[Dev] Asuswrt-Merlin 388.1 development

[RMerlin]

That's crazy. What all did you do to resolve the issue?

Asus increased nvram on the HND 5.04 devices to 192 KB.

 

[SwampKracker]

This is weaker hardware than AX86U.

But there is no challenge in buying stronger hardware to do the job (with the extra horsepower to handle whatever additional work you ask it to do). It just works.

 

[tinuz]

Any idea @RMerlin ?
Or do I just need to do a clean install (dirty flash over latest stable public merlin), gt-ax11000

Everything is working, just noticed these errors with the 388.1 alpha (dirty flash) :

Oct 15 21:25:01 kernel: potentially unexpected fatal signal 7.
Oct 15 21:25:01 kernel: CPU: 2 PID: 14530 Comm: awk Tainted: P O 4.1.51 #4
Oct 15 21:25:01 kernel: Hardware name: Broadcom-v8A (DT)
Oct 15 21:25:01 kernel: task: ffffffc02434a100 ti: ffffffc01f29c000 task.ti: ffffffc01f29c000
Oct 15 21:25:01 kernel: PC is at 0x738c0
Oct 15 21:25:01 kernel: LR is at 0x62d98
Oct 15 21:25:01 kernel: pc : [<00000000000738c0>] lr : [<0000000000062d98>] pstate: a0070010
Oct 15 21:25:01 kernel: sp : 00000000ffbce228
Oct 15 21:25:01 kernel: x12: 0000000000236008
Oct 15 21:25:01 kernel: x11: 00000000ffbce24c x10: 00000000ffbce250
Oct 15 21:25:01 kernel: x9 : 00000000ffbce254 x8 : 0000000000000003
Oct 15 21:25:01 kernel: x7 : 00000000ffbce404 x6 : 00000000ffbce418
Oct 15 21:25:01 kernel: x5 : 00000000000a0000 x4 : 00000000002361d8
Oct 15 21:25:01 kernel: x3 : 00000000ffbceeab x2 : 0000000000000000
Oct 15 21:25:01 kernel: x1 : 000000000023634c x0 : 00000000ffbce24c

 

[skeal]

Any idea @RMerlin ?
Or do I just need to do a clean install (dirty flash over latest stable public merlin), gt-ax11000

Everything is working, just noticed these errors with the 388.1 alpha (dirty flash) :

Oct 15 21:25:01 kernel: potentially unexpected fatal signal 7.
Oct 15 21:25:01 kernel: CPU: 2 PID: 14530 Comm: awk Tainted: P O 4.1.51 #4
Oct 15 21:25:01 kernel: Hardware name: Broadcom-v8A (DT)
Oct 15 21:25:01 kernel: task: ffffffc02434a100 ti: ffffffc01f29c000 task.ti: ffffffc01f29c000
Oct 15 21:25:01 kernel: PC is at 0x738c0
Oct 15 21:25:01 kernel: LR is at 0x62d98
Oct 15 21:25:01 kernel: pc : [<00000000000738c0>] lr : [<0000000000062d98>] pstate: a0070010
Oct 15 21:25:01 kernel: sp : 00000000ffbce228
Oct 15 21:25:01 kernel: x12: 0000000000236008
Oct 15 21:25:01 kernel: x11: 00000000ffbce24c x10: 00000000ffbce250
Oct 15 21:25:01 kernel: x9 : 00000000ffbce254 x8 : 0000000000000003
Oct 15 21:25:01 kernel: x7 : 00000000ffbce404 x6 : 00000000ffbce418
Oct 15 21:25:01 kernel: x5 : 00000000000a0000 x4 : 00000000002361d8
Oct 15 21:25:01 kernel: x3 : 00000000ffbceeab x2 : 0000000000000000
Oct 15 21:25:01 kernel: x1 : 000000000023634c x0 : 00000000ffbce24c

Sorry no support for this version it's a test alpha not an actual legit alpha. I'm sure however that the amount of times it has been mentioned he knows of the issue. Be patient and wait on the next file to be released, however, it too may not include support. Support picks up later in the development. add on scripts are not included in his testing, I dare say he doesn't use them anyway.

 

[tinuz]

Sorry no support for this version it's a test alpha not an actual legit alpha. I'm sure however that the amount of times it has been mentioned he knows of the issue. Be patient and wait on the next file to be released, however, it too may not include support. Support picks up later in the development. add on scripts are not included in his testing, I dare say he doesn't use them anyway.

Thx, did not know that.

 

[DJones]

“Q: An error message in my syslog mentions that something is tainted. What's wrong?
A: Nothing is wrong. The Asuswrt and Asuswrt-Merlin Linux kernels are flagged as tainted because they contain closed source kernel modules from Trend Micro, Broadcom and Tuxera. See this post for the lengthy explanation how to decode that tainted and the flags following it.”

May or may not be relevant for the alpha.

 

[RMerlin]

The awk crash was an upstream bug caused by a busybox backport. I already received a patch from Asus to fix it. But as pointed out, these things are expected, as these are early alpha builds.

