On Sat, Sep 30, 2017 at 11:04:35AM -0700, Deepa Dinamani wrote:
> On Sat, Sep 30, 2017 at 1:15 AM, Dan Carpenter <dan.carpenter@xxxxxxxxxx> wrote:
> > There used to be a test against "if (param->sglen > MAX_SGLEN)" but it
> > was removed during a refactor. It leads to an integer overflow and a
> > stack overflow in test_queue() if we try to create a too large urbs[]
> > array on the stack.
> >
> > There is a second integer overflow in test_queue() as well if
> > "param->iterations" is too high. I don't immediately see that it's
> > harmful but I've added a check to prevent it and silence the static
> > checker warning.
> >
> > Fixes: 18fc4ebdc705 ("usb: misc: usbtest: Remove timeval usage")
> > Signed-off-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> >
> > diff --git a/drivers/usb/misc/usbtest.c b/drivers/usb/misc/usbtest.c
> > index eee82ca55b7b..113e38bfe0ef 100644
> > --- a/drivers/usb/misc/usbtest.c
> > +++ b/drivers/usb/misc/usbtest.c
> > @@ -1964,6 +1964,9 @@ test_queue(struct usbtest_dev *dev, struct usbtest_param_32 *param,
> > int status = 0;
> > struct urb *urbs[param->sglen];
> >
> > + if (!param->sglen || param->iterations > UINT_MAX / param->sglen)
>
> Would adding a comment here that we are checking for overflow of
> context.count make sense?
> It takes some time to figure out that context.count is unsigned and
> hence the check is against UINT_MAX here.
To be honest, this is a very typical sort of integer overflow test and
it looks pretty self explanatory to me. We're avoiding a divide by zero
bug and we're making sure the result isn't more than UINT_MAX.
It has nothing to do with context.count. param->iterations and
param->sglen are both unsinged int, so if we do "param->iterations *
param->sglen" that will overflow.
regards,
dan carpenter
--
To unsubscribe from this list: send the line "unsubscribe kernel-janitors" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
