On Sun, Jun 11, 2017 at 02:58:47PM +0800, icen...@aosc.io wrote:
> 在 2017-06-07 17:42,Maxime Ripard 写道:
> > On Mon, Jun 05, 2017 at 12:01:48AM +0800, Icenowy Zheng wrote:
> > > + soc {
> > > +         display_clocks: clock@1000000 {
> > > +                 compatible = "allwinner,sun8i-a83t-de2-clk";
> > > +                 reg = <0x01000000 0x100000>;
> > > +                 clocks = <&ccu CLK_BUS_DE>,
> > > +                          <&ccu CLK_DE>;
> > > +                 clock-names = "bus",
> > > +                               "mod";
> > > +                 resets = <&ccu RST_BUS_DE>;
> > > +                 #clock-cells = <1>;
> > > +                 #reset-cells = <1>;
> > > +                 assigned-clocks = <&ccu CLK_DE>;
> > > +                 assigned-clock-parents = <&ccu CLK_PLL_DE>;
> > > +                 assigned-clock-rates = <432000000>;
> > > +         };
> > 
> > We discussed that already a few times, but there's no reason to do
> > so. If you need a downstream clock at a particular rate, call
> > clk_set_rate on it, period.
> > 
> > Whether its parent will be coming from PLL_DE or some other more
> > appriopriate clock is not relevant and doesn't make any difference.
> 
> The clock framework is not so smart to deal with these infomations:
> - CLK_PLL_PERIPH should always be 600MHz
> - CLK_TVE should always be 216MHz
> - CLK_DE (in fact CLK_MIXER{0,1}) should be larger than 300MHz (for 4K)

None of what you're doing guarantees what you state above, so I'm not
really sure what your point is.

> So we have to specify CLK_DE to be 432MHz, and then it will set
> CLK_PLL_DE to this value, then the CLK_TVE can be set to 216MHz with
> divider 2.

Yes, but it works by accident. Any clock change somewhere in the same
clock-tree might break whatever you have set in the DT.

Hence why you want to do it within the clock framework and your
driver, not here.

Maxime

-- 
Maxime Ripard, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

Attachment: signature.asc
Description: PGP signature

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Reply via email to