nits: add TODO issues; credits in README

a16z · Aug 9, 2023 · 9925e2f · 9925e2f
1 parent ae7b9e0
commit 9925e2f
Show file tree

Hide file tree

Showing 3 changed files with 3 additions and 4 deletions.
diff --git a/README.md b/README.md
@@ -2,7 +2,7 @@
 
 ![imgs/lasso_logo.png](imgs/lasso_logo.png)
 
-Lookup Arguments via Sum-check and Sparse polynomial commitments, including for Oversized tables.
+Lookup Arguments via Sum-check and Sparse polynomial commitments, including for Oversized tables. Code originally derived from [Spartan](https://github.com/microsoft/Spartan) by Srinath Setty.
 
 ## Overview
 
@@ -31,7 +31,7 @@ Lookup Arguments via Sum-check and Sparse polynomial commitments, including for
 -   `cargo build --release`
 -   `cargo run --release -- --name <bench_name>`
 -   `cargo run --release -- --name <bench_name> --chart`: Display performance gant chart
--   `cargo run --release --name <bench_name> --features ark-msm`: Run without MSM small field optimizations
+-   `cargo run --release --features ark-msm -- --name <bench_name>`: Run without MSM small field optimizations
 -   `sudo cargo flamegraph`
 
 _Note on benching / flamegraphing: Turn off the parallel feature in Cargo.toml (`multicore`) and / or `export RAYON_NUM_THREADS=1` to make flamegraph more interpretable. Turning off `multicore` and leaving `unset RAYON_NUM_THREADS` allows testing Arkworks MSM parallelism without Lasso parallelism._

diff --git a/src/poly/dense_mlpoly.rs b/src/poly/dense_mlpoly.rs
@@ -239,7 +239,6 @@ impl<F: PrimeField> DensePolynomial<F> {
   }
 
   pub fn extend(&mut self, other: &DensePolynomial<F>) {
-    // TODO: allow extension even when some vars are bound
     assert_eq!(self.Z.len(), self.len);
     let other_vec = other.vec();
     assert_eq!(other_vec.len(), self.len);

diff --git a/src/subprotocols/sumcheck.rs b/src/subprotocols/sumcheck.rs
@@ -215,7 +215,7 @@ impl<F: PrimeField> SumcheckInstanceProof<F> {
         accum
       }).collect();
 
-      // TODO: This can be parallelized now
+      // TODO(#31): Parallelize
       for poly_i in 0..(combined_degree + 1) {
         for mle_i in 0..mle_half {
           eval_points[poly_i] += accum[mle_i][poly_i];