A progress bar for static analyzers*

Woosuk Lee, Hakjoo Oh, Kwangkeun Yi

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

We present a technique for devising a progress indicator of static analyzers. Progress indicator is a useful user interface that shows how close a static analysis has progressed so far to its completion. Because static analysis’ progress depends on the semantic complexity, not on the code size, of the target software, devising an accurate progressindicator is not obvious. Our technique first combines a semantic-based pre-analysis and a statistical method to approximate how a main analysis progresses in terms of lattice height of the abstract domain. Then, we use this information during the main analysis and estimate the analysis’ current progress. We apply the technique to three existing analyses (interval, octagon, and pointer analyses) for C and show the technique estimates the actual analysis progress for various benchmarks.

Original languageEnglish
Pages (from-to)184-200
Number of pages17
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8723
Publication statusPublished - 2014 Jan 1
Externally publishedYes

Fingerprint

Static analysis
Semantics
Information use
User interfaces
Static Analysis
Statistical methods
Octagon
Estimate
Statistical method
User Interface
Completion
Benchmark
Target
Interval
Software

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

A progress bar for static analyzers*. / Lee, Woosuk; Oh, Hakjoo; Yi, Kwangkeun.

In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 8723, 01.01.2014, p. 184-200.

Research output: Contribution to journalArticle

@article{9c368024a0e1470fb28fc1c36b061525,
title = "A progress bar for static analyzers*",
abstract = "We present a technique for devising a progress indicator of static analyzers. Progress indicator is a useful user interface that shows how close a static analysis has progressed so far to its completion. Because static analysis’ progress depends on the semantic complexity, not on the code size, of the target software, devising an accurate progressindicator is not obvious. Our technique first combines a semantic-based pre-analysis and a statistical method to approximate how a main analysis progresses in terms of lattice height of the abstract domain. Then, we use this information during the main analysis and estimate the analysis’ current progress. We apply the technique to three existing analyses (interval, octagon, and pointer analyses) for C and show the technique estimates the actual analysis progress for various benchmarks.",
author = "Woosuk Lee and Hakjoo Oh and Kwangkeun Yi",
year = "2014",
month = "1",
day = "1",
language = "English",
volume = "8723",
pages = "184--200",
journal = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - A progress bar for static analyzers*

AU - Lee, Woosuk

AU - Oh, Hakjoo

AU - Yi, Kwangkeun

PY - 2014/1/1

Y1 - 2014/1/1

N2 - We present a technique for devising a progress indicator of static analyzers. Progress indicator is a useful user interface that shows how close a static analysis has progressed so far to its completion. Because static analysis’ progress depends on the semantic complexity, not on the code size, of the target software, devising an accurate progressindicator is not obvious. Our technique first combines a semantic-based pre-analysis and a statistical method to approximate how a main analysis progresses in terms of lattice height of the abstract domain. Then, we use this information during the main analysis and estimate the analysis’ current progress. We apply the technique to three existing analyses (interval, octagon, and pointer analyses) for C and show the technique estimates the actual analysis progress for various benchmarks.

AB - We present a technique for devising a progress indicator of static analyzers. Progress indicator is a useful user interface that shows how close a static analysis has progressed so far to its completion. Because static analysis’ progress depends on the semantic complexity, not on the code size, of the target software, devising an accurate progressindicator is not obvious. Our technique first combines a semantic-based pre-analysis and a statistical method to approximate how a main analysis progresses in terms of lattice height of the abstract domain. Then, we use this information during the main analysis and estimate the analysis’ current progress. We apply the technique to three existing analyses (interval, octagon, and pointer analyses) for C and show the technique estimates the actual analysis progress for various benchmarks.

UR - http://www.scopus.com/inward/record.url?scp=84921752179&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84921752179&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:84921752179

VL - 8723

SP - 184

EP - 200

JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SN - 0302-9743

ER -