Re: [Intel-gfx] [PATCH 07/11] drm/i915/display/vrr: Configure and enable VRR in modeset enable

From: "Navare, Manasi" <manasi.d.navare@intel.com>
To: Jani Nikula <jani.nikula@linux.intel.com>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [Intel-gfx] [PATCH 07/11] drm/i915/display/vrr: Configure and enable VRR in modeset enable
Date: Tue, 1 Dec 2020 14:56:02 -0800	[thread overview]
Message-ID: <20201201225602.GC22644@labuser-Z97X-UD5H> (raw)
In-Reply-To: <87imadscnr.fsf@intel.com>

On Tue, Nov 10, 2020 at 12:56:40PM +0200, Jani Nikula wrote:
> On Thu, 22 Oct 2020, Manasi Navare <manasi.d.navare@intel.com> wrote:
> > This patch computes the VRR parameters from VRR crtc states
> > and configures them in VRR registers during CRTC enable in
> > the modeset enable sequence.
> >
> > Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > Signed-off-by: Manasi Navare <manasi.d.navare@intel.com>
> > ---
> >  drivers/gpu/drm/i915/display/intel_ddi.c |  5 ++++
> >  drivers/gpu/drm/i915/display/intel_vrr.c | 38 ++++++++++++++++++++++++
> >  drivers/gpu/drm/i915/display/intel_vrr.h |  2 ++
> >  3 files changed, 45 insertions(+)
> >
> > diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c
> > index 09811be08cfe..391c51979334 100644
> > --- a/drivers/gpu/drm/i915/display/intel_ddi.c
> > +++ b/drivers/gpu/drm/i915/display/intel_ddi.c
> > @@ -49,6 +49,7 @@
> >  #include "intel_sprite.h"
> >  #include "intel_tc.h"
> >  #include "intel_vdsc.h"
> > +#include "intel_vrr.h"
> >  
> >  struct ddi_buf_trans {
> >  	u32 trans1;	/* balance leg enable, de-emph level */
> > @@ -4046,6 +4047,10 @@ static void intel_enable_ddi(struct intel_atomic_state *state,
> >  
> >  	intel_ddi_enable_transcoder_func(encoder, crtc_state);
> >  
> > +	/* Enable VRR if requested through CRTC property */
> 
> I don't think the comment is helpful, really.

Yes will remove

> 
> > +	if (crtc_state->vrr.enable)
> > +		intel_vrr_enable(encoder, crtc_state);
> 
> In the disable path you check the vrr.enable within the
> function. Perhaps we should do the same here, i.e. call vrr enable
> unconditionally and have it early return if not requested.

Okay will do

> 
> > +
> >  	intel_enable_pipe(crtc_state);
> >  
> >  	intel_crtc_vblank_on(crtc_state);
> > diff --git a/drivers/gpu/drm/i915/display/intel_vrr.c b/drivers/gpu/drm/i915/display/intel_vrr.c
> > index 56114f535f94..7f1353bac583 100644
> > --- a/drivers/gpu/drm/i915/display/intel_vrr.c
> > +++ b/drivers/gpu/drm/i915/display/intel_vrr.c
> > @@ -64,3 +64,41 @@ intel_vrr_compute_config(struct intel_dp *intel_dp,
> >  		 crtc_state->vrr.vtotalmax);
> >  }
> >  
> > +void intel_vrr_enable(struct intel_encoder *encoder,
> > +		      const struct intel_crtc_state *crtc_state)
> > +{
> > +	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
> > +	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
> > +	enum pipe pipe = crtc->pipe;
> > +	const struct drm_display_mode *adjusted_mode =
> > +		&crtc_state->hw.adjusted_mode;
> > +	u32 trans_vrr_ctl = 0, trans_vrr_vmax = 0, trans_vrr_vmin = 0, trans_vrr_flipline = 0, trans_push = 0;
> > +	u16 framestart_to_pipelinefull_linecnt = 0;
> 
> All the initializations to 0 are unnecessary.

Will remove

> 
> > +
> > +	framestart_to_pipelinefull_linecnt =
> > +		min_t(u16, 255, (crtc_state->vrr.vtotalmin - adjusted_mode->crtc_vdisplay));
> > +
> > +	trans_vrr_ctl = VRR_CTL_VRR_ENABLE |  VRR_CTL_IGN_MAX_SHIFT |
> > +		VRR_CTL_FLIP_LINE_EN | VRR_CTL_LINE_COUNT(framestart_to_pipelinefull_linecnt) |
> > +		VRR_CTL_SW_FULLLINE_COUNT;
> > +
> > +	/* Programming adjustments for 0 based regs */
> > +	trans_vrr_vmax = crtc_state->vrr.vtotalmax - 1;
> > +	trans_vrr_vmin = crtc_state->vrr.vtotalmin - 1;
> > +	trans_vrr_flipline = crtc_state->vrr.vtotalmin - 1;
> > +
> > +	trans_push = TRANS_PUSH_EN;
> 
> Frankly I'd just throw away the above four temp variables.

Okay and use them directly in de_write ?

> 
> > +
> > +	intel_de_write(dev_priv, TRANS_VRR_VMIN(pipe), trans_vrr_vmin);
> > +	intel_de_write(dev_priv, TRANS_VRR_VMAX(pipe), trans_vrr_vmax);
> > +	intel_de_write(dev_priv, TRANS_VRR_CTL(pipe), trans_vrr_ctl);
> > +	intel_de_write(dev_priv, TRANS_VRR_FLIPLINE(pipe), trans_vrr_flipline);
> > +	intel_de_write(dev_priv, TRANS_PUSH(pipe), trans_push);
> > +
> > +	drm_dbg(&dev_priv->drm,	"Enabling VRR on pipe (%c)\n", pipe_name(pipe));
> 
> drm_dbg_kms. "pipe %c" is the convention, not "pipe (%c)".

Got it will fix this

Thanks for the inputs

Manasi

> 
> > +	drm_dbg(&dev_priv->drm, "VRR Parameters: Vblank - Min = %d, Max = %d Flipline Count = %d, CTL Reg = 0x%08x, TRANS PUSH reg = 0x%08x",
> > +		crtc_state->vrr.vtotalmin, crtc_state->vrr.vtotalmax,
> > +		crtc_state->vrr.vtotalmin, trans_vrr_ctl,
> > +		trans_push);
> > +}
> > +
> > diff --git a/drivers/gpu/drm/i915/display/intel_vrr.h b/drivers/gpu/drm/i915/display/intel_vrr.h
> > index 1e6fe8fe92ec..05d982d6fbae 100644
> > --- a/drivers/gpu/drm/i915/display/intel_vrr.h
> > +++ b/drivers/gpu/drm/i915/display/intel_vrr.h
> > @@ -17,5 +17,7 @@ struct intel_dp;
> >  bool intel_is_vrr_capable(struct drm_connector *connector);
> >  void intel_vrr_compute_config(struct intel_dp *intel_dp,
> >  			      struct intel_crtc_state *crtc_state);
> > +void intel_vrr_enable(struct intel_encoder *encoder,
> > +		      const struct intel_crtc_state *crtc_state);
> >  
> >  #endif /* __INTEL_VRR_H__ */
> 
> -- 
> Jani Nikula, Intel Open Source Graphics Center
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx