On 12/08/2022 17:38, Rob Herring wrote: >>> No, the common property 'cd-debounce-delay-ms' should be used instead. >> >> So it's a case of "what's in the dts is wrong so it needs to change"? > > Yes. > >> For the future, where is the line drawn between "don't break the dts" >> & "don't use bad properties"? > > Depends... > >> Only if the property has consumers? >> I don't mind, just want to get it right next time ;) > > In this case, there shouldn't be any compatibility issues because the > mmc core has supported the common property longer than the dts > existed. If that was not the case, then we'd probably be stuck with > the property in the dts file. That still depends on usage in the > kernel or elsewhere, in dts files, stability of the platform support, > etc. Cool, thanks Rob!
