Summarizing Embedded Execution Traces through a Compact View

International audienceExecution traces are becoming a highly appreciated method to debug and analyze software programs, especially in the areas of embedded systems and high-performance parallel computing. However, execution traces can result in enormous volumes of data, making them difficult to manage and exploit. In this paper, we describe a graphical solution to help the user navigate such large traces, in order to localize the points of interest for the analysis