| | |
Summary: Speculative Prefetching of Induction Pointers
Artour Stoutchinin 1 , Jose Nelson Amaral 2 , Guang R. Gao 3 , James C.
Dehnert 4 , Suneel Jain 5 , and Alban Douillet 3
1 STMicroelectronics, Grenoble, France
2 Department of Computing Science, University of Alberta,
Edmonton, AB, T6G-2E8, Canada
amaral@cs.ualberta.ca, http://www.cs.ualberta.ca/ amaral
3 Computer Architecture and Parallel System Laboratory,
University of Delaware, Newark, DE, USA
fggao, douilletg@capsl.udel.edu, http://www.capsl.udel.edu
4 Transmeta Co., Santa Clara, CA, USA, dehnert@transmeta.com
5 Hewlett-Packard Co., Cupertino, CA, USA, sjain@cup.hp.com
6 Stoutchinin, Dehnert, and Jain were at SGI when most of this research was
conducted.
Abstract. We present an automatic approach for prefetching data for
linked list data structures. The main idea is based on the observation
that linked list elements are frequently allocated at constant distance
from one another in the heap. When linked lists are traversed, a regular
pattern of memory accesses with constant stride emerges. This regularity
in the memory footprint of linked lists enables the development of a
|
