mkWeak# has a very dangerous type

The mkWeak# primop has a type looking like this:

primop  MkWeakOp "mkWeak#" GenPrimOp
   o -> b -> c -> State# RealWorld -> (# State# RealWorld, Weak# b #)

But the third argument is allowed to be a null pointer (indicating no finaliser). That is why we see this code in GHC.TopHandler:

mkWeakThreadId :: ThreadId -> IO (Weak ThreadId)
mkWeakThreadId t@(ThreadId t#) = IO $ \s ->
   case mkWeak# t# t (unsafeCoerce# 0#) s of 
      (# s1, w #) -> (# s1, Weak w #)

But coercing a non-pointer to a pointer is dangerous. Very dangerous. Could crash the garbage collector. This is Not Right.

I found this when tidying up Core Lint. For now I have commented out the extra test, but it should be there!

lintCoercion (UnsafeCo ty1 ty2)
  = do { k1 <- lintType ty1
       ; _k2 <- lintType ty2
--       ; unless (k1 `eqKind` k2) $ 
--         failWithL (hang (ptext (sLit "Unsafe coercion changes kind"))
--                       2 (ppr co))
       ; return (k1, ty1, ty2) }

}

Trac metadata

Trac field	Value
Version	7.4.1
Type	Bug
TypeOfFailure	OtherFailure
Priority	normal
Resolution	Unresolved
Component	Compiler
Test case
Differential revisions
BlockedBy
Related
Blocking
CC
Operating system
Architecture