WIFI Connection Drops - Intel Centrino Advanced-N 6235 AGN

Been plagued with this bug for a while but the problem is more frequent now following a fresh install.
After a while of running fine, my wireless device will fail and drop its connection.

When it enters this state, only a reboot will bring it back up.

I tried reloading the iwlwifi module and restarting the network service, it won’t be able to reconnect without a reboot.
Thought it could be a problem with power management so I added pcie_aspm=force as boot parameter.
My installation of tumbleweed is pretty up to date as of now (openSUSE Tumbleweed 20180917).

This problem is present using NetworkManager and Wicked.


    6.954719] iwlwifi 0000:03:00.0: enabling device (0000 -> 0002)
    6.961733] iwlwifi 0000:03:00.0: loaded firmware version 18.168.6.1 op_mode iwldvm
    7.038882] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUG enabled
    7.038884] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUGFS enabled
    7.038886] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEVICE_TRACING disabled
    7.038888] iwlwifi 0000:03:00.0: Detected Intel(R) Centrino(R) Advanced-N 6235 AGN, REV=0xB0
    7.084645] iwlwifi 0000:03:00.0 wlp3s0: renamed from wlan0
    8.597848] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
    8.610845] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
    8.910037] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
    8.993037] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
    9.018911] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
    9.319003] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
    9.400779] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
    9.491496] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
   12.677234] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
   12.978844] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
   13.065374] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
   16.320297] wlp3s0: authenticate with foo:bar:mac
   16.322854] wlp3s0: send auth to foo:bar:mac (try 1/3)
   16.387583] wlp3s0: authenticated
   16.387926] wlp3s0: associate with foo:bar:mac (try 1/3)
   16.390851] wlp3s0: RX AssocResp from foo:bar:mac (capab=0x411 status=0 aid=1)
   16.408889] wlp3s0: associated
   16.419389] IPv6: ADDRCONF(NETDEV_CHANGE): wlp3s0: link becomes ready
   16.486517] wlp3s0: Limiting TX power to 27 (30 - 3) dBm as advertised by foo:bar:mac
[20045.045456] wlp3s0: deauthenticated from foo:bar:mac (Reason: 3=DEAUTH_LEAVING)
[20048.320873] wlp3s0: authenticate with foo:bar:mac
[20048.323585] wlp3s0: send auth to foo:bar:mac (try 1/3)
[20048.870022] wlp3s0: send auth to foo:bar:mac (try 2/3)
[20049.894090] wlp3s0: send auth to foo:bar:mac (try 3/3)
[20050.822074] wlp3s0: authentication with foo:bar:mac timed out
[20054.369577] wlp3s0: authenticate with foo:bar:mac
[20054.372411] wlp3s0: send auth to foo:bar:mac (try 1/3)
[20054.378212] wlp3s0: authenticated
[20054.385910] wlp3s0: associate with foo:bar:mac (try 1/3)
[20054.389665] wlp3s0: RX AssocResp from foo:bar:mac (capab=0x411 status=0 aid=1)
[20054.408519] wlp3s0: associated
[20054.477361] wlp3s0: Limiting TX power to 21 (24 - 3) dBm as advertised by foo:bar:mac
[21457.817494] wlp3s0: deauthenticated from foo:bar:mac (Reason: 6=CLASS2_FRAME_FROM_NONAUTH_STA)
[21458.012566] wlp3s0: authenticate with foo:bar:mac
[21458.015934] wlp3s0: send auth to foo:bar:mac (try 1/3)
[21458.096198] wlp3s0: authenticated
[21458.099436] wlp3s0: associate with foo:bar:mac (try 1/3)
[21458.113810] wlp3s0: associate with foo:bar:mac (try 2/3)
[21458.128091] wlp3s0: RX AssocResp from foo:bar:mac (capab=0x411 status=0 aid=1)
[21458.147379] wlp3s0: associated
[21458.194553] wlp3s0: Limiting TX power to 21 (24 - 3) dBm as advertised by foo:bar:mac
[21467.566776] wlp3s0: authenticate with foo:bar:mac
[21467.569905] wlp3s0: send auth to foo:bar:mac (try 1/3)
[21467.618325] wlp3s0: authenticated
[21467.619300] wlp3s0: associate with foo:bar:mac (try 1/3)
[21467.621934] wlp3s0: RX AssocResp from foo:bar:mac (capab=0x411 status=0 aid=1)
[21467.640937] wlp3s0: associated
[21467.717941] wlp3s0: Limiting TX power to 27 (30 - 3) dBm as advertised by foo:bar:mac

What does the following report?

