Benchmark: Named vs Explicit Returns in Go

crg@crg.eti.br (Cesar Gimenes) — Tue, 17 Dec 2024 03:00:24 -0300

A question came up in our Go study group: are named returns faster than explicit returns? That was the hypothesis.

Fortunately, Go has good tools to test this.

Let’s write some code with two sum functions: one with named returns and one with explicit returns.

package main

import (
 "testing"
)

// Função com retornos nomeados
func somaNomeados(a, b int) (resultado int, err error) {
 resultado = a + b
 return
}

// Função com retornos explícitos
func somaExplicitos(a, b int) (int, error) {
 return a + b, nil
}

Next, we create the benchmark functions. The names of these functions matter. Go’s testing tool looks for functions that start with “Benchmark”.

Benchmark on Cesar Gimenes

Benchmark: Named vs Explicit Returns in Go