> It's a real bug and a correct fix, as we've discussed before, > but if the kernel loses this race I believe it will crash dereferencing nil. > Lucio showed a kernel that was very much still running.
you are correct. i was confused. the bug reported looks like a missing waserror(). - erik
