This fixes the breakage mentioned by @paulwalker-arm in D111237. We should only be searching for Type Promotions of Integers for PHI + ZExt.
LGTM. But, now I'm wondering whether we could handle vectors... ;p
Hehe, I don't see why not but that feels like another potential improvement rather than a fix ;) I'll go ahead and commit this to unblock @paulwalker-arm if that's OK?