Upgraded gl-mt6000 rc4->rc5.
No issues on my humble soho setup.
Upgraded BPI R4, mt3000 and Redmi AX6s from rc4 to rc5 with owut, all completely smooth.
2 Likes
ASUS TUF AX4200/AX6000 bootloop on new u-boot versions still not fixed..
Upgraded my Linksys E8450 router successfully.
The IPv6 issue mentioned in the -rc4 post didn't surface.
Dynalink DL-WRX36
1./ System & corresponding Kernel log errors (there are no files in the /tmp/overlaydirectory).
IIRC I haven't noticed these before...
Wed Jan 8 21:13:40 2025 user.err kernel: [ 9.657573] mount_root: failed to launch kmodloader from internal overlay
Wed Jan 8 21:13:40 2025 user.err kernel: [ 9.558827] kmodloader: failed to open /tmp/overlay/upper/lib/modules/6.6.69/mbcache.ko
Wed Jan 8 21:13:40 2025 user.err kernel: [ 9.552766] kmodloader: failed to open /tmp/overlay/upper/lib/modules/6.6.69/jbd2.ko
Wed Jan 8 21:13:40 2025 user.err kernel: [ 9.535114] kmodloader: failed to open /tmp/overlay/upper/lib/modules/6.6.69/ext4.ko
[ 9.535114] kmodloader: failed to open /tmp/overlay/upper/lib/modules/6.6.69/ext4.ko
[ 9.552766] kmodloader: failed to open /tmp/overlay/upper/lib/modules/6.6.69/jbd2.ko
[ 9.558827] kmodloader: failed to open /tmp/overlay/upper/lib/modules/6.6.69/mbcache.ko
[ 9.657573] mount_root: failed to launch kmodloader from internal overlay
2. The wan rx/tx LED is only showing a consistent orange, green is off.
using default /etc/config/system LED settings...
config led 'led_wan_port_link_green'
option name 'WAN-PORT-LINK-GREEN'
option sysfs '90000.mdio-1:1c:green:wan'
option trigger 'netdev'
option mode 'link_2500'
option dev 'wan'
config led 'led_wan_port_link_yellow'
option name 'WAN-PORT-LINK-YELLOW'
option sysfs '90000.mdio-1:1c:yellow:wan'
option trigger 'netdev'
option mode 'tx rx link_10 link_100 link_1000'
option dev 'wan'
How to reconfigure, i.e. to have a solid green LED (link) permanently on, and the orange LED to flicker with tx/rx.
I played with a multitude of settings, still no luck! 
3. Same with the LAN port LED's, they're also showing a consistent blinking orange, green is off, when connected to a device.
Im currently on 24.10.0-rc2, Attended sysupgrade does not give me an option to install rc5 , using openwrt one device
After an upgrade for octeon, I'm told to fsck my /dev/sda1. What is the exact procedure?
root@OpenWrt:~# logread|grep fsck
Wed Jan 8 09:40:32 2025 kern.warn kernel: [ 23.078933] FAT-fs (sda1): Volume was not properly unmounted. Some data may be corrupt. Please run fsck.
root@OpenWrt:~# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
loop0 7:0 0 3.5G 0 loop /overlay
sda 8:0 1 3.8G 0 disk
ââsda1 8:1 1 142M 0 part
ââsda2 8:2 1 3.5G 0 part /rom
root@OpenWrt:~# fsck.vfat /dev/sda1
-ash: fsck.vfat: not found
root@OpenWrt:~# find /usr/ -name fsck*
/usr/sbin/fsck.ext2
/usr/sbin/fsck.ext3
/usr/sbin/fsck.ext4
Obviously missing something. Which package would be required?
root@OpenWrt:~# opkg install kmod-fs-vfat
Installing kmod-fs-vfat (6.6.69-r1) to root...
Downloading https://downloads.openwrt.org/releases/24.10.0-rc5/targets/octeon/generic/kmods/6.6.69-1-136e69716e1e3adc49f8197fb21277f1/kmod-fs-vfat_6.6.69-r1_mips64_octeonplus.ipk
Installing kmod-nls-cp437 (6.6.69-r1) to root...
Downloading https://downloads.openwrt.org/releases/24.10.0-rc5/targets/octeon/generic/kmods/6.6.69-1-136e69716e1e3adc49f8197fb21277f1/kmod-nls-cp437_6.6.69-r1_mips64_octeonplus.ipk
Installing kmod-nls-iso8859-1 (6.6.69-r1) to root...
