//===----------------------------------------------------------------------===//
//
// Part of libcu++, the C++ Standard Library for your entire system,
// under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES.
//
//===----------------------------------------------------------------------===//

#include <thrust/device_vector.h>
#include <thrust/partition.h>

#include <cuda/memory_pool>
#include <cuda/std/execution>
#include <cuda/stream>

#include "nvbench_helper.cuh"

template <typename T>
static void basic(nvbench::state& state, nvbench::type_list<T>)
{
  using select_op_t = less_then_t<T>;

  const auto elements       = static_cast<std::size_t>(state.get_int64("Elements"));
  const bit_entropy entropy = str_to_entropy(state.get_string("Entropy"));

  const T val = lerp_min_max<T>(entropy_to_probability(entropy));
  select_op_t select_op{val};

  thrust::device_vector<T> input = generate(elements);

  state.add_element_count(elements);
  state.add_global_memory_reads<T>(elements);
  state.add_global_memory_writes<T>(elements);

  caching_allocator_t alloc{};
  state.exec(nvbench::exec_tag::gpu | nvbench::exec_tag::no_batch | nvbench::exec_tag::sync,
             [&](nvbench::launch& launch) {
               do_not_optimize(cuda::std::partition(cuda_policy(alloc, launch), input.begin(), input.end(), select_op));
             });
}

NVBENCH_BENCH_TYPES(basic, NVBENCH_TYPE_AXES(fundamental_types))
  .set_name("base")
  .set_type_axes_names({"T{ct}"})
  .add_int64_power_of_two_axis("Elements", nvbench::range(16, 28, 4))
  .add_string_axis("Entropy", {"1.000", "0.544", "0.000"});