PROGRAML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations

Chris Cummins; Zacharias V. Fisches; Tal  Ben-Nun; Torsten Hoefler; Michael F P O'Boyle; Hugh Leather

PROGRAML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations

Chris Cummins^*, Zacharias V. Fisches, Tal Ben-Nun, Torsten Hoefler, Michael F P O'Boyle, Hugh Leather

^*Corresponding author for this work

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

Abstract

Machine learning (ML) is increasingly seen as a viable approach for building compiler optimization heuristics, but many ML methods cannot replicate even the simplest of the data flow analyses that are critical to making good optimization decisions. We posit that if ML cannot do that, then it is insufficiently able to reason about programs. We formulate data flow analyses as supervised learning tasks and introduce a large open dataset of programs and their corresponding labels from several analyses. We use this dataset to benchmark ML methods and show that they struggle on these fundamental program reasoning tasks. We propose ProGraML - Program Graphs for Machine Learning - a language-independent, portable representation of program semantics. ProGraML overcomes the limitations of prior works and yields improved performance on downstream optimization tasks.

Original language	English
Title of host publication	Proceedings of the 38th International Conference on Machine Learning
Publisher	PMLR
Pages	2244-2253
Number of pages	10
Publication status	Published - 18 Jul 2021
Event	Thirty-eighth International Conference on Machine Learning - Online Duration: 18 Jul 2021 → 24 Jul 2021 https://icml.cc/

Publication series

Name	Proceedings of Machine Learning Research
Volume	139
ISSN (Electronic)	2640-3498

Conference

Conference	Thirty-eighth International Conference on Machine Learning
Abbreviated title	ICML 2021
Period	18/07/21 → 24/07/21
Internet address	https://icml.cc/

Access to Document

PROGRAML_CUMMINS_DOA08052021_AFVAccepted author manuscript, 493 KBLicence: All Rights Reserved

http://proceedings.mlr.press/v139/cummins21a.htmlLicence: All Rights Reserved

Cite this

Cummins, C., Fisches, Z. V., Ben-Nun, T., Hoefler, T., O'Boyle, M. F. P., & Leather, H. (2021). PROGRAML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations. In Proceedings of the 38th International Conference on Machine Learning (pp. 2244-2253). (Proceedings of Machine Learning Research; Vol. 139). PMLR. http://proceedings.mlr.press/v139/cummins21a.html

@inproceedings{7d0c1ed8ef76471989f25db8764b68b2,

title = "PROGRAML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations",

abstract = "Machine learning (ML) is increasingly seen as a viable approach for building compiler optimization heuristics, but many ML methods cannot replicate even the simplest of the data flow analyses that are critical to making good optimization decisions. We posit that if ML cannot do that, then it is insufficiently able to reason about programs. We formulate data flow analyses as supervised learning tasks and introduce a large open dataset of programs and their corresponding labels from several analyses. We use this dataset to benchmark ML methods and show that they struggle on these fundamental program reasoning tasks. We propose ProGraML - Program Graphs for Machine Learning - a language-independent, portable representation of program semantics. ProGraML overcomes the limitations of prior works and yields improved performance on downstream optimization tasks. ",

author = "Chris Cummins and Fisches, {Zacharias V.} and Tal Ben-Nun and Torsten Hoefler and O'Boyle, {Michael F P} and Hugh Leather",

year = "2021",

month = jul,

day = "18",

language = "English",

series = "Proceedings of Machine Learning Research",

publisher = "PMLR",

pages = "2244--2253",

booktitle = "Proceedings of the 38th International Conference on Machine Learning",

note = "Thirty-eighth International Conference on Machine Learning, ICML 2021 ; Conference date: 18-07-2021 Through 24-07-2021",

url = "https://icml.cc/",

}

Cummins, C, Fisches, ZV, Ben-Nun, T, Hoefler, T, O'Boyle, MFP & Leather, H 2021, PROGRAML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations. in Proceedings of the 38th International Conference on Machine Learning. Proceedings of Machine Learning Research, vol. 139, PMLR, pp. 2244-2253, Thirty-eighth International Conference on Machine Learning, 18/07/21. <http://proceedings.mlr.press/v139/cummins21a.html>

PROGRAML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations. / Cummins, Chris; Fisches, Zacharias V.; Ben-Nun, Tal et al.

Proceedings of the 38th International Conference on Machine Learning. PMLR, 2021. p. 2244-2253 (Proceedings of Machine Learning Research; Vol. 139).

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

TY - GEN

T1 - PROGRAML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations

AU - Cummins, Chris

AU - Fisches, Zacharias V.

AU - Ben-Nun, Tal

AU - Hoefler, Torsten

AU - O'Boyle, Michael F P

AU - Leather, Hugh

PY - 2021/7/18

Y1 - 2021/7/18

N2 - Machine learning (ML) is increasingly seen as a viable approach for building compiler optimization heuristics, but many ML methods cannot replicate even the simplest of the data flow analyses that are critical to making good optimization decisions. We posit that if ML cannot do that, then it is insufficiently able to reason about programs. We formulate data flow analyses as supervised learning tasks and introduce a large open dataset of programs and their corresponding labels from several analyses. We use this dataset to benchmark ML methods and show that they struggle on these fundamental program reasoning tasks. We propose ProGraML - Program Graphs for Machine Learning - a language-independent, portable representation of program semantics. ProGraML overcomes the limitations of prior works and yields improved performance on downstream optimization tasks.

AB - Machine learning (ML) is increasingly seen as a viable approach for building compiler optimization heuristics, but many ML methods cannot replicate even the simplest of the data flow analyses that are critical to making good optimization decisions. We posit that if ML cannot do that, then it is insufficiently able to reason about programs. We formulate data flow analyses as supervised learning tasks and introduce a large open dataset of programs and their corresponding labels from several analyses. We use this dataset to benchmark ML methods and show that they struggle on these fundamental program reasoning tasks. We propose ProGraML - Program Graphs for Machine Learning - a language-independent, portable representation of program semantics. ProGraML overcomes the limitations of prior works and yields improved performance on downstream optimization tasks.

M3 - Conference contribution

T3 - Proceedings of Machine Learning Research

SP - 2244

EP - 2253

BT - Proceedings of the 38th International Conference on Machine Learning

PB - PMLR

T2 - Thirty-eighth International Conference on Machine Learning

Y2 - 18 July 2021 through 24 July 2021

ER -

PROGRAML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this