From: Martin Moosbauer <martinmoosbauer@vivaldi.net>
To: linux-wireless@vger.kernel.org
Cc: linux-mediatek@lists.infradead.org, stable@vger.kernel.org
Subject: • [REGRESSION 6.18.54 -> 6.18.63] cfg80211 / mt7925e: Invalid "Country: EU" beacon forces regdom to 98 and breaks AP/5GHz
Date: Sat, 14 Nov 2026 16:15:22 +0000 [thread overview]
Message-ID: <8db36875c56cd9703dee42c7a00b81c9@vivaldi.net> (raw)
Hi all,
I am reporting a regulatory regression on a MediaTek MT7925 (mt7925e)
wireless card caused by neighboring ISP routers (Vodafone Germany /
CommScope-Arris RDK-B) broadcasting an invalid 802.11d Country IE set to
"EU".
This worked cleanly on Arch Linux kernel `6.18.54-2-lts`, where the
system stayed on `DE` even after scanning the neighboring `EU` beacons.
After updating to `6.18.63-2-lts` (and also reproducible on
`7.2.8.zen1-2` and `7.3.4-zen2-1-zen`), the regulatory domain
immediately falls back to `98` (`DFS-UNSET`), enforcing `PASSIVE-SCAN`
across all 5 GHz channels and breaking AP/hotspot startup as well as
manual scans.
### Problem Description
1. Neighboring Vodafone Station routers (OUI 4c:12:65, CommScope/Arris)
broadcast `Country: EU` in their beacons (both on private SSIDs and
public "Vodafone Hotspot" / "Vodafone Homespot" virtual BSSs). Note:
Neighboring Vantiva/Technicolor Vodafone Stations (OUI 98:9d:5d) omit
the Country IE and do not trigger this issue.
2. On kernel `6.18.54-2-lts`, `cfg80211` / `mt7925e` stayed on `DE` when
seeing these beacons alongside valid `DE` routers. Both AP/hotspot and 5
GHz STA modes worked without issues, and manual scans (`iw dev wlp8s0
scan`) succeeded.
3. On kernel `6.18.63-2-lts` (as well as `7.2.8` and `7.3.4`), as soon
as background scans pick up the `Country: EU` beacon, the active
regulatory domain is overwritten with `98` (`DFS-UNSET`). All 5 GHz
channels are restricted to `PASSIVE-SCAN`, preventing AP startup and
causing manual scans (`iw dev ... scan`) to fail with `command failed:
Device or resource busy (-16)`. Manually running `iw reg set DE` is
immediately reverted on the next background scan.
### Expected Behavior
An invalid or non-ISO country code like "EU" received from an
unassociated/neighboring AP beacon should be ignored/discarded by
`cfg80211` (matching the `6.18.54-2-lts` behavior) rather than
overriding a valid active regulatory domain (`DE`) with `98`.
### System Information
- Distribution: Arch Linux
- Last known working kernel: `6.18.54-2-lts`
- Affected kernels: `6.18.63-2-lts`, `7.2.8.zen1-2`, `7.3.4-zen2-1-zen`
- Hardware / Driver: MediaTek MT7925 (`mt7925e`, ASIC revision
`79250000`)
- Firmware version: `HW/SW Version: 0x8a108a10, Build Time:
20260813113015a` / `WM Firmware Version: ____000000, Build Time:
20260813113118` (`linux-firmware 20261019-1`)
- `wireless-regdb` version: `2026.09.03-1`
- `iw` version: `6.17-1`
### Regulatory Output (`iw reg get` after boot on 6.18.63-2-lts)
global
country 98: DFS-UNSET
(2402 - 2472 @ 40), (N/A, 20), (N/A)
(2457 - 2482 @ 25), (N/A, 20), (N/A), PASSIVE-SCAN
(2474 - 2483 @ 9), (N/A, 20), (N/A), NO-OFDM, PASSIVE-SCAN
(5170 - 5250 @ 80), (N/A, 20), (N/A), NO-OUTDOOR, AUTO-BW, PASSIVE-SCAN
(5250 - 5330 @ 80), (N/A, 20), (0 ms), NO-OUTDOOR, DFS, AUTO-BW,
PASSIVE-SCAN
(5490 - 5725 @ 160), (N/A, 20), (0 ms), DFS, PASSIVE-SCAN
(5725 - 5730 @ 5), (N/A, 13), (0 ms), DFS, PASSIVE-SCAN
(5735 - 5835 @ 80), (N/A, 13), (N/A), PASSIVE-SCAN
(57240 - 63720 @ 2160), (N/A, 0), (N/A)
### Relevant Scan Dump Output
BSS 4c:12:65:ba:9b:8c(on wlp8s0_2)
SSID: *SSID Deleted for Privacy reasons*
Country: EU Environment: Indoor/Outdoor
* OBSS non-GF present: 0
* BSS Transition
BSS 4e:12:65:ba:9b:9c(on wlp8s0_2)
SSID: Vodafone Hotspot
Country: EU Environment: Indoor/Outdoor
* OBSS non-GF present: 0
BSS 4e:12:65:ba:9b:bc(on wlp8s0_2)
SSID: Vodafone Homespot
Country: EU Environment: Indoor/Outdoor
* OBSS non-GF present: 0
Any guidance or patch to restore the `6.18.54` behavior and prevent
unassociated invalid "EU" beacons from poisoning the active regulatory
domain would be greatly appreciated.
Best regards,
Martin Moosbauer
reply other threads:[~2026-11-14 16:15 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=8db36875c56cd9703dee42c7a00b81c9@vivaldi.net \
--to=martinmoosbauer@vivaldi.net \
--cc=linux-mediatek@lists.infradead.org \
--cc=linux-wireless@vger.kernel.org \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox