Date: Wed, 15 Jun 2005 22:48:31 -0700 From: Sam Leffler <sam@errno.com> To: Rong-En Fan <grafan@gmail.com> Cc: FreeBSD Current <freebsd-current@freebsd.org> Subject: Re: [Fwd: cvs commit: src/sys/modules Makefile src/sys/conf files src/sys/modules/ath_rate_sample Makefile src/sys/i386/conf NOTES] Message-ID: <42B112AF.8050708@errno.com> In-Reply-To: <6eb82e050615061535bcc722@mail.gmail.com> References: <42313286.2060206@errno.com> <423212CC.90800@centtech.com> <6eb82e050615061535bcc722@mail.gmail.com>
index | next in thread | previous in thread | raw e-mail
Rong-En Fan wrote: > On 3/12/05, Eric Anderson <anderson@centtech.com> wrote: > >>Sam Leffler wrote: >> >>>If you''re an ath user on current it'd be great to get John some >>>feedback on his rate control algorithm as it might be made the default. >>> I've not tested it with 5210 or 5211 cards and it's likely to have some >>>issues with them so beware. To use it be sure you have up to date code >>>and then specify >>> >>>device ath_rate_sample >>> >>>instead of the normal ath_rate_onoe. >> >> >>So far, I see no additional performance, in fact, upload speeds are about 15-20% less compared to the onoe stuff. It's stable and seems to work ok for me however. > > > I have an 5212 card on IBM X31. Setting at the same location, the windows > Atheros driver works very well. However, using FreeBSD the network sometimes > hang and saw some 'ath0 device timeout' at console. But, when I switched to > ath_sample (orig is onoe) it works very well. Althought I didn't > check the speed, > the ssh connection does not hang any more. BTW, ifconfig ath0 list ap shows > that S:N is about 8:0 to 13:0. My current is Jun 15. The sample rate control algorithm is much more responsive than the onoe algorithm. Can't comment on the timeouts w/o more info. Comparing the Atheros ndis driver isn't entirely fair as it will automatically switch between normal operation and XR mode at those signal levels. Samhome | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?42B112AF.8050708>