Vectorization is a powerful optimization technique that significantly boosts the performance of high performance computing applications operating on large data arrays. Despite decades of research on auto-vectorization, compilers frequently miss opportunities to vectorize code. On the other hand, writing vectorized code manually using compiler intrinsics is still a complex, error-prone task that demands deep knowledge of specific architecture and compilers. In this paper, we evaluate the potential of large-language models (LLMs) to generate vectorized (Single Instruction Multiple Data) code from scalar programs that process individual array elements. We propose a novel finite-state-machine multi-agents based approach that harnesses LLMs and test-based feedback to generate vectorized code. Our findings indicate that LLMs are capable of producing high-performance vectorized code with run-time speedup ranging from 1.1x to 9.4x as compared to the state-of-the-art compilers such as Intel Compiler, GCC, and Clang. To verify the correctness of vectorized code, we use Alive2, a leading bounded translation validation tool for LLVM IR. We describe a few domain-specific techniques to improve the scalability of Alive2 on our benchmark dataset. Overall, our approach is able to verify 38.2% of vectorizations as correct on the TSVC benchmark dataset.

Ensuring correctness is crucial for code generation. Formal verification offers a definitive assurance of correctness, but demands substantial human effort in proof construction and hence raises a pressing need for automation. The primary obsta- cle lies in the severe lack of data—there is much fewer proofs than code snippets for Large Language Models (LLMs) to train upon. In this paper, we introduce SAFE, a framework that overcomes the lack of human-written proofs to enable automated proof generation of Rust code. SAFE establishes a self-evolving cycle where data synthesis and fine-tuning collaborate to enhance the model capability, leveraging the definitive power of a symbolic verifier in telling correct proofs from incorrect ones. SAFE also re-purposes the large number of synthesized incorrect proofs to train the self-debugging capability of the fine-tuned models, empowering them to fix incorrect proofs based on the verifier’s feedback. SAFE demonstrates superior efficiency and precision compared to GPT-4o. Through tens of thousands of synthesized proofs and the self-debugging mechanism, we improve the capa- bility of open-source models, initially unacquainted with formal verification, to automatically write proofs for Rust code. This advancement leads to a signifi- cant improvement in performance, achieving a 52.52% accuracy rate in a bench- mark crafted by human experts, a significant leap over GPT-4o’s performance of 14.39%.

Proof-oriented programs mix computational content with proofs of program correctness. However, the human effort involved in programming and proving is still substantial, despite the use of Satisfiability Modulo Theories (SMT) solvers to automate proofs in languages such as F*. Seeking to spur research on using AI to automate the construction of proof-oriented programs, we curate a dataset of 600K lines of open-source F* programs and proofs, including software used in production systems ranging from Windows and Linux, to Python and Firefox. Our dataset includes around 32K top-level F* definitions, each representing a type-directed program and proof synthesis problem-producing a definition given a formal specification expressed as an F* type. We provide a program-fragment checker that queries F* to check the correctness of candidate solutions. We believe this is the largest corpus of SMT-assisted program proofs coupled with a reproducible program-fragment checker. Grounded in this dataset, we investigate the use of AI to synthesize programs and their proofs in F*, with promising results. Our main finding in that the performance of fine-tuned smaller language models (such as Phi-2 or StarCoder) compare favorably with large language models (such as GPT-4), at a much lower computational cost. We also identify various type-based retrieval augmentation techniques and find that they boost performance significantly. With detailed error analysis and case studies, we identify potential strengths and weaknesses of models and techniques and suggest directions for future improvements.

Improper parsing of attacker-controlled input is a leading source of software security vulnerabilities, especially when programmers transcribe informal format descriptions in RFCs into efficient parsing logic in low-level, memory unsafe languages. Several researchers have proposed formal specification languages for data formats from which efficient code can be extracted. However, distilling informal requirements into formal specifications is challenging and, despite their benefits, new, formal languages are hard for people to learn and use. In this work, we present 3DGen, a framework that makes use of AI agents to transform mixed informal input, including natural language documents (i.e., RFCs) and example inputs into format specifications in a language called 3D. To support humans in understanding and trusting the generated specifications, 3DGen uses symbolic methods to also synthesize test inputs that can be validated against an external oracle. Symbolic test generation also helps in distinguishing multiple plausible solutions. Through a process of repeated refinement, 3DGen produces a 3D specification that conforms to a test suite, and which yields safe, efficient, provably correct, parsing code in C. We have evaluated 3DGen on 20 Internet standard formats, demonstrating the potential for AI-agents to produce formally verified$C$code at a non-trivial scale. A key enabler is the use of a domain-specific language to limit AI outputs to a class for which automated. symbolic analysis is tractable.

