Re: [saag] another conflict review of some GOST stuff

["Stanislav V. Smyshlyaev" <[e-mail]>]

Hello, Tero!


>>That draft is not enough to specify how them are used in the
>>interoperable manner, as it does not allocate IANA numbers for them.
>>Of course if you have some other document specifying how to use
>>private use numbers for those algorithms then you could use those
>>algorithms, but then I think it would be better to move sections
>>4.2.2 to that document.

>>So if the algorithms are already specified in earlier RFCs, this
>>document tries to specify how to use them, but does not do that as it
>>does not even try to do IANA allocations, I do fail to see the meaning
>>of this document.

>>I.e. is this document trying to make GOST so that it can be used in
>>the IPsec or not. If yes, are you expecting to do IANA allocations
>>based on this document later, or what?






Let me clarify this. We don’t want any IANA allocations based on this
document. We fix algorithms for usage in protocols (IKEv1, TLS) – to cite
the document with algorithms, not describing protocols there. We don’t
think that this approach contradicts anything. This RFC is, in fact, an
international version of Russian standardization system document of
Technical Committee on Cryptography (TC 26) – that has been officially
approved and is currently used. All protocol definitions will be made (for
IKEv2 and TLS) independently in other RFCs.





>>From that point of view the IKEv2 section of the draft (4.2.2.2) is
>>not enough. The section 4.2.2.2 says that for this document specifies
>>HMAC_GOST3411_2012_256 and HMAC_GOST3411_2012_512 for PRFs for IKEv2,
>>but this document does not specify those at all. There is
>>HMAC_GOSTR3411_2012_256 and HMAC_GOSTR3411_2012_512, but I am not sure
>>whether they are actually same (GOST vs GOSTR), i.e. is this just typo
>>on one of those sections or something else?



Yes, this is a misprint. HMAC_GOSTR3411_2012_256 and
HMAC_GOSTR3411_2012_512 must be there. We’ll correct it in the final
revision, thank you very much.



>>The section 4.1.1 defining HMAC_GOSTR3411_2012_256 says that it "uses
>>H_256 as a hash function for HMAC", but does not specify what H_256
>>actually means. I assume it means n = 256 bits hash function from the
>>RFC6986, but that is not exactly clear.



H_256 is defined in the section «abbreviations and symbols» at page 4:

H_256 | GOST R 34.11-2012 hash function with 256-bit output |



>>Also text in this document only covers using those GOST as PRF, there
>>is not enough specification to use them as AUTH algorithm, or to use
>>the digital signatures in the IKEv2 authentication.



Yes, we know. If we make specifications of GOST usage in IKEv2, we’ll
prepare a specification for it.


Best regards,

Stanislav