Combining program recovery, auto-parallelisation and locality analysis for C programs on multi-processor embedded systems

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We develop a complete auto-parallelisation approach for multiple-address space digital signal processors (DSPs). It combines a pointer conversion technique with a new modulo elimination transformation. This is followed by a combined parallelisation and address resolution approach, which maps array references without introducing message-passing. Furthermore, as DSPs do not possess any cache structure, an optimisation is presented which transforms the program to both exploit remote data locality and local memory bandwidth. This parallelisation approach is applied to the DSPstone and UTDSP benchmark suites, giving an average speedup of 3.78 on a four processor analog devices TigerSHARC.
Original languageEnglish
Title of host publicationParallel Architectures and Compilation Techniques, 2003. PACT 2003. Proceedings. 12th International Conference on
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Pages104-113
Number of pages10
ISBN (Print)0-7695-2021-9
DOIs
Publication statusPublished - 1 Sep 2003

Keywords

  • C language
  • embedded systems
  • file organisation
  • multiprocessing systems
  • optimisation
  • parallel processing
  • parallelising compilers
  • program control structures
  • signal processing
  • C programs
  • DSP
  • TigerSHARC
  • UTDSP benchmark suites
  • address resolution approach
  • analog devices
  • array references mapping
  • auto-parallelisation approach
  • combined parallelisation
  • digital signal processors
  • locality analysis
  • memory bandwidth
  • modulo elimination transformation
  • multiprocessor embedded systems
  • pointer conversion technique
  • program recovery
  • remote data locality
  • Arithmetic
  • Bandwidth
  • Computer architecture
  • Costs
  • Digital signal processing
  • Digital signal processors
  • Embedded system
  • Hardware
  • Informatics
  • Signal resolution

Fingerprint Dive into the research topics of 'Combining program recovery, auto-parallelisation and locality analysis for C programs on multi-processor embedded systems'. Together they form a unique fingerprint.

Cite this