[Release 384/NG] Asuswrt-Merlin 384.5 is now available

[zoggy]

installed 384.5 on my rt-88u today. 12hours later, 2.4ghz wireless stops working. turning 2.4ghz radio off and on makes it work again. problem has existed since 384.x :/

 

[JIPG]

I don't touch language files, because keeping them in sync with Asus would become a nightmare. I recommend you contact Asus about it.

OK. I will try to contact Asus.
Thank you anyway.

 

[JIPG]

@JIPG, aren't the language files in the WWW directory? What I had done for something else (although it was admittedly not a language file) was, on the post-mount script for my USB stick, I copied WWW over to the stick, created a symlink to the "new" WWW directory, then used SED to change a few things. I'd post a sample but I'm at work and don't have the files available to me. You can change anything you want that way.

It sounds interesting. I have modified the translation files in several devices as ebooks, but never in a router. The most direct way, changing these specific words with a hexadecimal editor seems "dangerous". I prefer to change words in the translation files.

Where is the www directory?. I have no idea about the files you mean i have to change.

 

[kfp]

Upgraded all the way from 380.70 on my AC3200. Total factory reset and manual reentry of settings, of course. Almost everything just worked.

The 3200 has two 5 GHZ "radios" and I use that facility. One "radio" has a separate SSID (for a discrete network) and the other shares its SSID with the 2.4 GHz band. I have heretofore let the devices work out which band they prefer on that SSID, successfully.

My Amazon Echo Show refused to reconnect at all and my 2011 mac Mini on OSX 10.13.4 now prefers the 2.4 band instead of the 5 which it used to.

I had to switch the Echo Show to the discrete 5 Ghz band--which worked. It just cant discriminate between the 2.4 and 5 like it used to.

I have only guesses as to why this behavior is subtly altered. Love to hear your thoughts, all.

There have been numerous reports of instability between the two bands with the 384.4 update, a quick forum search would show as such. I have experienced what you've described: device preferring the 2.4 band rather than the 5. Fortunately it is a PC in my case so I could set the network adapter to prefer a certain band and all is well again.

Others have been reporting a hard limit on number of 2.4 Ghz clients connecting simultaneously, although I don't think anyone truly what the problem is. Looking at the Github commits, I would guess this commit between 384.3 and 384.4 is the culprit, but these are all binary blob updates, it would be really hard to tell what changed without reverse engineering them.

Edit: here is the Asus release for that particular release

ASUS RT-AC3200 Firmware version 3.0.0.4.382.50010
Security fixed
- Fixed AiCloud XSS vulnerabilities
- Fixed XSS vulnerability. Thanks for Joaquim's contribution.
- Fixed LAN RCE vulnerability. An independent security researcher has reported this vulnerability to Beyond Security’s SecuriTeam Secure Disclosure program
- Fixed remote code execution vulnerability. Thanks to David Maciejak of Fortinet's FortiGuard Labs
- Fixed Smart Sync Stored XSS vulnerabilities. Thanks fo Guy Arazi's contribution.
- Fixed CVE-2018-5721 Stack-based buffer overflow.

Bug fixed
- Fixed CTF related issues
- Fixed AiCloud smart sync issue.
- Fixed client icon modification issue when client name includes special characters.
- Fixed AiCloud smart sync problem.

 

[Quoc Huynh]

Thank you RMerlin for your awesome work!

 

[hackztor]

There have been numerous reports of instability between the two bands with the 384.4 update, a quick forum search would show as such. I have experienced what you've described: device preferring the 2.4 band rather than the 5. Fortunately it is a PC in my case so I could set the network adapter to prefer a certain band and all is well again.

Others have been reporting a hard limit on number of 2.4 Ghz clients connecting simultaneously, although I don't think anyone truly what the problem is. Looking at the Github commits, I would guess this commit between 384.3 and 384.4 is the culprit, but these are all binary blob updates, it would be really hard to tell what changed without reverse engineering them.

Edit: here is the Asus release for that particular release

ASUS RT-AC3200 Firmware version 3.0.0.4.382.50010
Security fixed
- Fixed AiCloud XSS vulnerabilities
- Fixed XSS vulnerability. Thanks for Joaquim's contribution.
- Fixed LAN RCE vulnerability. An independent security researcher has reported this vulnerability to Beyond Security’s SecuriTeam Secure Disclosure program
- Fixed remote code execution vulnerability. Thanks to David Maciejak of Fortinet's FortiGuard Labs
- Fixed Smart Sync Stored XSS vulnerabilities. Thanks fo Guy Arazi's contribution.
- Fixed CVE-2018-5721 Stack-based buffer overflow.