Downloading https://downloads.openwrt.org/releases/24.10.0-rc5/targets/octeon/generic/kmods/6.6.69-1-136e69716e1e3adc49f8197fb21277f1/kmod-nls-iso8859-1_6.6.69-r1_mips64_octeonplus.ipk
Installing kmod-nls-utf8 (6.6.69-r1) to root...
Downloading https://downloads.openwrt.org/releases/24.10.0-rc5/targets/octeon/generic/kmods/6.6.69-1-136e69716e1e3adc49f8197fb21277f1/kmod-nls-utf8_6.6.69-r1_mips64_octeonplus.ipk
Configuring kmod-nls-cp437.
Configuring kmod-nls-iso8859-1.
Configuring kmod-nls-utf8.
Configuring kmod-fs-vfat.
root@OpenWrt:~# fsck.vfat /dev/sda1
-ash: fsck.vfat: not found
root@OpenWrt:~# opkg install dosfstools
Installing dosfstools (4.2-r4) to root...
Downloading https://downloads.openwrt.org/releases/24.10.0-rc5/packages/mips64_octeonplus/packages/dosfstools_4.2-r4_mips64_octeonplus.ipk
Configuring dosfstools.
root@OpenWrt:~# find /usr/ -name fsck*
/usr/sbin/fsck.ext2
/usr/sbin/fsck.ext3
/usr/sbin/fsck.ext4
/usr/sbin/fsck.fat
root@OpenWrt:~# fsck.vfat /dev/sda1
-ash: fsck.vfat: not found
root@OpenWrt:~# ls -ls /usr/sbin/fsck.*
0 lrwxrwxrwx 1 root root 6 Jan 4 16:35 /usr/sbin/fsck.ext2 -> e2fsck
0 lrwxrwxrwx 1 root root 6 Jan 4 16:35 /usr/sbin/fsck.ext3 -> e2fsck
0 lrwxrwxrwx 1 root root 6 Jan 4 16:35 /usr/sbin/fsck.ext4 -> e2fsck
132 -rwxr-xr-x 1 root root 132368 Jan 6 06:12 /usr/sbin/fsck.fat
root@OpenWrt:~# fsck.fat /dev/sda1
fsck.fat 4.2 (2021-01-31)
There are differences between boot sector and its backup.
This is mostly harmless. Differences: (offset:original/backup)
65:01/00
1) Copy original to backup
2) Copy backup to original
3) No action
[123?q]? 3
Dirty bit is set. Fs was not properly unmounted and some data may be corrupt.
1) Remove dirty bit
2) No action
[12?q]? 2
/dev/sda1: 4 files, 50347/286274 clusters
Apart from this, which I reckon doesn't belong here, everything else in the upgrade worked as intended. Is there a standard place where I could upload upgrade logs for if one is interested, they would be available? They are too long IMO to append right here.
I updated Rock64 from rc4 to rc5. After the update, when I enter my login and password to access the web interface, I get an ubus error screen. The message says: «Error loading module 'ubus' from file '/usr/lib/lua/ubus.so'». There are many lines with errors...
If I enter a link to any other page of the web interface, for example, the firewall, everything works. Understanding that some package is causing the error, I started looking for it. In general, the error occurs after mounting an /overlay partition from an SD card as external.
I updated the ubus packages to the latest ones available. The critical error disappeared, but now instead of the overview page, the top menu appears and below it there is an error «SyntaxError JSON.parse: expected double-quoted property name at line 11 column 2 of the JSON data».
Your old rc4 libubus from overlay overrides the new rc5 from firmware.
You need to ensure that you have the same version on overlay / extroot / whatever as in the main firmware.
Or in the later error version, a similar mismatch related some package that you have installed to overlay.
On a MT6000 with vanilla rc5 I am seeing the WiFi clients connect at 20 MHz even when the channel is set to 80 MHz.
RX/TX Rate
1200.9 Mbit/s, 80 MHz, HE-MCS 11, HE-NSS 2
286.7 Mbit/s, 20 MHz, HE-MCS 11, HE-NSS 2
This is a Google pixel 6, I have seen this twice now on the new RC5. Speed test maxes out at about 140 Mbps which is in line with what I expect from a link speed of 286 Mbps.
If I turn off Wi-Fi for a few seconds on Pixel and connect again then it will go back to 80 MHz and 1200.9 Mbps line speed.