User-defined functions (UDFs) extend the capabilities of SQL by improving code reusability and encapsulating complex logic, but can hinder the performance due to optimization and execution inefficiencies. Prior approaches attempt to address this by rewriting UDFs into native SQL, which is then inlined into the SQL queries that invoke them. However, these approaches are either limited to simple pattern matching or require the synthesis of complex verification conditions from procedural code, a process that is brittle and difficult to automate. This limits coverage and makes the translation approaches less extensible to previously unseen procedural constructs. In this work, we present QURE, a framework that (1) leverages large language models (LLMs) to translate UDFs to native SQL, and (2) introduces a novel formal verification method to establish equivalence between the UDF and its translation. QURE uses the semantics of SQL operators to automate the derivation of verification conditions, in turn resulting in broad coverage and high extensibility. We model a large set of imperative constructs, particularly those common in Python and Pandas UDFs, in an intermediate verification language, allowing for the verification of their SQL translation. In our empirical evaluation of Python and Pandas UDFs, equivalence is successfully verified for 88% of UDF-SQL pairs (the rest lack semantically-equivalent SQLs) and LLMs correctly translate 84% of the UDFs. Executing the translated UDFs achieves median performance improvements of 23x on single-node clusters and 12x on 12-node clusters compared to the original UDFs, while also significantly reducing out-of-memory errors.

Generative AI has shown its value for many software engineering tasks. Still in its infancy, large language model (LLM)-based proof generation lags behind LLM-based code generation. In this paper, we present A uto V erus . A uto V erus uses LLMs to automatically generate correctness proof for Rust code. A uto V erus is designed to match the unique features of Verus, a verification tool that can prove the correctness of Rust code using proofs and specifications also written in Rust. A uto V erus consists of a network of agents that are crafted and orchestrated to mimic human experts’ three phases of proof construction: preliminary proof generation, proof refinement guided by generic tips, and proof debugging guided by verification errors. To thoroughly evaluate A uto V erus and help foster future research in this direction, we have built a benchmark suite of 150 non-trivial proof tasks, based on existing code-generation benchmarks and verification benchmarks. Our evaluation shows that A uto V erus can automatically generate correct proof for more than 90% of them, with more than half of them tackled in less than 30 seconds or 3 LLM calls.

Verification-aware programming languages such as Dafny and F* provide means to formally specify and prove properties of a program. Although the problem of checking an implementation against a specification can be defined mechanically, there is no algorithmic way of ensuring the correctness of the {\it user-intent formalization for programs}, expressed as a formal specification. This is because intent or requirement is expressed {\it informally} in natural language and the specification is a formal artefact. Despite, the advent of large language models (LLMs) has made tremendous strides bridging the gap between informal intent and formal program implementations recently, driven in large parts by benchmarks and automated metrics for evaluation. Recent work has proposed a framework for evaluating the {\it user-intent formalization} problem for mainstream programming languages~\cite{endres-fse24}. However, such an approach does not readily extend to verification-aware languages that support rich specifications (using quantifiers and ghost variables) that cannot be evaluated through dynamic execution. Previous work also required generating program mutants using LLMs to create the benchmark. We advocate an alternate, perhaps simpler approach of {\it symbolically testing specifications} to provide an intuitive metric for evaluating the quality of specifications for verification-aware languages. We demonstrate that our automated metric agrees closely on a human-labeled dataset of Dafny specifications for the popular MBPP code-generation benchmark, yet demonstrates cases where the human labeling is not perfect. We also outline formal verification challenges that need to be addressed to apply the technique more widely. We believe our work provides a stepping stone to enable the establishment of a benchmark and research agenda for the problem of user-intent formalization for programs.

Large language models (LLMs) have shown great potential in automating significant aspects of coding by producing natural code from informal natural language (NL) intent. However, given NL is informal, it does not lend easily to checking that the generated code correctly satisfies the user intent. In this paper, we propose a novel interactive workflowTiCoderfor guided intent clarification (i.e., partial formalization) through tests to support the generation of more accurate code suggestions. Through a mixed methods user study with 15 programmers, we present an empirical evaluation of the effectiveness of the workflow to improve code generation accuracy. We find that participants using the proposed workflow are significantly more likely to correctly evaluate AI generated code, and report significantly less task-induced cognitive load. Furthermore, we test the potential of the workflow at scale with four different state-of-the-art LLMs on two python datasets, using an idealized proxy for a user feedback. We observe an average absolute improvement of 45.97% in the pass@1 code generation accuracy for both datasets and across all LLMs within 5 user interactions, in addition to the automatic generation of accompanying unit tests.

