Peter Saint-Andre <[email protected]> writes:
>On 6/25/22 6:20 PM, Peter Gutmann wrote:
>> Yaron Sheffer <[email protected]> writes:
>>
>>> This revision addresses Ben's SecDir review, as well as several other 
>>> reviewers' comments. Thank you all!
>>
>> It doesn't have anything about EtM as per the recent discussion though...
>
>The conclusion of that discussion wasn't clear to me. Because none of the
>recommended ciphersuites have distinct encrypt and MAC operations, encrypt-
>then-MAC can't be anchored to the ciphersuite recommendations we currently
>make. Can you propose text?

The draft recommends only GCM suites (and nothing else) which many embedded
implementations don't support, so it's a bit of an ostrich-algorithm approach
to the issue... one approach would be to add:

   *  TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA256 and EtM

   *  TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA384 and EtM

   *  TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA256 and EtM

   *  TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA384 and EtM

Alternatively, I can create some text to say "if you use CBC then you MUST use
EtM", but it doesn't solve the problem of having only GCM as a recommended
algorithm, so I think adding CBC+EtM suites alongside GCM ones would be a
better fix.

Peter.

_______________________________________________
Uta mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/uta

Reply via email to