Extension field #44

lonerapier · 2024-05-07T00:19:30Z

This PR closes issue #40

It changes the following:

adds 2 degree extension field for PlutoField. I've not made field extension generic over degree. let me know if that is required.

0xJepsen · 2024-05-07T14:06:49Z

src/field/gf_101_2.rs

+use crate::field::FiniteField;
+
+/// Pluto curve with modulus 101 supports two degree extension field. This can be verified
+/// by finding out embedding degree of the curve, i.e. smallest k: r|q^k-1


When reading about the embedding degree I was trying to see this by hand but came across some questions. In our case r is our curve order which is 17 and p is field order which is 101. How does 17 | 101 ^2 - 1 which is 17 | 10200. 17 doesn't divide 10200. Also 10200 mod 101 is just 17 | 100. Do I have an error in my arithmetic? Are we working in a field here?

17 doesn't divide 10200.

Umm, Waylon 17 does divide 10200.

In our case r is our curve order which is 17

Also, our curve order is 102 (101 points + 1 point at infinity). 17 is the largest prime factor of $r$, thus, 17 is the order of the largest subgroup generated by the curve. Verified using this sage script.

E1 = EllipticCurve(F, [0, 3]) print(E1.order()) P1 = E1(1, 2) print(P1.order()) k = 17 cofactor = 6 # use cofactor clearing to find generator of prime order subgroup # take a random point, and multiply by cofactor until it's INF point G = E1.random_point() INF = E1(0) print(G, cofactor * G) while cofactor * G == INF: print("as") G = E1.random_point() Ps = cofactor * G print(cofactor * G, Ps.order())

that's a classic 😅

0xJepsen · 2024-05-07T14:08:06Z

src/field/gf_101_2.rs

+#[derive(Clone, Default, Copy, Eq, PartialEq, Hash, Debug, Serialize, Deserialize)]
+
+/// Quadratic Extension field element represented as polynomial of degree 1 in form:
+/// a_0 + a_1*t where {a_0, a_1} \in \mathhbb{F}. Uses irreducible poly of the form:


I think this is the right approach but curious how or if we should use this in our polynomial module. I know @Autoparallel you have been working on the FFTs and polynomial interface. Maybe you have some insight here?

Let me get the polynomial PR up.

0xJepsen · 2024-05-07T14:10:58Z

src/field/gf_101_2.rs

+
+/// Quadratic Extension field element represented as polynomial of degree 1 in form:
+/// a_0 + a_1*t where {a_0, a_1} \in \mathhbb{F}. Uses irreducible poly of the form:
+/// (X^2-K).


should this be X^2 -K or X^2 + K? X^2 + 2 is irreducible in F_{101} and should be used to generate our extension field.

I think both work as (correct me if I'm wrong) 2 is not a square in GF_{101}.

yeah, both of them are irreducible, added it in the field sage script as well

ronkathon/math/field.sage

Lines 47 to 51 in 9c0ba70

Ft.<t> = F[]

# irreducible element: t^2-2

P = Ft(t ^ 2 - 2)

assert P.is_irreducible()

src/field/gf_101_2.rs

Autoparallel

I'm good with this to go in as is :)

Autoparallel · 2024-05-07T15:02:03Z

src/field/gf_101.rs

@@ -83,6 +83,8 @@ impl FiniteField for GF101 {
  fn neg_one() -> Self { Self::new(Self::ORDER - 1) }

  fn generator() -> Self { Self::new(2) }
+
+  fn from_canonical_u32(n: u32) -> Self { Self::new(n) } // TODO: recheck


This is a nit, but we don't really need this if we just use GF101::new() right?

but then that's struct specific function right? We need a trait specific function to create new values, for example: in extension fields, each subvalue in extension field element is base field element.

ronkathon/src/field/gf_101_2.rs

Line 69 in 9c0ba70

Self { value: [F::from_canonical_u32(15), F::from_canonical_u32(20)] }

Is there a better way in Rust to do this? Sorry, still in my learning path for Rust.

Autoparallel · 2024-05-07T15:03:01Z

src/field/gf_101_2.rs

+#[derive(Clone, Default, Copy, Eq, PartialEq, Hash, Debug, Serialize, Deserialize)]
+
+/// Quadratic Extension field element represented as polynomial of degree 1 in form:
+/// a_0 + a_1*t where {a_0, a_1} \in \mathhbb{F}. Uses irreducible poly of the form:


Let me get the polynomial PR up.

Autoparallel · 2024-05-07T15:03:36Z

src/field/gf_101_2.rs

+
+/// Quadratic Extension field element represented as polynomial of degree 1 in form:
+/// a_0 + a_1*t where {a_0, a_1} \in \mathhbb{F}. Uses irreducible poly of the form:
+/// (X^2-K).