Search-based software testing (SBST) generates high-coverage test cases for programs under test with a combination of test case generation and mutation. SBST's performance relies on there being a reasonable probability of generating test cases that exercise the core logic of the program under test. Given such test cases, SBST can then explore the space around them to exercise various parts of the program. This paper explores whether Large Language Models (LLMs) of code, such as OpenAI's Codex, can be used to help SBST's exploration. Our proposed algorithm, CodaMosa, conducts SBST until its coverage improvements stall, then asks Codex to provide example test cases for under-covered functions. These examples help SBST redirect its search to more useful areas of the search space. On an evaluation over 486 benchmarks, CodaMosa achieves statistically significantly higher coverage on many more benchmarks (173 and 279) than it reduces coverage on (10 and 4), compared to SBST and LLM-only baselines.

Language models of code (LMs) work well when the surrounding code provides sufficient context. This is not true when it becomes necessary to use types, functionality or APIs defined elsewhere in the repository or a linked library, especially those not seen during training. LMs suffer from limited awareness of such global context and end up hallucinating. Integrated development environments (IDEs) assist developers in understanding repository context using static analysis. We extend this assistance, enjoyed by developers, to LMs. We propose monitor-guided decoding (MGD) where a monitor uses static analysis to guide the decoding. We construct a repository-level dataset PragmaticCode for method-completion in Java and evaluate MGD on it. On models of varying parameter scale, by monitoring for type-consistent object dereferences, MGD consistently improves compilation rates and agreement with ground truth. Further, LMs with fewer parameters, when augmented with MGD, can outperform larger LMs. With MGD, SantaCoder-1.1B achieves better compilation rate and next-identifier match than the much larger text-davinci-003 model. We also conduct a generalizability study to evaluate the ability of MGD to generalize to multiple programming languages (Java, C# and Rust), coding scenarios (e.g., correct number of arguments to method calls), and to enforce richer semantic constraints (e.g., stateful API protocols). Our data and implementation are available at https://github.com/microsoft/monitors4codegen.

A resource leak occurs when a program fails to free some finite resource after it is no longer needed. Such leaks are a significant cause of real-world crashes and performance problems. Recent work proposed an approach to prevent resource leaks based on checking resource management specifications. A resource management specification expresses how the program allocates resources, passes them around, and releases them; it also tracks the ownership relationship between objects and resources, and aliasing relationships between objects. While this specify-and-verify approach has several advantages compared to prior techniques, the need to manually write annotations presents a significant barrier to its practical adoption. This paper presents a novel technique to automatically infer a resource management specification for a program, broadening the applicability of specify-and-check verification for resource leaks. Inference in this domain is challenging because resource management specifications differ significantly in nature from the types that most inference techniques target. Further, for practical effectiveness, we desire a technique that can infer the resource management specification intended by the developer, even in cases when the code does not fully adhere to that specification. We address these challenges through a set of inference rules carefully designed to capture real-world coding patterns, yielding an effective fixed-point-based inference algorithm. We have implemented our inference algorithm in two different systems, targeting programs written in Java and C#. In an experimental evaluation, our technique inferred 85.5% of the annotations that programmers had written manually for the benchmarks. Further, the verifier issued nearly the same rate of false alarms with the manually-written and automatically-inferred annotations.

In collaborative software development, program merging isthemechanism to integrate changes from multiple programmers. Merge algorithms in modern version control systems report a conflict when changes interfere textually. Merge conflicts require manual intervention and frequently stall modern continuous integration pipelines. Prior work found that, although costly, a large majority of resolutions involve re-arranging text without writing any new code. Inspired by this observation we propose thefirst data-driven approachto resolve merge conflicts with a machine learning model. We realize our approach in a toolDeepMergethat uses a novel combination of (i) an edit-aware embedding of merge inputs and (ii) a variation of pointer networks, to construct resolutions from input segments. We also propose an algorithm to localize manual resolutions in a resolved file and employ it to curate a ground-truth dataset comprising 8,719 non-trivial resolutions in JavaScript programs. Our evaluation shows that, on a held out test set,DeepMergecan predict correct resolutions for 37% of non-trivial merges, compared to only 4% by a state-of-the-art semistructured merge technique. Furthermore, on the subset of merges with upto 3 lines (comprising 24% of the total dataset),DeepMergecan predict correct resolutions with 78% accuracy.

Testing is widely recognized as an important stage of the software development lifecycle. Effective software testing can provide benefits such as bug finding, preventing regressions, and documentation. In terms of documentation, unit tests express a unit's intended functionality, as conceived by the developer. A test oracle, typically expressed as an condition, documents the intended behavior of a unit under a given test prefix. Synthesizing a functional test oracle is a challenging problem, as it must capture the intended functionality rather than the implemented functionality.

