Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Wi-Fi broken on Dell E6520, Intel Centrino Advanced-N 6205 AGN, Debian 11.7 Xfce

65 views
Skip to first unread message

David Christensen

unread,
May 4, 2023, 4:20:08 PM5/4/23
to
debian-user:

I have a Dell Latituide E6520 laptop computer with Debian/ Xfce that is
used as for GUI desktop/ portable admin for my SOHO LAN:

https://www.dell.com/support/home/en-us/product-support/product/latitude-e6520/docs

2023-05-04 12:30:18 root@laalaa ~
# cat /etc/debian_version ; uname -a
11.7
Linux laalaa 5.10.0-22-amd64 #1 SMP Debian 5.10.178-3 (2023-04-22)
x86_64 GNU/Linux

2023-05-04 12:30:31 root@laalaa ~
# dpkg-query -W xfce4
xfce4 4.16


It has Ethernet that has been connected for years:

2023-05-04 12:36:23 root@laalaa ~
# dmesg | grep e1000
[ 1.806458] e1000e: Intel(R) PRO/1000 Network Driver
[ 1.806460] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[ 1.806688] e1000e 0000:00:19.0: Interrupt Throttling Rate (ints/sec)
set to dynamic conservative mode
[ 1.912063] e1000e 0000:00:19.0 0000:00:19.0 (uninitialized):
registered PHC clock
[ 2.009774] e1000e 0000:00:19.0 eth0: (PCI Express:2.5GT/s:Width x1)
5c:26:0a:7c:82:88
[ 2.009776] e1000e 0000:00:19.0 eth0: Intel(R) PRO/1000 Network
Connection
[ 2.009814] e1000e 0000:00:19.0 eth0: MAC: 10, PHY: 11, PBA No:
4041FF-0FF
[ 3.572969] e1000e 0000:00:19.0 eno1: renamed from eth0
[ 85.528065] e1000e 0000:00:19.0 eno1: NIC Link is Up 1000 Mbps Full
Duplex, Flow Control: Rx/Tx


It also has Wi-Fi, which has been switched off for many months (via the
wireless switch on the right side of the laptop). I switched on Wi-Fi
today, booted, logged in, and chose Xfce Panel -> right click
NetworkManager Applet -> Enable Wi-Fi. Wi-Fi does not start.


Powering off, unplugging Ethernet, booting, logging in, and chosing
Xfce Panel -> right click NetworkManager Applet -> Enable Wi-Fi produces
the same failure to start Wi-Fi.


Ethernet and Wi-Fi have both worked in the past; both separately and
simultaneously.


After failing to start Wi-Fi three times, I see:

2023-05-04 12:37:39 root@laalaa ~
# dmesg | grep -i wifi
[ 79.953085] Intel(R) Wireless WiFi driver for Linux
[ 79.953314] iwlwifi 0000:03:00.0: can't disable ASPM; OS doesn't have
ASPM control
[ 79.960679] iwlwifi 0000:03:00.0: firmware: direct-loading firmware
iwlwifi-6000g2a-6.ucode
[ 79.960863] iwlwifi 0000:03:00.0: loaded firmware version 18.168.6.1
6000g2a-6.ucode op_mode iwldvm
[ 79.960889] iwlwifi 0000:03:00.0: firmware: failed to load
iwl-debug-yoyo.bin (-2)
[ 80.057548] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUG disabled
[ 80.057551] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUGFS disabled
[ 80.057552] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEVICE_TRACING disabled
[ 80.057553] iwlwifi 0000:03:00.0: Detected Intel(R) Centrino(R)
Advanced-N 6205 AGN, REV=0xB0
[ 80.070510] iwlwifi 0000:03:00.0: reporting RF_KILL (radio disabled)
[ 80.070571] iwlwifi 0000:03:00.0: RF_KILL bit toggled to disable radio.
[ 80.095718] iwlwifi 0000:03:00.0 wlp3s0: renamed from wlan0



