rl_resize_terminal was added in readline 4.0, released in 1999.  I think we can rely on it. =)

I don't know editline, but it probably doesn't include the readline 6.3 change that makes this necessary, so your patch for it could just IFDEF out that line.

I think SIG_DFL is a pretty reasonable return value, actually.  It's clearly not SIG_IGN or a python handle, it's not really "unknown", and it can be described as default behavior: it is the default way readline is supposed to work.

Finally, while I agree that it is conceivable that one could try to use readline + other terminal mucking with SIGWINCH, (a) this didn't work before the 6.3 change, (b) it doesn't work now, (c) it's far less common than the standard case of using readline normally, and (d) supporting such usage would be complicated and require a readline module API change.  Let's start by supporting normal usage.