Hi Marek, > > > + * r8a779g3-sparrow-hawk$ grep -H . /sys/class/hwmon/hwmon?/name > > > + * /sys/class/hwmon/hwmon0/name:sensor1_thermal > > > + * /sys/class/hwmon/hwmon1/name:sensor2_thermal > > > + * /sys/class/hwmon/hwmon2/name:sensor3_thermal > > > + * /sys/class/hwmon/hwmon3/name:sensor4_thermal > > > + * /sys/class/hwmon/hwmon4/name:pwmfan > > > + * ^ ^^^^^^ > > > > For me, the output looks different, though: > > > > /sys/class/hwmon/hwmon0/name:pwmfan > > /sys/class/hwmon/hwmon1/name:sensor1_thermal > > /sys/class/hwmon/hwmon2/name:sensor2_thermal > > /sys/class/hwmon/hwmon3/name:sensor3_thermal > > /sys/class/hwmon/hwmon4/name:sensor4_thermal > > /sys/class/hwmon/hwmon5/name:pwmfan > > > > hwmon0 is the Argon, hwmon5 the (unpopulated) on board connector. > > > > I hope the naming is stable, but in any case, the docs need to be > > reworked a little, I guess? > > The hwmon devices are allocated first come first served, so the list can > look different for you. You need to look up the fan you want to control, of > course. Look up by name is simplest, if there are more fans, it might be a > bit more involved to find the right one. Exactly. This is why I would suggest to drop the '^^^^^' line from above. We definitely have two pwmfans with the above dtso. So finding the right one needs a second look anyhow. And while I think it is quite likely that the list of hwmon devices will look the same (I don't see a race condition which would make hwmon0 and hwmon5 swap), I agree we shouldn't rely on it. I would also think that we then maybe should change echo 2 > /sys/class/hwmon/hwmon4/pwm1_enable to echo 2 > /sys/class/hwmon/hwmon<X>/pwm1_enable or something? What do you think? Just suggestions, I am not really insisting. Well, maybe, about dropping the "^^^^" line ;) Happy hacking, Wolfram
