Re: [Gnash-dev] Releasing 0.7.2 RSN (i.e. on monday)...

[Rob Savoye]

strk wrote:
> Well, at this point I'm even more convinced about the stack corruption
> I guess 'wiping' out deleted memory might help us, but it's probably
> something that's better implemented as a delete wrapper itself ?

  You mean setting the memory before freeing it to zeros ? I'd actually
try dmalloc or efence in this case. Then is makes all allocations get
their own page, useful for tracking down some memory corruption
problems. The other hack we could do (probably only in HEAD), would I
could port over a memory allocator I wrote for a different GNU project
that allocates a big chunk of memory upfront, and then new/delete work
out of that pool. The pool can also be resized. This is the simplest way
to make sure all memory manipulation is thread safe.

        - rob -