NMI watchdog: BUG: soft lockup - CPU#1 stuck for 22s! [khugepaged:38]

[Corben]

On kernel 4.10 from the linux-hwe-edge package, I'm experiencing frequent freezes on my surface pro 3. This starts mostly when using firefox, where first firefox freezes (window is becoming grey). When I don't close firefox immediately (killing it, as its window greyed out), the complete system kinda locks up.
When I kill it, its process is still as defunct visible in the process list, and I can't start a new instance of firefox.

When I don't kill it, the NMI watchdog BUG occurs shortly after and I have to shut down the system via power button. It's not completely dead, as I can still ping the device and when I'm ssh'd into it, the shell is not completely dead. I'm having a tail -f /var/log/syslog open via ssh, as I can't look into it anymore when the desktop environment locked up. Here I can close the tail -f command, but can't start any new processes.

This did not happen with the linux-signed-image-generic-hwe-16.04 (4.8.0.51.22) yet.

apt-cache policy linux-image-generic-hwe-16.04-edge: 4.10.0.20.13
lsb_release -rd: Ubuntu 16.04.2 LTS

I attached the excerpt of /var/log/syslog.

ProblemType: Bug
DistroRelease: Ubuntu 16.04
Package: linux-generic-hwe-16.04-edge (not installed)
ProcVersionSignature: Ubuntu 4.8.0-51.54~16.04.1-generic 4.8.17
Uname: Linux 4.8.0-51-generic x86_64
ApportVersion: 2.20.1-0ubuntu2.5
Architecture: amd64
CurrentDesktop: Unity
Date: Fri May 5 16:41:03 2017
InstallationDate: Installed on 2015-11-12 (540 days ago)
InstallationMedia: Ubuntu 15.10 "Wily Werewolf" - Release amd64 (20151021)
SourcePackage: linux-meta-hwe-edge
UpgradeStatus: Upgraded to xenial on 2016-07-29 (279 days ago)

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[pmow (pmow)]

this seems very similar to #1689538.

[Adrian (adrian-b)]

@Corben check out #1674838 (#1689538 is marked a duplicate of #1674838). Per your syslog you're getting the swapops.h:129 error before the soft lockups, as I was.

Installing the kernel recommended there resolved the swapops.h:129 error for me. I believe the fix is coming into updates on the 5th of June.

I was stable for 20 days with that kernel however I've just got the following, so now I'm hunting around again...
"NMI watchdog: BUG: soft lockup - CPU#1 stuck for 22s! [pool:16907]"

[Corben (tobias-krummen)]

Thanks for your answer @Adrian.
I went back to the linux-signed-image-generic-hwe-16.04 package, which has kernel 4.8.0.52.23 inside.
Haven't had any issues like this with that kernel since my bug-report here.

With the 4.8 kernel I have other issues with my surface pro 3 related to wifi: #1690332

[Timo Aaltonen (tjaalton)]

assuming this got fixed since then, reopen if not