2c5e12d5f8
Workaround aliasing in Fp12[BLS12-377] inversion, fix #147
2021-02-02 12:53:36 +01:00
83dcd988b3
FpDbl revisited ( #144 ) - 7% perf improvement everywhere, up to 30% in double-width primitives
...
* reorg mul -> limbs_double_width, ConstantineASM CttASM
* Implement squaring specialized scalar path (22% faster than mul)
* Implement "portable" assembly for squaring
* stash part of the changes
* Reorg montgomery reduction - prepare to introduce Comba optimization
* Implement comba Montgomery reduce (but it's slower!)
* rename t -> a
* 30% performance improvement by avoiding toOpenArray!
* variable renaming
* Fix 32-bit imports
* slightly better assembly for sub2x
* There is an annoying bottleneck
* use out-of-place Fp assembly instead of in-place
* diffAlias is unneeded now
* cosmetic
* speedup fpDbl sub by 20%
* Fix Fp2 -> Fp6 -> Fp12 towering. It seems 5% faster
* Stash ADCX/ADOX squaring
2021-02-01 03:52:27 +01:00
d12d5faf21
Implement Jacobian mixed addition ( #142 )
2021-01-30 14:21:55 +01:00
b91ec1cb15
Metering ( #140 )
...
* Add metering facilities
* Metering reporting
* Add example report on metering BLS12-381 pairings
2021-01-29 22:21:19 +01:00
95e23339b2
Decimal conversion ( #139 )
...
* Add constant-time fromDecimal conversion. Add warnings on intended purposes of hex/decimals
* introduce setuint + cosmetic fixes Wordbitsize -> Wordbitwidth in comments
* Add decimal conversion (non-constant-time)
* fix comments [skip ci]
2021-01-29 20:42:36 +01:00
47daefde1f
forgot an import
2021-01-24 13:55:18 +01:00
98a4b2f91a
constant cosmetics
2021-01-24 12:57:13 +01:00
75493dfb5b
Fix #131 , inversion tests didn't take into account that the RNG can produce a 0 input and so a.inv can be different from 1
2021-01-24 12:37:02 +01:00
7e97cd4ac5
Fuzz fix - non-unique modular representation after Assembly negate ( #137 )
...
* Fix #114 - Negating 0 left the prime modulus, which is working most of the time for everything except for comparison. (also somehow triggers and workaround weird compiler bug where exceptions tracking is activated in macros and all the curve enums were stringified as their ordinal value)
* https://github.com/mratsim/constantine/issues/136 was also fixed, add to anti-regression
* add comment in test
* Fix the pure Nim fallback as well
2021-01-24 12:35:27 +01:00
5b1d280486
Fix 50% perf regression (2x with GCC) on binary GCD based inversion ( #135 )
...
* Fix 50% perf regresion Revert part of #95 , fix #134
* Deactivate inversion via addition chain for BW6-761. 2x slower than Euclid
2021-01-23 21:44:22 +01:00
82819b1b10
Square Root & Inversion addition chains - 20% perf increase ( #132 )
...
* Addition chain for sqrt BLS12-381: 20% perf improvement
* sqrt addchain for BN254_Snarks - 20% perf improvement as well
* Fix operation count [skip ci]
* BLS12-377 sqrt - 10% perf improvement
* sqrt addition chain for BW6-761 - 6% speedup
* BN254_Nogami inversion addchain
* sqrt addchain for BN254_Nogami
* Inversion addchain for BLS12-377
* inversion ddition chain for BW6-761
2021-01-23 20:55:40 +01:00
a02dd19d36
Compendium of pairing-based cryptography optimizations
2021-01-23 15:46:41 +01:00
638cb71e16
Fr: Finite Field parametrized by the curve order ( #115 )
...
* Introduce Fr type: finite field over curve order. Need workaround for https://github.com/nim-lang/Nim/issues/16774
* Split curve properties into core and derived
* Attach field properties to an instantiated field instead of the curve enum
* Workaround https://github.com/nim-lang/Nim/issues/14021 , yet another "working with types in macros" is difficult https://github.com/nim-lang/RFCs/issues/44
* Implement finite field over prime order of a curve subgroup
* skip OpenSSL tests on windows
2021-01-22 00:09:52 +01:00
ac6300555a
Fix test suite ( #116 )
...
* Pin nim-serialization. Workaround #113 and https://github.com/status-im/nim-serialization/issues/33
* Need to workaround nimble installing dependency multiple times
* non-interactive
* UB sanitizer missing on mingw
* Fix OpenSSL benchmark on non-Linux platforms
* Accelerate CI:
- Skip 32-bit on 64-bit tests
- Only test leaf functionality.
* Don't define -fstack-protector-all with MinGW
* skip line functions and cyclotomic tests (already tested in pairing) + only compile the benches don't run them.
2021-01-21 21:25:42 +01:00
023e690efc
Fix #111
2021-01-11 08:25:02 +01:00
29fffecc93
Update README.md
2021-01-06 19:27:16 +01:00
a5c1d077fb
deal with DLL mess for OpenSSL test
2021-01-03 21:50:22 +01:00
e89429e822
SHA256 Hash function
2020-12-15 19:18:36 +01:00
c89c78d2d9
Typo Borrow instead of Carry in return type
2020-12-13 18:57:23 +01:00
a022db1c08
Sqrt fp2 acceleration ( #109 )
...
* Use sqrt.square() == a instead of sqrt * invsqrt = -1 (Euler criterion) for sqrt existence.
