1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
// Copyright (C) 2019-2022 Aleo Systems Inc.
// This file is part of the snarkVM library.

// The snarkVM library is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.

// The snarkVM library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.

// You should have received a copy of the GNU General Public License
// along with the snarkVM library. If not, see <https://www.gnu.org/licenses/>.

pub mod batched;
pub mod standard;

#[cfg(all(feature = "cuda", target_arch = "x86_64"))]
mod cuda;

#[cfg(target_arch = "x86_64")]
pub mod prefetch;

use snarkvm_curves::{bls12_377::G1Affine, traits::AffineCurve};
use snarkvm_fields::PrimeField;

use core::any::TypeId;

#[cfg(all(feature = "cuda", target_arch = "x86_64"))]
use core::sync::atomic::{AtomicBool, Ordering};

#[cfg(all(feature = "cuda", target_arch = "x86_64"))]
static HAS_CUDA_FAILED: AtomicBool = AtomicBool::new(false);

pub struct VariableBase;

impl VariableBase {
    pub fn msm<G: AffineCurve>(bases: &[G], scalars: &[<G::ScalarField as PrimeField>::BigInteger]) -> G::Projective {
        // For BLS12-377, we perform variable base MSM using a batched addition technique.
        if TypeId::of::<G>() == TypeId::of::<G1Affine>() {
            #[cfg(all(feature = "cuda", target_arch = "x86_64"))]
            if !HAS_CUDA_FAILED.load(Ordering::SeqCst) {
                match cuda::msm_cuda(bases, scalars) {
                    Ok(x) => return x,
                    Err(_e) => {
                        HAS_CUDA_FAILED.store(true, Ordering::SeqCst);
                        eprintln!("CUDA failed, moving to the next MSM method");
                    }
                }
            }
            batched::msm(bases, scalars)
        }
        // For all other curves, we perform variable base MSM using Pippenger's algorithm.
        else {
            standard::msm(bases, scalars)
        }
    }

    #[cfg(test)]
    fn msm_naive<G: AffineCurve>(bases: &[G], scalars: &[<G::ScalarField as PrimeField>::BigInteger]) -> G::Projective {
        use itertools::Itertools;
        use snarkvm_utilities::BitIteratorBE;

        bases.iter().zip_eq(scalars).map(|(base, scalar)| base.mul_bits(BitIteratorBE::new(*scalar))).sum()
    }

    #[cfg(test)]
    fn msm_naive_parallel<G: AffineCurve>(
        bases: &[G],
        scalars: &[<G::ScalarField as PrimeField>::BigInteger],
    ) -> G::Projective {
        use rayon::prelude::*;
        use snarkvm_utilities::BitIteratorBE;

        bases.par_iter().zip_eq(scalars).map(|(base, scalar)| base.mul_bits(BitIteratorBE::new(*scalar))).sum()
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use snarkvm_curves::bls12_377::{Fr, G1Affine};
    use snarkvm_fields::PrimeField;
    use snarkvm_utilities::rand::TestRng;

    fn create_scalar_bases<G: AffineCurve<ScalarField = F>, F: PrimeField>(
        rng: &mut TestRng,
        size: usize,
    ) -> (Vec<G>, Vec<F::BigInteger>) {
        let bases = (0..size).map(|_| G::rand(rng)).collect::<Vec<_>>();
        let scalars = (0..size).map(|_| F::rand(rng).to_bigint()).collect::<Vec<_>>();
        (bases, scalars)
    }

    #[test]
    fn test_msm() {
        use snarkvm_curves::ProjectiveCurve;
        for msm_size in [1, 5, 10, 50, 100, 500, 1000] {
            let mut rng = TestRng::default();
            let (bases, scalars) = create_scalar_bases::<G1Affine, Fr>(&mut rng, msm_size);

            let naive_a = VariableBase::msm_naive(bases.as_slice(), scalars.as_slice()).to_affine();
            let naive_b = VariableBase::msm_naive_parallel(bases.as_slice(), scalars.as_slice()).to_affine();
            assert_eq!(naive_a, naive_b, "MSM size: {msm_size}");

            let candidate = standard::msm(bases.as_slice(), scalars.as_slice()).to_affine();
            assert_eq!(naive_a, candidate, "MSM size: {msm_size}");

            let candidate = batched::msm(bases.as_slice(), scalars.as_slice()).to_affine();
            assert_eq!(naive_a, candidate, "MSM size: {msm_size}");
        }
    }

    #[cfg(all(feature = "cuda", target_arch = "x86_64"))]
    #[test]
    fn test_msm_cuda() {
        let mut rng = TestRng::default();
        for _ in 0..100 {
            let (bases, scalars) = create_scalar_bases::<G1Affine, Fr>(&mut rng, 1 << 10);
            let rust = standard::msm(bases.as_slice(), scalars.as_slice());

            let cuda = cuda::msm_cuda(bases.as_slice(), scalars.as_slice()).unwrap();
            assert_eq!(rust, cuda);
        }
    }
}