search for: cryptol

...son semantics from a frontend/verification perspective. I saw there was a lot of interest in a proposed semantics in late 2016 and mid-2017. Perhaps this was due to the “Taming Undefined..." PLDI paper, which seemed like a nice writeup. I'd like to formalize this in a theorem prover or Cryptol. Since the mid 2017 discussion, is there an emerging consensus of the semantics of undef/poison? If so, is there a newer writeup than the PLDI paper? The paper semantics in Section 4 on bitvector types "isz” seem reasonable to me. However, I was surprised by the pointer semantics. I expect...