RT3573 wpa_supplicant: "nl80211: Delete station $my_ap" after "SME: Send 20/40 BSS Coexistence to $my_ap"

RT3573 wpa_supplicant: "nl80211: Delete station $my_ap" after "SME: Send 20/40 BSS Coexistence to $my_ap"

[Peter Stuge]

Xose Vazquez Perez wrote:
> 0x0846, 0x9012 NETGEAR WNDA4100 N900 Wireless Dual Band USB Adapter

I'm testing the WNDA4100 with torvalds/linux.git and as of a day or
two ago and wpa_supplicant-2.0.

I haven't measured throughput. Latency is really bad for interactive
connections, but at least the packets are flowing on all bands. Yay!

Ping to the access point is around 3 ms with the odd 20+ ms when the
kernel goes off to do something else.


Here's my problem:

After wpa_supplicant -Dnl80211 -ddd sees scan results and says
"SME: Send 20/40 BSS Coexistence to $my_ap" and that frame is acked,
wpa_supplicant seems to receive a "link up" event which causes it to
delete the existing association for no good reason. It blacklists my
ap so I need to reconnect manually with iw dev wlan1 connect -w $ssid.

(Oh, and iw dev wlan1 connect -w consistently says "Unspecified failure"
even though the connection is established fine by wpa_supplicant. Other
problem though.)

My AP is hidden, so it isn't included in the scan results. But
wpa_supplicant is connected to it fine when the scan results arrive.

wpa_supplicant does *not* disconnect if I manually initiate a scan
when there is no other link activity, this only happens for
"Scan requested (ret=0) - scan timeout 30 seconds" scans.

--8<-- wpa_supplicant -ddd log snippet with MAC address replaced with $my_ap.
wlan1: New scan results available
SME: Send 20/40 BSS Coexistence to $my_ap
nl80211: Send Action frame (ifindex=6, freq=2417 MHz wait=0 ms no_cck=0)
nl80211: CMD_FRAME freq=2417 wait=0 no_cck=0 no_ack=0 offchanok=1
nl80211: Frame TX command accepted; cookie 0xffff880039f70700
wlan1: Checking for other virtual interfaces sharing same radio (phy1) in event_scan_results
nl80211: Event message available
nl80211: MLME event 60
nl80211: MLME event frame - hexdump(len=34): d0 00 da 00 $my_ap 84 1b 5e 7e e1 bc $my_ap d0 72 04 00 48 01 00 49 03 00 06 05
nl80211: Frame TX status event
nl80211: Action TX status: cookie=0ffff880039f70700 (match) (ack=1)
wlan1: Event TX_STATUS (18) received
wlan1: EVENT_TX_STATUS dst=$my_ap type=0 stype=13
RTM_NEWLINK: operstate=1 ifi_flags=0x1003 ([UP])
RTM_NEWLINK, IFLA_IFNAME: Interface 'wlan1' added
nl80211: if_removed already cleared - ignore event
nl80211: Event message available
nl80211: Delete station $my_ap
nl80211: Event message available
nl80211: MLME event 39
nl80211: MLME event frame - hexdump(len=26): c0 00 00 00 $my_ap 84 1b 5e 7e e1 bc $my_ap 00 00 04 00
nl80211: Deauthenticate event
wlan1: Event DEAUTH (12) received
wlan1: Deauthentication notification
wlan1:  * reason 4 (locally generated)
wlan1:  * address $my_ap
Deauthentication frame IE(s) - hexdump(len=0): [NULL]
wlan1: CTRL-EVENT-DISCONNECTED bssid=$my_ap reason=4 locally_generated=1
wlan1: Auto connect enabled: try to reconnect (wps=0 wpa_state=9)
wlan1: Setting scan request: 0 sec 100000 usec
Added BSSID $my_ap into blacklist
wlan1: Blacklist count 1 --> request scan in 100 ms
wlan1: Setting scan request: 0 sec 100000 usec
wlan1: Disconnect event - remove keys
wpa_driver_nl80211_set_key: ifindex=6 alg=0 addr=(nil) key_idx=0 set_tx=0 seq_len=0 key_len=0
wpa_driver_nl80211_set_key: ifindex=6 alg=0 addr=(nil) key_idx=1 set_tx=0 seq_len=0 key_len=0
wpa_driver_nl80211_set_key: ifindex=6 alg=0 addr=(nil) key_idx=2 set_tx=0 seq_len=0 key_len=0
wpa_driver_nl80211_set_key: ifindex=6 alg=0 addr=(nil) key_idx=3 set_tx=0 seq_len=0 key_len=0
wpa_driver_nl80211_set_key: ifindex=6 alg=0 addr=0x112f828 key_idx=0 set_tx=0 seq_len=0 key_len=0
   addr=$my_ap
wlan1: State: COMPLETED -> DISCONNECTED
-->8--


What's the problem? wpa_supplicant or rt2800usb or mac80211 or all of
the above?


