Borked system boots to black screen, cant get to TTY

vecst

I'm not sure what happened. I probably had my solus up for a week, just suspending in-between sessions. I switched to my windows install for a couple of days and when I went back I was met with a black screen.

I tried to switch to a tty, and it would flash for a moment, then give me a blinking cursor. First I tried doing nomodeset, and I was met with the same issue.
I switched to my backup linux install chrooted into my install, rebuilt database, ran updates, checked for broken packages, reinstalled kernel, reinstalled nvidia drivers and switched to beta drivers, nothing seemed to fix the problem. I'm not sure what the issue is, so I'm not sure where to even begin to look. I'll provide any info you need, I just don't know what to provide.

Any help would be greatly appreciated. Thanks

brent

vecst I switched to my backup linux install chrooted into my install, rebuilt database, ran updates, checked for broken packages, reinstalled kernel, reinstalled nvidia drivers and switched to beta drivers, nothing seemed to fix the problem

https://help.getsol.us/docs/user/troubleshooting/boot-rescue/
did you have both /boot and /root mounted when you chroot-ed?

edit/if you switch out to nouveau drivers and it works than nvidia is a variable. you could rule it out.

I got nothing after this. I think you need to fix bootloader or Windows mugged you. When I had win and solus together it would beat the crap out of Solus regularly til I got rid of Win.

vecst

brent

Yes, I had both boot and mounted when I chrooted. I will try nouveau drivers and see what happens.
It doesn't feel like a boot loader thing to me, but I could be wrong. I've been running windows and solus side by side for over 10 years and never had an issue.
Like I said before when I try to switch to a tty I can see a login screen for half a second then it switches to a blinking cursor. I don't think I would be able to see that if it were a bootloader issue

pomon

Did you completely disable Windows and Secure Boot?
In the chroot, mount the boot partition using:
sudo clr-boot-manager mount-boot
Then check:

df -h | grep boot
ls /boot
bootctl status
lsblk -p

vecst

pomon
That is definitely disable on my windows. I don't really think its a bootloader issue. As I said before I can attempt to switch to a tty and I'll see the login message, then it switches to a blinking cursor.
I've never had a boot fail where I can attempt to access a tty, but I guess that's possible

vecst

pomon
df
/dev/nvme1n1p1 1022M 228M 795M 23% /boot

ls /boot
EFI loader solus-enroll-me.cer

bootctl status

systemd-boot not installed in ESP.
System:
      Firmware: UEFI 2.70 (American Megatrends 5.17)
 Firmware Arch: x64
   Secure Boot: enabled (user)
  TPM2 Support: no
  Boot into FW: supported

Current Boot Loader:
      Product: systemd-boot 254
     Features: ✓ Boot counting
               ✓ Menu timeout control
               ✓ One-shot menu timeout control
               ✓ Default entry control
               ✓ One-shot entry control
               ✓ Support for XBOOTLDR partition
               ✓ Support for passing random seed to OS
               ✓ Load drop-in drivers
               ✓ Support Type #1 sort-key field
               ✓ Support @saved pseudo-entry
               ✓ Support Type #1 devicetree field
               ✓ Enroll SecureBoot keys
               ✓ Retain SHIM protocols
               ✓ Boot loader sets ESP information
WARNING: The boot loader reports a different ESP UUID than detected (8d4bf089-63ef-4792-b0a0-0c6dc736a4ed vs.>
          ESP: /dev/disk/by-partuuid/8d4bf089-63ef-4792-b0a0-0c6dc736a4ed
         File: └─EFI/com.solus-project/grubx64.efi

Random Seed:
 System Token: not set
       Exists: no

Available Boot Loaders on ESP:
          ESP: /boot (/dev/disk/by-partuuid/9c582e75-81cf-4107-99b6-ecdd970fcdda)
         File: └─/EFI/BOOT/BOOTX64.EFI

Boot Loaders Listed in EFI Variables:
        Title: UEFI OS
           ID: 0x0001
       Status: active, boot-order
    Partition: /dev/disk/by-partuuid/2fdb4d46-12a6-4940-aae8-2768b9d47625
         File: └─/EFI/BOOT/BOOTX64.EFI

        Title: UEFI OS
           ID: 0x0002
       Status: active, boot-order
    Partition: /dev/disk/by-partuuid/8d4bf089-63ef-4792-b0a0-0c6dc736a4ed
         File: └─/EFI/BOOT/BOOTX64.EFI

        Title: Windows Boot Manager
           ID: 0x0000
       Status: active, boot-order
    Partition: /dev/disk/by-partuuid/6c5bb501-ac80-4b50-9e92-a1299805a757
         File: └─/EFI/MICROSOFT/BOOT/BOOTMGFW.EFI

        Title: Solus Linux Bootloader
           ID: 0x0003
       Status: active, boot-order
    Partition: /dev/disk/by-partuuid/8d4bf089-63ef-4792-b0a0-0c6dc736a4ed
         File: └─/EFI/COM.SOLUS-PROJECT/BOOTLOADERX64.EFI

