摘要：A thorough understanding of the timing behavior of embedded systems software has become very important. With the advent of ever more complex embedded software e.g. in autonomous driving, the size of this software is growing at a fast pace. Execution time profiles (ETP) have proven to be a useful way to understand the timing behavior of embedded software. Collecting these ETPs was either limited to small applications or required multiple runs of the same software for calibration processes. In this contribution, we present a novel method for collecting ETPs in a single shot of the software at very high quality even for large applications.
关键词：WCET; Execution Time Profiling; ARM CoreSight; Event Stream Processing
