# Dependence-Aware, Unbounded Sound Predictive Race Detection

## TL;DR

This paper presents two novel, sound predictive race detection methods that incorporate data and control dependence, enabling analysis of full program executions and detecting more data races in large Java programs.

## Contribution

It introduces two new approaches that improve sound predictive race detection by handling full executions and precisely modeling dependencies, surpassing prior methods.

## Key findings

- Detect more data races in large Java programs.
- Handle full program executions effectively.
- Incorporate data and control dependence precisely.

## Abstract

Data races are a real problem for parallel software, yet hard to detect. Sound predictive analysis observes a program execution and detects data races that exist in some other, unobserved execution. However, existing predictive analyses miss races because they do not scale to full program executions or do not precisely incorporate data and control dependence. This paper introduces two novel, sound predictive approaches that incorporate data and control dependence and handle full program executions. An evaluation using real, large Java programs shows that these approaches detect more data races than the closest related approaches, thus advancing the state of the art in sound predictive race detection.