Using goto for optimization

Question

Hi i was wandering if using goto is good practice for optimization. i know it's a barbarian thing to do but, seriously.

for instance write this:

switch(command[0].cmd)
{
    case 0: // turn off
        s::s_off();
        S_time = command[0].length;
    break;
    case 1: // turn on
        s::s_on();
        S_time = command[0].length;
    break;
    case 4: // nop

    break;
}

like this:

switch(command[0].cmd)
{
    case 0: // turn off
        s::s_off();
        goto a;
    break;
    case 1: // turn on
        s::s_on();
        goto a;
    break;
    case 4: // nop
        goto b;
    break;
}

a:
S_time = command[0].length;
b:

Answer 1

It is not recommended to use goto frequently. But it is not "evil" and in many cases is much easier to put 1 goto instead of more complex logic only to avoid the "evil". And in your example single goto is enough:

switch(command[0].cmd)
{
    case 0: // turn off
        s::s_off();
        break;

    case 1: // turn on
        s::s_on();
        break;

    case 4: // nop
        goto b; // break after goto is useless
}

S_time = command[0].length;
b: