OpenWrt 25.12.0-rc4 - Release Candidate

jopenz · February 3, 2026, 3:57pm

Thanks for the correction @dave14305 and @egc
You're right - the UCI options are endpoint_host and endpoint_port as separate values, not a combined endpoint. I was conflating the wg tool syntax (which uses endpoint ip:port) with the UCI layer.
Confirmed working:

uci set network.@wireguard_wghelsinki[0].endpoint_host='77.42.34.55'
uci set network.@wireguard_wghelsinki[0].endpoint_port='51820'
uci commit network
ifdown wghelsinki && ifup wghelsinki

The clue was there in uci show network all along - I just didn't read it carefully enough. Apologies for the noise.

Pico · February 3, 2026, 6:17pm

I had installed an unchanged headless x86/64 from 24.10.4 to 25.12rc4 via a new plain disk installation and then imported a 24.10 config backup. Both old and new image had used the same default PCI NIC kmods.

Works fine, but 25.10 had unexpectedly remapped eth to port mapping order.

old ethX→port mapping was 2103 on a 4x NIC + an onboard NIC was eth4.
new ethX port mapping now 3214 on a 4x NIC + an onboard NIC is now eth0.

I am hoping, it will stay stable, as I am worried a bit that this might toggle randomly on reboots occasionally.

The unplannable kind of random ethX assignment on x86 and single board computer sometimes is a bit tedious. Similar issue with USB Wifi and mobile dongles on Raspberrys running OpenWRT in any version. Wish we could find a better way, eg. linking config items to static MACs of NICs/Wifi/mobile adapters.

In case anyone runs into the same issue, to speedup debugging, plug in a display and then consecutively plug in a single network device into each port, and make a note of the default debug output on screen, as it shows the ethX number for the port you had just plugged your device in.

MarkyMark · February 3, 2026, 6:56pm

Just updated my EDUP EP-RT2960S from 24.10.5 to 25.12.0-rc4 using attended sysupgrade. The update went very smooth, without any problems.
What I have noticed though is my lan1 port is showing as lan3 and vice versa

plus my led is always blue.
I know these were previously known problems, but was sort of hoping they had been fixed by now. Not an major problem but it would be good if they were put to bed for anyone in the future using this type router.
Everything else was the same, my wifi and guest wifi set ups all transfered across, no issue.

faruktezcan · February 3, 2026, 7:51pm

owut gives error while upgrading : $(date +%H:%M:%S)-NanoPi:/etc/apk/repositories.d$ owut download -v -v
owut - OpenWrt Upgrade Tool 2026.01.13~2526d84b-r1 (/usr/bin/owut)
Downloaded https://sysupgrade.openwrt.org/json/v1/overview.json to /tmp/owut-overview.json (18217B at 0.069 Mbps)
Downloaded https://downloads.openwrt.org/releases/25.12.0-rc4/targets/rockchip/armv8/profiles.json to /tmp/owut-platform.json (146B at 0.002 Mbps)
ERROR: Response status 404 while downloading

https://downloads.openwrt.org/releases/25.12.0-rc4/targets/rockchip/armv8/profiles.json

ERROR: Unsupported target 'rockchip/armv8'

anon63541380 · February 3, 2026, 8:06pm

Did you read the release announcement? rockchip not bult for this rc.

DallasPaley · February 3, 2026, 11:49pm

I figured I would post this here. Not sure if there is anything that needs to be reported to developers.

Both Router and Access Point are GL.iNet Flint 2 (GL-MT6000)

Update wired access point 25.12.0-rc4

Router running: 24.10.5

System Log

Severity: Error

At Boot

[Feb 2, 2026, 11:21:27 AM EST] kern.err: [ 1.759271] u-boot-env-layout 11230000.mmc:card@0:block:partitions:block-partition-env:nvmem-layout: Invalid calculated CRC32: 0xc09e9b4c (expected: 0x00000000)
[Feb 2, 2026, 11:21:27 AM EST] kern.err: [ 1.773638] u-boot-env-layout 11230000.mmc:card@0:block:partitions:block-partition-env:nvmem-layout: probe with driver u-boot-env-layout failed with error -22
[Feb 2, 2026, 11:21:31 AM EST] daemon.err: procd: Got unexpected signal 1

After Boot

[Feb 2, 2026, 11:26:58 AM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 2, 2026, 12:22:51 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 2, 2026, 12:22:51 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 2, 2026, 12:22:56 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 2, 2026, 1:43:20 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 2, 2026, 2:30:04 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 2, 2026, 6:28:10 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 2, 2026, 6:34:50 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 2, 2026, 7:11:50 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 3, 2026, 12:05:15 AM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 3, 2026, 5:34:50 AM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed
[Feb 3, 2026, 3:18:19 PM EST] daemon.err: hostapd: phy1-ap0: nl80211: kernel reports: key addition failed

Severity: Warning

At Boot

[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 0.003638] cacheinfo: Unable to detect cache hierarchy for CPU 0
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.134069] GPT:Primary header thinks Alt. header is not at the end of the disk.
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.141567] GPT:1 != 15269887
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.144526] GPT:Alternate GPT header not at the end of the disk.
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.150527] GPT:1 != 15269887
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.153481] GPT: Use GNU Parted to correct GPT errors.
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.832253] gpio_button_hotplug: loading out-of-tree module taints kernel.
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 6.486923] overlayfs: null uuid detected in lower fs '/', falling back to xino=off,index=off,nfs_export=off.
[Feb 2, 2026, 11:21:27 AM EST] daemon.warn: dnsmasq[1]: no servers found in /tmp/resolv.conf.d/resolv.conf.auto, will retry

