[CCCL] 瘦身 + 补全: 移除 cudax/python/libcudacxx-tests 冗余文件, 新增 c2h 测试助手 + cmake 构建系统 + 8 个 CUDA thrust examples

变更摘要:
- 删除: cudax/ (783 files, 7.2M) — 实验性组件,竞赛不需要
- 删除: python/ (226 files, 2.0M) — Python 绑定,竞赛不需要
- 删除: libcudacxx/{test,benchmarks,codegen,cmake,share} (4432 files, 31M)
  保留: libcudacxx/include/ (1463 headers, cuda::std 编译依赖)
- 新增: c2h/ (27 files) — CUB Catch2 测试辅助头文件,编译 243 个测试必需
- 新增: cmake/ (29 files) — CCCL 原生 CMake 构建系统
- 新增: thrust/examples/cuda/ (7 files) + cpp_integration/ (1 file)
  async_reduce, custom_temporary_allocation, explicit_cuda_stream,
  global_device_vector, range_view, unwrap_pointer, wrap_pointer, device

结果: cccl_upstream 从 74M→35M (瘦身 53%), 核心内容 100% 保留:
  27/27 tuning headers, 78 benchmarks, 243 tests,
  60 thrust examples, 18 CUB examples, 全部编译头文件
This commit is contained in:
muh-bot
2026-08-03 12:39:26 +00:00
parent a2a5dd8f00
commit 24ef6a91b5
5439 changed files with 0 additions and 719516 deletions

View File

@@ -1,58 +0,0 @@
//===----------------------------------------------------------------------===//
//
// 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/std/cstddef>
#include <cuda/experimental/__cuco/capacity.cuh>
#include <cuda/experimental/__cuco/hash_functions.cuh>
#include <cuda/experimental/__cuco/probing_scheme.cuh>
#include <testing.cuh>
namespace cudax = cuda::experimental;
C2H_TEST("cuco make_valid_capacity rounding and validation", "[capacity]")
{
using probing = cudax::cuco::double_hashing<1, cudax::cuco::hash<int>>;
[[maybe_unused]] constexpr int bucket = 1;
static_assert(cudax::cuco::is_double_hashing_v<probing>, "scheme is double hashing");
// make_valid_capacity rounds up and is idempotent; is_valid_capacity is derived from it
constexpr auto valid = cudax::cuco::make_valid_capacity<probing, bucket>(::cuda::std::size_t{1000});
static_assert(valid >= 1000, "rounds up");
static_assert(cudax::cuco::is_valid_capacity<probing, bucket>(valid), "result is valid");
static_assert(cudax::cuco::make_valid_capacity<probing, bucket>(valid) == valid, "idempotent");
// 1000 is not a valid double-hashing capacity; it rounds up to a prime-cycle capacity
static_assert(!cudax::cuco::is_valid_capacity<probing, bucket>(::cuda::std::size_t{1000}), "1000 is not valid");
// equal-rounding requests produce the same valid capacity
static_assert(cudax::cuco::make_valid_capacity<probing, bucket>(::cuda::std::size_t{1000})
== cudax::cuco::make_valid_capacity<probing, bucket>(::cuda::std::size_t{1008}),
"requests that round to the same capacity agree");
// cuCollections extent_test parity: double hashing, cg_size 2, bucket_size 4.
// 1234 rounds up to next_prime(ceil(1234 / 8) = 155) = 157, times the stride 8 -> 1256.
using dh4 = cudax::cuco::double_hashing<2, cudax::cuco::hash<int>>;
[[maybe_unused]] constexpr int bucket4 = 4;
static_assert(cudax::cuco::make_valid_capacity<dh4, bucket4>(::cuda::std::size_t{1234}) == ::cuda::std::size_t{1256},
"compile-time valid capacity matches the cuCollections extent test");
REQUIRE(cudax::cuco::make_valid_capacity<dh4, bucket4>(::cuda::std::size_t{1234}) == ::cuda::std::size_t{1256});
// a desired load factor outside (0, 1] is rejected
using lp4 = cudax::cuco::linear_probing<2, cudax::cuco::hash<int>>;
auto bad_lf = [](double __lf) {
[[maybe_unused]] auto __r = cudax::cuco::make_valid_capacity<lp4, bucket4>(::cuda::std::size_t{1000}, __lf);
};
REQUIRE_THROWS(bad_lf(0.0));
REQUIRE_THROWS(bad_lf(-0.5));
REQUIRE_THROWS(bad_lf(1.5));
}