Re: Using irq-crossbar.c

[Sebastian Frias <sf84@laposte.net>]

Hi Marc,

Thanks for your input, please find my comments below.

On 06/11/2016 11:58 AM, Marc Zyngier wrote:
>> I think Sebastian is even more baffled by the DT mess
>> (sorry, intricacies) than I am.
> 
> This mess is what has saved us from the apocalypse 5 years ago, and
> describing a complex system is not easy (what a surprise...). If you
> just want to apply recipes without understanding the underlying
> constraints, you're in for a lot of pain.

Nobody said we wanted a recipe, but it would help to have examples of each of the configurations supported.
Indeed, there is effort in writing the DT bindings and even Documentation/IRQ-domain.txt yet I did not see example drivers for the different configurations.
Also, it is difficult to know what is the recommended way, since old APIs are kept for a while.

APIs are usually extended/designed around specific cases, so there must be examples, but they are not easy to find.

> 
>> The base file he was referring to is:
>>
>> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/arm/boot/dts/tango4-common.dtsi
> 
> I know which file that is, it is mentioned in the diff. I was merely
> trying to point out the glaring mistakes that could be enough for a
> interrupt controller hierarchy to be completely non-functional:
> 
> - Your crossbar doesn't have a #interrupt-cells property. How do you
>   expect the interrupt specifiers to be interpreted?

Thanks for the pointer, actually, after adding "#interrupt-cells = <3>;" I can see the driver getting requests for domain maps and xlates :-)

> - You've changed the default interrupt controller to be your crossbar.
>   Which means that all the sub-nodes are inheriting it. Have you
>   checked that this was valid for all of these nodes?

What do you mean with "valid for all these nodes"? All HW irq lines go to the crossbar.

> 
> And as it has been pointed out before, you seem to be reusing an
> existing driver. Do you know for sure that the usage constraints are
> similar?

It is not being reused as is, I just thought (incorrectly apparently) that it was a good example for our case.

Best regards,

Sebastian