* Accelerate sqrt_fp2 by 33%
2020-12-13 17:08:32 +01:00
f0b18ecfe0
Github Actions add-path is gone https://github.blog/changelog/2020-10-01-github-actions-deprecating-set-env-and-add-path-commands/
2020-12-13 15:28:51 +01:00
13e6b7dfe6
Travis pricing change, keep only ARM64 and PPC64
2020-12-12 19:47:45 +01:00
45ef3a65e0
Skip 32-bit tests on 64-bit machines (too long)
2020-10-31 14:51:17 +01:00
244f58350c
Implement BW6-761 Endomorphism acceleration ( #104 )
...
* Implement BW6-761 GLV on G1 + Psi Untwist-Frobenius-Twist
* Fix frobenius constants for embedding degree != 12
* Fix test type/parsing issues
* Generalize frobenius map coefficient formula
* Fix Frobenius Psi generalization
* Don't confuse t and trace of frobenius + update scalarMul to use Frobenius on Fp Twist
* Fix ec_sage type definition
* fix decription [skip ci]
* update comment [skip ci]
* typo
* restore frobenius tests iterations
2020-10-13 23:58:35 +02:00
7f0f511760
Fix Readme Github Action link [skip CI]
2020-10-11 22:04:13 +02:00
7826c40e26
Unify GCC and Clang ASM ( #103 )
...
* GCC-10 on Mac seems to require this syntax instead of the "+"
* update comment
2020-10-11 21:36:16 +02:00
1383aae105
Remove outdated TODOs [skip ci]
...
- noinline consts: https://github.com/nim-lang/RFCs/issues/257
2020-10-11 21:33:59 +02:00
6530596032
Endomorphism acceleration for BN254-Nogami ( #102 )
2020-10-10 18:53:48 +02:00
a2f46f77b7
Sage constants & tests codegen ( #101 )
...
* Implement a Sage codegenerator for frobenius constants
* Sage codegen for pairings
* Autogen of endomorphism acceleration constants
* The autogen fixed a copy-paste bug in lattice decomposition. We can use conditional negation now and save an add+dbl in scalar mul
* small fixes
* sage code for square root bls12-377 is not old
* readme updates
* Provide test suggestions for derive_frobenius
* indentation + add equation form to sage
* Sage test vector generator
* Use the json vectors
- includes type system workaround: generic sandwich https://github.com/nim-lang/Nim/issues/11225
- converting NimNode to typedesc: https://github.com/nim-lang/Nim/issues/6785
* Delete old sage code
* Install nim-serialization and nim-json-serialization in CI
* CI nimble install force yes
2020-10-10 16:19:23 +02:00
71bb4c799a
BW6-761 part 1 ( #100 )
...
* Add Fp, Fp2, Fp6 support for BW6-761
* Add G1 for BW6-761
* Prepare to support G2 twists on the same field as G1
* Remove a useless dependent type for lines
* Implement G2 for BW6-761
* Fix Line leftover
2020-10-09 07:51:47 +02:00
49164b66d8
fix testing canary
2020-10-05 22:20:29 +02:00
d622f48507
Unsed imports cleanup ( #97 )
2020-10-04 17:33:17 +02:00
fc1c3472ce
Fused projective line eval ( #96 )
...
* Reorg line functions to allow for Jacobian eval
* 2x faster Miller loop!!! with fused line eval double
* Support Line Double Fusion for D-Twists
* Implement fused line addition
2020-10-04 09:39:02 +02:00
986245b5c1
Jacobian coordinates ( #95 )
...
* Add projective-> affine bench
* Add conditional copy and div2 benches
* Fp4 benchmarks
* Constant-time Jacobian addition
* Jacobian doubling
* Use a simpler Add+Dbl complete formula
* Update tests
* Fix conditional negate
* Rollaback complete addition, we were only handling curve coef a == 0
2020-10-02 00:01:09 +02:00
0effd66dbd
SWei -> SHortW, weierstrass -> shortweierstrass
2020-09-27 23:02:48 +02:00
51586c7272
Merge pull request #94 from mratsim/reorg-curves-constants
...
Reorg curves constants
2020-09-27 22:31:13 +02:00
39ac014dfe
small fixes
2020-09-27 21:24:54 +02:00
00fa1ea7b6
Move pairings
2020-09-27 21:00:35 +02:00
2721131168
Move curve specific square root
2020-09-27 17:55:31 +02:00
204c72b811
again leftovers
2020-09-27 17:55:09 +02:00
48bab72d1f
cleanup leftover from BN curve inversion
2020-09-27 17:39:45 +02:00
8a456b84db
Move inversions curve-specific routines to the curve folder
2020-09-27 17:37:02 +02:00
3f48a590e8
Move assembly to their own folder
2020-09-27 17:25:21 +02:00
03d58ac1e7
Standardize constants names
2020-09-27 17:18:14 +02:00
f864fb20ec
Remove unused BN inversion
2020-09-27 16:17:13 +02:00
d04ccdd578
Move the cubic root to GLV files
2020-09-27 16:01:31 +02:00
34eb6fcfbd
Move GLV constants
2020-09-27 15:39:12 +02:00
e676e06959
move frobenius macros
2020-09-27 15:06:16 +02:00
fd45a790eb
Consolidated curve-specific Frobenius in a curve folder
2020-09-27 14:46:54 +02:00
92183c8b05
Remove unused curves
2020-09-27 13:13:45 +02:00
Mamy André-Ratsimbazafy