CPU Barycentric Evaluation and Test

[Gali-StarkWare]

No description provided.

[reviewable-StarkWare]

This change is 

[Gali-StarkWare]

• Replace OODS Eval Method #1074
• Barycentric Evaluation Bench #1073
• Parallel Barycentric and Weights #1072
• Save Barycentric Weights in Bit Reverse Order #1071
• Barycentric Evaluation in Poly Ops #1070
• SIMD Barycentric Evaluation and Test #1069
• CPU Barycentric Evaluation and Test #1068 👈 (View in Graphite)
• gali/barycentric

This stack of pull requests is managed by Graphite. Learn more about stacking.

[shaharsamocha7]

Reviewed 1 of 2 files at r1, all commit messages.
Reviewable status: 1 of 2 files reviewed, 3 unresolved discussions (waiting on @Alon-Ti)

---

crates/prover/src/core/constraints.rs line 37 at r1 (raw file):

}

pub fn coset_vanishing_derivative<F: ExtensionOf<BaseField>>(coset: Coset, p: CirclePoint<F>) -> F {

Which points are zeros for this function?
Why do we have exp = 4**(log_size - 1)?
document

Code quote:

pub fn coset_vanishing_derivative<F: ExtensionOf<BaseField>>(coset: Coset, p: CirclePoint<F>) -> F {

---

crates/prover/src/core/poly/circle/evaluation.rs line 165 at r1 (raw file):

// TODO(Gali): Remove.
#[allow(dead_code)]
fn weights(log_size: u32, sample_point: CirclePoint<SecureField>) -> Col<CpuBackend, SecureField> {

Is this going to be removed?
Or do we want to keep trivial weight calculation?

Code quote:

#[allow(dead_code)]
fn weights(log_size: u32, sample_point: CirclePoint<SecureField>) -> Col<CpuBackend, SecureField> {

---

crates/prover/src/core/poly/circle/evaluation.rs line 303 at r1 (raw file):

                .to_vec(),
        );
        let s = CanonicCoset::new(3);

Can we evaluate on a larger CanonicCoset here?

Code quote:

        let s = CanonicCoset::new(3);

[shaharsamocha7]

Reviewable status: 1 of 2 files reviewed, 3 unresolved discussions (waiting on @Alon-Ti and @Gali-StarkWare)

---

crates/prover/src/core/poly/circle/evaluation.rs line 165 at r1 (raw file):

Previously, shaharsamocha7 wrote…

Is this going to be removed?
Or do we want to keep trivial weight calculation?

also we would probably need the talk to understand this code

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 99.26471% with 1 line in your changes missing coverage. Please review.

Project coverage is 92.70%. Comparing base (61d1f6b) to head (bb9e718).

Files with missing lines	Patch %	Lines
crates/prover/src/core/poly/circle/evaluation.rs	99.16%	1 Missing ⚠️

Additional details and impacted files

@@                 Coverage Diff                  @@
##           gali/barycentric    #1068      +/-   ##
====================================================
+ Coverage             92.64%   92.70%   +0.06%     
====================================================
  Files                   106      106              
  Lines                 14273    14409     +136     
  Branches              14273    14409     +136     
====================================================
+ Hits                  13223    13358     +135     
- Misses                  969      970       +1     
  Partials                 81       81              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 2.

Benchmark suite	Current: bb9e718	Previous: 8b5436d	Ratio
iffts/simd ifft/27	604473857 ns/iter (± 10795101)	277030291 ns/iter (± 5776193)	2.18
iffts/simd ifft/28	1223931924 ns/iter (± 18791726)	586165718 ns/iter (± 12065155)	2.09

This comment was automatically generated by workflow using github-action-benchmark.

CC: @shaharsamocha7

[Gali-StarkWare]

Reviewable status: 0 of 2 files reviewed, 3 unresolved discussions (waiting on @Alon-Ti, @Gali-StarkWare, and @shaharsamocha7)

---

crates/prover/src/core/constraints.rs line 37 at r1 (raw file):

Previously, shaharsamocha7 wrote…

Which points are zeros for this function?
Why do we have exp = 4**(log_size - 1)?
document

Done.

---

crates/prover/src/core/poly/circle/evaluation.rs line 165 at r1 (raw file):

Previously, shaharsamocha7 wrote…

also we would probably need the talk to understand this code

I am in favor of keeping the cpu weight calculation, like you have the cpu version of eval_at_point

---

crates/prover/src/core/poly/circle/evaluation.rs line 303 at r1 (raw file):

Previously, shaharsamocha7 wrote…

Can we evaluate on a larger CanonicCoset here?

Will do

[shaharsamocha7]

Reviewed all commit messages.
Reviewable status: 0 of 2 files reviewed, 1 unresolved discussion (waiting on @Alon-Ti and @Gali-StarkWare)