Re: [PATCH v5 05/12] crypto: ecc - Implement vli_mmod_fast_521 for NIST p521

[Stefan Berger <stefanb@linux.ibm.com>]

On 3/11/24 01:07, Bharat Bhushan wrote:
> Minor nits
> 
>> -----Original Message-----
>> From: Stefan Berger <stefanb@linux.ibm.com>
>> Sent: Thursday, March 7, 2024 3:53 AM
>> To: keyrings@vger.kernel.org; linux-crypto@vger.kernel.org;
>> herbert@gondor.apana.org.au; davem@davemloft.net
>> Cc: linux-kernel@vger.kernel.org; saulo.alessandre@tse.jus.br;
>> lukas@wunner.de; Stefan Berger <stefanb@linux.ibm.com>
>> Subject: [EXTERNAL] [PATCH v5 05/12] crypto: ecc - Implement
>> vli_mmod_fast_521 for NIST p521
>>   
>> ----------------------------------------------------------------------
>> Implement vli_mmod_fast_521 following the description for how to calculate
>> the modulus for NIST P521 in the NIST publication "Recommendations for
>> Discrete Logarithm-Based Cryptography: Elliptic Curve Domain Parameters"
>> section G.1.4.
>>
>> NIST p521 requires 9 64bit digits, so increase the ECC_MAX_DIGITS so that
>> arrays fit the larger numbers.
>>
>> Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
>> Tested-by: Lukas Wunner <lukas@wunner.de>
>> ---
>>   crypto/ecc.c                  | 31 +++++++++++++++++++++++++++++++
>>   include/crypto/internal/ecc.h |  2 +-
>>   2 files changed, 32 insertions(+), 1 deletion(-)
>>
>> diff --git a/crypto/ecc.c b/crypto/ecc.c index f53fb4d6af99..373660e7b19d
>> 100644
>> --- a/crypto/ecc.c
>> +++ b/crypto/ecc.c
>> @@ -902,6 +902,31 @@ static void vli_mmod_fast_384(u64 *result, const
>> u64 *product,  #undef AND64H  #undef AND64L
>>
>> +/* Computes result = product % curve_prime
>> + * from "Recommendations for Discrete Logarithm-Based Cryptography:
>> + *       Elliptic Curve Domain Parameters" G.1.4
>> + */
>> +static void vli_mmod_fast_521(u64 *result, const u64 *product,
>> +				const u64 *curve_prime, u64 *tmp)
>> +{
>> +	const unsigned int ndigits = 9;
>> +	size_t i;
>> +
>> +	for (i = 0; i < ndigits; i++)
>> +		tmp[i] = product[i];
>> +	tmp[8] &= 0x1ff;
>> +
>> +	vli_set(result, tmp, ndigits);

I have also modified this here now to initialize 'result' from lowest 
521 bis of product without the detour through tmp.

>> +
>> +
>> +	for (i = 0; i < ndigits; i++)
>> +		tmp[i] = (product[8 + i] >> 9) | (product[9 + i] << 55);
>> +	tmp[8] &= 0x1ff;
>> +
>> +	vli_mod_add(result, result, tmp, curve_prime, ndigits); }
>> +
>> +
>>   /* Computes result = product % curve_prime for different curve_primes.
>>    *
>>    * Note that curve_primes are distinguished just by heuristic check and @@ -
>> 941,6 +966,12 @@ static bool vli_mmod_fast(u64 *result, u64 *product,
>>   	case 6:
>>   		vli_mmod_fast_384(result, product, curve_prime, tmp);
>>   		break;
>> +	case 9:
> 
> Can we use ECC_CURVE_NIST_P384_DIGITS, ECC_CURVE_NIST_P256_DIGITS  in this function?
>   
> And define ECC_CURVE_NIST_P521_DIGITS, which is same as ECC_MAX_DIGITS defined below in this patch?
> 
>> +		if (curve->nbits == 521) {

If I replace the numbers with these hash-defines's in here (in an 
additional patch on existing code) then I can just about remove the 
check on nbits here as well... ?


>> +			vli_mmod_fast_521(result, product, curve_prime,
>> tmp);
>> +			break;
>> +		}
>> +		fallthrough;
>>   	default:
>>   		pr_err_ratelimited("ecc: unsupported digits size!\n");
>>   		return false;
>> diff --git a/include/crypto/internal/ecc.h b/include/crypto/internal/ecc.h index
>> 4a556b41873e..de17bcdeb53a 100644
>> --- a/include/crypto/internal/ecc.h
>> +++ b/include/crypto/internal/ecc.h
>> @@ -33,7 +33,7 @@
>>   #define ECC_CURVE_NIST_P192_DIGITS  3
>>   #define ECC_CURVE_NIST_P256_DIGITS  4
>>   #define ECC_CURVE_NIST_P384_DIGITS  6
>> -#define ECC_MAX_DIGITS              (512 / 64) /* due to ecrdsa */
>> +#define ECC_MAX_DIGITS              DIV_ROUND_UP(521, 64) /* NIST P521 */
>>
>>   #define ECC_DIGITS_TO_BYTES_SHIFT 3
>>
>> --
>> 2.43.0
>>
>