Using Graph Neural Networks for Program Termination
This work addresses debugging nontermination bugs for programmers by providing a machine learning-based estimation tool, which is incremental as it builds on prior neural network approaches but introduces graph-based methods and adaptations like attention and segmentation.
The paper tackles the problem of estimating program termination behavior and identifying likely causes of nontermination for debugging purposes, using Graph Neural Networks (GNNs) to analyze graph representations of programs, achieving results that include classifiers for termination and semantic segmentation to localize problematic AST nodes.
Termination analyses investigate the termination behavior of programs, intending to detect nontermination, which is known to cause a variety of program bugs (e.g. hanging programs, denial-of-service vulnerabilities). Beyond formal approaches, various attempts have been made to estimate the termination behavior of programs using neural networks. However, the majority of these approaches continue to rely on formal methods to provide strong soundness guarantees and consequently suffer from similar limitations. In this paper, we move away from formal methods and embrace the stochastic nature of machine learning models. Instead of aiming for rigorous guarantees that can be interpreted by solvers, our objective is to provide an estimation of a program's termination behavior and of the likely reason for nontermination (when applicable) that a programmer can use for debugging purposes. Compared to previous approaches using neural networks for program termination, we also take advantage of the graph representation of programs by employing Graph Neural Networks. To further assist programmers in understanding and debugging nontermination bugs, we adapt the notions of attention and semantic segmentation, previously used for other application domains, to programs. Overall, we designed and implemented classifiers for program termination based on Graph Convolutional Networks and Graph Attention Networks, as well as a semantic segmentation Graph Neural Network that localizes AST nodes likely to cause nontermination. We also illustrated how the information provided by semantic segmentation can be combined with program slicing to further aid debugging.