1

System board: PCEngines APU2C4

uname -a
Linux demobox 4.9.0-0.bpo.8-amd64 #1 SMP Debian 4.9.110-3+deb9u5~deb8u1 (2018-10-03) x86_64 GNU/Linux

I am experiencing an issue with a specific system board where the system will show the following output on the serial terminal. The issue seems to be related to the lm-sensors package. On this system the "sensors" command is executed every 5 seconds to update a text file with the CPU temperature.

Here is the output which is displayed in the serial terminal.

[46032.528636] NMI watchdog: BUG: soft lockup - CPU#1 stuck for 23s! [sensors:27184] 
[46044.524697] NMI watchdog: BUG: soft lockup - CPU#0 stuck for 23s! [sensors:27181] [46052.536738] NMI watchdog: BUG: soft lockup - CPU#3 stuck for 22s! [sensors:27323] 
[46060.528779] NMI watchdog: BUG: soft lockup - CPU#1 stuck for 23s! [sensors:27184] 
[46061.480782] INFO: rcu_sched detected stalls on CPUs/tasks: 
[46061.486295] 2-...: (1 GPs behind) idle=b53/140000000000000/0 softirq=793153/793154 fqs=1536 
[46061.494819] (detected by 1, t=9741845 jiffies, g=655278, c=655277, q=6428) 
[46061.501934] rcu_sched kthread starved for 9738665 jiffies! g655278 c655277 f0x2 RCU_GP_WAIT_FQS(3) ->state=0x0 
[46072.524840] NMI watchdog: BUG: soft lockup - CPU#0 stuck for 22s! [sensors:27181] 
[46080.536881] NMI watchdog: BUG: soft lockup - CPU#3 stuck for 22s! [sensors:27323] 
[46088.528922] NMI watchdog: BUG: soft lockup - CPU#1 stuck for 23s! [sensors:27184]
0

First of all:

Are you experiencing performance issues?

If you type top, do you see anything wrong? Is there any service consuming a lot of CPU?

lm-sensors simply gather information of the CPU but sometimes can have problems detecting the HW; problems that may give you "false" errors, but that should not affect to the performance. In case of having performance issues consider the following:

-Update the kernel to the latest version.

-If the power supply is wrong, consider to test with another one (if possible) as a "broken" power supply can affect to the general motherboard/cpu performance.

  • Hi There, Its worth noting that once the messages above appear in the serial terminal the system is unresponsive, i cant SSH to it or ping its IP. While the system is accessible there is nothing obvious to suggest that there are any problems. This is a production system and i have others exactly the same to compare it with. The kernel is up to date and the PSU is above spec for the board. – Simon Dunning Jan 8 at 16:10
  • @SimonDunning, in that case I would suggest to do the previous steps to see if it solves the issue. – Dasel Jan 8 at 16:14
  • Can you think of anything else i can do to dig a bit deeper ? Thanks – Simon Dunning Jan 8 at 16:17
  • If the kernel update is not feasible, you can use dmesg to check which messages appear just before appearing those errors. You can also try to include the nomodeset parameter to the GRUB entry in order to load the drivers in a different way; that change can sometimes skip those kind of problems; here you have a link that explains how to add that parameter to GRUB: askubuntu.com/questions/38780/… – Dasel Jan 9 at 9:14

Your Answer

By clicking "Post Your Answer", you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.