Hello!
I’m facing a problem where my system freezes completely after a while.
I had a complete freeze after a few hours of playing (minecraft), only a reset button helped, and then soon after a freeze again.
The day before, only music was playing on the computer, did nothing, after a couple of hours I just hear the music repeating a tiny section, again my system froze.
I’ve been experiencing this problem for about 2-3 weeks. I have almost no logs in the /var/log/ folder, no syslog, no dmesg, no kernel, no samba. memtest ran without error.
However, I do have a journalctl log that I would like to share with you, and please, anyone who sees a fatal error, help me??
The error occurred around 21:30 on 2023.10.13 and 17:53 on 2023.10.14.
joutnalctl 1:
journalctl 2:
Please, someone who is an expert, please find out what is wrong!
Thank you!
(Excuse my English, I am Hungarian)
there is only zswap. Thus it loops, because it cannot allocate more RAM and didn’t kill Minecraft because the OOMD service (Out-Of-Memory Killer) is not running.
Add a swap partition/file or start/enable systemd-oomd.service.
Unfortunately, the solution is not quite right.
Yesterday during the game I had another freeze and I think there are worrying things in my log. I’ve attached it in case they can help:
These symptomatic errors can have various issues, but at least it has to do with the CPU scheduler from what I see in the logs. Without context or debugging the kernel, it’s hard to say anything.
At least pick_next_task_fair is a function of the CFS scheduler. Something, and I would say Minecraft, has crashed it.
When I see page_fault_oops+0x15a/0x2d0 then this refers to a problem where no RAM can be allocated (most likely, but not sure).
Limit Minecraft to use only the RAM you have, not more.
And when I see this:
okt 14 17:54:27 xenomanjaro kernel: total RAM covered: 3071M
Then you have just the bare minimum of RAM. Minecraft needs at least 4GB, you have 3GB.
I’ve never used a CPU scheduler, what can screw it up?
I have enough RAM, I think (16GB), plus I now have a 16GB swap file set up.
Where can I start to troubleshoot this, as I’m not very professional?
I have extra monitor, now I will monitor events, ram usage while playing…
Thank you!
You always use one. A scheduler just keeps sure that everything runs in the right order. While multi-tasking, it keeps priorities and balance, thus let the whole system be responsive.
Here your governor is schedutil, which clocks up the cpu based on kernel I/O activity (in simple words). Therefore you got this kernel panic and not another one.
I had the memtest before my last report, no errors, ran all night.
I didn’t monitor it, but today I’ll try to see how much RAM it uses.
There is no mod/texture package.
I start the game from terminal with mininum 2GB max 4GB RAM.
$ java -Xmx4096M -Xms2048M -jar /
Not sure, since I never played Minecraft, but when I read the system requirements, then you need at least 4 GB x86_64 (ARM64 is lower). I guess a limit of 4GB could produce problems, since Java doesn’t allow Minescraft to allocate more RAM, but normally it should not freeze the system (kernel panic), but rather crash the program.
Hi!
I am back again. The manjaro froze again, but now the game is not even running.
Only terminals to monitor the processes (htop, journalctl -f, watch free -h) plus a data recovery script (FTP to local HDD). About 1/5 of the memory was used.
However, htop Mem, although not much used, was orange in total. If I remember correctly, that means memory occupied?
Anyone got any more ideas?
journalctl log:
nov 06 20:33:20 xenomanjaro NetworkManager[1009]: <info> [1699299200.6997] dhcp6 (enp34s0): state changed new lease, address=2001:4c4c:109a:8e00:a52a:7140:97e5:b222
nov 06 20:33:20 xenomanjaro dbus-daemon[921]: [system] Activating via systemd: service name='org.freedesktop.nm_dispatcher' unit='dbus-org.freedesktop.nm-dispatcher.service' requested by ':1.5' (uid=0 pid=1009 comm="/usr/bin/NetworkManager --no-daemon")
nov 06 20:33:20 xenomanjaro systemd[1]: Starting Network Manager Script Dispatcher Service...
nov 06 20:33:20 xenomanjaro dbus-daemon[921]: [system] Successfully activated service 'org.freedesktop.nm_dispatcher'
nov 06 20:33:20 xenomanjaro systemd[1]: Started Network Manager Script Dispatcher Service.
nov 06 20:33:30 xenomanjaro systemd[1]: NetworkManager-dispatcher.service: Deactivated successfully.
nov 06 20:44:20 xenomanjaro NetworkManager[1009]: <info> [1699299860.9466] dhcp4 (enp34s0): state changed new lease, address=192.168.0.58
nov 06 20:44:20 xenomanjaro dbus-daemon[921]: [system] Activating via systemd: service name='org.freedesktop.nm_dispatcher' unit='dbus-org.freedesktop.nm-dispatcher.service' requested by ':1.5' (uid=0 pid=1009 comm="/usr/bin/NetworkManager --no-daemon")
nov 06 20:44:20 xenomanjaro systemd[1]: Starting Network Manager Script Dispatcher Service...
nov 06 20:44:20 xenomanjaro dbus-daemon[921]: [system] Successfully activated service 'org.freedesktop.nm_dispatcher'
nov 06 20:44:20 xenomanjaro systemd[1]: Started Network Manager Script Dispatcher Service.
nov 06 20:44:30 xenomanjaro systemd[1]: NetworkManager-dispatcher.service: Deactivated successfully.
nov 06 20:58:07 xenomanjaro kernel: BUG: kernel NULL pointer dereference, address: 0000000000000017
nov 06 20:58:07 xenomanjaro kernel: #PF: supervisor read access in kernel mode
-- Boot 596264090fc94065b26860e9fe649031 --
