What is the performance difference?

Using the following benchmark (https://play.golang.org/p/rCxWqmQXvq5), I get these numbers:

BenchmarkSumRange-8   	    3000	    437643 ns/op
BenchmarkSumIndex-8   	    3000	    482681 ns/op

So, using an explicit index seems to be slower than using "range".

Something I noticed is that, if I write "sumRange" as follows, the difference disappears:

func sumRange(slc []byte) byte {
    res := byte(0)
    for i := range slc {
        res += slc[i]
    }
    return res
}

/cc @randall77 @josharian @cherrymui

This appears to occur because of choices in the scheduler.
The inner loop in sumRange schedules like this:

v19 (5) = MOVBloadidx1 <byte> v26 v10 v1 (v[byte])
v24 (+5) = ADDQconst <int> [1] v10
v21 (+6) = ADDL <byte> v30 v19 (res[byte])

The inner loop in sumIndex schedules like this:

v27 (+13) = ADDQconst <int> [1] v11 (i[int])
v23 (+14) = MOVBloadidx1 <byte> v18 v11 v1
v24 (14) = ADDL <byte> v33 v23 (res[byte])

The schedule in sumIndex is a problem because the ADDQconst can't overwrite its argument, as the argument is not dead yet. So it has to target a different register and then get copied back.

This has been an issue for a while. I noted it in 2016: #16122 (comment)

Perhaps just a tweak to the scheduler is needed.

I'm unclear as to why the scheduler makes different choices here. Might just be how the values happened to be ordered on entry to the scheduler.