        Title: UEFI OS
           ID: 0x0009
       Status: active, boot-order
    Partition: /dev/disk/by-partuuid/9c582e75-81cf-4107-99b6-ecdd970fcdda
         File: └─/EFI/BOOT/BOOTX64.EFI

        Title: Solus Linux Bootloader
           ID: 0x0004
       Status: active, boot-order
    Partition: /dev/disk/by-partuuid/9c582e75-81cf-4107-99b6-ecdd970fcdda
         File: └─/EFI/COM.SOLUS-PROJECT/BOOTLOADERX64.EFI

Boot Loader Entries:
        $BOOT: /boot (/dev/disk/by-partuuid/9c582e75-81cf-4107-99b6-ecdd970fcdda)
        token: solus

Default Boot Loader Entry:
         type: Boot Loader Specification Type #1 (.conf)
        title: Solus 4.7 Endurance
           id: Solus-current-6.14.6-319.conf
       source: /boot//loader/entries/Solus-current-6.14.6-319.conf
        linux: /boot//EFI/com.solus-project/kernel-com.solus-project.current.6.14.6-319
       initrd: /boot//EFI/com.solus-project/initrd-com.solus-project.current.6.14.6-319
               /boot//EFI/com.solus-project/initrd-com.solus-project.current.6.14.6-319.nvidia
               /boot//EFI/com.solus-project/freestanding-com.solus-project.nvidia-firmware
      options: root=PARTUUID=386673df-c1ab-4469-b740-4c7cd95abf21 rd.vconsole.keymap=us rd.vconsole.font=ter->

lsblk -p

/dev/loop0         7:0    0 308.2M  1 loop 
/dev/loop1         7:1    0   5.5M  1 loop 
/dev/loop2         7:2    0  63.7M  1 loop 
/dev/loop3         7:3    0     4K  1 loop 
/dev/loop4         7:4    0  49.3M  1 loop 
/dev/loop5         7:5    0  91.7M  1 loop 
/dev/loop6         7:6    0   5.5M  1 loop 
/dev/loop7         7:7    0 308.6M  1 loop 
/dev/loop8         7:8    0  63.8M  1 loop 
/dev/loop9         7:9    0  44.4M  1 loop 
/dev/sda           8:0    0 465.8G  0 disk 
├─/dev/sda1        8:1    0   100M  0 part 
├─/dev/sda2        8:2    0    16M  0 part 
├─/dev/sda3        8:3    0 465.1G  0 part 
└─/dev/sda4        8:4    0   530M  0 part 
/dev/sdb           8:16   0   1.8T  0 disk 
├─/dev/sdb1        8:17   0    16M  0 part 
├─/dev/sdb2        8:18   0 886.4G  0 part 
├─/dev/sdb3        8:19   0 976.4G  0 part 
└─/dev/sdb4        8:20   0   210M  0 part 
/dev/sdc           8:32   0 111.8G  0 disk 
├─/dev/sdc1        8:33   0   489M  0 part 
├─/dev/sdc2        8:34   0   3.7G  0 part [SWAP]
└─/dev/sdc3        8:35   0 107.6G  0 part 
/dev/nvme0n1     259:0    0 953.9G  0 disk 
├─/dev/nvme0n1p1 259:1    0    16M  0 part 
└─/dev/nvme0n1p2 259:2    0 953.9G  0 part 
/dev/nvme1n1     259:3    0 465.8G  0 disk 
├─/dev/nvme1n1p1 259:4    0     1G  0 part /boot
├─/dev/nvme1n1p2 259:5    0   456G  0 part /
└─/dev/nvme1n1p3 259:6    0   8.8G  0 part

brent

vecst I've been running windows and solus side by side for over 10 years and never had an issue.

wow (I was never the lucky one who could ever get away with that).

vecst systemd-boot not installed in ESP.

https://discourse.nixos.org/t/could-not-find-any-previously-installed-systemd-boot/15627/9

you shouldn't have that error. couple bootctl commands took care of it ^^ in that instances.
little by little.

vecst

brent https://discourse.nixos.org/t/could-not-find-any-previously-installed-systemd-boot/15627/9

It seems the selected solution was to restart multiple times? I can promise you I've tried that. I'll look into running bootctl install and see what happens.

Do you think the TTY would be able to show the login if the boot was messed up? I'm just trying to wrap my head around it. I've used linux for a long time and never seen a tty flash the prompt then switch the a blinking cursor. I can spam ctrl + alt + f3 and it will repeatedly flash the login prompt.

stocc

brent systemd-boot not installed in ESP.

I have that same message on my fully functioning system, so it may be a red herring.

Staudey

This sounds like it might be an NVIDIA issue. Is your system fully up-to-date? Maybe there is a mismatch between kernel and nvidia driver packages (one of the more common causes of this blinking cursor issue).

