Asuswrt-Merlin 374.43 is out

[Builder71]

Jun 11 21:35:43 kernel: wl0.1: received packet with  own address as source address

Just tracked that one down myself....in my case it was a newly installed Intel 7260HW-AC mini PCI-e card. There is an advanced option 'Mixed mode protection' that is set to 'CTS-to-self enabled'.

I assume that is some feature on the NIC you use? (Not the router.)
Here I have a few Android phones and tablets using the guest WiFi network.
I have no idea where to look for this on the Android stuff.

 

[Builder71]

My guess would be that the closed source driver is outputting a cleaner signal as needed to the client in question. More (actual) Tx power is not always better. A better or cleaner signal is always better.

As the throughput tests show.

Yep, must be something like that.
Throughput is fine so I leave it alone.

 

[BenPope81]

Just checked, and HW acceleration is still shown as 'Enabled'.

I had this problem a few versions back. I had been playing with IPv6. Rebooting it would speed it up again... Until a Wi-Fi device connected. It was very strange.

I got it to work by resetting it and configuring from scratch.

 

[neil0311]

Just a suggestion: Maybe someone should open a separate thread for comcast issues? I'll butt out now.

I wasn't going to comment, but how can you call this a "Comcast Issue" when I can back rev to a previous version, and the problem goes away, and then I re-flash with .43 and the problem comes back....on cue and completely reproducible?

It may be a problem of integration between something to do with Comcast or their DHCP combined with a change of behavior in the firmware, but it's not a Comcast issue per se.

I'm grateful for Merlin's work, but just tired of tracking down these issues. The upgrades and fixes aren't worth the hassle.

 

[jim769]

I wasn't going to comment, but how can you call this a "Comcast Issue" when I can back rev to a previous version, and the problem goes away, and then I re-flash with .43 and the problem comes back....on cue and completely reproducible?

It may be a problem of integration between something to do with Comcast or their DHCP combined with a change of behavior in the firmware, but it's not a Comcast issue per se.

I'm grateful for Merlin's work, but just tired of tracking down these issues. The upgrades and fixes aren't worth the hassle.

That's all I was trying to say as well something has changed with 43 that affects v6 it is 100% reproducible. Like you i believe since every single time I install 43 v6 is broke every single time I revert back to an older version v6 works without a hitch it's not rocket science to determine that it's a issue with 43 at least in my environment. So I will stick with 42 because it just works.

 

[durd]

I seem to be having problems with upnp. Im using utorrent 1.8.4 with random port at start (and register with upnp ofc) and since 374.43 i havent been connectable. I rolled back to 374.42 and im suddenly connectable again.

Ive checked the Port Forwarding tab, my computer is the only one there, with the correct port, ip and in the VUPNP chain. And they disappear when i quit utorrent.

As i mentioned, everything works again in .42, no big configuration changes were made after i upgraded to .43 and i have not done a factory reset as that wasnt necessary for other users.

Im not quite sure how to log upnp as it usually just works. Let me know what info is needed (and maybe how) and ill find stuff.

 

[Bobcat00]

I upgraded my RT-N66U to 374.43 and my smart TV would continually drop its wireless connection, making it impossible to play videos streamed from my PC.

Went back to 374.35_4-sdk5 and all is well.

So I wasted a total of 4 hours trying to do upgrades to this router. I give up.

 

[RMerlin]

One thing that IPv6 users with issues can try is to disable the Comcast flood fix:

nvram set ipv6_neighsol_drop=0
nvram commit
reboot

See what happens after that.

 

[RMerlin]

I seem to be having problems with upnp. Im using utorrent 1.8.4 with random port at start (and register with upnp ofc) and since 374.43 i havent been connectable. I rolled back to 374.42 and im suddenly connectable again.

Ive checked the Port Forwarding tab, my computer is the only one there, with the correct port, ip and in the VUPNP chain. And they disappear when i quit utorrent.

