News for hackers. Stuff that matters.
Re:SeL4 on RISC-V Verified to Binary Code by archi42 on Wednesday May 05, 2021 @06:24AM (#27048199)
> This checker is proven correct, and because it is simpler, its proof of correctness is easier.

Still correct ;)

What I belief GP (fuklief) meant was: The SMT solver might still be incorrect. Like the register allocator itself in CompCert. But for CompCert, the result is checked by a formally verified checker. I do not know if that's true or not for the result of the SMT solver.

Re:SeL4 on RISC-V Verified to Binary Code by creata on Wednesday May 05, 2021 @07:50AM (#27048732)
Don't SMT solvers output proof objects so that the results can be easily independently verified?
Re:SeL4 on RISC-V Verified to Binary Code by syrak on Wednesday May 05, 2021 @09:21AM (#27049471)
Even if you had a verified SMT checker, you also need to prove that the encoding of the problem in SMT is correct, i.e., that a proof does translate to a valid register allocation solution, to achieve comparable guarantees to CompCert.