Re: [PATCH v4 13/31] clk: at91: sama7d65: switch to parent_hw and parent_data

[Claudiu Beznea <claudiu.beznea@tuxon.dev>]

Hi, Ryan,

In title: s/switch/switch system clocks"

On 9/19/25 00:15, Ryan.Wanner@microchip.com wrote:
> From: Ryan Wanner <Ryan.Wanner@microchip.com>
> 
> Switch the system clocks to use parent_hw and parent_data. Having this
> allows the driver to conform to the new clk-system API.
> 
> The parent registration is after the USBCK registration due to one of
> the system clocks being dependent on USBCK.
> 
> Signed-off-by: Ryan Wanner <Ryan.Wanner@microchip.com>
> ---
>  drivers/clk/at91/sama7d65.c | 36 +++++++++++++++++++++++-------------
>  1 file changed, 23 insertions(+), 13 deletions(-)
> 
> diff --git a/drivers/clk/at91/sama7d65.c b/drivers/clk/at91/sama7d65.c
> index 986e8ef57dea..38c44b5d5d42 100644
> --- a/drivers/clk/at91/sama7d65.c
> +++ b/drivers/clk/at91/sama7d65.c
> @@ -537,23 +537,23 @@ static struct {
>  /*
>   * System clock description
>   * @n:	clock name
> - * @p:	clock parent name
> + * @p:	clock parent hw
>   * @id: clock id
>   */
> -static const struct {
> +static struct {
>  	const char *n;
> -	const char *p;
> +	struct clk_hw *parent_hw;
>  	u8 id;
>  } sama7d65_systemck[] = {
> -	{ .n = "uhpck",		.p = "usbck", .id = 6 },
> -	{ .n = "pck0",		.p = "prog0", .id = 8, },
> -	{ .n = "pck1",		.p = "prog1", .id = 9, },
> -	{ .n = "pck2",		.p = "prog2", .id = 10, },
> -	{ .n = "pck3",		.p = "prog3", .id = 11, },
> -	{ .n = "pck4",		.p = "prog4", .id = 12, },
> -	{ .n = "pck5",		.p = "prog5", .id = 13, },
> -	{ .n = "pck6",		.p = "prog6", .id = 14, },
> -	{ .n = "pck7",		.p = "prog7", .id = 15, },
> +	{ .n = "uhpck",		.id = 6 },
> +	{ .n = "pck0",		.id = 8, },
> +	{ .n = "pck1",		.id = 9, },
> +	{ .n = "pck2",		.id = 10, },
> +	{ .n = "pck3",		.id = 11, },
> +	{ .n = "pck4",		.id = 12, },
> +	{ .n = "pck5",		.id = 13, },
> +	{ .n = "pck6",		.id = 14, },
> +	{ .n = "pck7",		.id = 15, },
>  };
>  
>  /* Mux table for programmable clocks. */
> @@ -1299,9 +1299,19 @@ static void __init sama7d65_pmc_setup(struct device_node *np)
>  		sama7d65_pmc->pchws[i] = hw;
>  	}
>  
> +	/* Set systemck parent hws. */
> +	sama7d65_systemck[0].parent_hw = usbck_hw;
> +	sama7d65_systemck[1].parent_hw = sama7d65_pmc->pchws[0];
> +	sama7d65_systemck[2].parent_hw = sama7d65_pmc->pchws[1];
> +	sama7d65_systemck[3].parent_hw = sama7d65_pmc->pchws[2];
> +	sama7d65_systemck[4].parent_hw = sama7d65_pmc->pchws[3];
> +	sama7d65_systemck[5].parent_hw = sama7d65_pmc->pchws[4];
> +	sama7d65_systemck[6].parent_hw = sama7d65_pmc->pchws[5];
> +	sama7d65_systemck[7].parent_hw = sama7d65_pmc->pchws[6];
> +	sama7d65_systemck[8].parent_hw = sama7d65_pmc->pchws[7];
>  	for (i = 0; i < ARRAY_SIZE(sama7d65_systemck); i++) {
>  		hw = at91_clk_register_system(regmap, sama7d65_systemck[i].n,
> -					      sama7d65_systemck[i].p, NULL,
> +					      NULL, &AT91_CLK_PD_HW(sama7d65_systemck[i].parent_hw),
>  					      sama7d65_systemck[i].id, 0);

Just saying: we could have been used parent_hw only for system clocks (and
some other clocks updated in this series) if we wouldn't have the dt-compat.c.

>  		if (IS_ERR(hw))
>  			goto err_free;