On 4/1/22 17:48, Laurent Pinchart wrote: Hi,
[...]
diff --git a/drivers/gpu/drm/panel/panel-lvds.c b/drivers/gpu/drm/panel/panel-lvds.c index 27a1c9923b09..65c6a6e9e223 100644 --- a/drivers/gpu/drm/panel/panel-lvds.c +++ b/drivers/gpu/drm/panel/panel-lvds.c @@ -30,7 +30,8 @@ struct panel_lvds { const char *label; unsigned int width; unsigned int height; - struct videomode video_mode; + struct drm_display_mode dmode;"dmode" sounds a bit weird, I would have gone for just "mode", or "display_mode", but I don't mind much.
That's how the of_get_drm_panel_display_mode() parameter is called in drivers/gpu/drm/drm_modes.c , so I'll just keep it for consistency.
[...]
- videomode_from_timing(&timing, &lvds->video_mode); - - ret = of_property_read_u32(np, "width-mm", &lvds->width); - if (ret < 0) { - dev_err(lvds->dev, "%pOF: invalid or missing %s DT property\n", - np, "width-mm"); - return -ENODEV; - } - ret = of_property_read_u32(np, "height-mm", &lvds->height); - if (ret < 0) { - dev_err(lvds->dev, "%pOF: invalid or missing %s DT property\n", - np, "height-mm"); - return -ENODEV; - }of_get_drm_panel_display_mode() doesn't consider missing width-mm or height-mm properties as an error. Should we check here that ->width_mm and ->height_mm are not 0 ?
I wonder whether we should always require valid width-mm and height-mm DT property in of_get_drm_panel_display_mode() instead.
[...]
