It is especially ill-defined when the input itself may have been rounded already, e.g. round(0.15, 1) = 0.1 because 0.15 is not exact in binary64 and the binary

It is especially ill-defined when the input itself may have been rounded already, e.g. round(0.15, 1) = 0.1 because 0.15 is not exact in binary64 and the binary64 value closest to 0.15 is a bit smaller than 0.15. If you have to retain the exact rounded value for some reason, you should switch to fixed point from that point on.