linux-kernel - Re: [PATCH] HID: logitech-hidpp: Check maxfield in hidpp_get_report

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]

Message-ID: <2026010956-anteater-pungent-d5b6@gregkh>
Date: Fri, 9 Jan 2026 12:14:43 +0100
From: Greg KH <gregkh@...uxfoundation.org>
To: Günther Noack <gnoack@...gle.com>
Cc: Filipe Laíns <lains@...eup.net>,
	Bastien Nocera <hadess@...ess.net>, Jiri Kosina <jikos@...nel.org>,
	Benjamin Tissoires <bentiss@...nel.org>, stable@...r.kernel.org,
	linux-input@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] HID: logitech-hidpp: Check maxfield in
 hidpp_get_report_length()

On Fri, Jan 09, 2026 at 11:59:12AM +0100, Günther Noack wrote:
> Do not crash when a report has no fields.
> 
> Fake USB gadgets can send their own HID report descriptors and can define report
> structures without valid fields.  This can be used to crash the kernel over USB.
> 
> Cc: stable@...r.kernel.org
> Signed-off-by: Günther Noack <gnoack@...gle.com>
> ---
>  drivers/hid/hid-logitech-hidpp.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/hid/hid-logitech-hidpp.c b/drivers/hid/hid-logitech-hidpp.c
> index 9ced0e4d46ae..919ba9f50292 100644
> --- a/drivers/hid/hid-logitech-hidpp.c
> +++ b/drivers/hid/hid-logitech-hidpp.c
> @@ -4316,6 +4316,9 @@ static int hidpp_get_report_length(struct hid_device *hdev, int id)
>  	if (!report)
>  		return 0;
>  
> +	if (!report->maxfield)
> +		return 0;

Combine this with the if() above this?

And if we are going to be handling "malicious" USB devices, be careful,
you are just moving the target lower down, you also need to audit ALL
data coming from the device, not just the descriptors.  I'm all for
this, just realize that this is a change in how Linux treats devices
(and all other operating systems as well.)

For now, we strongly recommend not allowing "untrusted" devices to bind
to your system if this is a threat model you care about.

Not to reject this, or your other patch like this, just letting you
know.

thanks,

greg k-h