Booting fails after update restart. What to do next?

I have just restarted Tumbleweed as directed after my update today and the machine stalled at grub. I have the notice:-

GNU GRUB version 2.02~beta2
Minimal BASH-like line editing is supported. For the first word, TAB list possible command completions. Anywhere else TAB lists possible device or file completions.

grub>
Possible commands are:
[there follows a long list]

I tried “boot” and was told:-
error: you need to load the kernel first.

I have tried to re-start from cold but same result.

Please can somebody suggest what I do next please.

There’s a problem already. Tumbleweed grub has been at version 2.04 for a while (I think December, but I’m not sure of the date of that change).

So how are you booting (or how were you booting before now)?

Booting from the MBR?
Booting from a boot sector (usually in partition containing “/boot”)?
Booting with UEFI?

And how did you install Tumbleweed? In particular, did you use Gecko or did you use an actual openSUSE Tumbleweed iso?

Hi and thanks for coming to my rescue. As far as I recall booting was from MBR but honestly do not recall.

Installation as far as I recall was on line upgrade from openSUSE.

This system has been running a year or so with only sudo zypper dup upgrades, which vary in quantity but arrive most days, and without any issues until now.

I tried a live boot disk (rescatux from LinuxFormat magazine) but this failed to start the X server so I cannot yet examine the drive. Will try what other options I have here but in process of moving house so can’t find anything quickly, even my memory! I probably have some openSUSE DVDs around so will try other options but if you need more infor please let me know and thanks again for getting back to me.
Regards,
Budge

Was that earlier linux installed with Gecko linux?

I ask, because it was a flaw of Gecko. It installed booting, but never set the pointers as to how booting was installed. So when grub was updated, booting was not properly reinstalled.

In any case, your best bet might be to intall booting properly.

Boot the rescue system from install media. Be sure to use the correct version (32-bit or 64-bit, depending on what you have installed).

Mount the root partition of your Tumbleweed system at “/mnt”


mount --bind /dev /mnt/dev
mount --bind /proc /mnt/proc
mount --bind /sys /mnt/sys
chroot /mnt
mount -a

While still in that chroot session, run “yast” (at the command line). Get to the bootloader section. And check the settings. If need be, change something (increase the timeout by 1 second). And that should reinstall the bootloader.

Then exit from the chroot session, and reboot. See if that fixes the problem.

I know nothing about Gecko and am sure i always go to the openSUSE website.

Well I found a Tumbleweed DVD and am in the rescue system. Logged in as root and received the tty1:rescue: prompt and ran your first command.
received the following:-
mount.bin: //mnt/dev: mount point does not exist.

tried to see where I was at by running ls. Found .bash_history .gnupg .profile bin

had a look at root directory and tried you command but still same message so am at your mercy.

I think you missed this part:

Mount the root partition of your Tumbleweed system at “/mnt”

And thus skipped that. @nrickert did not provide an exact command for that because he does not know where your / (root) partition is. It must be something like

mnt /dev/sd???? /mnt

Hi Henk,
Many thanks, I have it now and am looking at the Boot Loader Settings. I have:-

Boot Loader
GRUB2 for EFI
[x] Enable Secure Boot
 ] Enable Trusted Boot Support
Protective MBR flag
do not change
Kernel Parameters
Optional Kernel Command Line Parameter
splash=silent resume=dev/disk/by-uuid/46f191b6-7472-49c0-bd98-75824020d947 quiet
CPU Mitigations
Auto
[X] Use graphical console
Console resolution  Autodetect by grub2 Console theme /boot/grub2/themes/openSUSE/theme.txt
Boot Loader Settings
Timeout in seconds 8  [x] Probe Foreign OS
Default Boot Section openSUSE Tumbleweed
Protect Boot Loader with Password
[X} Protect Entry Modification Only]

I cannot see anything wrong with what I see so will change the time setting to 10 seconds and save the new settings and see what happens when I exit and reboot. Will report back either way.
Many thanks,
Budge

No joy I am afraid, I get back to GNU GRUB version 2.02~beta2 and Minimal BASH-like line editing is supported.
HaveI missed something as I ha d hoped the GRUB version would have been changed, based on Nickert’s advice.
Where next I ask?

