OpenWrt 24.10.0-rc6 - Sixth Release Candidate

RaylynnKnight · 2025-01-25T03:56:28.598Z

Additional data on the apm821xx/nand target. The Cisco MX60 and MX60W devices update from 24.10.0-rc4 to 24.10.0-rc6 without the image version message that the Netgear WNDAP620 and WNDR4700 had.

brada4 · 2025-01-25T04:22:10.559Z

It is about time to rebase tens of turned down dsa led patches for years literarily.
https://openwrt.org/releases/24.10/changelog-24.10.0-rc6#addressed_bugs

RaylynnKnight · 2025-01-25T07:49:59.361Z

I have successfully installed OpenWrt on a TP-Link Festa FR205 (which I discovered is identical hardware to the ER605v2, see detail on https://wikidevi.wi-cat.ru/TP-LINK_Festa_FR205) using the instructions for the TP-Link ER605v2. I was then able to update to 24.10.0-rc6.

McGarnagle · 2025-01-25T08:13:26.300Z

ipq806x checking in
24.10.0-rc5 transition to DSA was a bit rough in luci, VLAN config on the bridge wouldn't save without causing connection loss, automatic rollback saved it, finally figured out i needed to create the VLAN devices before saving the VLAN config on the bridge. Would benefit from more documentation or even better a UI helper.

24.10.0-rc6 attempted Attended System Upgrade but it would just keep searching for firmware, never failed or succeeded. Downloaded the sysupgrade .bin and flashed no problem.

Overall seems like a good stable release, I haven't benchmarked it though

ayanna · 2025-01-25T08:25:09.615Z

Upgraded via owut without issues a test device TP-Link RE200 V4 (MT76x8). I didn't try yet on my other "home network" devices Archer C6 V2 due to 5GHz firmware issue indicated in "known issues at first post" even if workaround exists. I will be waiting for it to be hopefully solved in later release candidate or final 24.10.
Thank you!

Dante · 2025-01-25T12:20:55.637Z

Might not be the best place to ask, but any reason why ath10k-ct firmware is still the default (at least on Archer C7 v2) and not the mainline one? The CT is 4 years older than the mainline according to the package manager, and it doesn't seem like it will be maintained anymore.

porkepik · 2025-01-25T14:57:09.263Z

ERLite 3. Is there something I should address in this log?

