RE: Running a "final" finaliser

The other thing that strikes me about this is don't we also have the same potential problem with weak pointer finalisers? Can they be supported in Haskell without pre-emptive concurrency?

Weak pointers are a GHC-only extension too, at the moment. They require some careful GC algorithms as well as support for Haskell finalisers. GHC uses weak pointers to implement ForeignPtr internally. Cheers, Simon