Modify

Opened 3 years ago

Closed 3 years ago

Last modified 19 months ago

#18794 closed defect (fixed)

ath10k: Spurious quick kickout for STA

Reported by: Anton van Bohemen <avbohemen@…> Owned by: developers
Priority: normal Milestone: Chaos Calmer 15.05
Component: base system Version: Trunk
Keywords: Cc:

Description

Running a TP-Link Archer C7 v2 with CC r43790, my kernel log is filling up with these messages, every second:

[168037.910000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA f4:f1:5a:cd:73:7c
[168038.910000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA f4:f1:5a:cd:73:7c
[168039.910000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA f4:f1:5a:cd:73:7c

System memory is also filling up slowly when these messages occur. CPU usage is negligible (according to collectd). The only way to clear this situation is to do "wifi down; wifi", after which the 'stale peer' MAC is removed:

[168040.800000] device wlan1 left promiscuous mode
[168040.800000] br-lan: port 3(wlan1) entered disabled state
[168040.840000] device wlan0 left promiscuous mode
[168040.840000] br-lan: port 2(wlan0) entered disabled state
[168040.850000] ath10k_pci 0000:01:00.0: removing stale peer f4:f1:5a:cd:73:7c from vdev_id 0
[168040.860000] ath10k_pci 0000:01:00.0: received addba event for invalid peer_id: 30
[168040.870000] ath10k_pci 0000:01:00.0: peer-unmap-event: unknown peer id 30
[168042.850000] ath10k_pci 0000:01:00.0: otp stream is empty, using board.bin contents
[168043.870000] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready
[168043.890000] device wlan0 entered promiscuous mode
[168043.950000] IPv6: ADDRCONF(NETDEV_UP): wlan1: link is not ready
[168043.970000] device wlan1 entered promiscuous mode
[168043.970000] br-lan: port 3(wlan1) entered forwarding state
[168043.980000] br-lan: port 3(wlan1) entered forwarding state
[168045.520000] br-lan: port 3(wlan1) entered disabled state
[168045.520000] br-lan: port 2(wlan0) entered forwarding state
[168045.530000] br-lan: port 2(wlan0) entered forwarding state
[168045.530000] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[168045.560000] br-lan: port 3(wlan1) entered forwarding state
[168045.560000] br-lan: port 3(wlan1) entered forwarding state
[168045.570000] IPv6: ADDRCONF(NETDEV_CHANGE): wlan1: link becomes ready
[168047.530000] br-lan: port 2(wlan0) entered forwarding state
[168047.560000] br-lan: port 3(wlan1) entered forwarding state

BTW, the device in question is an iPhone 5, running IOS 8.1.2. It probably connected and left the AP many times. My guess is the 'stale peer' should be cleaned more quickly (i.e. directly after it disconnects from the AP)?

Attachments (1)

ath10k.warn.log.gz (134.5 KB) - added by bogey 3 years ago.
ath10k warnings

Download all attachments as: .zip

Change History (19)

comment:1 Changed 3 years ago by anonymous

This is the same issue as #18699, it looks like that an upstream commit of the ath10k driver will fix it.

comment:2 follow-ups: Changed 3 years ago by anonymous

I have also the similar problem with Lumia 820 WP8 client. Other clients connect fine.
Here is some log entries

Sun Feb  1 20:43:30 2015 kern.warn kernel: [198934.390000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA 3c:c2:43:6f:cf:f8
Sun Feb  1 20:43:31 2015 daemon.info hostapd: wlan1: STA 3c:c2:43:6f:cf:f8 IEEE 802.11: disassociated
Sun Feb  1 20:43:31 2015 kern.warn kernel: [198935.390000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA 3c:c2:43:6f:cf:f8
Sun Feb  1 20:43:32 2015 daemon.info hostapd: wlan1: STA 3c:c2:43:6f:cf:f8 IEEE 802.11: deauthenticated due to inactivity (timer DEAUTH/REMOVE)
Sun Feb  1 20:48:06 2015 kern.warn kernel: [199210.010000] ath10k_pci 0000:01:00.0: failed to delete peer 3c:c2:43:6f:cf:f8 for vdev 0: -145
Sun Feb  1 20:48:06 2015 kern.warn kernel: [199210.170000] ath10k_pci 0000:01:00.0: SWBA overrun on vdev 0
Sun Feb  1 20:48:06 2015 kern.warn kernel: [199210.270000] ath10k_pci 0000:01:00.0: SWBA overrun on vdev 0
Sun Feb  1 20:48:06 2015 kern.warn kernel: [199210.380000] ath10k_pci 0000:01:00.0: SWBA overrun on vdev 0
Sun Feb  1 20:48:06 2015 kern.warn kernel: [199210.480000] ath10k_pci 0000:01:00.0: SWBA overrun on vdev 0

