This seems to be related to: "rfkill: make new event layout opt-in" in the -28 kernel. When I revert that a local reproduction run starts to work again. However I never got the same error pattern as in the ADT logs (those seem to get random data back from trying to read the MAC address). So a little doubt remains.
This seems to be related to: "rfkill: make new event layout opt-in" in the -28 kernel. When I revert that a local reproduction run starts to work again. However I never got the same error pattern as in the ADT logs (those seem to get random data back from trying to read the MAC address). So a little doubt remains.