Post #1403794

@jripley (In our case we were pretty sure it wasn't actual bad RAM cells because -- on a system with no MMU -- the actual faulting address and backtrace would be different each time on the same unit.)