System freezes randomly

kernel

#1

Good morning,

I have installed Manjaro few months ago, had OpenSUSE, Fedora and Ubuntu before. Manjaro keeps randomly hanging (full kernel freeze, not even magic sysrq help, only power off using power button on box). Hangs occur completely randomely, but quite seldom. Computer is stable for days and then suddenly freezes and sometimes it works 5 minutes from boot and then hangs. I’ve tried using different kernel version (currently on 4.13.4-3), different nVidia driver, nothing helps. None of previous distributions had this problem on this hardware. Any ideas how to debug? Nothing about that hang in dmesg or journalctl.

Output of inxi -b:

System: Host: manufactorum Kernel: 4.13.4-3-MANJARO x86_64 bits: 64 Desktop: KDE Plasma 5.10.5
Distro: Manjaro Linux
Machine: Device: desktop Mobo: Gigabyte model: B85M-D2V v: x.x serial: N/A UEFI: American Megatrends v: F3 date: 03/11/2014
CPU: Quad core Intel Core i5-4460 (-MCP-) speed/max: 3192/3400 MHz
Graphics: Card: NVIDIA GK107 [GeForce GTX 650]
Display Server: x11 (X.Org 1.19.3 ) driver: nvidia
Resolution: 1920x1080@60.00hz, 1920x1080@60.00hz, 1920x1080@60.00hz
OpenGL: renderer: GeForce GTX 650/PCIe/SSE2 version: 4.5.0 NVIDIA 384.90
Network: Card-1: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller driver: r8168
Card-2: Realtek RTL8192CE PCIe Wireless Network Adapter driver: rtl8192ce
Drives: HDD Total Size: 1240.3GB (27.4% used)
Info: Processes: 271 Uptime: 4:13 Memory: 6863.3/16013.3MB Client: Shell (fish) inxi: 2.3.40


My laptop keeps randomly freezing on Manjaro
#2

'LLo,

Depending of your machine, have you checked if ram memory cards are (always) correctly engaged ?


#3

How can I check that? I have executed both SMART tests and memchecks to rule out hardware faults, nothing came up in them.


#4

You take them out and put them back in. :slight_smile:


#5

But I haven’t modified hardware configuration at all in this year and OpenSUSE worked perfectly well (besides broken updates which convinced me it’s time to jump ship to something more archy).


#6

Seemingly random freezes are almost always down to one or more of: RAM, overheating, loose hardware, lack of power.

Cards can work themselves free over time. This is a known issue and why AGP and PCIe slots have clips.

Troubleshooting/debugging is a process of elimination, so if you follow a series of steps it narrows down the possible causes of the issue.

If you ignore a suggestion you leave that as a variable. If you ask for help then question the answer because it’s not the simple answer you want (“install this” or “make this specific configuration change”), then people aren’t going to bother giving any more help.

So, the first step is to ensure all of your hardware components are seated correctly.

Or, go back to SuSE and run it for the same length of time as Manjaro and see if the issues reoccur.


#7

Ok then, I will reinsert/reconnect all hardware components and see if that helps. I didn’t mean to sound like know it all smart ass, I have just tried to rule out hardware problems before.


#8

It’s OK, we’re chill here. :wink:

Nobody–ever–wants to anticipate a possible hardware problem. But it’s like @jonathon said, with problems such as you describe it is back to basics, 1, 2, 3…


#9

Ok, so I have dismantled my PC, took out my 2 8GB blue Kingston Fury’s out of Mobo, cleaned everything and put them in again. Then I have left it working for whole night and somewhere around 9 am it froze completely. What’s the next step in debug?

One thing that might be helpful is that I have this weird problem with SDDM segfaulting on login attempt occasionally and instantly restarting.


#10

Sadly, this still happens quite often (about once a day). Temperature of components is well below danger zones, as I rarely put system under big load (incremental compilers don’t really force the CPU or GPU). Does anyone know what would be the next step for diagnostics with this problem?


#11

Use an installation disc to run MemTest overnight. Replacing cheap RAM is better than replacing a computer…

Anyway, that would be my next test, to check against flaky RAM.


#12

Full overnight MemTest check did not yield anything, memory is corruption-free. HDD is corruption free too, at least full SMART tests claim that. It’s a Manjaro thing, I switched back to OpenSuse partition (still have it in my grub) and left machine working for some time while I used laptop, nothing happened. Same with windows. Lately full kernel hangs happen less often, but smaller, temporary kernel hangs occur during video playback (and I guess this must happen on kernel level as whole system becomes unresponsive, including TTY switching and sysrq). Could this be nvidia?


#13

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.