Cavium CN7322 / Marvell LiquidIO II CN2360 Thermal Sensors

hwmon0.md

cn7322:/usr/lib/check_mk_agent/plugins# ls -l /sys/class/hwmon/hwmon0/
total 0
-r--r--r-- 1 root root 4096 Mar 21 15:07 alarms
lrwxrwxrwx 1 root root    0 Mar 21 15:07 device -> ../../../0-004c
-r--r--r-- 1 root root 4096 Mar 21 15:07 name
lrwxrwxrwx 1 root root    0 Mar 21 15:07 of_node -> ../../../../../../../../firmware/devicetree/base/soc@0/i2c@1180000001000/tmp@4c
drwxr-xr-x 2 root root    0 Mar 21 15:07 power
lrwxrwxrwx 1 root root    0 Mar 21 15:07 subsystem -> ../../../../../../../../class/hwmon
-rw-r--r-- 1 root root 4096 Mar 21 15:07 temp1_crit
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp1_crit_alarm
-rw-r--r-- 1 root root 4096 Mar 21 15:07 temp1_crit_hyst
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp1_input
-rw-r--r-- 1 root root 4096 Mar 21 15:07 temp1_max
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp1_max_alarm
-rw-r--r-- 1 root root 4096 Mar 21 15:07 temp1_min
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp1_min_alarm
-rw-r--r-- 1 root root 4096 Mar 21 15:07 temp2_crit
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp2_crit_alarm
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp2_crit_hyst
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp2_fault
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp2_input
-rw-r--r-- 1 root root 4096 Mar 21 15:07 temp2_max
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp2_max_alarm
-rw-r--r-- 1 root root 4096 Mar 21 15:07 temp2_min
-r--r--r-- 1 root root 4096 Mar 21 15:07 temp2_min_alarm
-rw-r--r-- 1 root root 4096 Mar 21 15:07 temp2_offset
-rw-r--r-- 1 root root 4096 Mar 21 15:07 uevent
-rw-r--r-- 1 root root 4096 Mar 21 15:07 update_interval

lnx_thermal_hwmon.sh

#!/usr/bin/env bash

# angepasst auf basis von https://gitlab.com/pers-proj/monitoring-extensions/check_lnx_thermal_hwmon/-/blob/master/agents/plugins/lnx_thermal_hwmon.sh?ref_type=heads
# (der Treiber definiert kein label)

set -eu

echo "<<<lnx_thermal:sep(124)>>>"

for hwmon in /sys/class/hwmon/hwmon*; do
	for F in "${hwmon}"/temp*_input; do
		if ! [ -f "$F" ]; then
			continue
		fi
		sensor="${F##*/}"
		sensor_base="${F%_input}"

		label_file="${sensor_base}_label"
		if [ -f "$label_file" ]; then
			read -r label < "$label_file"
		else
			#continue
                        label=${sensor}
		fi

		device="$(readlink "${hwmon}/device")"

		# Ignore thermal_zone originated hwmon, since that's covered by
		# by the stock agent
		case "$device" in
			../../thermal_zone*)
				continue
				;;
		esac

		if ! read -r F < "$F"; then
			continue
		fi

		read -r device_name < "${hwmon}/name"

		# name | mode | type (unused) | temperature
		line="${device_name}_${sensor}|-|${hwmon##*/}_$sensor|${F}"

		max_temp_file="${sensor_base}_max" # max temp (if present)
		if [ -f "$max_temp_file" ]; then
			read -r max_temp < "$max_temp_file"
			line="${line}|${max_temp}|critical"
		fi

		max_hyst_temp_file="${sensor_base}_max_hyst" # "warn" temp (if present)
		if [ -f "$max_hyst_temp_file" ]; then
			read -r max_hyst_temp < "$max_hyst_temp_file"
			line="${line}|${max_hyst_temp}|hot"
		fi

		echo "$line"
	done
done
echo "<<<>>>"

output_example.md

<<<lnx_thermal:sep(124)>>>
sa56004_temp1_input|-|hwmon0_temp1_input|33625|80000|critical
sa56004_temp2_input|-|hwmon0_temp2_input|46375|85000|critical
<<<>>>

Achtung

• alarmwerte und ihre verarbeitung nicht getestet
• im u-boot environment irgendwo bei den I2C sachen findet man hinweise auf 3 verschiedene Sensorchips, hier nicht