You are not logged in.
Pages: 1
Hello everyone, I have a significant problem that unfortunately only happens to me with Arch.
At some point, the system appears as if it's going to shut down, but then the screens that I'm attaching start.
https://u.pcloud.link/publink/show?code … lXTSewiXy7
The problem is that the system cannot be turned off unless I hold down the power button for a long time, which is quite brutal.
If I press ctrl+Alt+F2 or F3, etc., the same screens continue to appear.
However, if I hold down ctrl+Alt+Del, the following screen appears, and then the same series of codes continues.
https://u.pcloud.link/publink/show?code … NPXmjyPn2k
I had a similar issue on Ubuntu that I resolved with "nvme_core.default_ps_max_latency_us=6000" in /etc/default/grub, which I also did here, but apparently it doesn't work or isn't the cause.
In Arch, the first entry in /etc/default/grub is "loglevel=3 quiet," and I don't remember if it was also in Ubuntu. I simply put "nvme_core.default_ps_max_latency_us=6000" right after it.
Am I doing something wrong, or is this not the issue at all, and the problem is something else?
Thank you all in advance.
GE75 Raider 8SF; 12 × Intel® Core™ i7-8750H CPU @ 2.20GHz; 31,1 GiB di RAM
KDE Plasma 5.27.0
Kernel Version 6.1.12-arch1-1 (64 bit)
X11 Nvidia (proprietary)
Offline
The problem is that the system cannot be turned off unless I hold down the power button for a long time, which is quite brutal.
For emergencies there is sysq+REISUB (give some time between letters, sync umount etc. are not instant operations).
But if that nvme drive is your only storage then it won't help either (if it's no longer operational, nothing can be unmounted anymore).
Your screenshots only show filesystem errors, but its unclear what triggered them, so your issue could be anything at this point.
With storage problems, any storage related errors don't make it into the logs anymore (can't be stored on nonfunctional storage after all). You could try switch to network logging to catch them.
Offline
Try
- "nvme_core.default_ps_max_latency_us=0", https://wiki.archlinux.org/title/Solid_ … ST_support
- "pcie_aspm=off" or to boot the lts kernel, https://bugs.archlinux.org/task/76918
- "iommu=soft", https://wiki.archlinux.org/title/Solid_ … nd_support
Edit: the system seems indeed to try to shutdown, is that voluntary or "spontanous"?
Is there a parallel windows?
Last edited by seth (2023-02-20 12:49:53)
Offline
For emergencies there is sysq+REISUB (give some time between letters, sync umount etc. are not instant operations).
I didn't know that, thank you very much!
You could try switch to network logging to catch them.
I don't know how to do it; I am ignorant on the subject.
Try
- "nvme_core.default_ps_max_latency_us=0", https://wiki.archlinux.org/title/Solid_ … ST_support
I had tried this before "nvme_core.default_ps_max_latency_us=6000"
- "pcie_aspm=off" or to boot the lts kernel, https://bugs.archlinux.org/task/76918
- "iommu=soft", https://wiki.archlinux.org/title/Solid_ … nd_support
I don't have the expertise to modify the kernel; I am also ignorant on the subject. How do I execute these two commands or instructions?
Edit: the system seems indeed to try to shutdown, is that voluntary or "spontanous"?
Is there a parallel windows?
Unfortunately, the issue occurs spontaneously.
Yes, there is a dual boot with Windows 11, and now that you mention it twice out of three times the issue occurred while running VirtualBox with Windows running, the other time it occurred out of nowhere.
By the way VirtualBox gave me some installation problems; in the end, in order to install it without errors, I had to install exactly these packages and in this exact order:
virtualbox-host-modules-arch
virtualbox-guest-iso
virtualbox
However, since it happened to me once without VirtualBox running, I'm afraid it may not be related to it.
GE75 Raider 8SF; 12 × Intel® Core™ i7-8750H CPU @ 2.20GHz; 31,1 GiB di RAM
KDE Plasma 5.27.0
Kernel Version 6.1.12-arch1-1 (64 bit)
X11 Nvidia (proprietary)
Offline
I had tried this before "nvme_core.default_ps_max_latency_us=6000"
0 != 6000 - learned that from Sesame Street.
How do I execute these two commands or instructions?
https://wiki.archlinux.org/title/Kernel_parameters
How do I execute these two commands or instructions?
3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.
Edit:
I don't know how to do it
https://wiki.archlinux.org/title/Genera … netconsole
Last edited by seth (2023-02-20 15:24:09)
Offline
0 != 6000 - learned that from Sesame Street.
Even though I'm using ChatGPT to translate, it seems like it's also making some mistakes.
What I meant to say was that before using '6000', I had already tried '0' and it had caused my system to freeze
So it seems to me that I have to modify /etc/default/grub
loglevel=3 quiet splash iommu=soft pcie_aspm=off nvme_core.default_ps_max_latency_us=0
In this way?
3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!)
I'm glad I learned this a long time ago, and yes, I understand why you're emphasizing it because in my entire life, 37 years so far, you're the only one who's telling me this instead of the other way around.
and reboot windows and linux twice for voodo reasons.
I don't know if you were making fun of me here (but it seems like you weren't), but I did it anyway (and if it were so
absurd)
Here I would have a NAS available that I am building with openmediavault, do you think it can be done or do I need a more "normal" machine?
GE75 Raider 8SF; 12 × Intel® Core™ i7-8750H CPU @ 2.20GHz; 31,1 GiB di RAM
KDE Plasma 5.27.0
Kernel Version 6.1.12-arch1-1 (64 bit)
X11 Nvidia (proprietary)
Offline
So it seems to me that I have to modify /etc/default/grub
loglevel=3 quiet splash iommu=soft pcie_aspm=off nvme_core.default_ps_max_latency_us=0
In this way?
You'll still have to run grub-mkconfig, but I highly recommend to experiment using the commandline editor in grub (select the boot, press "e" to edit it) first in order to have the change transient.
Notably because of
I had already tried '0' and it had caused my system to freeze
I'm glad I learned this a long time ago, and yes, I understand why you're emphasizing it because in my entire life, 37 years so far, you're the only one who's telling me this instead of the other way around.
I'm gonna say ChatGPT fucked up some more here.
=> https://www.deepl.com/translator
I don't know if you were making fun of me here
No. There're anecdotal reports for this to be required to "unconfuse" the ACPI table or whatever. It's "voodoo" because the process isn't understood.
do you think it can be done or do I need a more "normal" machine?
We'll first see whether we can take an easy victory w/ the kernel parameters reg. the known issues.
There's still time to worry about the netconsole afterwards.
Offline
Pages: 1