On Tue, Sep 24, 2024 at 6:46 PM Jeff Johnson <quic_jjohnson@xxxxxxxxxxx> wrote: > > On 9/24/2024 1:06 AM, Krzysztof Kozlowski wrote: > > On 20/09/2024 23:02, Jeff Johnson wrote: > >> Again, since I'm a DT n00b: > >> Just to make sure I understand, you are saying that with this change any > >> existing .dts/.dtb files will still work with an updated driver, so the new > >> properties are not required to be populated on existing devices. > > > > Did you folks rejected patches acked by DT maintainers on basis of not > > understanding DT at all? > > I personally have not rejected any DT patches. Nor have I accepted any. > I'm deferring to Kalle. > > >> However a new driver with support for these properties will utilize them when > >> they are present, and the current ath11k .dts files will need to be updated to > > > > It is not related to drivers at all. Nothing in this thread is related > > to drivers. > > > > Can we entirely drop the drivers from the discussion? > > I brought up drivers since in the discussion there was concern that this DT > change would potentially break existing devices that have a DTS that defines > qcom,ath11k-calibration-variant, and I wanted clarification on that point. > > /jeff How could this happen? DT schema is used to validate device-tree sources but is not used by the kernel code in any way. I've said it several times over the course of this and the previous (qca6390) discussion. Bartosz