fft/engines/crt.hpp
| Line | Branch | Exec | Source |
|---|---|---|---|
| 1 | #pragma once | ||
| 2 | |||
| 3 | #include <cassert> | ||
| 4 | #include <cstdint> | ||
| 5 | #include <span> | ||
| 6 | #include <utility> | ||
| 7 | |||
| 8 | #include "fft/engine.hpp" | ||
| 9 | #include "fft/engines/ntt.hpp" | ||
| 10 | #include "modnum.hpp" | ||
| 11 | |||
| 12 | namespace ecnerwala::fft::engines { | ||
| 13 | |||
| 14 | // Multiplies mod `mnum` by running NTTs modulo two FFT-friendly primes and CRT'ing. | ||
| 15 | // Inputs use balanced representatives (|v| <= MOD/2), so the true integer coefficients | ||
| 16 | // are bounded by n (MOD/2)^2. | ||
| 17 | template <typename mnum, typename num1 = mod_goldilocks, typename num2 = modnum<(15 << 27) + 1>> | ||
| 18 | struct crt { | ||
| 19 | static_assert(sizeof(decltype(mnum::MOD)) <= 4, "n (MOD/2)^2 must fit the CRT modulus product"); | ||
| 20 | using value_type = mnum; | ||
| 21 | static constexpr bool commutative = true; | ||
| 22 | static constexpr int unit_scale = 1; | ||
| 23 | using E1 = ntt<num1>; | ||
| 24 | using E2 = ntt<num2>; | ||
| 25 |
1/2✓ Branch 9 → 10 taken 48 times.
✗ Branch 9 → 14 not taken.
|
96 | template <int A = 1> struct transformed_t { |
| 26 | typename E1::transformed t1; | ||
| 27 | typename E2::transformed t2; | ||
| 28 |
2/2✓ Branch 18 → 19 taken 159 times.
✓ Branch 18 → 20 taken 58 times.
|
3469 | int size() const { return t1.size(); } |
| 29 | 156 | transformed_t() = default; | |
| 30 | 1062 | transformed_t(typename E1::transformed&& t1_, typename E2::transformed&& t2_) | |
| 31 | 2124 | : t1(std::move(t1_)), t2(std::move(t2_)) {} | |
| 32 | ✗ | template <int A2> requires (A2 != A) explicit(A2 > A) transformed_t(transformed_t<A2>&& o) | |
| 33 | ✗ | : t1(std::move(o.t1)), t2(std::move(o.t2)) {} | |
| 34 | }; | ||
| 35 | using transformed = transformed_t<1>; | ||
| 36 | 48 | template <int K> struct product_t { | |
| 37 | typename E1::product p1; | ||
| 38 | typename E2::product p2; | ||
| 39 | 3811 | int size() const { return sz(p1); } | |
| 40 | ✗ | product_t() = default; | |
| 41 | 4019 | product_t(typename E1::product&& p1_, typename E2::product&& p2_) | |
| 42 | 7990 | : p1(std::move(p1_)), p2(std::move(p2_)) {} | |
| 43 | 20 | template <int K2> requires (K2 != K) explicit(K2 > K) product_t(product_t<K2>&& o) | |
| 44 | 40 | : p1(std::move(o.p1)), p2(std::move(o.p2)) {} | |
| 45 | }; | ||
| 46 | using product = product_t<1>; | ||
| 47 | |||
| 48 | 895 | static transformed transform(std::span<const mnum> a, int n) { | |
| 49 | 895 | assert(sz(a) <= 2 * n); | |
| 50 |
1/1✓ Branch 7 → 8 taken 895 times.
|
895 | auto b1 = buffer_pool<num1>::get(sz(a)); |
| 51 |
1/1✓ Branch 10 → 42 taken 895 times.
|
895 | auto b2 = buffer_pool<num2>::get(sz(a)); |
| 52 |
2/2✓ Branch 43 → 11 taken 13618 times.
