Re: Exynos850 and ExynosAuto v9 pinctrl wakeup muxed interrupt

[Sam Protsenko]

On Sun, 16 Jan 2022 at 19:15, Krzysztof Kozlowski
<krzysztof.kozlowski@xxxxxxxxxxxxx> wrote:
>
> On 15/01/2022 21:38, Sam Protsenko wrote:
> > On Sat, 15 Jan 2022 at 17:46, Krzysztof Kozlowski
> > <krzysztof.kozlowski@xxxxxxxxxxxxx> wrote:
> >>
> >> This would mean that my dts patch removing all interrupts for alive and
> >> cmgp was correct:
> >> https://lore.kernel.org/linux-samsung-soc/66754058-187e-ffd5-71ba-4720101f5d98@xxxxxxxxxxxxx/T/#mf0b06ebdac554d57d8230dc546c3d57d59d7bd6b
> >> Was it?
> >>
> >
> > Yep. But patches [1,2] I sent recently should be probably applied
> > before yours -- they belong together. Please take those in your patch
> > series (when sending the next version).
> >
> > Thanks!
> >
> > [1] https://lkml.org/lkml/2022/1/14/861
> > [2] https://lkml.org/lkml/2022/1/3/680
> >
> >>> Will send the patch soon -- please add it to the beginning of your
> >>> series along with my other patch I already submitted.
>
> DTS and driver changes never go via same tree, so if you say there is
> such dependency, then my patch should wait till another release so your
> driver change will be in mainline.
>
> What is actually the dependency here between my patch removing incorrect
> interrupts and yours:
> 1. removing EINT for these GPIO banks from pinctrl driver,
> 2. adding missing GPIO banks?
>

No dependency really. I just assumed those can go through one tree,
and it would be nice to avoid error messages to appear between
commits. But those errors are actually already there anyway (just
masked by the code you remove), and nothing is broken functionally in
your commit. So if those patches go thru different trees -- no harm in
changing the order. I'm more concerned about time needed to get those
in mainline. So please send all the patches together. Sorry for
confusion :)

Thanks!

>
> Best regards,
> Krzysztof