Re: [PATCH v9 02/14] drm/msm/dp: drop deprecated .mode_set() and use .atomic_pre_enable

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] msm_dp_bridge_mode_valid() evaluates wide bus availability using the previously committed mode's YUV420 state rather than the state of the mode being validated. This causes valid high-bandwidth RGB modes to be incorrectly rejected if a YUV420 mode was previously committed.
--

--- Patch [2]: [PATCH v9 02/14] drm/msm/dp: drop deprecated .mode_set() and use .atomic_pre_enable ---
commit 9ae2d320a8d0fa2507140b418f1dff372f81454c
Author: Yongxing Mou <yongxing.mou@oss.qualcomm.com>

drm/msm/dp: drop deprecated .mode_set() and use .atomic_pre_enable
    
The bridge .mode_set() callback is deprecated. Remove it and move the
mode setup logic to .atomic_pre_enable(), where the adjusted_mode is
available from the atomic CRTC state. Drop msm_dp_mode from
msm_dp_display_private and store the mode directly in the panel.

> diff --git a/drivers/gpu/drm/msm/dp/dp_display.c b/drivers/gpu/drm/msm/dp/dp_display.c
> index bce165bac106b..c106aabda588e 100644
> --- a/drivers/gpu/drm/msm/dp/dp_display.c
> +++ b/drivers/gpu/drm/msm/dp/dp_display.c
[ ... ]
> @@ -1309,7 +1325,7 @@ bool msm_dp_wide_bus_available(const struct msm_dp *msm_dp_display)
>  
>  	dp = container_of(msm_dp_display, struct msm_dp_display_private, msm_dp_display);
>  
> -	if (dp->msm_dp_mode.out_fmt_is_yuv_420)
> +	if (dp->panel->msm_dp_mode.out_fmt_is_yuv_420)

[Severity: High]
This is a pre-existing issue, but does evaluating wide bus availability using
the previously committed mode's YUV420 state cause valid high-bandwidth RGB
modes to be incorrectly rejected?

When a user sets a YUV420 mode, out_fmt_is_yuv_420 is set to true during
atomic_pre_enable. If the driver later probes or validates modes from EDID,
msm_dp_bridge_mode_valid() calls msm_dp_wide_bus_available():

msm_dp_bridge_mode_valid()
    if (msm_dp_wide_bus_available(dp))
        mode_pclk_khz /= 2;

This function then reads the stale out_fmt_is_yuv_420 flag and returns false.
Consequently, the pixel clock for high-bandwidth RGB modes is not halved,
causing them to exceed the maximum pixel clock and be rejected.

While this patch addresses a similar state leak for the DPU encoder by moving
the mode state update to atomic_pre_enable, it seems mode validation (which
runs before atomic_pre_enable) still relies on this stale state.

>  		return false;
>  
>  	return dp->wide_bus_supported;

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260728-dp_mstclean-v9-0-f7779fce10f4@oss.qualcomm.com?part=2