[GHC] #12918: Make DefaultSignatures more particular about their types on the RHS of a context

[GHC]

#12918: Make DefaultSignatures more particular about their types on the RHS of a
context
-------------------------------------+-------------------------------------
        Reporter:  RyanGlScott       |                Owner:  simonpj
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.2.1
       Component:  Compiler (Type    |              Version:  8.0.2-rc1
  checker)                           |
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
                                     |  Unknown/Multiple
 Type of failure:  None/Unknown      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #12784            |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by RyanGlScott):

 OK, I've made some progress on this, but I'm once again stuck. I'm using
 `tcUnifyTy` once again to check if two type signatures coincide according
 to the criteria listed above. But there's a case it doesn't catch:

 {{{#!hs
 class Foo a where
   bar :: forall b. a -> b
   default bar :: a -> Int
   bar = ...
 }}}

 Here, `tcUnifyTy` happily unifies `b` with `Int`, yielding a false
 positive. So I need to find some way to prevent `b` from unifying with
 concrete types like `Int`. (Does skolemizing it accomplish this?) But we'd
 also want this to be accepted:

 {{{#!hs
 class Foo a where
   bar :: forall b. a -> b
   default bar :: forall c. a -> c
   bar = ...
 }}}

 Since `a -> c` is just a simple alpha-renaming of `a -> b`. So how can we
 tell `b` to unify with `c` but not with `Int`?

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12918#comment:5>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler