Ok, the problem is in `rnConDecl` in the `ConDeclGADT` case, which in
#11028: Refactor ConDecl -------------------------------------+------------------------------------- Reporter: simonpj | Owner: alanz Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.2 Resolution: | Keywords: 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 simonpj): Replying to [comment:22 alanz]: this branch simply calls `rnHsSigType`.
Instead it should perform as per the original, which splits the result
type,leaving it as `MaybeO ex t` in this case. Wait! Why? What would the difference be in the result of `rnConDecl`? I'm suspicious; I'd like to understand better. An example perhaps? -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/11028#comment:23 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler