What the correct device id is, how to find it out?

[naddika]

Arch Linux, Think Pad yoga 11e

It won't run because the fan device is incorrect.

sudo systemctl status thinkfan

 thinkfan.service - simple and lightweight fan control program
     Loaded: loaded (/usr/lib/systemd/system/thinkfan.service; disabled; vendor preset: disabled)
    Drop-In: /etc/systemd/system/thinkfan.service.d
             └─override.conf
     Active: failed (Result: exit-code) since Wed 2021-06-23 21:40:27 +03; 1s ago
    Process: 52494 ExecStart=/usr/bin/thinkfan $THINKFAN_ARGS (code=exited, status=1/FAILURE)
        CPU: 5ms

Jun 23 21:40:27 user1 systemd[1]: Starting simple and lightweight fan control program...
Jun 23 21:40:27 user1 thinkfan[52494]: ERROR: Error scanning /sys/devices/pci0000:00/0000:00:03.1/0000:27:00.0/hwmon: No such file or directory
Jun 23 21:40:27 user1 systemd[1]: thinkfan.service: Control process exited, code=exited, status=1/FAILURE
Jun 23 21:40:27 user1 systemd[1]: thinkfan.service: Failed with result 'exit-code'.
Jun 23 21:40:27 user1 systemd[1]: Failed to start simple and lightweight fan control program.

How do I find out what the correct device is?

---

cat /usr/share/doc/thinkfan/examples/thinkfan.yaml

sensors:
  - hwmon: /sys/class/hwmon/hwmon0/temp1_input

  - hwmon: /sys/devices/pci0000:00/0000:00:03.1/0000:27:00.0/hwmon
    indices: [1, 2, 5, 6] # adds 4 temperature sensors
    correction: [0, 0, 0, -5] # add -5 °C to temp6_input

  - hwmon: /sys/class/hwmon
    name: k10temp
    indices: [1]

  - hwmon: /sys/class/block/sdc/device/hwmon
    indices: [1]
    optional: true # don't exit if the sensor can't be read

  - atasmart: /dev/sda


  - tpacpi: /proc/acpi/ibm/thermal
    indices: [1, 2, 3, 4]


  - nvml: 27:00.0



##############################################################################
# Fan Drivers
# ===========
#

fans:
  - hwmon: /sys/class/hwmon/hwmon0/pwm1

  - hwmon: /sys/class/graphics/fb0/device/hwmon
    indices: [1] # Use pwm1

  - hwmon: /sys/class/hwmon
    name: amdgpu
    indices: [1]

  - tpacpi: /proc/acpi/ibm/fan



##############################################################################
# Fan Speeds (simple mode)

levels:
  - [0, 0, 50]
  - ["level auto", 45, 75]
  - ["level disengaged", 70, 255]



##############################################################################
# Fan Speeds (detailed mode)
# ==========================

levels:
  - speed: 0
    upper_limit: [50, 50, 50]

  - speed: 100
    lower_limit: [45, 45, 45]
    upper_limit: [65, 65, 65]

  - speed: 255
    lower_limit: [60, 60, 60]

[vmatare]

Try running sensors as root to get a first overview and then go ahead and explore your /sys/class/hwmon directory. In case you've never looked at sysfs before: You basically just wanna output the contents of the various files in there, that'll give you an idea what's going on. Basically you're looking for files called temp*_input. Take a look at the manpage (man thinkfan.conf) and PLEASE DON'T just copy an example config. Figuring out why it doesn't work for you (it never will) is probably harder than reading the manpage and writing a correct one from scratch.

[vmatare]

Also very important documentation on /sys/class/hwmon: https://www.kernel.org/doc/html/latest/hwmon/sysfs-interface.html

[naddika]

I've run sensors. And this:

 tree /sys/class/hwmon
/sys/class/hwmon
├── hwmon0 -> ../../devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0003:00/power_supply/ACAD/hwmon0
├── hwmon1 -> ../../devices/virtual/thermal/thermal_zone0/hwmon1
├── hwmon2 -> ../../devices/LNXSYSTM:00/LNXSYBUS:00/PNP0C0A:00/power_supply/BAT1/hwmon2
├── hwmon3 -> ../../devices/platform/thinkpad_hwmon/hwmon/hwmon3
├── hwmon4 -> ../../devices/virtual/thermal/thermal_zone1/hwmon4
├── hwmon5 -> ../../devices/platform/coretemp.0/hwmon/hwmon5
├── hwmon6 -> ../../devices/pci0000:00/0000:00:15.0/i2c_designware.0/i2c-0/i2c-WCOM5071:00/0018:056A:5074.0001/power_supply/wacom_battery_0/hwmon6
└── hwmon7 -> ../../devices/virtual/thermal/thermal_zone3/hwmon7

8 directories, 0 files

There's no temp*_input

[vmatare]

yeah, these are probably symlinks and it looks like your tree program isn't following them.

[vmatare]

take a look at the help with tree -h. You'll want to look at the -l option to make it follow symlinks and the -L option to limit the maximum depth. Try tree -l -L 2 /sys/class/hwmon for example.

