[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/thrust/testing/cuda/stream_wrapper.cu

@@ -0,0 +1,86 @@
+#include <thrust/execution_policy.h>
+#include <thrust/reduce.h>
+
+#include <cuda/stream>
+
+#include <unittest/unittest.h>
+
+// Simple non-owning stream wrapper that allows implicit conversion to cudaStream_t.
+struct stream_wrapper
+{
+  stream_wrapper(cudaStream_t s)
+      : stream(s)
+  {}
+
+  operator cudaStream_t() const
+  {
+    return stream;
+  }
+
+  cudaStream_t stream;
+};
+
+// Simple non-owning stream wrapper that allows implicit conversion to cudaStream_t and cuda::stream_ref.
+struct stream_wrapper_ref
+{
+  stream_wrapper_ref(cudaStream_t s)
+      : stream(s)
+  {}
+
+  operator cudaStream_t() const
+  {
+    return stream;
+  }
+  operator cuda::stream_ref() const
+  {
+    return cuda::stream_ref(stream);
+  }
+
+  cudaStream_t stream;
+};
+
+template <typename Wrapper, typename ExecutionPolicy>
+void TestOnStream(ExecutionPolicy policy)
+{
+  using Vector = thrust::device_vector<int>;
+
+  Vector v(3);
+  v[0] = 1;
+  v[1] = -2;
+  v[2] = 3;
+
+  cudaStream_t s;
+  cudaStreamCreate(&s);
+
+  Wrapper wrapper(s);
+
+  auto streampolicy = policy.on(wrapper);
+
+  ASSERT_EQUAL(thrust::reduce(streampolicy, v.begin(), v.end()), 2);
+
+  cudaStreamDestroy(s);
+}
+
+void TestCudartStreamSync()
+{
+  TestOnStream<stream_wrapper>(thrust::cuda::par);
+}
+DECLARE_UNITTEST(TestCudartStreamSync);
+
+void TestCudartStreamNoSync()
+{
+  TestOnStream<stream_wrapper>(thrust::cuda::par_nosync);
+}
+DECLARE_UNITTEST(TestCudartStreamNoSync);
+
+void TestCudaStreamRefSync()
+{
+  TestOnStream<stream_wrapper_ref>(thrust::cuda::par);
+}
+DECLARE_UNITTEST(TestCudaStreamRefSync);
+
+void TestCudaStreamRefNoSync()
+{
+  TestOnStream<stream_wrapper_ref>(thrust::cuda::par_nosync);
+}
+DECLARE_UNITTEST(TestCudaStreamRefNoSync);