Abstract Sound static analyses are an important ingredient for compiler optimizations and program verification tools. However, mathematically proving that a static analysis is sound is a difficult task due to two problems. First, soundness proofs relate two complicated program semantics (the static and the dynamic semantics) which are hard to reason about. Second, the more the static and dynamic semantics differ, the more work a soundness proof needs to do to bridge the impedance mismatch. These problems increase the effort and complexity of soundness proofs. Existing soundness theories address these problems by deriving both the dynamic and static semantics from the same artifact, often called generic interpreter. A generic interpreter provides a common structure along which a soundness proof can be composed, which avoids having to reason about the analysis as a whole. However, a generic interpreter restricts which analyses can be derived, as all derived analyses must roughly follow the program execution order. To lift this restriction, we develop a soundness theory for the blackboard analysis architecture, which is capable of describing backward, demand-driven, and summary-based analyses. The architecture describes static analyses with small independent modules, which communicate via a central store. Soundness of a compound analysis follows from soundness of all of its modules. Furthermore, modules can be proven sound independently, even though modules depend on each other. We evaluate our theory by proving soundness of four analyses: a pointer and call-graph analysis, a reflection analysis, an immutability analysis, and a demand-driven reaching definitions analysis.

Static call graphs are a fundamental building block of program analysis. However, differences in call-graph construction and the use of specific language features can yield unsoundness and imprecision. Call-graph analyses are evaluated using measures of precision and recall, but this is hard when a ground truth for real-world programs is generally unobtainable. In this work, we propose to use carefully constructed dynamic baselines based on fixed entry points and input corpora. The creation of this dynamic baseline is posed as an approximation of the ground truth---an optimization problem. We use manual extension and coverage-guided fuzzing for creating suitable input corpora. With these dynamic baselines, we study call-graph quality of multiple algorithms and implementations using four real-world Java programs. We find that our methodology provides valuable insights into call-graph quality and how to measure it. With this work, we provide a novel methodology to advance the field of static program analysis as we assess the computation of one of its core data structures---the call graph.

Traditional call-graph construction algorithms conflate the computation of possible runtime types with the actual resolution of (virtual) calls. This tangled design impedes supporting complex language features and APIs and making systematic trade-offs between precision, soundness, and scalability. It also impedes implementation of precise downstream analyses that rely on type information. To address the problem, we propose Unimocg, a modular architecture for call-graph construction that decouples the computation of type information from resolving calls. Due to its modular design, Unimocg can combine a wide range of different call-graph algorithms with algorithm-agnostic modules to support individual language features. Moreover, these modules operate at the same precision as the chosen call-graph algorithm with no further effort. Additionally, Unimocg allows other analyses to easily reuse type information from the call-graph construction at full precision. We demonstrate how Unimocg enables a framework of call-graph algorithms with different precision, soundness, and scalability trade-offs from reusable modules. Unimocg currently supports ten call-graph algorithms from vastly different families, such as CHA, RTA, XTA, and k-l-CFA. These algorithms show consistent soundness without sacrificing precision or performance. We also show how an immutability analysis is improved using Unimocg.

Modern software is often implemented in multiple interacting programming languages. When performing static analysis of such software, it is desirable to reuse existing single-language analyses to allow access to the results of decades of implementation effort.

Reasoning about immutability is important for pre-venting bugs, e.g., in multi-threaded software. So far, static analysis to infer immutability properties has mostly focused on individual objects and references. Reasoning about fields and entire classes, while significantly simpler, has gained less attention. Even a consistently used terminology is missing, which makes it difficult to implement analyses that rely on immutability information. We propose a model for class and field immutability that unifies terminology for immutability flavors considered by previous work and covers new levels of immutability to handle lazy initialization and immutability dependent on generic type parameters. Using the OPAL static analysis framework, we implement CiFi, a set of modular, collaborating analyses for different flavors of immutability, inferring the properties defined in our model. Additionally, we propose a benchmark of representative test cases for class and field immutability. We use the benchmark to showcase CiFi’s precision and recall in comparison to state of the art and use CiFi to study the prevalence of immutability in real-world libraries, showcasing the practical quality and relevance of our model.

Parallelization of static analyses is necessary to scale to real-world programs, but it is a complex and difficult task and, therefore, often only done manually for selected high-profile analyses. In this paper, we propose a programming model for semi-implicit parallelization of static analyses which is inspired by reactive programming. Reusing the domain-expert knowledge on how to parallelize anal- yses encoded in the programming framework, developers do not need to think about parallelization and concurrency issues on their own. The programming model supports stateful computations, only requires monotonic computations over lattices, and is independent of specific analyses. Our evaluation shows the applicability of the programming model to different analyses and the importance of user-selected scheduling strategies. We implemented an IFDS solver that was able to outperform a state-of-the-art, specialized parallel IFDS solver both in absolute performance and scalability.

Current approaches combining multiple static analyses deriving different, independent properties focus either on modularity or performance. Whereas declarative approaches facilitate modularity and automated, analysis-independent optimizations, imperative approaches foster manual, analysis-specific optimizations.

Call graphs are widely used; in particular for advanced control- and data-flow analyses. Even though many call graph algorithms with different precision and scalability properties have been proposed, a comprehensive understanding of sources of unsoundness, their relevance, and the capabilities of existing call graph algorithms in this respect is missing. To address this problem, we propose Judge, a toolchain that helps with understanding sources of unsoundness and improving the soundness of call graphs. In several experiments, we use Judge and an extensive test suite related to sources of unsoundness to (a) compute capability profiles for call graph implementations of Soot, WALA, DOOP, and OPAL, (b) to determine the prevalence of language features and APIs that affect soundness in modern Java Bytecode, (c) to compare the call graphs of Soot, WALA, DOOP, and OPAL – highlighting important differences in their implementations, and (d) to evaluate the necessary effort to achieve project-specific reasonable sound call graphs. We show that soundness-relevant features/APIs are frequently used and that support for them differs vastly, up to the point where comparing call graphs computed by the same base algorithms (e.g., RTA) but different frameworks is bogus. We also show that Judge can support users in establishing the soundness of call graphs with reasonable effort.

Analyzing methods in object-oriented programs whether they are side-effect free and also deterministic, i.e., mathematically pure, has been the target of extensive research. Identifying such methods helps to find code smells and security related issues, and also helps analyses detecting concurrency bugs. Pure methods are also used by formal verification approaches as the foundations for specifications and proving the pureness is necessary to ensure correct specifications. However, so far no common terminology exists which describes the purity of methods. Furthermore, some terms (e.g., pure or side-effect free) are also used inconsistently. Further, all current approaches only report selected purity information making them only suitable for a smaller subset of the potential use cases. In this paper, we present a fine-grained unified lattice model which puts the purity levels found in the literature into relation and which adds a new level that generalizes existing definitions. We have also implemented a scalable, modularized purity analysis which produces significantly more precise results for real-world programs than the best-performing related work. The analysis shows that all defined levels are found in real-world projects.