On Mon, Jan 21, 2019 at 11:37 AM Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote: > Looks like there may still be several drivers that would trigger this. > On next-20190121: (...) > Not all of them may have multiple instances, though. > I saw patches for some of them. If I read the code right the warning will not be printed until you try to register the same irqchip twice, on the second registration. If you just register one chip which is static const, nothing happens. I think we have fixed most ... but I need to fix some more still. Yours, Linus Walleij