DocumentCode
3638707
Title
Exploitation of scattered context grammars to model VLIW instruction constraints
Author
J. Křoustek;S. Židek;D. Kolář;A. Meduna
Author_Institution
Department of Information Systems, Faculty of Information Technology, Brno University of Technology, Bož
fYear
2010
Firstpage
165
Lastpage
168
Abstract
More and more nowadays data processing System-on-Chip (SoC) devices exploit the very long instruction word (VLIW) technology. The high performance of VLIW processors is achieved by a high instruction level parallelism. Program execution is scheduled statically at compilation time. Therefore, there is no need for run-time control mechanisms and hardware can be relatively simple. On the other hand, all constraints checks must be done by the compiler. This paper describes formal method for modeling instruction level limitations of these processors. This method is based on scattered context grammars that generate proper assembler code. This concept has two advantages - formal description of the dependency checking process and high reduction of description complexity over other methods.
Keywords
"VLIW","Grammar","Context","Computer architecture","Program processors","Registers","Context modeling"
Publisher
ieee
Conference_Titel
Electronics Conference (BEC), 2010 12th Biennial Baltic
ISSN
1736-3705
Print_ISBN
978-1-4244-7356-4
Electronic_ISBN
2382-820X
Type
conf
DOI
10.1109/BEC.2010.5630284
Filename
5630284
Link To Document