Enhancing cross-ISA DBT through automatically learned translation rules

Wenwen Wang; Stephen A McCamant; Antonia B Zhai; Pen-Chung Yew

doi:10.1145/3173162.3177160

Enhancing cross-ISA DBT through automatically learned translation rules

Wenwen Wang, Stephen A McCamant, Antonia B Zhai, Pen-Chung Yew

Computer Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

12 Scopus citations

Abstract

This paper presents a novel approach for dynamic binary translation (DBT) to automatically learn translation rules from guest and host binaries compiled from the same source code. The learned translation rules are then verified via binary symbolic execution and used in an existing DBT system, QEMU, to generate more efficient host binary code. Experimental results on SPEC CINT2006 showthat the average time of learning a translation rule is less than two seconds. With the rules learned from a collection of benchmark programs excluding the targeted program itself, an average 1.25X performance speedup over QEMU can be achieved for SPEC CINT2006. Moreover, the translation overhead introduced by this rule-based approach is very small even for shortrunning workloads.

Original language	English (US)
Title of host publication	ASPLOS 2018 - 23rd International Conference on Architectural Support for Programming Languages and Operating Systems
Publisher	Association for Computing Machinery
Pages	84-97
Number of pages	14
Volume	53
Edition	2
ISBN (Electronic)	9781450349116
DOIs	https://doi.org/10.1145/3173162.3177160
State	Published - Mar 19 2018
Event	23rd International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2018 - Williamsburg, United States Duration: Mar 24 2018 → Mar 28 2018

Publication series

Name	ACM SIGPLAN Notices
Publisher	Association for Computing Machinery (ACM)
ISSN (Print)	1523-2867

Other

Other	23rd International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2018
Country/Territory	United States
City	Williamsburg
Period	3/24/18 → 3/28/18

Bibliographical note

Funding Information:
We are very grateful to David Grove and the anonymous reviewers for their valuable comments and feedback. This material is based upon work supported by the National Science Foundation under Grant No. CNS-1514444. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Publisher Copyright:
© 2018 Copyright held by the owner/author(s).

Keywords

DBT
Rule learning
Symbolic execution

Access

10.1145/3173162.3177160

OpenUrl availability

Full text

Cite this

Wang, W., McCamant, S. A., Zhai, A. B., & Yew, P.-C. (2018). Enhancing cross-ISA DBT through automatically learned translation rules. In ASPLOS 2018 - 23rd International Conference on Architectural Support for Programming Languages and Operating Systems (2 ed., Vol. 53, pp. 84-97). (ACM SIGPLAN Notices). Association for Computing Machinery. https://doi.org/10.1145/3173162.3177160

Enhancing cross-ISA DBT through automatically learned translation rules. / Wang, Wenwen; McCamant, Stephen A ; Zhai, Antonia B et al.
ASPLOS 2018 - 23rd International Conference on Architectural Support for Programming Languages and Operating Systems. Vol. 53 2. ed. Association for Computing Machinery, 2018. p. 84-97 (ACM SIGPLAN Notices).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Wang, W, McCamant, SA , Zhai, AB & Yew, P-C 2018, Enhancing cross-ISA DBT through automatically learned translation rules. in ASPLOS 2018 - 23rd International Conference on Architectural Support for Programming Languages and Operating Systems. 2 edn, vol. 53, ACM SIGPLAN Notices, Association for Computing Machinery, pp. 84-97, 23rd International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2018, Williamsburg, United States, 3/24/18. https://doi.org/10.1145/3173162.3177160

