Messages in this thread | | | Date | Tue, 14 Jan 2014 18:52:53 +0000 | From | Jonathan Cameron <> | Subject | Re: [PATCH 2/3] iio: mxs-lradc: fix invalid channel number detection |
| |
On 13/01/14 23:43, Alexandre Belloni wrote: > On 13/01/2014 21:25, Marek Vasut wrote: >> On Monday, January 13, 2014 at 05:02:02 PM, Alexandre Belloni wrote: >>> 16 would be accepted as a channel number but it is invalid. It doesn't >>> really have any effect as mxs_lradc_read_raw is called from a "controlled" >>> environment so it it only gets values going from 0 to 15. >>> >>> Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com> >> >> Why don't you remove the check entirely then ? >> > > I'm not quite sure the inkernel API is sanitizing the input correctly > but maybe I didn't check enough. Maybe Jonathan can comment ? > Unless we have a bug (more than possible as this stuff isn't heavily used yet), it should be impossible to get the required reference to a channel that doesn't exist. Thus I don't 'think' the check is needed. Feel free to write a test case to prove me wrong ;)
Jonathan
| |