Re: [PATCH 2/2] drm/i915/dp: Decode DSC max delta bpp from sink DPCD

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

On 4/20/2026 4:56 PM, Nemesa Garg wrote:
> Add intel_dp_dsc_max_delta_bppx16() to parse sink DSC max
> delta bpp from DPCD when DP_DSC_MAX_BPP_DELTA_AVAILABILITY
> is set. The helper decodes RGB/YCbCr444 delta range and
> YCbCr420 delta range from DP_DSC_MAX_BPP_DELTA.


Looking from the spec, i think we need to check this support first, so 
overall to get the maximum compressed bpp supported for an output format:

Step 1: check if format specific range is given, if given use that for 
getting the max compressed bpp based on the output format. (DPCD 0x6E-6F)

Step 2: if above is not present, check for the maximum compressed bpp 
supported by sink and use that. (DPCD 0x67-68)

Step 3: if both of the above are not there, then go with the mandatory 
supported range given in the Table 2-157.


>
> Signed-off-by: Nemesa Garg <nemesa.garg@intel.com>
> ---
>   drivers/gpu/drm/i915/display/intel_dp.c | 40 +++++++++++++++++++++++++
>   1 file changed, 40 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
> index 35b8fb5740aa..7cc760aedd59 100644
> --- a/drivers/gpu/drm/i915/display/intel_dp.c
> +++ b/drivers/gpu/drm/i915/display/intel_dp.c
> @@ -2169,6 +2169,41 @@ static int dsc_compute_link_config(struct intel_dp *intel_dp,
>   	return -EINVAL;
>   }
>   
> +static u16 intel_dp_dsc_max_delta_bppx16(const struct intel_connector *connector,
> +					 enum intel_output_format output_format)
> +{
> +	const u8 *dsc_dpcd = connector->dp.dsc_dpcd;
> +
> +	if (dsc_dpcd[DP_DSC_MAX_BITS_PER_PIXEL_HI - DP_DSC_SUPPORT] &
> +	    DP_DSC_MAX_BPP_DELTA_AVAILABILITY) {


I think we can return early from here.


> +		int max_bpp_delta = 0;

Perhaps just use max_bpp.


> +
> +		switch (output_format) {
> +		case INTEL_OUTPUT_FORMAT_RGB:
> +		case INTEL_OUTPUT_FORMAT_YCBCR444:
> +			max_bpp_delta = dsc_dpcd[DP_DSC_MAX_BPP_DELTA - DP_DSC_SUPPORT] &

This can be kept in a variable to avoid computing samething again:

u8 max_bpp_delta_v1 = dsc_dpcd[DP_DSC_MAX_BPP_DELTA  - DP_DSC_SUPPORT];


> +				DP_DSC_RGB_YCbCr444_MAX_BPP_DELTA_MASK;
> +			if (max_bpp_delta >= 1 && max_bpp_delta <= 21)
> +				max_bpp_delta =  max_bpp_delta + MIN_DSC_BPP_DELTA_444 - 1;
> +			break;
> +		case INTEL_OUTPUT_FORMAT_YCBCR420:
> +			max_bpp_delta = (dsc_dpcd[DP_DSC_MAX_BPP_DELTA - DP_DSC_SUPPORT] &
> +					DP_DSC_RGB_YCbCr420_MAX_BPP_DELTA_MASK) >>
> +					BPP_DELTA_SHIFT_420;
> +			if (max_bpp_delta >= 1 && max_bpp_delta <= 7)
> +				max_bpp_delta = max_bpp_delta + MIN_DSC_BPP_DELTA_420 - 1;
> +			break;
> +		default:
> +			MISSING_CASE(output_format);
> +			return 0;
> +		}
> +
> +		return max_bpp_delta << 4;
> +	}
> +
> +	return 0;
> +}
> +
>   static
>   u16 intel_dp_dsc_max_sink_compressed_bppx16(const struct intel_connector *connector,
>   					    enum intel_output_format output_format,
> @@ -2176,6 +2211,11 @@ u16 intel_dp_dsc_max_sink_compressed_bppx16(const struct intel_connector *connec
>   {
>   	u16 max_bppx16 = drm_edp_dsc_sink_output_bpp(connector->dp.dsc_dpcd);
>   
> +	if (max_bppx16)
> +		return max_bppx16;
> +
> +	max_bppx16 = intel_dp_dsc_max_delta_bppx16(connector, output_format);


As mentioned, read this first, then fallback to the 
drm_edp_dsc_sink_output_bpp() and at last fall back to the table.

Also need to update the comment below.


Regards,

Ankit

> +
>   	if (max_bppx16)
>   		return max_bppx16;
>   	/*