comment:3 in reply to: ↑ 2 Changed 3 years ago by anonymous

Replying to anonymous:
And router is TP-LINK Archer C5 running with r44186

comment:4 in reply to: ↑ 2 Changed 3 years ago by anonymous

Replying to anonymous:
And router is TP-LINK Archer C5 running with r44186

comment:5 in reply to: ↑ 2 ; follow-up: Changed 3 years ago by anonymous

Replying to anonymous:
Ok now also other clients fails to connect:

Sun Feb  1 21:21:22 2015 kern.warn kernel: [201206.210000] ath10k_pci 0000:01:00.0: failed to delete peer 3c:c2:43:6f:cf:f8 for vdev 0: -145
Sun Feb  1 21:21:25 2015 kern.warn kernel: [201208.900000] ath10k_pci 0000:01:00.0: number of peers exceeded: peers number 127 (max peers 127)
Sun Feb  1 21:21:43 2015 kern.warn kernel: [201227.110000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA 3c:c2:43:6f:cf:f8
Sun Feb  1 21:22:02 2015 kern.warn kernel: [201245.860000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA 3c:c2:43:6f:cf:f8
Sun Feb  1 21:38:34 2015 kern.warn kernel: [202237.770000] ath10k_pci 0000:01:00.0: number of peers exceeded: peers number 127 (max peers 127)

comment:6 in reply to: ↑ 5 Changed 3 years ago by anonymous

Replying to anonymous:
after network restart

Sun Feb  1 22:21:54 2015 kern.warn kernel: [204838.310000] ath10k_pci 0000:01:00.0: failed to flush transmit queue (skip 0 ar-state 1): 0
Sun Feb  1 22:21:54 2015 kern.warn kernel: [204838.310000] ath10k_pci 0000:01:00.0: removing stale peer 3c:c2:43:6f:cf:f8 from vdev_id 0
Sun Feb  1 22:21:54 2015 kern.warn kernel: [204838.320000] ath10k_pci 0000:01:00.0: received addba event for invalid peer_id: 366
Sun Feb  1 22:21:54 2015 kern.warn kernel: [204838.330000] ath10k_pci 0000:01:00.0: peer-unmap-event: unknown peer id 366
Sun Feb  1 22:21:57 2015 kern.warn kernel: [204841.530000] ath10k_pci 0000:01:00.0: otp stream is empty, using board.bin contents

comment:7 Changed 3 years ago by bogey

Above anonymous comments 2-5 are mine.
Happened today again, I went out of range with my mobile and log started filling up with those spurious kickouts.
After network restart ath10 firmware crashed:

