26c04d1
another test issue found by scan-build by
2015-01-22 18:27:28 -0800
89165c7
fix test issue found by scan-build by
2015-01-22 18:25:17 -0800
6d798e2
fix test issue found by scan-build by
2015-01-22 18:25:17 -0800
d2e5e0f
const** related fixes for gcc by
2015-01-22 18:06:35 -0800
a59228c
const** related fixes for gcc by
2015-01-22 18:06:35 -0800
1b4072d
fix(?) perf regression in monty_aux_step by
2015-01-22 17:53:13 -0800
d274c64
fix performance regression for montgomery ladder by
2015-01-22 17:45:36 -0800
5965624
fix performance regression for montgomery ladder by
2015-01-22 17:45:36 -0800
5a0abe8
fixed undeclared function warning by
2015-01-22 17:34:10 -0800
93e866b
gmp-style foo_t[1] for points too by
2015-01-22 17:21:20 -0800
1f1836d
gmp-style foo_t[1] for points too by
2015-01-22 17:21:20 -0800
0a4593e
respace by
2015-01-22 16:22:37 -0800
4e926b0
respace by
2015-01-22 16:22:30 -0800
64ccdaf
field_t * -> field_a_t by
2015-01-22 16:08:29 -0800
e45137d
fix subx -> sub mistake by
2015-01-22 16:07:03 -0800
d383dfe
going to GMP-style element[1] types by
2015-01-22 15:35:58 -0800
b0a2110
remove probably-unnecessary optimizations (still needs benching to make sure) by
2015-01-22 14:37:45 -0800
33328a1
going to GMP-style element[1] types by
2015-01-22 15:35:58 -0800
942066a
remove probably-unnecessary optimizations (still needs benching to make sure) by
2015-01-22 14:37:45 -0800
825dd21
sync with master by
2015-01-22 13:23:03 -0800
16693df
port simplifying changes from decaf branch; going to make a few more as well by
2015-01-16 16:23:53 -0800
8abc24f
montgomery aux step is defined; probably doesnt work by
2015-01-16 16:52:56 -0800
a412e7b
simplify ec_point slightly by using subx,negx by
2015-01-16 16:23:53 -0800
ebb6b4d
both edwards and twisted edwards are in, and agree with one another. point comparison is in. by
2015-01-15 17:05:52 -0800
8e3ed52
sqrt(-d) by
2015-01-15 15:05:08 -0800
a9e1644
decaf in and working for untwisted curve by
2015-01-15 13:09:39 -0800
de113cb
working on experimental, cofactor-removing compression code by
2015-01-14 16:10:46 -0800
84abf97
remember to save HISTORY.txt; also fix some regressions related to -Werror dead code by
2014-10-27 17:07:19 -0700
1d07343
p521 testing, 803kcy ecdh by
2014-10-27 16:35:09 -0700
6546660
E-521-related changes. Not quite ready yet... by
2014-10-25 17:26:05 -0700
0dc1b0d
magic for p521. except not entirely, because there will probably be an impl with permuted limbs by
2014-10-23 18:12:54 -0700
1eaab1f
fix up conflict due to sync mistake by
2014-10-23 17:51:07 -0700
1f480b0
Big changes for curve flexibility. For details see HISTORY.txt. by
2014-10-23 17:41:51 -0700
b69f115
make sure _XOPEN_SOURCE gets defined before posix_memalign is used by
2014-10-21 11:18:22 -0700
edc6afe
no perf regression on haswell. Also, factored out field_cond_neg; restored p448_ prefixes in case of multiple fields in the same lib by
2014-10-20 18:20:35 -0700
c6d69de
WARNING: This commit is largely untested. by
2014-10-20 17:13:41 -0700
b4ce20d
merge in notes on elligator; move some point ops to ec_point by
2014-10-20 14:51:18 -0700
165510d
Working on demagication, so as to support other curves for the CFRG benchmarks in a month or so. by
2014-10-18 18:08:58 -0700
cc3c637
Changes to the eBAT build system, to bring the code and package closer to how DJB expects them. by
2014-10-12 17:22:10 -0700
3c30fb4
oops i forgot to save changes to HISTORY.txt by
2014-09-29 16:06:59 -0700
4433591
Forget yesterday's hack; just add an arch_config.h to each arch which says how many bits it is. by
2014-09-29 12:59:12 -0700
47b06f3
Spell __clang_minor__ correctly. Also, this and the prev patch should fix SUPERCOP problems by
2014-09-28 19:25:16 -0700
e692da0
Don't look at __SIZEOF_INT128__ on clang-3.2 or earlier, because they don't define it. by
2014-09-28 19:23:45 -0700
4f841cd
convert arch_neon_experimental to using vaguely cryptopp-style macroed asm. now it compiles on gcc in addition to\ clang, though it is still faster on clang by
2014-09-23 21:11:03 -0700
00b3b3b
unify all the ec_point.c files by
2014-09-22 22:26:52 -0700
a9c72b5
Begin ref impl, currently an arch option (arch_ref64). by
2014-09-18 21:21:01 -0700
4ca664e
adjust HISTORY by
2014-09-06 22:59:16 -0700
46c2fb4
A few adjustments on contributed patches. by
2014-09-06 22:54:57 -0700
a941e63
some minor changes to get the arch_32 stuff compiling, and a few tweaks for my use by
2014-08-29 18:10:06 +0100
c699cb2
Minor by
2014-09-06 22:05:41 -0400
c28723f
possibly the eBAT actually works now? by
2014-09-06 22:36:41 -0700
76b7562
Thanks again Samuel Neves: the "gcc bug" is actually undefined behavior caused by strict aliasing. For now just passing -fno-strict-aliasing, maybe unions to fix it later by
2014-08-09 14:34:50 -0700
94c51dd
Better patch for the bug Samuel Neves described by
2014-08-05 16:04:57 -0700
06502b4
Minor alignment fix in arch_neon_experimental by
2014-08-04 21:39:53 -0700
04b955e
Added really_memset, thanks David Leon Gil. by
2014-08-04 20:46:17 -0700
5847031
fix bug in word.h, thanks David Leon Gil by
2014-07-16 14:41:12 -0700
4eb210c
Mostly a cleanup release. Cleanup old code, improve documentation, improve GCC-cleanness, etc. by
2014-07-11 15:39:42 -0700
8ebdfae
gcc-clean, though the code is slow when compiled by gcc by
2014-05-07 13:35:24 -0700
d4085b9
Internal changes which break compatibility with previous versions (you knew this would happen). by
2014-05-03 20:26:33 -0700
d7f64fd
delete the text files, for real this time? by
2014-03-30 21:23:21 -0700
1eab9a3
New release. by
2014-03-29 15:58:41 -0700
ed6fdbf
unify error messages returned by goldilocks_verify to GOLDI_EINVAL by
2014-03-05 17:47:07 -0800
2265248
Change MUST_HAVE_AVX to MUST_HAVE_AVX2 since the other flags assume AVX2. by
2014-03-05 16:30:14 -0800
2712648
Second commit. Still very preliminary. by
2014-03-05 16:19:34 -0800
25697ca
Import the code by
2014-02-21 11:29:33 -0800
acd0950
Just the text files, until I get Legal approval to push the button by
2014-02-21 11:26:37 -0800