Re: [Cfrg] I-D Action: draft-irtf-cfrg-hpke-02.txt

[Stephen Farrell <[e-mail]>]

Hiya,

On 04/12/2019 11:29, Michael Scott wrote:
> Hello,
> 
> We have implemented an experimental version of HPKE, working from the
> latest draft (in the C++ language only for now)
> 
> https://github.com/miracl/core
> 
> Our experimental API supports all 4 modes of operation, but currently only
> for AES-GCM and the curves X25519 and P521.
> 
> Some feed-back: It would be a good idea to explicitly warn the reader that
> the X25519/X448 keys are actually in little-endian format, whereas for
> P256/521 they are big-endian.
> 
> The Npk sizes given for the KEMS for curves P256 and P521 appear to be for
> a compressed form of the points. In the test vectors uncompressed public
> keys are used.

Yep, I'd suggest changing to use uncompressed form and
changing the Npk, Nenc sizes accordingly. (To 65 for p256
and 133 for p521 if I got it right.) Anyway, my code [1]
does that and it matches the test vectors. BTW, if we did
stick with the compressed form, then 33 would seem more
useful than 32 for p256/Npk, but sticking with uncompressed
is likely better as that's what TLS does, so maybe code
supporting uncompressed will be easier to find.

> The first test vector provided (for mode 3) appears out-of-order.
> 
> Other than that a well written and easy to follow draft.

Not sure I'd agree with "easily" but I suspect your better
familiarity with the underlying crypto and your own miracl
code helped. FWIW though, even a relative ignoramus like me
thrashing about with OpenSSL managed to get it working:-)

Cheers,
S.

