dataflow analysis, node listings

Claus Rick <rick@puma.cs.uni-bonn.de>
17 Dec 1996 23:31:47 -0500

          From comp.compilers
Related articles
dataflow analysis, node listings rick@puma.cs.uni-bonn.de (Claus Rick) (1996-12-17)
| List of all articles for this month |
From: Claus Rick <rick@puma.cs.uni-bonn.de>
Newsgroups: comp.compilers
Date: 17 Dec 1996 23:31:47 -0500
Organization: University of Bonn
Keywords: analysis, dataflow, question

I am looking for references on the use and computation of so-called
node listings in dataflow analysis.


Node listings are used to speed up the iterative methods for
performing global dataflow analysis.


A node listing for a flow graph is an ordering of the nodes such that
every acyclic path is a subsequence thereof.


There is a rather complicated algorithm by Aho and Ullman (1976) which
computes a node listing of length at most n log n for a reducible flow
graph of n nodes. However, the node listings computed seem to be much
longer than necessary for a variety of graphs.


Are there any other algorithms for computing node listings?


Sincerely


Claus Rick


--
Claus Rick
Email: rick@cs.uni-bonn.de
WWW: http://eos.informatik.uni-bonn.de:8080/~rick
--

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.