[naddika]

% tree -l -L 2 /sys/class/hwmon
/sys/class/hwmon
├── hwmon0 -> ../../devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0003:00/power_supply/ACAD/hwmon0
│   ├── device -> ../../ACAD
│   ├── name
│   ├── power
│   ├── subsystem -> ../../../../../../../class/hwmon  [recursive, not followed]
│   └── uevent
├── hwmon1 -> ../../devices/virtual/thermal/thermal_zone0/hwmon1
│   ├── device -> ../../thermal_zone0
│   ├── name
│   ├── power
│   ├── subsystem -> ../../../../../class/hwmon  [recursive, not followed]
│   ├── temp1_crit
│   ├── temp1_input
│   └── uevent
├── hwmon2 -> ../../devices/LNXSYSTM:00/LNXSYBUS:00/PNP0C0A:00/power_supply/BAT1/hwmon2
│   ├── device -> ../../BAT1
│   ├── in0_input
│   ├── name
│   ├── power
│   ├── subsystem -> ../../../../../../../class/hwmon  [recursive, not followed]
│   └── uevent
├── hwmon3 -> ../../devices/virtual/thermal/thermal_zone1/hwmon3
│   ├── device -> ../../thermal_zone1
│   ├── name
│   ├── power
│   ├── subsystem -> ../../../../../class/hwmon  [recursive, not followed]
│   ├── temp1_input
│   └── uevent
├── hwmon4 -> ../../devices/platform/thinkpad_hwmon/hwmon/hwmon4
│   ├── device -> ../../../thinkpad_hwmon
│   ├── fan1_input
│   ├── name
│   ├── power
│   ├── pwm1
│   ├── pwm1_enable
│   ├── subsystem -> ../../../../../class/hwmon  [recursive, not followed]
│   ├── temp10_input
│   ├── temp11_input
│   ├── temp12_input
│   ├── temp13_input
│   ├── temp14_input
│   ├── temp15_input
│   ├── temp16_input
│   ├── temp1_input
│   ├── temp2_input
│   ├── temp3_input
│   ├── temp4_input
│   ├── temp5_input
│   ├── temp6_input
│   ├── temp7_input
│   ├── temp8_input
│   ├── temp9_input
│   └── uevent
├── hwmon5 -> ../../devices/pci0000:00/0000:00:15.0/i2c_designware.0/i2c-0/i2c-WCOM5071:00/0018:056A:5074.0001/power_supply/wacom_battery_0/hwmon5
│   ├── device -> ../../wacom_battery_0
│   ├── name
│   ├── power
│   ├── subsystem -> ../../../../../../../../../../class/hwmon  [recursive, not followed]
│   └── uevent
├── hwmon6 -> ../../devices/virtual/thermal/thermal_zone2/hwmon6
│   ├── device -> ../../thermal_zone2
│   ├── name
│   ├── power
│   ├── subsystem -> ../../../../../class/hwmon  [recursive, not followed]
│   ├── temp1_input
│   └── uevent
└── hwmon7 -> ../../devices/platform/coretemp.0/hwmon/hwmon7
    ├── device -> ../../../coretemp.0
    ├── name
    ├── power
    ├── subsystem -> ../../../../../class/hwmon  [recursive, not followed]
    ├── temp1_crit
    ├── temp1_crit_alarm
    ├── temp1_input
    ├── temp1_label
    ├── temp1_max
    ├── temp2_crit
    ├── temp2_crit_alarm
    ├── temp2_input
    ├── temp2_label
    ├── temp2_max
    ├── temp3_crit
    ├── temp3_crit_alarm
    ├── temp3_input
    ├── temp3_label
    ├── temp3_max
    └── uevent

[naddika]

there're multiple temp*_input

[vmatare]

Correct. They all have a name (in the file name) that tells you what they are. That name would also have been displayed by the sensors command.

[naddika]

That knowledge doesn't approximate me to solving the issue.

[vmatare]

Well... I deliberately keep thinkfan not too noob-friendly, and there is a good reason why it doesn't "just work" out of the box. If you can't tell by looking at the data which sensor is the correct one to use then it's very likely you either haven't been looking closely enough or you're missing information about the thermal behavior and limitations of your your hardware in order to be able to correctly interpret the data. In either case, I really shouldn't be telling you "put this in your config and you're good to go" because neither do I have that knowledge about your hardware, nor do I know what exact problem you're trying to solve (e.g. do you want to keep it cool or quiet? Hot or cold climate? etc.).

You currently have a fan control algorithm in place that your manufacturer deems safe. If you make a mistake and realize it too late, thinkfan can brick your hardware (in extreme cases) or at the very least cause damage and instability from overheating.

That being said, I can tell you that many people configure thinkfan to manage the CPU temperature only (e.g. by using the coretemp sensor) because that's the simplest to configure. It may let other devices (wifi, hard disk etc.) overheat if your room temperature is high and you CPU utilization is low, so it's certainly not safe under all conditions. So I'm not recommending it, but it can be a starting point to experiment if you keep an eye on the other temperatures as well.

[naddika]

I'm asking whether you recommend it or not.

You've not said anything that answers my question.

[vmatare]

I'm asking whether you recommend it or not.

Well no, I don't recommend using only the coretemp sensor and ignoring all others (assuming that is what you mean by "it").

You've not said anything that answers my question.

Okay, so my answer to your question "What is the correct device id?" is:

If you're unsure: All of them.