On Thu, Nov 24, 2022 at 10:20 PM Hans de Goede <hdegoede@xxxxxxxxxx> wrote:
> On 11/24/22 21:09, Andy Shevchenko wrote:
> > On Thu, Nov 24, 2022 at 10:00 PM Hans de Goede <hdegoede@xxxxxxxxxx> wrote:
...
> >> +static const char *int3472_dsm_type_to_func(u8 type)
> >> +{
> >> + switch (type) {
> >> + case INT3472_GPIO_TYPE_RESET:
> >> + return "reset";
> >> + case INT3472_GPIO_TYPE_POWERDOWN:
> >> + return "powerdown";
> >> + case INT3472_GPIO_TYPE_CLK_ENABLE:
> >> + return "clken";
> >> + case INT3472_GPIO_TYPE_PRIVACY_LED:
> >> + return "pled";
> >> + case INT3472_GPIO_TYPE_POWER_ENABLE:
> >> + return "power-enable";
> >
> > default:
> > return "unknown";
> >
> > ?
> >
> >> + }
> >> +
> >> + return "unknown";
> >> +}
> >
>
> In the passed some compiler versions complained about the non-void
> function not ending with a return statement.
>
> I guess I can give your variant a try (I agree it is more readable)
> and of we get compiler warnings we can switch back.
>
> I'll fix this up in the next version or when merging this,
> depending on further feedback on the series.
I believe it's not the case for a long time. (Esp. when the kernel
requires GCC 5.1 as bare minimum). We have a lot of (relatively) new
code that uses switch-cases like I suggested and I have heard none of
the complains from anybody, including all CIs crawling through the
kernel code.
--
With Best Regards,
Andy Shevchenko
