root@NBG7815:~# ethtool -s 10g speed 5000 duplex full autoneg on
Cannot advertise speed 5000 duplex full
But you still can:
root@NBG7815:~# ethtool -s 10g speed 1000 duplex full autoneg on
root@NBG7815:~# [276395.111688] nss-dp 3a007000.dp6-syn 10g: PHY Link is down
[276395.112045] br-lan: port 1(10g) entered disabled state
[276395.118677] br-home: port 1(10g.90) entered disabled state
[276395.122871] br-IoT: port 1(10g.91) entered disabled state
[276407.587340] nss-dp 3a007000.dp6-syn 10g: PHY Link up speed: 1000
[276407.587512] br-lan: port 1(10g) entered blocking state
[276407.592430] br-lan: port 1(10g) entered forwarding state
[276407.597920] br-home: port 1(10g.90) entered blocking state
[276407.603025] br-home: port 1(10g.90) entered forwarding state
[276407.608696] br-IoT: port 1(10g.91) entered blocking state
[276407.614307] br-IoT: port 1(10g.91) entered forwarding state
It is known there is a problem on nss support and 10g port.
Sorry, I'm discovering this issue... I limit the rate to 5000 or 2500 without any problem before ...
Limiting at 1G is not possible today for my usage, so, the port will stay at 10G, waiting for a patch or so. The device become very hot with the 10G + USB port utilisation (RTL433). I just pray for the fan to stay alive !
I wrote before that I had a problem with pppoe. I reinstalled the latest version of @asvio . But this time I let it reset the settings. All my problems were solved. It was probably a configuration error from the past.
Now I tried your 10g port speed setting and set it to 2500 and there was no problem. Although it could be because my lan card is also 2.5G. As additional information, I disabled the usteer service and tailscale service.
root@OpenWrt:~# ethtool 10g
Settings for 10g:
Supported ports: [ ]
Supported link modes: 10baseT/Full
100baseT/Full
1000baseT/Full
10000baseT/Full
1000baseKX/Full
10000baseKX4/Full
10000baseKR/Full
2500baseT/Full
5000baseT/Full
Supported pause frame use: Symmetric Receive-only
Supports auto-negotiation: Yes
Supported FEC modes: Not reported
Advertised link modes: 10baseT/Full
100baseT/Full
1000baseT/Full
10000baseT/Full
1000baseKX/Full
10000baseKX4/Full
10000baseKR/Full
2500baseT/Full
5000baseT/Full
Advertised pause frame use: Symmetric Receive-only
Advertised auto-negotiation: Yes
Advertised FEC modes: Not reported
Link partner advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
2500baseT/Full
Link partner advertised pause frame use: Symmetric Receive-only
Link partner advertised auto-negotiation: Yes
Link partner advertised FEC modes: Not reported
Speed: 2500Mb/s
Duplex: Full
Port: Twisted Pair
PHYAD: 8
Transceiver: external
Auto-negotiation: on
MDI-X: Unknown
Link detected: yes
I check my previous builds, I see that I have ethtool-full package installed, here in this build I have the ethtool ... Is there a link with this behaviour ?
I've to rebuild with the ethtool-full to see if that help. Do not have the time for now but I'll check in this way.
For the 5G AP, nothing to say today, all working fine, but at 80Mhz, still waiting for the bug.
When we want to install a new firmware, it asks if we want to keep the current settings. In my previous installations, I kept it. But this caused problems.
I'm testing on 160Mhz and channel 100 and usteer disable.
eight devices conneted (3 devices ax 2x2, 2 devices ac 2x2, 2 device ac 1x1 and 1 old device an).
No problem since last update (11h)
Has anyone had this problem on channel 100 and 160mhz?
Tue Nov 19 12:24:29 2024 daemon.err hostapd: 20/40 MHz: center segment 0 (=114) and center freq 1 (=5510) not in sync
Tue Nov 19 12:24:29 2024 daemon.err hostapd: Failed to set beacon parameters
Tue Nov 19 12:24:29 2024 kern.info kernel: [78225.846326] br-home: port 4(phy0-ap1) entered disabled state
Tue Nov 19 12:24:29 2024 kern.info kernel: [78225.883865] ath11k c000000.wifi phy0-ap1 (unregistering): left allmulticast mode
Tue Nov 19 12:24:29 2024 kern.info kernel: [78225.883919] ath11k c000000.wifi phy0-ap1 (unregistering): left promiscuous mode
Tue Nov 19 12:24:29 2024 kern.info kernel: [78225.890348] br-home: port 4(phy0-ap1) entered disabled state
Tue Nov 19 12:24:30 2024 daemon.notice netifd: Network device 'phy0-ap1' link is down
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.113534] ath11k c000000.wifi phy0-ap0: left allmulticast mode
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.113598] ath11k c000000.wifi phy0-ap0: left promiscuous mode
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.118887] br-lan: port 7(phy0-ap0) entered disabled state
Tue Nov 19 12:24:30 2024 daemon.notice netifd: Network device 'phy0-ap0' link is down
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.488785] br-lan: port 7(phy0-ap0) entered blocking state
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.488831] br-lan: port 7(phy0-ap0) entered disabled state
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.493239] ath11k c000000.wifi phy0-ap0: entered allmulticast mode
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.499262] ath11k c000000.wifi phy0-ap0: entered promiscuous mode
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.505405] br-lan: port 7(phy0-ap0) entered blocking state
Tue Nov 19 12:24:30 2024 kern.info kernel: [78226.511234] br-lan: port 7(phy0-ap0) entered forwarding state
