Re: [PATCH v1 1/1] mfd: intel-lpss: Fix the fractional clock divider flags

[Andy Shevchenko]

On Wed, Dec 13, 2023 at 05:46:05PM +0000, Lee Jones wrote:
> On Wed, 13 Dec 2023, Andy Shevchenko wrote:
> > On Wed, Dec 13, 2023 at 04:18:54PM +0000, Lee Jones wrote:
> > > On Wed, 13 Dec 2023, Andy Shevchenko wrote:
> > > > On Wed, Dec 13, 2023 at 04:13:52PM +0000, Lee Jones wrote:
> > > > > On Mon, 11 Dec 2023, Andy Shevchenko wrote:

...

> > > > > >  	tmp = clk_register_fractional_divider(NULL, name, __clk_get_name(tmp),
> > > > > > +					      0, lpss->priv, 1, 15, 16, 15,
> > > > > >  					      CLK_FRAC_DIVIDER_POWER_OF_TWO_PS,
> > > > > > -					      lpss->priv, 1, 15, 16, 15, 0,
> > > > > >  					      NULL);
> > > > > 
> > > > > What an ugly interface.  Intel-only too, right?
> > > > 
> > > > Nope, de facto way how custom clocks are being introduced.
> > > > See clk-provider.h for several similar APIs (that require an
> > > > additional, custom, flags to be supplied).
> > > 
> > > This call only has 2 call-sites, both Intel.
> > 
> > Yes, but the clock fractional divider is used wider.
> > 
> > And again, it's not related to Intel, as this how clock framework
> > does the custom clocks. I don't know how to say this clearer.
> 
> I'm not sure how you can say that.  Intel were the authors, hold the
> _only_ copyright and are the _only_ users.  If it were to be removed,
> there is only a single entity that would even notice.

_This_ API is indeed used by only Intel code right now, but the _design_
of the API is dictated by CCF, and not anyhow related to Intel.

> Anyway, it was just a passing comment.  Not positive, not negative.

Okay!

-- 
With Best Regards,
Andy Shevchenko