[Release] Asuswrt-Merlin 380.64_2 is now available

[Campigenus]

I have a TM-AC1900 currently on 380.63_2. Nothing I did so far allowed me to update it to 64, and now 64_1.

Normal web update ... unsuccessful message.

Rescue mode update ... no errors but upon reboot it's still on the old firmware.

Asus restoration utility ... there is no router in rescue mode to be found, message. This is not right as I can see the rescue mode page in my browser at 192.168.1.1.

Yes I factory reset the router. I am normally connecting to it though a GBE switch, but the results are the same if I plug in to it directly.

Wits end ... reached.

Thanks in advance.

 

[reerden]

380.64_1 has been uploaded to both mirrors, and published to the update server. It contains security fixes from Asus' GPL 4180 (now that I managed to fix what those fixes broke on the webui), as well as additional security fixes that implemented.

Easiest way to update is to use the Check button on the firmware upgrade page - it will take you directly to the appropriate download folder on Mediafire.

Any factory resetting recommend because of all those horror stories about exploits? Or can I just flash and forget?

 

[netware5]

Any factory resetting recommend because of all those horror stories about exploits? Or can I just flash and forget?

I do believe that factory reset is not needed if you have not been under attack and if you upgrade from 380.64

 

[s4mpson]

@vincom

@s4mpson
weird, possible nvram/hw failure, it did get corrupted for no apparent reason, maybe a power issue . maybe run stock for longer time b4 restting and installing rmerlin, or....
check md5 of all files and flash all files via asus restore tool
plug router in different wall plug
set static ip for nic in pc then
if u can reset nvram twice via telnet or
flash rmerlin twice or
flash older rmerlin versions twice or
flash johns fw twice or
older asus fw twice then
reset nvram twice via telnet or gui then
flash .64 via asus restore tool

ts can be a pita

I was/am leaning towards hardware failure during this issue and RMAing this thing (really don't want to deal with that, but if it comes to that so be it), but it's been stable on asus stock .4180 running a minimal 1-2 client setup. It was on a battery plug on a decent Tripp-lite UPS along with a cisco 2800 voip unit and a Netgear FS28Tp switch, with all of them humming along, no power loss or fluctuation etc. I moved the router to my desk plugged directly into a wall outlet and only connected to my laptop via cat5e and the gateway via cat5e. So I doubt power issues, though I eliminated it the best I could. After flashing any Merlin fw after boot>factory reset> boot it immediately goes into busted mode, wifi LEDs & radios not turning on, CPUs drawing cute sine curves on the utilization graph, everything I described etc.

I'm going to jump back into playing with this thing after it running stable with stock for most of the week. The md5 checks out on the newest merlin 64.1.
What is the nvram command.. "mtd-erase2 nvram" ? So wipe it, turn it off for 10-20 seconds, boot it up and go through basic config -> telnet/ssh -> wipe it -> shut down again -> boot -> config? Is nvram reset via telnet/ssh that much more reliable? If I wipe it like that will the router boot to config like a factory reset?

My plan is gui reboot in stock -> wait 5 minutes -> pull plug -> 5 minutes -> boot -> gui factory reset -> pull plug/5 mins -> into recovery -> flash merlin 64.1 -> boot -> factory reset or boot to config (recovery = factory reset, right) -> config -> check format JFFS at next boot -> gui reboot -> wait 5 mins to "settle" -> gui reboot again -> let it sit and see if anything works ( I have strong doubts the second the LEDs don't work). If it works hallelujah! If not I'll scrub and reboot it a whole bunch and then pick a random earlier 380 merlin build and flash in recovery. If that fails, then back to stock, then to an older version of merlin. If that fails... well I have some 10 gauge shells ready to go.

 

[RMerlin]

Any factory resetting recommend because of all those horror stories about exploits? Or can I just flash and forget?

No need to do a reset, unless you know for a fact you have been compromised. So far, the only people who have been had the webui open to the whole WAN...

 

[MarkRH]

I noticed this a bit with 380.64 (RT-AC68U) but seems to always happen with 380.64_1. I made a shortcut to the daily monitor page: (router)/Main_TrafficMonitor_daily.asp and in the past when I went to it, it asked me to login and then went to that page. Now, if it asks me to login, it goes to the main network map page instead.

