Re: [PATCH] Input-polldev:Check poll_interval range while registering input device

[Dmitry Torokhov <dmitry.torokhov@gmail.com>]

On Fri, Jul 17, 2015 at 07:11:04PM +0530, Anshul Garg wrote:
> Hi Mr. Dmitry
> ,
> 
> 
> 
> On Fri, Jul 17, 2015 at 1:05 AM, Dmitry Torokhov
> <dmitry.torokhov@gmail.com> wrote:
> > Hi Anshul,
> >
> > On Thu, Jul 16, 2015 at 10:57:13AM -0700, Anshul Garg wrote:
> >> Added check for poll_interval range in input_register_polled_device
> >> function similar to input_polldev_set_poll function.
> >>
> >> Signed-off-by: Anshul Garg <aksgarg1989@gmail.com>
> >> ---
> >>  drivers/input/input-polldev.c |    6 ++++++
> >>  1 file changed, 6 insertions(+)
> >>
> >> diff --git a/drivers/input/input-polldev.c b/drivers/input/input-polldev.c
> >> index 3664f81..0e88a9e 100644
> >> --- a/drivers/input/input-polldev.c
> >> +++ b/drivers/input/input-polldev.c
> >> @@ -309,6 +309,12 @@ int input_register_polled_device(struct input_polled_dev *dev)
> >>
> >>       input_set_drvdata(input, dev);
> >>       INIT_DELAYED_WORK(&dev->work, input_polled_device_work);
> >> +
> >> +     if (dev->poll_interval > dev->poll_interval_max ||
> >> +                     dev->poll_interval < dev->poll_interval_min) {
> >> +             pr_err("Poll interval Range incorrect\n");
> >> +             return -EINVAL;
> >> +     }
> >
> > What driver was this change tested with?
> I have not checked this change for any driver.
> 

Yes, I know that you did not test it because if you would have tested it
you would have noticed that it breaks existing devices, since hardly
any of them set dev->poll_interval_max and instead rely on the polldev
code to apply reasonable defaults.

There have been a few patches that had errors that would have been
easily caught if you tested the changes. Please next time you submit a
patch mention how you tested it to make sure it works as intended.

Thanks.

-- 
Dmitry