[INFRA] Import NVIDIA/CCCL upstream as optimization reference library

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

cccl_upstream/libcudacxx/codegen/generators/header.h

@@ -0,0 +1,88 @@
+//===----------------------------------------------------------------------===//
+//
+// Part of libcu++, the C++ Standard Library for your entire system,
+// 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 HEADER_H
+#define HEADER_H
+
+#include <string>
+
+inline void FormatHeader(std::ostream& out)
+{
+  constexpr auto header = R"XXX(//===----------------------------------------------------------------------===//
+//
+// Part of libcu++, the C++ Standard Library for your entire system,
+// 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.
+//
+//===----------------------------------------------------------------------===//
+
+// This is an autogenerated file, we want to ensure that it contains exactly the contents we want to generate
+// clang-format off
+
+#ifndef _CUDA_STD___ATOMIC_FUNCTIONS_CUDA_PTX_GENERATED_H
+#define _CUDA_STD___ATOMIC_FUNCTIONS_CUDA_PTX_GENERATED_H
+
+#include <cuda/std/detail/__config>
+
+#if defined(_CCCL_IMPLICIT_SYSTEM_HEADER_GCC)
+#  pragma GCC system_header
+#elif defined(_CCCL_IMPLICIT_SYSTEM_HEADER_CLANG)
+#  pragma clang system_header
+#elif defined(_CCCL_IMPLICIT_SYSTEM_HEADER_MSVC)
+#  pragma system_header
+#endif // no system header
+
+#include <cuda/std/cassert>
+#include <cuda/std/cstdint>
+
+#include <cuda/std/__type_traits/enable_if.h>
+#include <cuda/std/__type_traits/is_signed.h>
+#include <cuda/std/__type_traits/is_unsigned.h>
+
+#include <cuda/std/__atomic/scopes.h>
+#include <cuda/std/__atomic/order.h>
+#include <cuda/std/__atomic/functions/common.h>
+#include <cuda/std/__atomic/functions/cuda_ptx_generated_helper.h>
+#include <cuda/std/__atomic/functions/cuda_local.h>
+
+#include <cuda/std/__cccl/prologue.h>
+
+_CCCL_BEGIN_NAMESPACE_CUDA_STD
+
+#if _CCCL_CUDA_COMPILATION()
+
+extern "C" _CCCL_DEVICE void __atomic_cas_128b_unsupported_before_SM_90();
+extern "C" _CCCL_DEVICE void __atomic_exchange_128b_unsupported_before_SM_90();
+extern "C" _CCCL_DEVICE void __atomic_ldst_128b_unsupported_before_SM_70();
+)XXX";
+
+  out << header;
+}
+
+inline void FormatTail(std::ostream& out)
+{
+  constexpr auto tail = R"XXX(
+#endif // _CCCL_CUDA_COMPILATION()
+
+_CCCL_END_NAMESPACE_CUDA_STD
+
+#include <cuda/std/__cccl/epilogue.h>
+
+#endif // _CUDA_STD___ATOMIC_FUNCTIONS_CUDA_PTX_GENERATED_H
+
+// clang-format on
+)XXX";
+
+  out << tail;
+}
+
+#endif // HEADER_H