On Thu, 23 Jan 2025, Arnd Bergmann wrote:
> From: Arnd Bergmann <arnd@xxxxxxxx>
>
> In drivers/hid/, most drivers depend on CONFIG_HID, while a couple of the
> drivers in subdirectories instead depend on CONFIG_HID_SUPPORT and use
> 'select HID'. With the newly added INTEL_THC_HID, this causes a build
> warning for a circular dependency:
>
> WARNING: unmet direct dependencies detected for HID
> Depends on [m]: HID_SUPPORT [=y] && INPUT [=m]
> Selected by [y]:
> - INTEL_THC_HID [=y] && HID_SUPPORT [=y] && X86_64 [=y] && PCI [=y] && ACPI [=y]
>
> WARNING: unmet direct dependencies detected for INPUT_FF_MEMLESS
> Depends on [m]: INPUT [=m]
> Selected by [y]:
> - HID_MICROSOFT [=y] && HID_SUPPORT [=y] && HID [=y]
> - GREENASIA_FF [=y] && HID_SUPPORT [=y] && HID [=y] && HID_GREENASIA [=y]
> - HID_WIIMOTE [=y] && HID_SUPPORT [=y] && HID [=y] && LEDS_CLASS [=y]
> - ZEROPLUS_FF [=y] && HID_SUPPORT [=y] && HID [=y] && HID_ZEROPLUS [=y]
> Selected by [m]:
> - HID_ACRUX_FF [=y] && HID_SUPPORT [=y] && HID [=y] && HID_ACRUX [=m]
> - HID_EMS_FF [=m] && HID_SUPPORT [=y] && HID [=y]
> - HID_GOOGLE_STADIA_FF [=m] && HID_SUPPORT [=y] && HID [=y]
> - PANTHERLORD_FF [=y] && HID_SUPPORT [=y] && HID [=y] && HID_PANTHERLORD [=m]
>
> It's better to be consistent and always use 'depends on HID' for HID
> drivers. The notable exception here is USB_KBD/USB_MOUSE, which are
> alternative implementations that do not depend on the HID subsystem.
>
> Do this by extending the "if HID" section below, which means that a few
> of the duplicate "depends on HID" and "depends on INPUT" statements
> can be removed in the process.
>
> Fixes: 1b2d05384c29 ("HID: intel-thc-hid: Add basic THC driver skeleton")
> Signed-off-by: Arnd Bergmann <arnd@xxxxxxxx>
Applied to hid.git#for-6.14/upstream-fixes, thanks!
--
Jiri Kosina
SUSE Labs
