Open weedy opened 5 years ago
I'm not sure what's going on here. I'm guessing, but I think the device powers down, the kernel driver thinks the device has been removed and udevd removes the link. It gets woken up when udevadm test
or udevd --debug
causes the kernel to probe for it. I'm sorry I can't really help more without more diagnostics. Can you at least give me the model of the UPS and information about your kernel/distro.
kernel is anything from the past year, currently gentoo-sources-5.0.4 (Or .3, or .5, I forget) Gentoo ~amd64 UPS is a CyberPower full sine wave 1350PFC
So I caught it in the logs. Isn't this a full disconnect? The rules already use ACTION!="add|change", and it went from being device number 2 to device number 3...
I'll probably buy another cable, since CyberPower looks like they shop at the 59 cent store not even the full dollar store. But that's more because of how shit the current one looks and not because I think it will help.
Mar 31 12:47:23 localhost kernel: [68281.379558] usb usb5-port1: disabled by hub (EMI?), re-enabling...
Mar 31 12:47:23 localhost kernel: [68281.379564] usb 5-1: USB disconnect, device number 2
Mar 31 12:47:24 localhost kernel: [68281.636169] usb 5-1: new full-speed USB device number 3 using uhci_hcd
Mar 31 12:47:24 localhost kernel: [68281.823788] usb 5-1: New USB device found, idVendor=0764, idProduct=0501, bcdDevice= 0.01
Mar 31 12:47:24 localhost kernel: [68281.823791] usb 5-1: New USB device strings: Mfr=3, Product=1, SerialNumber=2
Mar 31 12:47:24 localhost kernel: [68281.823792] usb 5-1: Product: CP1350PFCLCD
Mar 31 12:47:24 localhost kernel: [68281.823794] usb 5-1: Manufacturer: CPS
Mar 31 12:47:24 localhost kernel: [68281.823795] usb 5-1: SerialNumber: 000000000000
Mar 31 12:47:24 localhost kernel: [68281.833887] hid-generic 0003:0764:0501.0004: hiddev96,hidraw0: USB HID v1.10 Device [CPS CP1350PFCLCD] on usb-0000:00:1a.2-1/input0
this looks like a normal disconnect + reconnect with a 1 second delay. i'd start looking at hardware issues first.
Can you please confirm if that is still the case with the latest release?
So this seems to be schrodinger's symlink. I'm running sys-fs/eudev-3.2.7-r1.
If I reboot the box it finds the device and makes a symlink. When I run udevadm test $(udevadm info -q path -n /dev/bus/usb/) it finds the device and makes a symlink. If I run udev from the console with --debug it finds the device and makes a symlink.
Everything is fine, unless the box is running for a while. Then suddenly the device drops off the face of the earth and no amount of hot-plugging or udev restarts bring it back. Running udevadm test or udevd --debug instantly brings it back.
How do I debug this when turning on logging stop the bug from happening? Considering it's only the nut/libusb device that breaks I feel like this might not be a 100% udevd bug.