On Tue, Mar 23 2021 18:01:27 +0100, Stefan Sperling wrote:
> This switches athn(4) to the new RA Tx rate adaptation module.
> Tests on athn(4) PCI devices are welcome.
> USB devices don't need to be tested in this case Tx rate adaptation
> is taken care of by firmware.
>
> I could only test on AR9285 so far, but the result looks promising.
Some numbers from quick tests on:
athn0 at pci1 dev 0 function 0 "Atheros AR9281" rev 0x01: apic 5 int 0
athn0: AR9280 rev 2 (2T2R), ROM rev 16, address <omitted>
in hostap mode (11n), sending data to a laptop on iwn(4).
tcpbench (tcp), 10sec:
- before: Mbps varied between 0.677 and 15.187, avg 4.647
- after: Mbps varied between 10.345 and 10.623, avg 10.588
tcpbench (udp), 10sec:
- before: Rx PPS between 1444 and 1461, about 17 Mbps
athn(4) sent: 495668032 bytes sent over 10.999 seconds
- after: Rx PPS between 1021 and 1056, about 12 Mbps
athn(4) sent: 462994048 bytes sent over 10.999 seconds
(I failed to record bytes received on iwn(4) for this test, sorry)
ping -f -c 10000 from athn(4) hostap to iwn(4):
before:
10000 packets transmitted, 9982 packets received, 0.2% packet loss
round-trip min/avg/max/std-dev = 0.397/1.029/100.599/1.400 ms
after:
10000 packets transmitted, 10000 packets received, 0.0% packet loss
round-trip min/avg/max/std-dev = 0.402/1.191/10.510/0.463 ms
Subjectively, the reduced loss & more predictable latency feels much better with
eg. interactive ssh sessions.
--
Lauri Tirkkonen | lotheac @ IRCnet
