[PATCH v4] ARM: prima2: move to generic reset controller driver framework

[arnd@arndb.de (Arnd Bergmann)]

On Monday 06 January 2014, Barry Song wrote:
> this moves to generic IP module reset framework, and make other drivers
> use common device_reset() API.
> 
> diff --git a/Documentation/devicetree/bindings/reset/sirf,rstc.txt b/Documentation/devicetree/bindings/reset/sirf,rstc.txt
> new file mode 100644
> index 0000000..0505de7
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/reset/sirf,rstc.txt
> @@ -0,0 +1,42 @@
> +CSR SiRFSoC Reset Controller
> +======================================
> +
> +Please also refer to reset.txt in this directory for common reset
> +controller binding usage.
> +
> +Required properties:
> +- compatible: Should be "sirf,prima2-rstc" or "sirf,marco-rstc"
> +- reg: should be register base and length as documented in the
> +  datasheet
> +- #reset-cells: 1, see below
> +
> +example:
> +
> +rstc: reset-controller at 88010000 {
> +	compatible = "sirf,prima2-rstc";
> +	reg = <0x88010000 0x1000>;
> +	#reset-cells = <1>;
> +};

Binding looks good to me.

> +void __init sirfsoc_of_rstc_init(void)
> +{
> +	struct device_node *np = of_find_matching_node(NULL, rstc_ids);
> +	if (!np) {
> +		pr_err("unable to find compatible sirf rstc node in dtb\n");
> +		return;
> +	}
> +
> +	sirfsoc_rstc_base = of_iomap(np, 0);
> +	if (!sirfsoc_rstc_base) {
> +		pr_err("unable to map rstc cpu registers\n");
> +		return;
> +	}
> +
> +	sirfsoc_reset_controller.of_node = np;
> +
> +	reset_controller_register(&sirfsoc_reset_controller);
> +}
> +

Can't this be a regular platform device? Drivers that get the reference to
the reset controller should respect a -EPROBE_DEFER error message if they
get initialized in the wrong order.

	Arnd