BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Title to be confirmedFlow- and Context-Sensitive Points-to Analysi
 s using Generalized Points-to Graphs - Pritam Gharat\, IIT Bombay
DTSTART:20160913T130000Z
DTEND:20160913T140000Z
UID:TALK67577@talks.cam.ac.uk
CONTACT:Alan Mycroft
DESCRIPTION:Many program analysis benefit from knowledge of aliasing or po
 ints-to information.\nComputing precise (FCPA\, fully flow-sensitive and c
 ontext-sensitive) and exhaustive (as against demand-driven) points-to info
 rmation is known to be computationally expensive.\nTherefore many practica
 l tools approximate the points-to information trading precision for effici
 ency.\nThis often has an adverse impact on computationally intensive analy
 ses such as model checking.\n\nPrevious approaches to FCPA (either top-dow
 n\, or bottom-up using placeholders to explicate unknown locations or by u
 sing multiple call-specific summary flow functions)\ndo not scale in pract
 ice.\nWe adopt a traditional approach (constructing bottom-up procedure su
 mmaries) involving pointers but use a novel representation of procedure su
 mmaries: the generalized points-to graph (GPG)\; these leave unknown locat
 ions implicit.\n\nBy design\, GPGs represent both memory (in terms of clas
 sical points-to facts) and memory transformers (in terms of generalized po
 ints-to facts).\nWe perform FCPA by progressively reducing generalized poi
 nts-to facts to classical points-to facts.\nGPGs distinguish between may a
 nd must pointer updates thereby facilitating strong updates within calling
  contexts.\nThe number of nodes in the GPG for a procedure is linearly bou
 nded by the number of variables and is independent of the number of statem
 ents in the procedure.\nEmpirical measurements on SPEC benchmarks show tha
 t GPGs are indeed compact in spite of large procedure sizes and scale to p
 rograms of up to 158kLoC.\n
LOCATION:GS15
END:VEVENT
END:VCALENDAR