Does anyone have ideas on how to resolve this issue?
Not seeing that issue at all on my MT6000. My iPhone and Thinkpad are both connected to 5GHz right now at 80MHz width, this is with WED disabled and modified AQL settings too.
RX Rate / TX Rate (FYI these numbers jump around)
640.3 Mbit/s, 80 MHz, HE-MCS 10, HE-NSS 1
864.8 Mbit/s, 80 MHz, HE-MCS 8, HE-NSS 2
iPhone speedtest.net just ran on 5GHz (~15ft from router): 842 Mbps down.
1 Like
After upgrading a new APs to a point on the 24.10-SNAPSHOT branch equivalent to RC5, I've noticed that my LuCI sessions get logged out very aggressively (ie, time out on the order of less than a couple of hours). Was some change made to session duration defaults as part of RC5 or previous 24.10 RCs? I don't think I noticed this in earlier 24.10 snapshot builds, nor in 23.05, but maybe I'm imagining things?
Thanks!
My windows WiFi 6 laptop connects normally.
1200.9 Mbit/s, 80 MHz, HE-MCS 11, HE-NSS 2
1200.9 Mbit/s, 80 MHz, HE-MCS 11, HE-NSS 2
Must be something with Android/pixel. I even have the random Mac address settings off..
Custom build from firmware-selector is not working for Dynalink DL-WRX36 (qualcommax/ipq807x) in version: 24.10.0-rc5 (r28304-6dacba30a7). 'init' in error box.
Working in all other versions: snapshot and stable. Worked fine in 24.10.0-rc4.
I'm seeing something similar on my Pixel 6 pro and a Netgear 3700v4 but in my case it isn't related to rc5 at all. I'm still running 23.05.5. Why do I post? To point out that it may not be an OpenWRT rc5 release issue that you are seeing. I suspect in my case it was related to neighbors getting new routers or AP's for the holidays and i'm seeing good old channel contention with the wifi code dropping back to the most narrow channel in order to blast through the interference. Try to test in the quietest time of day and see if it still happens.
3 Likes
Thank you so much! @wolfgangrupprecht This is very valuable information. It gives me another direction to think about.
I installed WiFi analyzer and checked, I am the only one using channel 36 in my area. The neighbors are using channel 161.
But as you said, maybe it is related to noise.
cat /sys/kernel/debug/ieee80211/phy1/statistics/dot11FCSErrorCount
4997101
cat /sys/kernel/debug/ieee80211/phy1/statistics/dot11RTSFailureCount
16117
The FCSErrorCount keeps rising indefinitely at a much faster rate. Almost as if all the packets have FCSError. Maybe a checksum related issue?
I do not understand the FCS Error count. Just putting the info out here in case someone knowledgeable in WiFi stack can chime in.
1 Like
Iâm a newbie to OpenWrt, and I upgraded from 24.10.0-rc4 to rc5 on a Banana Pi R3. Everything went smoothly! However, I couldnât select rc5 during the attended sysupgrade. So, I just updated using the regular sysupgrade and reinstalled my user-installed packages. It was a great learning experience, and I got to declutter my system by removing packages I no longer needed.
Are you using SQM?
I noticed some reports saying it's limited at around 500Mbps.
So, users with internet speed faster than 500Mbps will always be capped at this speed.
To test this I disabled HFO and tested SQM on MT6000. I have a 1 Gbit / 25 Mbit rated cable service (yes ISP is very asymmetrical).
Test with SQM from a couple months ago: https://www.waveform.com/tools/bufferbloat?test-id=fe841d79-96bf-4de2-bf27-17a34f5fff03
rc5 test with SQM: it's barely working the CPU cores (watching them with htop): https://www.waveform.com/tools/bufferbloat?test-id=fc6df2a2-a99f-4424-b0dc-c566bf24ed4e
EDIT: rc5 test with SQM + Packet Steering (all CPUs)/ RPS 128 enabled: https://www.waveform.com/tools/bufferbloat?test-id=abfaccee-47d0-4c45-a1cd-60f4593f91a3
EDIT2: ran it again on rc5 with my internet more quiet, SQM + PS seems fine:
https://www.waveform.com/tools/bufferbloat?test-id=905160ee-92c5-49df-a2f0-90ee2da3834c
I'd say SQM is ok, not seeing much regression with SQM + packet steering. Maybe just my congested internet I'm not the only one active at my home right now.