关键词：GameTime；时序分析；嵌入式系统；数据依赖
摘 要：Execution time analysis is central to the design and veri cation of real-time embedded systems. One approach to this problem is GameTime, a toolkit for timing analysis, which predicts the timing of various program paths using platform measurements of a special subset of paths. This thesis explores extensions to GameTime to handle two common sources of data-dependent timing behavior: instructions with variable timing and load/store dependencies. We present a technique for automatically learning a model of the data dependencies and encoding this model into the code under analysis for processing by GameTime. Using these extensions, we show that GameTime more accurately predicts the timing for a variety of benchmarks.