[Intel-gfx] [PATCH 2/2] drm/i915/psr: Disable PSR before disable pipe

[Lisovskiy, Stanislav]

On Fri, Jul 15, 2022 at 08:33:43AM +0300, Hogander, Jouni wrote:
> On Thu, 2022-07-14 at 08:07 -0700, José Roberto de Souza wrote:
> > The issue here was on for_each_intel_encoder_mask_with_psr() over the
> > new_crtc_state encoder mask, so if the CRTC was being disabled mask
> > would be zero and it would not have any chance to disable PSR.
> >
> > So here doing for_each_intel_encoder_mask_with_psr() over the
> > old_crtc_state encoder mask and then using the new_crtc_state to
> > check if PSR needs to be disabled.

Are we sure that using old_crtc_state mask is safe here?
Because currently we would be basically mixing a usage of 
encoder from old_crtc_state mask with new_crtc_state.

If you mention a specific scenario, when this happens(i.e crtc
is being disabled and new mask is 0) should we add a specific check 
instructing us to use old_crtc_state mask only?

Because currently you might be touching some other scenarios as
well, that is what I'm concerned about.


Stan

> >
> > Cc: Jouni Högander <jouni.hogander at intel.com>
> > Cc: Stanislav Lisovskiy <stanislav.lisovskiy at intel.com>
> > Signed-off-by: José Roberto de Souza <jose.souza at intel.com>
> > ---
> >  drivers/gpu/drm/i915/display/intel_psr.c | 14 ++++++++------
> >  1 file changed, 8 insertions(+), 6 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/display/intel_psr.c
> > b/drivers/gpu/drm/i915/display/intel_psr.c
> > index e6a870641cd25..98c3c8015a5c4 100644
> > --- a/drivers/gpu/drm/i915/display/intel_psr.c
> > +++ b/drivers/gpu/drm/i915/display/intel_psr.c
> > @@ -1863,7 +1863,9 @@ void intel_psr_pre_plane_update(struct
> > intel_atomic_state *state,
> >                               struct intel_crtc *crtc)
> >  {
> >       struct drm_i915_private *i915 = to_i915(state->base.dev);
> > -     const struct intel_crtc_state *crtc_state =
> > +     const struct intel_crtc_state *old_crtc_state =
> > +             intel_atomic_get_old_crtc_state(state, crtc);
> > +     const struct intel_crtc_state *new_crtc_state =
> >               intel_atomic_get_new_crtc_state(state, crtc);
> >       struct intel_encoder *encoder;
> >
> > @@ -1871,7 +1873,7 @@ void intel_psr_pre_plane_update(struct
> > intel_atomic_state *state,
> >               return;
> >
> >       for_each_intel_encoder_mask_with_psr(state->base.dev, encoder,
> > -                                          crtc_state-
> > >uapi.encoder_mask) {
> > +                                          old_crtc_state-
> > >uapi.encoder_mask) {
> 
> I would add comment here explaining why using encoder mask from
> old_crtc_state, but using new_crtc_state inside the loop. It's up to
> you:
> 
> Reviewed-by: Jouni Högander <jouni.hogander at intel.com>
> 
> >               struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
> >               struct intel_psr *psr = &intel_dp->psr;
> >               bool needs_to_disable = false;
> > @@ -1884,10 +1886,10 @@ void intel_psr_pre_plane_update(struct
> > intel_atomic_state *state,
> >                * - All planes will go inactive
> >                * - Changing between PSR versions
> >                */
> > -             needs_to_disable |=
> > intel_crtc_needs_modeset(crtc_state);
> > -             needs_to_disable |= !crtc_state->has_psr;
> > -             needs_to_disable |= !crtc_state->active_planes;
> > -             needs_to_disable |= crtc_state->has_psr2 != psr-
> > >psr2_enabled;
> > +             needs_to_disable |=
> > intel_crtc_needs_modeset(new_crtc_state);
> > +             needs_to_disable |= !new_crtc_state->has_psr;
> > +             needs_to_disable |= !new_crtc_state->active_planes;
> > +             needs_to_disable |= new_crtc_state->has_psr2 != psr-
> > >psr2_enabled;
> >
> >               if (psr->enabled && needs_to_disable)
> >                       intel_psr_disable_locked(intel_dp);
>