On 28/05/2024 14:00, Chukun Pan wrote: >> Please focus instead on why. >> >> https://lore.kernel.org/all/d096d9ea-39db-4a15-9c4d-ae228db970cb@xxxxxxxxxx/ > > Thanks for the link, I missed it. > But for the 4/5G modules, it usually has multiple gpio connections > and cannot be controlled through just one gpio. For example, when > rfkill unblock, if the module is not reset again, it will not work. Please send a patch with proper hardware description used as rationale. Your argument - driver has it - is not correct for the bindings. Bindings do not describe drivers. Read the old v1 (which had reset GPIO) and the arguments presented there. Include your counter-arguments to that old v1 arguments in your patch. > In this case, can rfkill-gpio be allowed to use reset-gpios property? Best regards, Krzysztof