[PATCH] WLAN: --power maxperf shot into the dark
freerunner at newkirk.us
Mon Jan 26 04:22:26 CET 2009
Have you run a cycle of your bouncing-ap-channel tests with maxperf?
On Sun, 25 Jan 2009 23:42:40 -0200, Werner Almesberger
<werner at openmoko.org> wrote:
> John Sullivan wrote:
>> What was happening for me is that I got a lease and could do dns lookups
>> but could not ping the gateway or any other host -- there would be one
>> ping attempt that did not return.
> Hmm, that's very odd. There shouldn't really be a difference between
> DHCP, DNS, and ping - at least not one that goes away if you use
> --power maxperf. Perhaps this depends also on time. Do DHCP and DNS
> continue to work after you had ping failures ?
> It would be interesting to know what your WLAN environment looks like
> (iwlist scan) and what "iwconfig eth0" says while you're experiencing
> this sort of problems. Also, if you know anything about the access
> point(s), such as brand or firmware, that would be useful.
> One known issue is that the AR6k sometimes tunes to the wrong
> frequency for some reason, which could explain lossy communication.
> - Werner
> By the way, the problem with maxperf is that it makes WLAN burn power
> like there's no tomorrow. It roughly doubles the total power consumption
> As I've mentioned before, telling the WLAN module not to sleep
> with "wmiconfig -i eth0 --power maxperf" reportedly improves
> network stability in some situations where otherwise disconnects
> were frequent. (Sorry if this sounds rather vague. I don't have
> much real data on this phenomenon. More details would be welcome.)
http://newkirk.us/om (FR stuff)
More information about the devel