On Wed, Dec 1, 2021 at 1:11 PM Sean Anderson <sean...@gmail.com> wrote:
>
> Hi Adam,
>
> On 6/4/21 1:26 PM, Adam Ford wrote:
> > +/*
> > + * The device references to the Versaclock point to the head, so xlate
> > needs to
> > + * redirect it to clk_out[idx]
> > + */
> > +static int vc5_clk_out_xlate(struct clk *hw, struct ofnode_phandle_args
> > *args)
> > +{
> > + unsigned int idx = args->args[0];
> > +
> > + if (args->args_count != 1) {
> > + debug("Invaild args_count: %d\n", args->args_count);
> > + return -EINVAL;
> > + }
> > +
> > + hw->id = idx;
> > +
> > + return 0;
> > +}
>
> Can you explain a bit more why this is needed? I looked it over a few
> times, and I'm not sure what it does differently from
> clk_of_xlate_default.
If memory serves, the indices were starting at 0 instead of 1. To
match what Lniux does we I need to switch the index to be 1,2,3,4
instead of 0,1,2,3 since there is an OUT0. It's been a while since I
did this, so I don't fully remember.
If you have good success removing it and using the
clk_of_xlate_default, I'm willing to test it to make sure it doesn't
break my board so we can remove it.
adam
>
> --Sean
