[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/c2h/generators_uniform_offsets.cu

@@ -0,0 +1,66 @@
+// SPDX-FileCopyrightText: Copyright (c) 2011-2025, NVIDIA CORPORATION. All rights reserved.
+// SPDX-License-Identifier: BSD-3-Clause
+
+#include <c2h/generators.h>
+// #include <c2h/detail/generators.cuh>
+
+#include <thrust/device_ptr.h>
+#include <thrust/find.h>
+#include <thrust/scan.h>
+
+#include <cuda/std/cstdint>
+#include <cuda/std/span>
+
+#include <c2h/device_policy.h>
+
+namespace c2h::detail
+{
+template <class T>
+struct greater_equal_op
+{
+  T val;
+
+  __device__ bool operator()(T x)
+  {
+    return x >= val;
+  }
+};
+
+template <typename T>
+std::size_t gen_uniform_offsets(
+  seed_t seed, cuda::std::span<T> segment_offsets, T total_elements, T min_segment_size, T max_segment_size)
+{
+  gen_values_between(seed, segment_offsets, min_segment_size, max_segment_size);
+  *thrust::device_ptr<T>(&segment_offsets[total_elements]) = total_elements + 1;
+  thrust::exclusive_scan(device_policy, segment_offsets.begin(), segment_offsets.end(), segment_offsets.begin());
+  const auto iter =
+    thrust::find_if(device_policy, segment_offsets.begin(), segment_offsets.end(), greater_equal_op<T>{total_elements});
+  *thrust::device_ptr<T>(&*iter) = total_elements;
+  return iter - segment_offsets.begin() + 1;
+}
+
+template std::size_t gen_uniform_offsets(
+  seed_t seed,
+  cuda::std::span<int32_t> segment_offsets,
+  int32_t total_elements,
+  int32_t min_segment_size,
+  int32_t max_segment_size);
+template std::size_t gen_uniform_offsets(
+  seed_t seed,
+  cuda::std::span<uint32_t> segment_offsets,
+  uint32_t total_elements,
+  uint32_t min_segment_size,
+  uint32_t max_segment_size);
+template std::size_t gen_uniform_offsets(
+  seed_t seed,
+  cuda::std::span<int64_t> segment_offsets,
+  int64_t total_elements,
+  int64_t min_segment_size,
+  int64_t max_segment_size);
+template std::size_t gen_uniform_offsets(
+  seed_t seed,
+  cuda::std::span<uint64_t> segment_offsets,
+  uint64_t total_elements,
+  uint64_t min_segment_size,
+  uint64_t max_segment_size);
+} // namespace c2h::detail