[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/cub/test/catch2_test_printing.cu

@@ -0,0 +1,39 @@
+// SPDX-FileCopyrightText: Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
+// SPDX-License-Identifier: BSD-3-Clause
+
+#include <sstream>
+
+#include "test_util.h"
+#include <c2h/catch2_test_helper.h>
+
+template <typename T>
+std::string print(T val)
+{
+  std::stringstream ss;
+  ss << val;
+  return ss.str();
+}
+
+#if TEST_INT128()
+TEST_CASE("Test utils can print __int128", "[test][utils]")
+{
+  REQUIRE(print(__int128_t{0}) == "0");
+  REQUIRE(print(__int128_t{42}) == "42");
+  REQUIRE(print(__int128_t{-1}) == "-1");
+  REQUIRE(print(__int128_t{-42}) == "-42");
+  REQUIRE(print(-1 * (__int128_t{1} << 120)) == "-1329227995784915872903807060280344576");
+}
+
+TEST_CASE("Test utils can print __uint128", "[test][utils]")
+{
+  REQUIRE(print(__uint128_t{0}) == "0");
+  REQUIRE(print(__uint128_t{1}) == "1");
+  REQUIRE(print(__uint128_t{42}) == "42");
+  REQUIRE(print(__uint128_t{1} << 120) == "1329227995784915872903807060280344576");
+}
+#endif
+
+TEST_CASE("Test utils can print KeyValuePair", "[test][utils]")
+{
+  REQUIRE(print(cub::KeyValuePair<int, int>{42, -42}) == "(42,-42)");
+}