> [Severity: Medium]
> Does this early return prevent the hardware register from being updated to
> clear the calibration offset when steps evaluates to zero?

Dunno but we have to stop somewhere. This should be fixed seperate of
this patch series.