Re: [PATCH 2/2] ARM: dts: rockchip: add dts for RK3288-Tinker board

[Eddie Cai]

2017-01-20 16:12 GMT+08:00 Heiko Stuebner <heiko@xxxxxxxxx>:
> Hi Eddie,
>
> Am Freitag, 20. Januar 2017, 15:10:47 CET schrieb Eddie Cai:
>> 2017-01-19 17:58 GMT+08:00 Heiko Stuebner <heiko@xxxxxxxxx>:
>> > Hi Eddie,
>> >
>> > Am Donnerstag, 19. Januar 2017, 10:11:59 CET schrieb Eddie Cai:
>> >> This patch add basic support for RK3288-Tinker board. We can boot in to
>> >> rootfs with this patch.
>> >>
>> >> Signed-off-by: Eddie Cai <eddie.cai@xxxxxxxxxxxxxx>
>> >
>> > looks good in general, just some small question down below.
>> >
>> > [...]
>> >
>> >> +     /*
>> >> +      * NOTE: vcc_sd isn't hooked up on v1.0 boards where power comes
>> >> from
>> >> +      * vcc_io directly.  Those boards won't be able to power cycle SD
>> >> cards +      * but it shouldn't hurt to toggle this pin there anyway.
>> >> +      */
>> >
>> > just to clarify, later board will have that pin connected, right?
>>
>> Copy from rk3288-evb.dtsi. forgot to delete it. I will remove it in next
>> version
>> >> +     vcc_sd: sdmmc-regulator {
>> >> +             compatible = "regulator-fixed";
>> >> +             gpio = <&gpio7 11 GPIO_ACTIVE_LOW>;
>> >> +             pinctrl-names = "default";
>> >> +             pinctrl-0 = <&sdmmc_pwr>;
>> >> +             regulator-name = "vcc_sd";
>> >> +             regulator-min-microvolt = <3300000>;
>> >> +             regulator-max-microvolt = <3300000>;
>> >> +             startup-delay-us = <100000>;
>> >> +             vin-supply = <&vcc_io>;
>> >> +     };
>> >> +};
>> >
>> > [...]
>> >
>> >> +&hdmi {
>> >> +     #address-cells = <1>;
>> >> +     #size-cells = <0>;
>> >> +     #sound-dai-cells = <0>;
>> >> +     ddc-i2c-bus = <&i2c5>;
>> >> +     status = "okay";
>> >> +     /* Don't use vopl for HDMI */
>> >> +     ports {
>> >> +             hdmi_in: port {
>> >> +                     /delete-node/ endpoint@1;
>> >> +             };
>> >
>> > what is the reason for this? You enable both VOPs below and the linux
>> > display subsystem should be able to select an appropriate VOP for output
>> > just fine on its own. So there should be no reason for capping the hdmi's
>> > connection to one of the vops.
>>
>> The VOP big support 4k display. is designed for HDMI  4K display. VOP
>> little is for other display(eDP, LVDS, Mipi etc)
>
> The hdmi _can_ talk to both vops, which is why it has the connection to both.
> Resolution-limitations and selecting a matching vop should be handled in the
> drm driver I'd think - but you'll need to talk to Mark Yao or some else
> knowledgable in graphics.
>
> The devicetree is about describing the _available_ hardware, not configuring
> how it is supposed to be used ;-) .
OK, agree with you, i will remove it in next version.
>
>
> Heiko