Re: [PATCH v2] genirq: Clarify that irq wake state is orthogonal to enable/disable

[Thomas Gleixner]

Doug Anderson <dianders@xxxxxxxxxxxx> writes:
>> Thomas also mentioned that hardware could work either way and probably
>> should not be assumed to work one way or the other.
>
> Right...
>
> ...and then (paraphrasing) Stephen pointed out that policy makes it
> really hard for clients of the API to work properly.
>
> ...and then (paraphrasing) Thomas said "Good point.  As long as you
> document that not all drivers _actually_ behave the way you describe,
> it's fine to add a comment saying that drivers _should_ behave the way
> you describe".
>
> Or, said another way: if a driver doesn't behave the way Stephen
> describes then it should be fixed unless there is some reason why
> there is no possible way to make it happen.

Yes, that's right.

Thanks,

        tglx