On 11/05/16 17:08, Jon Hunter wrote: > Hi Rob, > > On 11/05/16 16:51, Rob Herring wrote: >> On Sat, May 7, 2016 at 9:10 AM, Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote: >>> Hi Jon, >>> >>> On Fri, May 6, 2016 at 10:32 AM, Jon Hunter <jonathanh@xxxxxxxxxx> wrote: >>>>> The "nvidia,tegra210-agic" string can be taken as describing any >>>>> Tegra-210 specific integration quirks, though I agree that's also not >>>>> fantastic for extending PM support beyond Tegra 210 and variants >>>>> thereof. >>>>> >>>>> So maybe the best approach is bailing out in the presence of clocks >>>>> and/or power domains after all, on the assumption that nothing today has >>>>> those properties, though I fear we may have problems with that later >>>>> down the line if/when people describe those for the root GIC to describe >>>>> those must be hogged, even if not explicitly managed. >>>> >>>> On further testing, by bailing out in the presence of clocks and/or >>>> power-domains, the problem I now see is that although the primary gic-400 >>>> has been registered, we still try to probe it again later as it matches >>>> the platform driver. One way to avoid this would be ... >>>> >>>> diff --git a/drivers/of/irq.c b/drivers/of/irq.c >>>> index e7bfc175b8e1..631da7ad0dbf 100644 >>>> --- a/drivers/of/irq.c >>>> +++ b/drivers/of/irq.c >>>> @@ -556,6 +556,8 @@ void __init of_irq_init(const struct of_device_id *matches) >>>> * its children can get processed in a subsequent pass. >>>> */ >>>> list_add_tail(&desc->list, &intc_parent_list); >>>> + >>>> + of_node_set_flag(desc->dev, OF_POPULATED); >>>> } >>> >>> That sounds like the right thing to do to me... >> >> Seems fine to me, but it would be a problem since this is a global >> decision if you wanted to have some hand-off from an "early driver" to >> a platform driver. I guess setting the flag could move to drivers that >> need it although I don't think drivers should be touching the flags. > > Isn't this the other way around? Setting this flag means that I have > been populated and so don't bother creating a platform device for this > device as it isn't needed. A by-product if this, is that if we did > happen to have a platform driver for the irqchip that also has an early > driver, then the hand-off would never happen if the early init was > successful. > > The driver would still have to decide whether to hand-off and to do that > it would need to return an error from the early driver [0]. > >>>> If this is not appropriate then I guess I will just need to use >>>> "tegra210-agic" for the compatibility flag. >>> >>> As I want this for plain gic-400, I'd be unhappy ;-) >> >> IMO, the plain gic-400 should not have these dependencies and you >> should use SoC specific compatible strings should you need to deal >> with this problem. > > It is fine for my case, but it does mean I cannot say ... > > compatible = "tegra210-agic", "gic-400"; > > ... because this will always match the early driver (unless we do > something like I have suggested above). So I would have ... Sorry this is wrong. The above will always match the early driver. The problem with the above compatibility string is that, if the platform driver matches "gic-400" then it will try to probe all gic-400s even if they have been initialised early and this is definitely not what we want. This could be solved by setting the OF_POPULATED flag. Jon -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html