About Using the Test APP : ThermalPro_x64

Excuse me, I use the program ThermalPro_x64 for temperature control testin. In the 24.1 version of the normal use, but in the 28.1 version, as long as the temperature is higher than the set temperature of cpu_throttle, start adjusting, about five to six minutes there will be a crash, even if I set cpu_throttle below the sixty degrees.
The crash situation is the entire Ubuntu interface can not be used, but still receive the message through the UART and issued a directive, how this is ?
And I can not use this program on the public board, there will be the error message shown in the figure, what’s going on?

thermalpro_error.bmp (1.29 MB)

I know nothing of this software, but is there hardware attached for the software to work with? If so, how is the hardware connected? Were there any drivers added? Does any hardware get its power from the Jetson itself, or is it self-powered (assuming there is extra hardware)?

Hi michaelyang,

The kernel version of R24.1 is 3.10.96, however the R28.1 is 4.4.15, this might be the cause to have different result. BTW, where did you get this temperature control testing program - ThermalPro_x64?

Thanks

Hi,
This program is provided by the NVIDIA FAE.If the tool can’t not used on the version R28.1, do you have any suggested tool ?

Thanks.

Hi michaelyang,

AFAIK, we don not support that tool on L4T now even it can work in previous R24.1 version.

Please refer to Jetson TX1 Thermal Design Guide to do the experiment:
http://developer.nvidia.com/embedded/dlc/jetson-tx1-thermal-design-guide

You can try to increase CPU/GPU clock and keep the maximum freq if simply wants to have a high loading scenario.

echo 1 > /d/clock/gbus/floor.gbus/state

Check if it really runs at expected clock:
cat /sys/kernel/debug/clock/cpu/pto_rate

Thanks