arXiv subjects

Akash Lal

Publications and source records attributed to Akash Lal.

15 recordsLinked to original sources

arXiv2026

AFGNN: API Misuse Detection using Graph Neural Networks and Clustering

Application Programming Interfaces (APIs) are crucial to software development, enabling integration of existing systems with new applications by reusing tried and tested code, saving development time and increasing software safety. In particular, the Java standard library APIs, along with numerous third-party APIs, are extensively utilized in the development of enterprise application software. However, their misuse remains a significant source of bugs and vulnerabilities. Furthermore, due to the limited examples in the official API documentation, developers often rely on online portals and generative AI models to learn unfamiliar APIs, but using such examples may introduce unintentional errors in the software. In this paper, we present AFGNN, a novel Graph Neural Network (GNN)-based framework for efficiently detecting API misuses in Java code. AFGNN uses a novel API Flow Graph (AFG) representation that captures the API execution sequence, data, and control flow information present in the code to model the API usage patterns. AFGNN uses self-supervised pre-training with AFG representation to effectively compute the embeddings for unknown API usage examples and cluster them to identify different usage patterns. Experiments on popular API usage datasets show that AFGNN significantly outperforms state-of-the-art small language models and API misuse detectors.

cs.SE↗

arXiv2024

Enabling Memory Safety of C Programs using LLMs

Memory safety violations in low-level code, written in languages like C, continues to remain one of the major sources of software vulnerabilities. One method of removing such violations by construction is to port C code to a safe C dialect. Such dialects rely on programmer-supplied annotations to guarantee safety with minimal runtime overhead. This porting, however, is a manual process that imposes significant burden on the programmer and, hence, there has been limited adoption of this technique. The task of porting not only requires inferring annotations, but may also need refactoring/rewriting of the code to make it amenable to such annotations. In this paper, we use Large Language Models (LLMs) towards addressing both these concerns. We show how to harness LLM capabilities to do complex code reasoning as well as rewriting of large codebases. We also present a novel framework for whole-program transformations that leverages lightweight static analysis to break the transformation into smaller steps that can be carried out effectively by an LLM. We implement our ideas in a tool called MSA that targets the CheckedC dialect. We evaluate MSA on several micro-benchmarks, as well as real-world code ranging up to 20K lines of code. We showcase superior performance compared to a vanilla LLM baseline, as well as demonstrate improvement over a state-of-the-art symbolic (non-LLM) technique.

cs.SE↗

arXiv2023

Resource Leak Checker (RLC#) for C# Code using CodeQL

Resource leaks occur when a program fails to release a finite resource after it is no longer needed. These leaks are a significant cause of real-world crashes and performance issues. Given their critical impact on software performance and security, detecting and preventing resource leaks is a crucial problem. Recent research has proposed a specify-and-check approach to prevent resource leaks. In this approach, programmers write resource management specifications that guide how resources are stored, passed around, and released within an application. We have developed a tool called RLC#, for detecting resource leaks in C# code. Inspired by the Resource Leak Checker (RLC) from the Checker Framework, RLC# employs CodeQL for intraprocedural data flow analysis. The tool operates in a modular fashion and relies on resource management specifications integrated at method boundaries for interprocedural analysis. In practice, RLC# has successfully identified 24 resource leaks in open-source projects and internal proprietary Azure microservices. Its implementation is declarative, and it scales well. While it incurs a reasonable false positive rate, the burden on developers is minimal, involving the addition of specifications to the source code.

cs.PL↗

arXiv2023

Finding Inductive Loop Invariants using Large Language Models

Loop invariants are fundamental to reasoning about programs with loops. They establish properties about a given loop's behavior. When they additionally are inductive, they become useful for the task of formal verification that seeks to establish strong mathematical guarantees about program's runtime behavior. The inductiveness ensures that the invariants can be checked locally without consulting the entire program, thus are indispensable artifacts in a formal proof of correctness. Finding inductive loop invariants is an undecidable problem, and despite a long history of research towards practical solutions, it remains far from a solved problem. This paper investigates the capabilities of the Large Language Models (LLMs) in offering a new solution towards this old, yet important problem. To that end, we first curate a dataset of verification problems on programs with loops. Next, we design a prompt for exploiting LLMs, obtaining inductive loop invariants, that are checked for correctness using sound symbolic tools. Finally, we explore the effectiveness of using an efficient combination of a symbolic tool and an LLM on our dataset and compare it against a purely symbolic baseline. Our results demonstrate that LLMs can help improve the state-of-the-art in automated program verification.

