Re: [Intel-gfx] [PATCH v2] drm/i915/dp: Increase slice_height for DP

[Jani Nikula <jani.nikula@linux.intel.com>]

On Thu, 02 Feb 2023, Suraj Kandpal <suraj.kandpal@intel.com> wrote:
> According VDSC spec 1.2a Section 3.8 Options for Slice
> implies that 108 lines is an optimal slice height, but any
> size can be used as long as vertical active
> integer multiple and maximum vertical slice count requirements are met.
>
> Bspec: 49259
>
> Cc: Jani Nikula <jani.nikula@linux.intel.com>
> Cc: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
> Cc: Swati Sharma <swati2.sharma@intel.com>
> Signed-off-by: Suraj Kandpal <suraj.kandpal@intel.com>
>
> diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
> index 62cbab7402e9..cb4fbcd935db 100644
> --- a/drivers/gpu/drm/i915/display/intel_dp.c
> +++ b/drivers/gpu/drm/i915/display/intel_dp.c
> @@ -1415,6 +1415,30 @@ static int intel_dp_sink_dsc_version_minor(struct intel_dp *intel_dp)
>  		DP_DSC_MINOR_SHIFT;
>  }
>  
> +static int intel_dp_get_slice_height(int vactive)
> +{
> +	int slice_height;
> +
> +	/*
> +	 * VDSC1.2a spec in Section 3.8 Options for Slices implies that
> +	 * 108 lines is an optimal slice height,
> +	 * but any size can be used as long as vertical active integer
> +	 * multiple and maximum vertical slice count requirements are met.
> +	 */
> +	for (slice_height = 108; slice_height <= vactive; slice_height += 2)
> +		if (vactive % slice_height == 0)
> +			return slice_height;

I realize now that this loop will always find the slice_height as it
iterates to vactive inclusive. vactive % slice_height == 0 if
slice_height == vactive.

I did a bit of scripting on this. It seems pretty common to find a
slice_height < 200, especially for any of the more standard vactives I
checked, but maybe 8% go up to vactive/2, and 15% to vactive, for
arbitrary even vactives. But the ones that go big are typically fairly
obscure heights.

Maybe it's good enough.

> +
> +	if (vactive % 8 == 0)
> +		slice_height = 8;
> +	else if (vactive % 4 == 0)
> +		slice_height = 4;
> +	else
> +		slice_height = 2;
> +
> +	return slice_height;

Sorry to say, given what I said above, this fallback appears to be
unnecessary. Maybe just return 2 here instead of the if ladder?

BR,
Jani.


> +}
> +
>  static int intel_dp_dsc_compute_params(struct intel_encoder *encoder,
>  				       struct intel_crtc_state *crtc_state)
>  {
> @@ -1433,17 +1457,7 @@ static int intel_dp_dsc_compute_params(struct intel_encoder *encoder,
>  	vdsc_cfg->rc_model_size = DSC_RC_MODEL_SIZE_CONST;
>  	vdsc_cfg->pic_height = crtc_state->hw.adjusted_mode.crtc_vdisplay;
>  
> -	/*
> -	 * Slice Height of 8 works for all currently available panels. So start
> -	 * with that if pic_height is an integral multiple of 8. Eventually add
> -	 * logic to try multiple slice heights.
> -	 */
> -	if (vdsc_cfg->pic_height % 8 == 0)
> -		vdsc_cfg->slice_height = 8;
> -	else if (vdsc_cfg->pic_height % 4 == 0)
> -		vdsc_cfg->slice_height = 4;
> -	else
> -		vdsc_cfg->slice_height = 2;
> +	vdsc_cfg->slice_height = intel_dp_get_slice_height(vdsc_cfg->pic_height);
>  
>  	ret = intel_dsc_compute_params(crtc_state);
>  	if (ret)

-- 
Jani Nikula, Intel Open Source Graphics Center