Analysis of input-dependent program behavior using active profiling

Xipeng Shen, Michael L. Scott, Chengliang Zhang, Sandhya Dwarkadas, Chen Ding, Mitsunori Ogihara

Research output: Chapter in Book/Report/Conference proceedingChapter

3 Citations (Scopus)

Abstract

Utility programs, which perform similar and largely independent operations on a sequence of inputs, include such common applications as compilers, interpreters, and document parsers; databases; and compression and encoding tools. The repetitive behavior of these programs, while often clear to users, has been difficult to capture automatically. We present an active profiling technique in which controlled inputs to utility programs are used to expose execution phases, which are then marked, automatically, through binary instrumentation, enabling us to exploit phase transitions in production runs with arbitrary inputs. We demonstrate the effectiveness and programmability of active profiling via experiments with six utility programs from the SPEC benchmark suite; compare to code and interval phases; and describe applications of active profiling to memory management and memory leak detection.

Original languageEnglish (US)
Title of host publicationProceedings of the 2007 Workshop on Experimental Computer Science
DOIs
StatePublished - 2007
Externally publishedYes
Event2007 Workshop on Experimental Computer Science - San Diego, CA, United States
Duration: Jun 25 2007Jun 26 2007

Other

Other2007 Workshop on Experimental Computer Science
CountryUnited States
CitySan Diego, CA
Period6/25/076/26/07

Fingerprint

Utility programs
Data storage equipment
Leak detection
Phase transitions
Experiments

Keywords

  • Active profiling
  • Dynamic optimization
  • Memory management
  • Program phase analysis and prediction

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Shen, X., Scott, M. L., Zhang, C., Dwarkadas, S., Ding, C., & Ogihara, M. (2007). Analysis of input-dependent program behavior using active profiling. In Proceedings of the 2007 Workshop on Experimental Computer Science [5] https://doi.org/10.1145/1281700.1281705

Analysis of input-dependent program behavior using active profiling. / Shen, Xipeng; Scott, Michael L.; Zhang, Chengliang; Dwarkadas, Sandhya; Ding, Chen; Ogihara, Mitsunori.

Proceedings of the 2007 Workshop on Experimental Computer Science. 2007. 5.

Research output: Chapter in Book/Report/Conference proceedingChapter

Shen, X, Scott, ML, Zhang, C, Dwarkadas, S, Ding, C & Ogihara, M 2007, Analysis of input-dependent program behavior using active profiling. in Proceedings of the 2007 Workshop on Experimental Computer Science., 5, 2007 Workshop on Experimental Computer Science, San Diego, CA, United States, 6/25/07. https://doi.org/10.1145/1281700.1281705
Shen X, Scott ML, Zhang C, Dwarkadas S, Ding C, Ogihara M. Analysis of input-dependent program behavior using active profiling. In Proceedings of the 2007 Workshop on Experimental Computer Science. 2007. 5 https://doi.org/10.1145/1281700.1281705
Shen, Xipeng ; Scott, Michael L. ; Zhang, Chengliang ; Dwarkadas, Sandhya ; Ding, Chen ; Ogihara, Mitsunori. / Analysis of input-dependent program behavior using active profiling. Proceedings of the 2007 Workshop on Experimental Computer Science. 2007.
@inbook{966688e712c049b7b861fa862ca18c5c,
title = "Analysis of input-dependent program behavior using active profiling",
abstract = "Utility programs, which perform similar and largely independent operations on a sequence of inputs, include such common applications as compilers, interpreters, and document parsers; databases; and compression and encoding tools. The repetitive behavior of these programs, while often clear to users, has been difficult to capture automatically. We present an active profiling technique in which controlled inputs to utility programs are used to expose execution phases, which are then marked, automatically, through binary instrumentation, enabling us to exploit phase transitions in production runs with arbitrary inputs. We demonstrate the effectiveness and programmability of active profiling via experiments with six utility programs from the SPEC benchmark suite; compare to code and interval phases; and describe applications of active profiling to memory management and memory leak detection.",
keywords = "Active profiling, Dynamic optimization, Memory management, Program phase analysis and prediction",
author = "Xipeng Shen and Scott, {Michael L.} and Chengliang Zhang and Sandhya Dwarkadas and Chen Ding and Mitsunori Ogihara",
year = "2007",
doi = "10.1145/1281700.1281705",
language = "English (US)",
isbn = "9781595937513",
booktitle = "Proceedings of the 2007 Workshop on Experimental Computer Science",

}

TY - CHAP

T1 - Analysis of input-dependent program behavior using active profiling

AU - Shen, Xipeng

AU - Scott, Michael L.

AU - Zhang, Chengliang

AU - Dwarkadas, Sandhya

AU - Ding, Chen

AU - Ogihara, Mitsunori

PY - 2007

Y1 - 2007

N2 - Utility programs, which perform similar and largely independent operations on a sequence of inputs, include such common applications as compilers, interpreters, and document parsers; databases; and compression and encoding tools. The repetitive behavior of these programs, while often clear to users, has been difficult to capture automatically. We present an active profiling technique in which controlled inputs to utility programs are used to expose execution phases, which are then marked, automatically, through binary instrumentation, enabling us to exploit phase transitions in production runs with arbitrary inputs. We demonstrate the effectiveness and programmability of active profiling via experiments with six utility programs from the SPEC benchmark suite; compare to code and interval phases; and describe applications of active profiling to memory management and memory leak detection.

AB - Utility programs, which perform similar and largely independent operations on a sequence of inputs, include such common applications as compilers, interpreters, and document parsers; databases; and compression and encoding tools. The repetitive behavior of these programs, while often clear to users, has been difficult to capture automatically. We present an active profiling technique in which controlled inputs to utility programs are used to expose execution phases, which are then marked, automatically, through binary instrumentation, enabling us to exploit phase transitions in production runs with arbitrary inputs. We demonstrate the effectiveness and programmability of active profiling via experiments with six utility programs from the SPEC benchmark suite; compare to code and interval phases; and describe applications of active profiling to memory management and memory leak detection.

KW - Active profiling

KW - Dynamic optimization

KW - Memory management

KW - Program phase analysis and prediction

UR - http://www.scopus.com/inward/record.url?scp=37849042793&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=37849042793&partnerID=8YFLogxK

U2 - 10.1145/1281700.1281705

DO - 10.1145/1281700.1281705

M3 - Chapter

SN - 9781595937513

BT - Proceedings of the 2007 Workshop on Experimental Computer Science

ER -