As i mentioned, everything works again in .42, no big configuration changes were made after i upgraded to .43 and i have not done a factory reset as that wasnt necessary for other users.

Im not quite sure how to log upnp as it usually just works. Let me know what info is needed (and maybe how) and ill find stuff.

If the port does show under VUPNP them miniupnpd correctly forwarded that port. You could try disabling HW acceleration in case it would somehow conflict.

I have no problem here with uTorrent 3.4.1 (why are you using such an antique version BTW?) and a random port.

 

[RMerlin]

After upgrading from .42, Media Server Path setting had been wiped out, I had to reconfigure it. (upgraded with no reset)

And consistent with previous versions, if set ON, 'iTunes Server' (mt-daapd) is started with -m option, disabling its built-in mDNS function. Since Asuswrt-merlin's separate mDNS daemon does not start automatically for unknown reason, "iTunes Server" does not work out of box. This has been commented before. (A simple workaround is to install entware version of mt-daapd(firefly) and start it without -m option. Probably nobody is using iTunes Server(=mt-dappd=firefly) function though. But it's there and it's not working, so I report it anyway.)

AC66U

EDIT: The issue is that mDNSResponder is missing on the MIPS builds. ARM builds weren't affected as they use avahi.

 

[Trebuin]

RMerlin,
I keep hearing about comcast ipv6 issues. I just have mine disabled because I saw issue with it on my apple router (fixed now). What I want to know is this: my Asus is acting as a repeater with my apple router being connected to the modem. Will it have any of those issues with IPV6 if I were to re-enable them or will any issues be related to the Apple router?

BTW, all my previous issues were in relation to the 5GHz not reaching through walls despite the router saying it had a strong issue. Once I swapped, all firmware runs now. Thanks for all your help and great work you put into enhancing the Asus!

Respectfully,
Treb

 

[durd]

If the port does show under VUPNP them miniupnpd correctly forwarded that port. You could try disabling HW acceleration in case it would somehow conflict.

I have no problem here with uTorrent 3.4.1 (why are you using such an antique version BTW?) and a random port.

HW Acceleration is disabled because i use QoS (dns/dhcp/ntp - highest, http/https - high, ssh - medium).

I should of course have mentioned this in my first post, ive got a mac, and the latest version of uTorrent for mac is 1.8.4. Ive also got a n66u and ive got an ipv6 tunnel via HE, but the tracker isnt ipv6 enabled yet. Not sure why i didnt mention this, heck, i work in 3rd-line support

I rebooted the router of course, got a new IP on 2 of 4 reboots. Quit uTorrent nicely, waited 30min and started again with the same problem. Then rolled back to .42 and all is well. I didnt factory reset as i mentioned earlier.
Anything else i can try/log?

 

[ualdrivr]

Asuswrt-Merlin 374.43 is now available for RT-N16/RT-N66/RT-AC66.
EDIT:
Asuswrt-Merlin 374.43_2 is now available for the RT-AC68U
EDIT2:
AC56U build is now also available.


The focus of this release was fixing bugs and polishing things here and there.

• Traffic Monitoring for PPPoE users has been fixed, and should now report correct figures.
• A couple of webui-related issues were resolved, such as the problems with the new Media Server page, SSIDs with single quotes in their names breaking the webui, and a few others.
• dnsmasq now returns an empty string when queried for a WPAD proxy URL, which resolves log noise for broken clients such as Windows 7. Note that if you were using a custom dnsmasq.conf to provide a valid WPAD URL, you will need to change your config to use a Postconf script instead, to modify the default 252 option entry in dnsmasq.conf.
• A few component updates - miniupnpd and openssl were both updated to their latest versions. Additionally, a few additional ASM patches were backported from openssl 1.0.2, improving SHA performance for ARM-based routers.
• Guest clients were able to acess LAN devices when they weren't supposed to
• And a few other fixes, listed in the changelog.
• One new feature did trickle down into this release - you can now force a DDNS refresh after a configurable number of days.

Changelog is here.
Downloads are here.
More details can be found in the complete changelog.

