[ath9k-devel] ath9k_htc station never disconnects from AP

[ath9k-devel] ath9k_htc station never disconnects from AP

[Ebright, Dave]

Hello,

I am using the backports-3.10-2 to build/install the ath9k_htc drivers to use with the Ubiquiti WiFiStation Ext (ar9271).  I am using this device as a station to connect to an AP which is a D-Link DIR-825 Xtreme N Dual Band (2x2) Gigabit Router.  For testing purposes I am initially using no encryption (open).  For the client I am using wpa_supplicant-2.0.  I have no issue with connecting this station/client to the AP.  That is "iw" shows that the link is up/connected, I can ping the AP, send files via ssh/scp over the link, etc..

The issue I have is that when I power off the AP (unplug the AP from wall power) with the link up, the station never times out the connection to the now powered off AP.  That is I do "iw dev wlan0 link" and "iw dev wlan0 station dump" and the results returned look just like they did when the AP was powered on and connected (appears to think the link is still up).  For the station dumps of course the inactive time increases.  The tx packets/tx bytes are also increasing while the rx packets/rx bytes stay constant.  Also if I perform "iw dev wlan0 scan"  it still reports the powered off AP in the scan results???  Yes I verified there is not another AP using the same SSID (always checking the MAC).  Of course ping fails.

When I run wpa_supplicant in debug mode after powering off the AP, it rolls (every second)  in reporting NL80211_CQM_RSSI_BEACON_LOSS_EVENT.  Same if I run "iw event" which rolls in "wlan0 (phy #0): CQM event: Beacon loss detected.  The wpa_supplicant process never reports the link going down and I have waited sometimes an hour or more.  Power management is off.   Using firmware htc_9271.fw.

I did run a wifi monitor to collect the packets when this condition is occurring and the station is constantly transmitting null function packets (~9 or 10 per session, that is a null function packet (flags 0x1) and then several retries which is null function packet (flags 0x9 same session number).  The destination for these packets is the powered off AP.  I thought these null function packets from a station are generally for power savings which is turned off?

Same issue occurs with Alfa AWUS036NHA (ar9271).

So my question is what should trigger the station to give up on the AP and disconnect?  Why doesn't the beacon loss event trigger this to occur after some timeout period of time has gone by?  Is this issue with the drivers or firmware?

Thanks,

Dave

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20130718/dea90984/attachment.htm 

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

Hallo list:

i'm using openwrt with mac80211 and ath9k for mesh networks, but the 
iwconfig not permited using the sensitivity control,

i find this parameter on driver, but not have sucess

i'm need help

att:


Flavio Leonel

[ath9k-devel] sensitivity control for ath9k with mac80211

[Holger Schurig]

Hi Flavio,

"iwconfig" is prehistoric, try your luck with "iw".

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

Ok i know that but the command iw not permited set sensitivity limit

how i can seting this limit on atth9k , this question..

[ath9k-devel] sensitivity control for ath9k with mac80211

[Holger Schurig]

Hmm, now I remember. From the back of my mind, old iwconfig set
sensitivity was used with some Orininco cards (Hermes Chips) that
could associate and roam themselves. So called "full-mac" cards, which
are now a bit out of fashion (except on phones/embedded devices).

mac80211 based cards cannot do associate/roam by themselves. They just
send and receive 802.11 packets. Association and roaming is being done
at different levels, e.g. mac80211. So, if you want to roam with a
mac80211 based card, you'll need to use wpa_supplicant (even when
using WEP), e.g. with some of the bgscan modules.

Now you're using mesh, and here I don't have enought knowledge.
Probably the mesh implementation will have it's own tool to set mesh
properties, or?

[ath9k-devel] sensitivity control for ath9k with mac80211

[Ben Greear]

On 07/19/2013 06:36 AM, Flavio Leonel wrote:
> Ok i know that but the command iw not permited set sensitivity limit
>
> how i can seting this limit on atth9k , this question..

