Optimize Hive hash computation for nested types

src/main/cpp/benchmarks/CMakeLists.txt

@@ -1,5 +1,5 @@
 #=============================================================================
-# Copyright (c) 2022-2024, NVIDIA CORPORATION.
+# Copyright (c) 2022-2025, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -81,5 +81,8 @@ ConfigureBench(BLOOM_FILTER_BENCH
 ConfigureBench(GET_JSON_OBJECT_BENCH
   get_json_object.cu)
 
+ConfigureBench(HASH_BENCH
+  hash.cu)
+
 ConfigureBench(PARSE_URI_BENCH
   parse_uri.cpp)

src/main/cpp/benchmarks/hash.cu

@@ -0,0 +1,69 @@
+/*
+ * Copyright (c) 2024-2025, NVIDIA CORPORATION.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include <benchmarks/common/generate_input.hpp>
+
+#include <cudf_test/column_utilities.hpp>
+
+#include <cudf/io/types.hpp>
+
+#include <hash.hpp>
+#include <nvbench/nvbench.cuh>
+
+constexpr auto min_width = 10;
+constexpr auto max_width = 10;
+
+static void hash(nvbench::state& state)
+{
+  auto const size_bytes = static_cast<std::size_t>(state.get_int64("size_bytes"));
+  auto const max_depth  = static_cast<cudf::size_type>(state.get_int64("max_depth"));
+
+  auto const bench_structs = true;
+  auto const input_table   = [&] {
+    if (bench_structs) {
+      data_profile const table_profile =
+        data_profile_builder().no_validity().struct_depth(max_depth).struct_types(
+          std::vector<cudf::type_id>{
+            cudf::type_id::INT32, cudf::type_id::FLOAT32, cudf::type_id::STRING});
+      return create_random_table(
+        {cudf::type_id::STRUCT}, table_size_bytes{size_bytes}, table_profile);
+    } else {
+      data_profile const table_profile =
+        data_profile_builder()
+          .no_validity()
+          .distribution(cudf::type_id::STRING, distribution_id::NORMAL, min_width, max_width)
+          .distribution(cudf::type_id::LIST, distribution_id::NORMAL, min_width, max_width)
+          .list_depth(max_depth)
+          .list_type(cudf::type_id::STRING);
+      return create_random_table(
+        {cudf::type_id::LIST}, table_size_bytes{size_bytes}, table_profile);
+    }
+  }();
+
+  auto const stream = cudf::get_default_stream();
+  state.set_cuda_stream(nvbench::make_cuda_stream_view(stream.value()));
+  state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) {
+    // `hive_hash` can be substituted with other hash functions
+    spark_rapids_jni::hive_hash(*input_table);
+  });
+  state.add_global_memory_reads<nvbench::int8_t>(size_bytes);
+}
+
+NVBENCH_BENCH(hash)
+  .set_name("hash")
+  .add_int64_axis("size_bytes",
+                  {50'000'000, 100'000'000, 500'000'000, 1'000'000'000})  // 50MB, 100MB, 500MB, 1GB
+  .add_int64_axis("max_depth", {1, 2, 4, 8});