[Fork] FreshTomato-MIPS - development discussion only - for support always open your own thread


FL0WL0W

Network Noob
Member
Update... :( wl driver still appears to be broken in 2020.3. The driver works correctly in the asus firmware and also in the asuswrt-merlin firmware, but they don't have vlan support. is there anyway we can take the assuswrt driver and use it?

P.S. not making a thread because i already know the issue. looking for some discussion on how we can possibly fix the driver in the firmware. expect more, bigger donations ;)
 

pedro311

Addicted to LI
Member
You still haven't answered a few basic questions:
- was it a clean installation (cleaning nvram)?
- what exact version did you upload into your (I presume) RT-N66U (file name)?
- did you try the image from the other branch then?
- what does it mean "broken"? What is the symptom? How to reproduce it, step by step?
 

FL0WL0W

Network Noob
Member
You still haven't answered a few basic questions:
- was it a clean installation (cleaning nvram)?
- what exact version did you upload into your (I presume) RT-N66U (file name)?
- did you try the image from the other branch then?
- what does it mean "broken"? What is the symptom? How to reproduce it, step by step?
-clean install
-freshtomato-K26USB_RT-N5x-MIPSR2-2020.3-AIO-64K
-not sure what you mean from the other branch. i have tried both the stock asus firmware and the asuswrt-merlin and neither of them had any issues over ~6 months.
-Symptoms are the same as they were before. clients not dissasociating from the assoclist when disconnected. this issue seems to only occur after a couple hours of being on. which is why i mistakenly thought it was fixed

would you like me to try freshtomato-RT-N66U_RT-AC6x-2020.3-AIO-64K?

See page 80
This a longtime known bug with the entire MIPS branch of Tomato. The bug appears to exist in the wireless driver itself, so it is unlikely to be fixed as it is a pre-compiled binary.
page 81
Driver 5.110.27.20012 is also used in v1.28.0511 MIPSR2Toastman-RT-N K26 Std. However, the bug is not present.
Edit; the bug may be multiwan induced. Because, that was the difference between the items tested.
 

pedro311

Addicted to LI
Member
would you like me to try freshtomato-RT-N66U_RT-AC6x-2020.3-AIO-64K?
Exactly

-Symptoms are the same as they were before. clients not dissasociating from the assoclist when disconnected. this issue seems to only occur after a couple hours of being on. which is why i mistakenly thought it was fixed
So what script did you use to check it, cause I presume that Device List page doesn't control the gate opening itself?
 

Techie007

Addicted to LI
Member
-freshtomato-K26USB_RT-N5x-MIPSR2-2020.3-AIO-64K
<snip>
would you like me to try freshtomato-RT-N66U_RT-AC6x-2020.3-AIO-64K?
Exactly
If we've got two firmware branches that can both run a router, one with newer code and one with older code, we should eliminate the older one to reduce confusion.

-Symptoms are the same as they were before. clients not dissasociating from the assoclist when disconnected. this issue seems to only occur after a couple hours of being on. which is why i mistakenly thought it was fixed
If you just want it to work, I would strongly suggest upgrading to a Tenda AC15 router. If you don't mind shopping around and waiting a bit, they can often be had on eBay for $30-40.
 

FL0WL0W

Network Noob
Member
If we've got two firmware branches that can both run a r

If you just want it to work, I would strongly suggest upgrading to a Tenda AC15 router. If you don't mind shopping around and waiting a bit, they can often be had on eBay for $30-40.
I would like to figure out how to update the wl driver to the same one used on the stock asus firmware or asuswrt-merlin as that one did work but lacked features that tomato offers like vlans and virtual wireless. I have 3 routers total so i would have to get 3 Tenda AC15s. while im not completely opposed to that, it does lack some features i like with my n66us. like detachable antennas. i have a big yagi antenna at my gate aimed down my road so that my phone will pick up the ssid beacon and start trying to connect and be connected by the time i reach my gate so i don't have to sit there at the gate waiting for it to connect. + i would rather donate to the project then spend money on new routers.