I tried messing with this some months ago and got nowhere.  I could
not figure any way to make the NIC ignore fainter signals, and I am
not sure it is possible to make the hardware do this...

If you get something working, I'd like to learn about how you did it...

Thanks,
Ben

> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>


-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com

[ath9k-devel] sensitivity control for ath9k with mac80211

[Sergey Ryazanov]

2013/7/19 Ben Greear <greearb@candelatech.com>:
> On 07/19/2013 06:36 AM, Flavio Leonel wrote:
>> Ok i know that but the command iw not permited set sensitivity limit
>>
>> how i can seting this limit on atth9k , this question..
>
> I tried messing with this some months ago and got nowhere.  I could
> not figure any way to make the NIC ignore fainter signals, and I am
> not sure it is possible to make the hardware do this...
>

At least 5k chips have a special register for CCA threshold
configuration. Modern 9k chips have not inherited this register?

-- 
BR,
Sergey

[ath9k-devel] sensitivity control for ath9k with mac80211

[Felix Fietkau]

On 2013-07-19 4:58 PM, Sergey Ryazanov wrote:
> 2013/7/19 Ben Greear <greearb@candelatech.com>:
>> On 07/19/2013 06:36 AM, Flavio Leonel wrote:
>>> Ok i know that but the command iw not permited set sensitivity limit
>>>
>>> how i can seting this limit on atth9k , this question..
>>
>> I tried messing with this some months ago and got nowhere.  I could
>> not figure any way to make the NIC ignore fainter signals, and I am
>> not sure it is possible to make the hardware do this...
>>
> 
> At least 5k chips have a special register for CCA threshold
> configuration. Modern 9k chips have not inherited this register?
The CCA threshold is not the same as the threshold for signal detection.
In many ways, 5k and 9k cards have a somewhat similar set of registers
to configure various aspects of detection sensitivity. Some of those are
set in the initvals, some are controlled by ANI.
The main issue is that the driver does not expose any convenient knobs
to control this.

- Felix

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

Well I already suspected the problems that the ath9k would not nehuma 
tool that fizesseo control of a range limit signal strength.

Please'm fairly layman in chips and eeprom
clarify CCA & ANI.?

I need to create this algorithm but do not know where to start
person from the list to help here desenviolvermos this tool
running on the system RouterBoard mikrotik well.

help me

ATT:
Flavio Leonel

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

Well I already suspected the problems that the ath9k would not nehuma 
tool that fizesseo control of a range limit signal strength.

Please'm fairly layman in chips and eeprom
clarify CCA & ANI.?

I need to create this algorithm but do not know where to start
help me, we will develop this tool running in the system RouterBoard 
mikrotik well.

help me

ATT:
Flavio Leonel

[ath9k-devel] sensitivity control for ath9k with mac80211

[Sergey Ryazanov]

2013/7/19 Flavio Leonel <frleonel@gmail.com>:
> Please'm fairly layman in chips and eeprom
> clarify CCA & ANI.?
CCA is 802.11 standard abbreviation of "clear channel assessment".
Simplistically, CCA threshold is the signal level above which receiver
assumes that the channel is busy. Mainly affects the Tx procedure.
ANI stands for Automatic Noise Immunity. ANI is a software driven
method for controlling radio sensitivity in the face of noise sources
[1].

>
> I need to create this algorithm but do not know where to start
> help me, we will develop this tool running in the system RouterBoard
> mikrotik well.
Sensitivity is too general term. What exactly do you want to control?

1. https://wiki.freebsd.org/dev/ath_hal(4)/AutomaticNoiseImmunity

-- 
BR,
Sergey

[ath9k-devel] sensitivity control for ath9k with mac80211

[Sergey Ryazanov]

Hello Flavio,

2013/7/20, Flavio Leonel <frleonel@gmail.com>:
> 2013/7/19 Sergey Ryazanov <ryazanov.s.a@gmail.com>
>> Sensitivity is too general term. What exactly do you want to control?
>>
> I need to control the level snr customer having a minimum of -50 and a
> maximum of -70
>
Linux mesh realization has a non standard option, which blocks the
peer initiation, if remote station has the signal level below the
specified value. Is this suitable for you?

