gpu: nvgpu: Update gk20a pde bit coverage function

The mm_gk20a.c function that returns number of bits that a PDE covers
is very useful for determing PDE size for all chips. Copy this into
the common VM code since this applies to all chips/platforms.

Bug 200105199

Change-Id: I437da4781be2fa7c540abe52b20f4c4321f6c649
Signed-off-by: Alex Waterman <alexw@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1639730
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

1 files changed, 2 insertions, 0 deletions

diff --git a/drivers/gpu/nvgpu/include/nvgpu/vm.h b/drivers/gpu/nvgpu/include/nvgpu/vm.h
index 1689444c..e5d0e197 100644
--- a/drivers/gpu/nvgpu/include/nvgpu/vm.h
+++ b/drivers/gpu/nvgpu/include/nvgpu/vm.h
@@ -218,6 +218,8 @@ void nvgpu_vm_put(struct vm_gk20a *vm);
 int vm_aspace_id(struct vm_gk20a *vm);
 int nvgpu_big_pages_possible(struct vm_gk20a *vm, u64 base, u64 size);
 
+int nvgpu_vm_pde_coverage_bit_count(struct vm_gk20a *vm);
+
 /* batching eliminates redundant cache flushes and invalidates */
 void nvgpu_vm_mapping_batch_start(struct vm_gk20a_mapping_batch *batch);
 void nvgpu_vm_mapping_batch_finish(