Abstract | ||
---|---|---|
The Ball-Larus path-profiling algorithm is an efficient technique to collect acyclic path frequencies of a program. However, longer paths — those extending across loop iterations — describe the runtime behaviour of programs better. We generalize the Ball-Larus profiling algorithm for profiling k-iteration paths — paths that can span up to to k iterations of a loop. We show that it is possible to number such k-iteration paths perfectly, thus allowing for an efficient profiling algorithm for such longer paths. We also describe a scheme for mixed-mode profiling: profiling different parts of a procedure with different path lengths. Experimental results show that k-iteration profiling is realistic. |
Year | DOI | Venue |
---|---|---|
2009 | 10.1109/CGO.2009.11 | Seattle, WA |
Keywords | Field | DocType |
different part,ball-larus path-profiling algorithm,ball-larus profiling algorithm,different path length,k-iteration profiling,longer path,k-iteration path,efficient profiling algorithm,acyclic path frequency,k-iteration paths,mixed-mode profiling,reverse engineering,frequency,automation,compilers,probability density function,optimization,system monitoring,data mining,approximation algorithms,clustering,computer science | Computer science,Profiling (computer programming),Reverse engineering,Parallel computing,Algorithm,System monitoring,Compiler,Real-time computing,Cluster analysis,Probability density function | Conference |
ISSN | ISBN | Citations |
2164-2397 | 978-0-7695-3576-0 | 14 |
PageRank | References | Authors |
0.70 | 4 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Subhajit Roy | 1 | 45 | 10.84 |
Y. N. Srikant | 2 | 251 | 32.04 |