/usr/sbin/iwconfig wlan0

Not ideal, but some users have found disabling 802.11n is the only way to get a stable connection with some of these Intel Centrino N-series chipsets eg via /etc/modprobe.d/20-iwlwifi.conf

options iwlwifi 11n_disable=0

Worth trying on the fly before attempting to connect perhaps…

modprobe -r iwlwifi
modprobe iwlwifi 11n_disable=0

Any difference with connection stability?

iwconfig while connection is active

/usr/sbin/iwconfig wlp3s0wlp3s0    IEEE 802.11  ESSID:"foobar"  
          Mode:Managed  Frequency:5.805 GHz  Access Point: foo:bar:mac   
          Bit Rate=121.5 Mb/s   Tx-Power=15 dBm   
          Retry short limit:7   RTS thr:off   Fragment thr:off
          Power Management:off
          Link Quality=53/70  Signal level=-57 dBm  
          Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:0  Invalid misc:41   Missed beacon:0



dmesg after module reload with 11n_disable=0, link won’t come up in this condition.


[36474.677988] Intel(R) Wireless WiFi driver for Linux
[36474.677989] Copyright(c) 2003- 2015 Intel Corporation
[36474.678617] iwlwifi 0000:03:00.0: loaded firmware version 18.168.6.1 op_mode iwldvm
[36474.686199] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUG enabled
[36474.686201] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUGFS enabled
[36474.686202] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEVICE_TRACING disabled
[36474.686204] iwlwifi 0000:03:00.0: Detected Intel(R) Centrino(R) Advanced-N 6235 AGN, REV=0xB0
[36474.719162] ieee80211 phy1: Selected rate control algorithm 'iwl-agn-rs'
[36474.723425] iwlwifi 0000:03:00.0 wlp3s0: renamed from wlan0
[36474.777199] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36474.790276] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36475.090811] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36475.174326] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36475.199752] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36475.503349] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36475.586682] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36475.665558] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36478.851364] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36479.151855] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36479.234680] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36482.395893] wlp3s0: authenticate with foo:bar:mac
[36482.399024] wlp3s0: send auth to foo:bar:mac (try 1/3)
[36482.459877] wlp3s0: authenticated
[36482.460776] wlp3s0: associate with foo:bar:mac (try 1/3)
[36482.464581] wlp3s0: RX AssocResp from foo:bar:mac (capab=0x411 status=0 aid=1)
[36482.483344] wlp3s0: associated
[36482.487762] IPv6: ADDRCONF(NETDEV_CHANGE): wlp3s0: link becomes ready
[36482.558815] wlp3s0: Limiting TX power to 27 (30 - 3) dBm as advertised by foo:bar:mac
[36486.584463] wlp3s0: deauthenticating from foo:bar:mac by local choice (Reason: 3=DEAUTH_LEAVING)
[36487.085636] IPv6: ADDRCONF(NETDEV_UP): enp0s25: link is not ready
[36487.087682] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36487.100679] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36487.403666] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36487.486834] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36487.514034] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36487.816012] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36487.899743] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36487.956744] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36491.139356] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36491.443820] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36491.527202] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36491.537761] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36491.563524] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36491.867544] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36491.950587] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36506.490506] IPv6: ADDRCONF(NETDEV_UP): enp0s25: link is not ready
[36506.492779] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36506.505725] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36506.807978] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36506.890838] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36506.918191] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36507.219849] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36507.303739] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36507.384210] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36510.595309] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36510.899322] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36510.981660] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36510.991483] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[36511.017555] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36511.319063] iwlwifi 0000:03:00.0: Radio type=0x2-0x1-0x0
[36511.411701] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready



I’m no kernel expert but it seems to me that one module or service crashes or deadlocks and is not properly reset by reloading iwlwifi or restarting network. What else could be involved?

Are there options to get more debug logging from the iwlwifi module, or are there other firmwares I could try?

The problem just triggered now and I got logs from everything around the crash.

journalctl -u NetworkManager



Sep 19 05:51:48 s76 NetworkManager[1270]: <info>  [1537350708.6989] manager: startup complete
Sep 19 05:51:48 s76 NetworkManager[1270]: <info>  [1537350708.9381] manager: NetworkManager state is now CONNECTED_GLOBAL


Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6547] dhcp4 (wlp3s0):   address 192.168.0.137
Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6548] dhcp4 (wlp3s0):   plen 24 (255.255.255.0)
Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6548] dhcp4 (wlp3s0):   gateway 192.168.0.1
Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6549] dhcp4 (wlp3s0):   lease time 3600
Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6549] dhcp4 (wlp3s0):   hostname 's76'
Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6549] dhcp4 (wlp3s0):   nameserver '176.103.130.130'
Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6549] dhcp4 (wlp3s0):   nameserver '176.103.130.131'
Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6549] dhcp4 (wlp3s0):   domain name 'foobar'
Sep 19 06:15:56 s76 NetworkManager[1270]: <info>  [1537352156.6549] dhcp4 (wlp3s0): state changed bound -> bound