I think both work as (correct me if I'm wrong) 2 is not a square in GF_{101}.

src/field/gf_101_2.rs

lonerapier · 2024-05-07T18:28:59Z

@0xJepsen @Autoparallel changes done. let me know if any others are needed

commit 2637594 Author: Waylon Jepsen <57912727+0xJepsen@users.noreply.github.com> Date: Thu May 9 10:38:12 2024 -0600 curves: Second curve group and some test (#49) * chore: tests + scalar multiplication * fmt * wip: compiles, but generator point is broken * wip * fix mul, generator and add hardcoded tests (#51) * fix mul, generator and add hardcoded tests * fix generator test * wip * fix: test fails when rng has no inverse --------- Co-authored-by: Sambhav <lonerapier@proton.me> commit def9371 Author: Waylon Jepsen <57912727+0xJepsen@users.noreply.github.com> Date: Thu May 9 09:36:25 2024 -0600 docs: readme (#46) * docs: readme * wip: pairing check * docs * Update README.md Co-authored-by: Colin Roberts <colin@autoparallel.xyz> * Update README.md Co-authored-by: Colin Roberts <colin@autoparallel.xyz> --------- Co-authored-by: Colin Roberts <colin@autoparallel.xyz> commit ab0d374 Author: Sambhav <lonerapier@proton.me> Date: Wed May 8 03:00:15 2024 +0530 Extension field (#44) * feat: extension field init * add test * add more tests and comments * move pow impl default to trait commit f9aa66e Author: Waylon Jepsen <57912727+0xJepsen@users.noreply.github.com> Date: Tue May 7 09:00:35 2024 -0600 curves in sage (#39) commit 63ce62f Author: Waylon Jepsen <waylonjepsen1@gmail.com> Date: Mon May 6 17:46:22 2024 -0600 fix: lock commit d1c84eb Author: Colin Roberts <colin@autoparallel.xyz> Date: Mon May 6 16:39:25 2024 -0700 feat: home-baked `FiniteField` trait (#38) * feat: new `FiniteField` trait Now everything compiles again. Will work to clean this all up and get all the tests to pass. * fix: `GF101` tests pass * fix: reimplement monty optimizations * clean: udeps --------- Co-authored-by: Waylon Jepsen <57912727+0xJepsen@users.noreply.github.com> commit 96c8b66 Merge: fb27e5f 62a9a57 Author: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Date: Mon May 6 23:26:56 2024 +0000 Merge pull request #42 from pluto/dependabot/cargo/anyhow-1.0.83 Bump anyhow from 1.0.82 to 1.0.83 commit 62a9a57 Author: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Date: Mon May 6 23:26:43 2024 +0000 Bump anyhow from 1.0.82 to 1.0.83 Bumps [anyhow](https://github.com/dtolnay/anyhow) from 1.0.82 to 1.0.83. - [Release notes](https://github.com/dtolnay/anyhow/releases) - [Commits](dtolnay/anyhow@1.0.82...1.0.83) --- updated-dependencies: - dependency-name: anyhow dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com>

lonerapier added 2 commits May 7, 2024 05:48

feat: extension field init

37b8f66

add test

50651e1

0xJepsen requested a review from Autoparallel May 7, 2024 01:16

add more tests and comments

9c0ba70

lonerapier force-pushed the extension_field branch from 0524294 to 9c0ba70 Compare May 7, 2024 06:30

0xJepsen reviewed May 7, 2024

View reviewed changes

Autoparallel approved these changes May 7, 2024

View reviewed changes

0xJepsen linked an issue May 7, 2024 that may be closed by this pull request

Extension fields #40

Closed

move pow impl default to trait

8f93f8b

0xJepsen merged commit ab0d374 into pluto:main May 7, 2024
4 checks passed

lonerapier deleted the extension_field branch May 8, 2024 05:16

github-actions bot mentioned this pull request Jul 1, 2024

chore: release v0.1.0 #111

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Extension field #44

Extension field #44

lonerapier commented May 7, 2024 •

edited

Loading

0xJepsen May 7, 2024

lonerapier May 7, 2024

0xJepsen May 7, 2024 •

edited

Loading

0xJepsen May 7, 2024

Autoparallel May 7, 2024

0xJepsen May 7, 2024

Autoparallel May 7, 2024

lonerapier May 7, 2024

Autoparallel left a comment

Autoparallel May 7, 2024

lonerapier May 7, 2024

Autoparallel May 7, 2024

Autoparallel May 7, 2024

lonerapier commented May 7, 2024

	Ft.<t> = F[]

	# irreducible element: t^2-2
	P = Ft(t ^ 2 - 2)
	assert P.is_irreducible()

Extension field #44

Extension field #44

Conversation

lonerapier commented May 7, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

0xJepsen May 7, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Autoparallel left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

lonerapier commented May 7, 2024

lonerapier commented May 7, 2024 •

edited

Loading

0xJepsen May 7, 2024 •

edited

Loading