Probabilistic program analysis for parallelizing compilers

I. Forsythe, Peter Milligan, Paul Sage

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

Abstract

Parallelizing compilers have difficulty analysing and optimising complex code. To address this, some analysis may be delayed until run-time, and techniques such as speculative execution used. Furthermore, to enhance performance, a feedback loop may be setup between the compile time and run-time analysis systems, as in iterative compilation. To extend this, it is proposed that the run-time analysis collects information about the values of variables not already determined, and estimates a probability measure for the sampled values. These measures may be used to guide optimisations in further analyses of the program. To address the problem of variables with measures as values, this paper also presents an outline of a novel combination of previous probabilistic denotational semantics models, applied to a simple imperative language.
Original languageEnglish
Title of host publicationHigh Performance Computing for Computational Science: VECPAR 2004
Pages610-622
Number of pages13
Volume3402
DOIs
Publication statusPublished - Apr 2005
EventHigh Performance Computing for Computational Science - VECPAR 2004 - Valencia, Spain
Duration: 01 Apr 200501 Apr 2005

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume3402
ISSN (Electronic)1611-3349

Conference

ConferenceHigh Performance Computing for Computational Science - VECPAR 2004
CountrySpain
CityValencia
Period01/04/200501/04/2005

Fingerprint

Information analysis
Semantics
Feedback

Bibliographical note

ISSN: 0302-9743

Cite this

Forsythe, I., Milligan, P., & Sage, P. (2005). Probabilistic program analysis for parallelizing compilers. In High Performance Computing for Computational Science: VECPAR 2004 (Vol. 3402, pp. 610-622). (Lecture Notes in Computer Science ; Vol. 3402). https://doi.org/10.1007/11403937_46
Forsythe, I. ; Milligan, Peter ; Sage, Paul. / Probabilistic program analysis for parallelizing compilers. High Performance Computing for Computational Science: VECPAR 2004. Vol. 3402 2005. pp. 610-622 (Lecture Notes in Computer Science ).
@inproceedings{ca76718f287743f9909c138ad27003dd,
title = "Probabilistic program analysis for parallelizing compilers",
abstract = "Parallelizing compilers have difficulty analysing and optimising complex code. To address this, some analysis may be delayed until run-time, and techniques such as speculative execution used. Furthermore, to enhance performance, a feedback loop may be setup between the compile time and run-time analysis systems, as in iterative compilation. To extend this, it is proposed that the run-time analysis collects information about the values of variables not already determined, and estimates a probability measure for the sampled values. These measures may be used to guide optimisations in further analyses of the program. To address the problem of variables with measures as values, this paper also presents an outline of a novel combination of previous probabilistic denotational semantics models, applied to a simple imperative language.",
author = "I. Forsythe and Peter Milligan and Paul Sage",
note = "ISSN: 0302-9743",
year = "2005",
month = "4",
doi = "10.1007/11403937_46",
language = "English",
volume = "3402",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "610--622",
booktitle = "High Performance Computing for Computational Science: VECPAR 2004",

}

Forsythe, I, Milligan, P & Sage, P 2005, Probabilistic program analysis for parallelizing compilers. in High Performance Computing for Computational Science: VECPAR 2004. vol. 3402, Lecture Notes in Computer Science , vol. 3402, pp. 610-622, High Performance Computing for Computational Science - VECPAR 2004, Valencia, Spain, 01/04/2005. https://doi.org/10.1007/11403937_46

Probabilistic program analysis for parallelizing compilers. / Forsythe, I.; Milligan, Peter; Sage, Paul.

High Performance Computing for Computational Science: VECPAR 2004. Vol. 3402 2005. p. 610-622 (Lecture Notes in Computer Science ; Vol. 3402).

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

TY - GEN

T1 - Probabilistic program analysis for parallelizing compilers

AU - Forsythe, I.

AU - Milligan, Peter

AU - Sage, Paul

N1 - ISSN: 0302-9743

PY - 2005/4

Y1 - 2005/4

N2 - Parallelizing compilers have difficulty analysing and optimising complex code. To address this, some analysis may be delayed until run-time, and techniques such as speculative execution used. Furthermore, to enhance performance, a feedback loop may be setup between the compile time and run-time analysis systems, as in iterative compilation. To extend this, it is proposed that the run-time analysis collects information about the values of variables not already determined, and estimates a probability measure for the sampled values. These measures may be used to guide optimisations in further analyses of the program. To address the problem of variables with measures as values, this paper also presents an outline of a novel combination of previous probabilistic denotational semantics models, applied to a simple imperative language.

AB - Parallelizing compilers have difficulty analysing and optimising complex code. To address this, some analysis may be delayed until run-time, and techniques such as speculative execution used. Furthermore, to enhance performance, a feedback loop may be setup between the compile time and run-time analysis systems, as in iterative compilation. To extend this, it is proposed that the run-time analysis collects information about the values of variables not already determined, and estimates a probability measure for the sampled values. These measures may be used to guide optimisations in further analyses of the program. To address the problem of variables with measures as values, this paper also presents an outline of a novel combination of previous probabilistic denotational semantics models, applied to a simple imperative language.

UR - http://www.scopus.com/inward/record.url?scp=24144484691&partnerID=8YFLogxK

U2 - 10.1007/11403937_46

DO - 10.1007/11403937_46

M3 - Conference contribution

VL - 3402

T3 - Lecture Notes in Computer Science

SP - 610

EP - 622

BT - High Performance Computing for Computational Science: VECPAR 2004

ER -

Forsythe I, Milligan P, Sage P. Probabilistic program analysis for parallelizing compilers. In High Performance Computing for Computational Science: VECPAR 2004. Vol. 3402. 2005. p. 610-622. (Lecture Notes in Computer Science ). https://doi.org/10.1007/11403937_46