Re: dominator tree

cliffc@jaberwocky.Eng.Sun.COM (Cliff Click)
15 Mar 1998 00:19:22 -0500

          From comp.compilers

Related articles
dominator tree (1998-03-05)
Re: dominator tree (1998-03-07)
Re: dominator tree (David Chase) (1998-03-07)
Re: dominator tree (1998-03-08)
Re: dominator tree (1998-03-12)
Re: dominator tree (Vugranam Sreedhar) (1998-03-12)
Re: dominator tree (Richard F. Man) (1998-03-13)
Re: dominator tree cliffc@jaberwocky.Eng.Sun.COM (1998-03-15)
Re: dominator tree (1998-03-15)
| List of all articles for this month |

From: cliffc@jaberwocky.Eng.Sun.COM (Cliff Click)
Newsgroups: comp.compilers
Date: 15 Mar 1998 00:19:22 -0500
Organization: Compilers Central
References: 98-03-029 98-03-067
Keywords: optimize, practice

David Chase writes:

> In practice, I think most people use the O(n log n) Tarjan and Lengauer
> algorithm. I implemented the faster one once O(n inverse-Ackerman(n));
> I recall that the journal article describing it contains a typo. I
> also recall that a friend of mine laughed at me for going to all the
> trouble, since log N grows slowly enough, and the nlogn algorithm is
> simpler.

I use the faster version always, because the amount of extra code to do
so is quite trivial. Its like 4 or 5 more lines of C code.


Cliff Click Compiler Designer and Researcher
cliffc at JavaSoft
(408) 863-3266 MS UCUP02-302

Post a followup to this message

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