gpu: nvgpu: use u64 instead of dma_addr_t

The patch modifies a common dma api
(nvgpu_dma_alloc_flags_vid_at) to use u64 argument
(which is OS agnostic) instead of dma_addr_t as
the argument type which is Linux specific.

Change-Id: I74a694b08364b4d9e2826ffaf4620b113604d1cf
Signed-off-by: Sourab Gupta <sourabg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1567709
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

1 files changed, 1 insertions, 1 deletions

diff --git a/drivers/gpu/nvgpu/include/nvgpu/dma.h b/drivers/gpu/nvgpu/include/nvgpu/dma.h
index 22b40ac9..50681f8d 100644
--- a/drivers/gpu/nvgpu/include/nvgpu/dma.h
+++ b/drivers/gpu/nvgpu/include/nvgpu/dma.h
@@ -172,7 +172,7 @@ int nvgpu_dma_alloc_flags_vid(struct gk20a *g, unsigned long flags,
  *   %NVGPU_DMA_NO_KERNEL_MAPPING
  */
 int nvgpu_dma_alloc_flags_vid_at(struct gk20a *g, unsigned long flags,
-                size_t size, struct nvgpu_mem *mem, dma_addr_t at);
+                size_t size, struct nvgpu_mem *mem, u64 at);
 
 /**
  * nvgpu_dma_free - Free a DMA allocation