Wifi code #4078

jvonau · 2025-08-31T21:23:02Z

Fixes bug:

should workaround #4077

Description of changes proposed in this pull request:

add networkctl feedback
Skip matching channel and country_code just record detected values.
#4076 (comment)
#4076 (comment)
use nmcli radio wifi on

Smoke-tested on which OS or OS's:

pending

Mention a team member @username e.g. to help with code review:

jvonau · 2025-08-31T22:36:04Z

#4077 (comment)

jvonau · 2025-08-31T23:59:26Z

Seems to work.

jvonau · 2025-09-01T01:04:01Z

Clean wired install on Trixie
https://paste.centos.org/view/7d0dccb8
Toggle network_installed and network_enabled
sudo iiab-network
https://paste.centos.org/view/918a0b4f

holta · 2025-09-01T01:43:35Z

What OS's are most useful to test here?

Maybe Linux Mint 22.2 and a few others?

jvonau · 2025-09-01T02:08:23Z

The 'follows NetworkManager' commits removes the WiFi constraints around channel and just use the supplied country_code as it appears that hostapd just follows what NetworkManager has detected to be in use for both. If you are going to test Mint it should be on a hard metal WiFi based install, should be the same as RasPiOS. Ubuntu desktops should tested the same way but should behave the same way as NetworkManager is the backend. Ubuntu server images don't have iw installed and use systemd-network as the backend so those images won't work without setting the WiFi up in netplan first. WiFi installs will need to reboot to have hotspot active due to having no_net_restart set to True

jvonau · 2025-09-01T03:18:15Z

What OS's are most useful to test here?

The target of the nmcli radio wifi on change is just RasPiOS wired installs as both Stage2 and Stage4 images use NetworkManager as the backend on both but the WiFi is soft-blocked by default.

holta · 2025-09-01T05:26:52Z

Very helpful explanations; Thanks.

I (and others if they can?) will try to help with testing on Monday (tomorrow).

jvonau · 2025-09-01T13:44:53Z

Being curious I ran raspi-config setting the WiFi country to GB and rebooted
https://paste.centos.org/view/560438ef
Line 1230

COMMAND: /usr/sbin/iw reg get # Detected WiFi country code / legal frequencies
global
country US: DFS-FCC
(902 - 904 @ 2), (N/A, 30), (N/A)
(904 - 920 @ 16), (N/A, 30), (N/A)
(920 - 928 @ 8), (N/A, 30), (N/A)
(2400 - 2472 @ 40), (N/A, 30), (N/A)
(5150 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
(5250 - 5350 @ 80), (N/A, 24), (0 ms), DFS, AUTO-BW
(5470 - 5730 @ 160), (N/A, 24), (0 ms), DFS
(5730 - 5850 @ 80), (N/A, 30), (N/A), AUTO-BW
(5850 - 5895 @ 40), (N/A, 27), (N/A), NO-OUTDOOR, AUTO-BW, PASSIVE-SCAN
(5925 - 7125 @ 320), (N/A, 12), (N/A), NO-OUTDOOR, PASSIVE-SCAN
(57240 - 71000 @ 2160), (N/A, 40), (N/A)
phy#0
country 99: DFS-UNSET
(2402 - 2482 @ 40), (6, 20), (N/A)
(2474 - 2494 @ 20), (6, 20), (N/A)
(5140 - 5360 @ 160), (6, 20), (N/A)
(5460 - 5860 @ 160), (6, 20), (N/A)

Line 1488

Sep 01 08:14:24 pi500 hostapd[791]: ap0: interface state UNINITIALIZED->COUNTRY_UPDATE

Line 1898

cmdline.txt has cfg80211.ieee80211_regdom=GB

The morel of the story is testing over the years has shown the order from greatest to least is NetworkManager, hostapd, cmdline for WiFi Regulatory Domain Codes priority.

jvonau · 2025-09-01T18:26:28Z

4: ap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel master br0 state UP group default qlen 1000
link/ether 02:af:6c:1d:f4:df brd ff:ff:ff:ff:ff:ff permaddr 2e:cf:67:68:1c:4a
inet6 fe80::af:6cff:fe1d:f4df/64 scope link proto kernel_ll
valid_lft forever preferred_lft forever

ecbaa88

4: ap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel master br0 state UP group default qlen 1000
link/ether 02:af:6c:1d:f4:df brd ff:ff:ff:ff:ff:ff permaddr 2e:cf:67:68:1c:4a

https://github.com/iiab/iiab/blob/master/roles/network/tasks/install.yml#L74 is ineffective in disabling ipv6

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
link/ether 2c:cf:67:c8:1c:49 brd ff:ff:ff:ff:ff:ff
inet 10.0.0.35/24 brd 10.0.0.255 scope global dynamic noprefixroute eth0
valid_lft 172772sec preferred_lft 172772sec
inet6 2604:3d09:a57a:e000::a760/128 scope global dynamic noprefixroute
valid_lft 172774sec preferred_lft 172774sec
inet6 2604:3d09:a57a:e000:8d24:1a43:a40c:719/64 scope global dynamic noprefixroute
valid_lft 299sec preferred_lft 299sec
inet6 fe80::8f1b:16fe:8469:c257/64 scope link noprefixroute

holta · 2025-09-01T18:40:46Z

https://github.com/iiab/iiab/blob/master/roles/network/tasks/install.yml#L74 is ineffective in disabling ipv6

Some OS's / kernels changing maybe?

(Just FWIW 8 days ago I was working with an individual whose IPv6 was successfully disabled by IIAB as he watched the IIAB install proceed... but that was a "Trisquelized" version of Ubuntu Server 24.04 LTS.)

jvonau · 2025-09-02T00:25:01Z

Hope he didn't lose internet access, lol Whatever, kernel_LL on ap0 is the wrong thing to do, should be a dummy under br0, period.

holta · 2025-09-03T03:45:08Z

If you are going to test Mint it should be on a hard metal WiFi based install, should be the same as RasPiOS.

This PR (#4078) on Linux Mint 22.2 result — iiab-diagnostics: https://paste.centos.org/view/8a60bcff (does that help confirm assumptions?)

And I can test Ubuntu Desktop 24.04 (or Ubuntu Desktop 25.10) next in coming hours if that helps.

jvonau · 2025-09-03T05:09:30Z

If you are going to test Mint it should be on a hard metal WiFi based install, should be the same as RasPiOS.

This PR (#4078) on Linux Mint 22.2 result — iiab-diagnostics: https://paste.centos.org/view/8a60bcff (does that help confirm assumptions?)

And I can test Ubuntu Desktop 24.04 (or Ubuntu Desktop 25.10) next in coming hours if that helps.

Well that was a wired install, looks fine, does it work ok? I did ask for a wifi install but this is a start...

holta · 2025-09-03T07:52:20Z

Well that was a wired install, looks fine, does it work ok?

Looks fine: connecting to hotspot works fine, prior to and after reboot.

I did ask for a wifi install but this is a start...

Ok. On a completely new install of Mint 22.2 Beta, with IIAB installed over WiFi, the final result (below) isn't perfect — but perhaps tolerable ?

WiFi client could not connect to IIAB's hotspot prior to reboot — perhaps this is completely normal for WiFi installs of IIAB? Output of systemctl status hostapd was:

○ hostapd.service - Hostapd IEEE 802.11 AP, IEEE 802.1X/WPA/WPA2/EAP/RADIUS Authenticator
     Loaded: loaded (/etc/systemd/system/hostapd.service; enabled; preset: enabled)
     Active: inactive (dead)

Sep 03 02:52:07 box systemd[1]: hostapd.service - Access point and authentication server for Wi-Fi and Ethernet was skipped because of an unmet condition check (ConditionFileNotEmpty=/etc/hostapd/hostapd.conf).

...until after IIAB reboot, whereupon IIAB's hotspot worked — i.e. student WiFi client devices can connect to IIAB's hotspot, and browse to its http://10.10.10.10
iiab-diagnostics: https://paste.centos.org/view/979e3e7c
Linux Mint was (and is) however unable to "ping mit.edu" — even after after running iiab-hotspot-off then clicking Linux Mint's graphical widget in the bottom-right of its desktop (it had remembered the home's SSID password from earlier, temporarily allowing "ping mit.edu" to work), and then running iiab-hotspot-on then rebooting. RECAP: Upon reboot, WiFi client devices can once again connect to IIAB, but IIAB still can't "ping mit.edu" (i.e. no dice with simultaneous upstream WiFi).
Linux Mint beeps and shows this message about every minute: (for the first 5 minutes after each boot)

Is it possible the Wi-Fi firmware for this ThinkPad T14 gen 1 just doesn't support AP+STA mode? Even despite "* managed" and "* AP" being listed in the output of iw list ?
Would it be useful for me to do another Linux Mint 22.2 install (on ThinkPad T14 bare metal) with wifi_up_down: False ? (Or any other test(s) that would be more useful?) Given that Linux Mint is extremely popular in quite a number of schools in many low-income countries.
- Or, I could test this PR (or others?) with Ubuntu Desktop 25.10 next, if that's more immediately useful.

jvonau · 2025-09-03T12:19:20Z

Well that was a wired install, looks fine, does it work ok?

Looks fine: connecting to hotspot works fine, prior to and after reboot.

Good.

I did ask for a wifi install but this is a start...

Ok. On a completely new install of Mint 22.2 Beta, with IIAB installed over WiFi, the final result (below) isn't perfect — but perhaps tolerable ?