Bug fixed
- Fixed CTF related issues
- Fixed AiCloud smart sync issue.
- Fixed client icon modification issue when client name includes special characters.
- Fixed AiCloud smart sync problem.

SOunds like it. I have same issue with 385. I was previously on 384_2 and did not have any issue. AC88u.

 

[Mister Craft]

Running smoothly thanks

 

[RMerlin]

System log full of error message as below

This is usually caused by SNMP.

 

[cmkelley]

It sounds interesting. I have modified the translation files in several devices as ebooks, but never in a router. The most direct way, changing these specific words with a hexadecimal editor seems "dangerous". I prefer to change words in the translation files.

Where is the www directory?. I have no idea about the files you mean i have to change.

Well, I'm not sure exactly where the language files are, I'm sort of assuming they're buried in the /www directory somewhere, you may have to dig a bit to figure out which file it is. Note that the following worked on an AC3200 running 380.69 with entware installed on a usb stick - your mileage may vary, at the very least, the fstab file and post-mount script will need to be changed if you don't have entware installed.

First, in the /jffs/configs/ directory you need a file named fstab (i.e. /jffs/configs/fstab):

/opt/www    /www    none    bind    0    0

Then you need a /jffs/scripts/post-mount script (assumes your usb drive is mounted on /tmp/mnt/usb - should be similar depending on your usb drive name):

#!/bin/sh
# runs after a partition gets mounted
# $1 contains path to mount

if [ "$1" = "/tmp/mnt/usb" ] ; then
  ln -nsf $1/entware /tmp/opt

### Copy firmware /www to usb & mount
  rm -rf /opt/www
  tar -cf - /www | tar -C /opt -xf -
  mount /www

### Modify webgui to add NTP Status page
  cp /opt/var/spool/ntp/Tools_NtpdStats.asp /www
  sed -i '/"Tools_OtherSettings.asp", tabName: "Other Settings"/a {url: "Tools_NtpdStats.asp", tabName: "NTP Daemon"},' /www/require/modules/menuTree.js
  service restart_httpd

fi

The script must be executable

chmod a+rx /jffs/scripts/*

If you don't have entware installed you'll have to fool around with the fstab file and the script to get it to work, because there won't be an entware directory on the USB stick.

The last section of the script which added the NTP Status Page you need to remove and add sed statements editing your language file. There is an excellent article in the Wiki on custom scripts: https://github.com/RMerl/asuswrt-merlin/wiki/User-scripts - actually there's a ton of useful information on the Wiki, well worth your reading if you haven't already. There's plenty of guides on the web if you're not familiar with sed.

Of course, if the language files aren't in the /www directory tree (again, I don't know where they live), then your modification job is a lot more involved.

 

[cmkelley]

I'm sure it was ASUS that made changes to the code that sealed the deal. I'm sure that it's the carrier that put pressure on ASUS to do this. There was a lot of people getting that router just for the purpose of changing it over. It was the best deal. I got mine for free and it was a great router. We just made it better by modding it. Now it's done because the carrier won't update the firmware and that means ASUS won't either. I'm stopped at 384.4_2 and to me that is fine as now it's just an AP behind the router. There still may be ways to update it but i'm sure its not without headaches so why bother.

Obviously it was ASUS that changed the code, but it's doubtful the carriers care - they'd simply refuse to support any unofficial firmware and wash their hands of it. It's almost certainly a licensing issue.

From my limited knowledge, some other posts here, and a bit of inference, I can surmise that ASUS buys licenses from the likes of TrendMicro, Broadcomm, etc. to distribute their code with ASUS' routers. Some of these modules are not shipped with the TM- units; they are cheaper partly because there are fewer licensing fees built into them. So the owners of those routers have not paid the license fee for those modules - just like if you bought a computer with Linux pre-installed, you haven't bought a license to use Windows. So I'm pretty sure ASUS put the code in there to prevent flashing custom firmware to protect itself against lawsuits from its licensors - by allowing the licensed code to be flashed onto the TM- routers which ASUS had not paid a licensing fee for, they would be distributing that licensed code illegally. The licenses almost certainly include language obligating ASUS to take reasonable steps to keep the licensed code from being flashed onto routers that ASUS didn't pay the license fee for.

 

[Huib]

Well, I'm not sure exactly where the language files are, I'm sort of assuming they're buried in the /www directory somewhere, you may have to dig a bit to figure out which file it is. Note that the following worked on an AC3200 running 380.69 with entware installed on a usb stick - your mileage may vary, at the very least, the fstab file and post-mount script will need to be changed if you don't have entware installed.

