Beta Asuswrt-Merlin 386.7 Beta is now available

[Treadler]

Help!
These CONSOLE entries are flooding the syslog.
View attachment 41914

Jun 17 01:01:08 kernel: CONSOLE: 103101.238 wl0: wlc_ampdu_recv_addba_resp: Failed. status 1 wsize 0 policy 1
Yesterday it got any devices connected out of the network.
Anyone know it?
First time those CONSOLE entries are show up, in 386.5_2 and previously were not show up.

I saw a lot of that with the recent Alpha, all fixed for me with the current Beta.
Reverting to latest stable as you said, also fixed it.

[Edit] The Console errors have returned, in the Beta.
No apparent effect on operation though.

 

[Damit1]

I saw a lot of that with the recent Alpha, all fixed for me with the current Beta.
Reverting to latest stable as you said, also fixed it.

Guess it is Beta's issues.
Hope @RMerlin will solve it in his next Beta release.

 

[RMerlin]

Guess it is Beta's issues.
Hope @RMerlin will solve it in his next Beta release.

It's not a beta issue, it's because Asus compiled a few models with debug logging enabled in the wireless driver.

I was able to obtain recompiled versions of at least some of these drivers, dunno if it covers all affected models (I don't feel like plugging and testing 16 different routers one by one...). So expect to have at least some of them with debugging disabled in beta 2.

 

[RMerlin]

I need at least one owner from each of the following models running Beta 1 to provide me with some info.

GT-AXE11000
RT-AX68U
RT-AC68U_V4

Please run the following command over SSH, and post the output:

cat /proc/mtd

I just want to double check which of these three models may also have been impacted by the JFFS-related issues, beside the RT-AX86U. That will save me some time from checking it myself.

 

[supe]

I need at least one owner from each of the following models running Beta 1 to provide me with some info.

GT-AX11000
RT-AX68U
RT-AC68U_V4

Please run the following command over SSH, and post the output:

cat /proc/mtd

I just want to double check which of these three models may also have been impacted by the JFFS-related issues, beside the RT-AX86U. That will save me some time from checking it myself.

from a AX68U:

ASUSWRT-Merlin RT-AX68U 386.7_beta1 Sat Jun 11 00:12:12 UTC 2022
admin@RT-AX68U-0C20:/tmp/home/root# cat /proc/mtd
dev: size erasesize name
mtd0: 05f00000 00020000 "rootfs"
mtd1: 05f00000 00020000 "rootfs_update"
mtd2: 00800000 00020000 "data"
mtd3: 00100000 00020000 "nvram"
mtd4: 05f00000 00020000 "image_update"
mtd5: 05f00000 00020000 "image"
mtd6: 10000000 00020000 "dummy1"
mtd7: 10000000 00020000 "dummy2"
mtd8: 03000000 00020000 "misc2"
mtd9: 00800000 00020000 "misc1"
mtd10: 05619000 0001f000 "rootfs_ubifs"
mtd11: 0001f000 0001f000 "METADATA"
mtd12: 0001f000 0001f000 "METADATACOPY"
mtd13: 00355439 0001f000 "filestruct_full.bin"
mtd14: 006c8000 0001f000 "data"
admin@RT-AX68U-0C20:/tmp/home/root#

 

[RMerlin]

from a AX68U:

Thank you - that confirms this model is affected.

 

[RMerlin]

Jun 15 05:04:09 inadyn[3012]: /etc/inadyn.conf:3: unexpected token '='

You enabled the DDNS service, but didn't configure it (might be an empty hostname).

 

[alchemy]

Updated AC86U to 386.7_beta1 from 386.5_2, and all's well. Thanks @RMerlin!

 

[emwgee]

I need at least one owner from each of the following models running Beta 1 to provide me with some info.

GT-AX11000
RT-AX68U
RT-AC68U_V4

Please run the following command over SSH, and post the output:

cat /proc/mtd

I just want to double check which of these three models may also have been impacted by the JFFS-related issues, beside the RT-AX86U. That will save me some time from checking it myself.

on a GT-AX11000


ASUSWRT-Merlin GT-AX11000 386.7_beta1 Sat Jun 11 00:14:32 UTC 2022
Flyer@Moonrocket:/tmp/home/root# cat /proc/mtd
dev: size erasesize name
mtd0: 051c0000 00020000 "rootfs"
mtd1: 051c0000 00020000 "rootfs_update"
mtd2: 00800000 00020000 "data"
mtd3: 00100000 00020000 "nvram"
mtd4: 05700000 00020000 "image"
mtd5: 05700000 00020000 "image_update"
mtd6: 00520000 00020000 "bootfs"
mtd7: 00520000 00020000 "bootfs_update"
mtd8: 00100000 00020000 "misc3"
mtd9: 03f00000 00020000 "misc2"
mtd10: 00800000 00020000 "misc1"
mtd11: 04d23000 0001f000 "rootfs_ubifs"

 

[RMerlin]

My bad - I needed the gt-axe11000, not the gt-ax11000.

 

[Damit1]

It's not a beta issue, it's because Asus compiled a few models with debug logging enabled in the wireless driver.

