Hello, I think a good first step is to learn the theory of admissible representations of p-adic groups and for this Godement's notes on Jacquet-Langlands theory and then Casselman's unpublished book on p-adic groups (available from his website) are good starting points. A good way to read Casselman's notes is to rewrite the proof of every theorem explicitly for a small non-GL(2) split group, say GL(3) or Sp(4). If you want to see the big picture shape of the theory and how it is connected to Galois representations you can look at the Trieste notes by Prasad and Raghuram (available from Dipendra Prasad's homepage at Tata). And in your first attempt to learn the theory don't worry too much about supercuspidal representations; just treat them like black boxes or elementary particles. One very nice thing about Godement's notes is that the theory is immediately followed by applications to Hecke theory and L functions.

Let me add a couple of more points to address a question by Alex. You don't need much for Godement's notes; you do, however, need to be comfortable with Tate's thesis (p-adic integration, Haar measure, Poisson summation in the adelic setting, etc). I suppose that's the first thing you need to do if you haven't done already:

"READ TATE'S THESIS."

Tate's original writeup is amazing and much recommended. There is also the lovely book by Ramakrishnan and Valenza, as well as, of course, Bump's book. If you are already familiar with modular forms (and if not, why aren't you? :-) ) then Gelbart's classical book in the Princeton series is a good place to see the connections between the classical theory and the automorphic theory. When I was just starting to learn automorphic forms, I found Gelbart's treatment very nicely therapeutic.