On Tue, Jun 1, 2021 at 10:35 AM Daniel Thompson <daniel.thompson@xxxxxxxxxx> wrote: > IIUC if this code is prevented from progressing for 3ms or longer then > we risk deactivating the backlight instead of setting it. Does there need > to be a retrospective timeout check and recovery loop? Hm, could be another patch if we run into it. Newbyte (CC) has tested it and it works fine and stabilize his system without any problems in practice with the backlight setting. I would say if users run into this issue we can look into a timeout-and-restart loop. You can add: Tested-by: newbyte@xxxxxxxxxxx after discussion on IRQ. As this makes things not crash at least. Yours, Linus Walleij