On my 68 I see a bunch of "*.dict" files in /www, which appear to be regular text files, one translation per line.

 

[Emma1]

Just updated to 384.5 on RT-AC88u
Everything is working smoothly - except the two Raspberry PI 3 on WiFi. Cannot connect, cannot scan / find SSID anymore.
Worked perfectly well 30min ago with 384.4_2. Any idea's?

 

[joegreat]

Worked perfectly well 30min ago with 384.4_2. Any idea's?

Turn off the router - not just reboot - after every firmware update!

 

[DonnyJohnny]

I have a question. Is there a way to gracefully shut down a router than doing a power off button? Command based?
Gracefully as in shut down all process, dismount usb etc

 

[Emma1]

Turn off the router - not just reboot - after every firmware update!

Yes, sure. It even said so to do, so I did.
In fact, I've rebooted the router several times already - like power off, power on.
All the other WiFi devices are connecting perfectly, even a 10 years old Logitech Squeezebox 2...

 

[Eyerex]

Just updated to 384.5 on RT-AC88u
Everything is working smoothly - except the two Raspberry PI 3 on WiFi. Cannot connect, cannot scan / find SSID anymore.
Worked perfectly well 30min ago with 384.4_2. Any idea's?

Just setup a 86U plus have a Raspberry Pi 3 Model B+ running Picoreplayer and everything is working ok over 5ghz WiFi

 

[JIPG]

Well, I'm not sure exactly where the language files are, I'm sort of assuming they're buried in the /www directory somewhere, you may have to dig a bit to figure out which file it is. Note that the following worked on an AC3200 running 380.69 with entware installed on a usb stick - your mileage may vary, at the very least, the fstab file and post-mount script will need to be changed if you don't have entware installed.

First, in the /jffs/configs/ directory you need a file named fstab (i.e. /jffs/configs/fstab):
...
The script must be executable

chmod a+rx /jffs/scripts/*

If you don't have entware installed you'll have to fool around with the fstab file and the script to get it to work, because there won't be an entware directory on the USB stick.

The last section of the script which added the NTP Status Page you need to remove and add sed statements editing your language file. There is an excellent article in the Wiki on custom scripts: https://github.com/RMerl/asuswrt-merlin/wiki/User-scripts - actually there's a ton of useful information on the Wiki, well worth your reading if you haven't already. There's plenty of guides on the web if you're not familiar with sed.

Of course, if the language files aren't in the /www directory tree (again, I don't know where they live), then your modification job is a lot more involved.

Thank you for the explanation.This is completely new for me, but I will take a look on it.

If I cannot change the translation, at least I will learn a lot of things.

 

[mikelees2]

Obviously it was ASUS that changed the code, but it's doubtful the carriers care - they'd simply refuse to support any unofficial firmware and wash their hands of it. It's almost certainly a licensing issue.

From my limited knowledge, some other posts here, and a bit of inference, I can surmise that ASUS buys licenses from the likes of TrendMicro, Broadcomm, etc. to distribute their code with ASUS' routers. Some of these modules are not shipped with the TM- units; they are cheaper partly because there are fewer licensing fees built into them. So the owners of those routers have not paid the license fee for those modules - just like if you bought a computer with Linux pre-installed, you haven't bought a license to use Windows. So I'm pretty sure ASUS put the code in there to prevent flashing custom firmware to protect itself against lawsuits from its licensors - by allowing the licensed code to be flashed onto the TM- routers which ASUS had not paid a licensing fee for, they would be distributing that licensed code illegally. The licenses almost certainly include language obligating ASUS to take reasonable steps to keep the licensed code from being flashed onto routers that ASUS didn't pay the license fee for.

I Agree that is has to do with licenses but remember the TM rebranded version has a custom proprietary firmware that is modified for optimizing the TM Wi-Fi calling feature. This is why they don't want you to change the firmware. They don't want to provide support to someone that changes the firmware. They also want the router returned if you leave and if its modified they can't resell it or put it back out to be used by someone else. I don't remember what software the router came with. I do know the UI looked the same.
Like I said I modified it because they didn't update the firmware. I didn't want to get caught using a router that was as vulnerable because it wasn't up to date.

 

[jeff3820]

Installed 384.5 on both 68U and 86U. Went smoothly on both and both seem to be working well!!

 

[DaveMishSr]

Upgraded my main router from 384.4_2 last night. No issues so far. Using 1.1.1.1 DNS & their IPv6 addresses too. Have Cox GigaBlast service so we race through the Internet. All seems to be hunky dory. My two repeaters will be upgraded to 384.5_0 later today.

Great work as always Eric!