[CCCL] Add missing CCCL components: c2h, nvbench_helper, cmake, cudax, AGENTS.md

Added 863 files from NVIDIA/cccl sparse checkout:
- c2h/ (27 files): Catch2 test helpers — generators, validators, runner
- nvbench_helper/ (10 files): Benchmark harness utilities
- cmake/ (29 files): CMake presets and build helpers
- cudax/ (794 files): Experimental CUDA extensions
- AGENTS.md: NVIDIA's official AI agent instructions for CCCL
- CMakePresets.json: Standardized build configurations
- cccl-version.json: Version tracking

Also added CCCL_ASSET_MAP.md mapping all 4295 CCCL files to
competition value and PRD items.

cccl_upstream now covers 100% of competition-critical assets:
- 27 tuning headers (SM80/90/100 benchmark data)
- 32 dispatch headers (algorithm implementations)
- 60 Thrust examples (correctness verification)
- 217 CUB Catch2 tests (regression matrix)
- 153 CUB benchmarks (parameter space search)
- 18 CUB examples (API verification)
- 27 test helpers + benchmark harness
- 794 cudax experimental extensions

cccl_upstream/cudax/test/common/utility.cuh

@@ -0,0 +1,151 @@
+//===----------------------------------------------------------------------===//
+//
+// 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) 2024 NVIDIA CORPORATION & AFFILIATES.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef __COMMON_UTILITY_H__
+#define __COMMON_UTILITY_H__
+
+#include <cuda_runtime_api.h>
+// cuda_runtime_api needs to come first
+
+#include <cuda/__runtime/api_wrapper.h>
+#include <cuda/__stream/stream_ref.h>
+#include <cuda/atomic>
+#include <cuda/std/utility>
+
+#include <cuda/experimental/launch.cuh>
+
+#include <new> // IWYU pragma: keep (needed for placement new)
+
+#include "testing.cuh"
+
+namespace
+{
+namespace test
+{
+constexpr auto one_thread_dims = cuda::make_config(cuda::block_dims<1>(), cuda::grid_dims<1>());
+
+struct _malloc_pinned
+{
+private:
+  void* pv = nullptr;
+
+public:
+  explicit _malloc_pinned(std::size_t size)
+  {
+    cuda::__ensure_current_context guard(cuda::device_ref{0});
+    _CCCL_TRY_CUDA_API(::cudaMallocHost, "failed to allocate pinned memory", &pv, size);
+  }
+
+  ~_malloc_pinned()
+  {
+    cuda::__ensure_current_context guard(cuda::device_ref{0});
+    [[maybe_unused]] auto status = ::cudaFreeHost(pv);
+  }
+
+  template <class T>
+  T* get_as() const noexcept
+  {
+    return static_cast<T*>(pv);
+  }
+};
+
+template <class T>
+struct pinned
+{
+private:
+  _malloc_pinned _mem;
+
+public:
+  explicit pinned(T t)
+      : _mem(sizeof(T))
+  {
+    ::new (_mem.get_as<void>()) T(std::move(t));
+  }
+
+  ~pinned()
+  {
+    get()->~T();
+  }
+
+  T* get() noexcept
+  {
+    return _mem.get_as<T>();
+  }
+  const T* get() const noexcept
+  {
+    return _mem.get_as<T>();
+  }
+
+  T& operator*() noexcept
+  {
+    return *get();
+  }
+  const T& operator*() const noexcept
+  {
+    return *get();
+  }
+};
+
+template <int N>
+struct assign_n
+{
+  __device__ constexpr void operator()(int* pi) const noexcept
+  {
+    *pi = N;
+  }
+};
+
+template <int N>
+struct verify_n
+{
+  __device__ void operator()(int* pi) const noexcept
+  {
+    REQUIRE_DEVICE(*pi == N);
+  }
+};
+
+using assign_42 = assign_n<42>;
+using verify_42 = verify_n<42>;
+
+struct atomic_add_one
+{
+  __device__ void operator()(int* pi) const noexcept
+  {
+    cuda::atomic_ref atomic_pi(*pi);
+    atomic_pi.fetch_add(1);
+  }
+};
+
+struct atomic_sub_one
+{
+  __device__ void operator()(int* pi) const noexcept
+  {
+    cuda::atomic_ref atomic_pi(*pi);
+    atomic_pi.fetch_sub(1);
+  }
+};
+
+struct spin_until_80
+{
+  __device__ void operator()(int* pi) const noexcept
+  {
+    cuda::atomic_ref atomic_pi(*pi);
+    while (atomic_pi.load() != 80)
+      ;
+  }
+};
+
+struct empty_kernel
+{
+  __device__ void operator()() const noexcept {}
+};
+} // namespace test
+} // namespace
+#endif // __COMMON_UTILITY_H__