Fast Instruction Set Simulation Using LLVM-based Dynamic Translation

Vania Joloboff 1 Xinlei Zhou Claude Helmstetter 1 Xiaopeng Gao
1 FORMES - Formal Methods for Embedded Systems
LIAMA - Laboratoire Franco-Chinois d'Informatique, d'Automatique et de Mathématiques Appliquées, Inria Paris-Rocquencourt
Abstract : In the development of embedded systems, Instruction-Set Simulators (ISS) plays an important role. When using an ISS, simulation speed is a significant issue. In this paper, we present a dynamic translation technique that uses the LLVM open-source compiler infrastructure to increase the simulation speed. Our dynamic translation technique translates hot basic blocks of the target instruction set into LLVM bitcode, and compiles LLVM bitcode into host binary code using the LLVM Just-In-Time (JIT) compiler. We have simulated the same programs using LLVM-based dynamic translation and using traditional dynamic translation to compare their performance. The experiments show that the dynamic translation based on LLVM increases simulation speed.
Keywords : LLVM Instruction Set Simulation binary dynamic translation computer architecture compilation LLVM.
Type de document :
Communication dans un congrès
International MultiConference of Engineers and Computer Scientists 2011, Mar 2011, Hong Kong, China. Springer, 2188, pp.212-216, 2011, Lecture Notes in Engineering and Computer Science
Liste complète des métadonnées
https://hal.inria.fr/hal-00646947
Contributeur : Vania Joloboff <>
Soumis le : jeudi 1 décembre 2011 - 08:22:24
Dernière modification le : mardi 17 avril 2018 - 11:26:43

Identifiants

  • HAL Id : hal-00646947, version 1

Collections

INRIA | LIAMA | CIRAD | INRA

Citation

Vania Joloboff, Xinlei Zhou, Claude Helmstetter, Xiaopeng Gao. Fast Instruction Set Simulation Using LLVM-based Dynamic Translation. International MultiConference of Engineers and Computer Scientists 2011, Mar 2011, Hong Kong, China. Springer, 2188, pp.212-216, 2011, Lecture Notes in Engineering and Computer Science. 〈hal-00646947〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

372

Contact


archive-ouverte@inria.fr