On Fri, 2022-03-04 at 09:59 +0000, Matthew Auld wrote:
> If the vm doesn't request async binding, like for example with the
> dpt,
> then we should be able to skip the async path and avoid calling
> i915_vm_lock_objects() altogether. Currently if we have a moving
> fence
> set for the BO(even though it might have signalled), we still take
> the
> async patch regardless of the bind_async setting, and then later
> still
> end up just doing i915_gem_object_wait_moving_fence() anyway.
>
> Alternatively we would need to add dummy scratch object which can be
> locked, just for the dpt.
>
> Suggested-by: Thomas Hellström <thomas.hellstrom@xxxxxxxxxxxxxxx>
> Signed-off-by: Matthew Auld <matthew.auld@xxxxxxxxx>
> Cc: Stanislav Lisovskiy <stanislav.lisovskiy@xxxxxxxxx>
> ---
> drivers/gpu/drm/i915/i915_vma.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_vma.c
> b/drivers/gpu/drm/i915/i915_vma.c
> index 94fcdb7bd21d..4d4d3659c938 100644
> --- a/drivers/gpu/drm/i915/i915_vma.c
> +++ b/drivers/gpu/drm/i915/i915_vma.c
> @@ -1360,8 +1360,7 @@ int i915_vma_pin_ww(struct i915_vma *vma,
> struct i915_gem_ww_ctx *ww,
> if (flags & PIN_GLOBAL)
> wakeref = intel_runtime_pm_get(&vma->vm->i915-
> >runtime_pm);
>
> - moving = vma->obj ?
Is there a chance that "moving" will be used uninitialized later?
> i915_gem_object_get_moving_fence(vma->obj) : NULL;
> - if (flags & vma->vm->bind_async_flags || moving) {
> + if (flags & vma->vm->bind_async_flags) {
> /* lock VM */
> err = i915_vm_lock_objects(vma->vm, ww);
> if (err)
> @@ -1375,6 +1374,7 @@ int i915_vma_pin_ww(struct i915_vma *vma,
> struct i915_gem_ww_ctx *ww,
>
> work->vm = i915_vm_get(vma->vm);
>
> + moving = vma->obj ?
> i915_gem_object_get_moving_fence(vma->obj) : NULL;
IIRC, with Maarten's recent changes, vma->obj is always non-NULL.
Otherwise LGTM.
Reviewed-by: Thomas Hellström <thomas.hellstrom@xxxxxxxxxxxxxxx>
> dma_fence_work_chain(&work->base, moving);
>
> /* Allocate enough page directories to used PTE */
/Thomas