Tue Nov 19 12:24:30 2024 daemon.err hostapd: could not get valid channel
Tue Nov 19 12:24:30 2024 kern.info kernel: [78227.023493] br-lan: port 7(phy0-ap0) entered disabled state
---------------------------------------------
Tue Nov 19 12:54:53 2024 daemon.err hostapd: could not get valid channel
Tue Nov 19 12:54:53 2024 daemon.err hostapd: 20/40 MHz: center segment 0 (=114) and center freq 1 (=5510) not in sync
Tue Nov 19 12:54:53 2024 daemon.err hostapd: Can't set freq params
Tue Nov 19 12:54:53 2024 daemon.err hostapd: DFS start_dfs_cac() failed, -1
Tue Nov 19 12:54:53 2024 daemon.err hostapd: 20/40 MHz: center segment 0 (=114) and center freq 1 (=5510) not in sync
Tue Nov 19 12:54:53 2024 daemon.err hostapd: Can't set freq params
Tue Nov 19 12:54:53 2024 daemon.err hostapd: DFS start_dfs_cac() failed, -1
Tue Nov 19 12:54:53 2024 daemon.err hostapd: 20/40 MHz: center segment 0 (=114) and center freq 1 (=5510) not in sync
Tue Nov 19 12:54:53 2024 daemon.err hostapd: Can't set freq params
Tue Nov 19 12:54:53 2024 daemon.err hostapd: DFS start_dfs_cac() failed, -1
Tue Nov 19 12:54:53 2024 daemon.err hostapd: 20/40 MHz: center segment 0 (=114) and center freq 1 (=5510) not in sync
Tue Nov 19 12:54:53 2024 daemon.err hostapd: Can't set freq params
Tue Nov 19 12:54:53 2024 daemon.err hostapd: DFS start_dfs_cac() failed, -1
Tue Nov 19 12:54:53 2024 daemon.err hostapd: 20/40 MHz: center segment 0 (=114) and center freq 1 (=5510) not in sync
Tue Nov 19 12:54:53 2024 daemon.err hostapd: Can't set freq params
Tue Nov 19 12:54:53 2024 daemon.err hostapd: DFS start_dfs_cac() failed, -1
Tue Nov 19 12:54:53 2024 daemon.err hostapd: 20/40 MHz: center segment 0 (=114) and center freq 1 (=5510) not in sync
Tue Nov 19 12:54:53 2024 daemon.err hostapd: Can't set freq params
Tue Nov 19 12:54:53 2024 daemon.err hostapd: DFS start_dfs_cac() failed, -1
Tue Nov 19 12:54:53 2024 daemon.err hostapd: 20/40 MHz: center segment 0 (=114) and center freq 1 (=5510) not in sync
Tue Nov 19 12:54:53 2024 daemon.err hostapd: Can't set freq params
Tue Nov 19 12:54:53 2024 daemon.err hostapd: DFS start_dfs_cac() failed, -1
I had to reset the 5ghz radio to be able to get the phy0-ap* up again.
This is the first time i've catched it. I hope the problem is not as strong as the one with the r7800 and DFS channels
I'm starting to think that channel 100 is not going to be suitable for use on 160mhz.
The 10 minutes wait until the network is available and these radar detection problems that can leave you without WLAN without warning don't make me very happy...
I'm reflashing the router currently with your latest "official" build, but without restoring my conf. I've many mistakes with multiple services (usb hang, wireguard very slow, wifi 5G, apt refuse to work on my proxmox server (hugh !), the site "reddit" not accessible anymore (what ?? lol)
Are you referring to the latest version of the openwrt repository?
Or are you referring to the latest one I've published based on the 24.10 branch with nss support?
If it's the latter, I would ask you not to mention the word "official" to the firmware I publish because they are builds for people who either don't have the skills or don't have the resources to compile my repositories, which I've made public in case anyone is interested in taking something from there for their own repositories.
Taking advantage of this post I want to make you all see the following:
My skills do not go beyond recompiling data from various sources and joining them together to make them work according to my needs.
I don't test advanced things so I can't guarantee that they will work and also if something doesn't work and the solution isn't published somewhere on the internet I assure you that I won't be able to fix it.
I appreciate all of you posting your problems because at some point they may also be mine and it also allows me to reach out to people who really know about this the problems that arise.
That's the reason I put it in quotes, I know that you do all you can with your skills, I do the same with your work. Sorry if that hurt you, that's not my goal, but I think your work, and the work of all other here, can help to have the real official version.
I wish that all here know that official versions can't be find in a help community forum for a model.
I've finished the restore, because I don't want to be fired for a router, I've to go at work now ! I hope that wifi still working when coming back home.
To be continued ... And thank you a lot for your work and your precious help
Basically you need the AQR temperature and ATH11k temperature to create any usefull "thermal management" for this device.
ATH11k is not supporting an thermal interface to the kernel to read the temperature. You can only access the temperature via the driver itself. So as soon as someone is disabling wifi for any reason (like schedule) the temperature reading is dead.
For the AQR (if I understand the dev's right). Nobody wants a "dead" thermal zone within DTS files because the driver is not supporting it properly (due to incompatible data types).
For ath11k I doubt that there will be any solution. The same applies basically for the AQR as well. Both is beyond my skills to solve it and I didn't see here anyone able to take on this.
So only solution is doing your own builds modifying the DTS file and using a script to control the fan.