From 081a1c9efe817f1a039d1752bfa9b3a3277c0cc0 Mon Sep 17 00:00:00 2001
From: "Albert-Jan N. Yzelman" <albertjan.yzelman@huawei.com>
Date: Sat, 11 Jan 2025 11:38:40 +0100
Subject: [PATCH] Also better label naming for hand-written compiler-optimised
 kernels

---
 tests/performance/bench_kernels.c |  4 ++++
 tests/performance/bench_kernels.h |  9 ++++++++-
 tests/performance/dot.cpp         | 10 ++++++++--
 tests/performance/fma.cpp         |  9 +++++++--
 tests/performance/reduce.cpp      |  8 +++++++-
 5 files changed, 34 insertions(+), 6 deletions(-)

diff --git a/tests/performance/bench_kernels.c b/tests/performance/bench_kernels.c
index 83b5b72ec..6c32a8ee6 100644
--- a/tests/performance/bench_kernels.c
+++ b/tests/performance/bench_kernels.c
@@ -21,6 +21,8 @@
 
 #ifdef BENCH_KERNELS_OPENMP
 
+bool bench_kernels_parallel() { return true; }
+
 void bench_kernels_axpy(
 	double * restrict a,
 	const double alpha, const double * restrict x,
@@ -120,6 +122,8 @@ void bench_kernels_reduce(
 
 #else
 
+bool bench_kernels_parallel() { return false; }
+
 void bench_kernels_axpy(
 	double * restrict a,
 	const double alpha, const double * restrict x,
diff --git a/tests/performance/bench_kernels.h b/tests/performance/bench_kernels.h
index 29901a1dc..c58dd890f 100644
--- a/tests/performance/bench_kernels.h
+++ b/tests/performance/bench_kernels.h
@@ -18,7 +18,7 @@
 
 #include <omp.h>
 #include <assert.h>
-#include <stddef.h> //for size_t
+#include <stddef.h> // for size_t
 
 
 #ifdef __cplusplus
@@ -41,10 +41,14 @@ extern "C" {
 		double * __restrict__ const, const double * __restrict__, const size_t
 	);
 
+	bool bench_kernels_parallel();
+
 }
 
 #else
 
+#include <stdbool.h> // for bool
+
 /**
  * Executes \f$ a = \alpha x + y \f$ for \a a, \a x, and \a y vectors of
  * length \a n.
@@ -89,5 +93,8 @@ void bench_kernels_reduce(
 	double * restrict const alpha, const double * restrict x, const size_t n
 );
 
+/** @returns Whether the kernels defined here are (shared-memory) parallel. */
+bool bench_kernels_parallel();
+
 #endif
 
diff --git a/tests/performance/dot.cpp b/tests/performance/dot.cpp
index e76f0ca9e..35d23df78 100644
--- a/tests/performance/dot.cpp
+++ b/tests/performance/dot.cpp
@@ -463,8 +463,14 @@ int main( int argc, char ** argv ) {
 	}
 
 	// start benchmark test 1
-	std::cout << "\nBenchmark label: compiler-optimised dot product on raw "
-		<< "arrays of size " << in.n << std::endl;
+	std::cout << "\nBenchmark label: ";
+	if( bench_kernels_parallel() ) {
+		std::cout << "parallel (OpenMP) ";
+	} else {
+		std::cout << "sequential (C) ";
+	}
+	std::cout << "compiler-optimised dot product on raw arrays of size " << in.n
+		<< std::endl;
 	if( bench.exec( &bench_raw, in, out, 1, outer, true ) != SUCCESS ) {
 		std::cerr << "Error launching raw benchmark test.\nTest FAILED." << std::endl;
 		return 60;
diff --git a/tests/performance/fma.cpp b/tests/performance/fma.cpp
index 980a642c8..334721c81 100644
--- a/tests/performance/fma.cpp
+++ b/tests/performance/fma.cpp
@@ -382,8 +382,13 @@ int main( int argc, char ** argv ) {
 		return 50;
 	}
 
-	std::cout << "\nBenchmark label: compiler-optimised axpy of size " << in.n
-		<< std::endl;
+	std::cout << "\nBenchmark label: ";
+	if( bench_kernels_parallel() ) {
+		std::cout << "parallel (OpenMP) ";
+	} else {
+		std::cout << "sequential (C) ";
+	}
+	std::cout << "compiler-optimised axpy of size " << in.n << std::endl;
 	rc = bench.exec( &(test< RAW >), in, out, 1, outer, true );
 	if( rc != SUCCESS || out.error != SUCCESS ) {
 		std::cerr << "Functional test exits with nonzero exit code. "
diff --git a/tests/performance/reduce.cpp b/tests/performance/reduce.cpp
index 8305a08cb..31e5b48bc 100644
--- a/tests/performance/reduce.cpp
+++ b/tests/performance/reduce.cpp
@@ -266,7 +266,13 @@ int main( int argc, char ** argv ) {
 		rc = bench.exec( &(test< LAMBDA >), in, out, 1, outer, true );
 	}
 	if( rc == SUCCESS ) {
-		std::cout << "\nBenchmark label: compiler-optimised reduce-to-scalar of size "
+		std::cout << "\nBenchmark label: ";
+		if( bench_kernels_parallel() ) {
+			std::cout << "parallel (OpenMP) ";
+		} else {
+			std::cout << "sequential (C) ";
+		}
+		std::cout << "compiler-optimised reduce-to-scalar of size "
 			<< in.n << std::endl;
 		rc = bench.exec( &(test< RAW >), in, out, 1, outer, true );
 	}