Paul Berry <stereotype...@gmail.com> writes: > Changes since the previous version: > > - Because of objections, I've dropped the patch "intel: Create > intel_miptree_get_region() to prepare for fast color clear." > Instead of using an accessor function to ensure that appropriate > resolves are performed before any access to > intel_mipmap_tree::region, I've simply placed the resolves at the > points where they are needed, as we do for HiZ. > > - Rebased on top of Eric's intel_miptree_blit code--this reduced the > number of places where resolves had to be performed. > > - Added a perf_debug message if the client tries to clear to a color > that isn't compatible with the fast color clear mechanism. > > - Ensured that BRW_NEW_SURFACES is flagged when a fast clear occurs, > so that the MCS buffer pointer and the fast clear color will be > properly delievered next time the surface is used. > > - Fixed some build warnings. > > - Added patch 01/12, which avoids a corner case bug that was present > in the previous series.
Other than the significant question I had on patch 6/12 about miptree sizing, Reviewed-by: Eric Anholt <e...@anholt.net>
pgpd9SngZrGlv.pgp
Description: PGP signature
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
