start implementing permutation argument

This commit is contained in:
Alex Vlasov 2019-02-26 18:10:33 +03:00
parent b70af73621
commit c64438e800
3 changed files with 734 additions and 1 deletions

@ -22,6 +22,6 @@ Initial SONIC proof system integration using the code from the [original impleme
- [ ] Implement unhelped version - [ ] Implement unhelped version
- [x] Implement a part of S poly precomputation (S2) - [x] Implement a part of S poly precomputation (S2)
- [x] Implement a "well formed" argument - [x] Implement a "well formed" argument
- [ ] Implement a coefficients product argument - [x] Implement a coefficients product argument
- [ ] Implement a premutation argument - [ ] Implement a premutation argument
- [ ] Implement synthesizer for proper form of S polynomial - [ ] Implement synthesizer for proper form of S polynomial

@ -6,5 +6,6 @@
mod s2_proof; mod s2_proof;
mod wellformed_argument; mod wellformed_argument;
mod grand_product_argument; mod grand_product_argument;
mod permutation_argument;
pub use self::wellformed_argument::{WellformednessArgument, WellformednessProof}; pub use self::wellformed_argument::{WellformednessArgument, WellformednessProof};

@ -0,0 +1,732 @@
/// Permutation argument allows to prove that a commitment to a vector A is
/// actually a commitment to a vector of values that are equal to `(s^{perm})_i * y^{perm(i)}`
/// for some fixed permutation `perm`
use ff::{Field, PrimeField, PrimeFieldRepr, ScalarEngine};
use pairing::{Engine, CurveProjective, CurveAffine};
use std::marker::PhantomData;
use crate::sonic::srs::SRS;
use crate::sonic::util::*;
#[derive(Clone)]
pub struct SpecializedSRS<E: Engine> {
p_1: E::G1Affine,
p_2: Vec<E::G1Affine>,
p_3: E::G1Affine,
p_4: Vec<E::G1Affine>,
n: usize
}
#[derive(Clone)]
pub struct PermutationArgument<E: Engine> {
coefficients: Vec<Vec<E::Fr>>,
permutations: Vec<Vec<usize>>,
n: usize
}
#[derive(Clone)]
pub struct PermutationProof<E: Engine> {
t_opening: E::G1Affine,
e_zinv: E::Fr,
e_opening: E::G1Affine,
f_y: E::Fr,
f_opening: E::G1Affine,
}
fn permute<F: Field>(coeffs: &[F], permutation: & [usize]) -> Vec<F>{
assert_eq!(coeffs.len(), permutation.len());
let mut result: Vec<F> = vec![F::zero(); coeffs.len()];
for (i, j) in permutation.iter().enumerate() {
result[*j] = coeffs[i];
}
result
}
impl<E: Engine> PermutationArgument<E> {
pub fn new(coefficients: Vec<Vec<E::Fr>>, permutations: Vec<Vec<usize>>) -> Self {
assert!(coefficients.len() > 0);
assert_eq!(coefficients.len(), permutations.len());
let n = coefficients[0].len();
for (c, p) in coefficients.iter().zip(permutations.iter()) {
assert!(c.len() == p.len());
assert!(c.len() == n);
}
PermutationArgument {
coefficients: coefficients,
permutations: permutations,
n: n
}
}
pub fn make_specialized_srs(coefficients: &Vec<Vec<E::Fr>>, permutations: &Vec<Vec<usize>>, srs: &SRS<E>) -> SpecializedSRS<E> {
assert!(coefficients.len() > 0);
assert_eq!(coefficients.len(), permutations.len());
let n = coefficients[0].len();
// p1 is just a commitment to the powers of x
let p_1 = multiexp(srs.g_positive_x_alpha[0..n].iter(), vec![E::Fr::one(); n].iter()).into_affine();
let mut p_2 = vec![];
let p_3 = {
let values: Vec<E::Fr> = (1..=n).map(|el| {
