TMI: No, this was not a hazing ritual
Fortunately, building gcc 4.2 worked, so I spent most of today reconciling my changes with the updated version of the front-end. Eventually, all of the tests passed except a few, and those were a few that I wasn't going to fix (two that seemed to be failing for reasons unrelated to my changes, and two that failed because they were using new Function() to create generators -- it's not clear what the syntax should be for that now, or if it should be allowed; I'll leave that for someone else to decide).
I'm also not really sure why this had to take all day, gosh (actually, an 11-hour day). That doesn't matter, though, because it's done (at least up to receiving review for it). And tomorrow I work on Rust again.