On 09/04/2024 10:52, Javier Carrasco wrote: > On 4/9/24 09:34, Krzysztof Kozlowski wrote: >> On 08/04/2024 17:53, Javier Carrasco wrote: >>> This RTC requires a compatible, a reg and a single interrupt, >>> which makes it suitable for a direct conversion into trivial-rtc. >>> >>> Signed-off-by: Javier Carrasco <javier.carrasco.cruz@xxxxxxxxx> >>> --- >>> Documentation/devicetree/bindings/rtc/lpc32xx-rtc.txt | 15 --------------- >>> Documentation/devicetree/bindings/rtc/trivial-rtc.yaml | 2 ++ >>> 2 files changed, 2 insertions(+), 15 deletions(-) >> >> This one no... and if you tested DTS you would see errors, although you >> need to test specific lpc config, not multi_v7. >> >> It does not look like you tested the DTS against bindings. Please run >> `make dtbs_check W=1` (see >> Documentation/devicetree/bindings/writing-schema.rst or >> https://www.linaro.org/blog/tips-and-tricks-for-validating-devicetree-sources-with-the-devicetree-schema/ >> for instructions). >> >> Anyway, you *must* check all DTS before moving anything to trivial. >> >> Does it mean all other bindings were not checked against DTS at all? >> >> Best regards, >> Krzysztof >> > Hi, > > I did check the conversion against nxp/lpc/lpc3250-phy3250.dts, which > throws a message about the 'clocks' property. > > That property is not documented in the original binding, and even though > it could be missing, I could not find any function to get a clock (i.e. Old bindings are not really accurate. > any form of clk_get()) in rtc-lpc32xx.c, which is the only file where > the compatible can be found. > > Is therefore the property not useless in the dts? My apologies if I am > missing something here. Useless for whom? For Linux yes. For U-Boot or out-of-tree users of DTS, I don't know. Anyway the true question is if there is a clock or there is no. If there is a clock, then it should be in the binding even if Linux driver does not use it. I propose to add it and be done with it. Best regards, Krzysztof