2023-05-04 13:10:09 root@laalaa ~
# tail /var/log/syslog
May 4 13:07:38 laalaa systemd[1629]: Reached target Basic System.
May 4 13:07:38 laalaa systemd[1629]: Condition check resulted in
Multimedia Service being skipped.
May 4 13:07:38 laalaa systemd[1629]: Condition check resulted in Sound
Service being skipped.
May 4 13:07:38 laalaa systemd[1629]: Reached target Main User Target.
May 4 13:07:38 laalaa systemd[1629]: Startup finished in 118ms.
May 4 13:07:38 laalaa systemd[1]: Started User Manager for UID 0.
May 4 13:07:38 laalaa systemd[1]: Started Session 4 of user root.
May 4 13:07:39 laalaa systemd[1]: NetworkManager-dispatcher.service:
Succeeded.
May 4 13:08:43 laalaa NetworkManager[903]: <info> [1683230923.9058]
audit: op="radio-control" arg="wireless-enabled" pid=1483 uid=13250
result="success"
May 4 13:09:08 laalaa NetworkManager[903]: <info> [1683230948.0509]
audit: op="radio-control" arg="wireless-enabled" pid=1483 uid=13250
result="success"


The first attempt left no message in /var/log/syslog. The second and
third attempts left the last two entries shown above.


STFW "debian iwlwifi firmware: failed to load iwl-debug-yoyo.bin (-2)",
I see:

https://forums.debian.net/viewtopic.php?t=149817

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=966218


AIUI the posts indicate that the message "iwlwifi firmware: failed to
load iwl-debug-yoyo.bin (-2)" is a warning only; it did not affect
operation (?) (as of November 2022).


Is anyone else running an Intel(R) Centrino(R) Advanced-N 6205 AGN,
REV=0xB0 Wi-F- adapter?


Has anyone else seen this issue?


Comments or suggestions for gathering information and/or getting Wi-Fi
working?


TIA,

David

Christoph Brinkhaus

unread,
May 4, 2023, 4:40:06 PM5/4/23
to
Am Thu, May 04, 2023 at 01:14:12PM -0700 schrieb David Christensen:

Hello David,

[...]

> Ethernet and Wi-Fi have both worked in the past; both separately and
> simultaneously.
>
>
> After failing to start Wi-Fi three times, I see:
>
> 2023-05-04 12:37:39 root@laalaa ~
> # dmesg | grep -i wifi
> [ 79.953085] Intel(R) Wireless WiFi driver for Linux
> [ 79.953314] iwlwifi 0000:03:00.0: can't disable ASPM; OS doesn't have
> ASPM control
> [ 79.960679] iwlwifi 0000:03:00.0: firmware: direct-loading firmware
> iwlwifi-6000g2a-6.ucode
> [ 79.960863] iwlwifi 0000:03:00.0: loaded firmware version 18.168.6.1
> 6000g2a-6.ucode op_mode iwldvm
> [ 79.960889] iwlwifi 0000:03:00.0: firmware: failed to load
> iwl-debug-yoyo.bin (-2)

I see the same message with a different Intel chip, too.
And I also seen docs which state that this is ok.

> [ 80.057548] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUG disabled
> [ 80.057551] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUGFS disabled
> [ 80.057552] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEVICE_TRACING disabled
> [ 80.057553] iwlwifi 0000:03:00.0: Detected Intel(R) Centrino(R)
> Advanced-N 6205 AGN, REV=0xB0
> [ 80.070510] iwlwifi 0000:03:00.0: reporting RF_KILL (radio disabled)
> [ 80.070571] iwlwifi 0000:03:00.0: RF_KILL bit toggled to disable radio.
> [ 80.095718] iwlwifi 0000:03:00.0 wlp3s0: renamed from wlan0
>
> Comments or suggestions for gathering information and/or getting Wi-Fi
> working?

In the past I have had regular issues due to my fat fingers which
touched the Wi-Fi off button on my keyboard. At that time I have not
noticed that the issue has been me :-).

There is a directory /var/lib/systemd/rfkill/ which holds a file named
by the Wi-Fi device. Here it is pci-0000:04:00.0:wlan. The content is
just a 0 for normal operation or 1 for Wi-Fi off which means RF_Kill.

Please check that file on your system and replace a possible 1 by a 0.

There might be a better method involving systemd to achieve the same.
But I am not aware of that. There is also a configuration bit whether
to enable Wi-Fi after reboot or to keep the previous setting. But I do
not remember the details right now.

