Future Work?

Some kind of generalised hardware fault reporting?

static analysis of code?

some kind of mechanism to prevent driver errors hosing the kernel?

Might need to emphasise modules/LKMs for space reasons in future?

Make state/statistics more visible to userland?