I was able to obtain recompiled versions of at least some of these drivers, dunno if it covers all affected models (I don't feel like plugging and testing 16 different routers one by one...). So expect to have at least some of them with debugging disabled in beta 2.

Hopefully I'm one of them.
Thank you
I own AX88U

 

[thelonelycoder]

At this point, I have gone through this exercise so many times I am fairly certain that I have tried at least once to go through the installation of Diversion (with Entware) or just Entware first, prior to making a swap file, yet it fails every time for me under any version of 386.7, even after performing a very thorough "clean" reset/format and re-install. The error that I see when going through this process with just the Entware install scrolls off the screen very fast, but is similar (but perhaps not identical) to this one when I try to install Diversion first: (this screenshot is from an early exercise, when I wasn't doing a true new installation, just a forced re-install)

I see where it "could" go wrong, we'll have to thest this. See my PM.

Out of curiosity, was your test done on a similarly reset RT-AC86U, also runnning 386.7 Beta 1? As a reminder, I have no problems doing any of these installs on exactly the same router, and same USB drive (Ugreen enclosure SSD), as long as I am using 386.5_2.

Yup, 386.7_beta1 and full router reset on my RT-AC86U.

 

[JIPG]

Asus set the max length of that variable to 15 chars instead of 39, probably because it was copy/pasted from the IPv4 variable.

I cannot change length validation, so I reported the issue upstream.

Hello RMerlin,

Can you guess which is the practical usage limitation of this problem? (e.g. no ipv6 dns sever possible, ...)

 

[greggy101]

I need at least one owner from each of the following models running Beta 1 to provide me with some info.

GT-AXE11000
RT-AX68U
RT-AC68U_V4

Please run the following command over SSH, and post the output:

cat /proc/mtd

I just want to double check which of these three models may also have been impacted by the JFFS-related issues, beside the RT-AX86U. That will save me some time from checking it myself.

ASUSWRT-Merlin GT-AXE11000 386.7_beta1 Sat Jun 11 00:12:12 UTC 2022
admin@GT-AXE11000:/tmp/home/root# cat /proc/mtd
dev: size erasesize name
mtd0: 06300000 00020000 "rootfs"
mtd1: 06300000 00020000 "rootfs_update"
mtd2: 00800000 00020000 "data"
mtd3: 00100000 00020000 "nvram"
mtd4: 06300000 00020000 "image"
mtd5: 06300000 00020000 "image_update"
mtd6: 10000000 00020000 "dummy1"
mtd7: 10000000 00020000 "dummy2"
mtd8: 00100000 00020000 "misc3"
mtd9: 02700000 00020000 "misc2"
mtd10: 00800000 00020000 "misc1"
mtd11: 059da000 0001f000 "rootfs_ubifs"
mtd12: 0001f000 0001f000 "METADATA"
mtd13: 0001f000 0001f000 "METADATACOPY"
mtd14: 0035546f 0001f000 "filestruct_full.bin"
mtd15: 006c8000 0001f000 "data"

 

[netware5]

192168.50.1 has always been the default for all ASUS routers, AFAIK

Not "always", just newer routers. The older models default to 192.168.1.1 and author of the OP didn't know that, which is normal. I also remember my shock when I discovered this change in ASUS defaults a few years ago

 

[shabbs]

Not "always", just newer routers. The older models default to 192.168.1.1 and author of the OP didn't know that, which is normal. I also remember my shock when I discovered this change in ASUS defaults a few years ago

Like the old days of admin/admin and you were never forced to change the password. Good times.

 

[Kabouter Plop]

You guys keep default ip range ? i just use 10.0.10.1

 

[RMerlin]

Hello RMerlin,

Can you guess which is the practical usage limitation of this problem? (e.g. no ipv6 dns sever possible, ...)

It prevents you from specifying your own IPv6 DNS advertised to your clients over DHCP. This is a new feature, so it doesn't change anything for existing usage, you just can't use this new feature just yet (unless your IPv6 DNS address happens to be shorter than 15 chars).

ASUSWRT-Merlin GT-AXE11000 386.7_beta1 Sat Jun 11 00:12:12 UTC 2022
admin@GT-AXE11000:/tmp/home/root# cat /proc/mtd

Thank you. That's interesting, this device is not affected, despite being compiled from the same SDK as the two other affected models.

 

[kernol]

..... The error that I see when going through this process with just the Entware install scrolls off the screen very fast, but is similar (but perhaps not identical) to this one when I try to install Diversion first: (this screenshot is from an early exercise, when I wasn't doing a true new installation, just a forced re-install)

I can't help with your specific issue - but note that @thelonelycoder is on it and may have spotted something.

What I would like to suggest is that you give MobaXterm [free version] a whirl - because you can always scroll up in its terminal window to see what messages may have been generated by errors etc ...
https://mobaxterm.mobatek.net/

I have used it and its many other features for years.

 

[thelonelycoder]

@RMerlin do you see a way to prevent memaccess to put its sh command later in the $PATH. See https://github.com/RMerl/asuswrt-merlin.ng/blob/master/release/src/router/memaccess/memaccess.c

This is clearly what's happening for @jsbeddow here: https://www.snbforums.com/threads/asuswrt-merlin-386-7-beta-is-now-available.79282/post-769255

Looks like I have to use absoulute paths like /bin/sh.