[Solved] Help with diagnosing sudden system crashes

lennarttw · 2025-04-22 09:26:42

Hi,

I am experiencing a recurring issue when initiating a large download via Steam.
After a few minutes, the system suddenly crashes, and it does so consistently.

I checked the journal but I see no messages at all at the time of the crash, the system just shuts down instantly.

I am using a Asus laptop with Intel Lunar Lake hardware which is pretty new, so I think there might be some compatibility issue here.
But as there are no logs to give me some direction to look in, I am asking for help.

What can I do to further diagnose the issue? Is there any way to get more information from the system when this issue occurs?

Thanks for your time and understanding.

Edit: I am using `journalctl -b -1` to inspect the journal, I'm not sure if there is some way to get more output.

Last edited by lennarttw (2025-05-05 10:20:39)

twelveeighty · 2025-04-22 14:17:09

The last couple of months there have been several (at least 3-4 I can remember) "sudden shutdown/reboot" problems reported on the forums. See if you can locate those posts and if your symptoms are similar. Some have had to change some low level power/current settings, iirc.

seth · 2025-04-22 16:03:45

I'm not sure if there is some way to get more output.

Unless your in the adm/wheel group, "sudo", but a cold reboot will prevent the tail of the journal being synced to disk.

Unless there's a parallel windows (3rd link below) yuo're looking at a hardware problem (underpowered, overheated, bad RAM or CPU)

I am experiencing a recurring issue when initiating a large download via Steam.
After a few minutes, the system suddenly crashes, and it does so consistently.

How "large" is large?
Can we take steam out of the equation and control over the download destination? Does it crash if you download an equally "large" thing™ w/ wget/curl? Also if you're downloading it into /dev/null?
If not, keep an eye on the temperatures while running the steam download.
Do you haveother network devices (ie. can you switch from wifi to rj45 or vv.)? https://wiki.archlinux.org/title/Tethering ?

lennarttw · 2025-04-22 20:40:31

seth wrote:

I'm not sure if there is some way to get more output.
Unless your in the adm/wheel group, "sudo", but a cold reboot will prevent the tail of the journal being synced to disk.
Unless there's a parallel windows (3rd link below) yuo're looking at a hardware problem (underpowered, overheated, bad RAM or CPU)
I am experiencing a recurring issue when initiating a large download via Steam.
After a few minutes, the system suddenly crashes, and it does so consistently.
How "large" is large?
Can we take steam out of the equation and control over the download destination? Does it crash if you download an equally "large" thing™ w/ wget/curl? Also if you're downloading it into /dev/null?
If not, keep an eye on the temperatures while running the steam download.
Do you haveother network devices (ie. can you switch from wifi to rj45 or vv.)? https://wiki.archlinux.org/title/Tethering ?

Thanks for your feedback

By large I just mean that it takes several minutes to complete, generating some load on the system and ramping up the fans.
Every 5 minutes or so, it crashes.

I also noticed today it happens when I play games too, so long as they put the system under enough load (didn't happen on any 2D games).
Happens at roughly the same interval.

I kept an eye on the temps via sensor and they didn't go above 80, while the critical temp is 100.

Because I now noticed it doesn't happen only during downloads, I am pretty sure that it is not related to storage or networking. And I would expect some warnings in the journal if the issue was overheating, surely it would log a message when it gets unusually high before triggering a shutdown no?

Still not really any closer to a solution but they are some good pointers none the less.
It is possible it may be related to some power settings but the BIOS does not have any parameters that influence this. So the only variables that can be tuned are the kernel parameters I think.

Last edited by lennarttw (2025-04-22 20:42:29)

seth · 2025-04-22 21:59:01

And I would expect some warnings in the journal if the issue was overheating, surely it would log a message when it gets unusually high before triggering a shutdown no?

If the cpu went into an emergency reboot, nothing gets logged on disk, you might see MCE errors in the next boot.

By large I just mean that it takes several minutes to complete, generating some load on the system and ramping up the fans.
Every 5 minutes or so, it crashes.

ntfs? encrypted partition? Downloading a file from the interwebz should rather not be "generating some load on the system and ramping up the fans" - CPU or GPU and in what process?
But you're apparently rebooting in high-energy contexts.

I also noticed today it happens when I play games too, so long as they put the system under enough load (didn't happen on any 2D games).

