Revert "radeon: just don't map VRAM buffers at all"
authorLeo Liu <leo.liu@amd.com>
Wed, 2 Apr 2014 13:09:32 +0000 (09:09 -0400)
committerChristian König <christian.koenig@amd.com>
Fri, 4 Apr 2014 14:21:04 +0000 (16:21 +0200)
commit0817182b2f3ec73cc79fab72f791e14d98d5306f
tree0536c31153c531be2ce0061ef5351482a8c6f6bf
parentde1a59b7a71583ae4fb09f9e4ee049895f14ca5f
Revert "radeon: just don't map VRAM buffers at all"

This reverts commit 96e8b916a7a39a9ba58e92d1ad77b5501de63ac7.
In the case of VCE encoding with raw YUV file, CPU load directly
to VRAM is faster than combination of CPU writing to GTT and
then blit to VRAM with GPU.

Reviewed-by: Christian König <christian.koenig@amd.com>
src/gallium/drivers/radeon/r600_texture.c