P.S. Use bottom or inline posting, please.

-- 
BR,
Sergey

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

yes it would! and also need to control the level of the stations not 802.11s
quese connect to a simple station with 802.11n AP

PS: My APS works with a firmware based on openwrt being compiled that I 
use 2 interfaces 1 in doing 5ghz mesh and a 2.4 ghz doing in WIFI

[ath9k-devel] sensitivity control for ath9k with mac80211

[Sergey Ryazanov]

2013/7/21 Flavio Leonel <frleonel@gmail.com>:
> yes it would!
So you need to tune "mesh_rssi_threshold" parameter using iw utility.
Try to run "iw <iface> mesh_param" as starting point. I do not know
either the OpenWRT supports mesh configuration from the box or no.

> and also need to control the level of the stations not 802.11s
> quese connect to a simple station with 802.11n AP
>
You need to dig into hostapd configuration file. If I am not mistaken,
hostapd support something like that.

-- 
BR,
Sergey

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

Thanks for the help I will try to find this function in "iw" the openwrt 
but it is very limited I can not understand why, until I am studying to 
Uzar busybox - you have some stuff to help me compile it for cards based 
on Atheros mips?

the hotsapd not find anything about it but I'll keep searching

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

hallo my friend i am find fo iw tools & i see this command

root at BASE1:~# iw dev wlan0 cqm rssi on -70
command failed: Invalid argument (-22)
root at BASE1:~# iw dev wlan0 cqm rssi on 30
command failed: Invalid argument (-22)


but this parameters not running in openwrt

help me

[ath9k-devel] sensitivity control for ath9k with mac80211

[Sergey Ryazanov]

Hello Flavio,

2013/7/22 Flavio Leonel <frleonel@gmail.com>:
> hallo my friend i am find fo iw tools & i see this command
>
> root at BASE1:~# iw dev wlan0 cqm rssi on -70
> command failed: Invalid argument (-22)
> root at BASE1:~# iw dev wlan0 cqm rssi on 30
> command failed: Invalid argument (-22)
>
Again. You need to tune "mesh_param mesh_rssi_threshold" but not "cqm rssi".

-- 
BR,
Sergey

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

Em 28-07-2013 10:21, Sergey Ryazanov escreveu:
> Hello Flavio,
>
> 2013/7/22 Flavio Leonel <frleonel@gmail.com>:
>> hallo my friend i am find fo iw tools & i see this command
>>
>> root at BASE1:~# iw dev wlan0 cqm rssi on -70
>> command failed: Invalid argument (-22)
>> root at BASE1:~# iw dev wlan0 cqm rssi on 30
>> command failed: Invalid argument (-22)
>>
> Again. You need to tune "mesh_param mesh_rssi_threshold" but not "cqm rssi".
>
i'm sorry which the command as I can use to tune the mesh RSSI and treshold?

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

patch=

+ * @NL80211_ATTR_CQM_RSSI_HYST: RSSI hysteresis in dBm. This value specifies
+ *	the minimum amount the RSSI level must change after an event before a
+ *	new event may be issued (to reduce effects of RSSI oscillation).

but in openwrt no running this command


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20130721/e7e5ad40/attachment.htm 

[ath9k-devel] IWCONFIG - ubuntu

[Flavio Leonel]

1 - I already said here in the mailing list that the iwconfig archaic 
and that no one uses anymore, because such a response does not 
correspond with the current linux like ubuntu .. and the same still 
ultiliza iwconfig and has the possibility of setting the sensitivity on 
the card and both also utilize AR9xx equal to openwrt with radios common 
market ...

2 - can someone tell me what because openwrt uses and what is 
snapshot-driver.

3 - I wonder how to implement the driver ath9k pure in openwrt without 
the patches they offer ..

