Software Engineer at GoogleAbout Kostya: Kostya Serebryany is a Software Engineer at Google. His team develops and deploys dynamic testing tools, such as AddressSanitizer, MemorySanitizer, ThreadSanitizer, and libFuzzer. Prior to joining Google in 2007, Kostya spent 4 years at Elbrus/MCST working for Sun compiler lab and then 3 years at Intel Compiler Lab. Kostya holds a PhD from mesi.ru and a M.S. from msu.ru.
Talk Title: Hardware-assisted testing in production
Every software vendor is trying to "shift left", i.e. to move bug detection to earlier stages of software development. This is an important goal, which we are unlikely to ever achieve 100%, and thus we need to keep finding bugs when the software is already released. In this talk we will discuss three testing mechanisms that use special hardware features to enable sampled bug detection with near-zero overhead in production:
- GWP-ASan, detects heap corruption with hardware page protection.
- Per-allocation sampling with Arm Memory Tagging Extension.
- GWP-TSan, detects data races using hardware watchpoints.
Mohammad Reza Mousavi
King's College LondonAbout Mohammed: Mohammad Reza Mousavi is a professor of Software Engineering at King's College London (as of August 2021). Prior to that, he held the chair in Data-Oriented Software Engineering at the University of Leicester, UK. Before moving to the UK, he held positions in Sweden (Halmstad and Chalmers), The Netherlands (TU Eindhoven and TU Delft) and Iceland (Reykjavik). He has led various research grants from the British, Swedish, and Icelandic funding agencies on formal semantics, verification, and software and systems testing at the foundational and applied level, involving industrial sectors such as transport and healthcare. Currently, he is the principal investigator of the UKRI Trustworthy Autonomous Systems Node on Verifiability. His main research interest is in testing, particularly of variability-intensive and cyber-physical systems.
Catch Me If You Can: Doping Detection in Cyber-Physical Systems
We will start with a broad overview of our research philosophy on trust in autonomous systems. We will review some of the ongoing projects at our group in this context. Subsequently, we present a novel notion of doping cleanness for cyber-physical systems. This notion allows for perturbing the inputs and observing the perturbed outputs both in the time- and value-domains. We instantiate our definition using existing techniques for conformance testing and runtime monitoring for cyber-physical systems. We show that our generalised definitions are essential in a data-driven method for doping detection and apply our definitions to a case study concerning diesel emission tests. We report on the doping detection results on the NOx emission of a particular diesel vehicle.
The talk is based on joint work with the following people:
Sebastian Biewer (Saarland University), Rayna Dimitrova (CISPA Saarbrücken), Michael Fris (Automotive Powertrain HTW Saar), Maciej Gazda (University of Sheffield), Thomas Heinze (Automotive Powertrain HTW Saar), and Holger Hermanns (Saarland University)