I will try and upload a build by tomorrow .. His pull request has some changes suggested (especially the kmod-mt7603 driver) which I will try and fix and build using the latest main branch. I believe he has added some code for headers (I tried to upload a build without those headers from his previous pull request but it failed naturally) so ideally this should work based on the steps. MT7615E is the driver used here for 5GHz although the chip has mt7613e since that chip is apparently very similar with minor changes. The Archer A6 and some other similar model from tp-link uses the same chips but they seem to be stuck somewhere. lets see how this one works out.
Try this one. I haven't tested it just compiled it but its using the latest pull request from aashishkul so it should work and it comes with Luci so it should basically give you a fully functional router with wifi. Will test it in a couple of days can't have network downtime.
So, I have got this installed.
I see MT7613BE driver being used, but 5G band is definitely not working. Tried changing the channels here and there, but no use.
[UPDATE: 2G band is working as expected. Haven't tested WAN or internet connectivity yet.]
According to this post Mediatek MT7613E 5Ghz Wifi Driver the AP driver is missing in my build. I will recompile and upload a build. Just do a system upgrade and it should work. Check my repository in a few hours and let me know if it works.
That seems to have resolved the issue with 5Ghz band. Thanks!
I see one issue still persisting(this was there in initial build too):
Rebooting from LUCI or from Terminal(with "reboot now") seems to hang the router in some state. The Power LED doesn't glow, wired network is not accessible, WiFi SSIDs don't show up. Manually switching off and then switching on is the only option at this stage.
I am not sure how I can pull logs since the router goes into an inaccessible state. Is there a way I can pull logs from previous boot cycle?
Hmm... interesting. When you get time maybe switch to stock and retry the factory image. I will try and clean compile a new build today since you said this build works to see if there was any issue compiling. Otherwise how is the stability in terms of Wireless on both bands since you said both bands work fine? I plan on flashing mine (primary wireless router 100/100 connection) today so probably will know better.
This is my secondary router, and I haven't connected this to internet yet. I plan to eventually use this as a mesh-point(attempting to do this already).
That said, wlan connectivity on both bands seems to be stable at this point. I have been on continuous SSH sessions via my phone, and there were no hiccups. One point of note though: With country-code TW(Taiwan), 2G radio offers TX power up till 24dBm, but 5G radio is limited to 21dBm. Is this disparity normal? Or, is this something related to the driver choice for 5G radio?
Router seems to work great, except the reboot issue.
21 and 24dBm seems to be the limits? Afaik in IN the limits are 20dBm on any equipment if you set the country code to IN but these seem to be normal limits for such APs. That being said its too high but it really depends on your testing (In India the walls are brick walls so its hard to get any network with low transmit power sometimes). Also I don't know how accurate the transmit power is in the first place when it comes to the driver. Not that the driver is an issue but sometimes the chip just might not allow you to set a transmit power with third party firmware so I usually set it driver defaults and let the driver decide what's best. Though the mt76 driver is quite mature so its possible the transmit power is actually accurate in which case it seems like a great deal since the Archer A6 from Tplink has the same hardware but the transmit power on the 5G band had some bug and stuck on 3dBm (not sure its fixed now).
As far as the reboot bug is concerned I will know more once I have a build running but seems odd. Once snapshots are released for this (I believe the pull request needs to be accepted for that) then it would be best to run snapshots (or build based on the snapshots) so that such bugs can be tracked more easily and they don't depend on individual builds.
I seem to have bricked my router, while trying to reflash with factory image.
This is what I have done:
Reset 3 times to boot into Linksys stock image(something about a-b mode of this device)
Flashed stock image twice, assuming it will boot to alt_image
Flashed OpenWRT factory image(your 2nd build - with 5G AP driver included)
Here's what is happening now:
Power LED starts blinking.
Wired LAN1 blinks for 2-3 seconds(it is conneced to a laptop), and power LED stops blinking.
Laptop shows "ethernet" connected for these few seconds.
Cycle repeats.
Holding the reset button pressed doesn't seem to make any difference - probably because the entire cycle finishes in less time than it is required to recognize a reset.
I am thinking I might be able to recover it with TFTP. Any suggestions are welcome.