Program merging is standard practice when developers integrate their individual changes to a common code base. When the merge algorithm fails, this is called a merge conflict. The conflict either manifests as a textual merge conflict where the merge fails to produce code, or as a semantic merge conflict where the merged code results in compiler errors or broken tests. Resolving these conflicts for large code projects is expensive because it requires developers to manually identify the sources of conflicts and correct them. In this paper, we explore the feasibility of automatically repairing merge conflicts (both textual and semantic) using k-shot learning with pre-trained large neural language models (LM) such as GPT-3. One of the challenges in leveraging such language models is fitting the examples and the queries within a small prompt (2048 tokens). We evaluate LMs and k-shot learning for both textual and semantic merge conflicts for Microsoft Edge. Our results are mixed: on one-hand, LMs provide the state-of-the-art (SOTA) performance on semantic merge conflict resolution for Edge compared to earlier symbolic approaches; on the other hand, LMs do not yet obviate the benefits of special purpose domain-specific languages (DSL) for restricted patterns for program synthesis.

Large language models (LLMs) have demonstrated an impressive ability to generate code for various programming tasks. In many instances, LLMs can generate a correct program for a task when given numerous trials. Consequently, a recent trend is to do large scale sampling of programs using a model and then filtering/ranking the programs based on the program execution on a small number of known unit tests to select one candidate solution. However, these approaches assume that the unit tests are given and assume the ability to safely execute the generated programs (which can do arbitrary dangerous operations such as file manipulations). Both of the above assumptions are impractical in real-world software development. In this paper, we propose CodeRanker, a neural ranker that can predict the correctness of a sampled program without executing it. Our CodeRanker is fault-aware i.e., it is trained to predict different kinds of execution information such as predicting the exact compile/runtime error type (e.g., an IndexError or a TypeError). We show that CodeRanker can significantly increase the pass@1 accuracy of various code generation models (including Codex, GPT-Neo, GPT-J) on APPS, HumanEval and MBPP datasets.

Collaborative software development is an integral part of the modern software development life cycle, essential to the success of large-scale software projects. When multiple developers make concurrent changes around the same lines of code, a merge conflict may occur. Such conflicts stall pull requests and continuous integration pipelines for hours to several days, seriously hurting developer productivity. To address this problem, we introduce MergeBERT, a novel neural program merge framework based on token-level three-way differencing and a transformer encoder model. By exploiting the restricted nature of merge conflict resolutions, we reformulate the task of generating the resolution sequence as a classification task over a set of primitive merge patterns extracted from real-world merge commit data. Our model achieves 63–68% accuracy for merge resolution synthesis, yielding nearly a 3× performance improvement over existing semi-structured, and 2× improvement over neural program merge tools. Finally, we demonstrate that MergeBERT is sufficiently flexible to work with source code files in Java, JavaScript, TypeScript, and C# programming languages. To measure the practical use of MergeBERT, we conduct a user study to evaluate MergeBERT suggestions with 25 developers from large OSS projects on 122 real-world conflicts they encountered. Results suggest that in practice, MergeBERT resolutions would be accepted at a higher rate than estimated by automatic metrics for precision and accuracy. Additionally, we use participant feedback to identify future avenues for improvement of MergeBERT.

As smart contracts gain adoption in financial transactions, it becomes increasingly important to ensure that they are free of bugs and security vulnerabilities. Of particular relevance in this context are arithmetic overflow bugs, as integers are often used to represent financial assets like account balances. Motivated by this observation, this paper presents SolType, a refinement type system for Solidity that can be used to prevent arithmetic over- and under-flows in smart contracts. SolType allows developers to add refinement type annotations and uses them to prove that arithmetic operations do not lead to over- and under-flows. SolType incorporates a rich vocabulary of refinement terms that allow expressing relationships between integer values and aggregate properties of complex data structures. Furthermore, our implementation, called Solid, incorporates a type inference engine and can automatically infer useful type annotations, including non-trivial contract invariants. To evaluate the usefulness of our type system, we use Solid to prove arithmetic safety of a total of 120 smart contracts. When used in its fully automated mode (i.e., using Solid's type inference capabilities), Solid is able to eliminate 86.3% of redundant runtime checks used to guard against overflows. We also compare Solid against a state-of-the-art arithmetic safety verifier called VeriSmart and show that Solid has a significantly lower false positive rate, while being significantly faster in terms of verification time.

