On Tue, Jan 18, 2022 at 12:48:37PM +0200, Stanislav Lisovskiy wrote:
> There might be various logical contructs when we might want
> to enable async flip, so lets calculate those and set this
> flag, so that there is no need in long conditions in other
> places.
>
> v2: - Set do_async_flip flag to False, if no async flip needed.
> Lets not rely that it will be 0-initialized, but set
> explicitly, so that the logic is clear as well.
>
> Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@xxxxxxxxx>
> ---
> drivers/gpu/drm/i915/display/intel_atomic_plane.c | 2 +-
> drivers/gpu/drm/i915/display/intel_display.c | 5 +++++
> drivers/gpu/drm/i915/display/intel_display_types.h | 3 +++
> 3 files changed, 9 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_atomic_plane.c b/drivers/gpu/drm/i915/display/intel_atomic_plane.c
> index d1344e9c06de..87bad665a2c8 100644
> --- a/drivers/gpu/drm/i915/display/intel_atomic_plane.c
> +++ b/drivers/gpu/drm/i915/display/intel_atomic_plane.c
> @@ -491,7 +491,7 @@ void intel_plane_update_arm(struct intel_plane *plane,
>
> trace_intel_plane_update_arm(&plane->base, crtc);
>
> - if (crtc_state->uapi.async_flip && plane->async_flip)
> + if (plane_state->do_async_flip)
> plane->async_flip(plane, crtc_state, plane_state, true);
> else
> plane->update_arm(plane, crtc_state, plane_state);
> diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
> index 0964b2403e2d..f3ce29c42bc3 100644
> --- a/drivers/gpu/drm/i915/display/intel_display.c
> +++ b/drivers/gpu/drm/i915/display/intel_display.c
> @@ -5027,6 +5027,11 @@ int intel_plane_atomic_calc_changes(const struct intel_crtc_state *old_crtc_stat
> needs_scaling(new_plane_state))))
> new_crtc_state->disable_lp_wm = true;
>
> + if (new_crtc_state->uapi.async_flip && plane->async_flip)
> + new_plane_state->do_async_flip = true;
> + else
> + new_plane_state->do_async_flip = false;
Clearing the flag should probably be in intel_plane_duplicate_state().
The decision to call intel_crtc_{enable,disable}_flip_done() should
also be based on do_async_flip, otherwise we're going to try to complete
the flip using two different mechanisms when we decide to do a sync flip
instead of an async flip.
Otherwise lgtm.
> +
> return 0;
> }
>
> diff --git a/drivers/gpu/drm/i915/display/intel_display_types.h b/drivers/gpu/drm/i915/display/intel_display_types.h
> index 41e3dd25a78f..6b107872ad39 100644
> --- a/drivers/gpu/drm/i915/display/intel_display_types.h
> +++ b/drivers/gpu/drm/i915/display/intel_display_types.h
> @@ -634,6 +634,9 @@ struct intel_plane_state {
>
> struct intel_fb_view view;
>
> + /* Indicates if async flip is required */
> + bool do_async_flip;
> +
> /* Plane pxp decryption state */
> bool decrypt;
>
> --
> 2.24.1.485.gad05a3d8e5
--
Ville Syrjälä
Intel
