Given a general constraint against doing harm (suitably qualified), we can derive a large number of more specific constraints -- one for each way of doing harm. Which of these more narrow constraints we choose to identify with a special label will largely be a matter of convenience. Thus, depending upon our purposes, in some cases it might be useful to talk of a constraint against killing another, or a constraint against chopping off someone's leg, or even a constraint against punching someone in the nose, and so on. On the other hand, at other times and for other purposes it might be more helpful to subsume these three under a somewhat more general category; we might, for example, prefer to talk of a constraint against doing bodily damage.
So far as I can see, nothing significant turns on the question of which of these more specific constraints we isolate and identify in this way. Philosophically, the important point is simply that there are various ways of doing harm, so that if there is a constraint against doing harm, then there is a constraint against doing harm in each of these particular ways.
Of course, it may be helpful to remind ourselves of just how many ways there are to do harm. Not only can we inflict harm by doing bodily damage, we can also harm someone simply by causing them pain -- even if there is no ensuing damage to their body. Thus we can cause someone physical pain or mental anguish, humiliation, embarrassment, or offense. Similarly, we can harm others by interfering with their liberty, by enslaving them, confining them, or otherwise interfering with their autonomy -- their ability to control their own lives. And we can harm others by damaging or taking their resources, thus reducing their ability to promote their well-being.