Verified Functional Programming of an Abstract Interpreter

Abstract interpreters are complex pieces of software: even if the abstract interpretation theory and companion algorithms well understood, their implementations subject to bugs, that might question soundness computations. While some formally verified have been written in past, writing understanding them requires expertise use proof assistants, a non-trivial amount interactive proofs. This paper presents interpreter fully programmed proved correct F* programming environment. Thanks refinement types SMT prover capabilities we demonstrate substantial saving effort compared previous works based on assistants. Almost all code our implementation, proofs included, functional style, presented directly paper.