> > Yeah, LM93... four tachos, two PWM outputs. Odd, but it seems like > > one anticipated configuration is to drive fans 1-2 (tacho's 1-2) > > *both* from pwm1. I could be wrong about that though. > > > Your not. A real pending configuration looks like: > > Fan A1----\ > Fan A2----|---\ > >-------PWM1 > Fan B1----|---/ > Fan B2----/ > > Fan C1----|------------PWM2 > Fan C2----/ OK, I have to accept Mark's idea then. Would it make sense to accept both fanN_pwm and pwmN, and use the approriate entries depending on the chip? It would also mean two possible names for auto-fan contol files (it87 model), so there would be three possible sets of names instead of two. This is not good, but OTOH, using pwmN for SMSC-like chips will probably make the user wonder why fan1 will stop monitoring when pwm1 is set to 0% duty cycle... So I guess we better simply revert to pwmN as Mark suggested. -- Jean "Khali" Delvare http://khali.linux-fr.org/
