File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: CUDAsmith: A fuzzer for CUDA compilers

TitleCUDAsmith: A fuzzer for CUDA compilers
Authors
KeywordsCompiler
compute unified device architecture (CUDA)
differential testing
equivalence modulo inputs (EMI) testing
fuzzing
general purpose computing on graphics processing unit (GPGPU)
Issue Date2020
PublisherIEEE. The Proceedings' web site is located at https://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1000143
Citation
Proceedings of the 2020 IEEE 44th Annual Computers, Software, and Applications Conference (COMPSAC), Madrid, Spain, 13-17 July 2020, p. 861-871 How to Cite?
AbstractCUDA is a parallel computing platform and programming model for the graphics processing unit (GPU) of NVIDIA. With CUDA programming, general purpose computing on GPU (GPGPU) is possible. However, the correctness of CUDA programs relies on the correctness of CUDA compilers, which is difficult to test due to its complexity. In this work, we propose CUDAsmith, a fuzzing framework for CUDA compilers. Our tool can randomly generate deterministic and valid CUDA kernel code with several different strategies. Moreover, it adopts random differential testing and EMI testing techniques to solve the test oracle problems of CUDA compiler testing. In particular, we lift live code injection to CUDA compiler testing to help generate EMI variants. Our fuzzing experiments with both the NVCC compiler and the Clang compiler for CUDA have detected thousands of failures, some of which have been confirmed by compiler developers. Finally, the cost-effectiveness of CUDAsmith is also thoroughly evaluated in our fuzzing experiment.
DescriptionCOMPSAC 2020 was held virutally due to COVID-19
Persistent Identifierhttp://hdl.handle.net/10722/287190
ISSN
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorJiang, B-
dc.contributor.authorWang, X-
dc.contributor.authorChan, WK-
dc.contributor.authorTse, TH-
dc.contributor.authorLi, N-
dc.contributor.authorYin, Y-
dc.contributor.authorZhang, Z-
dc.date.accessioned2020-09-22T02:57:11Z-
dc.date.available2020-09-22T02:57:11Z-
dc.date.issued2020-
dc.identifier.citationProceedings of the 2020 IEEE 44th Annual Computers, Software, and Applications Conference (COMPSAC), Madrid, Spain, 13-17 July 2020, p. 861-871-
dc.identifier.issn0730-3157-
dc.identifier.urihttp://hdl.handle.net/10722/287190-
dc.descriptionCOMPSAC 2020 was held virutally due to COVID-19-
dc.description.abstractCUDA is a parallel computing platform and programming model for the graphics processing unit (GPU) of NVIDIA. With CUDA programming, general purpose computing on GPU (GPGPU) is possible. However, the correctness of CUDA programs relies on the correctness of CUDA compilers, which is difficult to test due to its complexity. In this work, we propose CUDAsmith, a fuzzing framework for CUDA compilers. Our tool can randomly generate deterministic and valid CUDA kernel code with several different strategies. Moreover, it adopts random differential testing and EMI testing techniques to solve the test oracle problems of CUDA compiler testing. In particular, we lift live code injection to CUDA compiler testing to help generate EMI variants. Our fuzzing experiments with both the NVCC compiler and the Clang compiler for CUDA have detected thousands of failures, some of which have been confirmed by compiler developers. Finally, the cost-effectiveness of CUDAsmith is also thoroughly evaluated in our fuzzing experiment.-
dc.languageeng-
dc.publisherIEEE. The Proceedings' web site is located at https://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1000143-
dc.relation.ispartofIEEE Annual Computer Software and Applications Conference (COMPSAC) Proceedings-
dc.rightsIEEE Annual Computer Software and Applications Conference (COMPSAC) Proceedings. Copyright © IEEE.-
dc.rights©2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.-
dc.subjectCompiler-
dc.subjectcompute unified device architecture (CUDA)-
dc.subjectdifferential testing-
dc.subjectequivalence modulo inputs (EMI) testing-
dc.subjectfuzzing-
dc.subjectgeneral purpose computing on graphics processing unit (GPGPU)-
dc.titleCUDAsmith: A fuzzer for CUDA compilers-
dc.typeConference_Paper-
dc.identifier.emailTse, TH: thtse@cs.hku.hk-
dc.identifier.authorityTse, TH=rp00546-
dc.identifier.doi10.1109/COMPSAC48688.2020.0-156-
dc.identifier.scopuseid_2-s2.0-85094122760-
dc.identifier.hkuros314198-
dc.identifier.spage861-
dc.identifier.epage871-
dc.identifier.isiWOS:000629086600113-
dc.publisher.placeUnited States-
dc.identifier.issnl0730-3157-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats