Asuswrt-Merlin 378.56_2 is now available

[ekms]

You have to upgrade in steps. Grab Merlins build 378.53 and flash that then you will be able to upgrade to the latest.

Worked perfectly, thank you!

One last question if anyone could assist with this one, is there anyway to renew my WAN IP? For some reason I can't load some websites but I can when I am on my phone's tethering.

 

[RMerlin]

@RMerlin - I believe you had fixed the long name issue. Could this still be the same issue? What is the max number of chars that can be in the name?

I fixed what I could retrace, but the whole networkmap code is a mess in need of a complete audit or rewrite, so it could still have buffer overruns left.

The limit is either 16 or 32, depending on which section of the code you are looking at (yes, it's that inconsistent).

You might simply have leftovers from a previous FW version. Try clearing the nvram setting.

nvram unset nmp_client_list
nvram commit

 

[alruf]

I fixed what I could retrace, but the whole networkmap code is a mess in need of a complete audit or rewrite, so it could still have buffer overruns left.

The limit is either 16 or 32, depending on which section of the code you are looking at (yes, it's that inconsistent).

You might simply have leftovers from a previous FW version. Try clearing the nvram setting.

nvram unset nmp_client_list
nvram commit

I reset the variable using the commands above but the problem persists. An IP scanner on my computer shows no name for the HIKVISION cameras, while one for my phone shows the serial number for the camera which is 36chrs long as its name. I finally got an answer from HIKVISION (wrote again today) and I am seeing if they can allow users to change the client name but even if they decide to do it, I am sure it will be a while.

 

[john9527]

I reset the variable using the commands above but the problem persists.

You might try making a /jffs/configs/dnsmasq.conf.add with the following lines in it, one for each camera, where xx:...and yy:... are the mac addresses for the cameras

dhcp-host=xx:xx:xx:xx:xx:xx,Camera1
dhcp-host=yy:yy:yy:yy:yy:yy,Camera2

Then with the gui closed

nvram set nmp_client_list=""
nvram set dhcp_staticlist=""
nvram set custom_clientlist=""
nvram commit
reboot

If you had any other manual DHCP entries, you'll need to set those up again.

 

[alruf]

You might try making a /jffs/configs/dnsmasq.conf.add with the following lines in it, one for each camera, where xx:...and yy:... are the mac addresses for the cameras

dhcp-host=xx:xx:xx:xx:xx:xx,Camera1
dhcp-host=yy:yy:yy:yy:yy:yy,Camera2

Then with the gui closed

nvram set nmp_client_list=""
nvram set dhcp_staticlist=""
nvram set custom_clientlist=""
nvram commit
reboot

If you had any other manual DHCP entries, you'll need to set those up again.

Not wanting to clear out my relatively extensive dhcp_staticlist (although I can restore it thanks to your utility), I only cleared the other two with the GUI closed. Once I rebooted the network map was no longer crashing. Looking up the variable in nvram over the past hour has shown a much smaller list of entries and the corrupted ones are not there. We'll see how long this lasts... hopefully it is a final fix however I now know where to look.

I really appreciate the time/help/patience you and @RMerlin have put into help me and others like me!

 

[withoutname]

Hi, after i updated my ac66u to 378.56, i'm facing a problem with wifi, on both 2,5 and 5GHz channels.
I have a QNAP (TS-251) plugged to a Gigabit port, and through the LAN, i can move files at over 80Mb/s.
Normally over WiFi i can reach only 11/12Mb/s with my Surface connected over 5GHz, due the limitations of the coverage. Actually, with 378.56_2 i can reach only 2,5Mb/s of transfer speed. The same thing is happening with other 2 laptop, and different phones (Nexus s, Nexus 4, Nexus 7(2012), Nexus 7(2013), Nexus 5, Nexus 6 (support ac speed but no speed increase), Xiaomi Mi4), connected to my wifi on both channels. No problems through LAN.
Any suggest?

 

[alruf]

@john9527 - The network map no longer works again. The nvram usage went down given I cleared the renaming of the clients, and nmp_client_list is now around 2.4k rather than >4k. I do not see any corruption in all the MAC/IP/client name lists and it appears that the Network Map has now picked up the custom names I used when assigning static IP addresses via DHCP. Everything looks normal to me except the network map page has crashed so I cannot access the GUI from the default main page.

This week end I will try your last suggestion in full. Could you let me know what I am actually doing by adding those 2 lines in the dnsmasq.conf.add ? Is it just a way to force a specific client name on a network tying it to a specific MAC? Right now it appears that the RT-AC3200 router picked up the DHCP names I entered (ie FrontDoorCam) as the custom_clientlist var which popped back up only has 3 devices and not the problematic ones that now show with their shorter name.

Olgiata_Router:/tmp/home/root# nvram show | awk '{print length(), $0 | "sort -n -r"}' | cut -d"=" -f 1 | head -n 20
size: 60560 bytes (4976 left)
2402 nmp_client_list
1331 dhcp_staticlist
624 sshd_dsskey
585 sshd_hostkey
540 vts_rulelist
384 rc_support
338 sshd_ecdsakey
323 webdav_smb_pc
163 custom_clientlist
120 qos_rulelist
109 wl2_acs_excl_chans
109 wl1_acs_excl_chans
92 2:pa5ga2
92 2:pa5ga1
92 2:pa5ga0
92 0:pa5ga2
92 0:pa5ga1
92 0:pa5ga0
85 asus_device_list
83 wollist

From my syslog - somewhen during this timeframe it stopped working again (I don't know how to interpret the httpd messages):

