The Revenge of Finalizers

[Alastair Reid]

> Does this run into problems with exceptions?

Good question.

I think the answer depends on what happens if a Haskell finalizer
raises an IOError, calls System.exit or (Hugs/GHC extension) raises an
Exception.  I don't think the ffi spec comments on these.  I'd guess
the answer is:

1) Returning an IOError aborts the program.  

   (In Hugs, we'd probably want the other finalizers to be run first,
   the same may be true of GHC.)

2) Calling exit aborts the program.

   (In Hugs, we'd probably want the other finalizers to be run first,
   but I don't know what GHC would want.)

3) Not specified to avoid unwanted dependency (but if specified it would
   be the same as for IOErrors).

> Also, this is a nested call to eval(), in a primitive, which can
> invoke an IO action and therefore re-enter Haskell without going
> through unsafePerformIO.  Is that safe?

Yes, I think so.  Most calls to IO actions from primitives are safe
and I believe these ones are too.  (In some ways, its the rarity of
race conditions, etc. biting you that disturbs me most about Haskell
finalizers: it's hard to remember to do the right thing when the wrong
thing happens so rarely.)

--
Alastair