Re: Register Allocation via Hiearachical Graph Coloring

"Richard F. Man" <richard@imagecraft.com>
24 Sep 2004 00:22:43 -0400

          From comp.compilers

Related articles
Register Allocation via Hiearachical Graph Coloring sapna@cse.iitb.ac.in (Sapna Jain) (2004-09-21)
Re: Register Allocation via Hiearachical Graph Coloring richard@imagecraft.com (Richard F. Man) (2004-09-24)
Re: Register Allocation via Hiearachical Graph Coloring jle@ural.owlnet.rice.edu (2004-09-24)
Re: Register Allocation via Hiearachical Graph Coloring koes@cmu.edu (2004-10-02)
| List of all articles for this month |

From: "Richard F. Man" <richard@imagecraft.com>
Newsgroups: comp.compilers
Date: 24 Sep 2004 00:22:43 -0400
Organization: ImageCraft Creations Inc.
References: 04-09-129
Keywords: registers
Posted-Date: 24 Sep 2004 00:22:43 EDT

Sapna Jain wrote:


>I have read David Callahan & Brian Koblenz's paper of "Register
>Allacation via hierarchical Graph Coloring, & find that it works
>better in all condition than Chaitin's & Brigg's method.


Do you mean "it claims to work better..." rather than "[you] find that
it works better...?" If it is from experience, then you ought to know
the answers? I am almost certain that I have heard that a commercial
compiler uses this technique. Whether I heard it from one of the
numerous ex-colleagues or in a compiler conference, I do not remember.
Sorry.


--
// richard
http://www.imagecraft.com


Post a followup to this message

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