Recently I installed openSUSE Tumbleweed on my ThinkPad X390. Everything works so far, except Suspend to Disk. Sending it to hibernate looks good, but after I restart, it does not restore the old state but makes a fresh boot (unsaved files are gone, some applications are not there).
I have four partitions, one for uefi boot, one for the system (btrfs), one for /home (ext4), and swap (8 GiB). My computer has 8 GiB memory.
How to debug my suspend-to-disk issue?
Upload complete output of
journalctl --boot=-1 --no-pager --full
after hibernate/resume to https://susepaste.org. This may provide some starting points.
Hi @gruenich
Out of curiosity, does your ThinkPad X390 have touchscreen? If so, does the touchscreen work?
Thank you in advance
No, it does not have a touchscreen.
Thanks for the answer. I executed the command and pasted it: openSUSE Paste
You have a swap partition big enough for the memory image???
Yes, 8 GiB of memory and 8 GiB of swap. First I had a swap that was too small. I moved my swap partition to a new one, that is 8 GiB large.
# cat /proc/meminfo
MemTotal: 7897820 kB
and
# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
nvme0n1 259:0 0 476,9G 0 disk
├─nvme0n1p1 259:1 0 333,2M 0 part /boot/efi
├─nvme0n1p2 259:2 0 4G 0 part
├─nvme0n1p3 259:3 0 150G 0 part /var
│ /usr/local
│ /srv
│ /root
│ /opt
│ /boot/grub2/x86_64-efi
│ /boot/grub2/i386-pc
│ /
├─nvme0n1p4 259:4 0 314,6G 0 part /home
└─nvme0n1p5 259:5 0 8G 0 part [SWAP]
Well, at least it did try to hibernate. If you did not reboot yet, could you also upload the output of
journalctl --boot --no-pager --full
Did not reboot in the meantime:
resume= parameter is missing on kernel command line. It is present in dracut stored command line, but systemd is not aware about it.
Add resume=... parameter referring to your swap partition, reboot, then try hibernate/resume again. I cannot tell you the value of this parameter because you did not provide information allowing to guess it. Better is to use UUID or LABEL, not device name.
Ok, but where do I add the resume= paramenter? I tried Yast Bootloader, but then the reboot failed until I manually deleted the parameter. I tried the UUID and PARTUUID of /dev/nvme0n1p5, same result.
> blkid
/dev/nvme0n1p5: UUID="ec326a00-0b02-4ab5-88e2-3f8291d6391a" TYPE="swap" PARTUUID="c171f784-9a31-4c24-ac6c-8300f1d5fca6"
/dev/nvme0n1p3: UUID="8036cc12-b9ff-4ce8-906e-53ebbfbd8a8e" UUID_SUB="d807525b-7eea-4c7f-a6ca-34f6ad6f2a3e" BLOCK_SIZE="4096" TYPE="btrfs" PARTUUID="39402d22-1dc3-4fb3-818c-1f89e1417d54"
/dev/nvme0n1p1: UUID="E38F-6C70" BLOCK_SIZE="512" TYPE="vfat" PARTLABEL="EFI system partition" PARTUUID="0000044b-9c90-e394-bbb8-d80196080000"
/dev/nvme0n1p4: UUID="3c8cd2d3-3ac9-4dd5-beed-f56c0113d691" BLOCK_SIZE="4096" TYPE="ext4" PARTUUID="d05b2f33-6f68-49d4-9c0c-51db580132ad"
/dev/nvme0n1p2: UUID="1aa70c8b-ce66-4506-9ad9-53cee64eaa6c" TYPE="swap" PARTUUID="06917d3f-f3b2-4ae8-a7fe-800fb84fed1b"
You neither show what you added not errors when you added it so no comments are possible.
You are right! I opened Yast Bootloader and added to the optional kernel parameter splash=silent quiet security=apparmor the resume command resume=ec326a00-0b02-4ab5-88e2-3f8291d6391aor resume=c171f784-9a31-4c24-ac6c-8300f1d5fca6.
In both cases the booting stopped with the error message a start job is running followed by the UUID. I waited for a couple of minutes and aborted.
This should be
resume=UUID=ec326a00-0b02-4ab5-88e2-3f8291d6391a
or
resume=/dev/disk/by-uuid/ec326a00-0b02-4ab5-88e2-3f8291d6391a
That’s it. Hiberating works now. Thank you so much for debugging, spotting the issue, and guiding me to fix it!
Thank you for your answer @gruenich