Re: [RFC v4 5/6] drm/i915/dp: Change a link bandwidth computation for DP YCbCr 4:2:0 output

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

On Wed, Mar 06, 2019 at 09:31:01PM +0200, Gwan-gyeong Mun wrote:
> All of the link bandwidth and Data M/N calculations were assumed a bpp as
> RGB format. But When we are using YCbCr 4:2:0 output format on DP,
> we should change bpp calculations as YCbCr 4:2:0 format.
> The pipe_bpp value was assumed RGB format, therefore, it was multiplied
> with 3. But YCbCr 4:2:0 requires a multiplier value to 1.5.
> Therefore we need to divide pipe_bpp to 2 while DP output uses YCbCr4:2:0
> format.
>  - RGB format bpp = bpc x 3
>  - YCbCr 4:2:0 format bpp = bpc x 1.5
> 
> And it adds missed bpc values for a programming of VSC Header.
> It only affects dp and edp port which use YCbCr 4:2:0 output format.
> And for now, it does not consider a use case of DSC + YCbCr 4:2:0.
> 
> v2:
>   Addressed review comments from Ville.
>   Remove a changing of pipe_bpp on intel_ddi_set_pipe_settings().
>   Because the pipe is running at the full bpp, keep pipe_bpp as RGB
>   even though YCbCr 4:2:0 output format is used.
>   Add a link bandwidth computation for YCbCr4:2:0 output format.
> 
> Signed-off-by: Gwan-gyeong Mun <gwan-gyeong.mun@intel.com>
> ---
>  drivers/gpu/drm/i915/intel_dp.c | 64 +++++++++++++++++++++++++++++----
>  1 file changed, 57 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
> index 74f051428fb2..c6826cda8c5f 100644
> --- a/drivers/gpu/drm/i915/intel_dp.c
> +++ b/drivers/gpu/drm/i915/intel_dp.c
> @@ -1723,7 +1723,8 @@ void intel_dp_compute_rate(struct intel_dp *intel_dp, int port_clock,
>  struct link_config_limits {
>  	int min_clock, max_clock;
>  	int min_lane_count, max_lane_count;
> -	int min_bpp, max_bpp;
> +	int min_bpp, max_bpp, bpp_step;
> +	bool is_ycbcr420;
>  };
>  
>  static bool intel_dp_source_supports_fec(struct intel_dp *intel_dp,
> @@ -1833,7 +1834,7 @@ intel_dp_compute_link_config_wide(struct intel_dp *intel_dp,
>  	int bpp, clock, lane_count;
>  	int mode_rate, link_clock, link_avail;
>  
> -	for (bpp = limits->max_bpp; bpp >= limits->min_bpp; bpp -= 2 * 3) {
> +	for (bpp = limits->max_bpp; bpp >= limits->min_bpp; bpp -= limits->bpp_step) {
>  		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
>  						   bpp);
>  
> @@ -1847,7 +1848,10 @@ intel_dp_compute_link_config_wide(struct intel_dp *intel_dp,
>  
>  				if (mode_rate <= link_avail) {
>  					pipe_config->lane_count = lane_count;
> -					pipe_config->pipe_bpp = bpp;
> +					if (limits->is_ycbcr420)
> +						pipe_config->pipe_bpp = bpp * 2;
> +					else
> +						pipe_config->pipe_bpp = bpp;

This approach feels needlessly complicated.

Can't we just do something like this?

static int intel_dp_output_bpp(const struct intel_crtc_state *crtc_state, int bpp)
{
	if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420)
		bpp /= 2;
 	return bpp;
}

intel_dp_compute_link_config_*(struct intel_dp *intel_dp,
{
...
	for (bpp = limits->max_bpp; bpp >= limits->min_bpp; bpp -= 2 * 3) {
+		int output_bpp = intel_dp_output_bpp(pipe_config, bpp);
+
                mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
-						   bpp);
+						   output_bpp);


