gpu: nvgpu: use get_base_addr() to get inst_block address

Since inst_block could reside either in sysmem or vidmem,
use gk20a_mem_get_base_addr() to get it's base address

Jira DNVGPU-22

Change-Id: Ic9b4370e0a88b585483e78ea81df0ec6ff799487
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: http://git-master/r/1212702
(cherry picked from commit ecdffa7664f48dba0bcbd15b1340af5bf3b45802)
Reviewed-on: http://git-master/r/1219700
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/gk20a/mm_gk20a.c b/drivers/gpu/nvgpu/gk20a/mm_gk20a.c
index 1b052a20..54d85267 100644
--- a/drivers/gpu/nvgpu/gk20a/mm_gk20a.c
+++ b/drivers/gpu/nvgpu/gk20a/mm_gk20a.c
@@ -4690,7 +4690,7 @@ u64 gk20a_mm_inst_block_addr(struct gk20a *g, struct mem_desc *inst_block)
         if (g->mm.has_physical_mode)
                 addr = gk20a_mem_phys(inst_block);
         else
-                addr = gk20a_mm_smmu_vaddr_translate(g, sg_dma_address(inst_block->sgt->sgl));
+                addr = gk20a_mem_get_base_addr(g, inst_block, 0);
 
         return addr;
 }