Am 28.02.24 um 06:04 schrieb Jesse.Zhang:
fix the issue when run clinfo:
"amdgpu: Failed to create process VM object".
when amdgpu initialized, seq64 do mampping and update bo mapping in vm page table.
But when clifo run. It also initializes a vm for a process device through the function kfd_process_device_init_vm
and ensure the root PD is clean through the function amdgpu_vm_pt_is_root_clean.
So they have a conflict, and clinfo always failed.
Big NAK for this, you removed the check but didn't solved the problem in
any way.
When Raven still needs the ats feature than it is intentional that this
fails.
Regards,
Christian.
Signed-off-by: Jesse Zhang <Jesse.Zhang@xxxxxxx>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ------
1 file changed, 6 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index ed4a8c5d26d7..0bc0bc75be15 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -2361,12 +2361,6 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm)
* changing any other state, in case it fails.
*/
if (pte_support_ats != vm->pte_support_ats) {
- /* Sanity checks */
- if (!amdgpu_vm_pt_is_root_clean(adev, vm)) {
- r = -EINVAL;
- goto unreserve_bo;
- }
-
vm->pte_support_ats = pte_support_ats;
r = amdgpu_vm_pt_clear(adev, vm, to_amdgpu_bo_vm(vm->root.bo),
false);
