linux-kernel - Re: [PATCH] media: uvcvideo: Modified uvc_ctrl_fill_xu

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:   Wed, 20 Sep 2023 10:27:59 +0800
From:   Jerry Liu <jerry.liu@...hnexion.com>
To:     Laurent Pinchart <laurent.pinchart@...asonboard.com>
Cc:     mchehab@...nel.org, linux-media@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] media: uvcvideo: Modified uvc_ctrl_fill_xu_info
 'kmalloc' to 'kzalloc'

Hi Laurent,

Thanks for your comment!
I'm sorry I have a confusing description for that.

You're right it's sure to return the error if it gets an error length 
from UVC. But I think it still can work despite receiving the length is 1.
In uvc_query_ctrl, it will return an error but the value is 1 not '-EPIPE', 
I think even though the length is less than 1, it still continues to execute 
the XU command.
However, I found it will receive the wrong data from uvc_query_ctrl because 
it only assigns 1-byte, not 2-byte value to the data array.

For example:
if data array is not allocated with zero bytes:

                        asssigned value of  
data[0]  |  data[1]       1 byte length       data[0]  |  data[1] 
---------------------  -------------------->  --------------------- 
  0xcc   |   0xcc                               0x01   |   0xcc

then in uvc_ctrl_fill_xu_info, 'info->size' will get wrong size from data array.

Somtimes, the data array is allocated with zero bytes:

                        asssigned value of  
data[0]  |  data[1]       1 byte length       data[0]  |  data[1] 
---------------------  -------------------->  --------------------- 
  0x00   |   0x00                               0x01   |   0x00

In this case, 'info->size' will get correct size from data array.

On Fri, Sep 15, 2023 at 10:04:49PM +0300, Laurent Pinchart wrote:
> Hi Jerry,
> 
> Thank you for the patch.
> 
> On Fri, Sep 15, 2023 at 09:12:14AM -0700, Jerry Liu wrote:
> > If the request length of UVC XU is 1 (even though this is illegal), due
> > to 'data' may be the non-zero value, UVC_GET_LEN could potentially result
> > in a length that is not 1 because of the high byte is not zero. In order
> > to ensure that 2-byte data array is set to 0, 'kmalloc' is modified to 'kzalloc'.
> 
> I don't think this can happen. The call to uvc_query_ctrl(UVC_GET_LEN)
> is given a length of 2. If the device responds with less than two bytes,
> the function will return an error, and uvc_ctrl_fill_xu_info() will
> propagate the error to the caller, without accessing the data array.
> 
> > 
> > Signed-off-by: Jerry Liu <jerry.liu@...hnexion.com>
> > ---
> >  drivers/media/usb/uvc/uvc_ctrl.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/media/usb/uvc/uvc_ctrl.c b/drivers/media/usb/uvc/uvc_ctrl.c
> > index 5e9d3da862dd..054bc14f7a58 100644
> > --- a/drivers/media/usb/uvc/uvc_ctrl.c
> > +++ b/drivers/media/usb/uvc/uvc_ctrl.c
> > @@ -2088,7 +2088,7 @@ static int uvc_ctrl_fill_xu_info(struct uvc_device *dev,
> >  	u8 *data;
> >  	int ret;
> >  
> > -	data = kmalloc(2, GFP_KERNEL);
> > +	data = kzalloc(2, GFP_KERNEL);
> >  	if (data == NULL)
> >  		return -ENOMEM;
> >  
> 
> -- 
> Regards,
> 
> Laurent Pinchart