Kind regards,
Christoph
--
Ist die Katze gesund
schmeckt sie dem Hund.
signature.asc

David Christensen

unread,
May 4, 2023, 5:30:05 PM5/4/23
to
Thank you for replying. :-)


There is no /var/lib/systemd directory on the laptop:

2023-05-04 14:23:25 root@taz ~
# ls -1 /var/lib/systemd/
catalog
coredump
deb-systemd-helper-enabled
deb-systemd-user-helper-enabled
linger
pstore
random-seed
timers
timesync


David

zithro

unread,
May 4, 2023, 6:11:54 PM5/4/23
to
On 04 May 2023 23:26, David Christensen wrote:
> On 5/4/23 13:36, Christoph Brinkhaus wrote:
>> Am Thu, May 04, 2023 at 01:14:12PM -0700 schrieb David Christensen:
>
>>> [   80.070510] iwlwifi 0000:03:00.0: reporting RF_KILL (radio disabled)
>>> [   80.070571] iwlwifi 0000:03:00.0: RF_KILL bit toggled to disable radio.

This indicates rfkill is effectively blocking the wlan: "radio disabled"

>> There is a directory /var/lib/systemd/rfkill/ which holds a file named
>> by the Wi-Fi device. Here it is pci-0000:04:00.0:wlan. The content is
>> just a 0 for normal operation or 1 for Wi-Fi off which means RF_Kill.
>>
>> Please check that file on your system and replace a possible 1 by a 0.
>>
>
> There is no /var/lib/systemd directory on the laptop:
>

Maybe it's elsewhere in /var ?
$ find /var -type d -name rfkill

Or in sysfs :
$ find /sys -type d -name rfkill

Else, look everywhere ^^

Things to check :

Check the status via systemctl (didn't know there was a service):
# systemctl --status rfkill

Is the rfkill *module* loaded ?
# lsmod | grep rfkill
If nothing, try loading the module :
# modprobe rfkill

Is the rfkill *tool* installed ?
# dpkg -l "*rfkill*"
If not, install it
# apt install rfkill

David Christensen

unread,
May 4, 2023, 6:30:06 PM5/4/23
to
On 5/4/23 14:59, zithro wrote:
> On 04 May 2023 23:26, David Christensen wrote:
>> On 5/4/23 13:36, Christoph Brinkhaus wrote:
>>> Am Thu, May 04, 2023 at 01:14:12PM -0700 schrieb David Christensen:
>>
>>>> [   80.070510] iwlwifi 0000:03:00.0: reporting RF_KILL (radio disabled)
>>>> [   80.070571] iwlwifi 0000:03:00.0: RF_KILL bit toggled to disable
>>>> radio.
>
> This indicates rfkill is effectively blocking the wlan: "radio disabled"
>
>>> There is a directory /var/lib/systemd/rfkill/ which holds a file named
>>> by the Wi-Fi device. Here it is pci-0000:04:00.0:wlan. The content is
>>> just a 0 for normal operation or 1 for Wi-Fi off which means RF_Kill.
>>>
>>> Please check that file on your system and replace a possible 1 by a 0.
>>>
>>
>> There is no /var/lib/systemd directory on the laptop:
>>
>
> Maybe it's elsewhere in /var ?
> $ find /var -type d -name rfkill
>
> Or in sysfs :
> $ find /sys -type d -name rfkill
>
> Else, look everywhere ^^


Thank you for replying. :-)


D'oh! The previous post was the wrong machine...


This is the laptop:

2023-05-04 15:10:58 root@laalaa ~
# find / -xdev -name rfkill
/var/lib/systemd/rfkill
/usr/src/linux-headers-5.10.0-22-amd64/include/config/rfkill
/usr/src/linux-headers-5.10.0-21-amd64/include/config/rfkill
/usr/lib/modules/5.10.0-21-amd64/kernel/net/rfkill
/usr/lib/modules/5.10.0-22-amd64/kernel/net/rfkill

2023-05-04 15:13:04 root@laalaa ~
# ls -l /var/lib/systemd/rfkill
total 12
-rw-r--r-- 1 root root 2 May 4 12:15 pci-0000:03:00.0:wlan
-rw-r--r-- 1 root root 2 May 24 2022 pci-0000:0b:00.0:wlan
-rw-r--r-- 1 root root 2 May 4 13:04 platform-dell-laptop:wlan