let mut repr = <<E as ScalarEngine>::Fr as PrimeField>::Repr::default();
repr.as_mut()[0] = el as u64;
let fe = E::Fr::from_repr(repr).unwrap();
fe
}).collect();
multiexp(srs.g_positive_x_alpha[0..n].iter(), values.iter()).into_affine()
};
let mut p_4 = vec![];
for (c, p) in coefficients.iter().zip(permutations.iter()) {
assert!(c.len() == p.len());
assert!(c.len() == n);
// p2 is a commitment to the s^{perm}_i * x^i
let permuted_coeffs = permute(&c[..], &p[..]);
let p2 = multiexp(srs.g_positive_x_alpha[0..n].iter(), permuted_coeffs.iter()).into_affine();
p_2.push(p2);
{
let values: Vec<E::Fr> = p.iter().map(|el| {
let mut repr = <<E as ScalarEngine>::Fr as PrimeField>::Repr::default();
repr.as_mut()[0] = *el as u64;
let fe = E::Fr::from_repr(repr).unwrap();
fe
}).collect();
let p4 = multiexp(srs.g_positive_x_alpha[0..n].iter(), values.iter()).into_affine();
p_4.push(p4);
}
}
SpecializedSRS {
p_1: p_1,
p_2: p_2,
p_3: p_3,
p_4: p_4,
n: n
}
}
// // Make a commitment to a polynomial in a form A*B^{x+1} = [a_1...a_{n}, 0, b_1...b_{n}]
// pub fn commit_for_grand_product(a: &[E::Fr], b: &[E::Fr], srs: &SRS<E>) -> E::G1Affine {
// assert_eq!(a.len(), b.len());
// let n = a.len();
// multiexp(
// srs.g_positive_x_alpha[0..(2*n+1)].iter(),
// a.iter()
// .chain_ext(Some(E::Fr::zero()).iter())
// .chain_ext(b.iter())
// ).into_affine()
// }
// pub fn open_commitments_for_grand_product(&self, y: E::Fr, z: E::Fr, srs: &SRS<E>) -> Vec<(E::Fr, E::G1Affine)> {
// let n = self.n;
// let mut yz = y;
// yz.mul_assign(&z);
// let mut results = vec![];
// for a_poly in self.a_polynomials.iter() {
// let a = & a_poly[0..n];
// let b = & a_poly[(n+1)..];
// assert_eq!(a.len(), n);
// assert_eq!(b.len(), n);
// let mut val = evaluate_at_consequitive_powers(a, yz, yz);
// {
// let tmp = yz.pow([(n+2) as u64]);
// let v = evaluate_at_consequitive_powers(b, tmp, yz);
// val.add_assign(&v);
// }
// let mut constant_term = val;
// constant_term.negate();
// let opening = polynomial_commitment_opening(
// 0,
// 2*n + 1,
// Some(constant_term).iter()
// .chain_ext(a.iter())
// .chain_ext(Some(E::Fr::zero()).iter())
// .chain_ext(b.iter()),
// yz,
// &srs);
// results.push((val, opening));
// }
// results
// }
// // Make a commitment for the begining of the protocol, returns commitment and `v` scalar
// pub fn commit_to_individual_c_polynomials(&self, srs: &SRS<E>) -> Vec<(E::G1Affine, E::Fr)> {
// let mut results = vec![];
// let n = self.c_polynomials[0].len();
// for (p, v) in self.c_polynomials.iter().zip(self.v_elements.iter()) {
// let c = multiexp(
// srs.g_positive_x_alpha[0..n].iter(),
// p.iter()
// ).into_affine();
// results.push((c, *v));
// }
// results
// }
// // Argument is based on an approach of main SONIC construction, but with a custom S(X,Y) polynomial of a simple form
// pub fn commit_to_t_polynomial(&mut self, challenges: & Vec<E::Fr>, y: E::Fr, srs: &SRS<E>) -> E::G1Affine {
// assert_eq!(challenges.len(), self.a_polynomials.len());
// let n = self.n;
// let mut t_polynomial: Option<Vec<E::Fr>> = None;
// for (((a, c), v), challenge) in self.a_polynomials.iter()
// .zip(self.c_polynomials.iter())
// .zip(self.v_elements.iter())
// .zip(challenges.iter())
// {
// let mut a_xy = a.clone();
