SOLVED: Shutdown and hibernation make the drive disapper from BIOS

Elmario · 2023-09-20 23:23:34

OK: As a last resort, I now installed linux-lts.
And the console window on my desktop came up like this:

[ladmin@250g8 ~]$ systemctl hibernate 
[ladmin@250g8 ~]$

I guess this means hibernation worked (despite the journal saying no...)?

The journal is here:

sudo journalctl -b -1 | curl -F 'file=@-' 0x0.st
http://0x0.st/HOEY.txt

sudo journalctl -b -0 | curl -F 'file=@-' 0x0.st
http://0x0.st/HOEg.txt

Yes, I tried again. The desktop reloads in identical state to how I left it when hibernating. It's not just a reloaded session fom XFCE.
1. I dont' understand why there's still 'PM: Image not found (code -22)' in the journals.
2. As an LTS-kernel is basically outdated ... doesn't this mean my notebook will be incompatible with upcoming kernel releases? Do I need to report this issue somewhere like Github for the kernel-devs?

Last edited by Elmario (2023-09-20 23:38:12)

seth · 2023-09-21 07:21:42

Sep 21 01:20:02 250g8 systemd-logind[551]: The system will hibernate now!
Sep 21 01:20:02 250g8 systemd[1]: Starting Hibernate...
Sep 21 01:20:02 250g8 systemd-sleep[5186]: Entering sleep state 'hibernate'...
Sep 21 01:20:02 250g8 kernel: PM: hibernation: hibernation entry
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Marking nosave pages: [mem 0x00000000-0x00000fff]
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Marking nosave pages: [mem 0x0009f000-0x000fffff]
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Marking nosave pages: [mem 0x29ebf000-0x29ed2fff]
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Marking nosave pages: [mem 0x3b4ff000-0x40bfefff]
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Marking nosave pages: [mem 0x40c00000-0xffffffff]
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Basic memory bitmaps created
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Preallocating image memory
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Allocated 2382642 pages for snapshot
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Allocated 9530568 kbytes in 1.32 seconds (7220.12 MB/s)
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Creating image:
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Need to copy 2360258 pages
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Normal pages needed: 2360258 + 1024, available pages: 14332566
Sep 21 01:21:26 250g8 kernel: PM: hibernation: Basic memory bitmaps freed
Sep 21 01:21:26 250g8 kernel: PM: hibernation: hibernation exit

There's no "PM: Image not found (code -22)", hibernation worked flawlessly.
The LTS kernel isn't "outdated", it's the long term supported kernel.
At some point the main kernel becomes the new LTS, by then you'll want to have this fixed but the 6.5 kernels had rather a lot of issues anyway…

Elmario · 2023-09-21 09:18:51

Thank you very much for all the help!

Do you think I shoudl report this issue somewhere?

seth · 2023-09-21 15:53:34

Given the nvme parameters seem to have had(?) some impact, it's the likely offender.
If this doesn't go away w/ the next kernel updates you'll probably have to bisect this.

You can get older kernels out of the https://wiki.archlinux.org/title/Arch_Linux_Archive (keep lts and try 6.3, then, depending on the success, 6.2 or 6.4 and keep narrowing this down until you've two adjacent kernel releases good and bad)

Elmario · 2023-09-25 22:26:43

OK, I am on kernel 6.4 now and moved this to Grub's first line so it boots as default.
Woudl you suggest trying it including 'nvme_core.default_ps_max_latency_us=0 iommu=soft' or without?

Thank you very much for all the hep!

seth · 2023-09-26 06:33:05

Do you (have to) use them w/ the working LTS kernel?
You could try that w/o before - otherwise you need to keep the environment/config as consistent as possible (ie. use the same kernel parameters for every test) to narrow down on the first bad kernel.

Since it seems easier to trigger the problem w/o them, I'd first check the LTS kernel without them and if that still works, not use them for the test to get the results as fast as possible.

Elmario · 2023-10-02 06:31:01

seth wrote:

Do you (have to) use them w/ the working LTS kernel?

Kernel 6.4 was working fine without since nearly a week now (without the additional kernel parameters). Of course that's not too much compared to how irregularly the issue occured before.
I wanted to systematically try out all the Kernels from 6.4 to 6.5.1, because that's the itimespan when the issue occured. But I just saw the system has updated itself to 6.5.5 in between .. I don't know when exactly this update came, but I will see if it's working like this (and still without the additional Kernel parameters!). Otherwise I would have to suspend Kernel updates for encircling the Kernel version responsible for the problem).

Last edited by Elmario (2023-10-02 20:03:10)

Elmario · 2023-10-07 02:18:22

Hello!
So the issue occured again using the current kernel 6.5.5 it had automatically installed.
I now need to find how to stop the system from automatically updating the kernelfor more testing older kernel versions.

seth · 2023-10-07 06:40:08

https://wiki.archlinux.org/title/Pacman … g_upgraded
You'll need to add linux, linux-headers and pot OOT packages (eg. nvidia) there.
For the kernel specifically - and for a while - that's relatively safe.
Alternatively use the LTS kernel

Elmario · 2023-10-07 21:01:59

Thank you!
Yes, I still want to se the older kernels instead of lts for encircling the kernel version which raised the issue.

Elmario · 2023-10-24 11:09:01

Just as a memo:
I am at kernel 6.5.1 for week now and so far there's no problem. I will test this version one or two weeks longer. As the issue first occured with 6.5.3 I will soon have identified the responsible update.

Edit
20231027: Ist just happened, still using 6.5.1. Going back half the distance towards the previously working Kernel (6.4) now, which results in Kernel 6.4.6
(And I thought I was close with 6.5.1 working for weeks now and only 6.5.2 remaining in between .. it's sad )

Edit:
6.4.6 just broke.
Now trying 6.4.3.arch1-1-x86_64

20231113:
6.4.3.arch1-1-x86_64 just broke
Now trying linux-6.4.2.arch1-1-x86_64

Last edited by Elmario (2023-11-13 00:12:49)

Elmario · 2024-01-21 14:18:39

Still broken:
Linux 250g8 6.7.0-arch3-1 #1 SMP PREEMPT_DYNAMIC Sat, 13 Jan 2024 14:37:14 +0000 x86_64 GNU/Linux

Last working Kernel:
linux-6.4.2.arch1-1-x86_64

( https://archive.archlinux.org/packages/l/linux/ )
( pacman -U file:///var/cache/pacman/pkg/package-old_version.pkg.tar.type )
( https://archive.archlinux.org/packages/l/linux-headers/ )
( https://www.reddit.com/r/archlinux/comm … ux_kernel/ )

Last edited by Elmario (2024-04-08 23:45:49)

Arch Linux

#51 2023-09-20 23:23:34

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#52 2023-09-21 07:21:42

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#53 2023-09-21 09:18:51

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#54 2023-09-21 15:53:34

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#55 2023-09-25 22:26:43

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#56 2023-09-26 06:33:05

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#57 2023-10-02 06:31:01

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#58 2023-10-07 02:18:22

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#59 2023-10-07 06:40:08

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#60 2023-10-07 21:01:59

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#61 2023-10-24 11:09:01

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

#62 2024-01-21 14:18:39

Re: SOLVED: Shutdown and hibernation make the drive disapper from BIOS

Board footer