[1] https://github.com/sftcd/happykey


> 
> 
> Mike
> 
> On Wed, Dec 4, 2019 at 1:17 AM Nasrul Zikri <[e-mail]> wrote:
> 
>>
>> Hello Richard,
>>
>> Thanks for looking at my use-case. I have been working recently on a quick
>> implementation of FFDH so I would be interested to see how this perform in
>> HPKE. If I can test this first and then request code points later that is
>> good enough. If you can find a way to re-use TLS cipher suite code points
>> (like Stephen say) that is good also because this defines
>> ffdhe2048(0x0100), ffdhe3072(0x0101), ffdhe4096(0x0102), ffdhe6144(0x0103),
>> ffdhe8192(0x0104), ffdhe_private_use(0x01FC..0x01FF).
>>
>> I think there are improvements that can be made for generating custom
>> group for better security and speed. Using custom group help defend against
>> an powerful adversary who would put a lot of work into breaking a fixed
>> group, so I disagree that they are "even worse idea", but I realise there
>> will be difficulties in negotiate parameters for custom group so I don't
>> want to delay your draft.
>>
>> Tk,
>> Nasrul
>>
>>
>> ------------------------------
>> *From:* Richard Barnes <[e-mail]>
>> *Sent:* Thursday, November 28, 2019 04:02
>> *To:* Nasrul Zikri <[e-mail]>
>> *Cc:* [e-mail] <[e-mail]>
>> *Subject:* Re: [Cfrg] I-D Action: draft-irtf-cfrg-hpke-02.txt
>>
>> Hi Nasrul,
>>
>> Thanks for taking a look at this draft.  Personally, I am disinclined to
>> define FFDH schemes, unless there are other folks in the RG who think they
>> would be useful.  That said, the registry policy on group IDs is
>> Specification Required, so you can get code points if you have a
>> specification; it doesn't have to be in this doc..  AFAIK, there should be
>> no technical barrier to doing FFDH.
>>
>> As far as custom parameters, I think the only reasonable way to
>> accommodate them here would be to reserve some space in the registry for
>> private, vendor-specific use.  But this seems like an even worse idea than
>> FFDH, so again, I'm inclined to do nothing here.
>>
>> If other folks are interested in these cases, please speak up.
>>
>> --Richard
>>
>>
>> On Wed, Nov 20, 2019 at 1:32 AM Nasrul Zikri <[e-mail]>
>> wrote:
>>
>> On your draft of Hybrid Public Key Encryption.
>>
>> The draft appears to be for any DH KEM, but I note, however that the
>> examples and test vectors it gives are only for the elliptic curves
>> P-256, Curve25519, P-521, Curve448.
>>
>> Would it be possible to define the algorithm identifiers and test
>> vectors for some FFDH groups as well as the elliptic curve? Or is there
>> some important reason why only ECDH methods are suitable?
>>
>> If FFDH groups are indeed correct for use in the draft, it would appear
>> that the table in section 8.1 could be extended to allocate identifiers
>> for at least the parameter ffdhe2048, ffdhe3072, ffdhe4096, ffdhe6144,
>> ffdhe8192 as stated in RFC 7919, and perhaps the MODP groups as stated
>> in RFC 3526 and RFC 5114.
>>
>> I would also like there to be a way of specifying the use of a custom
>> finite field for when the use of a defined elliptic curve or finite
>> field is not enough. I realise that stating a method for transporting
>> the parameters {p,q,g} is outside the scope of this draft, but could a
>> value for custom groups or private use be stated in this table also?
>>
>> Tk,
>> Nasrul
>>
>>
>>
>>> Hey all,
>>>
>>> Happy IETF 106 deadline day!
>>>
>>> The authors feel that this version of HPKE is substantially complete.
>> All
>>> of the functional parts are there, as well as test vectors to facilitate
>>> interop.  And I think we've got some formal proofs on the way.  Please
>> take
>>> a look and speak up if you see any gaps.
>>>
>>> Thanks,
>>> --Richard
>>>
>>> On Mon, Nov 4, 2019 at 3:47 PM <[e-mail]>; wrote:
>>>
>>>>
>>>> A New Internet-Draft is available from the on-line Internet-Drafts
>>>> directories.
>>>> This draft is a work item of the Crypto Forum RG of the IRTF.
>>>>
>>>>         Title           : Hybrid Public Key Encryption
>>>>         Authors         : Richard L. Barnes
>>>>                           Karthik Bhargavan
>>>>         Filename        : draft-irtf-cfrg-hpke-02.txt
>>>>         Pages           : 45
>>>>         Date            : 2019-11-04
>>>>
>>>> Abstract:
>>>>    This document describes a scheme for hybrid public-key encryption
>>>>    (HPKE).  This scheme provides authenticated public key encryption of
>>>>    arbitrary-sized plaintexts for a recipient public key.  HPKE works
>>>>    for any combination of an asymmetric key encapsulation mechanism
>>>>    (KEM), key derivation function (KDF), and authenticated encryption
>>>>    with additional data (AEAD) encryption function...  We provide
>>>>    instantiations of the scheme using widely-used and efficient
>>>>    primitives.
>>>>
>>>>
>>>> The IETF datatracker status page for this draft is:
>>>> https://datatracker.ietf.org/doc/draft-irtf-cfrg-hpke/
>>>>
>>>> There are also htmlized versions available at:
>>>> https://tools.ietf.org/html/draft-irtf-cfrg-hpke-02
>>>> https://datatracker.ietf.org/doc/html/draft-irtf-cfrg-hpke-02
>>>>
>>>> A diff from the previous version is available at:
>>>> https://www.ietf.org/rfcdiff?url2=draft-irtf-cfrg-hpke-02
>>>>
>>>>
>>>> Please note that it may take a couple of minutes from the time of
>>>> submission
>>>> until the htmlized version and diff are available at tools.ietf.org.
>>>>
>>>> Internet-Drafts are also available by anonymous FTP at:
>>>> ftp://ftp.ietf.org/internet-drafts/
>>>>
>>>> _______________________________________________
>>>> Cfrg mailing list
>>>> [e-mail]
>>>> https://www.irtf.org/mailman/listinfo/cfrg
>>
>>
>> _______________________________________________
>> Cfrg mailing list
>> [e-mail]
>> https://www.irtf.org/mailman/listinfo/cfrg
>>
>> _______________________________________________
>> Cfrg mailing list
>> [e-mail]
>> https://www.irtf.org/mailman/listinfo/cfrg
>>
> 
> 
> _______________________________________________
> Cfrg mailing list
> [e-mail]
> https://www.irtf.org/mailman/listinfo/cfrg
>