 

[RMerlin]

I now have XT12 and GT-AXE16000 running on 388, although on a newer codebase (388_21224).

 

[RamGuy]

I'm considering switching my RT-AX88U to a GT-AX11000 Pro. Does the GT-AX11000 Pro support the option for running Fullcone NAT? If memory serves me right, @RMerlin mentioned something in the past that only a few models had the option to switch NAT Type, so unless this option exists in the SDK there isn't much to be done about it even when running Merlin firmware.

And this new codebase won't do much for VPN performance? Running WireGuard on these BCM49408 and BCM4908 SoCs is still going to give us around ~250 Mbps?

 

[RMerlin]

Running WireGuard on these BCM49408 and BCM4908 SoCs is still going to give us around ~250 Mbps?

OpenVPN can already hit 220 Mbps on a bcm4908.

 

[Damit1]

I now have XT12 and GT-AXE16000 running on 388, although on a newer codebase (388_21224).

Which code base did you get for AX88U?
Curious here

 

[RedMaligator]

I was already using a GT-AXE16000 as my primary router for a while now, and the GT-AX6000 before that. The RT-AX88U was only used for development purposes now.

Seeing errors that either I haven't seen or started.

May 5 01:05:03 kernel: flow_ring_format : 1
May 5 01:05:03 kernel: idma_active : 1
May 5 01:05:03 kernel: Creating CPU ring for queue number 1 with 128 packets descriptor=0xffffffbffc2c9880, size_of_entry 16
May 5 01:05:03 kernel: Done initializing Ring 1 Base=0xffffff80003e4000 End=0xffffff80003e4800 calculated entries= 128 RDD Base=3e013000K descriptor=0xffffffbffc2c9880
May 5 01:05:03 kernel: ^[[0;33;40mRunner DHD Offload initialization complete^[[0m
May 5 01:05:03 kernel: Dongle Host Driver, version 17.10.121.41 (r796515)
May 5 01:05:03 kernel: Compiled in drivers/net/wireless/bcmdhd
May 5 01:05:03 kernel: wfd_registerdevice Successfully registered dev eth7 ifidx 0 wfd_idx 1
May 5 01:05:03 kernel: Register interface [eth7] MAC:
May 5 01:05:03 kernel: random: init urandom read with 15 bits of entropy available
May 5 01:05:03 kernel: proc_dostring_crashlogsave: crash log filename is /jffs/crashlog.log
May 5 01:05:03 kernel: crashFileSet: log signature invalid !
May 5 01:05:03 kernel: ret 0 retlen 65536 buff <6>ehci-platform ehci-platform.0: new USB bus registered, assigned bus number 3
May 5 01:05:03 kernel: ^[[0;33;41m[ERROR pktrunner] runnerUcast_inet6addr_event,187: Could not rdpa_system_ipv6_host_address_table_find ret=-5^[[0m
May 5 01:05:04 kernel: wfd_registerdevice Successfully registered dev wl0.1 ifidx 1 wfd_idx 0
May 5 01:05:04 kernel: Register interface [wl0.1] MAC:
May 5 01:05:04 kernel: wfd_registerdevice Successfully registered dev wl0.4 ifidx 4 wfd_idx 0
May 5 01:05:04 kernel: Register interface [wl0.4] MAC:
May 5 01:05:04 kernel: IGMP Query send failed
May 5 01:05:04 kernel: IGMP Query send failed
May 5 01:05:04 kernel: eth5 (Ext switch port: 7) (Logical Port: 15) (phyId: 1e) Link UP at 1000 mbps full duplex
May 5 01:05:04 kernel: <=== Deactivate Deep Green Mode
May 5 01:05:04 kernel: IGMP Query send failed
May 5 01:05:04 kernel: wfd_registerdevice Successfully registered dev wl1.1 ifidx 1 wfd_idx 1
May 5 01:05:04 kernel: Register interface [wl1.1] MAC:
May 5 01:05:04 kernel: IGMP Query send failed
May 5 01:05:04 kernel: IGMP Query send failed
May 5 01:05:04 kernel: netdev path : eth6.0 -> eth6
May 5 01:05:04 kernel: BCMVLAN : eth6 mode was set to RG
May 5 01:05:04 kernel: netdev path : eth6.501 -> eth6
May 5 01:05:04 kernel: BCMVLAN : eth6 mode was set to RG
May 5 01:05:04 kernel: netdev path : eth7.0 -> eth7
May 5 01:05:04 kernel: BCMVLAN : eth7 mode was set to RG
May 5 01:05:04 kernel: netdev path : eth7.501 -> eth7
May 5 01:05:04 kernel: BCMVLAN : eth7 mode was set to RG
May 5 01:05:04 kernel: netdev path : eth6.502 -> eth6
May 5 01:05:04 kernel: BCMVLAN : eth6 mode was set to RG
May 5 01:05:04 kernel: netdev path : eth7.502 -> eth7
May 5 01:05:04 kernel: BCMVLAN : eth7 mode was set to RG
May 5 01:05:05 kernel: ^[[0;33;41m[ERROR vlan] registerVlanDeviceByName,410: Failed to REGISTER VLAN Device eth6.501^[[0m
May 5 01:05:05 kernel: ^[[0;33;41m[ERROR vlan] vlanIoctl ,634: Failed to create VLAN device (eth6, eth6.501)^[[0m
May 5 01:05:05 kernel: ^[[0;33;41m[ERROR vlan] registerVlanDeviceByName,410: Failed to REGISTER VLAN Device eth7.501^[[0m
May 5 01:05:05 kernel: ^[[0;33;41m[ERROR vlan] vlanIoctl ,634: Failed to create VLAN device (eth7, eth7.501)^[[0m
May 5 01:05:05 kernel: ^[[0;33;41m[ERROR vlan] registerVlanDeviceByName,410: Failed to REGISTER VLAN Device eth6.502^[[0m
May 5 01:05:05 kernel: ^[[0;33;41m[ERROR vlan] vlanIoctl ,634: Failed to create VLAN device (eth6, eth6.502)^[[0m
May 5 01:05:05 kernel: ^[[0;33;41m[ERROR vlan] registerVlanDeviceByName,410: Failed to REGISTER VLAN Device eth7.502^[[0m
May 5 01:05:05 kernel: ^[[0;33;41m[ERROR vlan] vlanIoctl ,634: Failed to create VLAN device (eth7, eth7.502)^[[0m
May 5 01:05:06 haveged: haveged starting up
May 5 01:05:07 wlceventd: wlceventd Start...
May 5 01:05:08 avahi-daemon[1926]: WARNING: No NSS support for mDNS detected, consider installing nss-mdns!
May 5 01:05:09 avahi-daemon[1926]: Alias name "RT-AX88U" successfully established.
May 5 01:05:09 lldpd[2490]: cannot get ethtool link information with GLINKSETTINGS (requires 4.9+): Operation not permitted
May 5 01:05:11 kernel: port_generic_open 536 skip turnning on power on eth0 here
May 5 01:05:11 kernel: SCSI subsystem initialized
May 5 01:05:11 kernel: tfat: loading out-of-tree module taints kernel.
May 5 01:05:11 init: fwver: 386.8_0

