#include "./scalar_multiplication_fast.hpp"
#include "./pippenger_arena_layout.hpp"
#include "./pippenger_constantine.hpp"
#include "./pippenger_dedup.hpp"
#include "barretenberg/common/assert.hpp"
#include "barretenberg/common/thread.hpp"
#include "barretenberg/ecc/curves/bn254/bn254.hpp"
#include "barretenberg/ecc/curves/grumpkin/grumpkin.hpp"
#include "barretenberg/ecc/groups/affine_add_packed.hpp"
#include "barretenberg/ecc/groups/element_impl.hpp"
#include "barretenberg/numeric/bitop/get_msb.hpp"
#include <barretenberg/env/hardware_concurrency.hpp>
#include <algorithm>
#include <atomic>
#include <bit>
#include <cstddef>
#include <cstdint>
#include <limits>
#include <memory>
#include <span>
#include <vector>
#include "./pippenger_fallbacks.hpp"
#include "./pippenger_batched.hpp"

Namespaces
namespace	bb
	Entry point for Barretenberg command-line interface.

namespace	bb::scalar_multiplication

namespace	bb::scalar_multiplication::round_parallel_detail

Functions
size_t	bb::scalar_multiplication::window_bits_tuning_oversub_factor (size_t n_input)
	N-dependent oversubscription factor used ONLY for `choose_window_bits`' target_load formula (not for actual thread dispatch).

template<typename F >
void	bb::scalar_multiplication::round_parallel_detail::msm_parallel_for (size_t num_threads, F &&body) noexcept
	Small-N fast-path: per-thread Jacobian Pippenger over a partition of the input.

template<typename Curve >
Curve::Element	bb::scalar_multiplication::round_parallel_detail::pippenger_round_parallel_jacobian_fast (std::span< const typename Curve::ScalarField > scalars, std::span< const typename Curve::AffineElement > points, size_t min_pts_per_thread_override=0, size_t max_threads=0) noexcept
	Single-MSM_fast, no-affine-trick Pippenger over window_bits-wide windows.

template<typename Curve >
size_t	bb::scalar_multiplication::compute_arena_bytes_for_msm (size_t n_input, bool external_glv_provided, bool dedup_active, size_t max_threads) noexcept
	Round-parallel Pippenger MSM_fast. Windows process sequentially (high-to-low) but each window is fully parallel across threads. Windows are processed in batches of `windows_in_batch` to amortise parallel_for barriers; the batch count is sized at runtime to fit BATCH_MEM_BUDGET (~32 MiB).

template<typename Curve >
Curve::Element	bb::scalar_multiplication::pippenger_round_parallel (PolynomialSpan< const typename Curve::ScalarField > scalars, std::span< const typename Curve::AffineElement > points, size_t dedup_info=0, std::span< const typename Curve::AffineElement > external_glv_doubled={}, std::span< std::byte > external_arena={}, size_t max_threads=0) noexcept
	State of the art pippenger_fast multiscalar multiplication algorithm.

template<typename Curve >
Curve::Element	bb::scalar_multiplication::pippenger_unsafe_fast (PolynomialSpan< const typename Curve::ScalarField > scalars, std::span< const typename Curve::AffineElement > points, size_t dedup_info) noexcept

template<typename Curve >
Curve::Element	bb::scalar_multiplication::pippenger_fast (PolynomialSpan< const typename Curve::ScalarField > scalars, std::span< const typename Curve::AffineElement > points, bool handle_edge_cases, size_t dedup_info) noexcept

template curve::BN254::Element	bb::scalar_multiplication::pippenger_unsafe_fast< curve::BN254 > (PolynomialSpan< const curve::BN254::ScalarField > scalars, std::span< const curve::BN254::AffineElement > points, size_t dedup_info) noexcept

template curve::Grumpkin::Element	bb::scalar_multiplication::pippenger_unsafe_fast< curve::Grumpkin > (PolynomialSpan< const curve::Grumpkin::ScalarField > scalars, std::span< const curve::Grumpkin::AffineElement > points, size_t dedup_info) noexcept

template curve::BN254::Element	bb::scalar_multiplication::pippenger_fast< curve::BN254 > (PolynomialSpan< const curve::BN254::ScalarField > scalars, std::span< const curve::BN254::AffineElement > points, bool handle_edge_cases, size_t dedup_info) noexcept

template curve::Grumpkin::Element	bb::scalar_multiplication::pippenger_fast< curve::Grumpkin > (PolynomialSpan< const curve::Grumpkin::ScalarField > scalars, std::span< const curve::Grumpkin::AffineElement > points, bool handle_edge_cases, size_t dedup_info) noexcept

template curve::BN254::Element	bb::scalar_multiplication::pippenger_round_parallel< curve::BN254 > (PolynomialSpan< const curve::BN254::ScalarField > scalars, std::span< const curve::BN254::AffineElement > points, size_t dedup_info, std::span< const curve::BN254::AffineElement > external_glv_doubled, std::span< std::byte > external_arena, size_t max_threads) noexcept

template curve::Grumpkin::Element	bb::scalar_multiplication::pippenger_round_parallel< curve::Grumpkin > (PolynomialSpan< const curve::Grumpkin::ScalarField > scalars, std::span< const curve::Grumpkin::AffineElement > points, size_t dedup_info, std::span< const curve::Grumpkin::AffineElement > external_glv_doubled, std::span< std::byte > external_arena, size_t max_threads) noexcept

template curve::BN254::Element	bb::scalar_multiplication::trivial_msm< curve::BN254 > (PolynomialSpan< const curve::BN254::ScalarField > scalars_span, std::span< const curve::BN254::AffineElement > all_points) noexcept

template curve::Grumpkin::Element	bb::scalar_multiplication::trivial_msm< curve::Grumpkin > (PolynomialSpan< const curve::Grumpkin::ScalarField > scalars_span, std::span< const curve::Grumpkin::AffineElement > all_points) noexcept

template curve::BN254::Element	bb::scalar_multiplication::trivial_msm_threaded< curve::BN254 > (PolynomialSpan< const curve::BN254::ScalarField > scalars_span, std::span< const curve::BN254::AffineElement > all_points, size_t max_threads) noexcept

template curve::Grumpkin::Element	bb::scalar_multiplication::trivial_msm_threaded< curve::Grumpkin > (PolynomialSpan< const curve::Grumpkin::ScalarField > scalars_span, std::span< const curve::Grumpkin::AffineElement > all_points, size_t max_threads) noexcept

template curve::BN254::Element	bb::scalar_multiplication::round_parallel_detail::pippenger_round_parallel_jacobian_fast< curve::BN254 > (std::span< const curve::BN254::ScalarField > scalars, std::span< const curve::BN254::AffineElement > points, size_t min_pts_per_thread_override, size_t max_threads) noexcept

template curve::Grumpkin::Element	bb::scalar_multiplication::round_parallel_detail::pippenger_round_parallel_jacobian_fast< curve::Grumpkin > (std::span< const curve::Grumpkin::ScalarField > scalars, std::span< const curve::Grumpkin::AffineElement > points, size_t min_pts_per_thread_override, size_t max_threads) noexcept

template size_t	bb::scalar_multiplication::compute_arena_bytes_for_msm< curve::BN254 > (size_t, bool, bool, size_t) noexcept