upgraded the firmware and should know in a few hours whether its any different

So what script did you use to check it, cause I presume that Device List page doesn't control the gate opening itself?
They don't dissasociate on the device list page, the wl assoclist command, or in my script. So it shouldn't matter what script i am using, but because i like to talk about it, here goes.
I made a python script that would telnet into all of the routers and run wl assoclist, but that was for stock asus and the python script ran on one of the routers running assuswrt that would log everything in sqlite and serve up a webpage for configuration.
Now i have moved over to node-red and use the node-red dashboard for configuration. to get the device list now I call the update.cgi with exec=devlist&_http_id=TIDxxx with the proper authentication and parse the JSON. It would be nice if tomato just connected directly to MQTT that way i would not have to store the routers username and password in node red. Node-Red allows me to more easily integrate more automation, like turning my carport light on if i am arriving home and its night time. see attached files for some screenshots of the node-red flows and configuration pages. The actual gate has an ESP8266 (wifi chip) with MQTT on it that i modded into the gate controller to be able to check whether the gate is open or closed and be able to open or close the gate. carport light has a sonoff switch to turn it on and off. And my next endeavor is automatic door locks
 

Attachments

FL0WL0W

Network Noob
Member
Update. It appears to be working for the moment. I have now updated all of my routers, and i will come back in a few days to confirm it has been working.
It's not that simple: in RT-N you have a working bcm_nat module in RT-AC, you don't have it.
Are there any other differences i should take note of? Since these are only access points, NAT is not an issue. For others, could the RT-AC wl driver be used on the RT-N firmware, or does that driver also operate the bcm_nat module?

For now i have donated another 10 bucks. Thank you everyone for your help.
 

kernel-panic69

Networkin' Nut
Member
RT-N is v5.x driver with bcm_fastnat, RT-AC is v6.x driver without it, if I am not completely losing my mind (?). Because 802.11ac devices require a different driver. @pedro311: does the module not support the newer driver or is it that it requires the ctf module, but there is no MIPS binary blob for it?
 

duren11

Reformed Router
Member
Hi, I just upgraded to 2020.1 (freshtomato-RT-N66U_RT-AC6x-2020.1-AIO-64K.trx) trying the AC6x version this time and restored my config from the MIPSR2 version (it has 3 tagged VLANs on br1/2/3 on a single port) and untagged trafiic across all ports. The minute I restore my configure I lose complete access to it on my untagged LAN. Normally, an upgrade from the MIPSR2 is simply a restore of config and I'm good to go.

To troubleshoot, I figure to try to rebuild by config my hand and I noticed the minute that I configure a VLAN (basically check Tagged to "On"), I am unable to ping the router anymore on the untagged LAN.

Went back to MIPSR2 and no problem...
Maybe it is possible to figure out what happend to RT-AC?

Fortunately, there are RT-N versions for n66u: https://exotic.se/freshtomato-mips/v2020/2020.1/K26RT-N/Asus RT-N66u 64k/
Ran into this problem, not sure why it happens on the K26AC version, temporary fix tracked here.
 

kylejustknows

Networkin' Nut
Member
A strange bug:

Background:
rt-n66u+ 2020.3 (also tested in 2020.2)+ PPPoE connect to fibre modem.

Symptom:
Packet loss occasionally, MTR result showing micro-disconnection between my router <-> first ISP hop.
Drop a packet every 5~10 minutes on WAN side. (unnoticeable if I wasn't gaming)

My tests:
Complaint to ISP, ISP checked (they remotely test everything to the fibre modem), saying maybe something funny on my home router.
I turn on/off WAN related router features (including IPv6, fastnat), the problem persists.
I erase the router NVRAM, then ONLY setup PPPoE, and the problem is still there.
(No error in router's log. Tested using both wireless and wired, same result. The packet loss is not on the LAN side.)

I connect my COMPUTER direct to fibre, using Windows-build-in PPPoE, it worked fine! No packet loss!
At this stage, I assume something is broken in my router's hardware.

Temporary solution:
I randomly turn on the "Bandwidth Limiter" function and input a dummy rule there. (the rule doesn't do anything)
And the problem is suddenly gone!

Conclusion:
I don't know why. There is nothing useful in the system log.
Maybe something related to WAN/PPPoE/NAT internally freezes/crashes, and the "Bandwidth Limiter" function can bypass it.
 

Top