✓ Branch 43 → 44 taken 895 times.
|
55367 | for (int i = 0; i < sz(a); i++) { int64_t v = a[i].balanced(); b1[i] = num1(v); b2[i] = num2(v); } |
| 53 | return transformed{ | ||
| 54 |
1/1✓ Branch 54 → 55 taken 895 times.
|
2685 | E1::transform(std::span<const num1>(b1.span()), n), |
| 55 |
1/1✓ Branch 66 → 67 taken 895 times.
|
2685 | E2::transform(std::span<const num2>(b2.span()), n), |
| 56 | 1790 | }; | |
| 57 | 895 | } | |
| 58 |
1/2✗ Branch 2 → 3 not taken.
✓ Branch 2 → 4 taken 96 times.
|
3035 | static void extend_to(transformed& t, int m, std::span<const mnum> coeffs) { |
| 59 |
3/4✗ Branch 2 → 3 not taken.
✓ Branch 2 → 4 taken 96 times.
✓ Branch 10 → 11 taken 1000 times.
✓ Branch 10 → 13 taken 1939 times.
|
5974 | if (t.size() >= m) return; |
| 60 |
1/1✓ Branch 15 → 16 taken 1939 times.
|
2035 | auto b1 = buffer_pool<num1>::get(sz(coeffs)); |
| 61 |
2/2✓ Branch 5 → 15 taken 96 times.
✓ Branch 18 → 50 taken 1939 times.
|
2035 | auto b2 = buffer_pool<num2>::get(sz(coeffs)); |
| 62 |
10/10✓ Branch 6 → 7 taken 11556440 times.
✓ Branch 6 → 8 taken 10521129 times.
✓ Branch 8 → 9 taken 11556440 times.
✓ Branch 8 → 10 taken 10521129 times.
✓ Branch 11 → 12 taken 11556440 times.
✓ Branch 11 → 13 taken 10521129 times.
✓ Branch 15 → 6 taken 22077569 times.
✓ Branch 15 → 16 taken 96 times.
✓ Branch 51 → 19 taken 7639 times.
✓ Branch 51 → 52 taken 1939 times.
|
55744169 | for (int i = 0; i < sz(coeffs); i++) { int64_t v = coeffs[i].balanced(); b1[i] = num1(v); b2[i] = num2(v); } |
| 63 |
3/4✓ Branch 18 → 19 taken 96 times.
✗ Branch 19 → 20 not taken.
✓ Branch 19 → 21 taken 96 times.
✓ Branch 63 → 64 taken 1939 times.
|
3974 | E1::extend_to(t.t1, m, std::span<const num1>(b1.span())); |
| 64 |
2/2✓ Branch 21 → 22 taken 96 times.
✓ Branch 78 → 79 taken 1939 times.
|
3974 | E2::extend_to(t.t2, m, std::span<const num2>(b2.span())); |
| 65 | 2035 | } | |
| 66 | ✗ | template <int A> static transformed_t<A> downsample(const transformed_t<A>& t, int n, bool odd) { | |
| 67 | ✗ | return transformed_t<A>{E1::downsample(t.t1, n, odd), E2::downsample(t.t2, n, odd)}; | |
| 68 | } | ||
| 69 | 294 | template <int K> static product_t<K> downsample(const product_t<K>& p, int n, bool odd) { | |
| 70 |
2/2✓ Branch 6 → 7 taken 294 times.
✓ Branch 14 → 15 taken 294 times.
|
1176 | return product_t<K>{E1::downsample(p.p1, n, odd), E2::downsample(p.p2, n, odd)}; |
| 71 | } | ||
| 72 | 164 | template <int A> static transformed_t<A> negate_arg(const transformed_t<A>& t, int n) { | |
| 73 |
2/2✓ Branch 6 → 7 taken 164 times.
✓ Branch 14 → 15 taken 164 times.
|
656 | return transformed_t<A>{E1::negate_arg(t.t1, n), E2::negate_arg(t.t2, n)}; |
| 74 | } | ||
| 75 | // Exact per prime; the scale tracks the true (integer) coefficient growth. | ||
| 76 | 3 | template <int A, int B> static transformed_t<A + B> add(transformed_t<A>&& a, const transformed_t<B>& b) { | |
| 77 | 12 | return transformed_t<A + B>{E1::add(std::move(a.t1), b.t1), E2::add(std::move(a.t2), b.t2)}; | |
| 78 | } | ||
| 79 | 1860 | template <int A, int B> static product_t<A * B> mul(const transformed_t<A>& a, const transformed_t<B>& b, int n) { | |
| 80 |
5/5ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<(1)*(1)> ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::mul<1, 1>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::transformed_t<1> const&, ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::transformed_t<1> const&, int):
✓ Branch 3 → 4 taken 48 times.
