gpu: nvgpu: support in-kernel vidmem mappings

Propagate the buffer aperture flag in gk20a_locked_gmmu_map up so that
buffers represented as a mem_desc and present in vidmem can be mapped to
gpu.

JIRA DNVGPU-18
JIRA DNVGPU-76

Change-Id: I46cf87e27229123016727339b9349d5e2c835b3e
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: http://git-master/r/1169308
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>

1 files changed, 2 insertions, 1 deletions

diff --git a/drivers/gpu/nvgpu/gk20a/gk20a.h b/drivers/gpu/nvgpu/gk20a/gk20a.h
index 45e16ad9..b8a2fc3e 100644
--- a/drivers/gpu/nvgpu/gk20a/gk20a.h
+++ b/drivers/gpu/nvgpu/gk20a/gk20a.h
@@ -492,7 +492,8 @@ struct gpu_ops {
                                 bool clear_ctags,
                                 bool sparse,
                                 bool priv,
-                                struct vm_gk20a_mapping_batch *batch);
+                                struct vm_gk20a_mapping_batch *batch,
+                                enum gk20a_aperture aperture);
                 void (*gmmu_unmap)(struct vm_gk20a *vm,
                                 u64 vaddr,
                                 u64 size,