If I'm in the same browser session (so logged in) it does go to the daily monitor page. Is this by design now with some of recent security modifications or did something unintentional change?

Everything else seems fine. Thanks for all your work.

 

[vincom]

@vincom


I was/am leaning towards hardware failure during this issue and RMAing this thing (really don't want to deal with that, but if it comes to that so be it), but it's been stable on asus stock .4180 running a minimal 1-2 client setup. It was on a battery plug on a decent Tripp-lite UPS along with a cisco 2800 voip unit and a Netgear FS28Tp switch, with all of them humming along, no power loss or fluctuation etc. I moved the router to my desk plugged directly into a wall outlet and only connected to my laptop via cat5e and the gateway via cat5e. So I doubt power issues, though I eliminated it the best I could. After flashing any Merlin fw after boot>factory reset> boot it immediately goes into busted mode, wifi LEDs & radios not turning on, CPUs drawing cute sine curves on the utilization graph, everything I described etc.

I'm going to jump back into playing with this thing after it running stable with stock for most of the week. The md5 checks out on the newest merlin 64.1.
What is the nvram command.. "mtd-erase2 nvram" ? So wipe it, turn it off for 10-20 seconds, boot it up and go through basic config -> telnet/ssh -> wipe it -> shut down again -> boot -> config? Is nvram reset via telnet/ssh that much more reliable? If I wipe it like that will the router boot to config like a factory reset?

My plan is gui reboot in stock -> wait 5 minutes -> pull plug -> 5 minutes -> boot -> gui factory reset -> pull plug/5 mins -> into recovery -> flash merlin 64.1 -> boot -> factory reset or boot to config (recovery = factory reset, right) -> config -> check format JFFS at next boot -> gui reboot -> wait 5 mins to "settle" -> gui reboot again -> let it sit and see if anything works ( I have strong doubts the second the LEDs don't work). If it works hallelujah! If not I'll scrub and reboot it a whole bunch and then pick a random earlier 380 merlin build and flash in recovery. If that fails, then back to stock, then to an older version of merlin. If that fails... well I have some 10 gauge shells ready to go.

i dont think an rma will work if stock fw has no issues. like i said ts can be a pita, but when you figure it out its bliss.
since your having issues then yes telnet is the best option, the gui basically does what telnet does but, well, via gui, but yours is fracked up for some reason so telnet aleast bypasses the messed up fw in some way or at least removing the middle man, the browser.
you can also view whats being used by the cpu/ram via telnet but i dont remember the syntax/cli off hand, could be handy when router is acting weird. lots of telnet commands can be issued, just need members here to post them and google is your friend, basically linux commands.
theres also tftp, which i never got into understanding because i never needed it but could be similar to telnet, idk

 

[Gitsum]

Does this 364.1 contain the changes already made in the 365 alpha series?

 

[Morac]

I upgraded from 364 to 364.1 on my AC88U and everything appears to be working. The only odd thing I'm seeing is that on the System Log -> IPv6 page, host names are truncated to 3 characters. This was working fine in 364.

Hostnames on other pages display the full name.

Edit:

Found another issue:

The System Log -> Wireless page also stopped updating for 2 devices. I'm not sure why. One of them showed as not having an IP address and the other showed the wrong one. I turned them off and on which seems to have fixed that.

At this point the only thing I'm seeing wrong is all hostnames on the IPv6 log page (except for "<unknown>" are only showing the first 3 characters.

 

[netmik3]

I'm getting some weird mac cloning issues, every time I refresh the page, the mac changes. This may be happening only when my modem isn't responding on a mac change. Maybe aggressive is trying to flush the mac address? It seems to settle back on my setting when connected. Very odd.

Update: Yeah it's not even using my mac setting.

Like this old thread http://www.snbforums.com/threads/rt-n66u-strange-cloned-mac-address-presented-to-wan.13503/

 

[RMerlin]

Does this 364.1 contain the changes already made in the 365 alpha series?

It only contains a backport of the 4180 httpd code, and two additional security fixes. It does not contain any other 4180 code, neither OpenVPN 2.4.