✓ Branch 9 → 10 taken 1809 times.
✓ Branch 22 → 23 taken 1809 times.
ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<(2)*(1)> ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::mul<2, 1>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::transformed_t<2> const&, ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::transformed_t<1> const&, int):
✓ Branch 9 → 10 taken 3 times.
✓ Branch 22 → 23 taken 3 times.
|
7296 | return product_t<A * B>{E1::mul(a.t1, b.t1, n), E2::mul(a.t2, b.t2, n)}; |
| 81 | } | ||
| 82 | 23 | template <int A> static product_t<A * A> sq(const transformed_t<A>& a, int n) { return mul(a, a, n); } | |
| 83 | template <int A1, int B1, int A2, int B2> | ||
| 84 | 1761 | static product_t<A1 * B1 + A2 * B2> mul2( | |
| 85 | const transformed_t<A1>& a1, const transformed_t<B1>& b1, | ||
| 86 | const transformed_t<A2>& a2, const transformed_t<B2>& b2, | ||
| 87 | int n | ||
| 88 | ) { | ||
| 89 | return product_t<A1 * B1 + A2 * B2>{ | ||
| 90 |
1/1✓ Branch 15 → 16 taken 1761 times.
|
3522 | E1::mul2(a1.t1, b1.t1, a2.t1, b2.t1, n), |
| 91 |
1/1✓ Branch 38 → 39 taken 1761 times.
|
3522 | E2::mul2(a1.t2, b1.t2, a2.t2, b2.t2, n), |
| 92 | 3522 | }; | |
| 93 | } | ||
| 94 | 104 | template <int K1, int K2> static product_t<K1 + K2> add(product_t<K1>&& a, product_t<K2>&& b) { | |
| 95 | 416 | return product_t<K1 + K2>{E1::add(std::move(a.p1), b.p1), E2::add(std::move(a.p2), b.p2)}; | |
| 96 | } | ||
| 97 | 3523 | template <int K = 1, typename Op = assign_op> static void finish(product_t<K>&& p, std::span<mnum> out, Op op = {}) { | |
| 98 | // The reconstruction needs |c| < whole/2; balanced inputs bound each addend's | ||
| 99 | // true coefficients by n (MOD/2)^2, so the safe length is divided by the | ||
| 100 | // accumulated scale. K <= 2 is very conservative (~2^35 even for MOD ~ 2^30). | ||
| 101 | static_assert(K <= 2, "crt: accumulated scale too large"); | ||
| 102 | 3523 | int n = p.size(); | |
| 103 | 3523 | assert(sz(out) <= n); | |
| 104 |
10/10void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_twice_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_twice_op):
✓ Branch 12 → 13 taken 238 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 12 → 13 taken 558 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op>):
✓ Branch 12 → 13 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 12 → 13 taken 6 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 12 → 13 taken 12 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 12 → 13 taken 689 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 12 → 13 taken 1192 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 12 → 13 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 12 → 13 taken 5 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 12 → 13 taken 607 times.
|
3523 | auto o1 = buffer_pool<num1>::get(sz(out)); |
| 105 |
14/16void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_twice_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_twice_op):
✓ Branch 15 → 16 taken 238 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 15 → 16 taken 558 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op>):
✓ Branch 15 → 16 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 15 → 16 taken 6 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 5 → 6 taken 8 times.
✗ Branch 6 → 7 not taken.
✓ Branch 6 → 8 taken 8 times.
✓ Branch 15 → 16 taken 12 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 5 → 6 taken 40 times.
✗ Branch 6 → 7 not taken.
✓ Branch 6 → 8 taken 40 times.
