gpu: nvgpu: corrections to dma alloc API documentation

- Mention about NVGPU_DMA_PHYSICALLY_ADDRESSED flag and other not being
  used on safety.
- Correct the range for input size.

Jira NVGPU-6415

Change-Id: I6dc8b087edcb640a96e07c90016099cdcf456e60
Signed-off-by: Shashank Singh <shashsingh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2538377
(cherry picked from commit 5a492b0aec1a7bfe0ebf904c48814faecabef99c)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2543401
Reviewed-by: svc_kernel_abi <svc_kernel_abi@nvidia.com>
Reviewed-by: Dinesh T <dt@nvidia.com>
Reviewed-by: Vaibhav Kachore <vkachore@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This commit is contained in:
Shashank Singh
2021-06-02 09:34:44 +00:00
committed by mobile promotions
parent ee2aaef308
commit 82734765d8

View File

@@ -148,9 +148,12 @@ int nvgpu_dma_alloc_sys(struct gk20a *g, size_t size, struct nvgpu_mem *mem);
* @param g - The GPU.
* @param flags - Flags modifying the operation of the DMA allocation.
* The following flags are accepted:
* - %NVGPU_DMA_NO_KERNEL_MAPPING
* - %NVGPU_DMA_READ_ONLY
* - %NVGPU_DMA_NO_KERNEL_MAPPING (Not used on safety)
* - %NVGPU_DMA_READ_ONLY (Not used on safety)
* - %NVGPU_DMA_PHYSICALLY_ADDRESSED
* @param size - Size of the allocation in bytes.
* Range: 0 to (U64_MAX - PAGE_SIZE + 1). There is further
* restriction to upper bound on size by the OS.
* @param mem - Struct for storing the allocation information.
*
* Allocate memory suitable for doing DMA. Store the allocation info in #mem.