Michal Zalewski wrote:
> A quick test case that crashes while trying to follow partly
> user-dependent corrupted pointers near valid memory regions (can be forced
> to write, too):
>
> http://lcamtuf.coredump.cx/ietrap/testme.html
>
> Firefox problem is being tracked here:
> https://bugzilla.mozilla.org/show_bug.cgi?id=371321

This bug was fixed in 2.0.0.2, released Friday Feb 23.