Comment by aparent

What do mean by this? I don't think complexity classes give any provable security to asymmetric cryptography. It cannot be proven that the difficulty of breaking any of the currently used (or known?) asymmetric key cryptography is not in P (It can't even be proven to be NP-complete).