eb74c36bd2
Abstract out tweak logic to secp256k1_eckey_* functions
2014-10-27 03:27:55 -07:00
ffffc87855
Use internal secp256k1_eckey_ prefix for functions in eckey
2014-10-27 03:01:51 -07:00
e2f71f1efe
Move non-ECDSA operations from ecdsa to eckey
2014-10-27 02:58:09 -07:00
ae6bc76e32
[API CHANGE] Use secp256k1_ec_ prefix for non-ECDSA key operations
2014-10-27 02:51:58 -07:00
dc407ed48c
Avoid valgrind complaints in OpenSSL initialization check
2014-10-27 02:27:58 -07:00
79f599d37f
Tests take too long by default
2014-10-27 02:25:56 -07:00
03bfc07b9e
Avoid uninitialized access in secp256k1_gej_double
2014-10-26 06:51:38 -07:00
bff11e9112
Verify num=openssl initialization and check repeatability
2014-10-26 05:39:13 -07:00
949c1ebb5e
Split up ecmult and ecmult_gen entirely
2014-10-26 03:42:24 -07:00
3fd6253e13
Better randomization for tests
2014-10-17 16:04:00 -07:00
bae6a42be4
Make bench use external interface
2014-09-29 08:21:59 +02:00
c259a7cbc0
Set precomputation table late and unset early.
...
Set the global pointer to the precomputation table only after initializing
it completely, and unset it before doing any uninitialization.
This causes fail-fast behavior in case of race conditions between
initialization and operations using it.
2014-09-13 17:19:30 +02:00
04e34d18c3
Split up signing and verification initialization
2014-09-13 17:14:17 +02:00
da556eb772
Merge pull request #62
...
1136bed
2014-09-13 16:26:27 +02:00
1136bedbc2
Make secp256k1_ge_set_gej work with the point at infinity
2014-09-13 00:46:32 +02:00
62c3f55a9d
Nothing-up-my-sleeving blinding for a*G
2014-09-01 14:56:12 +02:00
fa492f059d
Fix a signedness mistake in secp256k1_num_set_hex
...
We were using a potentially signed char as index in an array.
2014-08-27 01:21:57 +02:00
e2beb0bd2d
Merge pull request #51
...
364fde6
2014-08-27 01:20:19 +02:00
1c7fa133a6
Add VERIFY_CHECK/DEBUG_CHECK and use CHECK macros more
2014-08-26 01:23:51 +02:00
87c782f632
Merge pull request #53
...
2f6c801
2014-08-23 14:13:10 +02:00
2f6c801911
Try to not leave secret data on the stack or heap.
...
This makes a basic effort and has not been audited.
Doesn't appear to have a measurable performance impact on bench.
It also adds a secp256k1_num_free to secp256k1_ecdsa_pubkey_create.
2014-08-14 07:06:36 -07:00
364fde65f6
fix unsigned warning in num_gmp_impl.h
2014-08-11 16:09:09 -04:00
1a749b4a6e
Add secp256k1_num_eq and use it in tests
2014-08-09 19:22:42 +02:00
25f4aec02f
Add high-level secp256k1.c tests
2014-08-03 19:54:41 +02:00
21f81a8469
Correct secp256k1_fe_verify and use it everywhere
2014-07-17 21:07:40 +02:00
59447da38d
Test demonstrating discrepancy in sqr output
2014-07-17 21:07:38 +02:00
f16be77ffc
Use batch inversion in G precomputation
2014-07-17 15:33:00 +07:00
42822baaa8
Improve normalization performance for 32bit
...
- Uses a similar approach to the latest 64bit _normalize.
- Add one useful optimization back into the 64bit _normalize too.
Performance of 'bench' improved by around 0.5% for the 32bit field (but tested on a 64-bit machine).
2014-06-23 12:12:58 +07:00
9849c6898c
Merge pull request #31
...
0592d11
2014-06-20 23:25:38 +02:00
0592d117da
Introduce CHECK() for tests that works with NDEBUG.
2014-06-16 01:30:17 +02:00
4d79bebdff
Do not free endomorphism constants when disabled
2014-06-15 23:31:17 +02:00
ebfa43b20a
Merge pull request #25
...
6d6102f
2014-06-03 22:01:47 +02:00
9037707720
Rewrite 5x52 normalize method to be faster
2014-06-03 18:59:21 +07:00
6d6102fee4
A few improvements to the sqrt tests
2014-06-03 17:50:27 +07:00
7740eff28a
Merge pull request #19
...
09ca4f3
2014-06-02 10:46:12 +02:00
3bc866b88f
Merge pull request #20
...
d7fd4d0
2014-06-02 10:45:44 +02:00
17eec032c8
Support 64bit_asm field on OSX
2014-06-02 15:33:35 +07:00
d7fd4d0f6b
Use endomorphism in precomputations
2014-05-22 10:17:30 +07:00
09ca4f32e2
secp256k1_fe_sqrt checks for success
...
- secp256k1_fe_sqrt now checks that the value it calculated is actually a square root.
- Add return values to secp256k1_fe_sqrt and secp256k1_ge_set_xo.
- Callers of secp256k1_ge_set_xo can use return value instead of explicit validity checks
- Add random value tests for secp256k1_fe_sqrt
2014-05-21 10:22:14 +07:00
78fb796997
Merge pull request #8
...
ba8fc0e
2014-05-08 20:09:09 +02:00
ac274bff23
Fix x64 gmp init on platforms that define UL as 32 bits.
2014-05-08 10:28:20 -07:00
ba8fc0e28f
Check signature nonces for validity
...
The documentation implies that this check is happening, so make it so.
Without this check, passing an invalid nonce will trigger an internal assertion.
2014-04-24 14:18:44 -07:00
11ab562203
Move implementations from impl/*.h to *_impl.h
2014-03-12 18:40:02 +01:00
65a79b300c
Slice bytes of G multiples to avoid cache timings
2014-03-07 00:32:41 +01:00
bb7bb81d07
Fix addition chain entry in comments
2014-01-31 11:43:57 +07:00
f8ccd9befd
Re-write secp256k1_fe_sqrt and secp256k1_fe_inv to use a sliding window over blocks of 1s
2014-01-30 19:16:20 +07:00
5a9989c52a
add missing include
2014-01-17 23:24:12 -05:00
78cd96b151
autotools: autotools'ify libsecp256k1
2014-01-17 23:24:12 -05:00
60442b835f
Remove 5x64 implementation - no actual benefits
2013-12-01 21:16:47 +01:00
399c03f227
Make endomorphism optimization optional
2013-12-01 21:06:20 +01:00
Pieter Wuille
Gregory Maxwell
caktux
Pieter Wuille
Peter Dettman
evoskuil
William Swanson
Cory Fields