[RESPONDED] Help debugging laptop crashes on Linux?

Hey All,

Apologies if this has been asked before, but I can’t seem to find a thread with an issue similar to mine.
I’m running a DIY FW laptop with Manjaro KDE, on an i5. Of late, and with increasing frequency, there’ll come a time when my laptop just ups and freezes, and then the caps lock blinks a few times before the laptop shuts down completely.

I unfortunately have no insight into what could be going wrong, or how I could debug this. I’ve tried looking at journalctl, but am again at a loss on how to parse the messages I see there. Any help could be appreciated, thank you!

I haven’t been monitoring the temperatures or CPU usage explicitly, but the fan (which is usually my proxy) hasn’t been too loud. Will monitor going forward, thanks!

1 Like

I haven’t seen these issues myself yet on Pop!_OS Linux, but there have been some posts mentioning similar freezing issues in Linux. I’ve listed some below for your convenience.

Since my post was linked in this thread, I would suggest trying to load Linux with nomodeset passed as an option from Grub. For reasons I don’t understand, loading the Intel graphics driver reliably results in freezes within 10 minutes; otherwise the system is stable. I cannot say that I’ve seen the blinking caps lock though. But try my suggestion and see if that helps.

Start with

journalctl -f -p err

to see if there’s anything obvious. The words panic or segmentation fault are usually big clues. Then look for relevant text nearby, and search for it online. Something like: “kernel panic” “blah blah type error content here” manjaro and see what comes up. The arch and manjaro forums might also be helpful.

sudo dmesg

is a good idea, too. Look for the same sort of stuff.

Good luck!

1 Like

Thank you @Starlight, @Jerry_Vinokurov and @Richard_Tango-Lowy ! I haven’t had any crashes again, but these are good steps to implement and look out for the next time it happens!

FWIW I’ve just experienced this for the first time, including the blinking caps lock. I’m thinking it might be a temperature/fan issue, since this is the first time it’s happened and I’ve been using it for a week with no issue. The main thing that’s changed is the weather ¯\_(ツ)_/¯.

I ended up solving my issue by reinstalling fedora on the internal ssd, rather than the thumb drive I was using previously. turns out the things can get pretty hot when in constant use, and fedora doesn’t like that very much.
I’d suggest you try a different drive if the problems keep happening. Sounds like you’ve got it well in hand however.

1 Like

I’ve just had this issue with a brand new SSD.
I’ve switched the SN850x 2TB for a SN850x 4TB, and I copied all the data over from the 2TB in the M.2 slot onto the 4TB in a USB enclosure.
Right after that I put the 4TB in the M.2 slot.
On rebooting, I had to take care of grub etc, and the following reboot, a few seconds after I logged into my Xfwm4 session (I had also time to open a xterm), the blinking-capslock freeze happened.

Now I have rebooted 3-4 more times and it didn’t happen again.
I’ve also used the laptop for several hours without incident.

I am worried of two possible causes:

  • either the SSD is not well seated (unlikely, because, on inspection, no copper of the connector is visible)
  • or the SSD is defective (when inserting it in the USB enclosure, I bumped laterally some of its protruding BGA chips onto the border of the enclosure, I really hope that didn’t weaken the BGA junctions)

The time the blinking freeze happened, I had also forgotten to change my swap partition UUID in fstab, but I don’t think that’s a problem (because with 64GB of RAM, I wouldn’t need any swap right after booting).

I have a cron entry that copies dmesg to a dmesg.old every minutes.
I saw nothing in it (which would indicate a SSD problem), but, thinking about it, I didn’t leave the machine a full minute before rebooting, so maybe it just hadn’t any chance to write the file.

A search on Google indicates that a frozen machine with a blinking caps lock is typical of a kernel panic.

I just hope it doesn’t happen again

Keep us posted. if it was working before the changing of the SSD and a clean install also does the same thing, they it may be worth testing the new SSD.

1 Like

@Matt_Hartley Thanks for the heads up!
I think I will go straight to the SSD testing, although I’ve also read that solder joint instability can be hard to predict and can worsen in time?

Which SSD testing utility would you recommend on Linux?

The problem has not reoccurred at all, it looks like it really was a one-off.
Also: this is not a SSD that I bought from Framework, so don’t worry for any responsibility nor any risk of RMA. It is still a SSD officially supported.

No problem! I think this is a good jumping off point considering what was experienced. Here are our recommended drives: What storage (SSD) parts are compatible with the Framework Laptop?