~/Benchmarking In Go

Aug 22, 2025

1. Structure of a Go Benchmark

func BenchmarkMyFunc(b *testing.B) {
    for i := 0; i < b.N; i++ {
        MyFunc()
    }
}

2. Running Benchmarks

Run all benchmarks: go test -bench .
Run specific benchmarks: go test -bench 'Pattern'
Change default time per benchmark: go test -bench . -benchtime 3s

3. Prevent Compiler Optimizations

Assign return values to a package-level variable to avoid function calls being optimized away:

var result int
func BenchmarkDoSomething(b *testing.B) {
    var r int
    for i := 0; i < b.N; i++ {
        r = DoSomething()
    }
    result = r
}

5. Expensive Setup

Use b.ResetTimer() after setup code to exclude setup from timing.

6. Measuring Allocations

Show memory usage: go test -bench . -benchmem

7. Example Use Case

Comparing set intersection using different types ([]string, map[string]bool, map[string]struct{}) shows how benchmarks yield data on speed and memory use.

8. Interpreting Results

More iterations = lower noise.
Look for execution time (ns/op) and allocations (B/op, allocs/op).
Run multiple times for reliability.

Tips

Use benchmarking in Go to guide performance decisions.
Structure functions with b *testing.B.
Prevent code from being optimized away.
Run with go test -bench . (add -benchmem for allocations).
Focus on real-world bottlenecks; optimize thoughtfully.

Tags: [go]