linux-kernel - Re: [PATCH] net: usb: dm9601: fix uninitialized variable use in dm9601_mdio

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]

Message-ID: <e3ad2ed3-8859-4740-9126-9757f8d13a0e@gmail.com>
Date:   Mon, 9 Oct 2023 23:48:16 +0200
From:   Javier Carrasco <javier.carrasco.cruz@...il.com>
To:     Peter Korsgaard <peter@...sgaard.com>
Cc:     "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        syzbot+1f53a30781af65d2c955@...kaller.appspotmail.com,
        netdev@...r.kernel.org, linux-usb@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] net: usb: dm9601: fix uninitialized variable use in
 dm9601_mdio_read

Hi Peter,

On 09.10.23 22:48, Peter Korsgaard wrote:
> 
>  > syzbot has found an uninit-value bug triggered by the dm9601 driver [1].
>  > This error happens because the variable res is not updated if the call
>  > to dm_read_shared_word returns an error or if no data is read (see
>  > __usbnet_read_cmd()). In this particular case -EPROTO was returned and
>  > res stayed uninitialized.
> 
>  > This can be avoided by checking the return value of dm_read_shared_word
>  > and returning an error if the read operation failed or no data was read.
> 
>  > [1] https://syzkaller.appspot.com/bug?extid=1f53a30781af65d2c955
> 
>  > Signed-off-by: Javier Carrasco <javier.carrasco.cruz@...il.com>
>  > Reported-and-tested-by: syzbot+1f53a30781af65d2c955@...kaller.appspotmail.com
>  > ---
>  >  drivers/net/usb/dm9601.c | 9 ++++++++-
>  >  1 file changed, 8 insertions(+), 1 deletion(-)
> 
>  > diff --git a/drivers/net/usb/dm9601.c b/drivers/net/usb/dm9601.c
>  > index 48d7d278631e..e223daa93229 100644
>  > --- a/drivers/net/usb/dm9601.c
>  > +++ b/drivers/net/usb/dm9601.c
>  > @@ -222,13 +222,20 @@ static int dm9601_mdio_read(struct net_device *netdev, int phy_id, int loc) !
>  >  	struct usbnet *dev = netdev_priv(netdev);
>  
>  >  	__le16 res;
>  > +	int err;
>  
>  >  	if (phy_id) {
>  >  		netdev_dbg(dev->net, "Only internal phy supported\n");
>  >  		return 0;
>  >  	}
>  
>  > -	dm_read_shared_word(dev, 1, loc, &res);
>  > +	err = dm_read_shared_word(dev, 1, loc, &res);
>  > +	if (err <= 0) {
>  > +		if (err == 0)
>  > +			err = -ENODATA;
> 
> Looking at dm_read(), it doesn't look like we can end up here with err
> == 0, but OK.You are right, I just simulated the err = 0 value from
__usbnet_read_cmd() and in this case it is harmless because dm_read is
called with length != 0 (in dm_read_shared_word either 1 or 2) and
therefore it returns -EINVAL. A silly case where this would fail would
be if length == 0, which would be wrong anyways.

So I will remove the err == 0 case for v2.

Thanks a lot for your feedback.

> Acked-by: Peter Korsgaard <peter@...sgaard.com>
> 
Best regards,
Javier Carrasco