Sep 19 06:25:46 s76 NetworkManager[1270]: <warn>  [1537352746.4388] sup-iface[0x55f9d83870d0,wlp3s0]: connection disconnected (reason 3)
Sep 19 06:25:46 s76 NetworkManager[1270]: <info>  [1537352746.4615] device (wlp3s0): supplicant interface state: completed -> disconnected
Sep 19 06:25:46 s76 NetworkManager[1270]: <info>  [1537352746.5441] device (wlp3s0): supplicant interface state: disconnected -> scanning
Sep 19 06:25:49 s76 NetworkManager[1270]: <info>  [1537352749.6388] device (wlp3s0): supplicant interface state: scanning -> authenticating
Sep 19 06:25:49 s76 NetworkManager[1270]: <info>  [1537352749.7046] device (wlp3s0): supplicant interface state: authenticating -> associating
Sep 19 06:25:49 s76 NetworkManager[1270]: <info>  [1537352749.7362] device (wlp3s0): supplicant interface state: associating -> completed

dmesg


[Wed Sep 19 06:25:45 2018] wlp3s0: deauthenticated from foo:bar:mac (Reason: 3=DEAUTH_LEAVING)
[Wed Sep 19 06:25:49 2018] wlp3s0: authenticate with foo:bar:mac
[Wed Sep 19 06:25:49 2018] wlp3s0: send auth to foo:bar:mac (try 1/3)
[Wed Sep 19 06:25:49 2018] wlp3s0: authenticated
[Wed Sep 19 06:25:49 2018] wlp3s0: associate with foo:bar:mac (try 1/3)
[Wed Sep 19 06:25:49 2018] wlp3s0: RX AssocResp from foo:bar:mac (capab=0x411 status=0 aid=1)
[Wed Sep 19 06:25:49 2018] wlp3s0: associated
[Wed Sep 19 06:25:49 2018] wlp3s0: Limiting TX power to 27 (30 - 3) dBm as advertised by foo:bar:mac

iwconfig


wlp3s0    IEEE 802.11  ESSID:"foobar"  
          Mode:Managed  Frequency:5.805 GHz  Access Point: foo:bar:mac   
          Bit Rate=121.5 Mb/s   Tx-Power=15 dBm   
          Retry short limit:7   RTS thr:off   Fragment thr:off
          Power Management:off
          Link Quality=51/70  Signal level=-59 dBm  
          Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:12  Invalid misc:1   Missed beacon:0

I really don’t know what else to advise here. Some of the Intel Centrino N-series chipsets seem to be notorious for this kind of behaviour. Have a read of this thread for further discussion on driver options…

https://ubuntuforums.org/showthread.php?t=2264585

If nothing helps with resolving this consider submitting a bug report.

This caught my attention though…

Sep 19 06:25:46 s76 NetworkManager[1270]: <warn>  [1537352746.4388] sup-iface[0x55f9d83870d0,wlp3s0]: connection disconnected (reason 3)

That NM reason code is attributed to 'Device is now unmanaged ’
https://wiki.ubuntu.com/DebuggingNetworkManager/ReasonCodes

Are you sure that you only have one network management service active ie is wicked.service active at the same time NetworkManager.service is active?

I’m sure of nothing, just hoping the clean install would have a valid configuration lol!

I’m on a lapstop so NM is enabled and Wicked is supposed to be disabled, here’s my list of services:


**net**work.service                                                        enabled   
**Net**workManager-dispatcher.service                                      enabled   
**Net**workManager-wait-online.service                                     enabled   
**Net**workManager.service                                                 enabled   
systemd-**net**workd-wait-online.service                                   disabled  
systemd-**net**workd.service                                               disabled  
**wick**ed.service                                                         disabled  
**wick**edd-auto4.service                                                  disabled  
**wick**edd-dhcp4.service                                                  disabled  
**wick**edd-dhcp6.service                                                  disabled  
**wick**edd-nanny.service                                                  disabled  
**wick**edd-pppd@.service                                                  static    
**wick**edd.service                                                        indirect  
systemd-**net**workd.socket                                                disabled  
**net**work-online.target                                                  static    
**net**work-pre.target                                                     static    
**net**work.target                                                         static   

Thanks for the help!

Ok, so wicked is disabled and NetworkManager active as expected. All good from that perspective. Did you get a chance to review the thread I linked in post #6?

Yes, tried adding:

options iwlwifi 11n_disable=1 iwlwifi bt_coex_active=0

to modprobe.d, it was actually worst with this, connection lasts only a couple minutes.
I’ll go ahead and log an issue in bugzilla.

Thanks for the help

Found an existing issue and added my case to it:
https://bugzilla.opensuse.org/show_bug.cgi?id=1097605

Following the recommendations, I’ve installed the latest 4.4 kernel from Leap 42.3:

http://download.opensuse.org/repositories/Kernel:/openSUSE-42.3/standard/

It’s looking good right now, I’ll report back on how it holds up.

Thanks for update.

The 4.4 kernel from Leap 42.3 does the job! Have not lost the connection since switching. I can live with that…

Yes, as expected (based on the bug report findings). At least it’s a workaround until they get the driver regression fixed.

Well after a few days of running 4.4 it started degrading, maybe I was just lucky at first, now it breaks at any point for any reason. I give up on this, I’ve ordered a USB adapter to replace the builtin one. This laptop is getting pretty old and it’s been falling to pieces, I wouldn’t be surprised that the hw itself is faulty.