Le mar. 1 sept. 2026 à 12:11, Heinrich Schuchardt <[email protected]> a écrit : > > Am 1. September 2026 10:52:42 MESZ schrieb Julien Stephan > <[email protected]>: > >show_splash() computed the result of video_bmp_display() but discarded > >it and always returned 0, so a failing splash display was silently > >ignored by video_post_probe(). Return the value directly instead. > > > >This has the side effect of actually making the whole video device fail > >to probe (no display), where before the console came up fine minus the > >logo. > > I don't think we should disable video if the splash screen or the logo is not > displayed. > > I still want be able to use the U-Boot command line and the GRUB menu in this > case.
Hi Heinrich, Thanks for the feedback! I wasn't sure whether to keep the current (unintentional) behavior or honor the error path that was already there but dead. Agreed, video shouldn't be disabled just because the logo failed. What about dropping the return at caller site and only log a message? maybe use log_warning? keep log_debug ? Cheers Julien > > Best regards > > Heinrich > > > > >Signed-off-by: Julien Stephan <[email protected]> > >--- > > drivers/video/video-uclass.c | 5 +---- > > 1 file changed, 1 insertion(+), 4 deletions(-) > > > >diff --git a/drivers/video/video-uclass.c b/drivers/video/video-uclass.c > >index 228d6bacc58..de161054d52 100644 > >--- a/drivers/video/video-uclass.c > >+++ b/drivers/video/video-uclass.c > >@@ -595,11 +595,8 @@ void *video_get_u_boot_logo(void) > > static int show_splash(struct udevice *dev) > > { > > u8 *data = SPLASH_START(u_boot_logo); > >- int ret; > >- > >- ret = video_bmp_display(dev, map_to_sysmem(data), -4, 4, true); > > > >- return 0; > >+ return video_bmp_display(dev, map_to_sysmem(data), -4, 4, true); > > } > > > > int video_default_font_height(struct udevice *dev) > > >
