Symbolic crosschecking of floating-point and SIMD code

  • Authors:
  • Peter Collingbourne;Cristian Cadar;Paul H.J. Kelly

  • Affiliations:
  • Imperial College London, London, United Kingdom;Imperial College London, London, United Kingdom;Imperial College London, London, United Kingdom

  • Venue:
  • Proceedings of the sixth conference on Computer systems
  • Year:
  • 2011

Quantified Score

Hi-index 0.02

Visualization

Abstract

We present an effective technique for crosschecking an IEEE 754 floating-point program and its SIMD-vectorized version, implemented in KLEE-FP, an extension to the KLEE symbolic execution tool that supports symbolic reasoning on the equivalence between floating-point values. The key insight behind our approach is that floatingpoint values are only reliably equal if they are essentially built by the same operations. As a result, our technique works by lowering the Intel Streaming SIMD Extension (SSE) instruction set to primitive integer and floating-point operations, and then using an algorithm based on symbolic expression matching augmented with canonicalization rules. Under symbolic execution, we have to verify equivalence along every feasible control-flow path. We reduce the branching factor of this process by aggressively merging conditionals, if-converting branches into select operations via an aggressive phi-node folding transformation. We applied KLEE-FP to OpenCV, a popular open source computer vision library. KLEE-FP was able to successfully crosscheck 51 SIMD/SSE implementations against their corresponding scalar versions, proving the bounded equivalence of 41 of them (i.e., on images up to a certain size), and finding inconsistencies in the other 10.