vc4: Simplify the load/store utile functions.
authorEric Anholt <eric@anholt.net>
Wed, 4 Jan 2017 20:40:37 +0000 (12:40 -0800)
committerEric Anholt <eric@anholt.net>
Fri, 6 Jan 2017 01:19:48 +0000 (17:19 -0800)
commite64b1169d37599a9ee1c5877aa457a41c5a8d726
tree2855721b26f67677303755f8a627e6874421b4cb
parent7b8c67b3cc60066ea55d8178a70ae035dac3dd81
vc4: Simplify the load/store utile functions.

They now have less of a dependency on the cpp, and don't have to do a
divide.

Hacking up mesa-demos teximage to do only one subtest and not draw
points, I saw 1024x1024 glTexSubImage2D() improve by 4.86939% +/-
1.40408% (n=30) and glGetTexImage() by 2.18978% +/- 0.140268% (n=5).
src/gallium/drivers/vc4/vc4_tiling.c