Microprogramming a Writeable Control Memory using Very Long Instruction Word (VLIW) Compilation Techniques.

Abstract

Microprogrammed Digital Signal Processors (DSP) are frequently used as a solution to embedded processor applications. These processors utilize a control memory which permits execution of the processor's instruction set architecture (ISA). The control memory can take the form of a static, read only memory (ROM) or a dynamic, writeable control memory (WCM), or both. Microcoding the WCM permits redefining the processor's ISA and provides speedup due to its instruction level parallelism (ILP) potential. In the past, code generation efforts for microprogrammable processors focused on creating assembly and microcode as two separate steps. In this thesis, an alternative approach was chosen which combines the separate code generation steps into one automated, dual-target compilation process using the advanced techniques of VLIW compiler technology. The architecture chosen for this effort is a microprogrammable DSP being developed by Rome Labs, New York. The prototype compiler developed in this effort has demonstrated the potential for speedup of microcoded program portions over their assembly code counterparts. Therefore, the feasibility of program speedup produced by a dual-target compiler using VLIW compilation techniques has been validated.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1997
Accession Number
ADA335547

Entities

People

  • Randall S. Whitman

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Assembly Languages
  • C Programming Language
  • Compilers
  • Computer Architecture
  • Computer Programming
  • Computer Programs
  • Computer Science
  • Computers
  • Computing System Architectures
  • Decoding
  • Engineering
  • Instruction Set Architecture
  • Microcode
  • Microprogramming
  • Programming Languages
  • Software Development
  • United States

Fields of Study

  • Engineering

Readers

  • Parallel and Distributed Computing.