linux-kernel - Re: [PATCH] Input: elantech

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] [day] [month] [year] [list]

Date:   Thu, 27 Feb 2020 13:59:49 +0100
From:   Benjamin Tissoires <benjamin.tissoires@...hat.com>
To:     Bernd Edlinger <bernd.edlinger@...mail.de>,
        "Dave.Wang" <dave.wang@....com.tw>, jingle <jingle.wu@....com.tw>
Cc:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-input@...r.kernel.org" <linux-input@...r.kernel.org>,
        Dmitry Torokhov <dmitry.torokhov@...il.com>,
        Kai-Heng Feng <kai.heng.feng@...onical.com>,
        Enrico Weigelt <info@...ux.net>,
        Allison Randal <allison@...utok.net>,
        Aaron Ma <aaron.ma@...onical.com>
Subject: Re: [PATCH] Input: elantech - fix x_max/y_max values

Hi Bernd,

[adding Dave and Jingle]


On Thu, Feb 27, 2020 at 12:59 PM Bernd Edlinger
<bernd.edlinger@...mail.de> wrote:
>
> Since 37548659bb2 moved the querying of the x_max/y_max

Hmm, I am pretty sure checkpatch.pl should have complained here: the
correct way of mentioning a previous commit is:
'commit <12+ chars of sha1> ("<title line>")' - ie: 'commit
37548659bb22 ("Input: elantech - query the min/max information
beforehand too")'

> values from elantech_set_input_params to elantech_query_info,
> the returned x_max/y_max values are different than before,
> at least for some firmware versions.
>
> The reason is likely that this is now done before
> elantech_set_absolute_mode does run.  So it may happen that
> the returned values are exactly half of what they used to be,

That is weird. Can we get confirmation from Dave or Jingle about that?

> which makes input_report_abs in PS/2 mode report ABS_X values which
> exceed the x_max value, which is very annoying since the mouse stops
> to move then, and ABS_Y value become negative, which is benign.
>
> This was observed with a MSI GX70 laptop:
>
> elantech: assuming hardware version 3 (with firmware version 0x250f01)
> elantech: Synaptics capabilities query result 0x18, 0x17, 0x0b.
> elantech: Elan sample query result 05, 0e, 00
> input: ETPS/2 Elantech Touchpad as /devices/platform/i8042/serio...
>
> Correct this by doubling the returned x_max and y_max
> value for this specific firmware version.
>
> Fixes: 37548659bb2 ("Input: elantech - query the min/max
> information beforehand too")

Stephen will complain here: see multiple google results when you
search for "lkml linux-next: Fixes tag needs some work"

Basically:
- SHA1 should be at least 12 digits long
- the commit title should not be split, even if it gets longer than 80 columns.

Cheers,
Benjamin

>
> Signed-off-by: Bernd Edlinger <bernd.edlinger@...mail.de>
> ---
>  drivers/input/mouse/elantech.c | 4 ++++
>  1 file changed, 4 insertions(+)
>
> diff --git a/drivers/input/mouse/elantech.c b/drivers/input/mouse/elantech.c
> index 2d8434b..3399db8 100644
> --- a/drivers/input/mouse/elantech.c
> +++ b/drivers/input/mouse/elantech.c
> @@ -1720,6 +1720,10 @@ static int elantech_query_info(struct psmouse *psmouse,
>
>                 info->x_max = (0x0f & param[0]) << 8 | param[1];
>                 info->y_max = (0xf0 & param[0]) << 4 | param[2];
> +               if (info->fw_version == 0x250f01) {
> +                       info->x_max <<= 1;
> +                       info->y_max <<= 1;
> +               }
>                 break;
>
>         case 4:
> --
> 1.9.1
>