1. WiFi client could not connect to IIAB's hotspot prior to reboot — perhaps this is completely normal for WiFi installs of IIAB?  Output of `systemctl status hostapd` was:

○ hostapd.service - Hostapd IEEE 802.11 AP, IEEE 802.1X/WPA/WPA2/EAP/RADIUS Authenticator
     Loaded: loaded (/etc/systemd/system/hostapd.service; enabled; preset: enabled)
     Active: inactive (dead)

Sep 03 02:52:07 box systemd[1]: hostapd.service - Access point and authentication server for Wi-Fi and Ethernet was skipped because of an unmet condition check (ConditionFileNotEmpty=/etc/hostapd/hostapd.conf).

2. ...until after IIAB reboot, whereupon IIAB's hotspot worked — i.e. student WiFi client devices can connect to IIAB's hotspot, and browse to its http://10.10.10.10

3. iiab-diagnostics: https://paste.centos.org/view/979e3e7c

By design #3886
no_net_restart = True

4. Linux Mint was (and is) however unable to "ping mit.edu" — even after after running `iiab-hotspot-off` then clicking Linux Mint's graphical widget in the bottom-right of its desktop (it had remembered the home's SSID password from earlier, temporarily allowing "ping mit.edu" to work), and then running `iiab-hotspot-on` then rebooting.  RECAP: Upon reboot, WiFi client devices can once again connect to IIAB, but IIAB still can't "ping mit.edu" (i.e. no dice with simultaneous upstream WiFi).

5. Linux Mint beeps and shows this message about every minute: (for the first 5 minutes after each boot)

Change in Netplan perhaps? Looks like the installer wrote the netplan file. Might want to try sudo netplan apply at this point and report back.

snip
6. Is it possible the Wi-Fi firmware for this ThinkPad T14 gen 1 just doesn't support AP+STA mode? Even despite "* managed" and "* AP" being listed in the output of iw list ?

Maybe

7. Would it be useful for me to do another Linux Mint 22.2 install (on ThinkPad T14 bare metal) with `wifi_up_down: False` ?  (Or any other test(s) that would be more useful?)  Given that Linux Mint is extremely popular in quite a number of schools in many low-income countries.

Can't hurt but iiab-hotspot-on and disabling netplan's wifi stanza would need to be preformed.

   * Or, I could test this PR (or others?) with Ubuntu Desktop 25.10 next, if that's more immediately useful.

Give it a go more info the better.

holta · 2025-09-03T16:49:48Z

Change in Netplan perhaps? Looks like the installer wrote the netplan file. Might want to try sudo netplan apply at this point and report back.

a. On the existing https://paste.centos.org/view/979e3e7c (IIAB installed over WiFi) I tried netplan apply
b. 'ping mit.edu' and 'ping 1.1.1.1' still do not work.
c. IIAB (ThinkPad T14 laptop) then rebooted.
d. 'ping mit.edu' and 'ping 1.1.1.1' still do not work. PS IIAB's hotspot continues to work fine.

➡️ Would a fresh install of Linux Mint somehow be useful, to run netplan apply earlier on? (I suspect not, but advise me of any recommendation...)

useful for me to do another Linux Mint 22.2 install (on ThinkPad T14 bare metal) with wifi_up_down: False ?

Can't hurt but iiab-hotspot-on and disabling netplan's wifi stanza would need to be [performed].

➡️ I can do that. Is a WiFi install or Ethernet install of IIAB most useful?

And Clarify If Possible: if a netplan file appears in /etc/netplan that looks roughly like 90-NM-a3100392-2aa6-47f5-b293-d3e227b4ce9a.yaml — which lines should be commented out or removed? (Almost all of the file's 25 lines [i.e. Lines 845-869 from iiab-diagnostics pasted in below] appear to be about connecting to the home's wifi — so should I maybe just move this entire 90-NM-...ce9a.yml file aside?)

test this PR (or others?) with Ubuntu Desktop 25.10

Give it a go more info the better.

➡️ Ok. I'll do that afterwards. And I can adjust towards testing anything you recommend.

jvonau · 2025-09-03T20:26:09Z

4. Linux Mint was (and is) however unable to "ping mit.edu" — even after after running `iiab-hotspot-off` then clicking Linux Mint's graphical widget in the bottom-right of its desktop (it had remembered the home's SSID password from earlier, temporarily allowing "ping mit.edu" to work), and then running `iiab-hotspot-on` then rebooting.  RECAP: Upon reboot, WiFi client devices can once again connect to IIAB, but IIAB still can't "ping mit.edu" (i.e. no dice with simultaneous upstream WiFi).

client_wifi_channel_found = 9
ssid=unittest
channel=6

Lets back up a second sudo iiab-hotspot-off then get ping to work as above and reboot then post sudo iw dev sudo iiab-test-wifi

This is new/different

phy#0 (self-managed)
country US: DFS-UNSET

jvonau · 2025-09-04T08:02:11Z

useful for me to do another Linux Mint 22.2 install (on ThinkPad T14 bare metal) with wifi_up_down: False ?

Can't hurt but iiab-hotspot-on and disabling netplan's wifi stanza would need to be [performed].

