You are not logged in.
Hello everyone.
I'm trying to figure out what's causing this message "cgroup: remount is not allowed" on shutdown. It just showed up a few days ago. After doing some searches, there was a bug with systemd 228 and the "systemd.unified_cgroup_hierarchy=1" option giving this error. There was also some mention of file privilege issue's in /sys/fs/cgroup/systemd/?, though I hadn't changed anything regarding that. Everything is still default. I'm using v233-7 at the moment with kernel 4.11.7-1.
In case this helps:
Output of ls -l /sys/fs/
total 0
dr-xr-xr-x 2 root root 0 Jul 5 07:58 bpf
drwxr-xr-x 13 root root 340 Jul 5 07:57 cgroup
drwxr-xr-x 6 root root 0 Jul 5 07:58 ext4
drwxr-xr-x 3 root root 0 Jul 5 07:57 fuse
drwxr-xr-x 2 root root 0 Jul 5 07:57 pstore
Is this just a cosmetic error I shouldn't worry about? The PC doesn't hang at all during shutdown and nothing seemed to stand out in journalctl, no errors in systemctl --failed either.
Offline
It seems to be a bug introduced in systemd v233. I just downgraded to v232 and the message is gone. I'm still curious if it is related to systemd.unified_cgroup_hierarchy=1. Will post back if I find anything.
Offline
I'm getting the same message after upgrading systemd. I agree, it appears to be cosmetic, though I can't confirm this.
Offline
Me too, I have the same message and I don't know why.
Last edited by SupKurtJ (2017-07-06 07:26:51)
Offline
Me too, " Cgroup : Remount is not allowed ", is this a problem with the kernel?
Offline
https://wiki.archlinux.org/index.php/Co … mpty_posts
Do your homework, try to find what causes the problem and submit bug reports when you reach a conclusion. Me too posts are not magically going to help solve the problem.
R00KIE
Tm90aGluZyB0byBzZWUgaGVyZSwgbW92ZSBhbG9uZy4K
Offline
I don't think it's related to the kernel as downgrading it did not remove the error.
Offline
I don't think it's related to the kernel as downgrading it did not remove the error.
Then the next place to look might be systemd. That said, if nothing is broken you might save some time by ignoring that message.
R00KIE
Tm90aGluZyB0byBzZWUgaGVyZSwgbW92ZSBhbG9uZy4K
Offline
Pillgar wrote:I don't think it's related to the kernel as downgrading it did not remove the error.
Then the next place to look might be systemd. That said, if nothing is broken you might save some time by ignoring that message.
Is this your full time job or?
Offline
https://wiki.archlinux.org/index.php/Co … mpty_posts
Do your homework, try to find what causes the problem and submit bug reports when you reach a conclusion. Me too posts are not magically going to help solve the problem.
https://bugs.archlinux.org/task/54746
There is nice vote-link which should help developers and maintainers guiding attention, without disturbing and furthermore give users the feeling that they are not alone. Likely we will put it further to upstream, hopefully it is not real problem or a message which should be stored in the system-log instead of makeing users worried/paniced. Nobody likes such messages, but we need them
Last edited by hoschi (2017-07-09 14:12:15)
Offline
Update:
Systemd people think it's caused by Archlinux
Ping ... Pong ... Ping ... Pong
Offline
I'm not sure if this helps but 2 of my systems are running the latest 64 bit Arch. A desktop (AMD Phenom, NVIDIA GeForce, wired network and SATA drives) and a laptop (Intel Core I7, Intel Ivybridge Mobile, wireless and an SSD drive). The laptop does show the message on restart/power off but not the desktop. There are only a few packages that differ between the two.
Offline
@eda2x can you run `systemctl halt` on the desktop to check the message is not just being displayed at the moment the system would power off.
As noted the message is displayed after all none virtual filesystems have been unmounted and systemd (process one) has terminated so it is difficult to see how this is anything but cosmetic.
If someone wants to pursue it they would probably need to switch to using the systemd hooks so systemd is the init system in early boot then bisect between 322 and 323.
That bisection process in itself I would consider more likely to induce breakage than the original issue.
After finding the commit that causes the behavior change then discuss the issue between systemd, linux kernel and arch developers.
Edit:
https://bugs.archlinux.org/task/54746#comment159336 so the bisect could be avoid if using systemd.legacy_systemd_cgroup_controller
demonstrates the issue is caused by the use of hybrid or unified cgroup hierarchy. So the message from the kernel is produced when a cgroup v2 virtual filesystem is remounted ro?
Edit2:
systemctl mask mkinitcpio-generate-shutdown-ramfs.service
Disables the shutdown service that causes the remount so no message. See https://bbs.archlinux.org/viewtopic.php?pid=1372562 for what the service does.
Last edited by loqs (2017-07-14 21:44:55)
Offline
loqs - executing 'systemctl halt' did cause the message to be displayed on the desktop.
Offline
There is a bug report here:
Offline
As the upstream bug report has been closed, the cosmetic message will no longer appear with linux 4.13 and there has been at least two workarounds provided, can this thread be marked as solved?
Offline
I have the same problem, how to solve it
Offline
I have the same problem, how to solve it
What have you tried so far?
Offline