Marie-Christine Jakobs

Just Test What You Cannot Verify

Today, software verification is an established analysis method which can provide high guarantees for software safety. However, the resources (time and/or memory) for an exhaustive verification are not always available, and analysis then has to resort to other techniques, like testing. Most often, the already achieved partial verification results are discarded in this case, and testing has to start from scratch.

Certification for configurable program analysis

Configurable program analysis (CPA) is a generic concept for the formalization of different software analysis techniques in a single framework. With the tool CPAchecker, this framework allows for an easy configuration and subsequent automatic execution of analysis procedures ranging from data-flow analysis to model checking. The focus of the tool CPAchecker is thus on analysis. In this paper, we study configurability from the point of view of software certification. Certification aims at providing (via a prior analysis) a certificate of correctness for a program which is (a) tamper-proof and (b) more efficient to check for validity than a full analysis. Here, we will show how, given an analysis instance of a CPA, to construct a corresponding sound certification instance, thereby arriving at configurable program certification. We report on experiments with certification based on different analysis techniques, and in particular explain which characteristics of an underlying analysis allow us to design an efficient (in the above (b) sense) certification procedure.

Speed Up Configurable Certificate Validation by Certificate Reduction and Partitioning

Before execution, users should formally validate the correctness of software received from untrusted providers. To accelerate this validation, in the proof carrying code (PCC) paradigm the provider delivers the software together with a certificate, a formal proof of the software’s correctness. Thus, the user only checks if the attached certificate shows correctness of the delivered software.

Compact Proof Witnesses

Proof witnesses are proof artifacts showing correctness of programs wrt. safety properties. The recent past has seen a rising interest in witnesses as (a) proofs in a proof-carrying-code context, (b) certificates for the correct functioning of verification tools, or simply (c) exchange formats for (partial) verification results. As witnesses in all theses scenarios need to be stored and processed, witnesses are required to be as small as possible. However, software verification tools – the prime suppliers of witnesses – do not necessarily construct small witnesses.

Integrating Software and Hardware Verification

Verification of hardware and software usually proceeds separately, software analysis relying on the correctness of processors executing instructions. This assumption is valid as long as the software runs on standard CPUs that have been extensively validated and are in wide use. However, for processors exploiting custom instruction set extensions to meet performance and energy constraints the validation might be less extensive, challenging the correctness assumption.

Validity of Software Verification Results on Approximate Hardware

Approximate computing (AC) is an emerging paradigm for energy-efficient computation. The basic idea of AC is to sacrifice high precision for low energy by allowing hardware to carry out “approximately correct” calculations. This provides a major challenge for software quality assurance: programs successfully verified to be correct might be erroneous on approximate hardware. In this letter, we present a novel approach for determining under what conditions a software verification result is valid for approximate hardware. To this end, we compute the allowed tolerances for AC hardware from successful verification runs. More precisely, we derive a set of constraints which—when met by the AC hardware—guarantees the verification result to carry over to AC. On the practical side, we furthermore: 1) show how to extract tolerances from verification runs employing predicate abstraction as verification technology and 2) show how to check such constraints on hardware designs. We have implemented all techniques, and exemplify them on example C programs and a number of recently proposed approximate adders.

PART\(_\mathrm {PW}\): From Partial Analysis Results to a Proof Witness

Today, verification tools do not only output yes or no, but also provide correctness arguments or counterexamples. While counterexamples help to fix bugs, correctness arguments are used to increase the trust in program correctness, e.g., in Proof-Carrying Code (PCC). Correctness arguments are well-studied for single analyses, but not when a set of analyses together verifies a program, each of the analyses checking only a particular part. Such a set of partial, complementary analyses is often used when a single analysis would fail or is inefficient on some program parts.

Predicting rankings of software verification tools

Today, software verification tools have reached the maturity to be used for large scale programs. Different tools perform differently well on varying code. A software developer is hence faced with the problem of choosing a tool appropriate for her program at hand. A ranking of tools on programs could facilitate the choice. Such rankings can, however, so far only be obtained by running all considered tools on the program. In this paper, we present a machine learning approach to predicting rankings of tools on programs. The method builds upon so-called label ranking algorithms, which we complement with appropriate kernels providing a similarity measure for programs. Our kernels employ a graph representation for software source code that mixes elements of control flow and program dependence graphs with abstract syntax trees. Using data sets from the software verification competition SV-COMP, we demonstrate our rank prediction technique to generalize well and achieve a rather high predictive accuracy (rank correlation > 0.6).

Programs from Proofs: A Framework for the Safe Execution of Untrusted Software

Today, software is traded worldwide on global markets, with apps being downloaded to smartphones within minutes or seconds. This poses, more than ever, the challenge of ensuring safety of software in the face of (1) unknown or untrusted software providers together with (2) resource-limited software consumers. The concept of Proof-Carrying Code (PCC), years ago suggested by Necula, provides one framework for securing the execution of untrusted code. PCC techniques attach safety proofs, constructed by software producers, to code. Based on the assumption that checking proofs is usually much simpler than constructing proofs, software consumers should thus be able to quickly check the safety of software. However, PCC techniques often suffer from the size of certificates (i.e., the attached proofs), making PCC techniques inefficient in practice. In this article, we introduce a new framework for the safe execution of untrusted code called Programs from Proofs (PfP). The basic assumption underlying the PfP technique is the fact that the structure of programs significantly influences the complexity of checking a specific safety property. Instead of attaching proofs to program code, the PfP technique transforms the program into an efficiently checkable form, thus guaranteeing quick safety checks for software consumers. For this transformation, the technique also uses a producer-side automatic proof of safety. More specifically, safety proving for the software producer proceeds via the construction of an abstract reachability graph (ARG) unfolding the control-flow automaton (CFA) up to the degree necessary for simple checking. To this end, we combine different sorts of software analysis: expensive analyses incrementally determining the degree of unfolding, and cheap analyses responsible for safety checking. Out of the abstract reachability graph we generate the new program. In its CFA structure, it is isomorphic to the graph and hence another, this time consumer-side, cheap analysis can quickly determine its safety. Like PCC, Programs from Proofs is a general framework instantiable with different sorts of (expensive and cheap) analysis. Here, we present the general framework and exemplify it by some concrete examples. We have implemented different instantiations on top of the configurable program analysis tool CPAchecker and report on experiments, in particular on comparisons with PCC techniques.