Hi Chris,
On Tue, Aug 30, 2016 at 5:13 AM, Chris Brandt <chris.brandt@xxxxxxxxxxx> wrote:
> Check the MD_CLK pin to determine the current clock mode in order to set
> the pll clock parent correctly.
>
> Signed-off-by: Chris Brandt <chris.brandt@xxxxxxxxxxx>
Thanks for your patch!
> ---
> v2:
> * Switched to reading MD_CLK pin to determine mode
> ---
> drivers/clk/renesas/clk-rz.c | 21 +++++++++++++++++----
> 1 file changed, 17 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/clk/renesas/clk-rz.c b/drivers/clk/renesas/clk-rz.c
> index f6312c6..2fc8aae 100644
> --- a/drivers/clk/renesas/clk-rz.c
> +++ b/drivers/clk/renesas/clk-rz.c
> @@ -25,6 +25,9 @@ struct rz_cpg {
> #define CPG_FRQCR 0x10
> #define CPG_FRQCR2 0x14
>
> +#define PPR0 0xFCFE3200
> +#define PIBC0 0xFCFE7000
> +
> /* -----------------------------------------------------------------------------
> * Initialization
> */
> @@ -37,10 +40,20 @@ rz_cpg_register_clock(struct device_node *np, struct rz_cpg *cpg, const char *na
> static const unsigned frqcr_tab[4] = { 3, 2, 0, 1 };
>
> if (strcmp(name, "pll") == 0) {
> - /* FIXME: cpg_mode should be read from GPIO. But no GPIO support yet */
> - unsigned cpg_mode = 0; /* hardcoded to EXTAL for now */
> - const char *parent_name = of_clk_get_parent_name(np, cpg_mode);
> -
> + unsigned int cpg_mode;
> + const char *parent_name;
> + void __iomem *ppr0, *pibc0;
> +
> + /* MD_CLK is on P0_2 */
> + ppr0 = ioremap_nocache(PPR0, 2);
> + pibc0 = ioremap_nocache(PIBC0, 2);
> + BUG_ON(!ppr0 || !pibc0);
> + iowrite16(4, pibc0); /* Enable input buffer */
> + cpg_mode = ioread16(ppr0) & 4;
> + iounmap(ppr0);
> + iounmap(pibc0);
> +
> + parent_name = cpg_mode ? "usb_x1" : "extal";
If you would use "cpg_mode = (ioread16(ppr0) >> 2) & 1;", the value of
cpg_mode would be in sync with Table 6.2. Then you can avoid relying on actual
clock names in DT, and keep
parent_name = of_clk_get_parent_name(np, cpg_mode);
The bindings already dictate the parent clocks must match clock modes in the
datasheet, i.e. Table 6.2.
For easier maintenance, I would factor out reading the mode pins in a separate
function. When a proper GPIO driver is added, the function can be removed,
and its callsite updated.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