✓ Branch 15 → 16 taken 689 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 15 → 16 taken 1192 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 15 → 16 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 15 → 16 taken 5 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 15 → 16 taken 607 times.
|
3523 | auto o2 = buffer_pool<num2>::get(sz(out)); |
| 106 |
12/12void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_twice_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_twice_op):
✓ Branch 23 → 24 taken 238 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 23 → 24 taken 558 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op>):
✓ Branch 23 → 24 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 23 → 24 taken 6 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 8 → 9 taken 8 times.
✓ Branch 23 → 24 taken 12 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 8 → 9 taken 40 times.
✓ Branch 23 → 24 taken 689 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 23 → 24 taken 1192 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 23 → 24 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 23 → 24 taken 5 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 23 → 24 taken 607 times.
|
3523 | E1::finish(std::move(p.p1), o1.span()); |
| 107 |
12/12void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_twice_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_twice_op):
✓ Branch 32 → 33 taken 238 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 32 → 33 taken 558 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op>):
✓ Branch 32 → 33 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 32 → 33 taken 6 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 11 → 12 taken 8 times.
✓ Branch 32 → 33 taken 12 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 11 → 12 taken 40 times.
✓ Branch 32 → 33 taken 689 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 32 → 33 taken 1192 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 32 → 33 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 32 → 33 taken 5 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 32 → 33 taken 607 times.
|
3523 | E2::finish(std::move(p.p2), o2.span()); |
| 108 | |||
| 109 | // TODO: Could hardcode these | ||
| 110 | 10473 | num1 inv_n2 = inv(num1(num2::MOD)); | |
| 111 | 10473 | num2 inv_n1 = inv(num2(num1::MOD)); | |
| 112 | 3523 | __int128_t whole = __int128_t(num1::MOD) * __int128_t(num2::MOD); | |
| 113 | |||
| 114 | 3523 | mnum m1_mod = mnum(num1::MOD); | |
| 115 | 3523 | mnum m2_mod = mnum(num2::MOD); | |
| 116 | 3523 | mnum whole_mod = m1_mod * m2_mod; | |
| 117 |
24/24void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_twice_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_twice_op):
✓ Branch 118 → 66 taken 938 times.
✓ Branch 118 → 119 taken 238 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 120 → 66 taken 12989 times.
✓ Branch 120 → 121 taken 558 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op>):
✓ Branch 118 → 66 taken 168 times.
✓ Branch 118 → 119 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 118 → 66 taken 12 times.
✓ Branch 118 → 119 taken 6 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 31 → 15 taken 20 times.
✓ Branch 31 → 32 taken 8 times.
✓ Branch 118 → 66 taken 111 times.
✓ Branch 118 → 119 taken 12 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 30 → 15 taken 22077493 times.
✓ Branch 30 → 31 taken 40 times.
✓ Branch 119 → 66 taken 16392 times.
✓ Branch 119 → 120 taken 689 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 120 → 66 taken 3296 times.
✓ Branch 120 → 121 taken 1192 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 118 → 66 taken 168 times.
✓ Branch 118 → 119 taken 84 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 118 → 66 taken 139 times.
✓ Branch 118 → 119 taken 5 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 119 → 66 taken 4631 times.
✓ Branch 119 → 120 taken 607 times.
|
22119880 | for (int i = 0; i < sz(out); i++) { |
| 118 |
2/4void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✗ Branch 15 → 16 not taken.
✓ Branch 15 → 17 taken 20 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✗ Branch 15 → 16 not taken.
✓ Branch 15 → 17 taken 22077493 times.
|
22155201 | num1 v1 = o1[i] * inv_n2; |
| 119 |
2/4void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✗ Branch 15 → 16 not taken.
✓ Branch 15 → 17 taken 20 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✗ Branch 15 → 16 not taken.
✓ Branch 15 → 17 taken 22077493 times.
|
22155201 | num2 v2 = o2[i] * inv_n1; |
| 120 |
6/8void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✗ Branch 15 → 16 not taken.
✓ Branch 15 → 17 taken 20 times.
✓ Branch 18 → 19 taken 7 times.
✓ Branch 18 → 20 taken 13 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✗ Branch 15 → 16 not taken.
✓ Branch 15 → 17 taken 22077493 times.
