|COOKIES: By using this website you agree that we can place Google Analytics Cookies on your device for performance monitoring.|
From Bounded to Unbounded Proofs of Correctness
If you have a question about this talk, please contact Microsoft Research Cambridge Talks Admins.
This event may be recorded and made available internally or externally via http://research.microsoft.com. Microsoft will own the copyright of any recordings made. If you do not wish to have your image/voice recorded please consider this before attending
We describe automated techniques for proving program safety, that is, proving that every execution of a program does not cause a crash (e.g., via division by zero) and satisfies intended functionality (e.g., a programmer-supplied assertion). The core idea underlying our approach is that by examining a bounded version of the program, with a finite number of execution paths, we can generalize the proof of correctness to the whole program. Our generalization capitalizes on advances in SMT solving for path enumeration, novel interpolant generation techniques for DAGs of formulas, and a tight integration with abstract domains in order to improve interpolant “quality”. Our approach forms the basis of UFO , a C verifier we built in LLVM , that has recently won numerous categories in the 2013 Competition on Software Verification (SV-COMP).
This talk is part of the Microsoft Research Cambridge, public talks series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
Other listsTheoretical, Modelling and Informatics - Chemistry Research Interest Group SBR Graduate Talks Cancer Metabolism Interest Group Seminars
Other talksEfficient Constrained Inference and Structured Neural Networks for Semantic Role Labeling Making sense of 20 billion data points a day TBC (SP Workshop) Pregnancy and Childbirth: Discussing current academic research and clinical expertise to ensure a healthy pregnancy. 'The cultural self-fashioning of Frederick the Great' The Game of Crime and Punishment