FITL: Extending LLVM for the translation of fault-injection directives

Joel E. Denny; Seyong Lee; Jeffrey S. Vetter

doi:10.1145/2833157.2833160

FITL: Extending LLVM for the translation of fault-injection directives

Joel E. Denny, Seyong Lee, Jeffrey S. Vetter

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

2 Scopus citations

Abstract

The frequency of hardware errors in HPC systems continues to grow as system designs evolve toward exascale. Tolerating these errors efficiently and effectively will require software-based resilience solutions. With this requirement in mind, recent research has increasingly employed LLVM-based tools to simulate transient hardware faults in order to study the resilience characteristics of specific applications. However, such tools require researchers to configure their experiments at the level of the LLVM intermediate representation (LLVM IR) rather than at the source level of the applications under study. In this paper, we present FITL¹ (Fault-Injection Toolkit for LLVM), a set of LLVM extensions to which it is straightforward to translate source-level pragmas that specify fault injection. While we have designed FITL not to be tied to any particular compiler front end or high-level language, we also describe how we have extended our OpenARC compiler to translate a novel set of fault-injection pragmas for C to FITL. Finally, we present several resilience studies we have conducted using FITL, including a comparison with a source-level fault injector we have built as part of OpenARC.

Original language	English
Title of host publication	Proceedings of LLVM-HPC 2015
Subtitle of host publication	2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis
Publisher	Association for Computing Machinery, Inc
ISBN (Electronic)	9781450340052
DOIs	https://doi.org/10.1145/2833157.2833160
State	Published - Nov 15 2015
Event	2nd Workshop on the LLVM Compiler Infrastructure in HPC, LLVM-HPC 2015 - Austin, United States Duration: Nov 15 2015 → …

Publication series

Name	Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis

Conference

Conference	2nd Workshop on the LLVM Compiler Infrastructure in HPC, LLVM-HPC 2015
Country/Territory	United States
City	Austin
Period	11/15/15 → …

Bibliographical note

Publisher Copyright:
© 2015 ACM.

Keywords

C pragma
Compiler intermediate representation
Directive-based programming
Fault injection
LLVM
Ope-nARC
Resiliency

Access to Document

10.1145/2833157.2833160

Cite this

Denny, J. E., Lee, S., & Vetter, J. S. (2015). FITL: Extending LLVM for the translation of fault-injection directives. In Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis Article a4 (Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis). Association for Computing Machinery, Inc. https://doi.org/10.1145/2833157.2833160

Denny, Joel E. ; Lee, Seyong ; Vetter, Jeffrey S. / FITL : Extending LLVM for the translation of fault-injection directives. Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis. Association for Computing Machinery, Inc, 2015. (Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis).

@inproceedings{0e95cc6879d04ccb919bae3ef71c503e,

title = "FITL: Extending LLVM for the translation of fault-injection directives",

abstract = "The frequency of hardware errors in HPC systems continues to grow as system designs evolve toward exascale. Tolerating these errors efficiently and effectively will require software-based resilience solutions. With this requirement in mind, recent research has increasingly employed LLVM-based tools to simulate transient hardware faults in order to study the resilience characteristics of specific applications. However, such tools require researchers to configure their experiments at the level of the LLVM intermediate representation (LLVM IR) rather than at the source level of the applications under study. In this paper, we present FITL1 (Fault-Injection Toolkit for LLVM), a set of LLVM extensions to which it is straightforward to translate source-level pragmas that specify fault injection. While we have designed FITL not to be tied to any particular compiler front end or high-level language, we also describe how we have extended our OpenARC compiler to translate a novel set of fault-injection pragmas for C to FITL. Finally, we present several resilience studies we have conducted using FITL, including a comparison with a source-level fault injector we have built as part of OpenARC.",

keywords = "C pragma, Compiler intermediate representation, Directive-based programming, Fault injection, LLVM, Ope-nARC, Resiliency",

author = "Denny, {Joel E.} and Seyong Lee and Vetter, {Jeffrey S.}",

note = "Publisher Copyright: {\textcopyright} 2015 ACM.; 2nd Workshop on the LLVM Compiler Infrastructure in HPC, LLVM-HPC 2015 ; Conference date: 15-11-2015",

