Re: [PATCH 1/5] arm64: dts: rockchip: Add rk3399 vop and display-subsystem

[Heiko Stuebner]

Hi Jacob,

Am Freitag, 14. Juli 2017, 09:52:30 CEST schrieb Jacob Chen:
> 2017-07-14 7:34 GMT+08:00 Heiko Stuebner <heiko@xxxxxxxxx>:
> > Am Donnerstag, 13. Juli 2017, 00:03:51 CEST schrieb Jacob Chen:
> >> Add devicetree nodes for rk3399 VOP (Video Output Processors), and the
> >> top level display-subsystem root node.
> >>
> >> Later patches add endpoints (eDP, HDMI, MIPI, etc) that attach to the
> >> VOPs' output ports.
> >>
> >> Signed-off-by: Mark Yao <mark.yao@xxxxxxxxxxxxxx>
> >> Signed-off-by: Yakir Yang <ykk@xxxxxxxxxxxxxx>
> >> Signed-off-by: Caesar Wang <wxt@xxxxxxxxxxxxxx>
> >> Signed-off-by: Jacob Chen <jacob-chen@xxxxxxxxxx>
> >> ---
> >>  arch/arm64/boot/dts/rockchip/rk3399.dtsi | 65 ++++++++++++++++++++++++++++++++
> >>  1 file changed, 65 insertions(+)
> >>
> >> diff --git a/arch/arm64/boot/dts/rockchip/rk3399.dtsi b/arch/arm64/boot/dts/rockchip/rk3399.dtsi
> >> index e795135..300e500 100644
> >> --- a/arch/arm64/boot/dts/rockchip/rk3399.dtsi
> >> +++ b/arch/arm64/boot/dts/rockchip/rk3399.dtsi
> >> @@ -1455,6 +1455,71 @@
> >>               status = "disabled";
> >>       };
> >>
> >> +     vopl: vop@ff8f0000 {
> >> +             compatible = "rockchip,rk3399-vop-lit";
> >> +             reg = <0x0 0xff8f0000 0x0 0x1ffc>, <0x0 0xff8f2000 0x0 0x400>;
> >
> > What is this second memory region doing? It looks like this is the meant
> > to cover VOP_GAMMA_LUT_ADDR, but can't you just map the whole area?
> > ChromeOS seems to be doing fine using the whole area as
> >         <0x0 0xff8f0000 0x0 0x3efc>
> > and I've also not seen any code changes actually mapping/using this
> > second area.
> >
> 
> 0x1c00 - 0x200 for cabc_lut
> 0x2000 - 0x300 for gama_lut
> 
> My mistakes, It seems mark havn't send gamma and cabc support to upstream.
> We shoud just using the whole area map.
> 
> 
> >
> >> +             interrupts = <GIC_SPI 119 IRQ_TYPE_LEVEL_HIGH 0>;
> >> +             clocks = <&cru ACLK_VOP1>, <&cru DCLK_VOP1>, <&cru HCLK_VOP1>, <&cru DCLK_VOP1_DIV>;
> >> +             clock-names = "aclk_vop", "dclk_vop", "hclk_vop", "dclk_source";
> >
> > While I know that this is based on my idea on handling the hdmi pll-rate
> > requirements, I haven't found the matching code- and dt-binding-changes
> > posted to a list yet.
> >
> 
> Yeah, should i collect mark's patches or remove it at first?

As we don't know how much discussion is necessary for that, removing
dclk_source in the first iteration and re-add it once the feature.


Heiko