[GHC] #11833: GHC can't derive polykinded instance of polykinded typeclass for newtype that requires a class constraint

13 Apr 2016

      #11833: GHC can't derive polykinded instance of polykinded typeclass for newtype
that requires a class constraint
-------------------------------------+-------------------------------------
           Reporter:  RyanGlScott    |             Owner:  RyanGlScott
               Type:  bug            |            Status:  new
           Priority:  normal         |         Milestone:
          Component:  Compiler       |           Version:  8.0.1
  (Type checker)                     |
           Keywords:                 |  Operating System:  Unknown/Multiple
       Architecture:                 |   Type of failure:  GHC rejects
  Unknown/Multiple                   |  valid program
          Test Case:                 |        Blocked By:
           Blocking:                 |   Related Tickets:  #8865
Differential Rev(s):                 |         Wiki Page:
-------------------------------------+-------------------------------------
 GHC chokes when trying to derive the following:

 {{{#!hs
 {-# LANGUAGE GeneralizedNewtypeDeriving #-}
 {-# LANGUAGE PolyKinds #-}
 module Example where

 class Category (cat :: k -> k -> *) where
   catId   :: cat a a
   catComp :: cat b c -> cat a b -> cat a c

 newtype T (c :: * -> * -> *) a b = MkT (c a b) deriving Category
 }}}

 with the following error:

 {{{
 $ /opt/ghc/8.0.1/bin/ghc Example.hs -fprint-explicit-kinds
 [1 of 1] Compiling Example          ( Example.hs, Example.o )

 Example.hs:9:57: error:
     • No instance for (Category * c)
         arising from the 'deriving' clause of a data type declaration
       Possible fix:
         use a standalone 'deriving instance' declaration,
           so you can specify the instance context yourself
     • When deriving the instance for (Category * (T c))
 }}}

 I know what is happening. Unlike in #8865, where we were deriving
 `Category` for a concrete type like `Either`, in the above example we are
 attempting to derive an instance of the form:

 {{{#!hs
 instance Category * c => Category (T * c) where ...
 }}}

 (using `-fprint-explicit-kinds` syntax). But `validDerivPred` is checking
 if `sizePred (Category * c)` equals the number of free type variables in
 `Category * c`. But note that `sizePred` counts both type variables
 //and// type constructors, and `*` is a type constructor! So
 `validDerivPred` erroneously rejects the above instance.

 To fix this behavior, I think we just need to change `validDerivPred` to
 only consider the //visible// arguments of `Category` (i.e., only `c`).
 This should be a pretty easy fix - patch incoming.

--
Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/11833
GHC http://www.haskell.org/ghc/
The Glasgow Haskell Compiler