Forking structure is widespread in the open-source repositories and that causes a significant number of merge conflicts. In this paper, we study the problem of textual merge conflicts from the perspective of Microsoft Edge, a large, highly collaborative fork off the main Chromium branch with significant merge conflicts. Broadly, this study is divided into two sections. First, we empirically evaluate textual merge conflicts in Microsoft Edge and classify them based on the type of files, location of conflicts in a file, and the size of conflicts. We found that ~28% of the merge conflicts are 1-2 line changes, and many resolutions have frequent patterns. Second, driven by these findings, we explore Program Synthesis (for the first time) to learn patterns and resolve structural merge conflicts. We propose a novel domain-specific language (DSL) that captures many of the repetitive merge conflict resolution patterns and learn resolution strategies as programs in this DSL from example resolutions. We found that the learned strategies can resolve 11.4% of the conflicts (~41% of 1-2 line changes) that arise in the C++ files with 93.2% accuracy.

Smart contracts are programs that run on the blockchain and digitally enforce the execution of contracts between parties. Because bugs in smart contracts can have serious monetary consequences, ensuring the correctness of such software is of utmost importance. In this paper, we present a novel technique, and its implementation in a tool called SMARTPULSE, for automatically verifying temporal properties in smart contracts. SMARTPULSE is the first smart contract verification tool that is capable of checking liveness properties, which ensure that "something good" will eventually happen (e.g., "I will eventually receive my refund"). We experimentally evaluate SMARTPULSE on a broad class of smart contracts and properties and show that (a) SMARTPULSE allows automatically verifying important liveness properties, (b) it is competitive with or better than state-of-the-art tools for safety verification, and (c) it can automatically generate attacks for vulnerable contracts.

Microsoft's Static Driver Verifier (SDV) pioneered the use of software model checking for ensuring that device drivers correctly use operating system (OS) APIs.However, the verification methodology has been difficult to extend in order to support either (a) new classes of drivers for which SDV does not already have a harness and stubs, or (b) memory-corruption properties.Any attempt to apply SDV out-of-the-box results in either false alarms due to the lack of environment modeling, or scalability issues when finding deeply nested bugs in the presence of a very large number of memory accesses.In this paper, we describe our experience designing and shipping a new class of checks known as angelic checks through SDV with the aid of angelic verification (AV) [1] technology, over a period of 4 years.AV pairs a precise inter-procedural assertion checker with automatic inference of likely specifications for the environment.AV helps compensate for the lack of environment modeling and regains scalability by making it possible to find deeply nested bugs, even for complex memorycorruption properties.These new rules have together found over a hundred confirmed defects during internal deployment at Microsoft, including several previously unknown high-impact potential security vulnerabilities.AV considerably increases the reach of SDV, both in terms of drivers as well as rules that it can support effectively.

With the growing use of DevOps tools and frameworks, there is an increased need for tools and techniques that support more than code. The current state-of-the-art in static developer assistance for tools like Docker is limited to shallow syntactic validation. We identify three core challenges in the realm of learning from, understanding, and supporting developers writing DevOps artifacts: (i) nested languages in DevOps artifacts, (ii) rule mining, and (iii) the lack of semantic rule-based analysis. To address these challenges we introduce a toolset, binnacle, that enabled us to ingest 900,000 GitHub repositories.

This paper aims to shed light on how loops are used in smart contracts. Towards this goal, we study various syntactic and semantic characteristics of loops used in over 20,000 Solidity contracts deployed on the Ethereum blockchain, with the goal of informing future research on program analysis for smart contracts. Based on our findings, we propose a small domain-specific language (DSL) that can be used to summarize common looping patterns in Solidity. To evaluate what percentage of smart contract loops can be expressed in our proposed DSL, we also design and implement a program synthesis toolchain called Solis that can synthesize loop summaries in our DSL. Our evaluation shows that at least 56% of the analyzed loops can be summarized in our DSL, and 81% of these summaries are exactly equivalent to the original loop.

Dockerfiles are one of the most prevalent kinds of DevOps artifacts used in industry. Despite their prevalence, there is a lack of sophisticated semantics-aware static analysis of Dockerfiles. In this paper, we introduce a dataset of approximately 178,000 unique Dockerfiles collected from GitHub. To enhance the usability of this data, we describe five representations we have devised for working with, mining from, and analyzing these Dockerfiles. Each Dockerfile representation builds upon the previous ones, and the final representation, created by three levels of nested parsing and abstraction, makes tasks such as mining and static checking tractable. The Dockerfiles, in each of the five representations, along with metadata and the tools used to shepard the data from one representation to the next are all available at: https://doi.org/10.5281/zenodo.3628771.

We present an approach for comparing two closely related concurrent programs, whose goal is to give feedback about interesting differences without relying on user-provided assertions. This approach compares two programs in terms of cross-thread interferences and data-flow, under a parametrized abstraction which can detect any difference in the limit. We introduce a partial order relation between these abstractions such that a program change that leads to a “smaller” abstraction is more likely to be regression-free from the perspective of concurrency. On the other hand, incomparable or bigger abstractions, which are an indication of introducing new, possibly undesired, behaviors, lead to succinct explanations of the semantic differences.

