[MLton] Non-exhaustive exn fn match

[Stephen Weeks]

> I couldn't remember whether we finally decided to 
> introduce 
>    _address "symbol" : ptrTy;
> or just expect people to use
>    #1 (_symbol "symbol" : ptrTy, cbTy;)

You preferred the former, I was ambivalent, and I didn't hear anything
else.  So, it looks like _address to me.