I just flashed my router (second update with the wifi driver) and it works fine actually I also tested the reboot and I seem to have no issues. I basically followed aashishkul’s method. When your router boots up it boots from A not B I believe. So when you flash stock you need to cold boot the router (also factory reset always when flashing openwrt and do a cold boot). Once that is done then you flash the stock image and boot from B. However you need to make sure that you have flashed both partitions with the same image (which I don’t believe is the case). I don’t know how to flash from tftp for linksys (I would suggest try resetting the router once and see if it boots back into openwrt since partition A should still be Openwrt).
I only flashed it few minutes ago and setup things (it actually works better than I expected so far) but as far as the blinking goes it is weird and takes time to boot so maybe reset the router (using the button) and wait for some time to see if it boots into some image.
Is there any reason why the USB3 kernel module is included? I built using your pull request (excellent work btw) and its running fine and I don't see anywhere USB3 being used other than the fact that a port is detected! (a physical one is not present) I first thought if something was connected to it via USB but its not. Also for the E8450/RT3200 there seems to be a flash layout replacement (probably to free up more space) to UBI here https://github.com/dangowrt/linksys-e8450-openwrt-installer . Is it possible to build this for this model as well? Backing up the flash might be a good idea obviously since this process is likely irreversible and will definitely brick the router if it goes wrong. Let me know, I am looking at the code to see if it can be compatible with E5600 but so far I am impressed with the hardware for the price. I hope this one is mainlined soon so that I can rely on regular snapshots instead of building my own and breaking things.
Not sure what went wrong with mine. After multiple attempts at resetting, trying to get into "management mode", trying the 1-minute reset button hold, etc - I still have no success.
The router sometimes stays on(when I try the 3-time on-off cycle, with the ethernet connected to laptop). At this point, the power button doesn't glow anymore, but the ethernet port responds to traffic. Router is not responsive to pings.
Apparently, TFTP advertisement technique doesn't work with Linksys routers. They have a different TFTP-utility based approach, and the router needs to respond to pings before we can try TFTP utility
At this point, I have 3 options:
Wait for somebody else to give me a better approach to try.
Call this a defective piece and utilize my return-window.
Use this opportunity to open up the router and try flashing via Serial - But I need instructions.
After doing build you need to flash firmware built with extension factory.bin. there is no tftp unless you do a highly invasive dissembly and get access to on board serial port. In case everything else fails, start the router wait for 15 seconds and start again wait for 15 seconds start again wait for 15 seconds start again and let the router boot. At this point, router will start in alt_firmware. Flash firmware taken from Linksys website. The router will be back to life again.
MT7621 has USB. However Linksys has decided not to provide it to user because then there will hardly be any difference between this router and much expense routers from Linksys. So there is nothing we can do about it. The Wifi are connected to PCIE interface.
Are you saying this router is capable of pulling firmware from Linksys website, and we should make sure it is able to connect to the internet?
I have a feeling I have flashed into alt_firmware - is this possible?
At the moment, the router doesn't even wait 15 seconds. It starts up, power LED and LAN LED(if connected to laptop) blink for 2-3 seconds, and it cycles. I could try your procedure keeping a gap of 3 seconds instead. Is it worth trying?
[EDIT]
I think you meant "flash firmware taken from Linksys website, after the router boots to alt_firmware".
When I do the power on/off for 3 times, the router ends up in a state where the power LED doesn't glow, but the network LEDs flicker based on them being connected/not. At this point, the router doesn't respond to pings, and the web GUI doesn't open up as well(even though the laptop has a static IP(192.168.1.1) setup.
So, I have got a replacement - assuming my router itself was faulty, since it doesn't boot to even alt_firmware successfully.
With my second attempt at using OpenWRT on the E5600, I can confirm the reboot bug exists.
Steps to reproduce:
Open LUCI.
Navigate to: System>Reboot
Click on the Reboot button.
Expected behavior: Router should restart and boot up successfully.
Actual behavior: Router tries to reboot. All LEDs stay off, router is unresponsive, LUCI is not accessible via LAN, WiFi SSIDs don't show up.
So, I have discovered that this behavior occurs only when the WAN port of the router is disconnected. Reboot works as expected when WAN is connected to the internet via my primary router.
