Reviewed 4 of 38 files at r1, 2 of 7 files at r2, 2 of 2 files at r3, all commit messages.
Reviewable status: 8 of 44 files reviewed, 2 unresolved discussions (waiting on @andrewmilson)

crates/prover/Cargo.toml line 57 at r3 (raw file):

rust-2018-idioms = "deny"
unused = "deny"

Rebase and remove:

[package.metadata.cargo-machete]
ignored = ["downcast-rs"]

crates/prover/src/core/backend/cpu/quotients.rs line 80 at r3 (raw file):

/// Precomputes the complex conjugate line coefficients for each column in each sample batch.
///
/// For the `i`-th (in a sample batch) column's numerator term `alpha^i * (c * F(p) - (a * p.y +

Why am I seeing changes in comments and other changes in the code? Is it somehow related to the toolchain update?

Reviewable status: 6 of 44 files reviewed, 2 unresolved discussions (waiting on @alonh5)

crates/prover/Cargo.toml line 57 at r3 (raw file):

Done.

crates/prover/src/core/backend/cpu/quotients.rs line 80 at r3 (raw file):

The new toolchain was throwing errors for comments.
Mainly complaining the summaries were too long and needing a empty line in places.

Reviewed 16 of 38 files at r1, 4 of 7 files at r2, 18 of 18 files at r4, all commit messages.
Reviewable status: all files reviewed, 6 unresolved discussions (waiting on @andrewmilson)

.github/workflows/ci.yaml line 163 at r4 (raw file):

    steps:
      - uses: actions/checkout@v4
<<<<<<< HEAD

Resolve.

Code quote:

<<<<<<< HEAD

crates/prover/src/lib.rs line 1 at r4 (raw file):

#![allow(incomplete_features)]

Can you pleae explain the changes here?

crates/prover/src/core/backend/simd/m31.rs line 288 at r4 (raw file):

}

cfg_if::cfg_if! {

Is this also something that the new toolchain required?

crates/prover/src/core/backend/simd/lookups/gkr.rs line 28 at r4 (raw file):

        // Start DP with CPU backend to avoid dealing with instances smaller than a SIMD vector.
        let (y_rem, y_last_chunk) = y.split_last_chunk::<{ LOG_N_LANES as usize }>().unwrap();

How did you catch this? Did tests fail?

crates/prover/src/core/pcs/mod.rs line 39 at r4 (raw file):

            pow_bits: 5,
            // fri_config: FriConfig::new(0, 1, 3),
            fri_config: FriConfig::new(0, 1, 50),

On purpose?

Code quote:

            // fri_config: FriConfig::new(0, 1, 3),
            fri_config: FriConfig::new(0, 1, 50),

crates/prover/src/core/vcs/blake2s_ref.rs line 34 at r4 (raw file):

#[inline(always)]
fn rot16(x: u32) -> u32 {
    x.rotate_left(16)

Why is this left and all the others right? is it because for 16 it's identical?

Reviewable status: 41 of 44 files reviewed, 6 unresolved discussions (waiting on @alonh5)

.github/workflows/ci.yaml line 163 at r4 (raw file):

Done.

crates/prover/src/lib.rs line 1 at r4 (raw file):

stdarch_x86_avx512 replaces stdsimd as the feature to use the avx512 intrinsics.
Added as cfg_attr since stdarch_x86_avx512 feature only exists on x86

crates/prover/src/core/backend/simd/m31.rs line 288 at r4 (raw file):

Not entirely but made sense to combine the change.
The error the toolchain had was about Swizzle, InterleaveEvens, InterleaveOdds being unused on non avx2/avx512

crates/prover/src/core/backend/simd/lookups/gkr.rs line 28 at r4 (raw file):

Yes exactly.
No compile issue so was a bit perplexed at the test failures initially.

crates/prover/src/core/pcs/mod.rs line 39 at r4 (raw file):

My bad. Thanks

crates/prover/src/core/vcs/blake2s_ref.rs line 34 at r4 (raw file):

Oh funny, didn't notice this. Yeah I guess they are.

Rust analyser detected the bit shifts as a rotate and offered a quick fix which I applied.
Just changed it to rotate_right to be consistent.

Reviewed 3 of 3 files at r5, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @andrewmilson)

crates/prover/src/core/backend/simd/lookups/gkr.rs line 28 at r4 (raw file):

Weird they would change the order like that.

crates/prover/src/core/vcs/blake2s_ref.rs line 34 at r4 (raw file):

Great thanks :)

Can this PR be closed?

Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @andrewmilson)

⚠️ Performance Alert ⚠️

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

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

CC: @shaharsamocha7