Re: [PATCH 1/6] cpufreq: Use existing stub functions instead of IS_ENABLED macro

[Rafael J. Wysocki]

On Monday, June 24, 2019 11:22:19 AM CEST Daniel Lezcano wrote:
> On 22/06/2019 11:12, Rafael J. Wysocki wrote:
> > On Fri, Jun 21, 2019 at 3:23 PM Daniel Lezcano
> > <daniel.lezcano@xxxxxxxxxx> wrote:
> >>
> >> The functions stub already exist for the condition the IS_ENABLED
> >> is trying to avoid.
> >>
> >> Remove the IS_ENABLED macros as they are pointless.
> > 
> > AFAICS, the IS_ENABLED checks are an optimization to avoid generating
> > pointless code (including a branch) in case CONFIG_CPU_THERMAL is not
> > set.
> > 
> > Why do you think that it is not useful?
> 
> I agree but I'm not a big fan of IS_ENABLED macros in the code when it
> is possible to avoid them.
> 
> What about adding a stub for that like:

Well,

> #ifdef CPU_THERMAL
> static inline int cpufreq_is_cooling_dev(struct cpufreq_driver *drv)
> {
> 	return drv->flags & CPUFREQ_IS_COOLING_DEV;
> }
> #else
> static inline int cpufreq_is_cooling_dev(struct cpufreq_driver *drv)
> {
> 	return 0;
> }
> #endif

This may as well be defined as

static inline int cpufreq_is_cooling_dev(struct cpufreq_driver *drv)
{
	return IS_ENABLED(CPU_THERMAL) && drv->flags & CPUFREQ_IS_COOLING_DEV;
}

which is fewer lines of code.

And I would call it something like cpufreq_thermal_control_enabled().