GADT bug in 20041124

[Simon Peyton-Jones]

Thanks for this.  Now fixed.

Simon

| -----Original Message-----
| From: cvs-all-bounces at haskell.org [mailto:cvs-all-bounces at haskell.org]
On Behalf Of Keith
| Wansbrough
| Sent: 26 November 2004 14:36
| To: cvs-ghc at haskell.org
| Subject: GADT bug in 20041124
| 
| Hi all (long time no see!)...
| 
| Keen to try out the new GADT stuff, I downloaded a CVS snapshot of
| ghc-6.3.20041124.  Having built it, I discovered a bug.  The program
| 
| > module GADTest where
| >
| > data Expr :: * -> * where
| >   EInt    :: Int                                 -> Expr Int
| >   EBool   :: Bool                                -> Expr Bool
| >   EIf     :: (Expr Bool) -> (Expr a) -> (Expr a) -> Expr a
| 
| yields
| 
| > kw217 at ged:~/Misc/Haskell$ ghc -fglasgow-exts -c GADTest.hs
| > ghc-6.3.20041124: panic! (the `impossible' happened, GHC version
| > 6.3.20041124):
| >         types/Type.lhs:(996,0)-(997,77): Non-exhaustive patterns in
| >         function zip_ty_env
| >
| >
| > Please report it as a compiler bug to
| > glasgow-haskell-bugs at haskell.org,
| > or http://sourceforge.net/projects/ghc/.
| 
| Note that removing the EIf line, or writing "Expr a" rather than "Expr
| :: * -> *", makes the error go away.  Writing "Expr :: *" instead
| gives a kind error, of course.
| 
| --KW 8-)
| _______________________________________________
| Cvs-ghc mailing list
| Cvs-ghc at haskell.org
| http://www.haskell.org/mailman/listinfo/cvs-ghc