I already reported this as an issue. For me its only happens with the OpenWRT firmware, MTK stock runs fine.
I suspect it is maybe a DFS problem or some WiFi client causing it.
Does anyone else have this issue with AX devices and MT6000?
Are you using AX or only AC?
I have it for several months now and would be happy to get a solution.
Thanks in advance!
No issues here with a house full of apple products (multiple apple TV’s, multiple iPad’s, an iMac, and multiple iPhones).
1 Like
@SiXX I've tested the 2024 firmwares extensively, I didn't see a single difference in throughput, latency, AQL performance, etc. so I went back to using the one baked into main snapshot builds without worrying about it.
@Delphius I have 2 AX devices both work flawless (ThinkPad X1 and iPhone 15) on 5GHz. I also have several AC devices working flawless including an old discontinued (but still amazing) Chromecast Audio plugged into Studio Monitors but I have those default to 2.4GHz spectrum. I'm using a single SSID for both so it's more seamless and it works great, zero crashes etc.
The new mt76 driver update nbd put in the snapshots from last week added spectrum management (spectrum_mgmt) not sure what this does, read MediaTek's post about the changes but in testing don't see any difference. Getting the same great range and throughput I've been getting for months, gigabit downloads over wifi 6 (internet speedtest at 800-950 Mbps) is typical with this target in my apt.
2 Likes
Thanks for the fast replies.
As far as I understand from your posts before the current snapshot is based on an older firmware. Do you think it is worth to try a newer version with the issues I described?
Do you have any information for me where to find and how to implement it?
Yes the firmware is simple to replace, but I suggest you just sysupgrade to @pesa1234 build here since he has it built in. Please report back on that thread if you see any improvement, I couldn't find any differences when I swapped them. I suspect it's minor bug fixes that are hard to detect but MT isn't posting a changelog for the fw.
2 Likes
Thanks, I will give it a try. Could please also let me know how to replace the firmware manually?
1 Like
You would clone OpenWrt repo etc, pull in mtk repo and build from source.
1 Like
or, the easy way, you can just download the firmware and simply replace the files on the router, but, as @phinn said, I used the new firmware too, and didn't notice any difference in performance. Maybe worth a try.
Look here @_FailSafe topic ; [GL.iNET Flint 2 (GL-MT6000) Snapshot / Experimental / Bleeding Edge - #10 by _FailSafe]
3 Likes
Any solution for unattended update? It doesn't work at all, only when you modify a package, but this happens in all versions, so every time you want to update you need a new package...
From recent issues I wouldn't recommend attendedsysupgrade. Its much safer using firmware selector.
2 Likes
That's fine, but it shouldn't interfere so it doesn't work.
You could always use the image builder as well... take all the online automated services out of the equation altogether. Extremely reliable and will produce exactly what you want every time.
4 Likes
nice words but nothing to do with it, it's not a question of whether it's pretty or not, reliable or not, it's simply something that should work and doesn't.
…so create a new thread in the appropriate section with your question, concerns, etc. You’ll reach a wider audience that way.
Welcome to the world of development 
5 Likes
Hello all,
I've moved to OpenWrt r26329-cffd3ad8d7 after WiFi frustrations with the stock firmware.
After doing so, I get an intermittent crash on eth1 (my WAN port attached to ISP ONT)
My setup is very minimal,
- HW / SW Offloading Enabled
- SQM Disabled
- Collectd Installed along with plugins
- PPPoe WAN Setup (eth1)
Has anyone seem similar issues on this hardware? Noticed a Github issue for similar hardware using the same driver.
Have added ethtool -K eth1 tso off to /etc/rc.local, but it's still crashing randomly - testing with HW NAT acceleration off now.
