Konsta Holtta 0bb1f45958 gpu: nvgpu: optimize barrier in batch pramin writes ...

Move wmb() before the loop in pramin-accessed batch writes and use
writel_relaxed() directly, instead of calling gk20a_writel() that would
do wmb() on each iteration separately.

Jira DNVGPU-24

Change-Id: I4c1375a819266727f97e2f109d3132b5b0974ac6
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: http://git-master/r/1213600
(cherry picked from commit 79e3e38e0c5384ababfd55b8e6cd9723eb8f7b66)
Reviewed-on: http://git-master/r/1184343
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

2016-10-06 23:34:57 -07:00

..

nvgpu

gpu: nvgpu: optimize barrier in batch pramin writes

2016-10-06 23:34:57 -07:00