Compiler optimizations with retrofitting transformations: Is there a semantic mismatch?

Jay P. Lim, Vinod Ganapathy, Santosh Nagarakatte

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations

Abstract

A retrofitting transformation modifies an input program by adding instrumentation to monitor security properties at runtime. These tools often transform the input program in complex ways. Compiler optimizations can erroneously remove the instrumentation added by a retrofitting transformation in the presence of semantic mismatches between the assumptions of retrofitting transformations and compiler optimizations. This paper proposes a strategy to ascertain that every event of interest that is checked in the retrofitted program is also checked after optimizations. Our initial experiments have identified bugs both in previously proposed retrofitting transformations and our implementations of retrofitting transformations.

Original languageEnglish (US)
Title of host publicationPLAS 2017 - Proceedings of the 2017 Workshop on Programming Languages and Analysis for Security, co-located with CCS 2017
PublisherAssociation for Computing Machinery, Inc
Pages37-42
Number of pages6
ISBN (Electronic)9781450350990
DOIs
StatePublished - Oct 30 2017
Event12th ACM SIGSAC Workshop on Programming Languages and Analysis for Security, PLAS 2017 - Dallas, United States
Duration: Oct 30 2017 → …

Publication series

NamePLAS 2017 - Proceedings of the 2017 Workshop on Programming Languages and Analysis for Security, co-located with CCS 2017
Volume2017-January

Other

Other12th ACM SIGSAC Workshop on Programming Languages and Analysis for Security, PLAS 2017
Country/TerritoryUnited States
CityDallas
Period10/30/17 → …

All Science Journal Classification (ASJC) codes

  • Computational Theory and Mathematics
  • Computer Science Applications
  • Software

Fingerprint

Dive into the research topics of 'Compiler optimizations with retrofitting transformations: Is there a semantic mismatch?'. Together they form a unique fingerprint.

Cite this