|Graph Coloring firstname.lastname@example.org (Robert Sherry) (2004-02-01)|
|Re: Graph Coloring email@example.com (2004-02-04)|
|Re: graph coloring Robert.Thorpe@antenova.com (Robert Thorpe) (2004-02-08)|
|Re: graph coloring firstname.lastname@example.org (John McEnerney) (2004-02-12)|
|Re: Graph Coloring email@example.com (TOUATI Sid) (2004-02-12)|
|graph coloring firstname.lastname@example.org (Ramesh B S) (1996-03-20)|
|Re: graph coloring email@example.com (David Gillies) (1996-03-22)|
|Re: graph coloring firstname.lastname@example.org (1996-03-25)|
|graph coloring email@example.com (1997-06-13)|
|From:||David Gillies <firstname.lastname@example.org>|
|Date:||22 Mar 1996 21:34:44 -0500|
|Organization:||Hewlett Packard California Language Lab|
Ramesh B S wrote:
> Are there any standard compilers that use any of the graph coloring
> techniques (like live range analysis, hierarchical coloring etc) to do
> register allocation?
Sure, there's lots of production compilers that use graph coloring
register allocators. I know from first hand experience that the HP
PA-Risc and IBM XL-series compilers use graph coloring techniques that
essentially boil down to Chaitin's method. However, I think gcc does
not in order to avoid any conflict with IBM's patent on the technique.
Maybe someone can confirm this.
Return to the
Search the comp.compilers archives again.