Mon Feb  2 18:32:36 2015 kern.warn kernel: [277480.350000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:36 2015 kern.warn kernel: [277480.450000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:36 2015 kern.warn kernel: [277480.550000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:36 2015 kern.warn kernel: [277480.650000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:37 2015 kern.warn kernel: [277480.750000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:37 2015 kern.warn kernel: [277480.860000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:37 2015 kern.warn kernel: [277480.960000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:37 2015 kern.warn kernel: [277481.060000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:37 2015 kern.warn kernel: [277481.160000] ath10k_pci 0000:01:00.0: could not get mac80211 beacon
Mon Feb  2 18:32:44 2015 kern.warn kernel: [277488.260000] ath10k_warn: 21 callbacks suppressed
Mon Feb  2 18:32:44 2015 kern.warn kernel: [277488.260000] ath10k_pci 0000:01:00.0: failed to flush transmit queue (skip 0 ar-state 1): 0
Mon Feb  2 18:32:44 2015 kern.warn kernel: [277488.300000] ath10k_pci 0000:01:00.0: removing stale peer 3c:c2:43:6f:cf:f8 from vdev_id 0
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.310000] ath10k_pci 0000:01:00.0: firmware crashed! (uuid 76fc3531-4a81-4090-a83a-1c7f2cb55cad)
Mon Feb  2 18:32:44 2015 kern.info kernel: [277488.320000] ath10k_pci 0000:01:00.0: qca988x hw2.0 (0x4100016c, 0x043202ff) fw 10.2-00082-4-2 api 3 htt 2.1 wmi 65.109.0.0 cal otp
Mon Feb  2 18:32:44 2015 kern.info kernel: [277488.330000] ath10k_pci 0000:01:00.0: debug 0 debugfs 1 tracing 0 dfs 1 testmode 1
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.340000] ath10k_pci 0000:01:00.0: firmware register dump:
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.350000] ath10k_pci 0000:01:00.0: [00]: 0x4100016C 0x00000000 0x009A9523 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.360000] ath10k_pci 0000:01:00.0: [04]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.360000] ath10k_pci 0000:01:00.0: [08]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.370000] ath10k_pci 0000:01:00.0: [12]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.380000] ath10k_pci 0000:01:00.0: [16]: 0x00000000 0x00000000 0x00000000 0x009A9523
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.390000] ath10k_pci 0000:01:00.0: [20]: 0x00000000 0x00401930 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.400000] ath10k_pci 0000:01:00.0: [24]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.400000] ath10k_pci 0000:01:00.0: [28]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.410000] ath10k_pci 0000:01:00.0: [32]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.420000] ath10k_pci 0000:01:00.0: [36]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.430000] ath10k_pci 0000:01:00.0: [40]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.440000] ath10k_pci 0000:01:00.0: [44]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.450000] ath10k_pci 0000:01:00.0: [48]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.450000] ath10k_pci 0000:01:00.0: [52]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:44 2015 kern.err kernel: [277488.460000] ath10k_pci 0000:01:00.0: [56]: 0x00000000 0x00000000 0x00000000 0x00000000
Mon Feb  2 18:32:51 2015 kern.warn kernel: [277495.180000] ath10k_pci 0000:01:00.0: otp stream is empty, using board.bin contents
Mon Feb  2 18:32:52 2015 kern.warn kernel: [277496.420000] [<8726fe80>] ath10k_peer_unmap_event+0x1b24/0x2628 [ath10k_core]
Mon Feb  2 18:32:52 2015 kern.warn kernel: [277496.570000] ath10k_pci 0000:01:00.0: failed to allocate memory chunk
Mon Feb  2 18:32:57 2015 kern.warn kernel: [277501.340000] ath10k_pci 0000:01:00.0: wmi service ready event not received
Mon Feb  2 18:32:57 2015 kern.err kernel: [277501.640000] ath10k_pci 0000:01:00.0: Could not init core: -145
Mon Feb  2 18:33:09 2015 kern.warn kernel: [277513.160000] ath10k_pci 0000:01:00.0: otp stream is empty, using board.bin contents
Mon Feb  2 18:33:10 2015 kern.warn kernel: [277514.250000] [<8726fe80>] ath10k_peer_unmap_event+0x1b24/0x2628 [ath10k_core]
Mon Feb  2 18:33:10 2015 kern.warn kernel: [277514.400000] ath10k_pci 0000:01:00.0: failed to allocate memory chunk
Mon Feb  2 18:33:15 2015 kern.warn kernel: [277519.170000] ath10k_pci 0000:01:00.0: wmi service ready event not received
Mon Feb  2 18:33:15 2015 kern.err kernel: [277519.470000] ath10k_pci 0000:01:00.0: Could not init core: -145
Last edited 3 years ago by bogey (previous) (diff)

comment:8 Changed 3 years ago by anonymous

Same here, Archer C5 r43757, clients iPhone5 & HTC One m8, both on 5Ghz, rebooting solves the issue for some time

Changed 3 years ago by bogey

ath10k warnings

comment:9 Changed 3 years ago by mymegabyte

Same issue here: TP-Link Archer C7 v2, r44484. The iPhone5 on my network keeps getting booted and is unable to reconnect until I reboot the AP. Other clients appear to be working fine. Will report back if I find anything informative/additional to add.

[37139.670000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA 0c:3e:9f:...
[37140.670000] ath10k_pci 0000:01:00.0: Spurious quick kickout for STA 0c:3e:9f:...
  • Edited to correct router model.
Last edited 3 years ago by mymegabyte (previous) (diff)

comment:10 Changed 3 years ago by anonymous

I have the same problem with OSx, usually after I open Macbook's lid (i.e. it wakes up from sleep). I use r44541.

comment:11 Changed 3 years ago by anatolik

And I have the same issue with OpenWRT (14.07, r42625) on TP-LINK Archer C7 V2.

