Re: [PATCH 1/2] drm/i915/dpcd_bl: Skip testing control capability with force DPCD quirk - Kai-Heng Feng

From: Kai-Heng Feng <kai.heng.feng@canonical.com>
To: lyude@redhat.com
Cc: "Satadru Pramanik" <satadru@gmail.com>,
	"Jani Nikula" <jani.nikula@linux.intel.com>,
	"Joonas Lahtinen" <joonas.lahtinen@linux.intel.com>,
	"Rodrigo Vivi" <rodrigo.vivi@intel.com>,
	"David Airlie" <airlied@linux.ie>,
	"Daniel Vetter" <daniel@ffwll.ch>,
	"Ville Syrjälä" <ville.syrjala@linux.intel.com>,
	"Juha-Pekka Heikkila" <juhapekka.heikkila@gmail.com>,
	intel-gfx@lists.freedesktop.org,
	"open list:DRM DRIVERS" <dri-devel@lists.freedesktop.org>,
	"open list" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 1/2] drm/i915/dpcd_bl: Skip testing control capability with force DPCD quirk
Date: Thu, 8 Oct 2020 10:32:09 +0800	[thread overview]
Message-ID: <D18BA369-25A0-4D5E-A171-9EACF7DC950B@canonical.com> (raw)
In-Reply-To: <268f495fbb7e3042eb613398a8513a83d28d3fd9.camel@redhat.com>

Hi Lyude,

> On Oct 8, 2020, at 05:53, Lyude Paul <lyude@redhat.com> wrote:
> 
> Hi! I thought this patch rang a bell, we actually already had some discussion
> about this since there's a couple of other systems this was causing issues for.
> Unfortunately it never seems like that patch got sent out. Satadru?
> 
> (if I don't hear back from them soon, I'll just send out a patch for this
> myself)
> 
> JFYI - the proper fix here is to just drop the
> DP_EDP_BACKLIGHT_BRIGHTNESS_PWM_PIN_CAP check from the code entirely. As long as
> the backlight supports AUX_SET_CAP, that should be enough for us to control it.

Does the proper fix include dropping DP_QUIRK_FORCE_DPCD_BACKLIGHT entirely?

Kai-Heng

> 
> 
> On Wed, 2020-10-07 at 14:58 +0800, Kai-Heng Feng wrote:
>> HP DreamColor panel needs to be controlled via AUX interface. However,
>> it has both DP_EDP_BACKLIGHT_BRIGHTNESS_AUX_SET_CAP and
>> DP_EDP_BACKLIGHT_BRIGHTNESS_PWM_PIN_CAP set, so it fails to pass
>> intel_dp_aux_display_control_capable() test.
>> 
>> Skip the test if the panel has force DPCD quirk.
>> 
>> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
>> ---
>> drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c | 10 ++++++----
>> 1 file changed, 6 insertions(+), 4 deletions(-)
>> 
>> diff --git a/drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c
>> b/drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c
>> index acbd7eb66cbe..acf2e1c65290 100644
>> --- a/drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c
>> +++ b/drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c
>> @@ -347,9 +347,13 @@ int intel_dp_aux_init_backlight_funcs(struct
>> intel_connector *intel_connector)
>> 	struct intel_panel *panel = &intel_connector->panel;
>> 	struct intel_dp *intel_dp = enc_to_intel_dp(intel_connector->encoder);
>> 	struct drm_i915_private *i915 = dp_to_i915(intel_dp);
>> +	bool force_dpcd;
>> +
>> +	force_dpcd = drm_dp_has_quirk(&intel_dp->desc, intel_dp->edid_quirks,
>> +				      DP_QUIRK_FORCE_DPCD_BACKLIGHT);
>> 
>> 	if (i915->params.enable_dpcd_backlight == 0 ||
>> -	    !intel_dp_aux_display_control_capable(intel_connector))
>> +	    (!force_dpcd &&
>> !intel_dp_aux_display_control_capable(intel_connector)))
>> 		return -ENODEV;
>> 
>> 	/*
>> @@ -358,9 +362,7 @@ int intel_dp_aux_init_backlight_funcs(struct
>> intel_connector *intel_connector)
>> 	 */
>> 	if (i915->vbt.backlight.type !=
>> 	    INTEL_BACKLIGHT_VESA_EDP_AUX_INTERFACE &&
>> -	    i915->params.enable_dpcd_backlight != 1 &&
>> -	    !drm_dp_has_quirk(&intel_dp->desc, intel_dp->edid_quirks,
>> -			      DP_QUIRK_FORCE_DPCD_BACKLIGHT)) {
>> +	    i915->params.enable_dpcd_backlight != 1 && !force_dpcd) {
>> 		drm_info(&i915->drm,
>> 			 "Panel advertises DPCD backlight support, but "
>> 			 "VBT disagrees. If your backlight controls "
> -- 
> Sincerely,
>      Lyude Paul (she/her)
>      Software Engineer at Red Hat