Comment # 2 on bug 1179869 from fcicq

reproducible: USB device descriptor related errors are more frequent (need
hours?) with older kernel but I'm happy with 5.9 (at the moment none?), but
that is uninteresting and no way to fix (I dont think I can easily power cycle
the internal usb port). so I learned the suspend-wake cycle could help to
restore the functionality and not forced to reboot. so this question is once.

regression: not sure.

---

maybe it is an uninteresting bug. oops can be resolved but hw failure still
needs some power cycle.

my guess:
some hw failure
-> bluetoothd stuck (but there is no related log!) and closing?
-> suspend/wake
-> oops on resumed cleanup? or any remaining work?