[patch 2/2] staging: line6: use after free bug requesting version

[Greg Kroah-Hartman]

On Sun, Jan 13, 2013 at 07:15:22PM +0100, Markus Grabner wrote:
> On Thursday 06 December 2012 06:18:02 Stefan Hajnoczi wrote:
> > On Wed, Dec 5, 2012 at 7:44 PM, Dan Carpenter <dan.carpenter at oracle.com> 
> wrote:
> > > diff --git a/drivers/staging/line6/driver.c
> > > b/drivers/staging/line6/driver.c index 8a5d89e..884e0d8 100644
> > > --- a/drivers/staging/line6/driver.c
> > > +++ b/drivers/staging/line6/driver.c
> > > @@ -110,7 +110,7 @@ struct message {
> > > 
> > >  */
> > >  static void line6_data_received(struct urb *urb);
> > >  static int line6_send_raw_message_async_part(struct message *msg,
> > > 
> > > -                                            struct urb *urb);
> > > +                                            struct urb *urb, int free);
> > 
> > s/int/bool/
> > 
> > >  /*
> > >  
> > >         Start to listen on endpoint.
> > > 
> > > @@ -219,24 +219,42 @@ static void line6_async_request_sent(struct urb
> > > *urb)
> > > 
> > >                 usb_free_urb(urb);
> > >                 kfree(msg);
> > >         
> > >         } else
> > > 
> > > -               line6_send_raw_message_async_part(msg, urb);
> > > +               line6_send_raw_message_async_part(msg, urb, 0);
> > > +}
> > 
> > I'd add a bool free_buffer field to struct message and simply modify
> > line6_async_request_sent() to do:
> > 
> > if (msg->free_buffer)
> >          kfree(msg->buffer);
> > 
> > Then you don't need line6_async_request_sent_free_buffer() and
> > line6_send_raw_message_async_part() doesn't need to take a bool free
> > argument since struct message already contains that information.  It
> > would make the code simpler.
> Considering the suggestions made so far, I came up with the following 
> solution: the function "line6_send_raw_message_async" now has an additional 
> argument "bool copy", which indicates whether the supplied buffer should be 
> copied into a dynamically allocated block of memory. The copy flag is also 
> stored in the "message" struct such that the temporary memory can be freed 
> when appropriate without intervention of the caller.
> 
> The patch is against linux-next since this bug should be fixed regardless of 
> the status of moving the Line6 driver out of the staging area.
> 
> Reported-by: Stefan Hajnoczi <stefanha at gmail.com>
> Signed-off-by: Markus Grabner <grabner at icg.tugraz.at>

Can you resend this in a format I can apply it in (proper Subject, not
line-wrapped, etc.)

thanks,

greg k-h