How do you write Go benchmarks and what does -benchmem tell you?

// Benchmark function: func BenchmarkXxx(b *testing.B)
func BenchmarkJSONMarshal(b *testing.B) {
    user := User{ID: 1, Name: "Alice", Email: "alice@example.com", Age: 30}

    b.ResetTimer() // start timing AFTER setup (exclude allocation of user)
    for i := 0; i < b.N; i++ { // b.N is calibrated by the framework
        _, err := json.Marshal(user)
        if err != nil { b.Fatal(err) }
    }
}

// Memory allocation benchmark
func BenchmarkStringConcat(b *testing.B) {
    words := []string{"hello", "world", "foo", "bar", "baz"}

    b.ReportAllocs() // same as -benchmem for this specific benchmark
    b.ResetTimer()

    b.Run("plus operator", func(b *testing.B) {
        for i := 0; i < b.N; i++ {
            s := ""
            for _, w := range words { s += w } // alloc per iteration
            _ = s
        }
    })

    b.Run("strings.Builder", func(b *testing.B) {
        for i := 0; i < b.N; i++ {
            var sb strings.Builder
            sb.Grow(50) // pre-allocate — zero allocations inside loop
            for _, w := range words { sb.WriteString(w) }
            _ = sb.String()
        }
    })
}

// Run commands:
// go test -bench=. -benchmem ./...
// go test -bench=BenchmarkStringConcat -benchtime=5s -count=3

// Sample output:
// BenchmarkStringConcat/plus_operator-8   2345678   512 ns/op   256 B/op  5 allocs/op
// BenchmarkStringConcat/strings.Builder-8 9876543   123 ns/op    64 B/op  1 allocs/op
// Columns: name, iterations, ns per op, bytes per op, allocs per op