2023-05-04 15:13:13 root@laalaa ~
# file /var/lib/systemd/rfkill/*
/var/lib/systemd/rfkill/pci-0000:03:00.0:wlan: ASCII text
/var/lib/systemd/rfkill/pci-0000:0b:00.0:wlan: ASCII text
/var/lib/systemd/rfkill/platform-dell-laptop:wlan: ASCII text

2023-05-04 15:13:39 root@laalaa ~
# cat /var/lib/systemd/rfkill/pci-0000\:03\:00.0\:wlan

0

2023-05-04 15:14:35 root@laalaa ~
# cat /var/lib/systemd/rfkill/pci-0000\:0b\:00.0\:wlan

0

2023-05-04 15:14:48 root@laalaa ~
# cat /var/lib/systemd/rfkill/platform-dell-laptop\:wlan

0

2023-05-04 15:17:53 root@laalaa ~
# hexdump /var/lib/systemd/rfkill/platform-dell-laptop\:wlan
00000000 30 0a |0.|
00000002


So, there are three (3) files matching the description offered by
Christoph Brinkhaus, and they all contain "0\n".

>
> Things to check :
>
> Check the status via systemctl (didn't know there was a service):
> # systemctl --status rfkill


2023-05-04 15:15:40 root@laalaa ~
# systemctl --status rfkill
systemctl: unrecognized option '--status'

>
> Is the rfkill *module* loaded ?
> # lsmod | grep rfkill


2023-05-04 15:15:49 root@laalaa ~
# lsmod | grep rfkill
rfkill 32768 4 dell_laptop,dell_rbtn,cfg80211


> If nothing, try loading the module :
> # modprobe rfkill
>
> Is the rfkill *tool* installed ?
> # dpkg -l "*rfkill*"


2023-05-04 15:17:28 root@laalaa ~
# dpkg -l "*rfkill*"
dpkg-query: no packages found matching *rfkill*


> If not, install it
> # apt install rfkill


Comments or suggestions?


David

zithro

unread,
May 4, 2023, 7:20:06 PM5/4/23
to
On 05 May 2023 00:24, David Christensen wrote:
> # ls -l /var/lib/systemd/rfkill
> total 12
> -rw-r--r-- 1 root root 2 May  4 12:15 pci-0000:03:00.0:wlan
> -rw-r--r-- 1 root root 2 May 24  2022 pci-0000:0b:00.0:wlan
> -rw-r--r-- 1 root root 2 May  4 13:04 platform-dell-laptop:wlan

Strange, looks like you have two wlan NICs ? Oh, maybe it's the
bluetooth module.
Dunno, wireless sucks anyways ^^
So, to determine the correct WLAN adapter run both :

lspci -s 03:00.0
lspci -s 0b:00.0

> So, there are three (3) files matching the description offered by
> Christoph Brinkhaus, and they all contain "0\n".

This is strange, he indicated that 0 is for normal ops, and 1 for blocked.

> # systemctl --status rfkill
> systemctl: unrecognized option '--status'

My bad, it's
# systemctl status rfkill

>> Is the rfkill *module* loaded ?
>> # lsmod | grep rfkill
> # lsmod | grep rfkill
> rfkill                 32768  4 dell_laptop,dell_rbtn,cfg80211

Good, it's loaded.
And there's a hint, the module "dell_rbtn", which should mean dell radio
button.

>> Is the rfkill *tool* installed ?
> # dpkg -l "*rfkill*"
> dpkg-query: no packages found matching *rfkill*
>> If not, install it
>> # apt install rfkill

You did not install it :p

David Christensen

unread,
May 4, 2023, 8:00:13 PM5/4/23
to
On 5/4/23 16:12, zithro wrote:
> On 05 May 2023 00:24, David Christensen wrote:
>> # ls -l /var/lib/systemd/rfkill
>> total 12
>> -rw-r--r-- 1 root root 2 May  4 12:15 pci-0000:03:00.0:wlan
>> -rw-r--r-- 1 root root 2 May 24  2022 pci-0000:0b:00.0:wlan
>> -rw-r--r-- 1 root root 2 May  4 13:04 platform-dell-laptop:wlan
>
> Strange, looks like you have two wlan NICs ? Oh, maybe it's the
> bluetooth module.
> Dunno, wireless sucks anyways ^^
> So, to determine the correct WLAN adapter run both :
>
> lspci -s 03:00.0
> lspci -s 0b:00.0