➡️ I can do that. Is a WiFi install or Ethernet install of IIAB most useful?

Given that wifi_up_down: False is not usually tested I'd say both, the desktop might try to scan for available WiFi's despite having a WiFi device not being managed by NetworkManager.

Sep 03 03:02:39 box NetworkManager[924]: [1756882959.1189] manager: (lo): new Loopback device (/org/freedesktop/NetworkManager/Devices/1)
Sep 03 03:02:39 box NetworkManager[924]: [1756882959.1210] manager: (br0): new Bridge device (/org/freedesktop/NetworkManager/Devices/2)
Sep 03 03:02:39 box NetworkManager[924]: [1756882959.1233] manager: (enp0s31f6): new Ethernet device (/org/freedesktop/NetworkManager/Devices/3)
ep 03 03:02:39 box NetworkManager[924]: [1756882959.3038] manager: (wlp0s20f3): new 802.11 Wi-Fi device (/org/freedesktop/NetworkManager/Devices/4)
Sep 03 03:02:39 box NetworkManager[924]: [1756882959.3076] manager: (ap0): new 802.11 Wi-Fi device (/org/freedesktop/NetworkManager/Devices/5)
Sep 03 03:02:39 box NetworkManager[924]: [1756882959.4331] manager: (p2p-dev-wlp0s20f3): new 802.11 Wi-Fi P2P device (/org/freedesktop/NetworkManager/Devices/6)
Sep 03 03:02:39 box NetworkManager[924]: [1756882959.5535] manager: (tailscale0): new Tun device (/org/freedesktop/NetworkManager/Devices/7)
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.4863] device (wlp0s20f3): Activation: (wifi) connection 'Auto magoun' has security, and secrets exist. No new secrets needed.
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.4864] Config: added 'ssid' value 'magoun'
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.4864] Config: added 'scan_ssid' value '1'
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.4864] Config: added 'bgscan' value 'simple:30:-70:86400'
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.4864] Config: added 'key_mgmt' value 'WPA-PSK WPA-PSK-SHA256 FT-PSK SAE FT-SAE'
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.4864] Config: added 'auth_alg' value 'OPEN'
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.4864] Config: added 'psk' value ''
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.6025] device (wlp0s20f3): supplicant interface state: disconnected -> scanning
Sep 03 03:02:49 box NetworkManager[924]: [1756882969.6026] device (p2p-dev-wlp0s20f3): supplicant management interface state: disconnected -> scanning
Sep 03 03:02:59 box NetworkManager[924]: [1756882979.6172] device (wlp0s20f3): supplicant interface state: scanning -> disconnected

Note for below

Sep 03 03:02:38 box systemd[1]: Started wpa_supplicant.service - WPA supplicant.
Sep 03 03:02:39 box wpa_supplicant[915]: nl80211: kernel reports: Match already configured
Sep 03 03:02:39 box wpa_supplicant[915]: nl80211: Could not configure driver mode
Sep 03 03:02:39 box wpa_supplicant[915]: nl80211: deinit ifname=ap0 disabled_11b_rates=0
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWPMKSA]: Invalid argument
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWMODE]: Device or resource busy
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODEEXT]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODE]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODEEXT]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODE]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODEEXT]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODE]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODEEXT]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODE]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODEEXT]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODE]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODEEXT]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWENCODE]: Operation not supported
Sep 03 03:02:39 box wpa_supplicant[915]: ioctl[SIOCSIWPMKSA]: Invalid argument
Sep 03 03:02:49 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:02:49 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1
Sep 03 03:02:49 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:02:49 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1
Sep 03 03:02:49 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:02:49 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1
Sep 03 03:02:49 box wpa_supplicant[915]: wlp0s20f3: WPS-CANCEL
Sep 03 03:02:49 box wpa_supplicant[915]: wlp0s20f3: SME: Trying to authenticate with 8c:90:2d:21:34:1e (SSID='magoun' freq=2452 MHz)
Sep 03 03:02:49 box wpa_supplicant[915]: wlp0s20f3: SME: Authentication request to the driver failed

This is where NM gave up (scanning -> disconnected) but the desktop continues scanning

