Hash functions for datatype signatures in MPI

Julien Langou; George Bosilca; Graham Fagg; Jack Dongarra

doi:10.1007/11557265_14

Hash functions for datatype signatures in MPI

Julien Langou, George Bosilca, Graham Fagg, Jack Dongarra

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

4 Scopus citations

Abstract

Detecting misuse of datatypes in an application code is a desirable feature for an MPI library. To support this goal we investigate the class of hash functions based on checksums to encode the type signatures of MPI datatype. The quality of these hash functions is assessed in terms of hashing, timing and comparing to other functions published for this particular problem (Gropp, 7th European PVM/MPI Users' Group Meeting, 2000) or for other applications (CRCs). In particular hash functions based on Galois Field enables good hashing, computation of the signature of unidatatype in script O sign(1) and computation of the concatenation of two datatypes in script O sign(1) additionally.

Original language	English
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages	76-83
Number of pages	8
DOIs	https://doi.org/10.1007/11557265_14
State	Published - 2005
Externally published	Yes
Event	12th European PVM/MPI Users' Group Meeting - Recent Advances in Parallel Virtual Machine and Message Passing Interface - Sorrento, Italy Duration: Sep 18 2005 → Sep 21 2005

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3666 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	12th European PVM/MPI Users' Group Meeting - Recent Advances in Parallel Virtual Machine and Message Passing Interface
Country/Territory	Italy
City	Sorrento
Period	09/18/05 → 09/21/05

Access to Document

10.1007/11557265_14

Cite this

Langou, J., Bosilca, G., Fagg, G., & Dongarra, J. (2005). Hash functions for datatype signatures in MPI. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 76-83). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3666 LNCS). https://doi.org/10.1007/11557265_14

Langou, Julien ; Bosilca, George ; Fagg, Graham et al. / Hash functions for datatype signatures in MPI. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2005. pp. 76-83 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{4da7ddcc02954aa89a16d9ce57d244dc,

title = "Hash functions for datatype signatures in MPI",

abstract = "Detecting misuse of datatypes in an application code is a desirable feature for an MPI library. To support this goal we investigate the class of hash functions based on checksums to encode the type signatures of MPI datatype. The quality of these hash functions is assessed in terms of hashing, timing and comparing to other functions published for this particular problem (Gropp, 7th European PVM/MPI Users' Group Meeting, 2000) or for other applications (CRCs). In particular hash functions based on Galois Field enables good hashing, computation of the signature of unidatatype in script O sign(1) and computation of the concatenation of two datatypes in script O sign(1) additionally.",

author = "Julien Langou and George Bosilca and Graham Fagg and Jack Dongarra",

year = "2005",

doi = "10.1007/11557265_14",

language = "English",

isbn = "3540290095",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "76--83",

booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

note = "12th European PVM/MPI Users' Group Meeting - Recent Advances in Parallel Virtual Machine and Message Passing Interface ; Conference date: 18-09-2005 Through 21-09-2005",

}

Langou, J, Bosilca, G, Fagg, G & Dongarra, J 2005, Hash functions for datatype signatures in MPI. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3666 LNCS, pp. 76-83, 12th European PVM/MPI Users' Group Meeting - Recent Advances in Parallel Virtual Machine and Message Passing Interface, Sorrento, Italy, 09/18/05. https://doi.org/10.1007/11557265_14

Hash functions for datatype signatures in MPI. / Langou, Julien; Bosilca, George; Fagg, Graham et al.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2005. p. 76-83 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3666 LNCS).

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

TY - GEN

T1 - Hash functions for datatype signatures in MPI

AU - Langou, Julien

AU - Bosilca, George

AU - Fagg, Graham

AU - Dongarra, Jack

PY - 2005

Y1 - 2005

N2 - Detecting misuse of datatypes in an application code is a desirable feature for an MPI library. To support this goal we investigate the class of hash functions based on checksums to encode the type signatures of MPI datatype. The quality of these hash functions is assessed in terms of hashing, timing and comparing to other functions published for this particular problem (Gropp, 7th European PVM/MPI Users' Group Meeting, 2000) or for other applications (CRCs). In particular hash functions based on Galois Field enables good hashing, computation of the signature of unidatatype in script O sign(1) and computation of the concatenation of two datatypes in script O sign(1) additionally.

AB - Detecting misuse of datatypes in an application code is a desirable feature for an MPI library. To support this goal we investigate the class of hash functions based on checksums to encode the type signatures of MPI datatype. The quality of these hash functions is assessed in terms of hashing, timing and comparing to other functions published for this particular problem (Gropp, 7th European PVM/MPI Users' Group Meeting, 2000) or for other applications (CRCs). In particular hash functions based on Galois Field enables good hashing, computation of the signature of unidatatype in script O sign(1) and computation of the concatenation of two datatypes in script O sign(1) additionally.

UR - http://www.scopus.com/inward/record.url?scp=33646125871&partnerID=8YFLogxK

U2 - 10.1007/11557265_14

DO - 10.1007/11557265_14

M3 - Conference contribution

AN - SCOPUS:33646125871

SN - 3540290095

SN - 9783540290094

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 76

EP - 83

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

T2 - 12th European PVM/MPI Users' Group Meeting - Recent Advances in Parallel Virtual Machine and Message Passing Interface

Y2 - 18 September 2005 through 21 September 2005

ER -

Langou J, Bosilca G, Fagg G, Dongarra J. Hash functions for datatype signatures in MPI. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2005. p. 76-83. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/11557265_14

Hash functions for datatype signatures in MPI

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this