Re: [GHC] #14917: Allow levity polymorhism in binding position

19 Mar 2018

      #14917: Allow levity polymorhism in binding position
-------------------------------------+-------------------------------------
        Reporter:  andrewthad        |                Owner:  (none)
            Type:  feature request   |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:  8.2.2
      Resolution:                    |             Keywords:
                                     |  LevityPolymorphism
Operating System:  Unknown/Multiple  |         Architecture:
                                     |  Unknown/Multiple
 Type of failure:  None/Unknown      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by goldfire):

 Sure we do. I have not explored ''how'' we do, but sure we do. This
 program compiles and runs:

 {{{#!hs
 data List (b :: TYPE (TupleRep [IntRep, DoubleRep, LiftedRep])) = Nil |
 Cons b (List b)

 mapUbx :: forall (a :: Type) (b :: TYPE (TupleRep [IntRep, DoubleRep,
 LiftedRep])). (a -> b) -> [a] -> List b
 mapUbx _ [] = Nil
 mapUbx f (x : xs) = Cons (f x) (mapUbx f xs)

 blargh :: forall a. Int# -> Double# -> a -> (# Int#, Double#, a #)
 blargh x y = (#,,#) x y

 strange = mapUbx (blargh 3# 2.78##) [True, False]

 printList :: Show b => List (# Int#, Double#, b #) -> IO ()
 printList Nil = return ()
 printList (Cons (# n, d, b #) xs) = do
   print (I# n)
   print (D# d)
   print b
   printList xs

 main = do
   printList strange
 }}}

 We probably just eta-expand `blargh`.

-- 
Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/14917#comment:10
GHC http://www.haskell.org/ghc/
The Glasgow Haskell Compiler