Using macbookpro with 5GHz band is a pain, it takes several retries to connect to 5ghz network. The router system log is filled with "ath10k: Spurious quick kickout for STA".

comment:12 Changed 3 years ago by bogey

New firmware committed
r44656
Anyone tested yet if it helps for this and does the recent builds work OK?

comment:13 follow-up: Changed 3 years ago by anonymous

I noticed that the bug is triggered when OSx is connected to 2.4ghz ath9k. The reason is because OSx and iOS they don't support HT40 on 2.4ghz and they force the whole network to switch to HT20. This transition is buggy on ath9k and makes ath9k interface unusable.
That's where "Spurious quick kickout" bug is triggered on ath10k.

I use r44898 and I don't get "Spurious quick kickout" anymore, but now I was getting "Failed to stop TX DMA" on ath9k because of the HT40 issue.

As a workaround I set HT20 in /etc/config/wireless for the ath9k radio and everything is stable now. I hope they will fix HT40 to HT20 transition soon.

For more information check:
http://svn.dd-wrt.com/ticket/2952#comment:110
/ticket/11862.html

comment:14 Changed 3 years ago by Anton van Bohemen <avbohemen@…>

I am running r44953 for a week now, and I don't see any more "spurious quick kickout" messages anymore. The current wireless-testing (2015-03-09, r44655) and updated ath10k firmware (r44656) seem to work ok for me.

comment:15 in reply to: ↑ 13 Changed 3 years ago by bogey

Replying to anonymous:

I noticed that the bug is triggered when OSx is connected to 2.4ghz ath9k. The reason is because OSx and iOS they don't support HT40 on 2.4ghz and they force the whole network to switch to HT20. This transition is buggy on ath9k and makes ath9k interface unusable.
That's where "Spurious quick kickout" bug is triggered on ath10k.

I use r44898 and I don't get "Spurious quick kickout" anymore, but now I was getting "Failed to stop TX DMA" on ath9k because of the HT40 issue.

As a workaround I set HT20 in /etc/config/wireless for the ath9k radio and everything is stable now. I hope they will fix HT40 to HT20 transition soon.

For more information check:
http://svn.dd-wrt.com/ticket/2952#comment:110
/ticket/11862.html

Thanks for finding out. I will test to set ath9k to HT20. I'm still at r44186.

comment:16 Changed 3 years ago by nbd

  • Resolution set to fixed
  • Status changed from new to closed

comment:17 follow-up: Changed 19 months ago by A.F.G.

just seen today.
system :
Model TP-Link Archer C7 v2
Firmware Version OpenWrt Chaos Calmer 15.05.1 / LuCI 15.05-149-g0d8bbd2 Release (git-15.363.78009-956be55)
Kernel Version 3.18.23
I have a small difference : I have some error -132 coming (E_OVERFLOW ?)
like this : [524234.330000] ath10k_pci 0000:01:00.0: failed to add peer 90:b9:31:xx:xx:xx for vdev 0 when adding a new sta: -132
devices failing are one iPhone 5S and one iPad 2, both running iOS 9.03, and a sony Xperia Z5, running android Marshmallow.

comment:18 in reply to: ↑ 17 Changed 19 months ago by Zeugma_fr

Just ask for kernel or system log if needed. ptf reboot solves the issue (but I'm not sure I won't have the issue again).

Replying to A.F.G.:

just seen today.
system :
Model TP-Link Archer C7 v2
Firmware Version OpenWrt Chaos Calmer 15.05.1 / LuCI 15.05-149-g0d8bbd2 Release (git-15.363.78009-956be55)
Kernel Version 3.18.23
I have a small difference : I have some error -132 coming (E_OVERFLOW ?)
like this : [524234.330000] ath10k_pci 0000:01:00.0: failed to add peer 90:b9:31:xx:xx:xx for vdev 0 when adding a new sta: -132
devices failing are one iPhone 5S and one iPad 2, both running iOS 9.03, and a sony Xperia Z5, running android Marshmallow.

Add Comment

Modify Ticket

Action
as closed .
The resolution will be deleted. Next status will be 'reopened'.
Author


E-mail address and user name can be saved in the Preferences.

 
Note: See TracTickets for help on using tickets.