| | |
Summary: Dynamizing Static Algorithms, with Applications to Dynamic Trees and
History Independence
Umut A. Acar Guy E. Blelloch Robert Harper Jorge L. Vittes Shan Leung Maverick Woo
Computer Science Department, Carnegie Mellon University, Pittsburgh, PA 15213
fumut,blelloch,rwh,jvittes,maverickg@cs.cmu.edu
Abstract
We describe a machine model for automatically dy-
namizing static algorithms and apply it to history-
independent data structures. Static programs expressed
in this model are dynamized automatically by keeping
track of dependences between code and data in the form
of a dynamic dependence graph. To study the perfor-
mance of such automatically dynamized algorithms we
present an analysis technique based on trace stability.
As an example of the use of the model, we dynamize
the Parallel Tree Contraction Algorithm of Miller and
Reif to obtain a history-independent data structure for
the dynamic trees problem of Sleator and Tarjan.
1 Introduction
We describe techniques for automatically generating dy-
|
