[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

cccl_upstream/cudax/test/stf/examples/01-axpy-places.cu

@@ -1,119 +0,0 @@
-//===----------------------------------------------------------------------===//
-//
-// Part of CUDASTF 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) 2022-2024 NVIDIA CORPORATION & AFFILIATES.
-//
-//===----------------------------------------------------------------------===//
-
-/**
- * @file
- *
- * @brief This example illustrates how to use the task construct with grids of
- *        places and composite data places
- */
-
-#include <cuda/experimental/__places/partitions/tiled_partition.cuh>
-#include <cuda/experimental/__stf/graph/graph_ctx.cuh>
-#include <cuda/experimental/__stf/stream/stream_ctx.cuh>
-
-using namespace cuda::experimental::stf;
-
-template <typename T>
-__global__ void axpy(size_t start, size_t cnt, T a, const T* x, T* y)
-{
-  int tid      = blockIdx.x * blockDim.x + threadIdx.x;
-  int nthreads = gridDim.x * blockDim.x;
-
-  for (int ind = tid; ind < cnt; ind += nthreads)
-  {
-    y[ind + start] += a * x[ind + start];
-  }
-}
-
-double X0(size_t i)
-{
-  return sin((double) i);
-}
-
-double Y0(size_t i)
-{
-  return cos((double) i);
-}
-
-template <typename Ctx>
-void run()
-{
-  Ctx ctx;
-
-  const int N = 1024 * 1024 * 32;
-  double *X, *Y;
-
-  X = new double[N];
-  Y = new double[N];
-  SCOPE(exit)
-  {
-    delete[] X;
-    delete[] Y;
-  };
-
-  for (size_t ind = 0; ind < N; ind++)
-  {
-    X[ind] = X0(ind);
-    Y[ind] = Y0(ind);
-  }
-
-  //    std::shared_ptr<execution_grid> all_devs = exec_place::all_devices();
-  // use grid [ 0 0 0 0 ] for debugging purpose
-  auto all_devs = exec_place::repeat(exec_place::device(0), 4);
-
-  // 512k doubles = 4MB (2 pages)
-  // A 1D blocking strategy over all devices with a block size of 32 and a round robin distribution of blocks across
-  // devices
-  //    data_place cdp = data_place(exec_place::all_devices().as_grid().get_grid(),
-  //            [](dim4 grid_dim, pos4 index_pos) { return pos4((index_pos.x / (512 * 1024ULL)) % grid_dim.x); });
-
-  data_place cdp = data_place::composite(tiled_partition<512 * 1024ULL>(), all_devs);
-
-  auto handle_X = ctx.logical_data(X, {N});
-  auto handle_Y = ctx.logical_data(Y, {N});
-
-  double alpha = 3.14;
-
-  /* Compute Y = Y + alpha X */
-  auto t = ctx.task(all_devs, handle_X.read(cdp), handle_Y.rw(cdp));
-  t->*[&](auto, auto sX, auto sY) {
-    size_t grid_size = t.grid_dims().size();
-
-    assert(N % grid_size == 0);
-
-    for (size_t i = 0; i < grid_size; i++)
-    {
-      auto active = t.activate_place(i);
-      axpy<<<16, 128, 0, t.get_stream(i)>>>(i * N / grid_size, N / grid_size, alpha, sX.data_handle(), sY.data_handle());
-    }
-  };
-
-  /* Check the result on the host */
-  ctx.host_launch(handle_X.read(), handle_Y.read())->*[&](auto sX, auto sY) {
-    for (size_t ind = 0; ind < N; ind++)
-    {
-      // Y should be Y0 + alpha X0
-      EXPECT(fabs(sY(ind) - (Y0(ind) + alpha * X0(ind))) < 0.0001);
-
-      // X should be X0
-      EXPECT(fabs(sX(ind) - X0(ind)) < 0.0001);
-    }
-  };
-
-  ctx.finalize();
-}
-
-int main()
-{
-  run<stream_ctx>();
-  // Disabled until composite data places are implemented with graphs
-  // run<graph_ctx>();
-}