> Hi Barnabás, > > I did another refactor using hwmon_device_register_with_info() and > HWMON_CHANNEL_INFO(). I'm unsure if this is what you were looking for > so I'm going to attach the patch instead of submitting as a V8 for now. > > My main concern as that the use of the above removes the > pwm1_auto_point1_pwm + pwm1_auto_point1_temp format and gives two > hwmon<num>, one per cpu/gpu fan with: > > device power > fan1_input subsystem > fan2_input temp1_input > fan3_input temp2_input > fan4_input temp3_input > fan5_input temp4_input > fan6_input temp5_input > fan7_input temp6_input > fan8_input temp7_input > in0_enable temp8_input > name uevent > > cat -p /sys/devices/platform/asus-nb-wmi/hwmon/hwmon7/name > asus_cpu_fan_custom_curve > > I've named the root name of each as descriptive as possible to convey > exactly what each is > > Oh and `sensors` now shows: > > asus_cpu_fan_curve-isa-0000 > Adapter: ISA adapter > fan1: 8 RPM > fan2: 10 RPM > fan3: 18 RPM > fan4: 20 RPM > fan5: 28 RPM > fan6: 34 RPM > fan7: 44 RPM > fan8: 56 RPM > temp1: +0.0°C > temp2: +0.1°C > temp3: +0.1°C > temp4: +0.1°C > temp5: +0.1°C > temp6: +0.1°C > temp7: +0.1°C > temp8: +0.1°C > > > > FYI, the pwmX_enable attributes can be created by the hwmon > > subsystem itself if you use [devm_]hwmon_device_register_with_info() > > with appropriately populated `struct hwmon_chip_info`. > > So when you say this, did you mean *just* for the pwmX_enable > attributes? > [...] Yes, that's right. Best regards, Barnabás Pőcze