[Voyage-linux] A madwifi story...

[Edwin Whitelaw]

All,

Some weird problems I've had over the last couple of days and their
solution may help someone else....

WRAP running Voyage 0.2 with svn 1527 updates
One CM9 and one Engenius 2511 radio card - note that the CM9 is wifi1

schaeffer:~# dmesg | grep wifi
hostap_pci: Registered netdevice wifi0
wifi0: Original COR value: 0x1
wifi0: NIC: id=0x8013 v1.0.0
wifi0: PRI: id=0x15 v1.1.1
wifi0: STA: id=0x1f v1.8.2
wifi0: Intersil Prism2.5 PCI: mem=0xa0000000, irq=9
wifi0: registered netdevice wlan0
wifi1: 11a rates: 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
wifi1: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps
wifi1: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps
24Mbps 36Mbps 48Mbps 54Mbps
wifi1: turboA rates: 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
wifi1: turboG rates: 6Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
wifi1: H/W encryption support: WEP AES AES_CCM TKIP
wifi1: mac 5.9 phy 4.3 radio 3.6
wifi1: Use hw queue 1 for WME_AC_BE traffic
wifi1: Use hw queue 0 for WME_AC_BK traffic
wifi1: Use hw queue 2 for WME_AC_VI traffic
wifi1: Use hw queue 3 for WME_AC_VO traffic
wifi1: Use hw queue 8 for CAB traffic
wifi1: Use hw queue 9 for beacons
wifi1: Atheros 5212: mem=0x80000000, irq=12

I originally configured the CM9 as ath0 on wifi1 and everything came up
fine, associated to the AP, etc.

Network diagram is:

Internet--eth0--------WRAP-AP-ath1(CM9)<----->ath0(CM9)-WRAP Client
        10.220.200.21         10.220.203.1    10.220.203.2

I could not ping the AP(.203.1) end of my link from the client but
*could* ping the eth0 port(.200.21) of the AP and ping out into my local
net and Internet.  After about 30 seconds, that stopped until I again
pinged the .200.21 address (from the client).  If I kept pinging, the
client worked fine indefinitely, ie, I was able to do apt-get update,
upgrade, etc. The other very strange thing was that if I did "ssh
10.220.203.1" from the client (10.220.203.2), I would end up logging
back into the lcient rather than the AP.  Apparently some strange
arp-ness going on.

As it turned out, even though there was no ath0 and wifi1 was detected
as the CM9 card, the fix was to assign ath1 to wifi1 and all the
weirdness went away.

I will leave it to others to explain why this happened but thought it
might spare someone considerable time to post this.  As far as I could
find, it's not in the Wiki or otherwise noted that the assignment of
ath0 to wifi1 should be a problem.  If there is only one Atheros card in
the radio, shouldn't either ath0 or ath1 been an acceptable assignment
to the wifi1 device?

Regards,

Edwin

--
<=+=+=+==+=+=+==+=+=+=+=+=+=+=+=>
Edwin Whitelaw, P.E.
New River Valley Unwired, LLC
2200 Lonesome Dove Dr
Christiansburg, VA 24073
540-239-0318