Re: [PATCH v2 10/11] mfd: motorola-cpcap: diverge configuration per-board

[Andy Shevchenko <andriy.shevchenko@intel.com>]

On Sun, Feb 08, 2026 at 05:19:49PM +0200, Svyatoslav Ryhel wrote:
> 8 лютого 2026 р. 14:54:53 GMT+02:00, Andy Shevchenko <andriy.shevchenko@intel.com> пише:
> >On Fri, Feb 06, 2026 at 07:28:44PM +0200, Svyatoslav Ryhel wrote:

...

> >> -static const struct of_device_id cpcap_of_match[] = {
> >> -	{ .compatible = "motorola,cpcap", },
> >> -	{ .compatible = "st,6556002", },
> >> -	{},
> >> -};
> >> -MODULE_DEVICE_TABLE(of, cpcap_of_match);
> >> -
> >>  static const struct spi_device_id cpcap_spi_ids[] = {
> >>  	{ .name = "cpcap", },
> >>  	{ .name = "6556002", },
> >> +	{ .name = "mapphone-cpcap", },
> >>  	{},
> >>  };
> >
> >This  also needs to be converted to use driver data.
> 
> I can do that, but is it needed?

For the consistency's sake.
Also brief look at MFD SPI drivers I see that most of them use driver data.
All that is redundancy? Maybe clean them up as well to show the point?

> This id table, if I understand correctly, is here solely to silence warnings
> about missing spi id table. Driver itself does not use it in any way and will
> never use since all available devices that use this driver rely solely on the
> device tree, all possible future devices will be relying on device tree too
> due to obvious reasons.

Then maybe SPI should be fixed to avoid this table to be present? (But don't
waste your time, it will be a rabbit hole.)

> >>  MODULE_DEVICE_TABLE(spi, cpcap_spi_ids);

...

> And please, may you contain all your reviewing in one iteration. Then both of
> us will spend less time working on the patch. Thank you!

It's an impossible request, you should understand that. If you want, use AI
or your robot colleagues for that, I'm just a human being, I can't see
everything at once.


-- 
With Best Regards,
Andy Shevchenko