Sep 03 03:02:54 box iiab-test-wifi[1051]: Couldn't find an UPSTREAM SSID in files like wpa_supplicant.conf -- so exiting.
Sep 03 03:02:59 box wpa_supplicant[915]: ioctl[SIOCGIWSCAN]: Resource temporarily unavailable
Sep 03 03:02:59 box wpa_supplicant[915]: wlp0s20f3: SME: Trying to authenticate with 8c:90:2d:21:34:1e (SSID='magoun' freq=2452 MHz)
Sep 03 03:02:59 box wpa_supplicant[915]: wlp0s20f3: SME: Authentication request to the driver failed
Sep 03 03:02:59 box wpa_supplicant[915]: BSSID 8c:90:2d:21:34:1e ignore list count incremented to 2, ignoring for 10 seconds
Sep 03 03:02:59 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:02:59 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1
Sep 03 03:03:09 box wpa_supplicant[915]: wlp0s20f3: CTRL-EVENT-REGDOM-CHANGE init=DRIVER type=COUNTRY alpha2=US
Sep 03 03:03:09 box wpa_supplicant[915]: wlp0s20f3: SME: Trying to authenticate with 8c:90:2d:21:34:1e (SSID='magoun' freq=2452 MHz)
Sep 03 03:03:09 box wpa_supplicant[915]: wlp0s20f3: SME: Authentication request to the driver failed
Sep 03 03:03:09 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:03:09 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1
Sep 03 03:03:13 box wpa_supplicant[915]: wlp0s20f3: SME: Trying to authenticate with 8c:90:2d:21:34:1e (SSID='magoun' freq=2452 MHz)
Sep 03 03:03:13 box wpa_supplicant[915]: wlp0s20f3: SME: Authentication request to the driver failed
Sep 03 03:03:13 box wpa_supplicant[915]: wlp0s20f3: CTRL-EVENT-SSID-TEMP-DISABLED id=0 ssid="magoun" auth_failures=1 duration=10 reason=CONN_FAILED

Sep 03 03:04:07 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:04:07 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1
Sep 03 03:04:13 box wpa_supplicant[915]: wlp0s20f3: WPS-CANCEL
Sep 03 03:04:13 box wpa_supplicant[915]: wlp0s20f3: SME: Trying to authenticate with 8c:90:2d:21:34:1e (SSID='magoun' freq=2452 MHz)
Sep 03 03:04:13 box wpa_supplicant[915]: wlp0s20f3: SME: Authentication request to the driver failed
Sep 03 03:04:13 box wpa_supplicant[915]: wlp0s20f3: CTRL-EVENT-SSID-TEMP-DISABLED id=0 ssid="magoun" auth_failures=1 duration=10 reason=CONN_FAILED
Sep 03 03:04:26 box wpa_supplicant[915]: wlp0s20f3: CTRL-EVENT-SSID-REENABLED id=0 ssid="magoun"
Sep 03 03:04:26 box wpa_supplicant[915]: wlp0s20f3: SME: Trying to authenticate with 8c:90:2d:21:34:1e (SSID='magoun' freq=2452 MHz)
Sep 03 03:04:26 box wpa_supplicant[915]: wlp0s20f3: SME: Authentication request to the driver failed
Sep 03 03:04:26 box wpa_supplicant[915]: BSSID 8c:90:2d:21:34:1e ignore list count incremented to 2, ignoring for 10 seconds
Sep 03 03:04:26 box wpa_supplicant[915]: wlp0s20f3: CTRL-EVENT-SSID-TEMP-DISABLED id=0 ssid="magoun" auth_failures=2 duration=20 reason=CONN_FAILED
Sep 03 03:04:38 box wpa_supplicant[915]: wlp0s20f3: Reject scan trigger since one is already pending
Sep 03 03:04:50 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:04:50 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1
Sep 03 03:05:54 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:05:54 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1
Sep 03 03:07:29 box wpa_supplicant[915]: ioctl[SIOCSIWSCAN]: Operation not supported
Sep 03 03:07:29 box wpa_supplicant[915]: ap0: CTRL-EVENT-SCAN-FAILED ret=-1

And Clarify If Possible: if a netplan file appears in /etc/netplan that looks roughly like 90-NM-a3100392-2aa6-47f5-b293-d3e227b4ce9a.yaml — which lines should be commented out or removed? (Almost all of the file's 25 lines [i.e. Lines 845-869 from iiab-diagnostics pasted in below] appear to be about connecting to the home's wifi — so should I maybe just move this entire 90-NM-...ce9a.yml file aside?)

I'd just move it for now, with an eye to see the result #4039

…i hardware. This reverts commit 335ba55.

holta · 2025-09-04T16:44:35Z

Lets back up a second sudo iiab-hotspot-off then get ping to work as above and reboot then post sudo iw dev sudo iiab-test-wifi

Done: results below... again on that original WiFi-installed IIAB on Mint 22.2:

root@box:~# iw dev
phy#0
	Unnamed/non-netdev interface
		wdev 0x2
		addr bc:54:2f:a6:60:fe
		type P2P-device
	Interface wlp0s20f3
		ifindex 3
		wdev 0x1
		addr bc:54:2f:a6:60:fd
		ssid magoun
		type managed
		channel 9 (2452 MHz), width: 20 MHz, center1: 2452 MHz
		txpower 22.00 dBm
		multicast TXQ:
			qsz-byt	qsz-pkt	flows	drops	marks	overlmt	hashcol	tx-bytes	tx-packets
			0	0	0	0	0	0	0	0		0

root@box:~# iiab-test-wifi
iiab-test-wifi called
running pid 2627
NetworkManager
ssid is magoun
frequency is 2452 for magoun
Using 2452 for magoun
channel is 9 for magoun
Hostapd set for 6
Editing Hostapd for channel 9
Restarted hostapd

root@box:~#

(Just FYI the iiab-test-wifi command took ~27 seconds to run... which appears completely normal... hope this helps.)

