[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/examples/stf/graph_scope.cu

@@ -1,105 +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) 2022-2025 NVIDIA CORPORATION & AFFILIATES.
-//
-//===----------------------------------------------------------------------===//
-
-/**
- * @file
- * @brief Demonstration of graph_scope RAII usage styles
- *
- * This example shows different ways to use stackable_ctx::graph_scope_guard
- * for automatic push/pop management in nested contexts.
- */
-
-#include <cuda/experimental/stf.cuh>
-
-using namespace cuda::experimental::stf;
-
-int main()
-{
-  stackable_ctx ctx;
-
-  int data[10] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10};
-  auto lA      = ctx.logical_data(data);
-
-  // Style 1: Direct constructor (like std::lock_guard)
-  // This is the most idiomatic C++ style
-  {
-    stackable_ctx::graph_scope_guard scope{ctx}; // Direct constructor - push() called
-
-    auto temp = ctx.logical_data(lA.shape());
-    ctx.parallel_for(temp.shape(), temp.write(), lA.read())->*[] __device__(size_t i, auto temp, auto a) {
-      temp(i) = a(i) * 2;
-    };
-
-    ctx.parallel_for(lA.shape(), lA.write(), temp.read())->*[] __device__(size_t i, auto a, auto temp) {
-      a(i) = temp(i);
-    };
-
-    // pop() called automatically when scope goes out of scope
-  }
-
-  // Style 2: Factory method (convenience)
-  // Useful when you prefer auto type deduction
-  {
-    auto scope = ctx.graph_scope(); // Factory method - push() called
-
-    ctx.parallel_for(lA.shape(), lA.rw())->*[] __device__(size_t i, auto a) {
-      a(i) += 1;
-    };
-
-    // pop() called automatically
-  }
-
-  // Style 3: Direct constructor with explicit type alias
-  // Useful for readability in complex scenarios
-  {
-    using scope_t = stackable_ctx::graph_scope_guard;
-    scope_t scope{ctx}; // Explicit type - push() called
-
-    ctx.parallel_for(lA.shape(), lA.rw())->*[] __device__(size_t i, auto a) {
-      a(i) *= 3;
-    };
-
-    // pop() called automatically
-  }
-
-  // Style 4: Iterative pattern (like in stackable2.cu)
-  // Demonstrates repeated nested contexts
-  for (int iter = 0; iter < 3; iter++)
-  {
-    stackable_ctx::graph_scope_guard iteration{ctx}; // New scope each iteration
-
-    auto temp = ctx.logical_data(lA.shape());
-
-    // tmp = a
-    ctx.parallel_for(temp.shape(), temp.write(), lA.read())->*[] __device__(size_t i, auto temp, auto a) {
-      temp(i) = a(i);
-    };
-
-    // a++
-    ctx.parallel_for(lA.shape(), lA.rw())->*[] __device__(size_t i, auto a) {
-      a(i) += 1;
-    };
-
-    // tmp *= 2
-    ctx.parallel_for(temp.shape(), temp.rw())->*[] __device__(size_t i, auto temp) {
-      temp(i) *= 2;
-    };
-
-    // a += tmp
-    ctx.parallel_for(lA.shape(), temp.read(), lA.rw())->*[] __device__(size_t i, auto temp, auto a) {
-      a(i) += temp(i);
-    };
-
-    // pop() called automatically at end of iteration
-  }
-
-  ctx.finalize();
-  return 0;
-}