Asus laptop with Intel Lunar Lake hardware

So unlikely underdimensioned PSU or missing power supply connection, do you in any way shape or form overclock the system or use XMP?

evorster · 2025-04-30 11:31:34

Just as a precaution, I would take apart the system and blow out all the dust. Even though the CPU temps did not rise that much does not mean that another part of the system was overheating.

lennarttw · 2025-05-01 11:42:45

evorster wrote:

Just as a precaution, I would take apart the system and blow out all the dust. Even though the CPU temps did not rise that much does not mean that another part of the system was overheating.

A good suggestion but as this is a brand new laptop I highly doubt that is the issue.

lennarttw · 2025-05-01 11:48:00

seth wrote:

And I would expect some warnings in the journal if the issue was overheating, surely it would log a message when it gets unusually high before triggering a shutdown no?
If the cpu went into an emergency reboot, nothing gets logged on disk, you might see MCE errors in the next boot.

By large I just mean that it takes several minutes to complete, generating some load on the system and ramping up the fans.
Every 5 minutes or so, it crashes.
ntfs? encrypted partition? Downloading a file from the interwebz should rather not be "generating some load on the system and ramping up the fans" - CPU or GPU and in what process?
But you're apparently rebooting in high-energy contexts.
I also noticed today it happens when I play games too, so long as they put the system under enough load (didn't happen on any 2D games).
Asus laptop with Intel Lunar Lake hardware
So unlikely underdimensioned PSU or missing power supply connection, do you in any way shape or form overclock the system or use XMP?

I have not changed any settings in the BIOS. When I looked, it it doesn't have a whole lot of settings either.
I imagine you can use XMP anyway since the RAM is directly integrated into the SoC chip. So it's not third party RAM.

Downloading the file caused some system load because Steam compresses the content in 1MB blobs and reconstructs the original files from it.
I also tried to do something else, compile the linux kernel (using make -j8) and the same issue occurred after a few minutes.

I also ran memory tests from the BIOS and they passed successfully.
Perhaps at this point (even though it sucks) I might have to put Windows back on and see if I can reproduce the issue there. To determine for sure it is a firmware/driver issue and not hardware related.
And then maybe I need to contact ASUS or Intel?

seth · 2025-05-01 13:12:41

I also ran memory tests from the BIOS and they passed successfully.

https://wiki.archlinux.org/title/Stress … MemTest86+ - the tests are measured in days, at least run it overnight (16+ h) to be sure it's not the RAM.

For a ryzen CPU I'd refer to PBO (which is a notorious issue there) - can you trigger this w/ https://wiki.archlinux.org/title/Stress_testing#stress?
(parallel kernel builds will charge the CPU but also eat up a lot of RAM)

lennarttw · 2025-05-05 10:20:13

seth wrote:

I also ran memory tests from the BIOS and they passed successfully.
https://wiki.archlinux.org/title/Stress … MemTest86+ - the tests are measured in days, at least run it overnight (16+ h) to be sure it's not the RAM.
For a ryzen CPU I'd refer to PBO (which is a notorious issue there) - can you trigger this w/ https://wiki.archlinux.org/title/Stress_testing#stress?
(parallel kernel builds will charge the CPU but also eat up a lot of RAM)

I did some further testing using the stress testing tool.

At this point I think that the issue only occurs when the lid is closed, as I often do while working. Choosing to use my external monitors instead.
I think the lid might be blocking some airflow, or perhaps the display itself is overheating.
I guess I should ask ASUS support about it.

Thanks for the help

Arch Linux

#1 2025-04-22 09:26:42

[Solved] Help with diagnosing sudden system crashes

#2 2025-04-22 14:17:09

Re: [Solved] Help with diagnosing sudden system crashes

#3 2025-04-22 16:03:45

Re: [Solved] Help with diagnosing sudden system crashes

#4 2025-04-22 20:40:31

Re: [Solved] Help with diagnosing sudden system crashes

#5 2025-04-22 21:59:01

Re: [Solved] Help with diagnosing sudden system crashes

#6 2025-04-30 11:31:34

Re: [Solved] Help with diagnosing sudden system crashes

#7 2025-05-01 11:42:45

Re: [Solved] Help with diagnosing sudden system crashes

#8 2025-05-01 11:48:00

Re: [Solved] Help with diagnosing sudden system crashes

#9 2025-05-01 13:12:41

Re: [Solved] Help with diagnosing sudden system crashes

#10 2025-05-05 10:20:13

Re: [Solved] Help with diagnosing sudden system crashes

Board footer