Does anyone have any issues with 6.5.7? I updated yesterday and could not boot. (1 of 2) A start job is running for /dev/mapper/root (30s / no limit) (2 of 2) A start job is running for /dev/disk/by-uuid/MY-UUID (30s / no limit)

Also during update my display resolution was set to 640x480 and I could not change it, so decided to reboot. I use a popular nowadays setup with LUKS encryption + unlock on TPM2, secure boot. I thought I messed with configs somewhere so started chasing that: changing configs and rebooting with no luck. The solution was to restore kernel 6.5.5 and everything booted back up without a hiccup. I am dreaded to see what happens during next kernel upgrade.

This is not asking for hep, more like a PSA if you have setup similar to mine, be aware


To those who stumble on this post in the future, I have found a solution that was in my case not knowing my system well enough. Since I decided to use Unified kernel images, I used mkinitcpio to compile those, but for some reason I used sbctl-bundle on top of that, which in itself is not any harm, just extra unnecessary work, and every single time I referenced an initramfs image from /boot which was an old one and was installed prior to me switching to UKI. When I read on Arch Wiki that I can delete those initramfs images from /boot - I deleted them, then had problems with sbctl bundle, and ONLY THEN it clicked - any new kernel install/upgrade doesn’t generate initramfs in /boot but instead directly in UKI.

This is also good news because sbctl author announced deprecation of the bundle feature in the future.

  • raunz@mander.xyz
    link
    fedilink
    arrow-up
    5
    ·
    edit-2
    1 year ago

    Yup I’m experiencing issues too. I can still boot, but after some time processes start dying

    /edit
    nvm it’s just my SSD dying…

  • AlmightySnoo 🐢🇮🇱🇺🇦@lemmy.world
    link
    fedilink
    arrow-up
    1
    arrow-down
    2
    ·
    edit-2
    1 year ago

    Have you tried to update to kernel 6.5.6 to bisect a bit further? This would help to find the commit that’s responsible for that, after which you can report the bug and usually get a swift response (and meanwhile you can then just revert that single commit). That’s what I did when I had keyboard issues with a recent Ryzen laptop (it also turned out the fix was surprisingly simple, just had to comment out a few lines that were inserted by the “faulty” commit and I even proposed my first patch back then, then an AMD guy did a more proper fix and it’s now in Linux stable).

    • Archy@lemmy.worldOP
      link
      fedilink
      arrow-up
      4
      ·
      1 year ago

      I have not - this is a work laptop, I was kind of stressed to fix it ASAP so that I can continue working. I might try that at home

      • thanksforallthefish@literature.cafe
        link
        fedilink
        arrow-up
        2
        ·
        1 year ago

        Just a suggestion, I always have both the latest and LTS versions installed concurrently. Have the current version as what boots and pacman will update both when you -Syu.

        Then if you have any problems just boot the LTS. Grub automatically offers it in the advanced menu or you can manually point to it

        • Archy@lemmy.worldOP
          link
          fedilink
          arrow-up
          1
          ·
          1 year ago

          I’m inclined to have both, I guess hiw does Debian based distros keep multiple versions? I know I have to clean them up because they are kept like 5 versions or so

          • thanksforallthefish@literature.cafe
            link
            fedilink
            arrow-up
            2
            ·
            1 year ago

            Would depend on the distro. Mint keeps a few older kernels and puts an option in grub to choose them. Debian family doesnt do rolling release in the same way as arch so it’s a different paradigm