论文信息 - Detecting Data Races on OpenCL Kernels with Symbolic Execution

Detecting Data Races on OpenCL Kernels with Symbolic Execution

We present an automatic analysis technique for checking data races on OpenCL kernels. Our method defines symbolic execution techniques based on separation logic with suitable abstractions to automatically detect non-benign racy behaviours on kernel

Jérémy Dubreil | Dino Distefano

[1] Dawson R. Engler,et al. KLEE: Unassisted and Automatic Generation of High-Coverage Tests for Complex Systems Programs , 2008, OSDI.

[2] Alastair F. Donaldson,et al. Interleaving and Lock-Step Semantics for Analysis and Verification of GPU Kernels , 2013, ESOP.

[3] John C. Reynolds,et al. Separation logic: a logic for shared mutable data structures , 2002, Proceedings 17th Annual IEEE Symposium on Logic in Computer Science.

[4] Paul H. J. Kelly,et al. Symbolic Testing of OpenCL Code , 2011, Haifa Verification Conference.

[5] Guodong Li,et al. Scalable SMT-based verification of GPU kernel functions , 2010, FSE '10.

[6] Peter W. O'Hearn,et al. A Local Shape Analysis Based on Separation Logic , 2006, TACAS.

[7] Peter W. O'Hearn,et al. Symbolic Execution with Separation Logic , 2005, APLAS.

[8] Adam Betts,et al. GPUVerify: a verifier for GPU kernels , 2012, OOPSLA '12.

[9] Patrick Cousot,et al. Abstract interpretation: a unified lattice model for static analysis of programs by construction or approximation of fixpoints , 1977, POPL.

[10] Peng Li,et al. GKLEE: concolic verification and test generation for GPUs , 2012, PPoPP '12.

[11] Sorin Lerner,et al. Verifying GPU kernels by test amplification , 2012, PLDI.

[12] Marieke Huisman,et al. Specification and Verification of GPGPU Programs using Permission-Based Separation Logic , 2013 .