On 1.11.2017 15:25, Mauro Carvalho Chehab wrote: > Em Wed, 1 Nov 2017 15:09:36 +0200 > Todor Tomov <todor.tomov@xxxxxxxxxx> escreveu: > >> On 1.11.2017 15:03, Mauro Carvalho Chehab wrote: >>> Hi Todor, >>> >>> Em Wed, 1 Nov 2017 14:38:02 +0200 >>> Todor Tomov <todor.tomov@xxxxxxxxxx> escreveu: >>> >>>> Hi Mauro, >>>> >>>> Thank you for pointing to this. >>>> >>>> On 1.11.2017 14:16, Mauro Carvalho Chehab wrote: >>>>> if output->wm_num is bigger than 1, the value for reg is >>>> If output->wn_num equals 2, we handle all cases (i == 0, i == 1) and set reg properly. >>>> If output->wn_num is bigger than 2, then reg will not be initialized. However this is something that "cannot happen" and because of this the case is not handled. >>>> >>>> So I think that there is nothing wrong really but we have to do something to remove the warning. I agree with your patch, it is technically not a right value for reg but any cases in which wm_num is bigger than 2 are not supported anyway and should not happen. >>> >>> Thanks for your promptly answer. Well, if i is always at the [0..1] range, >>> then I guess the enclosed patch is actually better. >> >> I don't think that there is a lot of difference practically. If this one fixes the warning too, then it is fine for me. Thank you for working on this. > > Yes, it fixes the warning. I guess the main advantage of version 2 is > that, if it would ever be possible to have more than 2 outputs, it > should be clearer that the logic at break need changes. > > So, if ok for you, I'll stick with version 2. It would be great if > you could give your ack on that. Sure, and you can add: Acked-by: Todor Tomov <todor.tomov@xxxxxxxxxx> Thank you. Todor
