Cannot install ROME plasma 6 X11 from live mode

ben79 · August 22, 2024, 12:54pm

The Resources Index has a lot of valuable information like this.

Translated the output of your sudo update-grub2, this part looks good:

Creating GRUB configuration file…
 Found theme: /boot/grub2/themes/OpenMandriva/theme.txt
 Found background: /boot/grub2/themes/OpenMandriva/background.png
 Found Linux image: /boot/vmlinuz-6.10.0-desktop-1omv2490
 Found initial memory image: /boot/initrd-6.10.0-desktop-1omv2490.img
 Found memtest image: /boot/memtest.bin
 Warning: the OS probe will be run to detect other bootable partitions.
 Its output will be used to detect the bootable binaries they contain and create new boot entries.

This part I have not seen before:

ERROR: sil: invalid metadata checksum in area 2 on /dev/dm-0
Adding boot menu entry for UEFI firmware configuration …
Adding boot menu entry for EFI firmware configuration
done

Edit: On my ROME P6 hw system I do not have a /dev/dm-0. Also adding boot menu entry for UEFI and EFI looks sketchy. But I am not a developer, user knowledge lever here.

So if you have problems with booting there is this article. I had to use this myself about 6 days ago due to a mistake I made.

ben79 · August 22, 2024, 1:01pm

@olivier we do appreciate your providing logs and other info to help us help you.

rugyada · August 22, 2024, 2:29pm

Thank you for the feedback.

Yes, definitely.

stephane · August 24, 2024, 6:44am

on install , option for UEFI was not good
you have this : UEFI-[Legacy]: Dell v: 1.35.3 date: 07/12/2022

that mean , this is NOT full UEFI , in this case , your disk are in GPT , BUT
with legacy , the command install failback to “bios grub” mode ( because this is NOT UEFI )

Aug 20 14:40:02 omv-3241 root[1845]: os-prober: debug: running /usr/libexec/os-probes/50mounted-tests on /dev/sda1
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: /dev/sda1 is a LUKS partition; skipping
Aug 20 14:40:02 omv-3241 root[1845]: os-prober: debug: running /usr/libexec/os-probes/50mounted-tests on /dev/sdb1
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: /dev/sdb1 type not recognised; skipping
Aug 20 14:40:02 omv-3241 root[1845]: os-prober: debug: running /usr/libexec/os-probes/50mounted-tests on /dev/sdb2
Aug 20 14:40:02 omv-3241 sh[1676]: 13:40:02 [6]: void Calamares::RequirementsChecker::reportProgress()
Aug 20 14:40:02 omv-3241 sh[1676]:     Remaining modules: 1
Aug 20 14:40:02 omv-3241 sh[1676]:      "partition"
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: mounted using GRUB fat filesystem driver
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/05efi
Aug 20 14:40:02 omv-3241 root[1845]: 05efi: debug: Not on UEFI platform
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/10freedos
Aug 20 14:40:02 omv-3241 root[1845]: 10freedos: debug: /dev/sdb2 is a FAT partition (mounted by GRUB)
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/10qnx
Aug 20 14:40:02 omv-3241 root[1845]: 10qnx: debug: /dev/sdb2 is not a QNX4 partition: exiting
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/20macosx
Aug 20 14:40:02 omv-3241 macosx-prober[1922]: debug: /dev/sdb2 is not an HFS+ partition: exiting
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/20microsoft
Aug 20 14:40:02 omv-3241 root[1845]: 20microsoft: debug: /dev/sdb2 is a FAT partition (mounted by GRUB)
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/30utility
Aug 20 14:40:02 omv-3241 root[1845]: 30utility: debug: /dev/sdb2 is a FAT partition (mounted by GRUB)
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/40lsb
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/70hurd
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/80minix
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/83haiku
Aug 20 14:40:02 omv-3241 root[1845]: 83haiku: debug: /dev/sdb2 is not a BeFS partition: exiting
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/90linux-distro
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/90solaris
Aug 20 14:40:02 omv-3241 root[1845]: 50mounted-tests: debug: running subtest /usr/libexec/os-probes/mounted/efi
....
Aug 20 14:40:02 omv-3241 sh[1676]:     Setting up actions for "/dev/sda" with 0 entries.
Aug 20 14:40:02 omv-3241 sh[1676]:     ..  Can not resize "/dev/sda1" , filesystem "luks" does not support resize.
Aug 20 14:40:02 omv-3241 sh[1676]:     ..  Checking if "/dev/sda1" can be replaced.
Aug 20 14:40:02 omv-3241 sh[1676]:     ..  Partition "/dev/sda1" authorized for replace install.
Aug 20 14:40:02 omv-3241 sh[1676]:     .. contains replaceable "/dev/sda1"
Aug 20 14:40:02 omv-3241 sh[1676]:     ..  Can not resize "/dev/sda" , filesystem "unknown" does not support resize.
Aug 20 14:40:02 omv-3241 sh[1676]:     ..  Checking if "/dev/sda" can be replaced.
Aug 20 14:40:02 omv-3241 sh[1676]:     ..  NO, insufficient storage
Aug 20 14:40:02 omv-3241 sh[1676]:      Required  storage B: 9126805504 "(8GiB)"
Aug 20 14:40:02 omv-3241 sh[1676]:      Available storage B: 7564800 "(0GiB)"
...

