DocumentCode :
1918881
Title :
K-scope: A Java-Based Fortran Source Code Analyzer with Graphical User Interface for Performance Improvement
Author :
Terai, Masaaki ; Murai, Hitoshi ; Minami, Kazuo ; Yokokawa, Mitsuo ; Tomiyama, Eiji
Author_Institution :
RIKEN Adv. Inst. for Comput. Sci., Kobe, Japan
fYear :
2012
fDate :
10-13 Sept. 2012
Firstpage :
434
Lastpage :
443
Abstract :
Given that scientific computer programs are becoming larger and more complicated, high performance application developers routinely examine the program structure of their source code to improve their performance. We have developed K-scope, a source code analysis tool that can be used to improve code performance. K-scope has graphical user interface that visualizes program structures of Fortran 90 and FORTRAN 77 source code and enables static data-flow analysis. To develop the tool, we adopted the filtered abstract syntax tree (filtered-AST) model with Java to visualize the program structure efficiently. Filtered-AST, which extends the AST in the structured programming model by abstract block structuring, is suitable for visualization program structures. Based on this model, K-scope has been developed as an experimental implementation. It constructs filtered-AST objects from both source and intermediate code generated by the front-end of the XcalableMP compiler. We provide illustrations of the graphical user interface and give detailed examples of the tool applied to an actual application code.
Keywords :
FORTRAN; Java; computational linguistics; data flow analysis; graphical user interfaces; natural sciences computing; program compilers; program visualisation; software performance evaluation; source coding; structured programming; tree data structures; FORTRAN 77 source code; Fortran 90 source code; Java-based Fortran source code analyzer; K-scope; XcalableMP compiler; abstract block structuring; code performance improvement; filtered abstract syntax tree model; filtered-AST model; graphical user interface; high performance application developers; program structure visualization; scientific computer programs; source code analysis tool; static data flow analysis; structured programming model; Analytical models; Computational modeling; Load modeling; Object oriented modeling; Syntactics; Visualization; XML; GUI; XcalableMP compiler; filtered-AST; single core performance improvement; static source code analysis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Parallel Processing Workshops (ICPPW), 2012 41st International Conference on
Conference_Location :
Pittsburgh, PA
ISSN :
1530-2016
Print_ISBN :
978-1-4673-2509-7
Type :
conf
DOI :
10.1109/ICPPW.2012.61
Filename :
6337511
Link To Document :
بازگشت