On Tue, Oct 26, 2021 at 03:12:28PM +0100, Ian Abbott wrote:
> On 25/10/2021 12:45, Johan Hovold wrote:
> > The driver uses endpoint-sized USB transfer buffers but had no sanity
> > checks on the sizes. This can lead to zero-size-pointer dereferences or
> > overflowed transfer buffers in ni6501_port_command() and
> > ni6501_counter_command() if a (malicious) device has smaller max-packet
> > sizes than expected (or when doing descriptor fuzz testing).
> >
> > Add the missing sanity checks to probe().
> >
> > Fixes: a03bb00e50ab ("staging: comedi: add NI USB-6501 support")
> > Cc: stable@xxxxxxxxxxxxxxx # 3.18
> > Cc: Luca Ellero <luca.ellero@xxxxxxxxxxxxxxxx>
> > Signed-off-by: Johan Hovold <johan@xxxxxxxxxx>
> > ---
> > drivers/comedi/drivers/ni_usb6501.c | 8 ++++++++
> > 1 file changed, 8 insertions(+)
> >
> > diff --git a/drivers/comedi/drivers/ni_usb6501.c b/drivers/comedi/drivers/ni_usb6501.c
> > index 5b6d9d783b2f..eb2e5c23f25d 100644
> > --- a/drivers/comedi/drivers/ni_usb6501.c
> > +++ b/drivers/comedi/drivers/ni_usb6501.c
> > @@ -144,6 +144,10 @@ static const u8 READ_COUNTER_RESPONSE[] = {0x00, 0x01, 0x00, 0x10,
> > 0x00, 0x00, 0x00, 0x02,
> > 0x00, 0x00, 0x00, 0x00};
> >
> > +/* Largest supported packets */
> > +static const size_t TX_MAX_SIZE = sizeof(SET_PORT_DIR_REQUEST);
> > +static const size_t RX_MAX_SIZE = sizeof(READ_PORT_RESPONSE);
> > +
> > enum commands {
> > READ_PORT,
> > WRITE_PORT,
> > @@ -486,12 +490,16 @@ static int ni6501_find_endpoints(struct comedi_device *dev)
> > ep_desc = &iface_desc->endpoint[i].desc;
> >
> > if (usb_endpoint_is_bulk_in(ep_desc)) {
> > + if (usb_endpoint_maxp(ep_desc) < RX_MAX_SIZE)
> > + continue;
> > if (!devpriv->ep_rx)
> > devpriv->ep_rx = ep_desc;
> > continue;
> > }
> >
> > if (usb_endpoint_is_bulk_out(ep_desc)) {
> > + if (usb_endpoint_maxp(ep_desc) < TX_MAX_SIZE)
> > + continue;
> > if (!devpriv->ep_tx)
> > devpriv->ep_tx = ep_desc;
> > continue;
> >
>
> Perhaps it should return an error if the first encountered bulk-in
> endpoint has the wrong size or the first encountered bulk-out endpoint
> has the wrong size. Something like:
>
> if (usb_endpoint_is_bulk_in(ep_desc)) {
> if (!devpriv->ep_rx) {
> if (usb_endpoint_maxp(ep_desc) < RX_MAX_SIZE)
> break;
> }
> continue;
This is too convoluted, but I can move the max-packet sanity checks
after the endpoint look-ups instead.
It doesn't really matter in the end as the real devices presumably only
have two bulk endpoints.
Johan