MD5 checksums:

6c0f3e96c30896b06de74e47a22b2338  RT-AC66U_3.0.0.4_374.43_0.trx
c950755cdbc5f8d1078d2b8d4f105ab2  RT-N16_3.0.0.4_374.43_0.trx
78ebdd8b4c48fcfc7bef63111ad7ec4e  RT-N66U_3.0.0.4_374.43_0.trx

aa8091c9ddeb2f74e3f8a61e3fd12e18  RT-AC68U_3.0.0.4_374.43_2.trx
f357623534a2fa4ad9d4d4195a3139c3  RT-AC56U_3.0.0.4_374.43_2.trx

Thanks Merlin. Well after 49 days without any issues with the .5656 Asus Firmware, I took the plunge today and upgraded to 374.43_2 for my 68U. So far so good. Booted smoothly and all working so far 20 minutes into it.

 

[zeus2machina]

zeus2machina - I was able to solve this by giving the IP address of the router as the default gateway for the AP. Time was picked up right away.

Thanks for the info, I will give it a try when I have a chance
Cheers,

 

[RocketJSquirrel]

People rarely post to say Comcast IPv6 works great. Not to brag (OK, maybe a little), but I have had perfect native v6 via Comcast for almost 2 years on my RT-N66U regardless of firmware. It's been as perfect with all Merlin releases as it was with Asus stock.

The blob called Comcast is a mish-mash of cable systems cobbled together. Kind of like Forrest Gump's box of chocolates, you never know what you're gonna get.

 

[Crazy]

Back on page 11 I stated I had a problem with my usb port no detecting my usb flash drive and I've thought I fixed it when I restored saved config. Anyways it died again on .43 after one day and I reset everything and still can't get the usb to read any usb drives, even on factory setting. Ive looked at the logs on the day it stopped reading usb drive and I found no error or usb disconnect. So I went back to .42 and use my save config and everything is fine. Now been almost 3 days and I've been fine. I'm just wondering if anyone have the same problem with the AC66U. Or something I can do for testing. Thanks

 

[4Leaf]

This may be a stupid question, but what are the md5sum codes for? Am I supposed to use it for something?

 

[Trebuin]

you can verify the files have not been modified or corrupted from the downloaded source using software that generates those codes.

 

[RMerlin]

HW Acceleration is disabled because i use QoS (dns/dhcp/ntp - highest, http/https - high, ssh - medium).

I should of course have mentioned this in my first post, ive got a mac, and the latest version of uTorrent for mac is 1.8.4. Ive also got a n66u and ive got an ipv6 tunnel via HE, but the tracker isnt ipv6 enabled yet. Not sure why i didnt mention this, heck, i work in 3rd-line support

I rebooted the router of course, got a new IP on 2 of 4 reboots. Quit uTorrent nicely, waited 30min and started again with the same problem. Then rolled back to .42 and all is well. I didnt factory reset as i mentioned earlier.
Anything else i can try/log?

Check the verbose output from iptables over SSH:

iptables -t nat -L -v

See if the first column shows a packet count other than zero after a while. If it does, it means you did get traffic on the port, and it got forwarded to the specified IP.

I would also test without the HE tunnel up. OS X uses PCP for port forwarding (not sure if uTorrent supports this however), which might be having trouble if there's an IPv6 available, tho you should be able to see with the iptables output if the port did get corrected forwarded or not.

 

[cmillar6]

There's an issue with 374.43_2 when transferring larger files to a usb drive attached to my AC68u. Windows just seems to hang at 0% transferred (calculating) and eventually shows a network error that it can't connect to the drive. However, when I click try again the file has actually transferred successfully and Windows offers the option to overwrite it. I've reproduced this with multiple usb drives on both the usb 2 and 3 ports. Transferring smaller files like word documents or pdf files works just fine. Anybody else experiencing this?


Nevermind, I fixed it. I just formatted the thumbdrive to ext4 and it works great now.