Re: [GHC] #12700: Don't warn about redundant constraints for type equalities