Keeps Repeating

[Feb 2, 2026, 11:22:43 AM EST] daemon.warn: odhcpd[1917]: No default route present, setting ra_lifetime to 0!
[Feb 2, 2026, 11:22:59 AM EST] daemon.warn: odhcpd[1917]: No default route present, setting ra_lifetime to 0!
[Feb 2, 2026, 11:26:59 AM EST] daemon.warn: odhcpd[1917]: No default route present, setting ra_lifetime to 0!
[Feb 2, 2026, 11:27:54 AM EST] daemon.warn: odhcpd[1917]: No default route present, setting ra_lifetime to 0!
[Feb 2, 2026, 11:27:55 AM EST] daemon.warn: odhcpd[1917]: No default route present, setting ra_lifetime to 0!

systemcrash · February 4, 2026, 5:01am

Is this consistent? Looks like a failure occurred with the flash. Have you been flashing it… alot?

fanat · February 4, 2026, 5:06am

I can install wireguard and tune?

d3vdays · February 4, 2026, 5:27am

Did an upgrade on my LN1301 (Linksys) AP and main router and worked really seamlessly. There as a minor issue where on my AP (I have VLANs), the IOT vlan interface "Device" was incorrectly setup.

jopenz · February 4, 2026, 8:19am

Your local circumstance might be a constraint, but your router isn't a barrier.

DallasPaley · February 4, 2026, 1:48pm

No, I only flashed it the one time using Attended Sysupgrade.

I just rebooted:

Error:

[Feb 2, 2026, 11:21:27 AM EST] kern.err: [ 1.823187] u-boot-env-layout 11230000.mmc:card@0:block:partitions:block-partition-env:nvmem-layout: Invalid calculated CRC32: 0xc09e9b4c (expected: 0x00000000)
[Feb 2, 2026, 11:21:27 AM EST] kern.err: [ 1.842981] u-boot-env-layout 11230000.mmc:card@0:block:partitions:block-partition-env:nvmem-layout: probe with driver u-boot-env-layout failed with error -22
[Feb 2, 2026, 11:21:31 AM EST] daemon.err: procd: Got unexpected signal 1

Warning:

[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 0.003639] cacheinfo: Unable to detect cache hierarchy for CPU 0
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.134257] GPT:Primary header thinks Alt. header is not at the end of the disk.
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.141720] GPT:1 != 15269887
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.144676] GPT:Alternate GPT header not at the end of the disk.
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.150672] GPT:1 != 15269887
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.153624] GPT: Use GNU Parted to correct GPT errors.
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 1.821759] gpio_button_hotplug: loading out-of-tree module taints kernel.
[Feb 2, 2026, 11:21:27 AM EST] kern.warn: [ 6.517210] overlayfs: null uuid detected in lower fs '/', falling back to xino=off,index=off,nfs_export=off.
[Feb 2, 2026, 11:21:27 AM EST] daemon.warn: dnsmasq[1]: no servers found in /tmp/resolv.conf.d/resolv.conf.auto, will retry
[Feb 2, 2026, 11:21:44 AM EST] daemon.warn: odhcpd[1913]: No default route present, setting ra_lifetime to 0!
[Feb 2, 2026, 11:21:44 AM EST] daemon.warn: odhcpd[1913]: rfc9096: br-lan.10: piofile updated
[Feb 2, 2026, 11:21:48 AM EST] daemon.warn: odhcpd[1913]: No default route present, setting ra_lifetime to 0!
[Feb 4, 2026, 8:46:38 AM EST] authpriv.warn: dispatcher.uc: luci: failed login on / for root from 192.168.10.12
[Feb 4, 2026, 8:46:41 AM EST] daemon.warn: odhcpd[1913]: No default route present, setting ra_lifetime to 0!
[Feb 4, 2026, 8:47:34 AM EST] daemon.warn: odhcpd[1913]: No default route present, setting ra_lifetime to 0!

skuzzer · February 4, 2026, 3:11pm

Have to restart Adblock on a daily basis. Never had to do this before! Xiaomi Redmi AX6000 Uboot layout. Also when powering up Windows the WiFi is very slow to get a connection

Sti · February 4, 2026, 3:24pm

I experienced the same problem. The issue is related to the client ID being automatically sent after a specific commit. There's regression.

anon63541380 · February 4, 2026, 4:19pm

What fdes it have to do with OpenWrt?

egc · February 4, 2026, 4:35pm

Have you tried to power down the isp modem en router and after 5 minutes power up the modem and after 5 more minutes power up the router?
Maybe then the modem will learn the clientid?

Sti · February 4, 2026, 4:44pm

I've tried many times but haven't been able to get an IP address.

ssmurf · February 4, 2026, 4:47pm

I never did get past 50% on the web interface regardless of how many times I tried. However, from command line “owut upgrade” fired off without issues and I am now running RC4.

skuzzer · February 4, 2026, 4:53pm

I have reworded it slightly so all can understand?

Edit: See what you mean. It started at the same time as rc4 was released. On my other computer I started Linux & it was quite quick loading WiFi. Also on Android phone connection is instant. Maybe a Windose issue? I wouldn’t mind but it has the fastest processor of all by some margin.

egc · February 4, 2026, 4:57pm

Then your ISP modem seems not be able to handle the presence of a clientid not sure who's fault that is but I see the problem
Did you actually check if a clientid is send (from the command line ps | grep [u]dhcpc) if not you can perhaps try to set one based on your MAC address

Sti · February 4, 2026, 5:14pm

The operator is establishing the connection based on the MAC address. It was already defined in the interface. I also tried entering the MAC address in the Client ID to send when requesting DHCP section, but I couldn't get an IP address. The operator's modem is at least 10 years old, and there's no alternative, so as you said, I think the problem stems from it. Many other users are likely experiencing this problem as well; it would be great if at least a backward compatibility option could be provided.