>  					pipe_config->port_clock = link_clock;
>  
>  					return 0;
> @@ -1869,7 +1873,7 @@ intel_dp_compute_link_config_fast(struct intel_dp *intel_dp,
>  	int bpp, clock, lane_count;
>  	int mode_rate, link_clock, link_avail;
>  
> -	for (bpp = limits->max_bpp; bpp >= limits->min_bpp; bpp -= 2 * 3) {
> +	for (bpp = limits->max_bpp; bpp >= limits->min_bpp; bpp -= limits->bpp_step) {
>  		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
>  						   bpp);
>  
> @@ -1883,7 +1887,10 @@ intel_dp_compute_link_config_fast(struct intel_dp *intel_dp,
>  
>  				if (mode_rate <= link_avail) {
>  					pipe_config->lane_count = lane_count;
> -					pipe_config->pipe_bpp = bpp;
> +					if (limits->is_ycbcr420)
> +						pipe_config->pipe_bpp = bpp * 2;
> +					else
> +						pipe_config->pipe_bpp = bpp;
>  					pipe_config->port_clock = link_clock;
>  
>  					return 0;
> @@ -2015,6 +2022,8 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
>  	struct link_config_limits limits;
>  	int common_len;
>  	int ret;
> +	bool is_ycbcr420 =
> +		pipe_config->output_format == INTEL_OUTPUT_FORMAT_YCBCR420 ? true : false;
>  
>  	common_len = intel_dp_common_len_rate_limit(intel_dp,
>  						    intel_dp->max_link_rate);
> @@ -2030,6 +2039,20 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
>  
>  	limits.min_bpp = 6 * 3;
>  	limits.max_bpp = intel_dp_compute_bpp(intel_dp, pipe_config);
> +	limits.bpp_step = 2 * 3;
> +
> +	limits.is_ycbcr420 = is_ycbcr420;
> +
> +	if (is_ycbcr420) {
> +		/*
> +		 * bpp value was assumed to RGB therefore it was multiplied
> +		 * with 3. But YCbCr 4:2:0 requires multiplier value to 1.5
> +		 * therefore it divides pipe_bpp to 2.
> +		 */
> +		limits.min_bpp /= 2;
> +		limits.max_bpp /= 2;
> +		limits.bpp_step /= 2;
> +	}
>  
>  	if (intel_dp_is_edp(intel_dp) && intel_dp->edp_dpcd[0] < DP_EDP_14) {
>  		/*
> @@ -2224,13 +2247,23 @@ intel_dp_compute_config(struct intel_encoder *encoder,
>  			intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_LIMITED;
>  	}
>  
> -	if (!pipe_config->dsc_params.compression_enable)
> -		intel_link_compute_m_n(pipe_config->pipe_bpp,
> +	if (!pipe_config->dsc_params.compression_enable) {
> +		/*
> +		 * Basically, pipe_bpp value is assumed to RGB.
> +		 * And on the calculation of Data M and Data N, YCbCr 4:2:0
> +		 * output format of the number of bytes per pixel will be half
> +		 * the number of bytes of RGB pixel.
> +		 */
> +		int pipe_bpp = pipe_config->output_format == INTEL_OUTPUT_FORMAT_YCBCR420 ?
> +			pipe_config->pipe_bpp / 2 : pipe_config->pipe_bpp;
> +
> +		intel_link_compute_m_n(pipe_bpp,
>  				       pipe_config->lane_count,
>  				       adjusted_mode->crtc_clock,
>  				       pipe_config->port_clock,
>  				       &pipe_config->dp_m_n,
>  				       constant_n);
> +	}
>  	else
>  		intel_link_compute_m_n(pipe_config->dsc_params.compressed_bpp,
>  				       pipe_config->lane_count,
> @@ -4458,6 +4491,23 @@ intel_pixel_encoding_setup_vsc(struct intel_dp *intel_dp,
>  	 * 100b = 16bpc.
>  	 */
>  	vsc_sdp.DB17 = 0x1;
> +	switch (crtc_state->pipe_bpp / 3) {
> +	case 8: /* 8bpc */
> +		vsc_sdp.DB17 = 0x1;
> +		break;
> +	case 10: /* 10bpc */
> +		vsc_sdp.DB17 = 0x2;
> +		break;
> +	case 12: /* 12bpc */
> +		vsc_sdp.DB17 = 0x3;
> +		break;
> +	case 16: /* 16bpc */
> +		vsc_sdp.DB17 = 0x4;
> +		break;
> +	default:
> +		DRM_DEBUG_KMS("Invalid bpp value '%d'\n", crtc_state->pipe_bpp);
> +		break;
> +	}
>  
>  	/*
>  	 * Content Type (Bits 2:0)
> -- 
> 2.21.0

-- 
Ville Syrjälä
Intel
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx