Re: [PATCH] HID: wacom: stop hardware after post-start probe failures
From: Dmitry Torokhov
Date: Wed May 27 2026 - 15:35:18 EST
Hi Myeonghun,
On Mon, May 25, 2026 at 02:53:33AM +0900, Myeonghun Pak wrote:
> wacom_parse_and_register() starts HID hardware before registering inputs
> and initializing pad LEDs/remotes. Those later steps can fail, but their
> error paths currently release Wacom resources without stopping the HID
> hardware.
>
> Route post-hid_hw_start() failures through hid_hw_stop() before
> releasing driver resources.
>
> This issue was identified during our ongoing static-analysis research while
> reviewing kernel code.
>
> Fixes: c1d6708bf0d3 ("HID: wacom: Do not register input devices until after hid_hw_start")
> Cc: stable@xxxxxxxxxxxxxxx
> Co-developed-by: Ijae Kim <ae878000@xxxxxxxxx>
> Signed-off-by: Ijae Kim <ae878000@xxxxxxxxx>
> Signed-off-by: Myeonghun Pak <mhun512@xxxxxxxxx>
> ---
> drivers/hid/wacom_sys.c | 7 ++++---
> 1 file changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/hid/wacom_sys.c b/drivers/hid/wacom_sys.c
> index 0d1c6d90fe..c824d9c224 100644
> --- a/drivers/hid/wacom_sys.c
> +++ b/drivers/hid/wacom_sys.c
> @@ -2456,16 +2456,16 @@ static int wacom_parse_and_register(struct wacom *wacom, bool wireless)
>
> error = wacom_register_inputs(wacom);
> if (error)
> - goto fail;
> + goto fail_hw_stop;
>
> if (wacom->wacom_wac.features.device_type & WACOM_DEVICETYPE_PAD) {
> error = wacom_initialize_leds(wacom);
> if (error)
> - goto fail;
> + goto fail_hw_stop;
>
> error = wacom_initialize_remotes(wacom);
> if (error)
> - goto fail;
> + goto fail_hw_stop;
> }
>
> if (!wireless) {
> @@ -2496,6 +2496,7 @@ static int wacom_parse_and_register(struct wacom *wacom, bool wireless)
> return 0;
>
> fail_quirks:
> +fail_hw_stop:
> hid_hw_stop(hdev);
> fail:
> wacom_release_resources(wacom);
I'd get rid of 'fail_quirks' and use 'fail_hw_stop' everywhere,
otherwise:
Reviewed-by: Dmitry Torokhov <dmitry.torokhov@xxxxxxxxx>
Thanks.
--
Dmitry