
>
> Of course they mention, just to get more exposure...
>
> > Their example attack in Section 4.4 on (allegedly) IKEv2 in fact
> > uses secondary responder supporting IKEv1 Public Key Encryption
> > mode, without which the attack is impossible (as far as I
> > understand). So, in my opinion, the authors are at least not
> > accurate in claiming that IKEv2 itself is susceptible. Or am I
> > missing something?
>
> That is my understanding also.
>
> On the other hand breaking signature method of the IKEv2 using IKEv1
> oracle for the same RSA key is quite clever, and just points out that
> we must not use same key pair with different protocols / modes etc.
>
> We had this dicussion in when writing RFC8247 and recommending RFC7427
> Digital Signatures using RSASSA-PSS. Some implementors used same key
> pair for both RSASSA-PKCS1-v1_5 and RSASSA-PSS. This paper just points
> you that you must not do that kind of things.
>
> The text about the IKEv2 pre-shared keys is just already known fact,
> and the RFC7296 warns about that several times, i.e., it does say use
> EAP for those, and for EAP you should not use the weak EAP methods,
> but instead you should use key generating methods. Also because it is
> known fact that IKEv2 shared secret authentication is vulnerable to
> the off-line dictionary attacks, we created RFC6467 for secure
> password methods. I.e., PACE (RFC6631), AugPAKE (RFC6628), abd Secure
> PSK Authentication (RFC6617) should all be protected against offline
> dictionary attacks.
>
> Anyways, I think proper fix would be to disable IKEv1 support
> completely, not to just disable RSA encryption mode...
> --
> [email protected]
>
> _______________________________________________
> IPsec mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/ipsec
>
_______________________________________________
IPsec mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ipsec

Reply via email to