Florida International University
Edit Your Profile
FIU Discovery
Toggle navigation
Browse
Home
People
Organizations
Scholarly & Creative Works
Research Facilities
Support
Edit Your Profile
McPatom: A predictive analysis tool for atomicity violation using model checking
Conference
Zeng, R, Sun, Z, Liu, S
et al
. (2012). McPatom: A predictive analysis tool for atomicity violation using model checking .
EURO-PAR 2011 PARALLEL PROCESSING, PT 1,
7385 LNCS 191-207. 10.1007/978-3-642-31759-0_14
Share this citation
Twitter
Email
Zeng, R, Sun, Z, Liu, S
et al
. (2012). McPatom: A predictive analysis tool for atomicity violation using model checking .
EURO-PAR 2011 PARALLEL PROCESSING, PT 1,
7385 LNCS 191-207. 10.1007/978-3-642-31759-0_14
Copy Citation
Share
Overview
Identifiers
Additional Document Info
View All
Overview
cited authors
Zeng, R; Sun, Z; Liu, S; He, X
authors
He, Xudong
abstract
Multi-thread programs are prone to bugs due to concurrency. Concurrency bugs are hard to find and reproduce because of the large number of interleavings. Most non-deadlock concurrency bugs are atomicity violation bugs due to unprotected accesses of shared variables by multiple threads. This paper presents a dynamic prediction tool named McPatom for predicting atomicity violation bugs involving a pair of threads accessing a shared variable using model checking. McPatom uses model checking to ensure the completeness in predicting any possible atomicity violation captured in the abstract thread model extracted from an interleaved execution. McPatom can predict atomicity violations involving more than three accesses and multiple subroutines, and supports all synchronization primitives. We have applied McPatom in predicting several known bugs in real world systems including one that evades several other existing tools. We provide evaluations of McPatom in terms of atomicity violation predictability and performance with additional improvement strategies. © 2012 Springer-Verlag Berlin Heidelberg.
publication date
August 15, 2012
published in
DISTRIBUTED COMPUTING (DISC 2014)
Book
Identifiers
Digital Object Identifier (DOI)
https://doi.org/10.1007/978-3-642-31759-0_14
Additional Document Info
start page
191
end page
207
volume
7385 LNCS