Hi Ryan,

On Thu, Oct 15, 2026 at 4:26 PM Ryan Roberts <[email protected]> wrote:

> * Driver interaction with architectural support: I've opted to treat the CLA 
> as
>   a device rather than a CPU extension, so it's implemented as a (mostly)
>   self-contained driver.

What the current proposal does is to create nodes in the base of the device
tree, which makes the device tree parser instantiate those into
platform_device:s,
then each cla_device piggy-backs that platform device.

I don't think this is a good idea, I wouldn't use platform_device for this
since you can just look up the nodes and instantiate the cla_device:s
with a proper struct device inside all of them.

I think it's a better idea to put these inside the CPU nodes themselves,
they are there after all, intuitive right? Init code can loop over the
CPUs and instantiate a cla_device for each CPU directly without
having to detour over an opaque platform_device.

Devices inside CPU nodes will not (I think??) be populated
automatically, so you can add code to loop over the CPU nodes
at device_initcall level or any time during the boot process deemed
helpful. This also saves you from adding devices for CPU cores
that are flagged with status = "disabled"; which could happen.

Yours,
Linus Walleij

Reply via email to