Re: [PATCH net-next v5 1/4] net: wangxun: Add support for PTP clock

[Vadim Fedorenko <vadim.fedorenko@linux.dev>]

On 17/01/2025 15:51, Richard Cochran wrote:
> On Fri, Jan 17, 2025 at 02:15:01PM +0000, Vadim Fedorenko wrote:
> 
>> there is no way ptp_clock_register() will return NULL,
> 
> Really?
> 
> include/linux/ptp_clock_kernel.h:
> 
>   400 static inline struct ptp_clock *ptp_clock_register(struct ptp_clock_info *info,
>   401                                                    struct device *parent)
>   402 { return NULL; }
> 
> Also, sometimes the kernelDoc comments are correct, like in this case:
> 
>   304 /**
>   305  * ptp_clock_register() - register a PTP hardware clock driver
>   306  *
>   307  * @info:   Structure describing the new clock.
>   308  * @parent: Pointer to the parent device of the new clock.
>   309  *
>   310  * Returns a valid pointer on success or PTR_ERR on failure.  If PHC
>   311  * support is missing at the configuration level, this function
>   312  * returns NULL, and drivers are expected to gracefully handle that
>   313  * case separately.
>   314  */
> 
> 
> Thanks,
> Richard

Well, yes, this case is a special one. Then maybe it's better to adjust
Kconfig and Makefile to avoid it?