ARROW-14479: [C++] Hash Join Microbenchmarks

save-buffer · westonpace · commit ab86daf3f7c8 · 2022-01-12T16:35:33.000-10:00
Added microbenchmarks for Hash Join (using OpenMP for parallelism for now). Also added a python script that analyzes the benchmark file and makes a lot of pretty graphs! Closes apache#11876 from save-buffer/sasha_benchmark Authored-by: Sasha Krassovsky <krassovskysasha@gmail.com> Signed-off-by: Weston Pace <weston.pace@gmail.com>
diff --git a/cpp/CMakePresets.json b/cpp/CMakePresets.json
@@ -40,6 +40,8 @@
       "cacheVariables": {
         "ARROW_BUILD_BENCHMARKS": "ON",
         "ARROW_BUILD_BENCHMARKS_REFERENCE": "ON",
+        "ARROW_BUILD_OPENMP_BENCHMARKS": "ON",
+        "ARROW_BUILD_DETAILED_BENCHMARKS": "OFF",
         "CMAKE_BUILD_TYPE": "RelWithDebInfo"
       }
     },
diff --git a/cpp/cmake_modules/BuildUtils.cmake b/cpp/cmake_modules/BuildUtils.cmake
@@ -609,6 +609,10 @@ function(ADD_BENCHMARK REL_BENCHMARK_NAME)
     set(ARG_LABELS benchmark)
   endif()
 
+  if(ARROW_BUILD_DETAILED_BENCHMARKS)
+    target_compile_definitions(${BENCHMARK_NAME} PRIVATE ARROW_BUILD_DETAILED_BENCHMARKS)
+  endif()
+
   add_test(${BENCHMARK_NAME}
            ${BUILD_SUPPORT_DIR}/run-test.sh
            ${CMAKE_BINARY_DIR}
diff --git a/cpp/cmake_modules/DefineOptions.cmake b/cpp/cmake_modules/DefineOptions.cmake
@@ -168,6 +168,12 @@ if("${CMAKE_SOURCE_DIR}" STREQUAL "${CMAKE_CURRENT_SOURCE_DIR}")
   define_option(ARROW_BUILD_BENCHMARKS_REFERENCE
                 "Build the Arrow micro reference benchmarks" OFF)
 
+  define_option(ARROW_BUILD_OPENMP_BENCHMARKS
+                "Build the Arrow benchmarks that rely on OpenMP" OFF)
+
+  define_option(ARROW_BUILD_DETAILED_BENCHMARKS
+                "Build benchmarks that do a longer exploration of performance" OFF)
+
   if(ARROW_BUILD_SHARED)
     set(ARROW_TEST_LINKAGE_DEFAULT "shared")
   else()
diff --git a/cpp/src/arrow/compute/exec/CMakeLists.txt b/cpp/src/arrow/compute/exec/CMakeLists.txt
@@ -32,6 +32,19 @@ add_arrow_compute_test(util_test PREFIX "arrow-compute")
 
 add_arrow_benchmark(expression_benchmark PREFIX "arrow-compute")
 
+if(ARROW_BUILD_OPENMP_BENCHMARKS)
+  find_package(OpenMP REQUIRED)
+  add_arrow_benchmark(hash_join_benchmark
+                      PREFIX
+                      "arrow-compute"
+                      EXTRA_LINK_LIBS
+                      OpenMP::OpenMP_CXX)
+  if(MSVC)
+    target_compile_options(arrow-compute-hash-join-benchmark
+                           PRIVATE "-openmp:experimental -openmp:llvm")
+  endif()
+endif()
+
 add_arrow_compute_test(ir_test
                        PREFIX
                        "arrow-compute"
diff --git a/cpp/src/arrow/compute/exec/hash_join_benchmark.cc b/cpp/src/arrow/compute/exec/hash_join_benchmark.cc
diff --git a/cpp/src/arrow/compute/exec/hash_join_graphs.py b/cpp/src/arrow/compute/exec/hash_join_graphs.py

Original file line number	Diff line number	Diff line change
`@@ -40,6 +40,8 @@`
`40`	`40`	`"cacheVariables": {`
`41`	`41`	`"ARROW_BUILD_BENCHMARKS": "ON",`
`42`	`42`	`"ARROW_BUILD_BENCHMARKS_REFERENCE": "ON",`
	`43`	`+ "ARROW_BUILD_OPENMP_BENCHMARKS": "ON",`
	`44`	`+ "ARROW_BUILD_DETAILED_BENCHMARKS": "OFF",`
`43`	`45`	`"CMAKE_BUILD_TYPE": "RelWithDebInfo"`
`44`	`46`	`}`
`45`	`47`	`},`