[PATCH RESEND v6 1/2] arm64: dts: Add the arasan sdhci nodes in apm-storm.dtsi

[Ulf Hansson]

On 12 May 2015 at 09:28, Arnd Bergmann <arnd at arndb.de> wrote:
> On Tuesday 12 May 2015 11:03:46 Suman Tripathi wrote:
>> > @@ -533,6 +567,15 @@
>> >                         interrupts = <0x0 0x4f 0x4>;
>> >                 };
>> >
>> > +               sdhci0: sdhci at 1c000000 {
>> > +                       compatible = "arasan,sdhci-4.9a";
>> > +                       reg = <0x0 0x1c000000 0x0 0x100>;
>> > +                       interrupts = <0x0 0x49 0x4>;
>> > +                       dma-coherent;
>> > +                       clock-names = "clk_xin", "clk_ahb";
>> > +                       clocks = <&sdioclk 0>, <&ahbclk 0>;
>> > +               };
>> > +
>> >                 phy1: phy at 1f21a000 {
>> >                         compatible = "apm,xgene-phy";
>> >                         reg = <0x0 0x1f21a000 0x0 0x100>;
>> > --
>> > 1.8.2.1
>> >
>>
>> Can anyone from dt community review this patch ? I have changed the dts node
>> names from sdhc to sdhci as per Arnd, Michael comments .
>
> I was actually asking for it to be named 'mmc', not 'sdhci', because the
> name is supposed to indicate the purpose of the device, not the
> implementation. I realize that we are inconsistent here, just as with
> 'uart' vs 'serial', and that ePAPR does not define what to do.

Then we need a common name to address the following purposes: SD, MMC,
eMMC, SDIO.

>
> We should probably add something to Documentation/devicetree/bindings/mmc/mmc.txt
> about this topic and change all the dts files accordingly (unless there
> is a risk for regressions). At the moment, the mmc.txt file also includes
> an example with 'sdhci', not 'mmc'.
>

I am happy to change, as long as we can come up with something better.

Kind regards
Uffe