2023-05-04 16:23:44 root@laalaa ~
# lspci -s 03:00.0
03:00.0 Network controller: Intel Corporation Centrino Advanced-N 6205
[Taylor Peak] (rev 34)

2023-05-04 16:23:50 root@laalaa ~
# lspci -s 0b:00.0
0b:00.0 FireWire (IEEE 1394): O2 Micro, Inc. 1394 OHCI Compliant Host
Controller (rev 05)

>
>> So, there are three (3) files matching the description offered by
>> Christoph Brinkhaus, and they all contain "0\n".
>
> This is strange, he indicated that 0 is for normal ops, and 1 for blocked.
>
>> # systemctl --status rfkill
>> systemctl: unrecognized option '--status'
>
> My bad, it's
> # systemctl status rfkill


2023-05-04 16:23:57 root@laalaa ~
# systemctl status rfkill
Unit rfkill.service could not be found.

>
>>> Is the rfkill *module* loaded ?
>>> # lsmod | grep rfkill
>> # lsmod | grep rfkill
>> rfkill                 32768  4 dell_laptop,dell_rbtn,cfg80211
>
> Good, it's loaded.
> And there's a hint, the module "dell_rbtn", which should mean dell radio
> button.
>
>>> Is the rfkill *tool* installed ?
>> # dpkg -l "*rfkill*"
>> dpkg-query: no packages found matching *rfkill*
>>> If not, install it
>>> # apt install rfkill
>
> You did not install it :p


2023-05-04 16:24:49 root@laalaa ~
# apt install rfkill
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
The following NEW packages will be installed:
rfkill
0 upgraded, 1 newly installed, 0 to remove and 0 not upgraded.
Need to get 88.6 kB of archives.
After this operation, 151 kB of additional disk space will be used.
Get:1 http://deb.debian.org/debian bullseye/main amd64 rfkill amd64
2.36.1-8+deb11u1 [88.6 kB]
Fetched 88.6 kB in 0s (445 kB/s)
Selecting previously unselected package rfkill.
(Reading database ... 167563 files and directories currently installed.)
Preparing to unpack .../rfkill_2.36.1-8+deb11u1_amd64.deb ...
Unpacking rfkill (2.36.1-8+deb11u1) ...
Setting up rfkill (2.36.1-8+deb11u1) ...
Processing triggers for man-db (2.9.4-2) ...


Reboot. Login. Xfce Panel -> right click NetworkManager -> Enable
Wi-Fi -- Wi-Fi did not start. Try Enable Wi-Fi a couple three more
times -- nope:

2023-05-04 16:32:26 root@laalaa ~
# tail /var/log/syslog
May 4 16:30:37 laalaa systemd[1552]: Reached target Main User Target.
May 4 16:30:37 laalaa systemd[1552]: Startup finished in 125ms.
May 4 16:30:37 laalaa systemd[1]: Started User Manager for UID 0.
May 4 16:30:37 laalaa systemd[1]: Started Session 5 of user root.
May 4 16:31:17 laalaa NetworkManager[850]: <info> [1683243077.3070]
audit: op="radio-control" arg="wireless-enabled" pid=1418 uid=13250
result="success"
May 4 16:32:21 laalaa NetworkManager[850]: <info> [1683243141.1786]
audit: op="radio-control" arg="wireless-enabled" pid=1418 uid=13250
result="success"
May 4 16:32:21 laalaa systemd[1]: Started Run anacron jobs.
May 4 16:32:21 laalaa anacron[1593]: Anacron 2.3 started on 2023-05-04
May 4 16:32:21 laalaa anacron[1593]: Normal exit (0 jobs run)
May 4 16:32:21 laalaa systemd[1]: anacron.service: Succeeded.


Try Debian 9 -- nope.


Try Windows 7 Pro -- nope.


It looks like I have a dead Wi-Fi adapter.