Just a thought but I recall there being a snapshot being mentioned a while ago but it might have been btrfs only. Whatever I hope my system is OK and it is just the booting that is now snafu. How can I revert from last upgrade? My rescue system seems to work even if I don’t know how to use it all. Grateful for more help please when you have time.
Regards,
Budge

Your Yast bootloader shows you using grub2 for EFI.

My best guess is that your BIOS is somehow now using BIOS booting instead of UEFI booting, and there is an old grub2 installed in the MBR.

Can you try getting into the BIOS settings, and check.

Hi, I am not sure what I am looking at with the BIOS settings. Very technical and no mention of UEFI but many IBM options. Could I just create a new MBR or change the booting to root and force a new installation of grub. I could also try a reset of bios but have no idea where that would get me. Need to look at the options offered in more detail. Will try again tomorrow night. Busy day tomorrow.
Regards and thanks again.
Budge

Yes, that could be done. But whether that will work depends on other details.

Can you go to rescue mode, and then run

parted -l

and post the output.

Hi,
I I do not know my way around using the rescue system, as you will have noticed earlier so please help me get into where I can run parted. Should I mount the drive as I did earlier? Will try anyhow but let me know if I am going wrong.

parted -l

will only show all your disks and their partitions, etc. Thus no need to mount anything. Just run any rescue (or stand-alone) system, login at the console and run it.

OK I have it now. I have mounted sda3 which may not be the right partition but I do have some information. Will have to type it in here so here goes:-
Model: IBM ServeRAID M5014 (SCSI)
Disk /dev/sda: 5994GB
Sector size (logical/physical): 512B/4096B
Partition Table: gpt
Disk Flags:

Number Start End Size File system Name Flags
1 1049kB 165MB 164MB fat16 primary boot, esp
2 165MB 2319MB 2155MB linux-swap(v1) primary msftdata
3 2319MB 45.3GB 43.0GB btrfs
4 45.3GB 595GB 550GB xfs primary msftdata
5 595GB 5983GB 5388GB xfs primary msftdata

Error: Can’t have a partition outside the disk!
Ignore/Cancel
Model: IBM SATA DEVICE 81Y3681 (scsi)
Disk /dev/sr0: 4684MB
Sector size (logical/physical): 2048B/2048B
Partition Table: unknown

Hope this makes sense to you. I assume sda3 might be root after all as it is the btrfs and size is correct. What happened to snapshots here or could it be they have just filled up the drive like on a previous occasion when I had to do a new install!!!
Grateful for any help you can give.
Regards,
Budge

Yes, it does.

What happened to snapshots here or could it be they have just filled up the drive like on a previous occasion when I had to do a new install!!!

Snapshots won’t help with this problem.

It looks to me that you have a UEFI system and installed for UEFI booting. But, somehow, your BIOS is trying to use MBR booting.

There is a known cause for this. Some BIOS use the “pmbr_boot” flag to decide how to boot. It is possible that somehow the pmbr_boot flag got set by something that you did with your computer.

I suggest that you google “pmbr_boot”. You will find an explanation of how to unset that flag (can be done with “parted” and no, you do not need to mount partitions for that).

Try that before we look into other possibilities. Because everything else is harder.

Hi and thanks. Raised my hopes for a bit there but when I look at the partition flags I do not have pmbr_boot flag set. All I have is

boot, esp

AFAIK esp refers to uefi and is as it should be.
Seems it is more complicated but I have your attention if you can continue.
(FYI one of my broadband connections just went down due I suspect to flooding. Fortunately backup still available!)
Regards,
Budge

What is the output of

echo $grub_cpu
echo $grub_platform

in GRUB?

Hi,
I have:-

x86_64

form your first command and

efi

from the second.

So it boots in EFI mode. What is output of “efibootmgr -v” when booting live? Also “ls -lR” from ESP would be certainly interesting.