// let mut c_xy = c.clone();
// let v = *v;
// assert_eq!(a_xy.len(), 2*n + 1);
// assert_eq!(c_xy.len(), 2*n + 1);
// // make a T polynomial
// let r: Vec<E::Fr> = {
// // p_a(X,Y)*Y
// let mut tmp = y;
// tmp.square();
// mut_distribute_consequitive_powers(&mut a_xy[..], tmp, y);
// // add extra terms
// //v*(XY)^{n+1}*Y + X^{n+2} + X^{n+1}Y X^{2n+2}*Y
// // n+1 term v*(XY)^{n+1}*Y + X^{n+1}Y
// let tmp = y.pow(&[(n+2) as u64]);
// let mut x_n_plus_one_term = v;
// x_n_plus_one_term.mul_assign(&tmp);
// x_n_plus_one_term.add_assign(&y);
// a_xy[n].add_assign(&x_n_plus_one_term);
// // n+2 term
// a_xy[n+1].add_assign(&E::Fr::one());
// // 2n+2 term
// let mut tmp = y;
// tmp.negate();
// a_xy.push(tmp);
// assert_eq!(a_xy.len(), 2*n + 2);
// let mut r = vec![E::Fr::zero(); 2*n + 3];
// r.extend(a_xy);
// r
// };
// let r_prime: Vec<E::Fr> = {
// let mut c_prime: Vec<E::Fr> = c_xy.iter().rev().map(|el| *el).collect();
// c_prime.push(E::Fr::one());
// c_prime.push(E::Fr::zero());
// assert_eq!(c_prime.len(), 2*n + 3);
// c_prime
// };
// // multiply polynomials with powers [-2n-2, -1] and [1, 2n+2],
// // expect result to be [-2n+1, 2n+1]
// let mut t: Vec<E::Fr> = multiply_polynomials::<E>(r, r_prime);
// assert_eq!(t.len(), 6*n + 7);
// // drain first powers due to the padding and last element due to requirement of being zero
// for (i, el) in t[0..(2*n+3)].iter().enumerate() {
// assert_eq!(*el, E::Fr::zero(), "{}", format!("Element {} is non-zero", i));
// }
// t.drain(0..(2*n+3));
// let last = t.pop();
// assert_eq!(last.unwrap(), E::Fr::zero(), "last element should be zero");
// assert_eq!(t.len(), 4*n + 3);
// let mut val = {
// let mut tmp = y;
// tmp.square();
// evaluate_at_consequitive_powers(&c_xy, tmp, y)
// };
// val.add_assign(&E::Fr::one());
// // subtract at constant term
// assert_eq!(t[2*n+1], val);
// t[2*n+1].sub_assign(&val);
// if t_polynomial.is_some() {
// if let Some(t_poly) = t_polynomial.as_mut() {
// mul_add_polynomials(&mut t_poly[..], &t, *challenge);
// }
// } else {
// mul_polynomial_by_scalar(&mut t, *challenge);
// t_polynomial = Some(t);
// }
// }
// let t_polynomial = t_polynomial.unwrap();
// let c = multiexp(srs.g_negative_x_alpha[0..(2*n+1)].iter().rev()
// .chain_ext(srs.g_positive_x_alpha[0..(2*n+1)].iter()),
// t_polynomial[0..(2*n+1)].iter()
// .chain_ext(t_polynomial[(2*n+2)..].iter())).into_affine();
// self.t_polynomial = Some(t_polynomial);
// c
// }
// // Argument is based on an approach of main SONIC construction, but with a custom S(X,Y) polynomial of a simple form
// pub fn make_argument(self, a_zy: & Vec<E::Fr>, challenges: & Vec<E::Fr>, y: E::Fr, z: E::Fr, srs: &SRS<E>) -> GrandProductProof<E> {
// assert_eq!(a_zy.len(), self.a_polynomials.len());
// assert_eq!(challenges.len(), self.a_polynomials.len());
// let n = self.n;
// let c_polynomials = self.c_polynomials;
// let mut e_polynomial: Option<Vec<E::Fr>> = None;
// let mut f_polynomial: Option<Vec<E::Fr>> = None;
// let mut yz = y;
// yz.mul_assign(&z);
// let z_inv = z.inverse().unwrap();
// for (((a, c), challenge), v) in a_zy.iter()
// .zip(c_polynomials.into_iter())
// .zip(challenges.iter())
// .zip(self.v_elements.iter())
// {
