Re: [PATCH 03/14] drm/i915/dp: Add a helper to decide if AS SDP is needed

["Nautiyal, Ankit K" <ankit.k.nautiyal@intel.com>]

On 2/26/2026 9:31 PM, Ville Syrjälä wrote:
> On Thu, Feb 26, 2026 at 05:29:48PM +0200, Ville Syrjälä wrote:
>> On Mon, Feb 23, 2026 at 07:14:20PM +0530, Ankit Nautiyal wrote:
>>> Add a helper to check if AS SDP is needed. Currently AS SDP is only
>>> required when VRR is enabled. However, there are other use cases, along
>>> with the missing case of CMRR, dealt in subsequent changes.
>>>
>>> Signed-off-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
>>> ---
>>>   drivers/gpu/drm/i915/display/intel_dp.c | 11 ++++++++++-
>>>   1 file changed, 10 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
>>> index b999d8c085c7..65764ab0bb9b 100644
>>> --- a/drivers/gpu/drm/i915/display/intel_dp.c
>>> +++ b/drivers/gpu/drm/i915/display/intel_dp.c
>>> @@ -3111,6 +3111,12 @@ static void intel_dp_compute_vsc_colorimetry(const struct intel_crtc_state *crtc
>>>   	vsc->content_type = DP_CONTENT_TYPE_NOT_DEFINED;
>>>   }
>>>   
>>> +static bool intel_dp_needs_as_sdp(struct intel_dp *intel_dp,
>>> +				  struct intel_crtc_state *crtc_state)
>>> +{
>>> +	return crtc_state->vrr.enable;
>> Since we don't support VRR+PCON I don't think we actually
>> need the AS SDP for VRR.
> Hmm. DP 2.1a does say
> "When the plugged DP device has DPCD 02214h[0] = 1, a DP Source
>   device shall transmit an Adaptive-Sync SDP before enabling an
>   Adaptive-Sync video transmission."
>
> So maybe we do need to transmit it whenever the DP device
> supports it.

Yeah I think that will be a better way.


> In that case someone will probably need to figure
> out the fastset stuff.

Hmm.. currently we are avoiding checking of AS SDP during fastset.

The fields will change as VRR will get enabled/disabled on the fly, so 
we are avoiding check for fastset.


>
>> And I don't think we even have a fastset path for this stuff
>> (or where is it?) so isn't this already causing fastset
>> failures for VRR enable/disable? Or maybe we don't have
>> displays with both VRR and AS SDP support anywhere?

VRR with AS SDP supported panels are there, (not in CI, but some resume 
machines have them). Till now we didnt see any specific issues.

There is a test kms_vrr@flip-basic-fastset, I will re-check this on such 
a panel.


Regards,

Ankit

>>
>>> +}
>>> +
>>>   static void intel_dp_compute_as_sdp(struct intel_dp *intel_dp,
>>>   				    struct intel_crtc_state *crtc_state)
>>>   {
>>> @@ -3118,7 +3124,10 @@ static void intel_dp_compute_as_sdp(struct intel_dp *intel_dp,
>>>   	const struct drm_display_mode *adjusted_mode =
>>>   		&crtc_state->hw.adjusted_mode;
>>>   
>>> -	if (!crtc_state->vrr.enable || !intel_dp->as_sdp_supported)
>>> +	if (!intel_dp->as_sdp_supported)
>>> +		return;
>>> +
>>> +	if (!intel_dp_needs_as_sdp(intel_dp, crtc_state))
>>>   		return;
>>>   
>>>   	crtc_state->infoframes.enable |= intel_hdmi_infoframe_enable(DP_SDP_ADAPTIVE_SYNC);
>>> -- 
>>> 2.45.2
>> -- 
>> Ville Syrjälä
>> Intel