2022-11-18 11:07:18

by Benjamin Tissoires

Subject: [PATCH for-6.2/hid-bpf] HID: force HID depending on INPUT

In most configurations, INPUT is actually a boolean: either y or disabled,
but when it's disabled, you can't do much on your average laptop.

But it turns out that there is a possibility to have INPUT as a module:
you have to disable VT and TTY (of course), but also enable EXPERT.
I'll leave how to disable VT and TTY as an exercise for the bravest.

Anyway, if INPUT is m, we can still configure HID as y, which is not
correct because hid-input.c depends on the input API, meaning that
vmlinuz can not link.

So: add depends on INPUT too at the HID level, to ensure that if INPUT=m,
HID can only be m or disabled.

Fixes: 25621bcc8976 ("HID: Kconfig: split HID support and hid-core compilation")
Reported-by: kernel test robot <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Benjamin Tissoires <[email protected]>
---
drivers/hid/Kconfig | 1 +
1 file changed, 1 insertion(+)

diff --git a/drivers/hid/Kconfig b/drivers/hid/Kconfig
index ff40c18f4a2b..b345bbfb57bf 100644
--- a/drivers/hid/Kconfig
+++ b/drivers/hid/Kconfig
@@ -15,6 +15,7 @@ if HID_SUPPORT
config HID
tristate "HID bus core support"
default y
+ depends on INPUT
help
A human interface device (HID) is a type of computer device that
interacts directly with and takes input from humans. The term "HID"
--
2.38.1

2022-11-21 18:17:42

by Jiri Kosina

Subject: Re: [PATCH for-6.2/hid-bpf] HID: force HID depending on INPUT

On Fri, 18 Nov 2022, Benjamin Tissoires wrote:

> In most configurations, INPUT is actually a boolean: either y or disabled,
> but when it's disabled, you can't do much on your average laptop.
>
> But it turns out that there is a possibility to have INPUT as a module:
> you have to disable VT and TTY (of course), but also enable EXPERT.
> I'll leave how to disable VT and TTY as an exercise for the bravest.
>
> Anyway, if INPUT is m, we can still configure HID as y, which is not
> correct because hid-input.c depends on the input API, meaning that
> vmlinuz can not link.
>
> So: add depends on INPUT too at the HID level, to ensure that if INPUT=m,
> HID can only be m or disabled.
>
> Fixes: 25621bcc8976 ("HID: Kconfig: split HID support and hid-core compilation")
> Reported-by: kernel test robot <[email protected]>
> Link: https://lore.kernel.org/r/[email protected]
> Signed-off-by: Benjamin Tissoires <[email protected]>

Applied as well to for-6.2/hid-bpf. Thanks,

--
Jiri Kosina
SUSE Labs