Aug 20 14:40:40 omv-3241 sh[1676]:     InstallChoice Config::Erase "erase"
Aug 20 14:40:40 omv-3241 sh[1676]: 14:40:40 [6]: PartitionTable *CreatePartitionTableJob::createTable()
Aug 20 14:40:40 omv-3241 sh[1676]:     CreatePartitionTableJob::createTable trying to make table for device "/dev/sda"
Aug 20 14:40:40 omv-3241 sh[1676]: 14:40:40 [6]: void PartitionCoreModule::dumpQueue() const
Aug 20 14:40:40 omv-3241 sh[1676]:     # Queue:
Aug 20 14:40:40 omv-3241 sh[1676]:     .. ## Device: "/dev/sda"
Aug 20 14:40:40 omv-3241 sh[1676]:     .. - CreatePartitionTableJob
Aug 20 14:40:40 omv-3241 sh[1676]:     .. - CreatePartitionJob
Aug 20 14:40:40 omv-3241 sh[1676]:     .. - SetPartFlagsJob
Aug 20 14:40:40 omv-3241 sh[1676]: 14:40:40 [6]: void ChoicePage::updateActionChoicePreview(InstallChoice)
Aug 20 14:40:40 omv-3241 sh[1676]:     Updating partitioning preview widgets.
....

Aug 20 14:41:07 omv-3241 sh[1676]: 14:41:07 [6]: PartitionTable *CreatePartitionTableJob::createTable()
Aug 20 14:41:07 omv-3241 sh[1676]:     CreatePartitionTableJob::createTable trying to make table for device "/dev/sda"
Aug 20 14:41:07 omv-3241 sh[1676]: 14:41:07 [6]: void PartitionCoreModule::setBootLoaderInstallPath(const QString &)
Aug 20 14:41:07 omv-3241 sh[1676]:     PCM::setBootLoaderInstallPath "/dev/sda"
Aug 20 14:41:07 omv-3241 sh[1676]: 14:41:07 [6]: void PartitionCoreModule::setBootLoaderInstallPath(const QString &)
Aug 20 14:41:07 omv-3241 sh[1676]:     PCM::setBootLoaderInstallPath "/dev/sda"
Aug 20 14:41:07 omv-3241 sh[1676]: 14:41:07 [6]: void PartitionCoreModule::dumpQueue() const
Aug 20 14:41:07 omv-3241 sh[1676]:     # Queue:
Aug 20 14:41:07 omv-3241 sh[1676]:     .. ## Device: "/dev/sda"
Aug 20 14:41:07 omv-3241 sh[1676]:     .. - CreatePartitionTableJob
Aug 20 14:41:07 omv-3241 sh[1676]:     .. - CreatePartitionJob
Aug 20 14:41:07 omv-3241 sh[1676]:     .. - SetPartFlagsJob
Aug 20 14:41:10 omv-3241 irqbalance[759]: 
                                          
