How to enable cpu1, cpu2 cpu3 in T30


In T30, i can see only CPU0 is on line, other CPU’s are off line.
If i am trying to write “1” to cpu1, getting message “Switched to NOHz mode on CPU #1” and CPU1 is still off line(means on line value is “0”)

So how to make all CPU’s on line once the Linux is booted?


root@apalis-t30:~# cat /sys/devices/system/cpu/cpu0/online 
root@apalis-t30:~# cat /sys/devices/system/cpu/cpu1/online 
root@apalis-t30:~# cat /sys/devices/system/cpu/cpu2/online 
root@apalis-t30:~# cat /sys/devices/system/cpu/cpu3/online 
root@apalis-t30:~# echo 1 > /sys/devices/system/cpu/cpu1/online 
root@apalis-t30:~# cat /sys/devices/system/cpu/cpu1/online 
root@apalis-t30:~# dmesg | tail
[   13.096190] android_usb gadget: using random self ethernet address
[   13.103788] android_usb gadget: using random host ethernet address
[   15.041787] ADDRCONF(NETDEV_UP): enp1s0: link is not ready
[   15.047274] 8021q: adding VLAN 0 to HW filter on device enp1s0
[   15.091743] rndis0: MAC ea:b6:dc:d6:db:00
[   15.120898] rndis0: HOST MAC ee:dc:60:56:d4:bb
[   15.153362] ADDRCONF(NETDEV_UP): rndis0: link is not ready
[   15.179807] Switched to NOHz mode on CPU #1

Hi @deepak.singh

The NVIDIA © Tegra 3 embedded System-on-Chip ncorporates DVFS (Dynamic Voltage and Frequency Switching) and Thermal Throttling which enables the system to continuously adjust operating frequency and voltage in response to changes in workload and temperature to achieve the best performance with the lowest power consumption.

So if there are no tasks running which demand high cpu power, then only one cpu will run. However if there is an application, which creates a big workload, then all other cpus will be also active.

This can be tested by using stress ( stress -c 4 &) and then checking the state ( cat /sys/devices/system/cpu/cpu*/online) of all cpus.

Best regards, Jaski

Thank you Jaski for the clarification.

You are welcome.