root@OpenWrt:~# logread|grep -E "(warn|err|debug)"
Fri Jan 24 20:30:09 2025 kern.debug kernel: [    0.000000] pcpu-alloc: s29232 r8192 d32208 u69632 alloc=17*4096
Fri Jan 24 20:30:09 2025 kern.debug kernel: [    0.000000] pcpu-alloc: [0] 0 [0] 1
Fri Jan 24 20:30:09 2025 kern.debug kernel: [    8.334956] libata version 3.00 loaded.
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    8.523795] gpio gpiochip0: Static allocation of GPIO base is deprecated, use dynamic allocation.
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    8.629653] Driver 'pata_octeon_cf' needs updating - please use bus_type methods
Fri Jan 24 20:30:09 2025 kern.err kernel: [    8.637934] physmap-flash 1f400000.nor: Can't get bank width from device tree
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    8.645220] physmap-flash: probe of 1f400000.nor failed with error -22
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    8.654902] [Firmware Warn]: /soc@0/mdio@1180000001800/ethernet-phy@7: Whitelisted compatible string. Please remove
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    8.669120] [Firmware Warn]: /soc@0/mdio@1180000001800/ethernet-phy@6: Whitelisted compatible string. Please remove
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    8.681104] [Firmware Warn]: /soc@0/mdio@1180000001800/ethernet-phy@5: Whitelisted compatible string. Please remove
Fri Jan 24 20:30:09 2025 kern.info kernel: [    8.770156] EDAC DEVICE1: Giving out device to module octeon-l2c controller octeon_l2c_err: DEV octeon_l2c_edac (POLLED)
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    8.792851] Interface 0 has 3 ports (RGMII)
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    8.868818] phys_mapped_flash: Swapping erase regions for top-boot CFI table.
Fri Jan 24 20:30:09 2025 kern.warn kernel: [    9.032269] OF: fdt: not creating '/sys/firmware/fdt': CRC check failed
Fri Jan 24 20:30:09 2025 kern.debug kernel: [   10.723934] sd 0:0:0:0: [sda] Mode Sense: 23 00 00 00
Fri Jan 24 20:30:09 2025 kern.warn kernel: [   10.724291] sd 0:0:0:0: [sda] No Caching mode page found
Fri Jan 24 20:30:09 2025 kern.warn kernel: [   10.729732] sd 0:0:0:0: [sda] Assuming drive cache: write through
Fri Jan 24 20:30:09 2025 kern.warn kernel: [   10.770600] This architecture does not have kernel memory protection.
Fri Jan 24 20:30:09 2025 kern.debug kernel: [   10.781283]   with arguments:
Fri Jan 24 20:30:09 2025 kern.debug kernel: [   10.781298]     /sbin/init
Fri Jan 24 20:30:09 2025 kern.debug kernel: [   10.781308]   with environment:
Fri Jan 24 20:30:09 2025 kern.debug kernel: [   10.781315]     HOME=/
Fri Jan 24 20:30:09 2025 kern.debug kernel: [   10.781322]     TERM=linux
Fri Jan 24 20:30:09 2025 user.err kernel: [   17.910035] block: unable to load configuration (fstab: Entry not found)
Fri Jan 24 20:30:09 2025 user.err kernel: [   17.924048] block: unable to load configuration (fstab: Entry not found)
Fri Jan 24 20:30:09 2025 user.err kernel: [   23.250860] block: unable to load configuration (fstab: Entry not found)
Fri Jan 24 20:30:09 2025 user.err kernel: [   23.264895] block: unable to load configuration (fstab: Entry not found)
Fri Jan 24 20:30:09 2025 kern.warn kernel: [   23.317791] overlayfs: null uuid detected in lower fs '/', falling back to xino=off,index=off,nfs_export=off.
Fri Jan 24 20:30:09 2025 user.warn kernel: [   26.121907] urandom-seed: Seed file not found (/etc/urandom.seed)

phinn · 2025-01-25T16:59:29.596Z

After not seeing Packet Steering - Enabled (All CPU) option in LuCI on PC Chrome (incognito mode so no cache), I enabled it on my iPhone Chrome (which doesn't use chromium but apple webkit on iOS I believe). This enabled fine, but amusingly now that it's enabled it shows up in PC Chrome too like it was all previous snapshot/24.10 versions.

fakemanhk · 2025-01-25T17:04:34.053Z

Can confirm that LAN port never works when using RC6 with my R4S.

zytos · 2025-01-25T18:04:55.390Z

raspberry pi 5 8 go + hat GeeekPi U2500 M.2 NVME Dual USB3.0 to 2.5g Ethernet
change rc5 ok to rc6 with configuration reset ok
rc6 wifi no test it will come

dttsr · 2025-01-25T18:24:55.709Z

Upgraded Netgear WAX206 to rc6 from rc5. Smart TV (wired) dont connect anymore to the internet. The tv detect the connection correctly but no app (youtube etc) works, all the apps says are offline. I think can be related to this maybe? kernel reports: key addition failed
I dont know what else can be, no changes in configuration or cables.
Btw downgraded to rc5 still not working. I had to downgrade + restore an old config backup to get back online on smart tv

Sat Jan 25 19:17:53 2025 daemon.notice netifd: Network device 'lan2' link is up
Sat Jan 25 19:17:58 2025 daemon.info dnsmasq-dhcp[1]: DHCPDISCOVER(br-lan) mac-addredd-edited
Sat Jan 25 19:17:58 2025 daemon.info dnsmasq-dhcp[1]: DHCPOFFER(br-lan) 192.168.1.152 mac-addredd-edited
Sat Jan 25 19:17:58 2025 daemon.info dnsmasq-dhcp[1]: DHCPREQUEST(br-lan) 192.168.1.152 mac-addredd-edited
Sat Jan 25 19:17:58 2025 daemon.info dnsmasq-dhcp[1]: DHCPACK(br-lan) 192.168.1.152 mac-addredd-edited LGwebOSTV
Sat Jan 25 19:18:10 2025 daemon.err hostapd: nl80211: kernel reports: key addition failed

stevewaffler · 2025-01-25T22:15:22.251Z

Bonding/link aggregation performance is not performing as expected in active backup mode. In prior versions of this rc I had 10gb and 1gb links setup as active-backup with the 10gb as primary and the 1gb as failover.

