[uxp100]

I don't think there are any present, but fault injection mitigations (besides constant time execution, which does help) are something I'd be interested seeing their approach to.

I think it would be very useful if there were a way to formalize and prove that a particular mitigation was effective against a particular model of a fault. So if you assume a fault looks like X (single instruction skip, bit flips, multiple instruction skip, etc.) you could prove that a given SW mitigation was effective at retaining the other properties that are desired for the library.

This would be specific to a particular architecture of course.