[PATCHv4 2/2] clk: keystone: Add sci-clk driver support

[Tero Kristo]

On 13/06/17 05:04, Stephen Boyd wrote:
> On 05/22, Tero Kristo wrote:
>> +
>> +/**
>> + * ti_sci_clk_probe - Probe function for the TI SCI clock driver
>> + * @pdev: platform device pointer to be probed
>> + *
>> + * Probes the TI SCI clock device. Allocates a new clock provider
>> + * and registers this to the common clock framework. Also applies
>> + * any required flags to the identified clocks via clock lists
>> + * supplied from DT. Returns 0 for success, negative error value
>> + * for failure.
>> + */
>> +static int ti_sci_clk_probe(struct platform_device *pdev)
>> +{
>> +	struct device *dev = &pdev->dev;
>> +	struct device_node *np = dev->of_node;
>> +	struct sci_clk_provider *provider;
>> +	const struct ti_sci_handle *handle;
>> +	struct sci_clk_data *data;
>> +	int ret;
>> +
>> +	data = (struct sci_clk_data *)of_device_get_match_data(dev);
> 
> It would be nice to avoid this cast, and also check for NULL
> here. I tried to do it in a few minutes but then that caused a
> problem with throwing away const on sci_clk_data. Can this be
> resolved? We probably shouldn't be assigning the
> sci_clk_data::clocks member anyway if the structure is const
> assuming it goes into the read-only section. It may cause some
> restructuring of the code though.

Yeah, thats a good catch. I just cooked up a fix for that, will post v5 
of this patch alone in a bit.

Basically I moved the sci_clk_data->clocks to the provider itself so the 
sci_clk_data can remain const.

-Tero

> 
>> +
>> +	handle = devm_ti_sci_get_handle(dev);
>> +	if (IS_ERR(handle))
>> +		return PTR_ERR(handle);