A dynamic data dependence analysis approach for software pipelining

Authors:
Lin Qiao;Weitong Huang;Zhizhong Tang
Affiliations:
Department of Computer Science and Technology, Tsinghua University, Beijing, P.R.China;Department of Computer Science and Technology, Tsinghua University, Beijing, P.R.China;Department of Computer Science and Technology, Tsinghua University, Beijing, P.R.China
Venue:
NPC'05 Proceedings of the 2005 IFIP international conference on Network and Parallel Computing
Year:
2005

Citing 5
Cited 2

Instruction-level parallel processing: history, overview, and perspective

The Journal of Supercomputing - Special issue on instruction-level parallelism
Run-Time Disambiguation: Coping with Statically Unpredictable Dependencies

IEEE Transactions on Computers
Single-Dimension Software Pipelining for Multi-Dimensional Loops

Proceedings of the international symposium on Code generation and optimization: feedback-directed and runtime optimization
A static data dependence analysis approach for software pipelining

NPC'05 Proceedings of the 2005 IFIP international conference on Network and Parallel Computing
Coping with data dependencies of multi-dimensional array references

NPC'05 Proceedings of the 2005 IFIP international conference on Network and Parallel Computing

A static data dependence analysis approach for software pipelining

NPC'05 Proceedings of the 2005 IFIP international conference on Network and Parallel Computing
Coping with data dependencies of multi-dimensional array references

NPC'05 Proceedings of the 2005 IFIP international conference on Network and Parallel Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a run-time pointer aliasing disambiguation method for software pipelining techniques. By combining hardware with software, the method is better than run-time checking method or run-time compensation method, which is capable of dealing with irreversible code, and has limited compensation code space without serious rerollability problem. The new method solves pointer aliasing problem efficiently and makes it possible to obtain potential instruction-level parallel speedup. In this paper instruction-level parallel speedups of the new method are analyzed in detail. Three theoretical speedups, i.e., general speedup, probabilistic speedup and mean speedup with probability, are given, which will be helpful for studying and evaluating instruction-level parallelism of the new method.