o1vm/pickles: Fix various FIXMEs from review

o1vm/src/pickles/column_env.rs

@@ -13,12 +13,12 @@ use super::proof::WitnessColumns;
 
 type Evals<F> = Evaluations<F, Radix2EvaluationDomain<F>>;
 
-/// The collection f polynomials (all in evaluation form) and constants
+/// The collection of polynomials (all in evaluation form) and constants
 /// required to evaluate an expression as a polynomial.
 ///
 /// All are evaluations.
 pub struct ColumnEnvironment<'a, F: FftField> {
-    /// The witness coluomn polynomials. Includes relation columns and dynamic
+    /// The witness column polynomials. Includes relation columns and dynamic
     /// selector columns.
     pub witness: &'a WitnessColumns<Evals<F>, [Evals<F>; N_MIPS_SEL_COLS]>,
     /// The value `prod_{j != 1} (1 - ω^j)`, used for efficiently

o1vm/src/pickles/prover.rs

@@ -258,8 +258,9 @@ where
         quotient
     };
 
-    // FIXME: Add this to absorb here
     let quotient_commitment = srs.commit_non_hiding(&quotient_poly, 7);
+    // FIXME: Add this to absorb here (DONE)
+    absorb_commitment(&mut fq_sponge, &quotient_commitment);
 
     ////////////////////////////////////////////////////////////////////////////
     // Round 3: Evaluations at ζ and ζω

o1vm/src/pickles/verifier.rs

@@ -2,7 +2,7 @@
 #![allow(clippy::boxed_local)]
 
 use ark_ec::{AffineRepr, Group};
-use ark_ff::{PrimeField, Zero};
+use ark_ff::{PrimeField, Zero /* One */};
 use rand::thread_rng;
 
 use kimchi::{
@@ -174,9 +174,9 @@ where
     let combined_expr =
         Expr::combine_constraints(0..(constraints.len() as u32), constraints.clone());
 
-    // FIXME: Add these to the final check!!!!!
+    // FIXME: Add these to the final check!!!!! (DONE)
 
-    // FIXME: Fixup absorbs so they match in prover.rs
+    // FIXME: Fixup absorbs so they match in prover.rs (DONE)
 
     let quotient_eval_zeta = PolishToken::evaluate(
         combined_expr.to_polish().as_slice(),
@@ -198,18 +198,18 @@ where
     )
     .unwrap_or_else(|_| panic!("Could not evaluate quotient polynomial at zeta_omega"));
 
-    // Check the actual quotient works. combined_expr(eval) [ == quotient_eval_*] = quotient(eval) [== Given by prover (new field) -- chunked] * vanishing_poly(eval) [== x^n - 1 == zeta^(d1.size()) - 1]
+    // Check the actual quotient works.
+    //
+    // combined_expr(eval) [ == quotient_eval_*]
+    // =
+    // quotient(eval) [== Given by prover (new field) -- chunked]
+    // *
+    // vanishing_poly(eval) [== x^n - 1 == zeta^(d1.size()) - 1]
+    // FIXME: This should probably use some sort of proof assert, not just panic.
+    /* assert!(quotient_eval_zeta == quotient_evaluations * (pow(zeta, d1.size()) - G::ScalarField::one()), "The prover lied!"); */
 
-    // Fixme add ft eval to the proof
-    /*     coms_and_evaluations.push(Evaluation {
-           commitment: ft_comm,
-           evaluations: vec![vec![ft_eval0], vec![zeta_omega_evaluations.ft]],
-       });
-    */
     fr_sponge.absorb(&quotient_eval_zeta);
     fr_sponge.absorb(&quotient_eval_zeta_omega);
-    // fr_sponge.absorb(zeta_omega_evaluations.ft_eval1);
-    // -- End absorb all coms_and_evaluations
 
     let v_chal = fr_sponge.challenge();
     let v = v_chal.to_field(endo_r);
@@ -241,6 +241,14 @@ where
             })
         });
 
+        evaluations.push(Evaluation {
+            commitment: quotient_commitment.clone(),
+            evaluations: vec![
+                vec![quotient_eval_zeta],
+                vec![quotient_eval_zeta_omega],
+            ],
+        });
+
         evaluations
     };