On Fri, Dec 01, 2023 at 01:51:06PM -0600, Rob Herring wrote: > On Fri, Dec 1, 2023 at 10:56 AM Johan Hovold <johan@xxxxxxxxxx> wrote: > > On Fri, Dec 01, 2023 at 08:43:20AM -0600, Rob Herring wrote: > > > This is fine, but I prefer these MFDs have 1 complete example rather > > > than piecemeal examples for each child device. > > > > Yeah, this is not ideal. The closest thing we've got are the examples > > in: > > > > Documentation/devicetree/bindings/mfd/qcom,spmi-pmic.yaml > > > > Are you suggesting eventually dropping the examples from the child node > > bindings and adding (several) complete examples in the parent one? > > Yes, but if the child nodes are truly reused across multiple PMICs > then, it is probably a worthwhile exception. There's not a great deal > of reuse on most MFDs. Yes, they are indeed reused by multiple PMICs in this case. > > I guess there would need to be more than one if you want to cover all > > the various child nodes with real examples. > > We don't want examples to be exhaustive permutations of every > possibility either. Not every possible permutation but I guess we'd want coverage of all the various child nodes still (i.e. the child node examples that would have been removed). Johan
