CCCL (CUDA C++ Core Libraries) provides: - CUB: device/block/warp-level GPU primitives (reduce, scan, sort, topk) - Thrust: high-level parallel algorithms (transform_reduce, sort, scan) - libcudacxx: CUDA C++ standard library (atomics, barriers, memory) - cudax: experimental features (memory resources, allocators) - Tuning policies: per-SM hardware-specific algorithm parameters Competition optimization vectors mapped to CCCL: - Output TPS (83% weight): warp_reduce, block_reduce, device_topk - Input TPS (14% weight): device_scan, block_load, prefetch - Cache TPS (3% weight): prefix caching strategy patterns - Memory (0.9 util): pooled/cached/buddy allocators Source: https://github.com/NVIDIA/cccl (shallow clone, HEAD only) License: Apache-2.0
243 lines
9.5 KiB
Plaintext
243 lines
9.5 KiB
Plaintext
//===----------------------------------------------------------------------===//
|
|
//
|
|
// Part of CUDA Experimental in CUDA C++ Core Libraries,
|
|
// 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 <cuda/devices>
|
|
#include <cuda/functional>
|
|
#include <cuda/hierarchy>
|
|
#include <cuda/launch>
|
|
#include <cuda/std/algorithm>
|
|
#include <cuda/std/type_traits>
|
|
#include <cuda/stream>
|
|
|
|
#include <cuda/experimental/coop.cuh>
|
|
#include <cuda/experimental/group.cuh>
|
|
|
|
#include <testing.cuh>
|
|
|
|
#include <c2h/catch2_test_helper.h>
|
|
#include <c2h/extended_types.h>
|
|
#include <c2h/generators.h>
|
|
#include <catch2/matchers/catch_matchers_floating_point.hpp>
|
|
|
|
/***********************************************************************************************************************
|
|
* Thread Reduce Wrapper Kernels
|
|
**********************************************************************************************************************/
|
|
|
|
template <bool Broadcasted>
|
|
struct ReduceKernel
|
|
{
|
|
template <class Config, int NumItems, class T, class RedOp>
|
|
__device__ void operator()(
|
|
Config config,
|
|
cuda::std::integral_constant<int, NumItems>,
|
|
const T* __restrict__ d_in,
|
|
T* __restrict__ d_out,
|
|
RedOp red_op)
|
|
{
|
|
cudax::this_thread thread{config};
|
|
|
|
T thread_data[NumItems];
|
|
for (int i = 0; i < NumItems; ++i)
|
|
{
|
|
thread_data[i] = d_in[i];
|
|
}
|
|
|
|
if constexpr (Broadcasted)
|
|
{
|
|
const auto result = cudax::coop::reduce(cudax::broadcasted, thread, thread_data, red_op);
|
|
|
|
*d_out = result;
|
|
}
|
|
else
|
|
{
|
|
const auto result = cudax::coop::reduce(thread, thread_data, red_op);
|
|
|
|
REQUIRE(result.has_value());
|
|
*d_out = result.value();
|
|
}
|
|
}
|
|
};
|
|
|
|
/***********************************************************************************************************************
|
|
* Type list definition
|
|
**********************************************************************************************************************/
|
|
|
|
using integral_type_list =
|
|
c2h::type_list<cuda::std::int8_t, cuda::std::int16_t, cuda::std::uint16_t, cuda::std::int32_t, cuda::std::int64_t>;
|
|
|
|
using fp_type_list = c2h::type_list<float, double>;
|
|
|
|
using operator_integral_list =
|
|
c2h::type_list<cuda::std::plus<>,
|
|
cuda::std::multiplies<>,
|
|
cuda::std::bit_and<>,
|
|
cuda::std::bit_or<>,
|
|
cuda::std::bit_xor<>,
|
|
cuda::minimum<>,
|
|
cuda::maximum<>>;
|
|
|
|
using operator_fp_list = c2h::type_list<cuda::std::plus<>, cuda::std::multiplies<>, cuda::minimum<>, cuda::maximum<>>;
|
|
|
|
/***********************************************************************************************************************
|
|
* Verify results and kernel launch
|
|
**********************************************************************************************************************/
|
|
|
|
template <class T>
|
|
void verify_results(const T& expected_data, const T& test_results)
|
|
{
|
|
if constexpr (cuda::std::is_floating_point_v<T>)
|
|
{
|
|
REQUIRE_THAT(expected_data, Catch::Matchers::WithinRel(test_results, T{0.05}));
|
|
}
|
|
else
|
|
{
|
|
REQUIRE(expected_data == test_results);
|
|
}
|
|
}
|
|
|
|
template <class T, class RedOp, bool Broadcasted = false>
|
|
void run_reduce_kernel(
|
|
cuda::stream_ref stream,
|
|
int num_items,
|
|
const c2h::device_vector<T>& in,
|
|
c2h::device_vector<T>& out,
|
|
RedOp red_op,
|
|
cuda::std::bool_constant<Broadcasted> = {})
|
|
{
|
|
const auto config = cuda::make_config(cuda::grid_dims<1>(), cuda::block_dims<1>());
|
|
const auto in_ptr = thrust::raw_pointer_cast(in.data());
|
|
const auto out_ptr = thrust::raw_pointer_cast(out.data());
|
|
const ReduceKernel<Broadcasted> kernel{};
|
|
|
|
switch (num_items)
|
|
{
|
|
