Re: [GHC] #13117: Derived functor instance for void types handles errors badly