wbrawner/openssl
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

ec_cvt.c: avoid EC_GFp_nist_method on platforms with bn_mul_mont [see

Browse source 
commentary for details].
This commit is contained in:
Andy Polyakov 2011-05-20 20:31:37 +00:00
parent f37f20ffd3
commit fdf6dac859
1 changed files with 19 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

19
crypto/ec/ec_cvt.c
View file

@ -80,10 +80,29 @@ EC_GROUP *EC_GROUP_new_curve_GFp(const BIGNUM *p, const BIGNUM *a, const BIGNUM
const EC_METHOD *meth;
EC_GROUP *ret;
#if defined(OPENSSL_BN_ASM_MONT) && !defined(__sparc)
/*
* This might appear controversial, but the fact is that generic
* prime method was observed to deliver better performance even
* for NIST primes on a range of platforms, e.g.: 60%-15%
* improvement on IA-64, 50%-20% on ARM, 30%-90% on P4, 20%-25%
* in 32-bit build and 35%--12% in 64-bit build on Core2...
* Coefficients are relative to optimized bn_nist.c for most
* intensive ECDSA verify and ECDH operations for 192- and 521-
* bit keys respectively. What effectively happens is that loop
* with bn_mul_add_words is put against bn_mul_mont, and latter
* wins on short vectors. Correct solution should be implementing
* dedicated NxN multiplication subroutines for small N. But till
* it materializes, let's stick to generic prime method...
* <appro>
*/
meth = EC_GFp_mont_method();
#else
if (BN_nist_mod_func(p))
meth = EC_GFp_nist_method();
else
meth = EC_GFp_mont_method();
#endif
ret = EC_GROUP_new(meth);
if (ret == NULL)