Re: Low hanging fruit for optimizing loops

[Marco Leise]

GCC even cosiders cache line length and CPU cache size.
This alignment improvement should consistently work on
x86/amd64.

-- 
Marco