Predictive Data Race Detection for GPUs

TL;DR

This paper introduces GWCP, a new predictive partial order for GPU kernels that improves data race detection by considering GPU-specific features, implemented in the PreDataR tool with optimizations for better coverage and efficiency.

Contribution

It extends the weak-causally-precedes relation to GPU kernels, enabling more effective data race detection tailored to GPU programming models.

Findings

PreDataR achieves higher data race coverage than prior methods.

PreDataR operates with practical runtime overheads.

The approach effectively handles GPU-specific synchronization mechanisms.

Abstract

The high degree of parallelism and relatively complicated synchronization mechanisms in GPUs make writing correct kernels difficult. Data races pose one such concurrency correctness challenge, and therefore, effective methods of detecting as many data races as possible are required. Predictive partial order relations for CPU programs aim to expose data races that can be hidden during a dynamic execution. Existing predictive partial orders cannot be na\"ively applied to analyze GPU kernels because of the differences in programming models. This work proposes GWCP, a predictive partial order for data race detection of GPU kernels. GWCP extends a sound and precise relation called weak-causally-precedes (WCP) proposed in the context of multithreaded shared memory CPU programs to GPU kernels. GWCP takes into account the GPU thread hierarchy and different synchronization semantics such as barrier synchronization and scoped atomics and locks. We implement a tool called PreDataR that tracks the GWCP relation using binary instrumentation. PreDataR includes three optimizations and a novel vector clock compression scheme that are readily applicable to other partial order based analyses. Our evaluation with several microbenchmarks and benchmarks shows that PreDataR has better data race coverage compared to prior techniques at practical run-time overheads.