fix: benchmark with incorrect minimum value

example/setup_test.go

@@ -18,12 +18,35 @@ func BenchmarkLargeSetup(b *testing.B) {
 
 func BenchmarkLargeSetupInLoop(b *testing.B) {
 	var result uint64
+
+	// The `testing` package starts the timer automatically at the beginning of the benchmark. If we try to
+	// stop the timer before the expensiveOperation, and then start it again afterwards we'll run into an issue
+	// where the difference from the function start to the first b.StopTimer() call is included in the timing.
+	// ```
+	// 	for i := 0; i < b.N; i++ {
+	// 		b.StopTimer()
+	// 		expensiveOperation()
+	// 		b.StartTimer()
+	// 		b.Log("Running iteration", i, b.Elapsed())
+	//
+	// 		result = actualWork()
+	// 		b.Log("Running iteration", i, b.Elapsed())
+	// 	}
+	// ```
+	// This will print two timings:
+	// setup_test.go:26: Running iteration 0 782ns
+	// setup_test.go:29: Running iteration 0 7.560911ms
+	//
+	// However, the first timing (782ns) is misleading because it's not related to the benchmark. To avoid this, we have to
+	// stop the timer in the beginning of the benchmark function.
+	b.StopTimer()
+	b.ResetTimer()
 	for i := 0; i < b.N; i++ {
-		b.StopTimer()
 		expensiveOperation()
-		b.StartTimer()
 
+		b.StartTimer()
 		result = actualWork()
+		b.StopTimer()
 	}
 	runtime.KeepAlive(result)
 }

go-runner/src/main.rs

@@ -4,7 +4,8 @@ use std::path::Path;
 fn main() -> anyhow::Result<()> {
     env_logger::builder()
         .parse_env(env_logger::Env::new().filter_or("CODSPEED_LOG", "info"))
-        .filter_module("handlebars", log::LevelFilter::Off)
+        .filter_module("handlebars", log::LevelFilter::Warn)
+        .filter_module("dircpy", log::LevelFilter::Warn)
         .format_timestamp(None)
         .init();