 

[bluepoint]

The router's(AX88U) speed test is slower by 200/100 Mbps when IPV6 is enabled but when tested through windows 11 ethernet it tested just fine. It tells me the router's speed test is broken.

 

[DJones]

The router's(AX88U) speed test is slower by 200/100 Mbps when IPV6 is enabled but when tested through windows 11 ethernet it tested just fine. It tells me the router's speed test is broken.

Each time you do a speed test it can pick a server at random that is either closer or farther away and might skew results. I also prefer to run Speedtest from ssh. If you install rtrmon from amtm it has Okala speed test for ssh built in to its monitoring, you just have to enable it in the apps setup config.

Windows has different tcp settings then a Linux system will and may also change results. I’d only run the test when their is no activity on the network and never over wifi unless your testing wifi.

 

[bluepoint]

Each time you do a speed test it can pick a server at random that is either closer or farther away and might skew results. I also prefer to run Speedtest from ssh. If you install rtrmon from amtm it has Okala speed test for ssh built in to its monitoring, you just have to enable it in the apps setup config.

Windows has different tcp settings then a Linux system will and may also change results. I’d only run the test when their is no activity on the network and never over wifi unless your testing wifi.

The router's speedtest let you choose the closest test servers manually and they work fine if the router is not in dual stack(IPv4 only). However, when IPv6 is enabled with IPv4 the speedtest is slower by 200Mbps. This is an apple to apple test that's why I tested the speed in some other way like window's connected through one of the router's ethernet port and proven that the connection's speed from my ISP and router is just fine even if dual stack is on. That's why I think the router's speed test is not working properly in dual stack mode.

 

[Treadler]

Skynet & uiDivstats are back!

 

[RMerlin]

Which code base did you get for AX88U?

As per the changelog:

  - NEW: Merged with GPL 388_20566 (only RT-AX88U and GT-AX11000)
  - NEW: Merged with GPL 388_21224 (only XT12 and GT-AXE16000)

 

[DJones]

Dirty upgrade to 388.1_alpha1-g8ea471fa9e from GT-AX11000_388.1_alpha1-gacf408e082

System log looks much cleaner, Skynet is working correctly no awk issue.

 

[shabbs]

Did a filthy load of the new Alpha's on my two RT-AX88U APs... no issues to report. Upgrade was as expected, all devices re-connected and happy.

 

[Intrepid2007]

Ok, fair enough..


Another thing I noticed is that in the VPN status screen, the 'Client status' statistics are the same for the each Wireguard connection (different connections though):
View attachment 44820

In my case, I am connected to 3 different Wireguard servers.


However, each Wireguard client screen shows different statistics:
View attachment 44821

Regards

I just installed the latest 3.88 alpha version (RT-AX88U_388.1_alpha1-g8ea471fa9e)..
Installation went smooth and everything is running without problems.

The issue above has been fixed in this version, thank you for that.
Each VPN connection now displays the correct statistics