Re: [Ksummit-discuss] Devicetree Workshop at Kernel Summit Prague (26 Oct 2017)

[Tom Rini <trini@konsulko.com>]

[-- Attachment #1: Type: text/plain, Size: 1484 bytes --]

On Wed, Oct 18, 2017 at 12:09:58PM +0100, Mark Brown wrote:
> On Wed, Oct 18, 2017 at 06:35:24PM +0800, Chen-Yu Tsai wrote:
> 
> > I'd like to add something on the topic of non-Linux projects. In this
> > case it's diverging DT bindings from U-boot:
> 
> > https://patchwork.ozlabs.org/patch/823158/
> 
> > U-boot already has a set of devicetree binding additions:
> > https://github.com/u-boot/u-boot/tree/master/doc/device-tree-bindings
> 
> > The patch in question wants to ab(use) the regulator-name property for
> > driver instance binding. In my opinion this is not going to fly, as
> > boards are free to define the names. This either sees no use other than
> > as a dirty workaround for dts files that aren't following the PMIC
> > regulator bindings (regulator node names should follow well defined,
> > identifying names), or results in divergence of the DT files.
> 
> One meta issue I'm seeing here is that the u-boot people appear to have
> their own divergent copy of some of the binding documents.

Putting on my U-Boot hat now, it's mostly unintentional and something in
general (yes, the initial topic here is not such an example) we try and
avoid, or use u-boot, as the prefix on as it's something that had been
previously rejected or deemed inappropriate to be in the upstream
version of the binding.

But perhaps it's time to try and force the issue again, given what Rob
and others have said in other parts of the thread.

-- 
Tom

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 819 bytes --]