Re: [PATCH 2/2] ARM: Tegra: Seaboard: Add devicetree support

[Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org>]

On Sun, May 1, 2011 at 8:38 AM, Olof Johansson <olof-nZhT3qVonbNeoWH0uzbU5w@public.gmane.org> wrote:
> Hi,
>
> On Fri, Apr 29, 2011 at 10:12:31PM -0600, Stephen Warren wrote:
>> Signed-off-by: Stephen Warren <swarren-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
>> ---
>>  arch/arm/boot/dts/tegra-seaboard.dts |   33 +++++++++++++++++++++++++++++++++
>>  arch/arm/mach-tegra/Makefile.boot    |    1 +
>>  2 files changed, 34 insertions(+), 0 deletions(-)
>>  create mode 100644 arch/arm/boot/dts/tegra-seaboard.dts
>>
>> diff --git a/arch/arm/boot/dts/tegra-seaboard.dts b/arch/arm/boot/dts/tegra-seaboard.dts
>> new file mode 100644
>> index 0000000..d5c9ad7
>> --- /dev/null
>> +++ b/arch/arm/boot/dts/tegra-seaboard.dts
>> @@ -0,0 +1,33 @@
>> +/dts-v1/;
>> +/include/ "tegra250.dts"
>> +
>> +/ {
>> +     model = "NVIDIA Harmony";
>
> NVIDIA Seaboard?
>
>> +     compatible = "nvidia,seaboard", "nvidia,tegra250";
>> +
>> +     chosen {
>> +             bootargs = "vmalloc=192M video=tegrafb console=ttyS0,115200n8 root=/dev/mmcblk1p3 rw rootdelay=2";
>
> I'm partial to rootwait instead of rootdelay, a little less fragile for slow
> devices.
>
>> +     memory {
>> +             device_type = "memory";
>> +             reg = < 0x00000000 0x1C000000
>> +                     0x20000000 0x20000000 >;
>> +     };
>
> I guess a node for the graphics memory could be warranted

I've modified this in devicetree test to specify the entire memory
range, and carve out 0x1c000000..0x1fffffff with a memreserve section.

>
>> +
>> +     serial@70006300 {
>> +             status = "ok";
>> +             clock-frequency = < 216000000 >;
>> +     };
>> +
>> +     sdhci@c8000400 {
>> +             status = "ok";
>> +             gpios = <&gpio 69 0>, /* cd, gpio PI5 */
>> +                     <&gpio 57 0>, /* wp, gpio PH1 */
>> +                     <&gpio 70 0>; /* power, gpio PI6 */
>
> How would you represent just one missing gpio here? It might be better
> to have separate properties for each.

The gpio binding allows for empty entries by using '0' for the phandle.

>
> Also, it'd be nice to document these assumptions where other device tree
> bindings are (Documentation/devicetree/bindings).
>
>
> -Olof
>



-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.