304956.265307] mt7530-mdio mdio-bus:1f lan5: Link is Up - 100Mbps/Full - flow control rx/tx
[304956.273514] br-lan: port 5(lan5) entered blocking state
[304956.278826] br-lan: port 5(lan5) entered forwarding state
[305686.834382] ------------[ cut here ]------------
[305686.839086] NETDEV WATCHDOG: eth1 (mtk_soc_eth): transmit queue 1 timed out 7880 ms
[305686.846849] WARNING: CPU: 1 PID: 0 at dev_watchdog+0x240/0x248
[305686.852764] Modules linked in: pppoe ppp_async nft_fib_inet nf_flow_table_inet pppox ppp_generic nft_reject_ipv6 nft_reject_ipv4 nft_reject_inet nft_reject nft_redir nft_quota nft_numgen nft_nat nft_masq nft_log nft_limit nft_hash nft_flow_offload nft_fib_ipv6 nft_fib_ipv4 nft_fib nft_ct nft_chain_nat nf_tables nf_nat nf_flow_table nf_conntrack mt7915e(O) mt76_connac_lib(O) mt76(O) mac80211(O) iptable_mangle iptable_filter ipt_REJECT ipt_ECN ip_tables cfg80211(O) xt_time xt_tcpudp xt_tcpmss xt_statistic xt_multiport xt_mark xt_mac xt_limit xt_length xt_hl xt_ecn xt_dscp xt_comment xt_TCPMSS xt_LOG xt_HL xt_DSCP xt_CLASSIFY x_tables slhc sch_cake nfnetlink nf_reject_ipv6 nf_reject_ipv4 nf_log_syslog nf_defrag_ipv6 nf_defrag_ipv4 libcrc32c compat(O) crypto_safexcel sch_tbf sch_ingress sch_htb sch_hfsc em_u32 cls_u32 cls_route cls_matchall cls_fw cls_flow cls_basic act_skbedit act_mirred act_gact ifb tun sha512_arm64 sha1_ce sha1_generic seqiv md5 geniv des_generic libdes authencesn authenc leds_gpio xhci_plat_hcd
[305686.852945] xhci_pci xhci_mtk_hcd xhci_hcd gpio_button_hotplug(O) usbcore usb_common aquantia
[305686.951067] CPU: 1 PID: 0 Comm: swapper/1 Tainted: G O 6.6.30 #0
[305686.958529] Hardware name: GL.iNet GL-MT6000 (DT)
[305686.963300] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[305686.970327] pc : dev_watchdog+0x240/0x248
[305686.974411] lr : dev_watchdog+0x240/0x248
[305686.978491] sp : ffffffc08000be50
[305686.981876] x29: ffffffc08000be50 x28: ffffffc080b805f8 x27: ffffffc080b950c0
[305686.989079] x26: ffffffc080b969c0 x25: ffffffc08000bf10 x24: 0000000000001ec8
[305686.996282] x23: ffffffc080b96000 x22: 0000000000000001 x21: ffffff8001ad9000
[305687.003485] x20: ffffff8001ae0140 x19: ffffff8001ad94c0 x18: 000000000000015a
[305687.010688] x17: 2074756f2064656d x16: 6974203120657565 x15: ffffffc080baa1e8
[305687.017890] x14: 000000000000040e x13: 000000000000015a x12: 00000000ffffffea
[305687.025093] x11: 00000000ffffefff x10: ffffffc080c021e8 x9 : ffffffc080baa190
[305687.032296] x8 : 0000000000017fe8 x7 : c0000000ffffefff x6 : 0000000000000001
[305687.039498] x5 : 0000000000000000 x4 : 0000000000000000 x3 : 0000000000000000
[305687.046702] x2 : ffffff803fd9c490 x1 : ffffffbfbf219000 x0 : 0000000000000047
[305687.053905] Call trace:
[305687.056424] dev_watchdog+0x240/0x248
[305687.060162] call_timer_fn.isra.0+0x20/0x78
[305687.064419] __run_timers.part.0+0x1e4/0x248
[305687.068761] run_timer_softirq+0x38/0x6c
[305687.072755] __do_softirq+0xf8/0x25c
[305687.076402] ____do_softirq+0xc/0x14
[305687.080049] call_on_irq_stack+0x24/0x34
[305687.084042] do_softirq_own_stack+0x18/0x24
[305687.088294] irq_exit_rcu+0xb4/0xcc
[305687.091856] el1_interrupt+0x34/0x4c
[305687.095505] el1h_64_irq_handler+0x14/0x1c
[305687.099672] el1h_64_irq+0x68/0x6c
[305687.103145] default_idle_call+0x24/0x34
[305687.107141] do_idle+0xa8/0xf4
[305687.110271] cpu_startup_entry+0x30/0x38
[305687.114265] cpu_die_early+0x0/0x84
[305687.117826] __secondary_switched+0x6c/0x70
[305687.122081] ---[ end trace 0000000000000000 ]---
Seeing the same on my Mercusys MR90x, PPPoE Authentication, same Filogic chipset.
Are you an IPoE or PPPoE authentication for your WAN port?
Might be worth creating a new thread.
I've gone back to my Netgear WAX206 which just works.
this is pretty off topic, if you want to create a new thread regarding this new mt7915 firmware, I'll happily reply.
Thank you, yes, using PPPoE....unfortunately.
I've just done a fresh install with the snapshot from last night, which has stayed stable for 24 hours.
https://forum.openwrt.org/t/openwrt-23-05-3-service-release/192587/2
mediatek: MT7981/MT7986: fix Ethernet rx hang issue
This fix affects our chipset doesn't it?
I've also started a shell script so I can track what changes may be affecting the stability, without manual GUI changes.
3 Likes
I got a lot to read on this thread, to anyone who reads this, is it worth installing 23.05.3 on the MT6000? Are there important bugs that affect basic options?
Since that's the first stable release, I'd like to know how's running, thanks.