Thank you for the help. :-)

David

zithro

unread,
May 4, 2023, 9:20:07 PM5/4/23
to
On 05 May 2023 01:55, David Christensen wrote:
> Try Enable Wi-Fi a couple three more times -- nope
>
> Try Debian 9 -- nope.
>
> Try Windows 7 Pro -- nope.
>
> It looks like I have a dead Wi-Fi adapter.

Or a dead Wifi killswitch (the physical radio button) ?

Last thought, don't you have :
- a mechanical ON/OFF button (I mean like on your walls for the lights)
- AND a Fn+keyboard key combo to toggle Wifi (enable/disable) ?

On my old Thinkpad, I have both, whereas on a newer model I only have
the Fn+key combo.

> Thank you for the help.  :-)

Np, sorry for your loss ^^

David Christensen

unread,
May 4, 2023, 11:00:06 PM5/4/23
to
On 5/4/23 18:12, zithro wrote:
> On 05 May 2023 01:55, David Christensen wrote:
>> Try Enable Wi-Fi a couple three more  times -- nope
>>
>> Try Debian 9 -- nope.
>>
>> Try Windows 7 Pro -- nope.
>>
>> It looks like I have a dead Wi-Fi adapter.
>
> Or a dead Wifi killswitch (the physical radio button) ?


I tested the wireless switch and was seeing messages in /var/log/syslog.
Without a known good Wi-Fi adapter and a known good switch, I cannot
determine if one or both are good or bad.


The "Dell Latitude E6520 Owner's Manual" includes instructions for
replacing the Wireless Local Area Network (WLAN) Card. It's 12 steps
and I've done those before.


The "Dell Latitude E6520 Owner's Manual" includes instructions for
removing and installing the Wireless Switch Board. You have to dig deep
-- 26 steps.

>
> Last thought, don't you have :
> - a mechanical ON/OFF button (I mean like on your walls for the lights)
> - AND a Fn+keyboard key combo to toggle Wifi (enable/disable) ?


The keyboard has a dedicated "Fn" key and a few other keys with extra
symbols in the same color. I looked hard for a Wi-Fi symbol, but never
found one. I seem to recall that the wireless switch on the side is the
only hardware user interface on this laptop.


"Dell Latitude E6420/E6520 Setup and Features Information" documents the
"wireless switch" on the right side, but does not document the "Fn" and
related keys on the keyboard.


"Dell Latitude E6520 Owner's Manual" only mentions using "Fn" key to
enable/disable the keyboard embedded keyboard.

>
> On my old Thinkpad, I have both, whereas on a newer model I only have
> the Fn+key combo.


My older Dell Inspiron E1505 has the the Fn+key combo.

>
>> Thank you for the help.  :-)
>
> Np, sorry for your loss ^^


I was going to surf eBay and find another Wi-Fi card, but then I had one
last idea -- double-checking the the CMOS/NVRAM settings via Setup. I
cut the settings down to the bare minimum for Wi-Fi:

Settings
+ Wireless
+ Wireless Switch
| WWAN unchecked
| WLAN checked
| Bluetooth unchecked
+ Wireless Device Enable
WWAN unchecked
WLAN checked
Bluetooth unchecked
-> Apply
-> Exit


Now Wi-Fi works with my AT&T residential gateway.


I know the machine has Wi-Fi and it does not have WWAN. I don't know if
the Wi-Fi card includes Bluetooth. But, "if it ain't broke, don't fix it'.


That said, Wi-Fi does not work with my Ubuiquitti Networks UniFi setup
-- it doesn't like the MAC address.


Damn infernal computers and networks...


David

David Christensen

unread,
May 5, 2023, 12:10:06 AM5/5/23
to
On 5/4/23 19:52, David Christensen wrote:
> That said, Wi-Fi does not work with my Ubuiquitti Networks UniFi setup
> -- it doesn't like the MAC address.


A further problem -- Debian is changing the MAC address of the Wi-Fi
adapter ("MAC address spoofing", a security "feature") to MAC addresses
that my networking equipment (Ubiquiti Networks UniFi) detects as invalid:

"This field must be a valid MAC address"


The solution was to change the Wi-Fi settings in Debian:

Xfce Panel
-> right-click NetworkManager Applet
-> Edit Connections
-> wifi.tracy.holgerdanske.com
-> gear (Edit the selected connection"
-> Wi-Fi
-> Cloned MAC Address -> Permanent.


