Jean Delvare wrote: > Hi Chunhao, > > >>I noticed the subclient registration when I started to write 792 >>driver, I'm not >>familiar with subclient and sent lm_sensors group a mail for your >>introduction >>about it about several month ago, but do not receive any response. >>Since I'm not familiar with it, I discard it just like the other chip >>drivers: >>lm90.c and lm83.c, finding it can work well. > > > Just a couple clarifications: > > 1* Not all chips have subclients, in fact most haven't, only Winbond > chips and Asus clones have. > > 2* In our current implementation, subclients are registered only to > prevent other drivers to use them. They are not accessed by their > address (the same functions are available from the main address). > Khali is correct. It doesn't affect operation of the driver, only prevent other drivers from interfering. I added, tested, and checked in the subclient registration code. I also changed the init default to 0, which is our standard. mds
