p384: const fn field arithmetic
#589
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Uses the `fiat-constify` tool to mechanically transform `fiat-crypto` generated field implementations into `const fn`: https://github.com/RustCrypto/utils/tree/master/fiat-constify Resulting code has been run through `rustfmt` prior to checking it in: rustfmt 1.4.38-stable (fe5b13d6 2022-05-18) The following are criterion benchmarks after the translation. While these values are within the threshold of noise I experience running them on a semi-busy desktop computer (MacBook w\ 2.3 GHz 8-Core Intel Core i9) they do paint a picture of a slight performance regression. field element operations/mul time: [60.361 ns 60.755 ns 61.170 ns] change: [+3.0611% +4.1805% +5.1398%] (p = 0.00 < 0.05) Performance has regressed. field element operations/square time: [59.143 ns 59.508 ns 59.924 ns] change: [+4.8440% +5.5441% +6.2283%] (p = 0.00 < 0.05) Performance has regressed. field element operations/invert time: [49.713 us 50.069 us 50.481 us] change: [+14.227% +15.283% +16.277%] (p = 0.00 < 0.05) Performance has regressed. field element operations/sqrt time: [24.936 us 25.064 us 25.206 us] change: [+3.0874% +3.9261% +4.7805%] (p = 0.00 < 0.05) Performance has regressed. point operations/point-scalar mul time: [532.24 us 535.29 us 538.62 us] change: [+2.4804% +3.2221% +3.9748%] (p = 0.00 < 0.05) Performance has regressed. scalar operations/sub time: [7.1834 ns 7.2272 ns 7.2688 ns] change: [+13.036% +13.817% +14.551%] (p = 0.00 < 0.05) Performance has regressed. scalar operations/add time: [9.5314 ns 9.5797 ns 9.6287 ns] change: [+1.2331% +1.9875% +2.6816%] (p = 0.00 < 0.05) Performance has regressed. scalar operations/mul time: [59.303 ns 59.530 ns 59.779 ns] change: [+5.0340% +5.6533% +6.2711%] (p = 0.00 < 0.05) Performance has regressed. scalar operations/negate time: [6.6648 ns 6.6915 ns 6.7244 ns] change: [+0.1937% +1.0485% +1.8021%] (p = 0.01 < 0.05) Change within noise threshold. scalar operations/invert time: [49.255 us 49.365 us 49.489 us] change: [+12.196% +13.079% +13.937%] (p = 0.00 < 0.05) Performance has regressed.
tarcieri
force-pushed
the
p384/const-fn-field-arithmetic
branch
from
9c4b922
to
309984a
Compare
This was referenced Jun 2, 2022
tarcieri
added a commit
that referenced
this pull request
Jun 3, 2022
Now that #589 is landed, it's possible to add `const fn` implementations of all core field operations, including conersions to/from Montgomery form as well as arithmetic operations: add, double, sub, mul, neg, square
tarcieri
added a commit
that referenced
this pull request
Jun 3, 2022
Merged
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Uses the
fiat-constifytool to mechanically transformfiat-cryptogenerated field implementations intoconst fn:https://github.com/RustCrypto/utils/tree/master/fiat-constify
Resulting code has been run through
rustfmtprior to checking it in:The following are criterion benchmarks after the translation. While these values are within the threshold of noise I experience running them on a semi-busy desktop computer (MacBook w\ 2.3 GHz 8-Core Intel Core i9) they do paint a picture of a slight performance regression (~5%).
field element operations/mul
time: [60.361 ns 60.755 ns 61.170 ns]
change: [+3.0611% +4.1805% +5.1398%] (p = 0.00 < 0.05)
Performance has regressed.
field element operations/square
time: [59.143 ns 59.508 ns 59.924 ns]
change: [+4.8440% +5.5441% +6.2283%] (p = 0.00 < 0.05)
Performance has regressed.
field element operations/invert
time: [49.713 us 50.069 us 50.481 us]
change: [+14.227% +15.283% +16.277%] (p = 0.00 < 0.05)
Performance has regressed.
field element operations/sqrt
time: [24.936 us 25.064 us 25.206 us]
change: [+3.0874% +3.9261% +4.7805%] (p = 0.00 < 0.05)
Performance has regressed.
point operations/point-scalar mul
time: [532.24 us 535.29 us 538.62 us]
change: [+2.4804% +3.2221% +3.9748%] (p = 0.00 < 0.05)
Performance has regressed.
scalar operations/sub
time: [7.1834 ns 7.2272 ns 7.2688 ns]
change: [+13.036% +13.817% +14.551%] (p = 0.00 < 0.05)
Performance has regressed.
scalar operations/add
time: [9.5314 ns 9.5797 ns 9.6287 ns]
change: [+1.2331% +1.9875% +2.6816%] (p = 0.00 < 0.05)
Performance has regressed.
scalar operations/mul
time: [59.303 ns 59.530 ns 59.779 ns]
change: [+5.0340% +5.6533% +6.2711%] (p = 0.00 < 0.05)
Performance has regressed.
scalar operations/negate
time: [6.6648 ns 6.6915 ns 6.7244 ns]
change: [+0.1937% +1.0485% +1.8021%] (p = 0.01 < 0.05)
Change within noise threshold.
scalar operations/invert
time: [49.255 us 49.365 us 49.489 us]
change: [+12.196% +13.079% +13.937%] (p = 0.00 < 0.05)
Performance has regressed.