// // cj = ((aj + vj(yz)n+1)y + zn+2 + zn+1y z2n+2y)z1
// let mut c_zy = yz.pow([(n + 1) as u64]);
// c_zy.mul_assign(v);
// c_zy.add_assign(a);
// c_zy.mul_assign(&y);
// let mut z_n_plus_1 = z.pow([(n + 1) as u64]);
// let mut z_n_plus_2 = z_n_plus_1;
// z_n_plus_2.mul_assign(&z);
// let mut z_2n_plus_2 = z_n_plus_1;
// z_2n_plus_2.square();
// z_2n_plus_2.mul_assign(&y);
// z_n_plus_1.mul_assign(&y);
// c_zy.add_assign(&z_n_plus_1);
// c_zy.add_assign(&z_n_plus_2);
// c_zy.sub_assign(&z_2n_plus_2);
// c_zy.mul_assign(&z_inv);
// let mut rc = c_zy;
// rc.mul_assign(challenge);
// let mut ry = y;
// ry.mul_assign(challenge);
// if e_polynomial.is_some() && f_polynomial.is_some() {
// if let Some(e_poly) = e_polynomial.as_mut() {
// if let Some(f_poly) = f_polynomial.as_mut() {
// mul_add_polynomials(&mut e_poly[..], &c, rc);
// mul_add_polynomials(&mut f_poly[..], &c, ry);
// }
// }
// } else {
// let mut e = c.clone();
// let mut f = c;
// mul_polynomial_by_scalar(&mut e, rc);
// mul_polynomial_by_scalar(&mut f, ry);
// e_polynomial = Some(e);
// f_polynomial = Some(f);
// }
// }
// let e_polynomial = e_polynomial.unwrap();
// let f_polynomial = f_polynomial.unwrap();
// // evaluate e at z^-1
// let mut e_val = evaluate_at_consequitive_powers(&e_polynomial, z_inv, z_inv);
// e_val.negate();
// // evaluate f at y
// let mut f_val = evaluate_at_consequitive_powers(&f_polynomial, y, y);
// f_val.negate();
// let e_opening = polynomial_commitment_opening(
// 0,
// 2*n + 1,
// Some(e_val).iter().chain_ext(e_polynomial.iter()),
// z_inv,
// srs);
// let f_opening = polynomial_commitment_opening(
// 0,
// 2*n + 1,
// Some(f_val).iter().chain_ext(f_polynomial.iter()),
// y,
// srs);
// e_val.negate();
// f_val.negate();
// let mut t_poly = self.t_polynomial.unwrap();
// assert_eq!(t_poly.len(), 4*n + 3);
// // largest negative power of t is -2n-1
// let t_zy = {
// let tmp = z_inv.pow([(2*n+1) as u64]);
// evaluate_at_consequitive_powers(&t_poly, tmp, z)
// };
// t_poly[2*n + 1].sub_assign(&t_zy);
// let t_opening = polynomial_commitment_opening(
// 2*n + 1,
// 2*n + 1,
// t_poly.iter(),
// z,
// srs);
// GrandProductProof {
// t_opening: t_opening,
// e_zinv: e_val,
// e_opening: e_opening,
// f_y: f_val,
// f_opening: f_opening,
// }
// }
// pub fn verify_ab_commitment(n: usize,
// randomness: & Vec<E::Fr>,
// a_commitments: &Vec<E::G1Affine>,
// b_commitments: &Vec<E::G1Affine>,
// openings: &Vec<(E::Fr, E::G1Affine)>,
// y: E::Fr,
// z: E::Fr,
// srs: &SRS<E>
// ) -> bool {
// assert_eq!(randomness.len(), a_commitments.len());
// assert_eq!(openings.len(), a_commitments.len());
// assert_eq!(b_commitments.len(), a_commitments.len());
// let d = srs.d;
// // e(Dj,hαx)e(Dyz,hα) = e(Aj,h)e(Bj,hxn+1)e(gaj ,hα)
// let g = srs.g_positive_x[0];
// let h_alpha_x_precomp = srs.h_positive_x_alpha[1].prepare();
// let h_alpha_precomp = srs.h_positive_x_alpha[0].prepare();
// let mut h_x_n_plus_one_precomp = srs.h_positive_x[n];
// h_x_n_plus_one_precomp.negate();
// let h_x_n_plus_one_precomp = h_x_n_plus_one_precomp.prepare();
// let mut h_prep = srs.h_positive_x[0];
// h_prep.negate();
// let h_prep = h_prep.prepare();
// let a = multiexp(
// a_commitments.iter(),
// randomness.iter(),
