gpu: nvgpu: Fix ordering of wmb() in gk20a_writel()

Putting the wmb() before the write only ensures that any previous
writes are done. But this doesn't really do anything for the
writel_relaxed(). The point of the wmb() here is to ensure that
the write performed by the writel_relaxed() is actually done
before proceeding.

Bug 1811382

Change-Id: I7250ea074b8548c899acfd34d816de466cf53b6f
Signed-off-by: Alex Waterman <alexw@nvidia.com>
Reviewed-on: http://git-master/r/1216434
(cherry picked from commit c9aa02dc61138615d971902fe58dc6a113cdf00a)
Reviewed-on: http://git-master/r/1223457
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>

1 files changed, 1 insertions, 1 deletions

diff --git a/drivers/gpu/nvgpu/gk20a/gk20a.h b/drivers/gpu/nvgpu/gk20a/gk20a.h
index 83d388a7..72170362 100644
--- a/drivers/gpu/nvgpu/gk20a/gk20a.h
+++ b/drivers/gpu/nvgpu/gk20a/gk20a.h
@@ -1028,8 +1028,8 @@ int gk20a_restore_registers(struct gk20a *g);
 static inline void gk20a_writel(struct gk20a *g, u32 r, u32 v)
 {
         gk20a_dbg(gpu_dbg_reg, " r=0x%x v=0x%x", r, v);
-        wmb();
         writel_relaxed(v, g->regs + r);
+        wmb();
 }
 static inline u32 gk20a_readl(struct gk20a *g, u32 r)
 {