[FLASH-USERS] Re: Patch - Multigrid relaxation operator implementation

[Colin McNally]

Hi,
	I pointed out an issue with the FLASH2.5 multigrid back in July and sent out 
a patch. Not long after then I realized that patch, while giving good 
performance, will not work in all situations. I've gotten a chance finally to 
set this right with a different tactic and further patch the underlying issue 
of block artifacts showing up in the multigrid residual. I've posted the 
whole story with explanations and tests at:

http://imp.mcmaster.ca/~colinm/gpatch/index.html

--
Colin McNally
ABB 347
Physics & Astronomy, McMaster University
colin at mcnally.ca