// ).into_affine();
// let a = a.prepare();
// let b = multiexp(
// b_commitments.iter(),
// randomness.iter(),
// ).into_affine();
// let b = b.prepare();
// let mut yz_neg = y;
// yz_neg.mul_assign(&z);
// yz_neg.negate();
// let mut ops = vec![];
// let mut value = E::Fr::zero();
// for (el, r) in openings.iter().zip(randomness.iter()) {
// let (v, o) = el;
// ops.push(o.clone());
// let mut val = *v;
// val.mul_assign(&r);
// value.add_assign(&val);
// }
// let value = g.mul(value.into_repr()).into_affine().prepare();
// let openings = multiexp(
// ops.iter(),
// randomness.iter(),
// ).into_affine();
// let openings_zy = openings.mul(yz_neg.into_repr()).into_affine().prepare();
// let openings = openings.prepare();
// // e(Dj,hαx)e(Dyz,hα) = e(Aj,h)e(Bj,hxn+1)e(gaj ,hα)
// E::final_exponentiation(&E::miller_loop(&[
// (&openings, &h_alpha_x_precomp),
// (&openings_zy, &h_alpha_precomp),
// (&a, &h_prep),
// (&b, &h_x_n_plus_one_precomp),
// (&value, &h_alpha_precomp)
// ])).unwrap() == E::Fqk::one()
// }
// pub fn verify(
// n: usize,
// randomness: & Vec<E::Fr>,
// a_zy: & Vec<E::Fr>,
// challenges: &Vec<E::Fr>,
// t_commitment: E::G1Affine,
// commitments: &Vec<(E::G1Affine, E::Fr)>,
// proof: &GrandProductProof<E>,
// y: E::Fr,
// z: E::Fr,
// srs: &SRS<E>
// ) -> bool {
// assert_eq!(randomness.len(), 3);
// assert_eq!(a_zy.len(), challenges.len());
// assert_eq!(commitments.len(), challenges.len());
// let d = srs.d;
// let g = srs.g_positive_x[0];
// let h_alpha_x_precomp = srs.h_positive_x_alpha[1].prepare();
// let h_alpha_precomp = srs.h_positive_x_alpha[0].prepare();
// let mut h_prep = srs.h_positive_x[0];
// h_prep.negate();
// let h_prep = h_prep.prepare();
// // first re-calculate cj and t(z,y)
// let mut yz = y;
// yz.mul_assign(&z);
// let z_inv = z.inverse().unwrap();
// let mut t_zy = E::Fr::zero();
// t_zy.add_assign(&proof.e_zinv);
// t_zy.sub_assign(&proof.f_y);
// let mut commitments_points = vec![];
// let mut rc_vec = vec![];
// let mut ry_vec = vec![];
// for ((r, commitment), a) in challenges.iter()
// .zip(commitments.iter())
// .zip(a_zy.iter()) {
// let (c, v) = commitment;
// commitments_points.push(c.clone());
// // cj = ((aj + vj(yz)n+1)y + zn+2 + zn+1y z2n+2y)z1
// let mut c_zy = yz.pow([(n + 1) as u64]);
// c_zy.mul_assign(v);
// c_zy.add_assign(a);
// c_zy.mul_assign(&y);
// let mut z_n_plus_1 = z.pow([(n + 1) as u64]);
// let mut z_n_plus_2 = z_n_plus_1;
// z_n_plus_2.mul_assign(&z);
// let mut z_2n_plus_2 = z_n_plus_1;
// z_2n_plus_2.square();
// z_2n_plus_2.mul_assign(&y);
// z_n_plus_1.mul_assign(&y);
// c_zy.add_assign(&z_n_plus_1);
// c_zy.add_assign(&z_n_plus_2);
// c_zy.sub_assign(&z_2n_plus_2);
// c_zy.mul_assign(&z_inv);
// let mut rc = c_zy;
// rc.mul_assign(&r);
// rc_vec.push(rc);
// let mut ry = y;
// ry.mul_assign(&r);
// ry_vec.push(ry);
// let mut val = rc;
// val.sub_assign(r);
// t_zy.add_assign(&val);
// }
// let c_rc = multiexp(
// commitments_points.iter(),
// rc_vec.iter(),
// ).into_affine();
// let c_ry = multiexp(
// commitments_points.iter(),
// ry_vec.iter(),
// ).into_affine();
// let mut minus_y = y;
// minus_y.negate();
// let mut f_y = proof.f_opening.mul(minus_y.into_repr());
// let g_f = g.mul(proof.f_y.into_repr());
