gpu: nvgpu: Do not use NO_KERNEL_MAPPING for GR

With arm64 mapping buffers to CPU is free. Remove the use of
NO_KERNEL_MAPPING.

Change-Id: Ic99ca6b7f8c698e2d3cb1d61a821939798bb237b
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1753709
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Vinod Gopalakrishnakurup <vinodg@nvidia.com>
Reviewed-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-by: Lakshmanan M <lm@nvidia.com>
Reviewed-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

1 files changed, 4 insertions, 8 deletions

diff --git a/drivers/gpu/nvgpu/gk20a/gr_gk20a.c b/drivers/gpu/nvgpu/gk20a/gr_gk20a.c
index 7c51afca..28ccb896 100644
--- a/drivers/gpu/nvgpu/gk20a/gr_gk20a.c
+++ b/drivers/gpu/nvgpu/gk20a/gr_gk20a.c
@@ -1780,8 +1780,7 @@ int gr_gk20a_update_hwpm_ctxsw_mode(struct gk20a *g,
         if (enable_hwpm_ctxsw) {
                 /* Allocate buffer if necessary */
                 if (pm_ctx->mem.gpu_va == 0) {
-                        ret = nvgpu_dma_alloc_flags_sys(g,
-                                        NVGPU_DMA_NO_KERNEL_MAPPING,
+                        ret = nvgpu_dma_alloc_sys(g,
                                         g->gr.ctx_vars.pm_ctxsw_image_size,
                                         &pm_ctx->mem);
                         if (ret) {
@@ -2530,8 +2529,7 @@ int gk20a_gr_alloc_ctx_buffer(struct gk20a *g,
         if (nvgpu_mem_is_valid(&desc->mem))
                 return 0;
 
-        err = nvgpu_dma_alloc_flags_sys(g, NVGPU_DMA_NO_KERNEL_MAPPING,
-                                    size, &desc->mem);
+        err = nvgpu_dma_alloc_sys(g, size, &desc->mem);
         if (err)
                 return err;
 
@@ -2828,9 +2826,7 @@ int gr_gk20a_alloc_gr_ctx(struct gk20a *g,
         gr->ctx_vars.buffer_size = gr->ctx_vars.golden_image_size;
         gr->ctx_vars.buffer_total_size = gr->ctx_vars.golden_image_size;
 
-        err = nvgpu_dma_alloc_flags(g, NVGPU_DMA_NO_KERNEL_MAPPING,
-                                        gr->ctx_vars.buffer_total_size,
-                                        &gr_ctx->mem);
+        err = nvgpu_dma_alloc(g, gr->ctx_vars.buffer_total_size, &gr_ctx->mem);
         if (err)
                 return err;
 
@@ -2933,7 +2929,7 @@ static int gr_gk20a_alloc_channel_patch_ctx(struct gk20a *g,
         nvgpu_log(g, gpu_dbg_info, "patch buffer size in entries: %d",
                 alloc_size);
 
-        err = nvgpu_dma_alloc_map_flags_sys(ch_vm, NVGPU_DMA_NO_KERNEL_MAPPING,
+        err = nvgpu_dma_alloc_map_sys(ch_vm,
                         alloc_size * sizeof(u32), &patch_ctx->mem);
         if (err)
                 return err;