25.12.2 wifi-scripts: Bug: PHY is undefined for device

Installing and Using OpenWrt
1

As I mentioned here, I have a WAX202 running 25.12.2. It appears to have went offline for about 20 minutes. After I reconnected, WiFi doesn't work and I have the following in the log, no other errors:

daemon.notice: netifd: radio0 (2178): wifi-scripts: Bug: PHY is undefined for device
daemon.notice: netifd: radio1 (2265): wifi-scripts: Bug: PHY is undefined for device

Oddly, reverting to even 24.10 didn’t help, (which was odd). I found this in the log:

[    0.660044] mt7621-pci 1e140000.pcie:   No bus range found for /pcie@1e140000, using [bus 00-ff]
[    0.668777] mt7621-pci 1e140000.pcie:      MEM 0x0060000000..0x006fffffff -> 0x0060000000
[    0.676905] mt7621-pci 1e140000.pcie:       IO 0x001e160000..0x001e16ffff -> 0x0000000000
[    1.019320] mt7621-pci 1e140000.pcie: pcie0 no card, disable it (RST & CLK)
[    1.026227] mt7621-pci 1e140000.pcie: pcie1 no card, disable it (RST & CLK)
[    1.033197] mt7621-pci 1e140000.pcie: pcie2 no card, disable it (RST & CLK)
[    1.040090] mt7621-pci 1e140000.pcie: nothing connected in virtual bridges

This is from a working WAX202:

[    0.652023] mt7621-pci 1e140000.pcie: host bridge /pcie@1e140000 ranges:
[    0.658721] mt7621-pci 1e140000.pcie:   No bus range found for /pcie@1e140000, using [bus 00-ff]
[    0.667451] mt7621-pci 1e140000.pcie:      MEM 0x0060000000..0x006fffffff -> 0x0060000000
[    0.675584] mt7621-pci 1e140000.pcie:       IO 0x001e160000..0x001e16ffff -> 0x0000000000
[    1.018426] mt7621-pci 1e140000.pcie: pcie2 no card, disable it (RST & CLK)
[    1.025336] mt7621-pci 1e140000.pcie: PCIE0 enabled
[    1.030223] mt7621-pci 1e140000.pcie: PCIE1 enabled

I found this previous issue by @f00b4r0:

It seems to state that the solution is a cold restart.

A full power cycle is required to restore normal operation, "reboot" is not enough. Also this doesn't seem to occur after a cold boot, suggesting the hardware may be left in some invalid state upon reboot. I wonder if it may be related to the firmware crashes workarounds on this platform as it seems to occur primarily when rebooting after a firmware crash. Can't confirm it occurs only in that case though.

I currently cannot perform a full power cycle, so in the meantime, I was seeking suggestions and ideas - maybe someone knows how to recover from this state?

Also, I'm curious what caused the crash into this PCI state?

2

I got the same error message on a Zyxel P-2812HNU-F1 with 25.12.2. Tried a cold boot, but it was no fix.

3

I'm sorry to hear that.

As an update - a cold boot worked for me.

  • removing the barrel power connector for 5-10 seconds
  • plugging the connector back in
1 Like
4

Made a workaround, added 'echo 1 > /sys/bus/pci/rescan' in /etc/rc.local :blush:

1 Like