Dec 10 16:09:13 disk_monitor: Got SIGALRM...
Dec 10 16:28:12 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 10 16:28:12 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 10 17:44:35 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 10 17:44:38 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 10 17:44:40 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 10 17:44:43 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 10 17:52:37 rc_service: httpd 520:notify_rc start_lltdc
Dec 10 17:52:37 rc_service: httpd 520:notify_rc start_miniupnpc
Dec 10 17:52:37 rc_service: waitting "start_lltdc" via httpd ...
Dec 10 23:09:14 disk_monitor: Got SIGALRM...
Dec 11 00:09:14 disk_monitor: Got SIGALRM...
Dec 11 01:09:14 disk_monitor: Got SIGALRM...
Dec 11 06:40:23 rc_service: httpd 520:notify_rc start_webs_update
Dec 11 07:55:10 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 11 07:55:11 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 11 07:55:14 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 11 07:55:19 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 11 07:55:20 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 11 08:13:07 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 11 08:13:07 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 11 08:14:23 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118
Dec 11 08:15:12 dnsmasq-dhcp[514]: Ignoring domain hp.local for DHCP host name HPPC118

HPPC118 is why wife's work laptop in case it matters.

I also noticed that when the network map crashes, so does the bandwidth display under QOS. See image.

 

[uglyslate]

Just wanted to get clarification. Can I upgrade the AC66U from 378_54_2 to 56_2 directly and without wiping?

 

[Henk59]

Yesterday I updated the FW. AC66U 378.56 to 56_2 and activate for the very first time OpenVPN (as I typed this message),
it looks everthing works, even I switch On/Off the VPN and reconnect fine.
But I have 'important' question, at ipleak.net I'll see my LAN IP adress from my PC. Is this normal?
My ISP IP adress not visable, instead of this I'll see the VPN IP Adress.
Is this something to worry about.

Thank you, this is also my first message I'll dropped here.

 

[john9527]

@alruf.....sorry to heat things went south again.

Try to check the browser console again for any errors.

PS - I admit I included the dhcp_staticlist var just to make a double check that you didn't also have a client name with a character that could cause a problem. For safeties sake, they should be only alphanumeric and dash and underscore (also exclude any accented language characters if that applies). I'm speculating that some client connected that had a 'bad' name that broke things again.

 

[octopus]

If you not have turner off WebRTC. YES

 

[RMerlin]

Just wanted to get clarification. Can I upgrade the AC66U from 378_54_2 to 56_2 directly and without wiping?

In theory it should be fine, but be prepared to do a factory default reset if you do encounter issues. Sometimes it's not about the firmware version from which you are upgrading, but what previous versions you also ran before it. Therefore, it's not always possible to know for sure. For example, there were a few versions in the past that corrupted a certain nvram setting. That corrupted setting only started causing issues with newer firmware versions, as they were starting to rely on that setting to determine if your router had a 5 GHz radio.

 

[Henk59]

If you not have turner off WebRTC. YES

Thank you.
FireFox turned that on after update's.
(media.peerconnection.enabled set again to False)

 

[octopus]

Thank you.
FireFox turned that on after update's.
(media.peerconnection.enabled set again to False)

There is a adon to Firefox to toggele it on/off

 

[Henk59]

..Or
media.peerconnection.ice.default_address_only=true
Then there is no need to switch off/on.
Again thank you.

 

[pr0jects]

today I updated from 378.55 to 378.56_2 on the AC66-U
Factory Reset after update done - settings uploaded with the nvram-utility.

Router is working (at the moment) but it looks like the wireless mac-filter list is messed up
I can add entries, but I cannot delete the old / messed up entries (pictures).
I can connect with a device, which one is in the wireless mac-filter list .. so I assume, it´s still working, but I´m not able to test it, because all my devices are in the list.





any ideas how I can delete all entries from the wireless mac-filter list?
I would setup it again afterwards...

 

[john9527]

any ideas how I can delete all entries from the wireless mac-filter list?
I would setup it again afterwards...

Yes, sorry, they did something that changed the format of the maclist from .55 and before to .56 and later.

From telnet/ssh

nvram set wl_maclist=""
nvram set wl0_maclist=""
nvram set wl1_maclist=""
nvram set wl_maclist_x=""
nvram set wl0_maclist_x=""
nvram set wl1_maclist_x=""
nvram commit
reboot

I'll be taking a look at what needs to be changed in the utility over the weekend.

 

[pr0jects]

Yes, sorry, they did something that changed the format of the maclist from .55 and before to .56 and later.

From telnet/ssh

nvram set wl_maclist=""
nvram set wl0_maclist=""
nvram set wl1_maclist=""
nvram set wl_maclist_x=""
nvram set wl0_maclist_x=""
nvram set wl1_maclist_x=""
nvram commit
reboot

I'll be taking a look at what needs to be changed in the utility over the weekend.

ahhh oki - no problem

so this commands will delete all entries, right?

 

[john9527]

ahhh oki - no problem

so this commands will delete all entries, right?

Yes, that should wipe the maclist filters clean....if not, it means they REALLY changed something.

 

[pr0jects]

Yes, that should wipe the maclist filters clean....if not, it means they REALLY changed something.

thanks - the commands wiped alls entries from the maclist filters
Added all mac-addresses again and now everything is fine.

I found only another issue with the maclist.
the old mac-addresses were still visible on the guest-network page. As soon as I tried to delete an entry from this page, the router-login page was displayed.
In this case I have deactivated the guest-network and configured a new one - now I can add and delete mac-addresses from this page also.