Re: dev.ath.0.slottime changing and why ?

Daniel Dvorak wrote:
> On 2 servers, where atheros is installed, there are problem with unexpected
> changing "slottime" IFS parametr.
>  
> If I set this parametr manually or by new way with using of athctrl.sh for
> our correct destination between them, it changes to slottime=12 for example,
> but after undefined period it changes back to slottime=9 which is default.

Something is likely changing the slot time at the net80211 layer. 
Values passed to the hal will be preserved across resets.

The ability to override the standard IFS parameters has always been 
experimental.  Feel free to fix issues you find.

>  
> Anothers 2 parametrs ack and cts timeouts is okay.
>  
> habr# ./athctrl.sh -i ath0 -d 681
> Setup IFS parameters on interface ath0 for 681 meter p-2-p link
> dev.ath.0.slottime: 9 -> 12
> dev.ath.0.acktimeout: 27 -> 27
> dev.ath.0.ctstimeout: 27 -> 27
> habr# ifconfig ath0 media OFDM/24Mbps mode 11a mediaopt adhoc
> habr# ./athctrl.sh -i ath0 -d 681
> Setup IFS parameters on interface ath0 for 681 meter p-2-p link
> dev.ath.0.slottime: 9 -> 12
> dev.ath.0.acktimeout: 27 -> 27
> dev.ath.0.ctstimeout: 27 -> 27
> 
> And another question, could somebody recommend me some ath_rate from 3
> available options and why ?
>  
> I remind that there are available these:
>  
> ath_rate_onoe - which is default for fbsd driver
> ath_rate_amrr - http://www.marlow.dk/tech/madwifi/lacage04.pdf
> ath_rate_sample - which is default for madwifi (linux driver)
>  
> After reading the pdf file, ath_rate_amrr is seems to be the best one.
>  
> Any comments ?

ath_rate_sample is the only one to use.  It's still got some issues 
(John did some fixes that are in the Linux version that I haven't 
backported yet) but outperfoms the others hands down.

	Sam