Half of the 32GB AGX Orins I am working with are not reporting temperatures for CPU, GPU, etc thermal zones on the Orin AGX module, when I run tegrastats.
Kernel boot logs show errors in initialization of the thermal zones.
$ sudo dmesg | grep thermal
[ 0.256937] thermal_sys: Registered thermal governor 'step_wise'
[ 0.256939] thermal_sys: Registered thermal governor 'power_allocator'
[ 0.256944] thermal_sys: Registered thermal governor 'pid_thermal_gov'
[ 0.257291] OF: /thermal-zones/CPU-therm/cooling-maps/map0: could not find phandle
[ 0.257297] thermal_sys: Add a cooling_device property with at least one device
[ 0.257303] thermal_sys: failed to build thermal zone CPU-therm: -2
[ 0.257368] OF: /thermal-zones/GPU-therm/cooling-maps/map0: could not find phandle
[ 0.257372] thermal_sys: Add a cooling_device property with at least one device
[ 0.257376] thermal_sys: failed to build thermal zone GPU-therm: -2
[ 0.257418] OF: /thermal-zones/CV0-therm/cooling-maps/map0: could not find phandle
[ 0.257422] thermal_sys: Add a cooling_device property with at least one device
[ 0.257426] thermal_sys: failed to build thermal zone CV0-therm: -2
[ 0.257465] OF: /thermal-zones/CV1-therm/cooling-maps/map0: could not find phandle
[ 0.257469] thermal_sys: Add a cooling_device property with at least one device
[ 0.257472] thermal_sys: failed to build thermal zone CV1-therm: -2
[ 0.257511] OF: /thermal-zones/CV2-therm/cooling-maps/map0: could not find phandle
[ 0.257515] thermal_sys: Add a cooling_device property with at least one device
[ 0.257519] thermal_sys: failed to build thermal zone CV2-therm: -2
[ 0.257556] OF: /thermal-zones/SOC0-therm/cooling-maps/map0: could not find phandle
[ 0.257561] thermal_sys: Add a cooling_device property with at least one device
[ 0.257564] thermal_sys: failed to build thermal zone SOC0-therm: -2
[ 0.257602] OF: /thermal-zones/SOC1-therm/cooling-maps/map0: could not find phandle
[ 0.257606] thermal_sys: Add a cooling_device property with at least one device
[ 0.257610] thermal_sys: failed to build thermal zone SOC1-therm: -2
[ 0.257648] OF: /thermal-zones/SOC2-therm/cooling-maps/map0: could not find phandle
[ 0.257652] thermal_sys: Add a cooling_device property with at least one device
[ 0.257655] thermal_sys: failed to build thermal zone SOC2-therm: -2
This is 100% reproducible i.e the issue persists after power-cycles.