Re: [PATCH 5/7] ARM: sun8i: a33: Add display pipeline

From: Chen-Yu Tsai
Date: Mon Sep 05 2016 - 22:53:52 EST


On Tue, Sep 6, 2016 at 4:21 AM, Maxime Ripard
<maxime.ripard@xxxxxxxxxxxxxxxxxx> wrote:
> Hi,
>
> On Fri, Sep 02, 2016 at 02:28:54PM +0800, Chen-Yu Tsai wrote:
>> > + be0: display-backend@01e60000 {
>> > + compatible = "allwinner,sun8i-a33-display-backend";
>> > + reg = <0x01e60000 0x10000>;
>>
>> Please also list the interrupt, even though we don't use it yet.
>> The manual says it's 127 - 32 = 95.
>
> Yep, you're right.
>
>> > + sat0: sat@01e80000 {
>> > + compatible = "allwinner,sun8i-a33-sat";
>> > + reg = <0x01e80000 0x1000>;
>> > + clocks = <&ccu CLK_BUS_SAT>;
>> > + resets = <&ccu RST_BUS_SAT>;
>> > +
>> > + ports {
>> > + #address-cells = <1>;
>> > + #size-cells = <0>;
>> > +
>> > + sat0_in: port@0 {
>> > + #address-cells = <1>;
>> > + #size-cells = <0>;
>> > + reg = <0>;
>> > +
>> > + sat0_in_fe0: endpoint@0 {
>> > + reg = <0>;
>> > + remote-endpoint = <&fe0_out_sat0>;
>> > + };
>> > + };
>> > +
>> > + sat0_out: port@1 {
>> > + #address-cells = <1>;
>> > + #size-cells = <0>;
>> > + reg = <1>;
>> > +
>> > + sat0_out_be0: endpoint@0 {
>> > + reg = <0>;
>> > + remote-endpoint = <&be0_in_sat0>;
>> > + };
>> > + };
>>
>> I'm worried about the representation here.
>>
>> In the user manuals, the SAT is shown as part of the BE. Look at it
>> this way: if it did come before the BE and is independent, we
>> shouldn't have to bring the SAT out of reset for simplefb to work.
>
> Indeed.
>
>> For comparison, a similar function unit called "CMU" found on the
>> other post-sun6i SoCs has the same function description as SAT on
>> the A33. It uses the reserved registers at the beginning of the BE
>> address space.
>
> Hmm, ok, so you would essentially, merge the backend and sat nodes?
>
> That wouldn't be very hard to do, i'll do it.

Yes. That is what I propose.

ChenYu