holta · 2025-09-09T02:56:58Z

What is the result of cat /var/run/NetworkManager/no-stub-resolv.conf

FWIW there is no /var/run/NetworkManager

(And there is no no-stub-resolv.conf — according to the results of the 'updatedb' then 'locate no-stub-resolv.conf' commands.)

jvonau · 2025-09-09T10:37:23Z

What is the result of cat /var/run/NetworkManager/no-stub-resolv.conf

FWIW there is no /var/run/NetworkManager

(And there is no no-stub-resolv.conf — according to the results of the 'updatedb' then 'locate no-stub-resolv.conf' commands.)

Think the /run directory is excluded from 'updatedb'. Right, NM isn't running. How about cat /run/systemd/resolve/resolv.conf, there should be entries under /run/systemd/resolve/netif/ if each upstream connection but you need sudo to view them.

holta · 2025-09-09T14:36:09Z

root@box:~# cat /run/systemd/resolve/resolv.conf
# This is /run/systemd/resolve/resolv.conf managed by man:systemd-resolved(8).
# Do not edit.
#
# This file might be symlinked as /etc/resolv.conf. If you're looking at
# /etc/resolv.conf and seeing this text, you have followed the symlink.
#
# This is a dynamic resolv.conf file for connecting local clients directly to
# all known uplink DNS servers. This file lists all configured search domains.
#
# Third party programs should typically not access this file directly, but only
# through the symlink at /etc/resolv.conf. To manage man:resolv.conf(5) in a
# different way, replace this symlink by a static file or a different symlink.
#
# See man:systemd-resolved.service(8) for details about the supported modes of
# operation for /etc/resolv.conf.
nameserver 192.168.0.1
search lan

root@box:/run/systemd/resolve/netif# ls -l
total 4
-rw-r--r-- 1 systemd-resolve systemd-resolve 97 Sep  8 23:03 6

root@box:/run/systemd/resolve/netif# cat 6
# This is private data. Do not parse.
LLMNR=no
MDNS=no
DNSSEC=no
DNSOVERTLS=no
DEFAULT_ROUTE=yes

holta · 2025-09-09T20:27:29Z

Just FYI I'd like to wipe the ThinkPad T14 very soon — e.g. to reconfirm this PR with testing on fresh installs of the most common OS's (Linux distros). And a couple RPi tests would be wise too no doubt (:

holta · 2025-09-09T22:45:18Z

Probably not worth worrying about — but just FYI systemctl status dnsmasq after apt updates were applied to Ubuntu Server 25.10:

And then it's more "back to normal" after reboot:

holta · 2025-09-10T19:29:36Z

Fresh install test of this PR (#4078) on a new WiFi-install on the very latest pre-release of Ubuntu Desktop 25.10 -- looks good here with "Dual WiFi" (AP+STA mode):

iiab-diagnostics: https://paste.centos.org/view/fb6b0669

jvonau · 2025-09-10T19:49:19Z

iiab-diagnostics after reboot correct? Looks pretty good from here.

holta · 2025-09-10T20:02:32Z

iiab-diagnostics after reboot correct?

Yes.

PS I'll try to test another OS or two later this evening. [make that Thur evening, tonight]

holta · 2025-09-12T03:50:16Z

Quick tests of this PR on Linux Mint 22.2 all look good, FYI.

holta · 2025-09-13T00:48:22Z

Quick test of this PR on RasPiOS 12 (Bookworm) with desktop... on RPi 500... looks good:

"PR4078--WiFi-installed--AP+STA-mode--RasPiOS-12-Desktop"
https://paste.centos.org/view/e13c1db7

holta · 2025-09-13T02:44:49Z

Quick test of this PR on RasPiOS 13 (Trixie) with desktop... on RPi 500... looks good:

"PR4078--WiFi-installed--AP+STA-mode--RasPiOS-13-Desktop"
https://paste.centos.org/view/cd2310fa

holta · 2025-09-13T04:39:00Z

@jvonau

Any idea why "Dual WiFi" (AP+STA Mode) is not working on Debian 13 "Trixie" on x86_64 below? This might not be the most important OS/distro among IIAB implementers, but FWIW:

The output of ip a unexpectedly indicates that WiFi interface wlp0s20f3 is DOWN instead of UP.
The "Dual WiFi" (AP+STA Mode) capability works on the very same laptop, with other OS's like Ubuntu & Mint.
Doesn't seem like a WiFi firmware issue, but just in case...