@inproceedings{910c3b091a354bbbab5a279e550683f0,

title = "Enhancing cross-ISA DBT through automatically learned translation rules",

abstract = "This paper presents a novel approach for dynamic binary translation (DBT) to automatically learn translation rules from guest and host binaries compiled from the same source code. The learned translation rules are then verified via binary symbolic execution and used in an existing DBT system, QEMU, to generate more efficient host binary code. Experimental results on SPEC CINT2006 showthat the average time of learning a translation rule is less than two seconds. With the rules learned from a collection of benchmark programs excluding the targeted program itself, an average 1.25X performance speedup over QEMU can be achieved for SPEC CINT2006. Moreover, the translation overhead introduced by this rule-based approach is very small even for shortrunning workloads.",

keywords = "DBT, Rule learning, Symbolic execution",

author = "Wenwen Wang and McCamant, {Stephen A} and Zhai, {Antonia B} and Pen-Chung Yew",

note = "Funding Information: We are very grateful to David Grove and the anonymous reviewers for their valuable comments and feedback. This material is based upon work supported by the National Science Foundation under Grant No. CNS-1514444. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. Publisher Copyright: {\textcopyright} 2018 Copyright held by the owner/author(s).; 23rd International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2018 ; Conference date: 24-03-2018 Through 28-03-2018",

year = "2018",

month = mar,

day = "19",

doi = "10.1145/3173162.3177160",

language = "English (US)",

volume = "53",

series = "ACM SIGPLAN Notices",

publisher = "Association for Computing Machinery",

pages = "84--97",

booktitle = "ASPLOS 2018 - 23rd International Conference on Architectural Support for Programming Languages and Operating Systems",

edition = "2",

}

TY - GEN

T1 - Enhancing cross-ISA DBT through automatically learned translation rules

AU - Wang, Wenwen

AU - McCamant, Stephen A

AU - Zhai, Antonia B

AU - Yew, Pen-Chung

N1 - Funding Information: We are very grateful to David Grove and the anonymous reviewers for their valuable comments and feedback. This material is based upon work supported by the National Science Foundation under Grant No. CNS-1514444. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. Publisher Copyright: © 2018 Copyright held by the owner/author(s).

PY - 2018/3/19

Y1 - 2018/3/19

N2 - This paper presents a novel approach for dynamic binary translation (DBT) to automatically learn translation rules from guest and host binaries compiled from the same source code. The learned translation rules are then verified via binary symbolic execution and used in an existing DBT system, QEMU, to generate more efficient host binary code. Experimental results on SPEC CINT2006 showthat the average time of learning a translation rule is less than two seconds. With the rules learned from a collection of benchmark programs excluding the targeted program itself, an average 1.25X performance speedup over QEMU can be achieved for SPEC CINT2006. Moreover, the translation overhead introduced by this rule-based approach is very small even for shortrunning workloads.

AB - This paper presents a novel approach for dynamic binary translation (DBT) to automatically learn translation rules from guest and host binaries compiled from the same source code. The learned translation rules are then verified via binary symbolic execution and used in an existing DBT system, QEMU, to generate more efficient host binary code. Experimental results on SPEC CINT2006 showthat the average time of learning a translation rule is less than two seconds. With the rules learned from a collection of benchmark programs excluding the targeted program itself, an average 1.25X performance speedup over QEMU can be achieved for SPEC CINT2006. Moreover, the translation overhead introduced by this rule-based approach is very small even for shortrunning workloads.

KW - DBT

KW - Rule learning

KW - Symbolic execution

UR - http://www.scopus.com/inward/record.url?scp=85045376114&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85045376114&partnerID=8YFLogxK

U2 - 10.1145/3173162.3177160

DO - 10.1145/3173162.3177160

M3 - Conference contribution

AN - SCOPUS:85060086164

VL - 53

T3 - ACM SIGPLAN Notices

SP - 84

EP - 97

BT - ASPLOS 2018 - 23rd International Conference on Architectural Support for Programming Languages and Operating Systems

PB - Association for Computing Machinery

T2 - 23rd International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2018

Y2 - 24 March 2018 through 28 March 2018

ER -

Enhancing cross-ISA DBT through automatically learned translation rules

Abstract

Publication series

Other

Bibliographical note

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this