Please post the output of uname -a and eopkg info nvidia-glx-driver-common (or whatever different driver branch you use)

vecst

Staudey

I am fully updated. It's a pretty new install, just a few months old, I think it was back in February or March.

uname -a returns the same result whether I'm chrooted or not. I'm guessing from mounting proc. I don't think it will be useful, but just in case
Linux tundra 6.6.66-262.lts #1 SMP PREEMPT_DYNAMIC 2024-12-18 x86_64 GNU/Linux

Looking in /boot/loader/entries I have Solus-current-6.14.6-319.conf
and catting that give me

title Solus 4.7 Endurance
linux /EFI/com.solus-project/kernel-com.solus-project.current.6.14.6-319
initrd /EFI/com.solus-project/initrd-com.solus-project.current.6.14.6-319
initrd /EFI/com.solus-project/initrd-com.solus-project.current.6.14.6-319.nvidia
initrd /EFI/com.solus-project/freestanding-com.solus-project.nvidia-firmware
options root=PARTUUID=386673df-c1ab-4469-b740-4c7cd95abf21 rd.vconsole.keymap=us rd.vconsole.font=ter-v32b quiet splash rw nvidia-drm.modeset=1 resume=UUID=4b29a3ac-a274-4ee8-8e51-b8b631b4713e

eopkg info nvidia-glx-driver-current

Name                : nvidia-glx-driver-current, version: 570.169, release: 572
Summary             : NVIDIA Binary Driver (Current Kernel)
Description         : NVIDIA Binary Driver (Current Kernel)
Licenses            : EULA
Component           : kernel.drivers
Dependencies        : nvidia-glx-driver-common 
Distribution        : Solus, Dist. Release: 1
Architecture        : x86_64, Installed Size: 144.74 MB
Reverse Dependencies: 

Package found in Solus repository:
Name                : nvidia-glx-driver-current, version: 570.169, release: 572
Summary             : NVIDIA Binary Driver (Current Kernel)
Description         : NVIDIA Binary Driver (Current Kernel)
Licenses            : EULA
Component           : kernel.drivers
Dependencies        : nvidia-glx-driver-common 
Distribution        : Solus, Dist. Release: 1
Architecture        : x86_64, Installed Size: 144.74 MB, Package Size: 144.40 MB
Reverse Dependencies:

If there's anything else I can get you let me know. Thanks for your help

brent

stocc I have that same message on my fully functioning system, so it may be a red herring.

lotta red herrings for sure

banger

vecst uname -a returns the same result whether I'm chrooted or not. I'm guessing from mounting proc. I don't think it will be useful, but just in case
Linux tundra 6.6.66-262.lts #1 SMP PREEMPT_DYNAMIC 2024-12-18 x86_64 GNU/Linux

Looking in /boot/loader/entries I have Solus-current-6.14.6-319.conf
and catting that give me

I am no expert but the above looks odd. You seem to be running lts kernel but the boot files you catted are current. Something doesn't look right with this config.

Staudey

vecst uname -a returns the same result whether I'm chrooted or not. I'm guessing from mounting proc. I don't think it will be useful, but just in case

Ah sorry, I misread. I thought you were able to actually enter the TTY (for more than a second ^^).

So from those boot entries it seems you have a mismatch between kernel version and nvidia driver version. release 572 of the nvidia driver was built for kernel 6.14.11. It's weird that you can't enter TTY though, ususally that should work even with a mismatched driver package.
Such a mismatch often suggests having insufficient space on your ESP or an incomplete update. As a first measure you could try removing the nvidia driver and seeing if it boots correctly then.

vecst

banger
My backup linux is my old solus which is running that LTS kernel. That's how I knew the uname output was not from the chroot, but from my system and it was the same as a none chrooted terminal.

I'm guessing when I mount --types proc /proc /target/proc it adds my proc to target. I think I should have done mount --types proc none /target/proc but I'm not sure. I don't do much in the chroot universe, so I just followed the guide.

vecst

Staudey

No worries so the ESP is 1gb large, and only 23% is being used., surely almost 800MB is enough.
I tried uninstalling the nvidia driver and it's still behaving the same way. I tried booting with and without nomodeset

Here's a video of dmesg on boot and where it hangs. before it hangs the monitor starts flashing. That's not an effect from the camera.

vecst

Staudey So from those boot entries it seems you have a mismatch between kernel version and nvidia driver version. release 572 of the nvidia driver was built for kernel 6.14.11. It's weird that you can't enter TTY though, ususally that should work even with a mismatched driver package.

That was it. I looked into my boot partition bit closer and the kernel was there, just not being used. I ran clr-boot-manager and set the new kernel, rebooted to a black screen, then remembered I uninstalled drivers, went back and installed drivers anddd I'm in!

Thanks so much for the help

Staudey

vecst No worries so the ESP is 1gb large, and only 23% is being used., surely almost 800MB is enough.

Well, the fact of the matter is that the nvidia driver was apparently updated, but the kernel boot entry apparently was not. So something went wrong there. If it's not lack of space, then I don't know what the next likely candidate is.

stekte

A friend of mine had the same issue on his laptop a few months ago. He had inastalled
Arch + Gnome .
GPU = amd
The problem was solved after he disabled gdm.

vecst

Staudey
I'm not sure honestly, before it happened the last thing I installed was chrome to test out some firefox specific issue I was having. I know I wasn't updated at that point. I didn't run an update until I chrooted hoping that would fix the issue. In the past I've had to do that, but usually I can access it through a TTY.
What can I do to fix it? Do I need to make a new ESP and make it larger?