gpu: nvgpu: Use pgsz_idx instead of page_size

Alloc space writes the page size to a field that requires pgsz_idx.
That can cause corruption in internal kernel structures.

Clear_sparse treated a parameter as page size instead of index.

Bug 1549451

Change-Id: I73ce17b99aae6865056facce72d2ab9ca8b3f81d
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/495692

1 files changed, 3 insertions, 11 deletions

diff --git a/drivers/gpu/nvgpu/gm20b/mm_gm20b.c b/drivers/gpu/nvgpu/gm20b/mm_gm20b.c
index 0f9749f7..afb04cae 100644
--- a/drivers/gpu/nvgpu/gm20b/mm_gm20b.c
+++ b/drivers/gpu/nvgpu/gm20b/mm_gm20b.c
@@ -272,19 +272,11 @@ fail:
 }
 
 void gm20b_vm_clear_sparse(struct vm_gk20a *vm, u64 vaddr,
-                               u64 size, u32 pgsz) {
-        int pgsz_idx;
+                               u64 size, u32 pgsz_idx) {
         u64 vaddr_hi;
         u32 pde_lo, pde_hi, pde_i;
 
         gk20a_dbg_fn("");
-        /* determine pagesz idx */
-        for (pgsz_idx = gmmu_page_size_small;
-             pgsz_idx < gmmu_nr_page_sizes;
-             pgsz_idx++) {
-                if (gmmu_page_sizes[pgsz_idx] == pgsz)
-                        break;
-        }
         vaddr_hi = vaddr + size - 1;
         pde_range_from_vaddr_range(vm,
                                    vaddr,
@@ -292,8 +284,8 @@ void gm20b_vm_clear_sparse(struct vm_gk20a *vm, u64 vaddr,
                                    &pde_lo, &pde_hi);
 
         gk20a_dbg_info("vaddr: 0x%llx, vaddr_hi: 0x%llx, pde_lo: 0x%x, "
-                        "pde_hi: 0x%x, pgsz: %d, pde_stride_shift: %d",
-                        vaddr, vaddr_hi, pde_lo, pde_hi, pgsz,
+                        "pde_hi: 0x%x, pgsz_idx: %d, pde_stride_shift: %d",
+                        vaddr, vaddr_hi, pde_lo, pde_hi, pgsz_idx,
                         vm->mm->pde_stride_shift);
 
         for (pde_i = pde_lo; pde_i <= pde_hi; pde_i++) {