[Intel-gfx] [PATCH 2/4] drm/i915/display: Fix state of PSR2 sub features

[Ville Syrjälä]

On Mon, Sep 14, 2020 at 08:56:12PM +0000, Souza, Jose wrote:
> On Mon, 2020-09-14 at 23:30 +0300, Ville Syrjälä wrote:
> > On Mon, Sep 14, 2020 at 07:57:34PM +0000, Souza, Jose wrote:
> > > On Mon, 2020-09-14 at 17:24 +0300, Ville Syrjälä wrote:
> > > > On Mon, Aug 31, 2020 at 06:09:22PM -0700, José Roberto de Souza wrote:
> > > > > In case PSR2 is disabled by debugfs dc3co_enabled and
> > > > > psr2_sel_fetch_enabled were still being set causing some code paths
> > > > > to be executed were it should not.
> > > > > We have tests for PSR1 and PSR2 so keep those features disabled when
> > > > > PSR1 is active but PSR2 is supported is important.
> > > > > 
> > > > > Cc: Gwan-gyeong Mun <
> > > > > gwan-gyeong.mun at intel.com
> > > > > 
> > > > > 
> > > > > Cc: Ville Syrjälä <
> > > > > ville.syrjala at linux.intel.com
> > > > > 
> > > > > 
> > > > > Signed-off-by: José Roberto de Souza <
> > > > > jose.souza at intel.com
> > > > > 
> > > > > 
> > > > > ---
> > > > >  drivers/gpu/drm/i915/display/intel_psr.c | 11 +++++++----
> > > > >  1 file changed, 7 insertions(+), 4 deletions(-)
> > > > > 
> > > > > diff --git a/drivers/gpu/drm/i915/display/intel_psr.c b/drivers/gpu/drm/i915/display/intel_psr.c
> > > > > index 4e09ae61d4aa..6698d0209879 100644
> > > > > --- a/drivers/gpu/drm/i915/display/intel_psr.c
> > > > > +++ b/drivers/gpu/drm/i915/display/intel_psr.c
> > > > > @@ -962,12 +962,14 @@ static void intel_psr_enable_locked(struct drm_i915_private *dev_priv,
> > > > >  	dev_priv->psr.psr2_enabled = intel_psr2_enabled(dev_priv, crtc_state);
> > > > >  	dev_priv->psr.busy_frontbuffer_bits = 0;
> > > > >  	dev_priv->psr.pipe = to_intel_crtc(crtc_state->uapi.crtc)->pipe;
> > > > > -	dev_priv->psr.dc3co_enabled = !!crtc_state->dc3co_exitline;
> > > > > +	dev_priv->psr.dc3co_enabled = !!crtc_state->dc3co_exitline &&
> > > > > +				      dev_priv->psr.psr2_enabled;
> > > > >  	dev_priv->psr.transcoder = crtc_state->cpu_transcoder;
> > > > >  	/* DC5/DC6 requires at least 6 idle frames */
> > > > >  	val = usecs_to_jiffies(intel_get_frame_time_us(crtc_state) * 6);
> > > > >  	dev_priv->psr.dc3co_exit_delay = val;
> > > > > -	dev_priv->psr.psr2_sel_fetch_enabled = crtc_state->enable_psr2_sel_fetch;
> > > > > +	dev_priv->psr.psr2_sel_fetch_enabled = crtc_state->enable_psr2_sel_fetch &&
> > > > > +					       dev_priv->psr.psr2_enabled;
> > > > >  
> > > > >  	/*
> > > > >  	 * If a PSR error happened and the driver is reloaded, the EDP_PSR_IIR
> > > > > @@ -1178,7 +1180,7 @@ void intel_psr2_program_trans_man_trk_ctl(const struct intel_crtc_state *crtc_st
> > > > >  	struct i915_psr *psr = &dev_priv->psr;
> > > > >  
> > > > >  	if (!HAS_PSR2_SEL_FETCH(dev_priv) ||
> > > > > -	    !crtc_state->enable_psr2_sel_fetch)
> > > > > +	    !dev_priv->psr.psr2_sel_fetch_enabled)
> > > > >  		return;
> > > > >  
> > > > >  	intel_de_write(dev_priv, PSR2_MAN_TRK_CTL(psr->transcoder),
> > > > > @@ -1189,8 +1191,9 @@ void intel_psr2_sel_fetch_update(struct intel_atomic_state *state,
> > > > >  				 struct intel_crtc *crtc)
> > > > >  {
> > > > >  	struct intel_crtc_state *crtc_state = intel_atomic_get_new_crtc_state(state, crtc);
> > > > > +	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
> > > > >  
> > > > > -	if (!crtc_state->enable_psr2_sel_fetch)
> > > > > +	if (!dev_priv->psr.psr2_sel_fetch_enabled)
> > > > 
> > > > This looks rather sketchy. AFAICS this gets called during atomic_check()
> > > > so looking at stuff outside the crtc state is very suspicious.
> > > 
> > > This is called after the functions that change the PSR state so no issues, also we can't really on information in CRTC state, as PSR is only enabled
> > > if supported by state, i915 PSR parameter and PSR debug fs value.
> > 
> > I see it getting called from intel_crtc_atomic_check(). Confused.
> > Am I missing some other patches?
> 
> It is set from intel_psr_disable(), intel_psr_enable() and intel_psr_update() all executed before intel_psr2_sel_fetch_update()
> 
> intel_enable_ddi()
> 	intel_enable_ddi_dp()
> 		intel_psr_enable()
> 
> intel_update_crtc() {
> 	if (!modeset) {
> 		intel_encoders_update_pipe()
> 			encoder->update_pipe() / intel_ddi_update_pipe()
> 				intel_ddi_update_pipe_dp()
> 					intel_psr_update()
> 	}
> 
> 	...
> 		
> 	skl_update_planes_on_crtc(state, crtc);
> 		intel_update_plane()
> 			plane->update_plane() / skl_update_plane()
> 				skl_program_plane()
> 					intel_psr2_sel_fetch_update()

That's not what I see at all. The only caller I see is
intel_crtc_atomic_check().


> }
> 
> 
> > 
> > > > >  		return;
> > > > >  
> > > > >  	crtc_state->psr2_man_track_ctl = PSR2_MAN_TRK_CTL_ENABLE |

And if it would be called from there then this part would be
kinda bad. We should not mutate the state during the commit phase.

> > > > > -- 
> > > > > 2.28.0
> > > > 
> > > > 
> > 
> > 

-- 
Ville Syrjälä
Intel