On Tue, May 10, 2016 at 06:13:55PM +0900, Magnus Damm wrote: > On Tue, May 10, 2016 at 2:52 PM, Simon Horman > <horms+renesas@xxxxxxxxxxxx> wrote: > > Add the sd-uhs-sdr104 property to SDHI0. > > > > Signed-off-by: Simon Horman <horms+renesas@xxxxxxxxxxxx> > > --- > > arch/arm/boot/dts/r8a7790-lager.dts | 1 + > > 1 file changed, 1 insertion(+) > > > > diff --git a/arch/arm/boot/dts/r8a7790-lager.dts b/arch/arm/boot/dts/r8a7790-lager.dts > > index 749ba02b6a53..05d1ff7acee2 100644 > > --- a/arch/arm/boot/dts/r8a7790-lager.dts > > +++ b/arch/arm/boot/dts/r8a7790-lager.dts > > @@ -559,6 +559,7 @@ > > vqmmc-supply = <&vccq_sdhi0>; > > cd-gpios = <&gpio3 6 GPIO_ACTIVE_LOW>; > > sd-uhs-sdr50; > > + sd-uhs-sdr104; > > status = "okay"; > > Hi Simon, > > Thanks for this - interesting to see!! From what I can tell this code > is targeting r8a7790 Lager, and based on the data sheet and the DTS > there are four SDHI channels for the r8a7790 SoC. All R-Car Gen2 SDHI > channels are not identical, so on r8a7790 it looks like SDHI0 and > SDHI1 have extended capabilities for the clocks (and probably support > SDR104) while SDHI2 and SDHI3 do not have this hardware feature. How > is this difference handled today? In my mind it would make sense to > have different compat strings, but I think we differentiate with > resource size today? Probably I am missing the point somehow but my understanding is that that with this and other recent changes these differences are described by DT properties rather than the compatibility string. In particular the presence or absence of sd-uhs-sdr50 and sd-uhs-sdr104, and the value of max-frequency. Now you mention this I do wonder if sd-uhs-sdr50 and sd-uhs-sdr104 are SoC properties and thus should be present in r8a7790.dtsi rather than r8a7790-lager.dts. > So with this patches I can see that you enable SDR104 on SDHI0 on > Lager which makes sense, but is SDHI2 support still OK? I will double check but I think so.