In rc6, with this configuration and the 10gb link active, I'm seeing 500MB lan-lan speeds over the bond. Unbonded, the 10gb and 1gb connections perform at wire speed.

Monitoring the traffic through the switch shows traffic flowing through both bonded NICs instead of only the primary similar to what would be expected in LACP mode.

This is the only possible relevant system log entry related
daemon.notice netifd: wan (11558): Command failed: ubus call network.interface notify_proto { "action": 0, "link-up": false, "keep": false, "interface": "wan" } (Permission denied)

Running x86_64 generic, both are intel NICs using ixgbe and igb drivers

wjwj · 2025-01-25T22:35:50.029Z

Strange, I can not confirm this. I was curious and gave RC6 (r28388-58d0057481) a test run on my NanoPi R4S. Everything works normal, LAN port is up and running without any problems.

We discussed this here a couple of times already. The R4S seems to be a bit "picky" when it comes to power supplies and also SD cards. I for example recently had issues booting from SD cards that actually worked before. I then used a different one and everything worked again ... it actually looks like the LAN port does not come online, but actually I believe the device does not boot at all ... unfortunately never had time to connect an UART cable and take a deeper look.

Can you try a different SD card?

stevewaffler · 2025-01-25T22:57:16.358Z

In other news, upgrade of TP-Link Archer C7 v2 with auc went flawlessly. Everything seems to be working great.

wokkocher · 2025-01-25T23:39:30.260Z

This is the first time I upgraded with owut on command line and kept my config. My TP-Link TL-WR902ACv3 is working fine, so far.

I am still unhappy with mbedtls and its MBEDTLS_SSL_TLS1_3_COMPATIBILITY_MODE compile option. I will try free some time and compile my own version and see if this enables openvpn to do real TLSv1.3 . For now I am using wolfssl additionally to achieve this, but it is a waste of space on my tiny travel router.

JustAnotherEndUser · 2025-01-25T23:55:36.578Z

I am seeing this error for the first time with -rc6 on x86/64 Generic
( This is a fresh install. )

[   37.553634] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[   37.762593] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[   37.911412] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[   38.051659] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[   39.863401] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[   40.080742] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left

EDIT: It first appears at startup between 37 - 40 seconds of uptime.
In searching, not finding anything on OpenWRT, but did find this similar error on Stack Overflow:
https://stackoverflow.com/questions/18230714/percpu-allocation-failed-size-256-align-256-failed-to-allocate-new-chunk

EDIT: These errors appeared in the log after about 19 minutes uptime:

[ 1140.604277] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[ 1140.829176] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[ 1141.066370] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[ 1141.279960] percpu: allocation failed, size=16 align=8 atomic=1, atomic alloc failed, no space left
[ 1141.281332] percpu: limit reached, disable warning

After installing -rc6, I did restore a backup from 24.10.0-rc5. Not sure if that makes any difference.

devexxis · 2025-01-26T00:15:14.101Z

LAN and WAN stop working on the R5S. Had the FriendlyWRT 23.05.5 installed on eMMC. Upgrading using this or rc5 kills LAN and WAN access.

brada4 · 2025-01-26T00:38:31.043Z

You cannot migrate their fullcone firewall, you must reset config (extfs) or at least restore default /etc/config/firewall from /rom/ to / (squash)

wjwj · 2025-01-26T07:18:37.760Z

Not in my case. I specifically rebooted the device twice in my test and it worked just fine. We discussed a similar reboot issue in the past, but it does not seem to be a general problem…. without UART information close to impossible to debug.

ManuVice · 2025-01-26T09:54:13.018Z

flashed from rc2 to rc6 on NanoPi R4S no problems so far. Reboot is working.

Last time I had issues with reboot I changed my power supply to 5V/3.5A and problem was gone