On Fri November 19 2010 21:28:52 Jonathan Guerin wrote:
> I'm not sure if I'm understanding the way the data is represented. I'm
> pulling it via nl80211 (or even if just using iw to poll the values
> regularly). They just appear to go back and forwards in time, as well
> as being very large values
The old survey implementation was broken and returned nonsense data.
Clear cycle counters and survey data on reset. Since the cycle counters easily
overflow it's better to keep a local version of collected survey data (in ms
resolution, instead of clockrate) and update this every time survey is
re