ATT:

Flavio Leonel

[ath9k-devel] sensitivity control for ath9k with mac80211

[Sergey Ryazanov]

2013/7/19 Felix Fietkau <nbd@openwrt.org>:
> On 2013-07-19 4:58 PM, Sergey Ryazanov wrote:
>> 2013/7/19 Ben Greear <greearb@candelatech.com>:
>>> On 07/19/2013 06:36 AM, Flavio Leonel wrote:
>>>> Ok i know that but the command iw not permited set sensitivity limit
>>>>
>>>> how i can seting this limit on atth9k , this question..
>>>
>>> I tried messing with this some months ago and got nowhere.  I could
>>> not figure any way to make the NIC ignore fainter signals, and I am
>>> not sure it is possible to make the hardware do this...
>>>
>>
>> At least 5k chips have a special register for CCA threshold
>> configuration. Modern 9k chips have not inherited this register?
> The CCA threshold is not the same as the threshold for signal detection.
> In many ways, 5k and 9k cards have a somewhat similar set of registers
> to configure various aspects of detection sensitivity. Some of those are
> set in the initvals, some are controlled by ANI.
> The main issue is that the driver does not expose any convenient knobs
> to control this.
>
Yeah. It would be useful for developers to get delicate and operative
control over Tx/Rx process (at least via debugfs). But I never faced
with situations where this would be useful for end user. So, IMHO,
even if somebody submit appropriate patches, they will not be
accepted.

-- 
BR,
Sergey

[ath9k-devel] sensitivity control for ath9k with mac80211

[Ben Greear]

On 07/19/2013 05:53 PM, Sergey Ryazanov wrote:
> 2013/7/19 Felix Fietkau <nbd@openwrt.org>:
>> On 2013-07-19 4:58 PM, Sergey Ryazanov wrote:
>>> 2013/7/19 Ben Greear <greearb@candelatech.com>:
>>>> On 07/19/2013 06:36 AM, Flavio Leonel wrote:
>>>>> Ok i know that but the command iw not permited set sensitivity limit
>>>>>
>>>>> how i can seting this limit on atth9k , this question..
>>>>
>>>> I tried messing with this some months ago and got nowhere.  I could
>>>> not figure any way to make the NIC ignore fainter signals, and I am
>>>> not sure it is possible to make the hardware do this...
>>>>
>>>
>>> At least 5k chips have a special register for CCA threshold
>>> configuration. Modern 9k chips have not inherited this register?
>> The CCA threshold is not the same as the threshold for signal detection.
>> In many ways, 5k and 9k cards have a somewhat similar set of registers
>> to configure various aspects of detection sensitivity. Some of those are
>> set in the initvals, some are controlled by ANI.
>> The main issue is that the driver does not expose any convenient knobs
>> to control this.
>>
> Yeah. It would be useful for developers to get delicate and operative
> control over Tx/Rx process (at least via debugfs). But I never faced
> with situations where this would be useful for end user. So, IMHO,
> even if somebody submit appropriate patches, they will not be
> accepted.

I do a lot of testing in cases where there are a few APs with signal level of around -70 to -85,
primarily due to not-that-great shielding, as well as APs in surrounding buildings.

I'd like to be able to tune a NIC so that it just plain ignored those weak signals
to increase throughput with systems with much better signal quality.

Thanks,
Ben


-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com

[ath9k-devel] sensitivity control for ath9k with mac80211

[Flavio Leonel]

We Can testing the IW command with txpower limit, but not the ar92xx 
response for this parameter ...

i 'm looking finding one solution for this problem too in my network ..!
can someone help us?

