Hi, On 7/28/21 4:55 PM, Limonciello, Mario wrote: > On 7/28/2021 07:31, Hans de Goede wrote: >> Hi, >> >> On 7/20/21 4:53 AM, Mario Limonciello wrote: >>> This driver can be built on a kernel without rfkill, but events >>> won't work which causes an unexpected experience. >>> >>> Signed-off-by: Mario Limonciello <mario.limonciello@xxxxxxx> >> >> Mario I can understand where you are coming from here, but Kconfig >> dependencies are meant to express true in kernel dependencies. >> >> The wireless-hotkey driver is purely an ACPI based input driver >> from the kernel pov and as such it works fine without RFKILL >> being enabled. >> >> So adding a RFKILL dependency is the wrong thing to do here IMHO. >> > > Thanks, I get your point. Is there another type of relationship that can be expressed for this? No I'm afraid not. > Is SELECT a better idea perhaps? No select should only be used when enabling blocks of kernel code which in essence function as a library used by other code. In general either all users of a Kconfig symbol should either use depends on, or select. If different consumers of the functionality mix depends on and select then the Kconfig tools will typically exit with some error due to e.g. circular dependencies. Regards, Hans >>> --- >>> drivers/platform/x86/Kconfig | 1 + >>> 1 file changed, 1 insertion(+) >>> >>> diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig >>> index 7d385c3b2239..22b6e7e3da13 100644 >>> --- a/drivers/platform/x86/Kconfig >>> +++ b/drivers/platform/x86/Kconfig >>> @@ -419,6 +419,7 @@ config WIRELESS_HOTKEY >>> tristate "Wireless hotkey button" >>> depends on ACPI >>> depends on INPUT >>> + depends on RFKILL >>> help >>> This driver provides supports for the wireless buttons found on some AMD, >>> HP, & Xioami laptops. >>> >> >