// f_y.add_assign(&g_f);
// let mut minus_z = z;
// minus_z.negate();
// let mut t_z = proof.t_opening.mul(minus_z.into_repr());
// let g_tzy = g.mul(t_zy.into_repr());
// t_z.add_assign(&g_tzy);
// let mut minus_z_inv = z_inv;
// minus_z_inv.negate();
// let mut e_z_inv = proof.e_opening.mul(minus_z_inv.into_repr());
// let g_e = g.mul(proof.e_zinv.into_repr());
// e_z_inv.add_assign(&g_e);
// let h_alpha_term = multiexp(
// vec![e_z_inv.into_affine(), f_y.into_affine(), t_z.into_affine()].iter(),
// randomness.iter(),
// ).into_affine();
// let h_alpha_x_term = multiexp(
// Some(proof.e_opening).iter()
// .chain_ext(Some(proof.f_opening).iter())
// .chain_ext(Some(proof.t_opening).iter()),
// randomness.iter(),
// ).into_affine();
// let h_term = multiexp(
// Some(c_rc).iter()
// .chain_ext(Some(c_ry).iter())
// .chain_ext(Some(t_commitment).iter()),
// randomness.iter(),
// ).into_affine();
// E::final_exponentiation(&E::miller_loop(&[
// (&h_alpha_x_term.prepare(), &h_alpha_x_precomp),
// (&h_alpha_term.prepare(), &h_alpha_precomp),
// (&h_term.prepare(), &h_prep),
// ])).unwrap() == E::Fqk::one()
// }
}
// #[test]
// fn test_grand_product_argument() {
// use pairing::bls12_381::{Fr, G1Affine, G1, Bls12};
// use rand::{XorShiftRng, SeedableRng, Rand, Rng};
// use crate::sonic::srs::SRS;
// let srs_x = Fr::from_str("23923").unwrap();
// let srs_alpha = Fr::from_str("23728792").unwrap();
// let srs = SRS::<Bls12>::dummy(830564, srs_x, srs_alpha);
// let n: usize = 1 << 8;
// let rng = &mut XorShiftRng::from_seed([0x3dbe6259, 0x8d313d76, 0x3237db17, 0xe5bc0654]);
// let coeffs = (0..n).map(|_| Fr::rand(rng)).collect::<Vec<_>>();
// let mut permutation = coeffs.clone();
// rng.shuffle(&mut permutation);
// let a_commitment = multiexp(srs.g_positive_x_alpha[0..n].iter(), coeffs.iter()).into_affine();
// let b_commitment = multiexp(srs.g_positive_x_alpha[0..n].iter(), permutation.iter()).into_affine();
// let mut argument = GrandProductArgument::new(vec![(coeffs, permutation)]);
// let commitments_and_v_values = argument.commit_to_individual_c_polynomials(&srs);
// assert_eq!(commitments_and_v_values.len(), 1);
// let y : Fr = rng.gen();
// let challenges = (0..1).map(|_| Fr::rand(rng)).collect::<Vec<_>>();
// let t_commitment = argument.commit_to_t_polynomial(&challenges, y, &srs);
// let z : Fr = rng.gen();
// let grand_product_openings = argument.open_commitments_for_grand_product(y, z, &srs);
// let randomness = (0..1).map(|_| Fr::rand(rng)).collect::<Vec<_>>();
// let valid = GrandProductArgument::verify_ab_commitment(n,
// &randomness,
// &vec![a_commitment],
// &vec![b_commitment],
// &grand_product_openings,
// y,
// z,
// &srs);
// assert!(valid, "grand product commitments should be valid");
// let a_zy: Vec<Fr> = grand_product_openings.iter().map(|el| el.0.clone()).collect();
// let proof = argument.make_argument(&a_zy, &challenges, y, z, &srs);
// let randomness = (0..3).map(|_| Fr::rand(rng)).collect::<Vec<_>>();
// let valid = GrandProductArgument::verify(
// n,
// &randomness,
// &a_zy,
// &challenges,
// t_commitment,
// &commitments_and_v_values,
// &proof,
// y,
// z,
// &srs);
// assert!(valid, "t commitment should be valid");
// }