Information Flow Control
with Program Dependence Graphs

Speaker:
Jens Krinke, CREST, King's College London

Date and Time:
April 13, 16:15

Location:
Institute for Software Systems, SBS 95 E, Rm. 4.042

Abstract:
In classic information flow control (IFC), noninterference
guarantees that no information flows from secret input channels to
public output channels. However, this notion turned out to be overly
restrictive as many intuitively secure programs do allow some release,
and thus intransitive noninterference allows declassification. We
developed a static analysis that allows intransitive noninterference
based on Program Dependence Graphs. In contrast to type systems that
annotate variables, our approach annotates information sources and
sinks. Our approach is flow- and context-sensitive and because it is
based on an industrial strength Program Dependence Generator (CodeSurfer
from GrammaTech), it allows IFC for realistic C programs.

Bio:
Jens Krinke is lecturer in the Department of Computer Science,
King's College London, and Deputy Director of CREST - the Centre for
Research on Evolution, Search and Testing. His current research topics
include program analysis for software engineering purposes, in
particular for bug detection, taint analysis and information flow
control. Currently, he also works on Clone Detection and Code Provenance.