Re: [PATCH] drm/i915/scaler: Prune YUV420 modes against the max source width

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] drm/i915/scaler: missing vdisplay check against scaler max_h for YUV420
--

commit 3cf07aa18b1a8b7c70fe1209cb20db9d2d636b86
Author: Arun R Murthy <arun.r.murthy@intel.com>

drm/i915/scaler: Prune YUV420 modes against the max source width

This commit fixes skl_scaler_mode_valid() to compare the mode's horizontal
resolution against the scaler's maximum width instead of the maximum height.
This prevents single-pipe YUV420 modes from being improperly pruned against
the wrong limit.

> diff --git a/drivers/gpu/drm/i915/display/skl_scaler.c b/drivers/gpu/drm/i915/display/skl_scaler.c
> index 7994b983d509d..7e65c4fffbc0c 100644
> --- a/drivers/gpu/drm/i915/display/skl_scaler.c
> +++ b/drivers/gpu/drm/i915/display/skl_scaler.c
> @@ -145,7 +145,7 @@ skl_scaler_mode_valid(struct intel_display *display,
>  
>  	if (num_joined_pipes < 2 && output_format == INTEL_OUTPUT_FORMAT_YCBCR420) {
>  		skl_scaler_max_src_size(display, &max_w, &max_h);
> -		if (mode->hdisplay > max_h)
> +		if (mode->hdisplay > max_w)
>  			return MODE_NO_420;
>  	}

[Severity: Low]
Does this code also need to validate the vertical display size? It looks like
max_h is fetched from skl_scaler_max_src_size() but never validated against
mode->vdisplay.

If a YCbCr420 mode has a vdisplay that exceeds the scaler's maximum supported
height, it appears it would pass early mode pruning here, but any attempt to
set it would fail later with -EINVAL during the atomic commit phase in
skl_update_scaler(). For example, on Gen11 platforms, max_h is capped at
4096, but intel_mode_valid_max_plane_size() permits vdisplay up to 4320.
Should a check against max_h be added here as well?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261008041809.3674538-1-arun.r.murthy@intel.com?part=1