linux-kernel - Re: [PATCH v2] HID: input: fix a4tech horizontal wheel custom usage

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 11 Jun 2019 16:42:37 +0200
From:   Benjamin Tissoires <benjamin.tissoires@...hat.com>
To:     Nicolas Saenz Julienne <nsaenzjulienne@...e.de>
Cc:     Jiri Kosina <jikos@...nel.org>,
        Dmitry Torokhov <dmitry.torokhov@...il.com>, wbauer@....at,
        "open list:HID CORE LAYER" <linux-input@...r.kernel.org>,
        lkml <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2] HID: input: fix a4tech horizontal wheel custom usage

On Tue, Jun 11, 2019 at 2:13 PM Nicolas Saenz Julienne
<nsaenzjulienne@...e.de> wrote:
>
> Some a4tech mice use the 'GenericDesktop.00b8' usage to inform whether
> the previous wheel report was horizontal or vertical. Before
> c01908a14bf73 ("HID: input: add mapping for "Toggle Display" key") this
> usage was being mapped to 'Relative.Misc'. After the patch it's simply
> ignored (usage->type == 0 & usage->code == 0). Which ultimately makes
> hid-a4tech ignore the WHEEL/HWHEEL selection event, as it has no
> usage->type.
>
> We shouldn't rely on a mapping for that usage as it's nonstandard and
> doesn't really map to an input event. So we bypass the mapping and make
> sure the custom event handling properly handles both reports.
>
> Fixes: c01908a14bf73 ("HID: input: add mapping for "Toggle Display" key")
> Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@...e.de>
> ---

Yep, I like this one much better.

>
> NOTE: I CC'd Wolfgang as he's the one who can test this.

I'll wait for Wolfram to confirm that the patch works before pushing then.

>
> Changes since v1:
>   - new approach, moved fix into hid-a4tech
>
>  drivers/hid/hid-a4tech.c | 30 +++++++++++++++++++++++++++---
>  1 file changed, 27 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/hid/hid-a4tech.c b/drivers/hid/hid-a4tech.c
> index 98bf694626f7..3a8c4a5971f7 100644
> --- a/drivers/hid/hid-a4tech.c
> +++ b/drivers/hid/hid-a4tech.c
> @@ -23,12 +23,36 @@
>  #define A4_2WHEEL_MOUSE_HACK_7 0x01
>  #define A4_2WHEEL_MOUSE_HACK_B8        0x02
>
> +#define A4_WHEEL_ORIENTATION   (HID_UP_GENDESK | 0x000000b8)
> +
>  struct a4tech_sc {
>         unsigned long quirks;
>         unsigned int hw_wheel;
>         __s32 delayed_value;
>  };
>
> +static int a4_input_mapping(struct hid_device *hdev, struct hid_input *hi,
> +                           struct hid_field *field, struct hid_usage *usage,
> +                           unsigned long **bit, int *max)
> +{
> +       struct a4tech_sc *a4 = hid_get_drvdata(hdev);
> +
> +       if (a4->quirks & A4_2WHEEL_MOUSE_HACK_B8 &&
> +           usage->hid == A4_WHEEL_ORIENTATION) {
> +               /*
> +                * We do not want to have this usage mapped to anything as it's
> +                * nonstandard and doesn't really behave like an HID report.
> +                * It's only selecting the orientation (vertical/horizontal) of
> +                * the previous mouse wheel report. The input_events will be
> +                * generated once both reports are recorded in a4_event().
> +                */
> +               return -1;

You seem to be extra precocious here. This is basically what the
current hid-input.c mapping does. But I think it is for the best,
given that this time you are locking the behavior of this reserved
usage.

Cheers,
Benjamin

> +       }
> +
> +       return 0;
> +
> +}
> +
>  static int a4_input_mapped(struct hid_device *hdev, struct hid_input *hi,
>                 struct hid_field *field, struct hid_usage *usage,
>                 unsigned long **bit, int *max)
> @@ -52,8 +76,7 @@ static int a4_event(struct hid_device *hdev, struct hid_field *field,
>         struct a4tech_sc *a4 = hid_get_drvdata(hdev);
>         struct input_dev *input;
>
> -       if (!(hdev->claimed & HID_CLAIMED_INPUT) || !field->hidinput ||
> -                       !usage->type)
> +       if (!(hdev->claimed & HID_CLAIMED_INPUT) || !field->hidinput)
>                 return 0;
>
>         input = field->hidinput->input;
> @@ -64,7 +87,7 @@ static int a4_event(struct hid_device *hdev, struct hid_field *field,
>                         return 1;
>                 }
>
> -               if (usage->hid == 0x000100b8) {
> +               if (usage->hid == A4_WHEEL_ORIENTATION) {
>                         input_event(input, EV_REL, value ? REL_HWHEEL :
>                                         REL_WHEEL, a4->delayed_value);
>                         input_event(input, EV_REL, value ? REL_HWHEEL_HI_RES :
> @@ -131,6 +154,7 @@ MODULE_DEVICE_TABLE(hid, a4_devices);
>  static struct hid_driver a4_driver = {
>         .name = "a4tech",
>         .id_table = a4_devices,
> +       .input_mapping = a4_input_mapping,
>         .input_mapped = a4_input_mapped,
>         .event = a4_event,
>         .probe = a4_probe,
> --
> 2.21.0
>