When an evolving program is modified to address issues related to thread synchronization, there is a need to confirm the change is correct, i.e., it does not introduce unexpected behavior. However, manually comparing two programs to identify the semantic difference is labor intensive and error prone, whereas techniques based on model checking are computationally expensive. To fill the gap, we develop a fast and approximate static analysis for computing synchronization differences of two programs. The method is fast because, instead of relying on heavy-weight model checking techniques, it leverages a polynomial-time Datalog-based program analysis framework to compute differentiating data-flow edges, i.e., edges allowed by one program but not the other. Although approximation is used our method is sufficiently accurate due to careful design of the Datalog inference rules and iterative increase of the required data-flow edges for representing a difference. We have implemented our method and evaluated it on a large number of multithreaded C programs to confirm its ability to produce, often within seconds, the same differences obtained by human; in contrast, prior techniques based on model checking take minutes or even hours and thus can be 10x to 1000x slower.

With the rise of machine learning, there is a great deal of interest in treating programs as data to be fed to learning algorithms. However, programs do not start off in a form that is immediately amenable to most off-the-shelf learning techniques. Instead, it is necessary to transform the program to a suitable representation before a learning technique can be applied.

This paper addresses the problem of verifying equivalence between a pair of programs that operate over databases with different schemas. This problem is particularly important in the context of web applications, which typically undergo database refactoring either for performance or maintainability reasons. While web applications should have the same externally observable behavior before and after schema migration, there are no existing tools for proving equivalence of such programs. This paper takes a first step towards solving this problem by formalizing the equivalence and refinement checking problems for database-driven applications. We also propose a proof methodology based on the notion of bisimulation invariants over relational algebra with updates and describe a technique for synthesizing such bisimulation invariants. We have implemented the proposed technique in a tool called Mediator for verifying equivalence between database-driven applications written in our intermediate language and evaluate our tool on 21 benchmarks extracted from textbooks and real-world web applications. Our results show that the proposed methodology can successfully verify 20 of these benchmarks.

Even though many programmers rely on 3-way merge tools to integrate changes from different branches, such tools can introduce subtle bugs in the integration process. This paper aims to mitigate this problem by defining a semantic notion of conflict-freedom, which ensures that the merged program does not introduce new unwanted behaviors. We also show how to verify this property using a novel, compositional algorithm that combines lightweight summarization for shared program fragments with precise relational reasoning for the modifications. Towards this goal, our method uses a 4-way differencing algorithm on abstract syntax trees to represent different program versions as edits applied to a shared program with holes. This representation allows our verification algorithm to reason about different edits in isolation and compose them to obtain an overall proof of conflict freedom. We have implemented the proposed technique in a new tool called SafeMerge for Java and evaluate it on 52 real-world merge scenarios obtained from Github. The experimental results demonstrate the benefits of our approach over syntactic conflict-freedom and indicate that SafeMerge is both precise and practical.

Modern build systems help increase developer productivity by performing incremental building and testing. These build systems view a software project as a group of interdependent modules and perform regression test selection at the module level. However, many large software projects have imprecise dependency graphs that lead to wasteful test executions. If a test belongs to a module that has more dependencies than the actual dependencies of the test, then it is executed unnecessarily whenever a code change impacts those additional dependencies. In this paper, we formulate the problem of wasteful test executions due to suboptimal placement of tests in modules. We propose a greedy algorithm to reduce the number of test executions by suggesting test movements while considering historical build information and actual dependencies of tests. We have implemented our technique, called TestOptimizer, on top of CloudBuild, the build system developed within Microsoft over the last few years. We have evaluated the technique on five large proprietary projects. Our results show that the suggested test movements can lead to a reduction of 21.66 million test executions (17.09%) across all our subject projects. We received encouraging feedback from the developers of these projects; they accepted and intend to implement ≈80% of our reported suggestions.

Change-impact analysis (CIA) is the task of determining the set of program elements impacted by a program change. Precise CIA has great potential to avoid expensive testing and code reviews for (parts of) changes that are refactorings (semantics-preserving). However most statement-level CIA techniques suffer from imprecision as they do not incorporate the semantics of the change.

Code Reviews, an important and popular mechanism for quality assurance, are often performed on a change set, a set of modified files that are meant to be committed to a source repository as an atomic action. Understanding a code review is more difficult when the change set consists of multiple, independent, code differences. We introduce CLUSTERCHANGES, an automatic technique for decomposing change sets and evaluate its effectiveness through both a quantitative analysis and a qualitative user study.