✓ Branch 18 → 19 taken 10618902 times.
✓ Branch 18 → 20 taken 11458591 times.
|
44271558 | mnum o_mod = mnum(uint64_t(v1)) * m2_mod + mnum(int(v2)) * m1_mod; |
| 121 | 22116357 | __int128_t o_exact = __int128_t(uint64_t(v1)) * __int128_t(num2::MOD) + __int128_t(int(v2)) * __int128_t(num1::MOD); | |
| 122 |
27/28void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_twice_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_twice_op):
✓ Branch 106 → 107 taken 483 times.
✓ Branch 106 → 108 taken 455 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 106 → 107 taken 5319 times.
✓ Branch 106 → 108 taken 7670 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op>):
✓ Branch 106 → 107 taken 81 times.
✓ Branch 106 → 108 taken 87 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 106 → 107 taken 11 times.
✓ Branch 106 → 108 taken 1 time.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 21 → 22 taken 10 times.
✓ Branch 21 → 25 taken 10 times.
✗ Branch 22 → 23 not taken.
✓ Branch 22 → 24 taken 10 times.
✓ Branch 106 → 107 taken 55 times.
✓ Branch 106 → 108 taken 56 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 21 → 22 taken 17321289 times.
✓ Branch 21 → 25 taken 4756204 times.
✓ Branch 22 → 23 taken 1520120 times.
✓ Branch 22 → 24 taken 15801169 times.
✓ Branch 106 → 107 taken 7124 times.
✓ Branch 106 → 108 taken 9268 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 106 → 107 taken 1653 times.
✓ Branch 106 → 108 taken 1643 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 106 → 107 taken 86 times.
✓ Branch 106 → 108 taken 82 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 106 → 107 taken 101 times.
✓ Branch 106 → 108 taken 38 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 106 → 107 taken 2003 times.
✓ Branch 106 → 108 taken 2628 times.
|
22116357 | if (o_exact >= whole) { o_exact -= whole; o_mod -= whole_mod; } |
| 123 | // Balanced representatives: |o| <= whole/2 | ||
| 124 |
24/24void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_twice_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_twice_op):
✓ Branch 108 → 109 taken 455 times.
✓ Branch 108 → 110 taken 483 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 108 → 109 taken 5234 times.
✓ Branch 108 → 110 taken 7755 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_twice_op>):
✓ Branch 108 → 109 taken 87 times.
✓ Branch 108 → 110 taken 81 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 108 → 109 taken 1 time.
✓ Branch 108 → 110 taken 11 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 25 → 26 taken 10 times.
✓ Branch 25 → 29 taken 10 times.
✓ Branch 108 → 109 taken 56 times.
✓ Branch 108 → 110 taken 55 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<1, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<1>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 25 → 26 taken 4744924 times.
✓ Branch 25 → 29 taken 17332569 times.
✓ Branch 108 → 109 taken 6936 times.
✓ Branch 108 → 110 taken 9456 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::add_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::add_op):
✓ Branch 108 → 109 taken 1643 times.
✓ Branch 108 → 110 taken 1653 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::add_op>):
✓ Branch 108 → 109 taken 82 times.
✓ Branch 108 → 110 taken 86 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op> >(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::detail::cut_op<modnum<1000000007>, ecnerwala::fft::assign_op>):
✓ Branch 108 → 109 taken 38 times.
✓ Branch 108 → 110 taken 101 times.
void ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::finish<2, ecnerwala::fft::assign_op>(ecnerwala::fft::engines::crt<modnum<1000000007>, mod_goldilocks, modnum<2013265921> >::product_t<2>&&, std::span<modnum<1000000007>, 18446744073709551615ul>, ecnerwala::fft::assign_op):
✓ Branch 108 → 109 taken 1932 times.
✓ Branch 108 → 110 taken 2699 times.
|
22116357 | if (o_exact > whole / 2) o_mod -= whole_mod; |
| 125 | 22153665 | op(out[i], o_mod); | |
| 126 | } | ||
| 127 | 6998 | } | |
| 128 | }; | ||
| 129 | |||
| 130 | /* namespace ecnerwala::fft::engines */ } | ||
| 131 |