Re: [Intel-gfx] [PATCH 4/6] drm/i915: Try to use the correct power sequencer intiially on bxt/glk

["Ville Syrjälä" <ville.syrjala@linux.intel.com>]

On Thu, Nov 10, 2022 at 04:41:24PM +0200, Ville Syrjälä wrote:
> On Thu, Nov 10, 2022 at 01:56:46PM +0000, Manna, Animesh wrote:
> > 
> > 
> > > -----Original Message-----
> > > From: Ville Syrjala <ville.syrjala@linux.intel.com>
> > > Sent: Wednesday, November 9, 2022 4:47 PM
> > > To: intel-gfx@lists.freedesktop.org
> > > Cc: Manna, Animesh <animesh.manna@intel.com>
> > > Subject: [PATCH 4/6] drm/i915: Try to use the correct power sequencer intiially
> > > on bxt/glk
> > > 
> > > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > 
> > > Currently on bxt/glk we just grab the power sequencer index from the VBT data
> > > even though it may not have been parsed yet. That could lead us to using the
> > > incorrect power sequencer during the initial panel probe.
> > > 
> > > To avoid that let's try to read out the current state of the power sequencer from
> > > the hardware. Unfortunately the power sequencer no longer has anything in its
> > > registers to associate it with the port, so the best we can do is just iterate
> > > through the power sequencers and pick the first one. This should be sufficient
> > > for single panel cases.
> > > 
> > > For the dual panel cases we probably need to go back to parsing the VBT before
> > > the panel probe (and hope that panel_type=0xff is never a thing in those cases).
> > > To that end the code always prefers the VBT panel sequencer, if available.
> > > 
> > > TODO: Deal with all the modern platforms too
> > >       Maybe add checks to make sure the same power
> > >       sequencer doesn't get assigned to multiple ports?
> > > 
> > > Cc: Animesh Manna <animesh.manna@intel.com>
> > > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > ---
> > >  .../drm/i915/display/intel_display_types.h    |  8 +-
> > >  drivers/gpu/drm/i915/display/intel_panel.c    |  1 +
> > >  drivers/gpu/drm/i915/display/intel_pps.c      | 78 +++++++++++++++++--
> > >  3 files changed, 80 insertions(+), 7 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/display/intel_display_types.h
> > > b/drivers/gpu/drm/i915/display/intel_display_types.h
> > > index aec06cb24e23..25165110142b 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_display_types.h
> > > +++ b/drivers/gpu/drm/i915/display/intel_display_types.h
> > > @@ -330,7 +330,7 @@ struct intel_vbt_panel_data {
> > >  		bool present;
> > >  		bool active_low_pwm;
> > >  		u8 min_brightness;	/* min_brightness/255 of max */
> > > -		u8 controller;		/* brightness controller number */
> > > +		s8 controller;		/* brightness controller number */
> > >  		enum intel_backlight_type type;
> > >  	} backlight;
> > > 
> > > @@ -1571,9 +1571,13 @@ struct intel_pps {
> > >  	enum pipe active_pipe;
> > >  	/*
> > >  	 * Set if the sequencer may be reset due to a power transition,
> > > -	 * requiring a reinitialization. Only relevant on BXT.
> > > +	 * requiring a reinitialization. Only relevant on BXT+.
> > >  	 */
> > >  	bool pps_reset;
> > > +	/*
> > > +	 * Power sequencer index. Only relevant on BXT+.
> > > +	 */
> > > +	s8 pps_idx;
> > >  	struct edp_power_seq pps_delays;
> > >  	struct edp_power_seq bios_pps_delays;
> > >  };
> > > diff --git a/drivers/gpu/drm/i915/display/intel_panel.c
> > > b/drivers/gpu/drm/i915/display/intel_panel.c
> > > index 918b3b9d9ebe..1794e5eecf90 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_panel.c
> > > +++ b/drivers/gpu/drm/i915/display/intel_panel.c
> > > @@ -665,6 +665,7 @@ void intel_panel_init_alloc(struct intel_connector
> > > *connector)
> > >  	struct intel_panel *panel = &connector->panel;
> > > 
> > >  	connector->panel.vbt.panel_type = -1;
> > > +	connector->panel.vbt.backlight.controller = -1;
> > >  	INIT_LIST_HEAD(&panel->fixed_modes);
