3 ms·
One of the cool proofs that comes with sel4 IIRC is that it doesn't invoke undefined behavior. And since it does end to end validation, the proof would fail if
by monocasa 9y ago
One of the cool proofs that comes with sel4 IIRC is that it doesn't invoke undefined behavior. And since it does end to end validation, the proof would fail if the generated binary doesn't perform stuff like null checks when they are specified by the abstract specification written along side the kernel itself.
IMO taking the compiler out of the trusted computing base is the way to go.
- fmap 9y ago> IMO taking the compiler out of the trusted computing base is the way to go. That is the point of compiler verification after all. I'm biased here, but I think that another important point is to work out what the specification of a compiler is in the first place. Obviously, we've made a poor job of communicating this for C/C++, since so many people write broken code...