 

[RMerlin]

confirmed

Known issue, already fixed locally (buffer size was incorrectly thought to be only 4 bytes large instead of 64).

 

[Barilla]

After i updated to the latest version my Sonos system "stopped working", it works for maybe 5-15 minutes. Then the music stops. When i check what units that are connected i see that one isnt connected anymore. I have 3 but i can only see 2. It changes wich one it is. I can get it to work if i only use 2, doesnt matter what combination. But as soon as i add the third, its stops playting within max 15 minutes usually 5-10.

Ive factory resetted my router (AC68U) and set ut up again, it doesnt change channels and i had the speakers on both dhcp and fixed IP.

Any tips on why this could be? Its really annoying.

 

[joegreat]

After i updated to the latest version my Sonos system "stopped working", ... Any tips on why this could be? Its really annoying.

Discussed and answered here - and further down in the thread.
Basically you need only to re-connect your Sonos devices...

 

[Barilla]

Discussed and answered here - and further down in the thread.
Basically you need only to re-connect your Sonos devices...

Thank you... Was it really solved though? The onther guy in the thread connected a Bridge and then it worked. Thats kinda how i "solved" it now, i connected one speaker with cabled net and now they all works via Sonosnet. But its not the preferred solution and it worked perfect with a older release. My clients all pop up on the net and works as long as i only use 2 and not all 3.
Kinda seems like another problem?

 

[nielsrune]

I use https over 8443 for local administration and just upgraded my AC66U from 380.63_2 to 380.64_1.

After the successful upgrade and reboot (from ssh), a new https-cert was created (valid from just then and 10 years forward), so I had to export the cert chain and put it on in trust list in Chrome. That is all expected behaviour.

But after a few changes in the DDNS section I rebooted the router from the webui button. Afterwards the cert was not recognized anymore. It turned out, the https cert served from the router was an old one valid from August 1 2015. Since, to my knowledge, a new cert is generated upon every firmware upgrade, this seems like a bug.

Anyone experiencing the same?

 

[Boy1979]

Hello,

make an update on AC66U from 380.64 to 380.64_1 without any problems!

Thanks in advance to RMerlin

 

[Rylan]

I tried downloading RT-AC68U_380.64_1.zip from all 3 download sources and get the same SHA256 signature, but that signature doesn't match what's published on the lostrealm downloads page.

9920793c75f7ee87fb65d2e8f7848c2c7eb91cb61d74c85e9bb9076db8c00531

 

[PapaBear]

Finally updated my AC3200 from 380.60_Beta (I know, couldn't be bothered to keep up to date before now, plus it was stable for my setup). Had to flash 380.61 to get 380.64_1 to take. With that, now I have a flashing alert that says my NVRAM usage is high. I checked and it's 65544 / 65536 bytes ... which confuses me. I've factory reset and redid the setup from scratch manually.

So two questions:

1. What is up with my router now and how to fix it (btw, it's performing fine. Speeds are normal and my 20 connected devices are humming along with no issues)?
2. Is there a way to have the router just do DHCP for wireless clients and allow my RasPi's handle the DHCP for wired devices and simply use the router as a gateway to the net?

Thank you to anyone who can assist!

 

[przemekwawa]

I believe there's something wrong with UI in the last couple releases. It is broken on my several macOS and iOS devices rinning Safari. Desktop Chrome works OK but desktop Safari and mobile Safari and Chrome on iPad display glitches all over the place, pages constantly reloaded after crashes and so on. Does anyone use Safari to work with web UI?
AC5300 if important.
What is also curious I cannot update the router from Safari. It uploads firmware file, indicates updating progress bar but after reboot the fw version stays the same. In Chrome though all works like a charm.

There is definitely something wrong with last 2 releases and Safari: using ios and Mac OS, very often cannot open page, need to reload it many times, and during browsing it hangs sometimes, displaying fragment of webui or sometimes fragments of code. After some time, Safari card with webui opened takes all memmory.
I can upgrade using Safari, but browsing is buggy.
Ah, details: RT-AC68U (router mode) and 2xRT-AC66U (ap mode), all have problems (it started probably 2 releases ago).