#16411: Inconsistent -Wpartial-fields warnings with (~) vs. (~~)

--===============0510253714070801593==
Content-Type: multipart/related;
 boundary="===============4466499276756513896=="

--===============4466499276756513896==
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 8bit

#16411: Inconsistent -Wpartial-fields warnings with (~) vs. (~~)
-------------------------------------+-------------------------------------
           Reporter:  RyanGlScott    |             Owner:  (none)
               Type:  bug            |            Status:  new
           Priority:  normal         |         Milestone:
          Component:  Compiler       |           Version:  8.6.4
           Keywords:                 |  Operating System:  Unknown/Multiple
       Architecture:                 |   Type of failure:  Incorrect
  Unknown/Multiple                   |  error/warning at compile-time
          Test Case:                 |        Blocked By:
           Blocking:                 |   Related Tickets:
Differential Rev(s):                 |         Wiki Page:
-------------------------------------+-------------------------------------
 In this program:

 {{{#!hs
 {-# LANGUAGE GADTs #-}
 {-# LANGUAGE TypeOperators #-}
 {-# OPTIONS_GHC -Wpartial-fields #-}
 module Bug where

 import Data.Type.Equality

 data T1 z where
   MkT1a :: { rec1 :: () } -> T1 Int
   MkT1b :: (z ~ Bool) => T1 z

 data T2 z where
   MkT2a :: { rec2 :: () } -> T2 Int
   MkT2b :: (z ~~ Bool) => T2 z
 }}}

 I get no `-Wpartial-fields`–related warnings for `rec1`, but I do for
 `rec2`:

 {{{
 GHCi, version 8.6.4: http://www.haskell.org/ghc/  :? for help
 Loaded GHCi configuration from /home/rgscott/.ghci
 [1 of 1] Compiling Bug              ( Bug.hs, interpreted )

 Bug.hs:13:14: warning: [-Wpartial-fields]
     Use of partial record field selector: ‘rec2’
    |
 13 |   MkT2a :: { rec2 :: () } -> T2 Int
    |              ^^^^
 }}}

 This feels strangely inconsistent, since if GHC doesn't warn for `rec1`,
 which uses `(~)`, then I feel like it shouldn't warn for `rec2` either, as
 the only difference is that the surrounding constructor uses a slightly
 different equality constraint `(~~)`.

 The reason this happens is because the `dataConCannotMatch` function has
 special reasoning for `(~)`, but not `(~~)`. Patch incoming.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/16411>