Re: List container strawman

[Darren New <dnew@san.rr.com>]

Jeffrey Carter wrote:
> > a) The memory at that location gets reallocated to something else in the
> > meantime, which then puts what looks like a valid value in there.
> 
> It's certainly possible that the memory could be reallocated to the same
> list, so the dangling reference becomes valid again. I did say *fairly*
> foolproof. :)

If you really want it to be foolproof, then for each "Location" pointer,
you keep a pointer to the list, a pointer to the element being indexed,
and a list number counter. In each list head, you keep a list number
counter, a counter of the number of Location pointers outstanding, and
the head and tail of the list.

Creating a Location copies the list number counter into the Location
record and otherwise initializes it. It also increments the reference
count in the List head. Deleting a Location decrements the ref count in
the list head. Deleting the list is defered (via finalization) until the
reference count goes to zero.

Each access to the pointer in the Location can run thru the linked list
that the Location claims it points to, to make sure the node it points
to is still on the list. This makes indirection become O(N) instead of
O(1), but it's also the kind of check that would be easy to turn on and
off at runtime, let alone compile time.

Did this for a Pascal compiler, so students using dangling pointers
would be told that instead of just dumping core.

-- 
Darren New 
San Diego, CA, USA (PST). Cryptokeys on demand.
     Sore feet from standing in line at airport
                 security checkpoints: Jet Leg.