David

zithro

unread,
May 5, 2023, 6:31:50 AM5/5/23
to
On 05 May 2023 04:52, David Christensen wrote:
> I was going to surf eBay and find another Wi-Fi card, but then I had one
> last idea -- double-checking the the CMOS/NVRAM settings via Setup.  I
> cut the settings down to the bare minimum for Wi-Fi:
>
>   Settings
>   + Wireless
>     + Wireless Switch
>     |    WWAN           unchecked
>     |    WLAN           checked
>     |    Bluetooth      unchecked
>     +  Wireless Device Enable
>          WWAN           unchecked
>          WLAN           checked
>          Bluetooth      unchecked
>   -> Apply
>   -> Exit
>

Wow, good job, didn't think about the firmware.
For info, what were the settings BEFORE you changed them ?

> I know the machine has Wi-Fi and it does not have WWAN.  I don't know if
> the Wi-Fi card includes Bluetooth.  But, "if it ain't broke, don't fix it'.

Don't touch anything, and look up your card on the net ;)
If you don't use it now, maybe you'll need it one day, and you may have
forgotten you disabled it in the FW.

> Damn infernal computers and networks...

Too easy, not funny ^^

zithro

unread,
May 5, 2023, 6:40:07 AM5/5/23
to
On 05 May 2023 06:08, David Christensen wrote:
> On 5/4/23 19:52, David Christensen wrote:
>> That said, Wi-Fi does not work with my Ubuiquitti Networks UniFi setup
>> -- it doesn't like the MAC address.
>
>
> A further problem -- Debian is changing the MAC address of the Wi-Fi
> adapter ("MAC address spoofing", a security "feature") to MAC addresses
> that my networking equipment (Ubiquiti Networks UniFi) detects as invalid:
>
>     "This field must be a valid MAC address"

That would be nice for them to tell you what THEY consider a valid MAC
addr, as I don't think Debian produces MACs with z or x in it ...
Maybe they check against a valid manufacturer (which is the 1st half of
the MAC addr, iirc called OUI) ?
I'm curious if the Xen OUI is valid for them (00:16:3e:nn:nn:nn).

David Christensen

unread,
May 5, 2023, 6:51:29 AM5/5/23
to
On 5/5/23 03:25, zithro wrote:
> On 05 May 2023 04:52, David Christensen wrote:
>> I was going to surf eBay and find another Wi-Fi card, but then I had
>> one last idea -- double-checking the the CMOS/NVRAM settings via
>> Setup.  I cut the settings down to the bare minimum for Wi-Fi:
>>
>>    Settings
>>    + Wireless
>>      + Wireless Switch
>>      |    WWAN           unchecked
>>      |    WLAN           checked
>>      |    Bluetooth      unchecked
>>      +  Wireless Device Enable
>>           WWAN           unchecked
>>           WLAN           checked
>>           Bluetooth      unchecked
>>    -> Apply
>>    -> Exit
>>
>
> Wow, good job, didn't think about the firmware.
> For info, what were the settings BEFORE you changed them ?


Settings
+ Wireless
+ Wireless Switch
| WWAN checked
| WLAN checked
| Bluetooth checked
+ Wireless Device Enable
WWAN checked
WLAN checked
Bluetooth checked

>
>> I know the machine has Wi-Fi and it does not have WWAN.  I don't know
>> if the Wi-Fi card includes Bluetooth.  But, "if it ain't broke, don't
>> fix it'.
>
> Don't touch anything, and look up your card on the net ;)
> If you don't use it now, maybe you'll need it one day, and you may have
> forgotten you disabled it in the FW.


Agreed.

>
>> Damn infernal computers and networks...
>
> Too easy, not funny ^^


David

Charles Curley

unread,
May 5, 2023, 9:41:07 AM5/5/23
to
On Thu, 4 May 2023 21:08:07 -0700
David Christensen <dpch...@holgerdanske.com> wrote:

> A further problem -- Debian

Well, it isn't really Debian, it's Network Manager.