year = "2015",

month = nov,

day = "15",

doi = "10.1145/2833157.2833160",

language = "English",

series = "Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis",

publisher = "Association for Computing Machinery, Inc",

booktitle = "Proceedings of LLVM-HPC 2015",

}

Denny, JE , Lee, S & Vetter, JS 2015, FITL: Extending LLVM for the translation of fault-injection directives. in Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis., a4, Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis, Association for Computing Machinery, Inc, 2nd Workshop on the LLVM Compiler Infrastructure in HPC, LLVM-HPC 2015, Austin, United States, 11/15/15. https://doi.org/10.1145/2833157.2833160

FITL: Extending LLVM for the translation of fault-injection directives. / Denny, Joel E.; Lee, Seyong ; Vetter, Jeffrey S.
Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis. Association for Computing Machinery, Inc, 2015. a4 (Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis).

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

TY - GEN

T1 - FITL

T2 - 2nd Workshop on the LLVM Compiler Infrastructure in HPC, LLVM-HPC 2015

AU - Denny, Joel E.

AU - Lee, Seyong

AU - Vetter, Jeffrey S.

PY - 2015/11/15

Y1 - 2015/11/15

N2 - The frequency of hardware errors in HPC systems continues to grow as system designs evolve toward exascale. Tolerating these errors efficiently and effectively will require software-based resilience solutions. With this requirement in mind, recent research has increasingly employed LLVM-based tools to simulate transient hardware faults in order to study the resilience characteristics of specific applications. However, such tools require researchers to configure their experiments at the level of the LLVM intermediate representation (LLVM IR) rather than at the source level of the applications under study. In this paper, we present FITL1 (Fault-Injection Toolkit for LLVM), a set of LLVM extensions to which it is straightforward to translate source-level pragmas that specify fault injection. While we have designed FITL not to be tied to any particular compiler front end or high-level language, we also describe how we have extended our OpenARC compiler to translate a novel set of fault-injection pragmas for C to FITL. Finally, we present several resilience studies we have conducted using FITL, including a comparison with a source-level fault injector we have built as part of OpenARC.

AB - The frequency of hardware errors in HPC systems continues to grow as system designs evolve toward exascale. Tolerating these errors efficiently and effectively will require software-based resilience solutions. With this requirement in mind, recent research has increasingly employed LLVM-based tools to simulate transient hardware faults in order to study the resilience characteristics of specific applications. However, such tools require researchers to configure their experiments at the level of the LLVM intermediate representation (LLVM IR) rather than at the source level of the applications under study. In this paper, we present FITL1 (Fault-Injection Toolkit for LLVM), a set of LLVM extensions to which it is straightforward to translate source-level pragmas that specify fault injection. While we have designed FITL not to be tied to any particular compiler front end or high-level language, we also describe how we have extended our OpenARC compiler to translate a novel set of fault-injection pragmas for C to FITL. Finally, we present several resilience studies we have conducted using FITL, including a comparison with a source-level fault injector we have built as part of OpenARC.

KW - C pragma

KW - Compiler intermediate representation

KW - Directive-based programming

KW - Fault injection

KW - LLVM

KW - Ope-nARC

KW - Resiliency

UR - http://www.scopus.com/inward/record.url?scp=84955471442&partnerID=8YFLogxK

U2 - 10.1145/2833157.2833160

DO - 10.1145/2833157.2833160

M3 - Conference contribution

AN - SCOPUS:84955471442

T3 - Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis

BT - Proceedings of LLVM-HPC 2015

PB - Association for Computing Machinery, Inc

Y2 - 15 November 2015

ER -

Denny JE , Lee S , Vetter JS. FITL: Extending LLVM for the translation of fault-injection directives. In Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis. Association for Computing Machinery, Inc. 2015. a4. (Proceedings of LLVM-HPC 2015: 2nd Workshop on the LLVM Compiler Infrastructure in HPC - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis). doi: 10.1145/2833157.2833160

FITL: Extending LLVM for the translation of fault-injection directives

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this