We introduce Verification Modulo Versions (VMV), a new static analysis technique for reducing the number of alarms reported by static verifiers while providing sound semantic guarantees. First, VMV extracts semantic environment conditions from a base program P. Environmental conditions can either be sufficient conditions (implying the safety of P) or necessary conditions (implied by the safety of P). Then, VMV instruments a new version of the program, P', with the inferred conditions. We prove that we can use (i) sufficient conditions to identify abstract regressions of P' w.r.t. P; and (ii) necessary conditions to prove the relative correctness of P' w.r.t. P. We show that the extraction of environmental conditions can be performed at a hierarchy of abstraction levels (history, state, or call conditions) with each subsequent level requiring a less sophisticated matching of the syntactic changes between P' and P. Call conditions are particularly useful because they only require the syntactic matching of entry points and callee names across program versions. We have implemented VMV in a widely used static analysis and verification tool. We report our experience on two large code bases and demonstrate a substantial reduction in alarms while additionally providing relative correctness guarantees.

Modular assertion checkers are plagued with false alarms due to the need for precise environment specifications (preconditions and callee postconditions). Even the fully precise checkers report assertion failures under the most demonic environments allowed by unconstrained or partial specifications. The inability to preclude overly adversarial environments makes such checkers less attractive to developers and severely limits the adoption of such tools in the development cycle.

This paper describes a cross-version compiler validator and measures its effectiveness on the CLR JIT compiler. The validator checks for semantically equivalent assembly language output from various versions of the compiler, including versions across a seven-month time period, across two architectures (x86 and ARM), across two compilation scenarios (JIT and MDIL), and across optimizations levels. For month-to-month comparisons, the validator achieves a false alarm rate of just 2.2%. To help understand reported semantic differences, the validator performs a root-cause analysis on the counterexample traces generated by the underlying automated theorem proving tools. This root-cause analysis groups most of the counterexamples into a small number of buckets, reducing the number of counterexamples analyzed by hand by anywhere from 53% to 96%. The validator ran on over 500,000 methods across a large suite of test programs, finding 12 previously unknown correctness and performance bugs in the CLR compiler.

Previous version of a program can be a powerful enabler for program analysis by defining new relative specifications and making the results of current program analysis more relevant. In this paper, we describe the approach of differential assertion checking (DAC) for comparing different versions of a program with respect to a set of assertions. DAC provides a natural way to write relative specifications over two programs. We introduce a novel modular approach to DAC by reducing it to safety checking of a composed program, which can be accomplished by standard program verifiers. In particular, we leverage automatic invariant generation to synthesize relative specifications for pairs of loops and procedures. We provide a preliminary evaluation of a prototype implementation within the SymDiff tool along two directions (a) soundly verifying bug fixes in the presence of loops and (b) providing a knob for suppressing alarms when checking a new version of a program.

This paper describes our experience of performing reactive security audit of known security vulnerabilities in core operating system and browser COM components, using an extended static checker HAVOCLITE. We describe the extensions made to the tool to be applicable on such large C++ components, along with our experience of using an extended static checker in the large. We argue that the use of such checkers as a configurable static analysis in the hands of security auditors can be an effective tool for finding variations of known vulnerabilities. The effort has led to finding and fixing around 70 previously unknown security vulnerabilities in over 10 millions lines operating system and browser code.

Static assertion checking of open programs requires setting up a precise harness to capture the environment assumptions. For instance, a library may require a file handle to be properly initialized before it is passed into it. A harness is used to set up or specify the appropriate preconditions before invoking methods from the program. In the absence of a precise harness, even the most precise automated static checkers are bound to report numerous false alarms. This often limits the adoption of static assertion checking in the hands of a user.

We present a unified approach to type checking and property checking for low-level code. Type checking for low-level code is challenging because type safety often depends on complex, program-specific invariants that are difficult for traditional type checkers to express. Conversely, property checking for low-level code is challenging because it is difficult to write concise specifications that distinguish between locations in an untyped program's heap. We address both problems simultaneously by implementing a type checker for low-level code as part of our property checker.

Embedded software are becoming more and more pervasive in our lives, and many application domains have very high reliability requirements. Ensuring high software quality while still maintaining software productivity is a challenging task. In order to address this challenge, more formal analysis and automated verification techniques are needed in addition to standard software testing.

We present a case study in which a team of test engineers at Microsoft applied a feedback-directed random testing tool to a critical component of the .NET architecture. Due to its complexity and high reliability requirements, the component had already been tested by 40 test engineers over five years, using manual testing and many automated testing techniques.

