Also, making a copy of non-cached memory can be painful. VRAM is typically write combined.
The very advantage wfb is supposed to give is copyless access, taking that away could make it unbearable, instead of just slow. This should definately be tested first. Maarten. On Tue, Jun 9, 2009 at 6:58 PM, Maarten Maathuis<[email protected]> wrote: > Have you benchmarked this? I know it wasn't fast, but if it becomes > even (significantly) slower that's not very nice. I'm not near a > system where i can test this for the forseeable future. > > Maarten. > _______________________________________________ xorg-devel mailing list [email protected] http://lists.x.org/mailman/listinfo/xorg-devel
