PrETi: Predicting Execution Time in Early Stage with LLVM and Machine Learning
Risheng Xu, Philipp Sieweck, Hermann von Hasseln, Dirk Nowotka
TL;DR
Preti addresses the need for early-stage, accurate execution-time prediction in safety-critical real-time software by combining LLVM IR instrumentation with execution-flow, data-cache, and branch simulations, followed by ML-based time prediction trained on historical measurements. It introduces a two-component pipeline: preti_llvm for instrumented IR-based traces and preti_ml for learning from those traces, evaluated on a dataset of 1348 samples and achieving approximately 12% APE on public benchmarks. The approach outperforms state-of-the-art timing-prediction methods such as CATREEN and ITHEMAL while drastically reducing measurement workload. The work provides a scalable, hardware-independent pathway for ESTA, enabling faster iteration, better scheduling, and earlier debugging in real-time software development.
Abstract
We introduce preti, a novel framework for predicting software execution time during the early stages of development. preti leverages an LLVM-based simulation environment to extract timing-related runtime information, such as the count of executed LLVM IR instructions. This information, combined with historical execution time data, is utilized to train machine learning models for accurate time prediction. To further enhance prediction accuracy, our approach incorporates simulations of cache accesses and branch prediction. The evaluations on public benchmarks demonstrate that preti achieves an average Absolute Percentage Error (APE) of 11.98\%, surpassing state-of-the-art methods. These results underscore the effectiveness and efficiency of preti as a robust solution for early-stage timing analysis.