root@box:~# dmesg | grep iwlwifi
[    4.736894] iwlwifi 0000:00:14.3: enabling device (0000 -> 0002)
[    4.738966] iwlwifi 0000:00:14.3: Detected crf-id 0x3617, cnv-id 0x20000302 wfpm id 0x80000000
[    4.738987] iwlwifi 0000:00:14.3: PCI dev 02f0/0070, rev=0x351, rfid=0x10a100
[    4.738990] iwlwifi 0000:00:14.3: Detected Intel(R) Wi-Fi 6 AX201 160MHz
[    4.749933] iwlwifi 0000:00:14.3: TLV_FW_FSEQ_VERSION: FSEQ Version: 89.3.35.37
[    4.750225] iwlwifi 0000:00:14.3: firmware: failed to load iwl-debug-yoyo.bin (-2)
[    4.750233] iwlwifi 0000:00:14.3: firmware: failed to load iwl-debug-yoyo.bin (-2)
[    4.750239] iwlwifi 0000:00:14.3: firmware: failed to load iwl-debug-yoyo.bin (-2)
[    4.750241] iwlwifi 0000:00:14.3: loaded firmware version 77.2753b721.0 QuZ-a0-hr-b0-77.ucode op_mode iwlmvm
[    5.316447] iwlwifi 0000:00:14.3: Detected RF HR B3, rfid=0x10a100
[    5.400977] iwlwifi 0000:00:14.3: base HW address: bc:54:2f:a6:60:fd
[    5.452525] iwlwifi 0000:00:14.3 wlp0s20f3: renamed from wlan0
[    6.130611] iwlwifi 0000:00:14.3: Registered PHC clock: iwlwifi-PTP, with index: 1
[    6.658138] iwlwifi 0000:00:14.3 ap0: entered allmulticast mode
[    6.658203] iwlwifi 0000:00:14.3 ap0: entered promiscuous mode

iiab-diagnostics after several reboots: (Ethernet cable inserted right afterwards, to create this pastebin...)
https://paste.centos.org/view/97427bc2

jvonau · 2025-09-13T10:44:02Z

Given I don't see the usual channel 9 embedded within hostapd.conf when connected to your AP while installing, I suggest the workaround you used before

holta · 2025-09-13T16:21:23Z

I suggest the workaround you used #4078 (comment)

Thanks: I finally got it to work, great. Earlier I'd overlooked a step. So I should write down the steps for everyone -- presumably in the order below? (Here is a DRAFT 'How do i turn on "Dual WiFi" a.k.a. "AP+STA Mode" ?' for https://FAQ.IIAB.IO ...)

WARNING: Not all WiFi firmware supports "Dual WiFi" a.k.a. "AP+STA Mode", and WiFi performance can sometimes be severely degraded even when it is working. But if you want to give it a shot, here are 9 suggested step:

On your IIAB, run iw list and look at the output. If see * AP and * managed (each on their own line), then there's a good shot it will work. (Similar to: "Can I create a Wi-Fi hotspot using an old laptop?")

Verify that your IIAB's /etc/iiab/local_vars.yml contains wifi_up_down: true

Run: sudo iiab-hotspot-off

Confirm IIAB is connected to your home's WiFi Internet (e.g. use your IIAB OS's graphical desktop widget, or equivalent). Test it by running: ping mit.edu

OPTIONAL: Run iw dev just to be aware of your upstream WiFi channel.

Run: sudo iiab-test-wifi (to align channels for "Dual WiFi" a.k.a. AP+STA Mode; this takes about 30 seconds, but is much faster than running iiab-network !)

Run: sudo iiab-hotspot-on

Run: sudo reboot

Try to confirm that both are working:

Test your IIAB's upstream WiFi by running: ping mit.edu

Test your IIAB's downstream WiFi by running: systemctl status hostapd -- and with actual, functional tests using WiFi-capable "student devices".

ASIDE: Ubuntu 25.10 ISOs are a real mess in recent days (recent dailies completely fail to install) so I'm glad I got that pre-release testing done last week -- and I'll look at that OS again later in September &/or October.

holta · 2025-09-13T17:16:16Z

Here is a DRAFT

Cleaned up & posted:

"Can IIAB also connect to a school's Wi-Fi internet?"
https://wiki.iiab.io/go/FAQ#Can_IIAB_also_connect_to_a_school's_Wi-Fi_internet%3F

(If something's explained very poorly, please LMK!)

holta · 2025-09-13T21:25:51Z

Let's give this PR a shot, as it's now been tested in many different scenarios.

I'd like to merge it (or very similar) overnight if there are no objections?

iiab-network

holta · 2025-09-14T17:09:32Z

Either way, I'll merge this PR in coming hours:

➡️ Thanks to @jvonau for making it happen! ⬅️

holta · 2025-09-14T19:11:12Z

Good suggestions above thanks to @jvonau for linting + cleaner output, that are now part of:

PR (1) Lint command -v conditionals (2) Clean iiab-network output w/ echo lines (3) Tighten comments [in support of PR #4078] #4083

I'll let those 4 CI [GitHub Actions] workflows on PR #4083 run to completion, then merge it all.

