Binary Analysis for Performance, 
Security, and Correctness Tools

2018-04-24T05:04:23Z (GMT) by John Mellor-Crummey Barton Miller
The attached document is a one-slide lightning talk for a companion poster that describes work during 2018 as part of an NSF-funded collaboration between Rice University and the University of Wisconsin-Madison. Among other things, the companion poster describes adding task-based parallelism to Wisconsin's Dyninst binary analyzer, a new dynamic data race detector for OpenMP programs, and new capabilities in Rice's HPCToolkit performance tools to measure and analyze the the performance of GPU-accelerated code and generate insights into parallel application performance using automated analysis of program traces.<br>