> is changing the MAC address of the Wi-Fi
> adapter ("MAC address spoofing", a security "feature") to MAC
> addresses

This is a security feature for road warriors. Many access points (APs)
track users by MAC address, and use that data to feed ads, and
otherwise be rude. So if you are using public APs, feel free to use
random MAC addresses.

> that my networking equipment (Ubiquiti Networks UniFi)
> detects as invalid:
>
> "This field must be a valid MAC address"

On your home or work network you may have good reason to use the
manufacturer's MAC address. For example, your DCP server may assign
host name, IP address and other features based on MAC address. So you
should use the permanent address on known networks.

>
>
> The solution was to change the Wi-Fi settings in Debian:

Right. You can also do it via the command line, using nmcli. Which
makes it easy to script.

https://fedoramagazine.org/randomize-mac-address-nm/

--
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

Charles Curley

unread,
May 5, 2023, 9:51:43 AM5/5/23
to
On Fri, 5 May 2023 12:29:53 +0200
zithro <sl...@rabbit.lu> wrote:

> That would be nice for them to tell you what THEY consider a valid
> MAC addr, as I don't think Debian produces MACs with z or x in it ...
> Maybe they check against a valid manufacturer (which is the 1st half
> of the MAC addr, iirc called OUI) ?

See the Debian package ieee-data.


> I'm curious if the Xen OUI is valid for them (00:16:3e:nn:nn:nn).

It appears to be.

root@freeman:~# grep 16:3e /var/lib/arpwatch/ethercodes.db
0:16:3e Xensource, Inc.
root@freeman:~#

zithro

unread,
May 5, 2023, 11:30:05 AM5/5/23
to
On 05 May 2023 15:46, Charles Curley wrote:
> On Fri, 5 May 2023 12:29:53 +0200
> zithro <sl...@rabbit.lu> wrote:
>
>> That would be nice for them to tell you what THEY consider a valid
>> MAC addr, as I don't think Debian produces MACs with z or x in it ...
>> Maybe they check against a valid manufacturer (which is the 1st half
>> of the MAC addr, iirc called OUI) ?
>
> See the Debian package ieee-data.

ieee-data, another discovery today \o/
But note that in my sentence, "THEY" did refer to the Ubiquity AP the OP
was using, *not* Debian or NM.

>> I'm curious if the Xen OUI is valid for them (00:16:3e:nn:nn:nn).
>
> It appears to be.
>
> root@freeman:~# grep 16:3e /var/lib/arpwatch/ethercodes.db
> 0:16:3e Xensource, Inc.

Again, them == Ubiquity AP ;)

Xen recommends using its OUI on vifs, but does not enforce it, in part
for the reasons you wrote in your other post.

David Christensen

unread,
May 6, 2023, 5:01:56 PM5/6/23
to
debian-user:

I did some more trouble-shooting and found a better solution:

* Dell Latitude E6520 used as desktop/admin on SOHO network:

2023-05-06 13:51:42 root@laalaa ~
# cat /etc/debian_version;uname -a;dpkg-query -W xfce4;lspci|grep Centrino
11.7
Linux laalaa 5.10.0-22-amd64 #1 SMP Debian 5.10.178-3 (2023-04-22)
x86_64 GNU/Linux
xfce4 4.16
03:00.0 Network controller: Intel Corporation Centrino Advanced-N 6205
[Taylor Peak] (rev 34)

* Default Setup/CMOS/NVRAM settings:

Settings
+ Wireless
+ Wireless Switch
| WWAN checked
| WLAN checked
| Bluetooth checked
+ Wireless Device Enable
WWAN checked
WLAN checked
Bluetooth checked

* Network Manager settings:

Xfce panel
-> right click NetworkManager Applet
-> Edit Connections
-> Wi-Fi
-> wifi.tracy.holgerdanske.com
-> gear (Edit the selected connection)

General
Connect automatically with priority 0
Wi-Fi
SSID wifi.tracy.holgerdanske.com
Device wlp3s0
Wi-Fi Security
Security WPA & WPA2 Personal
Password ********
Proxy
Method None
IPv4 Settings
Method Automatic (DHCP)
IPv6 Settings
Method Automatic


Now when I boot and login, Wi-Fi connects automatically. :-)


David
0 new messages