This paper takes a fresh look at the problem of precise verification of heap-manipulating programs using first-order Satisfiability-Modulo-Theories (SMT) solvers. We augment the specification logic of such solvers by introducing the Logic of Interpreted Sets and Bounded Quantification for specifying properties of heap-manipulating programs. Our logic is expressive, closed under weakest preconditions, and efficiently implementable on top of existing SMT solvers. We have created a prototype implementation of our logic over the solvers Simplify and Z3 and used our prototype to verify many programs. Our preliminary experience is encouraging; the completeness and the efficiency of the decisionprocedure is clearly evident in practice and has greatly improved the user experience of the verifier.

We present a technique that improves random test generation by incorporating feedback obtained from executing test inputs as they are created. Our technique builds inputs incrementally by randomly selecting a method call to apply and finding arguments from among previously-constructed inputs. As soon as an input is built, it is executed and checked against a set of contracts and filters. The result of the execution determines whether the input is redundant, illegal, contract-violating, or useful for generating more inputs. The technique outputs a test suite consisting of unit tests for the classes under test. Passing tests can be used to ensure that code contracts are preserved across program changes; failing tests (that violate one or more contract) point to potential errors that should be corrected. Our experimental results indicate that feedback-directed random test generation can outperform systematic and undirected random test generation, in terms of coverage and error detection. On four small but nontrivial data structures (used previously in the literature), our technique achieves higher or equal block and predicate coverage than model checking (with and without abstraction) and undirected random generation. On 14 large, widely-used libraries (comprising 780KLOC), feedback-directed random test generation finds many previously-unknown errors, not found by either model checking or undirected random generation.

We present a new approach for performing predicate abstraction based on symbolic decision procedures. Intuitively, a symbolic decision procedure for a theory takes a set of predicates in the theory and symbolically executes a decision procedure on all the subsets over the set of predicates. The result of the symbolic decision procedure is a shared expression (represented by a directed acyclic graph) that implicitly represents the answer to a predicate abstraction query. We present symbolic decision procedures for the logic of Equality and Uninterpreted Functions (EUF) and Difference logic (DIFF) and show that these procedures run in pseudo-polynomial (rather than exponential) time. We then provide a method to construct symbolic decision procedures for simple mixed theories (including the two theories mentioned above) using an extension of the Nelson-Oppen combination method. We present preliminary evaluation of our Procedure on predicate abstraction benchmarks from device driver verification in SLAM.

Predicate abstraction provides a powerful tool for verifying properties of infinite-state systems using a combination of a decision procedure for a subset of first-order logic and symbolic methods originally developed for finite-state model checking. We consider models containing first-order state variables, where the system state includes mutable functions and predicates. Such a model can describe systems containing arbitrarily large memories, buffers, and arrays of identical processes. We describe a form of predicate abstraction that constructs a formula over a set of universally quantified variables to describe invariant properties of the first-order state variables. We provide a formal justification of the soundness of our approach and describe how it has been used to verify several hardware and software designs, including a directory-based cache coherence protocol.

We describe a novel method for verifying programs that manipulate linked lists, based on two new predicates that characterize reachability of heap cells. These predicates allow reasoning about both acyclic and cyclic lists uniformly with equal ease. The crucial insight behind our approach is that a circular list invariably contains a distinguished head cell that provides a handle on the list. This observation suggests a programming methodology that requires the heap of the program at each step to be well-founded, i.e., for any field f in the program, every sequence u.f, u.f.f,... contains at least one head cell. We believe that our methodology captures the most common idiom of programming with linked data structures. We enforce our methodology by automatically instrumenting the program with updates to two auxiliary variables representing these predicates and adding assertions in terms of these auxiliary variables.To prove program properties and the instrumented assertions, we provide a first-order axiomatization of our two predicates. We also introduce a novel induction principle made possible by the well-foundedness of the heap. We use our induction principle to derive from two basic axioms a small set of additional first-order axioms that are useful for proving the correctness of several programs.We have implemented our method in a tool and used it to verify the correctness of a variety of nontrivial programs manipulating both acyclic and cyclic singly-linked lists and doubly-linked lists. We also demonstrate the use of indexed predicate abstraction to automatically synthesize loop invariants for these examples.

SAT-based decision procedures for quantifier-free fragments of first-order logic have proved to be useful in formal verification. These decision procedures are either based on encoding atomic subformulas with Boolean variables, or by encoding integer variables as bit-vectors. Based on evaluating these two encoding methods on a diverse set of hardware and software benchmarks, we conclude that neither method is robust to variations in formula characteristics. We therefore propose a new hybrid technique that combines the two methods. We give experimental results showing that the hybrid method can significantly outperform either approach as well as other decision procedures.