On Sat, Oct 10, 2015 at 6:12 PM, Axel Davy <axel.d...@ens.fr> wrote: > On 10/10/2015 17:49, Marek Olšák wrote: >> >> On Sat, Oct 10, 2015 at 4:15 PM, Bas Nieuwenhuizen >> <b...@basnieuwenhuizen.nl> wrote: >>> >>> Hi Marek, >>> >>> The revised series is mostly done. I wanted to do more testing and to >>> try to make sure that the added cache flushes I am doing now (a >>> CACHE_FLUSH event before a fast clear and on switching framebuffers) >>> are the minimal needed. >>> >>>> Also, it looks like we don't need DCC decompression at all, right? It >>>> might be better to get rid of it and only use the 3D engine to access >>>> DCC-encoded surfaces. >>> >>> I still use it for flush_resource. I could make this redundant by >>> sharing the DCC buffer by appending the DCC buffer to the texture >>> resource similarly to how the CMASK is appended to the resource of a >>> MSAA buffer. This has the secondary benefit of not needing to >>> reference as many resources for command submission. >> >> IIRC, flush_resource is only used for shared (scanout) surfaces where >> DCC is always disabled. >> >> Marek >> _______________________________________________ >> mesa-dev mailing list >> mesa-dev@lists.freedesktop.org >> http://lists.freedesktop.org/mailman/listinfo/mesa-dev > > I think it's not a very good idea to rely on that. > > It may be true for now, but may change in the future: > For example, perhaps some day wayland will tell egl > the app is not fullscreen and that a non-scanoutable buffer > can be used.
If that happens, we'll implement DCC sharing between processes. Marek _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