case 1:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 1>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 2:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 2>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 3:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 3>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 4:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 4>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 5:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 5>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 6:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 6>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 7:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 7>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 8:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 8>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 9:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 9>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 10:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 10>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 11:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 11>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 12:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 12>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 13:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 13>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 14:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 14>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 15:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 15>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
case 16:
|
|
cuda::launch(stream, config, kernel, cuda::std::integral_constant<int, 16>{}, in_ptr, out_ptr, red_op);
|
|
break;
|
|
default:
|
|
FAIL("Unsupported number of items");
|
|
}
|
|
stream.sync();
|
|
}
|
|
|
|
constexpr int max_size = 16;
|
|
constexpr int num_seeds = 10;
|
|
|
|
/***********************************************************************************************************************
|
|
* Test cases
|
|
**********************************************************************************************************************/
|
|
|
|
_CCCL_DIAG_SUPPRESS_MSVC(4244) // warning C4244: '=': conversion from 'int' to '_Tp', possible loss of data
|
|
|
|
C2H_TEST("reduce/this_thread Integral Type Tests", "[reduce][this_thread]", integral_type_list, operator_integral_list)
|
|
{
|
|
using value_t = c2h::get<0, TestType>;
|
|
using op_t = c2h::get<1, TestType>;
|
|
constexpr auto reduce_op = op_t{};
|
|
constexpr auto operator_identity = cuda::identity_element<op_t, value_t>();
|
|
CAPTURE(c2h::type_name<value_t>(), max_size, c2h::type_name<decltype(reduce_op)>());
|
|
c2h::device_vector<value_t> d_in(max_size);
|
|
c2h::device_vector<value_t> d_out(1);
|
|
c2h::gen(C2H_SEED(num_seeds), d_in, cuda::std::numeric_limits<value_t>::min());
|
|
c2h::host_vector<value_t> h_in = d_in;
|
|
cuda::stream stream{cuda::devices[0]};
|
|
for (int num_items = 1; num_items <= max_size; ++num_items)
|
|
{
|
|
auto reference_result = cuda::std::accumulate(h_in.begin(), h_in.begin() + num_items, operator_identity, reduce_op);
|
|
run_reduce_kernel(stream, num_items, d_in, d_out, reduce_op);
|
|
verify_results(reference_result, c2h::host_vector<value_t>(d_out)[0]);
|
|
}
|
|
}
|
|
|
|
C2H_TEST("reduce/this_thread Floating-Point Type Tests", "[reduce][this_thread]", fp_type_list, operator_fp_list)
|
|
{
|
|
using value_t = c2h::get<0, TestType>;
|
|
using op_t = c2h::get<1, TestType>;
|
|
constexpr auto reduce_op = op_t{};
|
|
const auto operator_identity = cuda::identity_element<op_t, value_t>();
|
|
CAPTURE(c2h::type_name<value_t>(), max_size, c2h::type_name<decltype(reduce_op)>());
|
|
c2h::device_vector<value_t> d_in(max_size);
|
|
c2h::device_vector<value_t> d_out(1);
|
|
c2h::gen(C2H_SEED(num_seeds), d_in, cuda::std::numeric_limits<value_t>::min());
|
|
c2h::host_vector<value_t> h_in = d_in;
|
|
cuda::stream stream{cuda::devices[0]};
|
|
for (int num_items = 1; num_items <= max_size; ++num_items)
|
|
{
|
|
auto reference_result = cuda::std::accumulate(h_in.begin(), h_in.begin() + num_items, operator_identity, reduce_op);
|
|
run_reduce_kernel(stream, num_items, d_in, d_out, reduce_op);
|
|
verify_results(reference_result, c2h::host_vector<value_t>(d_out)[0]);
|
|
}
|
|
}
|
|
|
|
C2H_TEST("reduce/this_thread Broadcasted", "[reduce][this_thread]", integral_type_list)
|
|
{
|
|
using value_t = c2h::get<0, TestType>;
|
|
using op_t = cuda::std::plus<>;
|
|
constexpr auto reduce_op = op_t{};
|
|
constexpr auto operator_identity = cuda::identity_element<op_t, value_t>();
|
|
CAPTURE(c2h::type_name<value_t>(), max_size, c2h::type_name<decltype(reduce_op)>());
|
|
c2h::device_vector<value_t> d_in(max_size);
|
|
c2h::device_vector<value_t> d_out(1);
|
|
c2h::gen(C2H_SEED(num_seeds), d_in, cuda::std::numeric_limits<value_t>::min());
|
|
c2h::host_vector<value_t> h_in = d_in;
|
|
cuda::stream stream{cuda::devices[0]};
|
|
for (int num_items = 1; num_items <= max_size; ++num_items)
|
|
{
|
|
auto reference_result = cuda::std::accumulate(h_in.begin(), h_in.begin() + num_items, operator_identity, reduce_op);
|
|
run_reduce_kernel(stream, num_items, d_in, d_out, reduce_op, cuda::std::true_type{});
|
|
verify_results(reference_result, c2h::host_vector<value_t>(d_out)[0]);
|
|
}
|
|
}
|