[PATCH] Poly1305-AEAD: update to match latest Internet-Draft version
jussi.kivilinna at iki.fi
Sun Jun 29 13:18:35 CEST 2014
On 23.06.2014 19:50, Werner Koch wrote:
> On Mon, 23 Jun 2014 18:01, jussi.kivilinna at iki.fi said:
>> places, so it might be good idea to support both the old Poly1305-AEAD
>> and the new. But what would be the best way to select between the two
>> Poly1305-AEAD modes?
> We had the same probloem with the TIGER algorithm. Eventually we ended
> up with 3 algorithm ids to support all variants and implementation bugs.
> Thus what about:
> GCRY_MAC_POLY1305X = 501,
> GCRY_MAC_POLY1305X_AES = 502,
> GCRY_MAC_POLY1305X_CAMELLIA = 503,
> GCRY_MAC_POLY1305X_TWOFISH = 504,
> GCRY_MAC_POLY1305X_SERPENT = 505,
> GCRY_MAC_POLY1305X_SEED = 506,
> GCRY_MAC_POLY1305 = 511,
> GCRY_MAC_POLY1305_AES = 512,
> GCRY_MAC_POLY1305_CAMELLIA = 513,
> GCRY_MAC_POLY1305_TWOFISH = 514,
> GCRY_MAC_POLY1305_SERPENT = 515,
> GCRY_MAC_POLY1305_SEED = 516
> or maybe better "OLD" instead of "X".
Currently the MAC algorithm has not changed, so we'd be looking at cipher modes:
GCRY_CIPHER_MODE_POLY1305OLD = 10,
GCRY_CIPHER_MODE_POLY1305 = 11, /* or ..._POLY1305RFC */
Maybe it's better to wait for the actually RFC before adding new algorithm.
More information about the Gcrypt-devel