Re: [PATCH] drm/panel: kingdisplay-kd097d04: Use mipi_dsi_*_multi() functions

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential
issues to consider:
- [Medium] drm/panel: state imbalance in kingdisplay_panel_unprepare()
  on DSI error
- [Medium] drm/panel: kingdisplay_panel_disable() error return leaves
  panel enabled
--

commit ce1c20733cce84cf65434f3ef6732b56a7367cda
Author: Archit More <architmore303@gmail.com>

drm/panel: kingdisplay-kd097d04: Use mipi_dsi_*_multi() functions

This commit converts the Kingdisplay KD097D04 panel driver's disable,
unprepare, and prepare callbacks to use the mipi_dsi_*_multi() helpers.

> diff --git a/drivers/gpu/drm/panel/panel-kingdisplay-kd097d04.c b/drivers/gpu/drm/panel/panel-kingdisplay-kd097d04.c
> index 893af9b167567..01f5a96f1cfbf 100644
> --- a/drivers/gpu/drm/panel/panel-kingdisplay-kd097d04.c
> +++ b/drivers/gpu/drm/panel/panel-kingdisplay-kd097d04.c
> @@ -180,28 +180,21 @@ struct kingdisplay_panel *to_kingdisplay_panel(struct drm_panel *panel)
>  static int kingdisplay_panel_disable(struct drm_panel *panel)
>  {
>  	struct kingdisplay_panel *kingdisplay = to_kingdisplay_panel(panel);
> -	int err;
> +	struct mipi_dsi_multi_context dsi_ctx = { .dsi = kingdisplay->link };
>  
> -	err = mipi_dsi_dcs_set_display_off(kingdisplay->link);
> -	if (err < 0)
> -		dev_err(panel->dev, "failed to set display off: %d\n", err);
> +	mipi_dsi_dcs_set_display_off_multi(&dsi_ctx);
>  
> -	return 0;
> +	return dsi_ctx.accum_err;
>  }

[Severity: Medium]
Should kingdisplay_panel_disable() still return 0 here?

In drm_panel_disable(), the backlight is disabled before calling
panel->funcs->disable(), and a negative return value skips clearing
panel->enabled.  If mipi_dsi_dcs_set_display_off_multi() fails,
panel->enabled remains true and the next drm_panel_enable() returns early
without re-enabling the backlight.

>  static int kingdisplay_panel_unprepare(struct drm_panel *panel)
>  {
>  	struct kingdisplay_panel *kingdisplay = to_kingdisplay_panel(panel);
> +	struct mipi_dsi_multi_context dsi_ctx = { .dsi = kingdisplay->link };
>  	int err;
>  
> -	err = mipi_dsi_dcs_enter_sleep_mode(kingdisplay->link);
> -	if (err < 0) {
> -		dev_err(panel->dev, "failed to enter sleep mode: %d\n", err);
> -		return err;
> -	}
> -
> -	/* T15: 120ms */
> -	msleep(120);
> +	mipi_dsi_dcs_enter_sleep_mode_multi(&dsi_ctx);
> +	mipi_dsi_msleep(&dsi_ctx, 120);
>  
>  	gpiod_set_value_cansleep(kingdisplay->enable_gpio, 0);
>  
> @@ -209,12 +202,13 @@ static int kingdisplay_panel_unprepare(struct drm_panel *panel)
>  	if (err < 0)
>  		return err;
>  
> -	return 0;
> +	return dsi_ctx.accum_err;
>  }

[Severity: Medium]
Does returning dsi_ctx.accum_err after disabling kingdisplay->supply leave
panel->prepared out of sync with the hardware state?

When mipi_dsi_dcs_enter_sleep_mode_multi() fails, this function still
deasserts enable_gpio and disables kingdisplay->supply, but returns a
negative error code to drm_panel_unprepare(), which then skips setting
panel->prepared = false.  The next drm_panel_prepare() sees panel->prepared
as true and skips powering the panel back on, while a subsequent
drm_panel_unprepare() calls regulator_disable() a second time and
underflows the regulator enable count.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261110082535.7998-1-architmore303@gmail.com?part=1