[RFC 4/4] drm/i915/display/dp: On LT failure retry LT

[Murthy, Arun R]

> -----Original Message-----
> From: Nikula, Jani <jani.nikula at intel.com>
> Sent: Tuesday, February 13, 2024 11:45 PM
> To: Murthy, Arun R <arun.r.murthy at intel.com>; intel-gfx at lists.freedesktop.org
> Cc: Deak, Imre <imre.deak at intel.com>; Syrjala, Ville <ville.syrjala at intel.com>;
> Shankar, Uma <uma.shankar at intel.com>; Murthy, Arun R
> <arun.r.murthy at intel.com>
> Subject: Re: [RFC 4/4] drm/i915/display/dp: On LT failure retry LT
> 
> On Tue, 06 Feb 2024, Arun R Murthy <arun.r.murthy at intel.com> wrote:
> > On link training failure retry link training with a lesser link
> > rate/lane count as specified in the DP spec.
> >
> > Signed-off-by: Arun R Murthy <arun.r.murthy at intel.com>
> > ---
> >  drivers/gpu/drm/i915/display/intel_ddi.c | 10 +++++++++-
> >  1 file changed, 9 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c
> > b/drivers/gpu/drm/i915/display/intel_ddi.c
> > index ed7620e7f763..29d785a4b904 100644
> > --- a/drivers/gpu/drm/i915/display/intel_ddi.c
> > +++ b/drivers/gpu/drm/i915/display/intel_ddi.c
> > @@ -2502,6 +2502,7 @@ static void mtl_ddi_pre_enable_dp(struct
> intel_atomic_state *state,
> >  				 crtc_state->port_clock,
> >  				 crtc_state->lane_count);
> >
> > +retry:
> >  	/*
> >  	 * We only configure what the register value will be here.  Actual
> >  	 * enabling happens during link training farther down.
> > @@ -2586,7 +2587,14 @@ static void mtl_ddi_pre_enable_dp(struct
> intel_atomic_state *state,
> >  	 *     Pattern, wait for 5 idle patterns (DP_TP_STATUS Min_Idles_Sent)
> >  	 *     (timeout after 800 us)
> >  	 */
> > -	intel_dp_start_link_train(intel_dp, crtc_state);
> > +	if (!intel_dp_start_link_train(intel_dp, crtc_state)) {
> > +		/* Link Training failed, retain */
> > +		intel_dp->link_trained = false;
> > +		intel_dp_stop_link_train(intel_dp, crtc_state);
> > +		encoder->post_disable(state, encoder,
> > +				   crtc_state, conn_state);
> > +		goto retry;
> > +	}
> 
> As said, the retry needs to go via userspace.

If within the supported mode range then also do we need to send uevent to user and should it come via userspace?
The fallback mandates in DP2.1 spec does this fallback in a loop.

The present fallback structure
Struct dp_fallback {
	U32 link rate;
	U8 lane_count;
	U32 resolution;
}

In the same fallback code, the present mode will be verified to see if its less than or equal to the resolution in dp_fallback. If so proceed within the fallback loop else set the max link_rate/lane count values and sent uevent.

Thanks and Regards,
Arun R Murthy
--------------------
> 
> BR,
> Jani.
> 
> 
> >
> >  	/* 6.n Set DP_TP_CTL link training to Normal */
> >  	if (!is_trans_port_sync_mode(crtc_state))
> 
> --
> Jani Nikula, Intel