cs.PL↗

arXiv2023

Ranking LLM-Generated Loop Invariants for Program Verification

Synthesizing inductive loop invariants is fundamental to automating program verification. In this work, we observe that Large Language Models (such as gpt-3.5 or gpt-4) are capable of synthesizing loop invariants for a class of programs in a 0-shot setting, yet require several samples to generate the correct invariants. This can lead to a large number of calls to a program verifier to establish an invariant. To address this issue, we propose a {\it re-ranking} approach for the generated results of LLMs. We have designed a ranker that can distinguish between correct inductive invariants and incorrect attempts based on the problem definition. The ranker is optimized as a contrastive ranker. Experimental results demonstrate that this re-ranking mechanism significantly improves the ranking of correct invariants among the generated candidates, leading to a notable reduction in the number of calls to a verifier. The source code and the experimental data for this paper are available in \url{https://github.com/microsoft/NeuralInvariantRanker}.

cs.PL↗

arXiv2023

Fixing Rust Compilation Errors using LLMs

The Rust programming language, with its safety guarantees, has established itself as a viable choice for low-level systems programming language over the traditional, unsafe alternatives like C/C++. These guarantees come from a strong ownership-based type system, as well as primitive support for features like closures, pattern matching, etc., that make the code more concise and amenable to reasoning. These unique Rust features also pose a steep learning curve for programmers. This paper presents a tool called RustAssistant that leverages the emergent capabilities of Large Language Models (LLMs) to automatically suggest fixes for Rust compilation errors. RustAssistant uses a careful combination of prompting techniques as well as iteration with an LLM to deliver high accuracy of fixes. RustAssistant is able to achieve an impressive peak accuracy of roughly 74% on real-world compilation errors in popular open-source Rust repositories. We plan to release our dataset of Rust compilation errors to enable further research.

cs.SE↗

arXiv2023

Inference of Resource Management Specifications

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.

cs.SE↗

arXiv2021

MonkeyDB: Effectively Testing Correctness against Weak Isolation Levels

Modern applications, such as social networking systems and e-commerce platforms are centered around using large-scale storage systems for storing and retrieving data. In the presence of concurrent accesses, these storage systems trade off isolation for performance. The weaker the isolation level, the more behaviors a storage system is allowed to exhibit and it is up to the developer to ensure that their application can tolerate those behaviors. However, these weak behaviors only occur rarely in practice, that too outside the control of the application, making it difficult for developers to test the robustness of their code against weak isolation levels. This paper presents MonkeyDB, a mock storage system for testing storage-backed applications. MonkeyDB supports a Key-Value interface as well as SQL queries under multiple isolation levels. It uses a logical specification of the isolation level to compute, on a read operation, the set of all possible return values. MonkeyDB then returns a value randomly from this set. We show that MonkeyDB provides good coverage of weak behaviors, which is complete in the limit. We test a variety of applications for assertions that fail only under weak isolation. MonkeyDB is able to break each of those assertions in a small number of attempts.

cs.PL↗

arXiv2020

Distributed Bounded Model Checking

Program verification is a resource-hungry task. This paper looks at the problem of parallelizing SMT-based automated program verification, specifically bounded model-checking, so that it can be distributed and executed on a cluster of machines. We present an algorithm that dynamically unfolds the call graph of the program and frequently splits it to create sub-tasks that can be solved in parallel. The algorithm is adaptive, controlling the splitting rate according to available resources, and also leverages information from the SMT solver to split where most complexity lies in the search. We implemented our algorithm by modifying CORRAL, the verifier used by Microsoft's Static Driver Verifier (SDV), and evaluate it on a series of hard SDV benchmarks.

cs.PL↗

arXiv2020

Building Reliable Cloud Services Using P# (Experience Report)

Cloud services must typically be distributed across a large number of machines in order to make use of multiple compute and storage resources. This opens the programmer to several sources of complexity such as concurrency, order of message delivery, lossy network, timeouts and failures, all of which impose a high cognitive burden. This paper presents evidence that technology inspired by formal-methods, delivered as part of a programming framework, can help address these challenges. In particular, we describe the experience of several engineering teams in Microsoft Azure that used the open-source P# programming framework to build multiple reliable cloud services. P# imposes a principled design pattern that allows writing formal specifications alongside production code that can be systematically tested, without deviating from routine engineering practices. Engineering teams that have been using P# have reported dramatically increased productivity (in time taken to push new features to production) as well as services that have been running live for months without any issues in features developed and tested with P#.

cs.PL↗

arXiv2019

CONFLLVM: A Compiler for Enforcing Data Confidentiality in Low-Level Code

We present an instrumenting compiler for enforcing data confidentiality in low-level applications (e.g. those written in C) in the presence of an active adversary. In our approach, the programmer marks secret data by writing lightweight annotations on top-level definitions in the source code. The compiler then uses a static flow analysis coupled with efficient runtime instrumentation, a custom memory layout, and custom control-flow integrity checks to prevent data leaks even in the presence of low-level attacks. We have implemented our scheme as part of the LLVM compiler. We evaluate it on the SPEC micro-benchmarks for performance, and on larger, real-world applications (including OpenLDAP, which is around 300KLoC) for programmer overhead required to restructure the application when protecting the sensitive data such as passwords. We find that performance overheads introduced by our instrumentation are moderate (average 12% on SPEC), and the programmer effort to port OpenLDAP is only about 160 LoC.

cs.PL↗

arXiv2019

Reliable State Machines: A Framework for Programming Reliable Cloud Services

Building reliable applications for the cloud is challenging because of unpredictable failures during a program's execution. This paper presents a programming framework called Reliable State Machines (RSMs), that offers fault-tolerance by construction. Using our framework, a programmer can build an application as several (possibly distributed) RSMs that communicate with each other via messages, much in the style of actor-based programming. Each RSM is additionally fault-tolerant by design and offers the illusion of being "always-alive". An RSM is guaranteed to process each input request exactly once, as one would expect in a failure-free environment. The RSM runtime automatically takes care of persisting state and rehydrating it on a failover. We present the core syntax and semantics of RSMs, along with a formal proof of failure-transparency. We provide an implementation of the RSM framework and runtime on the .NET platform for deploying services to Microsoft Azure. We carried out an extensive performance evaluation on micro-benchmarks to show that one can build high-throughput applications with RSMs. We also present a case study where we rewrote a significant part of a production cloud service using RSMs. The resulting service has simpler code and exhibits production-grade performance.

cs.PL↗

arXiv2017

Precise Null Pointer Analysis Through Global Value Numbering

Precise analysis of pointer information plays an important role in many static analysis techniques and tools today. The precision, however, must be balanced against the scalability of the analysis. This paper focusses on improving the precision of standard context and flow insensitive alias analysis algorithms at a low scalability cost. In particular, we present a semantics-preserving program transformation that drastically improves the precision of existing analyses when deciding if a pointer can alias NULL. Our program transformation is based on Global Value Numbering, a scheme inspired from compiler optimizations literature. It allows even a flow-insensitive analysis to make use of branch conditions such as checking if a pointer is NULL and gain precision. We perform experiments on real-world code to measure the overhead in performing the transformation and the improvement in the precision of the analysis. We show that the precision improves from 86.56% to 98.05%, while the overhead is insignificant.

cs.PL↗

arXiv2014

Automatically finding atomic regions for fixing bugs in Concurrent programs

This paper presents a technique for automatically constructing a fix for buggy concurrent programs: given a concurrent program that does not satisfy user-provided assertions, we infer atomic blocks that fix the program. An atomic block protects a piece of code and ensures that it runs without interruption from other threads. Our technique uses a verification tool as a subroutine to find the smallest atomic regions that remove all bugs in a given program. Keeping the atomic regions small allows for maximum concurrency. We have implemented our approach in a tool called AtomicInf. A user of AtomicInf can choose between strong and weak atomicity semantics for the inferred fix. While the former is simpler to find, the latter provides more information about the bugs that got fixed. We ran AtomicInf on several benchmarks and came up with the smallest and the most precise atomic regions in all of them. We implemented an earlier technique to our setting and observed that AtomicInf is 1.7 times faster on an average as compared to an earlier approach.

cs.SE↗

arXiv2013

Variable and Thread Bounding for Systematic Testing of Multithreaded Programs

Previous approaches to systematic state-space exploration for testing multi-threaded programs have proposed context-bounding and depth-bounding to be effective ranking algorithms for testing multithreaded programs. This paper proposes two new metrics to rank thread schedules for systematic state-space exploration. Our metrics are based on characterization of a concurrency bug using v (the minimum number of distinct variables that need to be involved for the bug to manifest) and t (the minimum number of distinct threads among which scheduling constraints are required to manifest the bug). Our algorithm is based on the hypothesis that in practice, most concurrency bugs have low v (typically 1- 2) and low t (typically 2-4) characteristics. We iteratively explore the search space of schedules in increasing orders of v and t. We show qualitatively and empirically that our algorithm finds common bugs in fewer number of execution runs, compared with previous approaches. We also show that using v and t improves the lower bounds on the probability of finding bugs through randomized algorithms. Systematic exploration of schedules requires instrumenting each variable access made by a program, which can be very expensive and severely limits the applicability of this approach. Previous work [5, 19] has avoided this problem by interposing only on synchronization operations (and ignoring other variable accesses). We demonstrate that by using variable bounding (v) and a static imprecise alias analysis, we can interpose on all variable accesses (and not just synchronization operations) at 10-100x less overhead than previous approaches.

cs.PL↗