ath0 performence issues "ar9300_Stub_GetCTSTimeout" "ar9300_Stub_GetCTSTimeout

[Miguel Clara]

I've just upgrade to the latest 10/stable.

trying to git clone a repo was giving me horrible speed, and I honestly
tough it was at their side, until I noticed issues while browsing and this
in dmesg:

ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetCTSTimeout: called
ar9300_Stub_GetAntennaSwitch: called
ar9300_Stub_GetAntennaSwitch: called

Speedtest either gives me and ok result or very poor, and pings also prove
that something is not quite right:
64 bytes from 216.58.210.110: icmp_seq=7 ttl=59 time=18.096 ms
64 bytes from 216.58.210.110: icmp_seq=8 ttl=59 time=149.083 ms
64 bytes from 216.58.210.110: icmp_seq=9 ttl=59 time=18.376 ms
64 bytes from 216.58.210.110: icmp_seq=10 ttl=59 time=19.084 ms
64 bytes from 216.58.210.110: icmp_seq=11 ttl=59 time=18.305 ms
64 bytes from 216.58.210.110: icmp_seq=12 ttl=59 time=18.140 ms
64 bytes from 216.58.210.110: icmp_seq=13 ttl=59 time=16.700 ms
64 bytes from 216.58.210.110: icmp_seq=14 ttl=59 time=105.554 ms
64 bytes from 216.58.210.110: icmp_seq=15 ttl=59 time=34.336 ms
64 bytes from 216.58.210.110: icmp_seq=16 ttl=59 time=76.410 ms
64 bytes from 216.58.210.110: icmp_seq=17 ttl=59 time=34.400 ms
64 bytes from 216.58.210.110: icmp_seq=18 ttl=59 time=73.521 ms
64 bytes from 216.58.210.110: icmp_seq=19 ttl=59 time=154.728 ms
64 bytes from 216.58.210.110: icmp_seq=20 ttl=59 time=126.120 ms
64 bytes from 216.58.210.110: icmp_seq=21 ttl=59 time=170.920 ms
64 bytes from 216.58.210.110: icmp_seq=22 ttl=59 time=137.330 ms
64 bytes from 216.58.210.110: icmp_seq=23 ttl=59 time=17.424 ms

(tried other host with similar results)

It seems to settle for a while but randomly comes back..

uname -a
FreeBSD r2d2 10.1-STABLE FreeBSD 10.1-STABLE #7 r277979M: Sat Jan 31
16:05:30 WET 2015     root at r2d2:/usr/obj/usr/src/sys/GENERIC  amd64