Thanks

//Peter

Re: RT3573 wpa_supplicant: "nl80211: Delete station $my_ap" after "SME: Send 20/40 BSS Coexistence to $my_ap"

[Xose Vazquez Perez]

Peter Stuge wrote:

> I'm testing the WNDA4100 with torvalds/linux.git and as of a day or
> two ago and wpa_supplicant-2.0.
> 
> I haven't measured throughput. Latency is really bad for interactive
> connections, but at least the packets are flowing on all bands. Yay!
> 
> Ping to the access point is around 3 ms with the odd 20+ ms when the
> kernel goes off to do something else.

Support for RT3573 was added recently, you should use wireless-next:
http://git.kernel.org/cgit/linux/kernel/git/linville/wireless-next.git

You can also try latest fw(v0.33):
http://rt2x00.serialmonkey.com/pipermail/users_rt2x00.serialmonkey.com/attachments/20130111/f8ebe4d3/attachment.bin

md5sum:
======
ac4f6d8b679945208a978e397c016aa7 rt2870.bin (8192 bytes)

# mv /lib/firmware/rt2870.bin /lib/firmware/rt2870.bin.orig
# mv attachment.bin /lib/firmware/rt2870.bin

Re: RT3573 wpa_supplicant: "nl80211: Delete station $my_ap" after "SME: Send 20/40 BSS Coexistence to $my_ap"

[Peter Stuge]

Xose Vazquez Perez wrote:
> > I'm testing the WNDA4100 with torvalds/linux.git and as of a day or
> > two ago and wpa_supplicant-2.0.
> > 
> > I haven't measured throughput. Latency is really bad for interactive
> > connections, but at least the packets are flowing on all bands. Yay!
> > 
> > Ping to the access point is around 3 ms with the odd 20+ ms when the
> > kernel goes off to do something else.
> 
> Support for RT3573 was added recently, you should use wireless-next:
> http://git.kernel.org/cgit/linux/kernel/git/linville/wireless-next.git

I made a typo in my first email - it should have read:

"I'm testing the WNDA4100 with torvalds/linux.git as of a day or two ago"

I am on linux.git commit d8efd82eece89f8a5790b0febf17522affe9e1f1
which contains commit f212781, current wireless-next/master HEAD.

I also looked on the rt2x00 list for proposed 3573 patches but didn't
see any. I hope I already have the latest code.

Do you suspect that my problem (as described briefly in the email
subject and with more detail after the part of my email which you
quoted) is related to the device driver rather than higher-up parts
of the stack?


> You can also try latest fw(v0.33):
> http://rt2x00.serialmonkey.com/pipermail/users_rt2x00.serialmonkey.com/attachments/20130111/f8ebe4d3/attachment.bin

Thanks! Will try this and get back to you.

But why isn't this file in linux-firmware.git? Especially if it was
sent already in January?


Kind regards

//Peter

Re: [rt2x00-users] RT3573 wpa_supplicant: "nl80211: Delete station $my_ap" after "SME: Send 20/40 BSS Coexistence to $my_ap"

[Peter Stuge]

Peter Stuge wrote:
> I am on linux.git commit d8efd82eece89f8a5790b0febf17522affe9e1f1
> which contains commit f212781, current wireless-next/master HEAD.
..
> Do you suspect that my problem (as described briefly in the email
> subject and with more detail after the part of my email which you
> quoted) is related to the device driver rather than higher-up parts
> of the stack?
..
> > You can also try latest fw(v0.33):
> > http://rt2x00.serialmonkey.com/pipermail/users_rt2x00.serialmonkey.com/attachments/20130111/f8ebe4d3/attachment.bin
> 
> Thanks! Will try this and get back to you.

With the above firmware the problem described briefly in Subject and
in more detail in my original email occurs much more easily, even as
I am actively generating traffic over the interface. It happened
twice while I was composing this very short email.


//Peter