On Wed, Jan 12, 2011 at 11:54 AM, John Clark <jcl...@metricsystems.com>wrote:
> Roback, Joe schrieb: > >> I would agree, lately, I am seeing a bunch of these in my logs: >> >> ath: Failed to stop TX DMA in 100 msec after killing last frame >>> >> >> but my 802.11n 40MHz performance has been the best its been in 6 months >> ;-) >> >> >> On Tue, 11 Jan 2011 16:35:10 -0600, "J. Ryan Earl" <o...@jryanearl.us> >> wrote: >> >>> >>> >>> What's the exact error? I've see something similar in my dmesg: >>> >>> ath: Failed to stop TX DMA in 100 msec after killing last frame >>> >>> However, I get great performance regardless. Able to pull 150 mbit/s >>> actual >>> in a 2x2 MIMO configuration. >>> >> > > Are you two using 'station' mode and associating with a OpenWRT AP. Because > I'm getting abysmal performance. > Running iPerf in UDP mode, frequently gives 10% packet loss at 10 mbs... > > Theplatform I'm using is pretty close to the Atheros AP83 reference design. > I also have tested with a non-OpenWRT > station, and get much better performance. It is in "Access Point" mode. I measured throughput not with iperf, but by transferring data from a server on the LAN-side to an associated wireless device (a laptop in my case with a Intel 5100 AGN card) and inspecting the network traffic on both ends of the transport. I would suggest generating the traffic outside of OpenWRT so as to not impose any additional overhead that could affect your performance results. -JR
_______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
