Re: [PATCH 2/2] staging: vt6656: Make 'rx_rate' u8 instead of u8 *

[Malcolm Priestley]

On 28/09/16 08:40, Dan Carpenter wrote:
> On Tue, Sep 27, 2016 at 02:52:49PM -0300, Martin Alonso wrote:
> > Change the type and uses of rx_rate.
> >
> > Signed-off-by: Martin Alonso <martin.alonso@xxxxxxxx>
> > ---
> >  drivers/staging/vt6656/dpc.c | 9 +++++----
> >  1 file changed, 5 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/staging/vt6656/dpc.c b/drivers/staging/vt6656/dpc.c
> > index 655f000..782b7d7 100644
> > --- a/drivers/staging/vt6656/dpc.c
> > +++ b/drivers/staging/vt6656/dpc.c
> > @@ -46,7 +46,8 @@ int vnt_rx_data(struct vnt_private *priv, struct vnt_rcb *ptr_rcb,
> >  	__le64 *tsf_time;
> >  	u32 frame_size;
> >  	int ii, r;
> > -	u8 *rx_rate, *sq, *sq_3;
> > +	u8 rx_rate;
> > +	u8 *sq, *sq_3;
> >  	u32 wbk_status;
> >  	u8 *skb_data;
> >  	u16 *pay_load_len;
> > @@ -75,7 +76,7 @@ int vnt_rx_data(struct vnt_private *priv, struct vnt_rcb *ptr_rcb,
> >
> >  	skb_data = (u8 *)skb->data;
> >
> > -	rx_rate = skb_data + 5;
> > +	rx_rate = *(skb_data + 5);
>
> It occurs to me that we don't check that skb->len is large enough here.
> We just assume it has at least 5 bytes.
>
> TODO: vt6656: verify skb->len before using it.

skb->len is always set to the tail room then trimmed by this function.

Regards

Malcolm