An official website of the United States government
Intermediate verification languages like Why3 and Boogie have made it much easier to build program verifiers, transforming the process into a logic compilation problem rather than a proof automation one. Why3 in particular implements a rich logic for program specification with polymorphism, algebraic data types, recursive functions and predicates, and inductive predicates; it translates this logic to over a dozen solvers and proof assistants. Accordingly, it serves as a backend for many tools, including Frama-C, EasyCrypt, and GNATProve for Ada SPARK. But how can we be sure that these tools are correct? The alternate foundational approach, taken by tools like VST and CakeML, provides strong guarantees by implementing the entire toolchain in a proof assistant, but these tools are harder to build and cannot directly take advantage of SMT solver automation. As a first step toward enabling automated tools with similar foundational guarantees, we give a formal semantics in Coq for the logic fragment of Why3. We show that our semantics are useful by giving a correct-by-construction natural deduction proof system for this logic, using this proof system to verify parts of Why3’s standard library, and proving sound two of Why3’s transformations used to convert terms and formulas into the simpler logics supported by the backend solvers.
more »
« less
A Mechanized First-Order Theory of Algebraic Data Types with Pattern Matching
Algebraic data types (ADTs) and pattern matching are widely used to write elegant functional programs and to specify program behavior. These constructs are critical to most general-purpose interactive theorem provers (e.g. Lean, Rocq/Coq), first-order SMT-based deductive verifiers (e.g. Dafny, VeriFast), and intermediate verification languages (e.g. Why3). Such features require layers of compilation - in Rocq, pattern matches are compiled to remove nesting, while SMT-based tools further axiomatize ADTs with a first-order specification. However, these critical steps have been omitted from prior formalizations of such toolchains (e.g. MetaRocq). We give the first proved-sound sophisticated pattern matching compiler (based on Maranget’s compilation to decision trees) and first-order axiomatization of ADTs, both based on Why3 implementations. We prove the soundness of exhaustiveness checking, extending pen-and-paper proofs from the literature, and formulate a robustness property with which we find an exhaustiveness-related bug in Why3. We show that many of our proofs could be useful for reasoning about any first-order program verifier supporting ADTs.
more »
« less
- Award ID(s):
- 2219758
- PAR ID:
- 10654533
- Editor(s):
- Forster, Yannick; Keller, Chantal
- Publisher / Repository:
- Schloss Dagstuhl – Leibniz-Zentrum für Informatik
- Date Published:
- Volume:
- 352
- ISSN:
- 1868-8969
- Page Range / eLocation ID:
- 5:1-5:20
- Subject(s) / Keyword(s):
- Pattern Matching Compilation Algebraic Data Types First-Order Logic Software and its engineering → Semantics Theory of computation → Logic and verification
- Format(s):
- Medium: X Size: 20 pages; 955296 bytes Other: application/pdf
- Size(s):
- 20 pages 955296 bytes
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract This paper presents , a new intermediate verification language and deductive verification tool that provides inbuilt support for concurrency reasoning. ’s meta-theory is based on the higher-order concurrent separation logic Iris, incorporating core features such as user-definable ghost state and thread-modular reasoning via shared-state invariants. To achieve better accessibility and enable proof automation via SMT solvers, restricts Iris to its first-order fragment. The entailed loss of expressivity is mitigated by a higher-order module system that enables proof modularization and reuse. We provide an overview of the language and describe key aspects of the supported proof automation. We evaluate on a benchmark suite of verification tasks comprising linearizability and memory safety proofs for common concurrent data structures and clients as well as one larger case study. Our evaluation shows that improves over existing proof automation tools for Iris in terms of verification times and usability. Moreover, the tool significantly reduces the proof overhead compared to proofs constructed using the Iris/Rocq proof mode.more » « less
-
Dillig, Isil; Jhala, Ranjit (Ed.)We present Leapfrog, a Coq-based framework for verifying equivalence of network protocol parsers. Our approach is based on an automata model of P4 parsers, and an algorithm for symbolically computing a compact representation of a bisimulation, using "leaps." Proofs are powered by a certified compilation chain from first-order entailments to low-level bitvector verification conditions, which are discharged using off-the-shelf SMT solvers. As a result, parser equivalence proofs in Leapfrog are fully automatic and push-button. We mechanically prove the core metatheory that underpins our approach, including the key transformations and several optimizations. We evaluate Leapfrog on a range of practical case studies, all of which require minimal configuration and no manual proof. Our largest case study uses Leapfrog to perform translation validation for a third-party compiler from automata to hardware pipelines. Overall, Leapfrog represents a step towards a world where all parsers for critical network infrastructure are verified. It also suggests directions for follow-on efforts, such as verifying relational properties involving security.more » « less
-
Verifying relations between programs arises as a task in various verification contexts such as optimizing transformations, relating new versions of programs with older versions (regression verification), and noninterference. However, relational verification for programs acting on dynamically allocated mutable state is not well supported by existing tools, which provide a high level of automation at the cost of restricting the programs considered. Auto-active tools, on the other hand, require more user interaction but enable verification of a broader class of programs. This article presents WhyRel, a tool for the auto-active verification of relational properties of pointer programs based on relational region logic. WhyRel is evaluated through verification case studies, relying on SMT solvers orchestrated by the Why3 platform on which it builds. Case studies include establishing representation independence of ADTs, showing noninterference, and challenge problems from recent literature.more » « less
-
Verification of program safety is often reducible to proving the unsatisfiability (i.e., validity) of a formula in Satisfiability Modulo Theories (SMT): Boolean logic combined with theories that formalize arbitrary first-order fragments. Zeroknowledge (ZK) proofs allow SMT formulas to be validated without revealing the underlying formulas or their proofs to other parties, which is a crucial building block for proving the safety of proprietary programs. Recently, Luo et al. studied the simpler problem of proving the unsatisfiability of pure Boolean formulas but does not support proofs generated by SMT solvers. This work presents ZKSMT, a novel framework for proving the validity of SMT formulas in ZK. We design a virtual machine (VM) tailored to efficiently represent the verification process of SMT validity proofs in ZK. Our VM can support the vast majority of popular theories when proving program safety while being complete and sound. To demonstrate this, we instantiate the commonly used theories of equality and linear integer arithmetic in our VM with theory-specific optimizations for proving them in ZK. ZKSMT achieves high practicality even when running on realistic SMT formulas generated by Boogie, a common tool for software verification. It achieves a three-order-of-magnitude improvement compared to a baseline that executes the proof verification code in a general ZK system.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.4230/lipics.itp.2025.5