(1) Lint `command -v` conditionals (2) Clean `iiab-network` output w/ echo lines (3) Tighten comments [in support of PR #4078]

jvonau added 4 commits August 31, 2025 15:51

channel follows NetworkManager

335ba55

country follows NetworkManager

b99e48d

nmcli radio wifi on

fed59ff

networkctl s/b a different PR

eadfa47

jvonau changed the title ~~Wifi code~~ DRAFT: Wifi code Aug 31, 2025

holta added the enhancement label Aug 31, 2025

holta added this to the 8.3 milestone Aug 31, 2025

networkctl to log file

e3c4631

jvonau changed the title ~~DRAFT: Wifi code~~ Wifi code Sep 1, 2025

disable kernel_ll ipv6 address on ap0

ecbaa88

Revert "channel follows NetworkManager" still needed on some non-RasP…

c6ac43d

…i hardware. This reverts commit 335ba55.

holta reviewed Sep 14, 2025

View reviewed changes

iiab-network Show resolved Hide resolved

holta reviewed Sep 14, 2025

View reviewed changes

iiab-network Show resolved Hide resolved

holta mentioned this pull request Sep 14, 2025

(1) Lint command -v conditionals (2) Clean iiab-network output w/ echo lines (3) Tighten comments [in support of PR #4078] #4083

Merged

holta added a commit that referenced this pull request Sep 14, 2025

Merge pull request #4083 from holta/wifi_code

e6b8f72

(1) Lint `command -v` conditionals (2) Clean `iiab-network` output w/ echo lines (3) Tighten comments [in support of PR #4078]

holta merged commit ff9da99 into iiab:master Sep 14, 2025
4 checks passed

jvonau mentioned this pull request Sep 24, 2025

no dhcpd on new install #4090

Closed

This was referenced Sep 25, 2025

Unable to Host Hotspot on IIAB [when implementer switches RasPiOS desktop Wayland / labwc to LXDE] #4077

Open

Networking cleanup, building on #4064 & #4067 #4068

Closed

This was referenced Oct 30, 2025

all bridge members should not have IPv6 enabled [Bluetooth fix] #4126

Merged

BUG: bluethooth networking not working when hostapd is not running. #4128

Open

jvonau mentioned this pull request Nov 8, 2025

"Unable to restart service dnsmasq" during "20 min" IIAB "MEDIUM" on Ubuntu 24.04 on x86-64 / test-install [race condition?] #4099

Open

jvonau mentioned this pull request Nov 16, 2025

Use systemd-resolve with NetworkManager, use netplan for bridging, Fix DEPRECATION WARNINGs #4093

Open

Wifi code #4078

Wifi code #4078

Uh oh!

Conversation

jvonau commented Aug 31, 2025

Fixes bug:

Description of changes proposed in this pull request:

Smoke-tested on which OS or OS's:

Mention a team member @username e.g. to help with code review:

Uh oh!

jvonau commented Aug 31, 2025

Uh oh!

jvonau commented Aug 31, 2025

Uh oh!

jvonau commented Sep 1, 2025

Uh oh!

holta commented Sep 1, 2025

Uh oh!

jvonau commented Sep 1, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jvonau commented Sep 1, 2025

Uh oh!

holta commented Sep 1, 2025

Uh oh!

jvonau commented Sep 1, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jvonau commented Sep 1, 2025

Uh oh!

holta commented Sep 1, 2025

Uh oh!

jvonau commented Sep 2, 2025

Uh oh!

holta commented Sep 3, 2025

Uh oh!

jvonau commented Sep 3, 2025

Uh oh!

holta commented Sep 3, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jvonau commented Sep 3, 2025

Uh oh!

holta commented Sep 3, 2025

Uh oh!

jvonau commented Sep 3, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jvonau commented Sep 4, 2025

Uh oh!

holta commented Sep 4, 2025

Uh oh!

holta commented Sep 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jvonau commented Sep 9, 2025

Uh oh!

holta commented Sep 9, 2025

Uh oh!

holta commented Sep 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

holta commented Sep 9, 2025

Uh oh!

holta commented Sep 10, 2025

Uh oh!

jvonau commented Sep 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

holta commented Sep 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

holta commented Sep 12, 2025

Uh oh!

holta commented Sep 13, 2025

Uh oh!

holta commented Sep 13, 2025

Uh oh!

holta commented Sep 13, 2025

Uh oh!

jvonau commented Sep 1, 2025 •

edited

Loading

jvonau commented Sep 1, 2025 •

edited

Loading

holta commented Sep 3, 2025 •

edited

Loading

jvonau commented Sep 3, 2025 •

edited

Loading

holta commented Sep 9, 2025 •

edited

Loading

holta commented Sep 9, 2025 •

edited

Loading

jvonau commented Sep 10, 2025 •

edited

Loading

holta commented Sep 10, 2025 •

edited

Loading

holta commented Sep 14, 2025 •

edited

Loading