Thanks for the advice! That solved the problem in the and.
It turned out that the option wps_pushbutton '1' line is the culprit. After deleting that line in both interfaces wifi now works as expected.
Somehow that line got into the wireless configuration file in the pre v22.03 snapshot.
Closing the issue.
i'm use wps button on 22.03 do you think may create a problem ? for the moment my wifi is solid like the roch
Anybody else getting no channel noise values from iw dev wlan1 survey dump off the mt7915 radio?
Seems to have last been mentioned here in 2021:
Here's the survey results from the snapshot a few days back:
Survey data from wlan1
frequency: 5500 MHz [in use]
channel active time: 38042376 ms
channel busy time: 444871 ms
channel receive time: 206400 ms
channel BSS receive time: 34374 ms
channel transmit time: 216508 ms
Looks like the survey is implemented in mt76, (and was working when that functionality was added?) but after inserting a few printks it seems like the register values read out in mt7915_phy_get_nf() are all zeroes.
Casually looking through git doesn't reveal anything interesting, but the same code path was recently fixed for mt7916, so it should have been working for that radio - but somehow broken for the 7915 variant.
Regarding the boot loop you are observing: It reboots in the place where U-Boot should be starting, eventhough apparently BL31 loaded and worked (and that's in the same FIP structure as U-Boot). Maybe the flash is weirdly corrupted or some other rare hardware oddity? To know more you will have to wire up JTAG
I don't see any signal values when entering that command:
root@OpenWrt:~# iw dev wlan1 survey dump
Survey data from wlan1
frequency: 5180 MHz [in use]
channel active time: 599863390 ms
channel busy time: 3713363 ms
channel receive time: 7925474 ms
channel BSS receive time: 1131285 ms
channel transmit time: 5437944 ms
Survey data from wlan1
frequency: 5200 MHz
channel active time: 120 ms
channel busy time: 0 ms
channel receive time: 0 ms
channel BSS receive time: 0 ms
channel transmit time: 0 ms
Can I upgrade to the latest snapshot build without any issues? I don't have any firewall rules or vlans currently configured but I have two bridged interfaces. Also looking to bond these interfaces to a L2 switch. Appreciate if anyone could let me know if I can upgrade this safely to the latest snapshot.
I have a chance to buy used raspberry pi 4 in my place not particulary for this purpose, but it seems that it's possible to use this board as JTAG debugger:
Is this a viable option for this router or RPi GPIO interface is no go for this kind of mission and I need that FT2232-like board?
And if so my next steps will be just flashing recovery image via JTAG or I can somehow (I am a newbie in this JTAG\UART thing , just connecting pins here and there avoiding 5V pins ) examine a source of issues from my first reflash attempt?
As OpenOCD supports running on the Raspberry Pi this should work well. Both the Raspberry and the MT7622 SoC operate on 3.3V level, so connecting directly as described in the link you posted should be fine. I wonder if you will need pull-up/pull-down resistors for some signals to be fast enough, but maybe RasbPi's pinconf can also already do that for you in the SoC as in that link posted there isn't any mention of that.
Afaik rc5 is currently being build and I don't see Mediatek yet.
If rc5 is the latest release candidate before the new release depends on the amount and severness of the issues that will come out of rc5.
Don't know if anyone here has run into this. I bought an RT3200. Started with the installer instructions. everything worked, but due to some issues went back to my old router. With the release of RC5 decided to try again. Reflashed the device but noticed no change to the kernel version. A little investigation showed that the installed version was the initial flash from dangowrt. In addition no configuration file changes are preserved after reboot.