thank for this

ATT: Flavio Leonel
> On 07/19/2013 05:53 PM, Sergey Ryazanov wrote:
>> 2013/7/19 Felix Fietkau <nbd@openwrt.org>:
>>> On 2013-07-19 4:58 PM, Sergey Ryazanov wrote:
>>>> 2013/7/19 Ben Greear <greearb@candelatech.com>:
>>>>> On 07/19/2013 06:36 AM, Flavio Leonel wrote:
>>>>>> Ok i know that but the command iw not permited set sensitivity limit
>>>>>>
>>>>>> how i can seting this limit on atth9k , this question..
>>>>> I tried messing with this some months ago and got nowhere.  I could
>>>>> not figure any way to make the NIC ignore fainter signals, and I am
>>>>> not sure it is possible to make the hardware do this...
>>>>>
>>>> At least 5k chips have a special register for CCA threshold
>>>> configuration. Modern 9k chips have not inherited this register?
>>> The CCA threshold is not the same as the threshold for signal detection.
>>> In many ways, 5k and 9k cards have a somewhat similar set of registers
>>> to configure various aspects of detection sensitivity. Some of those are
>>> set in the initvals, some are controlled by ANI.
>>> The main issue is that the driver does not expose any convenient knobs
>>> to control this.
>>>
>> Yeah. It would be useful for developers to get delicate and operative
>> control over Tx/Rx process (at least via debugfs). But I never faced
>> with situations where this would be useful for end user. So, IMHO,
>> even if somebody submit appropriate patches, they will not be
>> accepted.
> I do a lot of testing in cases where there are a few APs with signal level of around -70 to -85,
> primarily due to not-that-great shielding, as well as APs in surrounding buildings.
>
> I'd like to be able to tune a NIC so that it just plain ignored those weak signals
> to increase throughput with systems with much better signal quality.
>
> Thanks,
> Ben
>
>

[ath9k-devel] sensitivity control for ath9k with mac80211

[Sergey Ryazanov]

2013/7/22 Ben Greear <greearb@candelatech.com>:
> On 07/19/2013 05:53 PM, Sergey Ryazanov wrote:
>>
>> 2013/7/19 Felix Fietkau <nbd@openwrt.org>:
>>>
>>> On 2013-07-19 4:58 PM, Sergey Ryazanov wrote:
>>>>
>>>> 2013/7/19 Ben Greear <greearb@candelatech.com>:
>>>>>
>>>>> On 07/19/2013 06:36 AM, Flavio Leonel wrote:
>>>>>>
>>>>>> Ok i know that but the command iw not permited set sensitivity limit
>>>>>>
>>>>>> how i can seting this limit on atth9k , this question..
>>>>>
>>>>>
>>>>> I tried messing with this some months ago and got nowhere.  I could
>>>>> not figure any way to make the NIC ignore fainter signals, and I am
>>>>> not sure it is possible to make the hardware do this...
>>>>>
>>>>
>>>> At least 5k chips have a special register for CCA threshold
>>>> configuration. Modern 9k chips have not inherited this register?
>>>
>>> The CCA threshold is not the same as the threshold for signal detection.
>>> In many ways, 5k and 9k cards have a somewhat similar set of registers
>>> to configure various aspects of detection sensitivity. Some of those are
>>> set in the initvals, some are controlled by ANI.
>>> The main issue is that the driver does not expose any convenient knobs
>>> to control this.
>>>
>> Yeah. It would be useful for developers to get delicate and operative
>> control over Tx/Rx process (at least via debugfs). But I never faced
>> with situations where this would be useful for end user. So, IMHO,
>> even if somebody submit appropriate patches, they will not be
>> accepted.
>
>
> I do a lot of testing in cases where there are a few APs with signal level
> of around -70 to -85,
> primarily due to not-that-great shielding, as well as APs in surrounding
> buildings.
>
Do you tried to tune CCA or ANI related stuff, as Felix suggested?

> I'd like to be able to tune a NIC so that it just plain ignored those weak
> signals
> to increase throughput with systems with much better signal quality.
>
Even if you force NIC to completely ignore weak signals during clear
channel detection procedure you can't avoid interference, which could
lead to rise of packet error rate, what lead to drop of the
throughput. On other hand, if your NIC transmit in the same time as
weak APs, you are jamming them.
Shielding is the single way for correct testing, until you do not try
to create a jammer.

-- 
BR,
Sergey