On Wed, Feb 03, 2016 at 06:23:13PM +0100, Maarten Lankhorst wrote:
> Op 03-02-16 om 17:07 schreef Ville Syrjälä:
> > On Wed, Feb 03, 2016 at 04:53:24PM +0100, Maarten Lankhorst wrote:
> >> This can be derived from the atomic state in pre_plane_update,
> >> which makes it more clear when it's supposed to be called.
> >>
> >> Reviewed-by: Ander Conselvan de Oliveira <conselvan2@xxxxxxxxx>
> >> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@xxxxxxxxxxxxxxx>
> >> ---
> >> drivers/gpu/drm/i915/intel_display.c | 25 +++++++++++++++++++------
> >> drivers/gpu/drm/i915/intel_drv.h | 1 -
> >> 2 files changed, 19 insertions(+), 7 deletions(-)
> >>
> >> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> >> index de94ef165fbe..c7f2a18ab34e 100644
> >> --- a/drivers/gpu/drm/i915/intel_display.c
> >> +++ b/drivers/gpu/drm/i915/intel_display.c
> >> @@ -4807,19 +4807,33 @@ static void intel_post_plane_update(struct intel_crtc *crtc)
> >> memset(atomic, 0, sizeof(*atomic));
> >> }
> >>
> >> -static void intel_pre_plane_update(struct intel_crtc *crtc)
> >> +static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state)
> >> {
> >> + struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
> >> struct drm_device *dev = crtc->base.dev;
> >> struct drm_i915_private *dev_priv = dev->dev_private;
> >> struct intel_crtc_atomic_commit *atomic = &crtc->atomic;
> >> struct intel_crtc_state *pipe_config =
> >> to_intel_crtc_state(crtc->base.state);
> >> + struct drm_atomic_state *old_state = old_crtc_state->base.state;
> >> + struct drm_plane *primary = crtc->base.primary;
> >> + struct drm_plane_state *old_pri_state =
> >> + drm_atomic_get_existing_plane_state(old_state, primary);
> >> + bool modeset = needs_modeset(&pipe_config->base);
> >>
> >> if (atomic->update_fbc)
> >> intel_fbc_pre_update(crtc);
> >>
> >> - if (atomic->pre_disable_primary)
> >> - intel_pre_disable_primary(&crtc->base);
> >> + if (old_pri_state) {
> > When might we not have and old state for the primary plane?
> >
> When updating some state unrelated to the primary plane, for example changing cursor.
Hmm. I thought we added all the planes to the state currently...
I guess that only happens due to watermark calculation on specific
platforms.
>
> modeset and update_pipe should always add all planes.
>
> ~Maarten
--
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
