linux-kernel - Re: [PATCH v1 1/2] HID: generic: Add ->match() check to __check_hid

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:   Wed, 7 Dec 2022 10:27:37 +0100
From:   Benjamin Tissoires <benjamin.tissoires@...hat.com>
To:     "Rafael J. Wysocki" <rjw@...ysocki.net>
Cc:     Jiri Kosina <jikos@...nel.org>, Bastien Nocera <hadess@...ess.net>,
        Filipe Laíns <lains@...eup.net>,
        linux-input@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>,
        "Rafael J. Wysocki" <rafael@...nel.org>,
        Thorsten Leemhuis <regressions@...mhuis.info>
Subject: Re: [PATCH v1 1/2] HID: generic: Add ->match() check to __check_hid_generic()

On Wed, Dec 7, 2022 at 10:13 AM Rafael J. Wysocki <rjw@...ysocki.net> wrote:
>
> From: Rafael J. Wysocki <rafael.j.wysocki@...el.com>
>
> Some special HID drivers (for example, hid-logitech-hidpp) use ->match()
> callbacks to reject specific devices that otherwise would match the
> driver's device ID list, with the expectation that those devices will
> be handled by some other drivers.  However, this doesn't work if
> hid-generic is expected to bind to the given device, because its
> ->match() callback, hid_generic_match(), rejects all devices that match
> device ID lists of the other HID drivers regardless of what is returned
> by the other drivers' ->match() callbacks.

Thanks Rafael for spotting that corner case in the ->match() processing.

>
> To make it work, amend the function used by hid_generic_match() for
> checking an individual driver, __check_hid_generic(), with a check
> involving the given driver's ->match() callback, so 0 is returned
> when that callback rejects the device in question.

Shouldn't we add that logic to hid_match_device() directly in
hid-core.c instead?
It feels wrong to have a function named "hid_match_device()" and have
to manually call later "->match()" on the driver itself.

Ack on the general idea anyway.

>
> Fixes: 532223c8ac57 ("HID: logitech-hidpp: Enable HID++ for all the Logitech Bluetooth devices")
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@...el.com>
> ---
>  drivers/hid/hid-generic.c |    8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
>
> Index: linux-pm/drivers/hid/hid-generic.c
> ===================================================================
> --- linux-pm.orig/drivers/hid/hid-generic.c
> +++ linux-pm/drivers/hid/hid-generic.c
> @@ -31,7 +31,13 @@ static int __check_hid_generic(struct de
>         if (hdrv == &hid_generic)
>                 return 0;
>
> -       return hid_match_device(hdev, hdrv) != NULL;
> +       if (!hid_match_device(hdev, hdrv))
> +               return 0;
> +
> +       if (hdrv->match)
> +               return hdrv->match(hdev, false);
> +
> +       return 1;
>  }
>
>  static bool hid_generic_match(struct hid_device *hdev,
>
>
>