> > >  }
> > > 
> > > diff --git a/drivers/gpu/drm/i915/display/intel_pps.c
> > > b/drivers/gpu/drm/i915/display/intel_pps.c
> > > index 84265096f751..ff4f1def59d2 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_pps.c
> > > +++ b/drivers/gpu/drm/i915/display/intel_pps.c
> > > @@ -211,8 +211,7 @@ static int
> > >  bxt_power_sequencer_idx(struct intel_dp *intel_dp)  {
> > >  	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
> > > -	struct intel_connector *connector = intel_dp->attached_connector;
> > > -	int backlight_controller = connector->panel.vbt.backlight.controller;
> > > +	int pps_idx = intel_dp->pps.pps_idx;
> > > 
> > >  	lockdep_assert_held(&dev_priv->display.pps.mutex);
> > > 
> > > @@ -220,7 +219,7 @@ bxt_power_sequencer_idx(struct intel_dp *intel_dp)
> > >  	drm_WARN_ON(&dev_priv->drm, !intel_dp_is_edp(intel_dp));
> > > 
> > >  	if (!intel_dp->pps.pps_reset)
> > > -		return backlight_controller;
> > > +		return pps_idx;
> > > 
> > >  	intel_dp->pps.pps_reset = false;
> > > 
> > > @@ -230,7 +229,7 @@ bxt_power_sequencer_idx(struct intel_dp *intel_dp)
> > >  	 */
> > >  	pps_init_registers(intel_dp, false);
> > > 
> > > -	return backlight_controller;
> > > +	return pps_idx;
> > >  }
> > > 
> > >  typedef bool (*pps_check)(struct drm_i915_private *dev_priv, int pps_idx); @@
> > > -310,6 +309,54 @@ vlv_initial_power_sequencer_setup(struct intel_dp
> > > *intel_dp)
> > >  		    pipe_name(intel_dp->pps.pps_pipe));
> > >  }
> > > 
> > > +static int
> > > +bxt_initial_pps_idx(struct drm_i915_private *i915, pps_check check) {
> > > +	int pps_idx, pps_num = 2;
> > > +
> > > +	for (pps_idx = 0; pps_idx < pps_num; pps_idx++) {
> > > +		if (check(i915, pps_idx))
> > > +			return pps_idx;
> > > +	}
> > > +
> > > +	return -1;
> > > +};
> > > +
> > > +static void
> > > +bxt_initial_power_sequencer_setup(struct intel_dp *intel_dp) {
> > > +	struct intel_encoder *encoder = &dp_to_dig_port(intel_dp)->base;
> > > +	struct intel_connector *connector = intel_dp->attached_connector;
> > > +	struct drm_i915_private *i915 = to_i915(encoder->base.dev);
> > > +
> > > +	lockdep_assert_held(&i915->display.pps.mutex);
> > > +
> > > +	/* first ask the VBT */
> > > +	intel_dp->pps.pps_idx = connector->panel.vbt.backlight.controller;
> > > +  
> > > +	/* VBT wasn't parsed yet? pick one where the panel is on */
> > > +	if (intel_dp->pps.pps_idx < 0)
> > > +		intel_dp->pps.pps_idx = bxt_initial_pps_idx(i915,
> > > pps_has_pp_on);
> > 
> > Always we will get 0 here even if bios enabled correctly two pps instance for dual EDP.
> 
> Yeah, dual eDP isn't really 100% possible to get right here since the
> hardware doesn't tell us which port is associated with which pps.
> 
> > Can pps be mapped with port number, like pps1 for portA and pps2 for portB?
> 
> I guess we could try something like that as the initial pass. Dunno if
> that sort of behaviour is really documented anywhere though, or is it
> just assumed that VBT will tell us the mapping explicitly in dual eDP
> cases? Also it's not clear what mapping we should use for port != A|B.
> 
> Another idea I was thinking is to mark each PPS as in use once we manage
> to probe something with it, and then skip those on subsequent probes.
> But that still doesn't guarantee the mapping is correct if both PPSes
> are enable when we do the EDID read.

But that's probably the direction we should go with the pps core.
Ie. make each pps a freestanding object just representing the
corresponding hw block, and just associate each with the
appropriate eDP port. This way if we don't know any better each
port would just grab the first plausible looking pps in probe
order. Or maybe we could even plug in some kind of nop pps
instance for the initial EDID read, and only do the real
eDP<->PPS assignment once everything has been probed.

-- 
Ville Syrjälä
Intel