                                          
                                          -----------------------------------------------------------------------------
Aug 20 14:41:10 omv-3241 sh[1676]: 14:41:10 [6]: void PartitionCoreModule::setBootLoaderInstallPath(const QString &)
Aug 20 14:41:10 omv-3241 sh[1676]:     PCM::setBootLoaderInstallPath "/dev/sda"
...
Aug 20 14:42:02 omv-3241 sh[1676]:     Checking LVM use on 1 devices
Aug 20 14:42:02 omv-3241 sh[1676]: 14:42:02 [6]: QVariantList FillGlobalStorageJob::createPartitionList() const
Aug 20 14:42:02 omv-3241 sh[1676]:     Building partition information map
Aug 20 14:42:02 omv-3241 sh[1676]:     .. partitions on "/dev/sda"
Aug 20 14:42:02 omv-3241 sh[1676]:     ..  mapping for "" "New Partition"
Aug 20 14:42:02 omv-3241 sh[1676]:      partlabel :   "root"
Aug 20 14:42:02 omv-3241 sh[1676]:      partition-uuid (partuuid) :   "PartUUID$af3e48b4"
Aug 20 14:42:02 omv-3241 sh[1676]:      parttype :   ""
Aug 20 14:42:02 omv-3241 sh[1676]:      partattrs :   "0"
Aug 20 14:42:02 omv-3241 sh[1676]:      mountPoint: :   "/"
Aug 20 14:42:02 omv-3241 sh[1676]:      fs: :   "ext4"
Aug 20 14:42:02 omv-3241 sh[1676]:      fsName :   "luks"
Aug 20 14:42:02 omv-3241 sh[1676]:      filesystem-uuid (uuid) :   "FSUUID$4a49fe31"
Aug 20 14:42:02 omv-3241 sh[1676]:      claimed :   "true"
Aug 20 14:42:02 omv-3241 sh[1676]:      luksMapperName: :   ""
Aug 20 14:42:02 omv-3241 sh[1676]: 14:42:02 [6]: QVariantList FillGlobalStorageJob::createPartitionList() const
Aug 20 14:42:02 omv-3241 sh[1676]:     Building partition information map
Aug 20 14:42:02 omv-3241 sh[1676]:     .. partitions on "/dev/sda"
Aug 20 14:42:02 omv-3241 sh[1676]:     ..  mapping for "" "New Partition"
Aug 20 14:42:02 omv-3241 sh[1676]:      partlabel :   "root"
Aug 20 14:42:02 omv-3241 sh[1676]:      partition-uuid (partuuid) :   "PartUUID$af3e48b4"
Aug 20 14:42:02 omv-3241 sh[1676]:      parttype :   ""
Aug 20 14:42:02 omv-3241 sh[1676]:      partattrs :   "0"
Aug 20 14:42:02 omv-3241 sh[1676]:      mountPoint: :   "/"
Aug 20 14:42:02 omv-3241 sh[1676]:      fs: :   "ext4"
Aug 20 14:42:02 omv-3241 sh[1676]:      fsName :   "luks"
Aug 20 14:42:02 omv-3241 sh[1676]:      filesystem-uuid (uuid) :   "FSUUID$4a49fe31"
Aug 20 14:42:02 omv-3241 sh[1676]:      claimed :   "true"
Aug 20 14:42:02 omv-3241 sh[1676]:      luksMapperName: :   ""
...
Aug 20 14:42:06 omv-3241 sh[1676]:     Starting job "Création d'une nouvelle table de partitions gpt sur /dev/sda ..." ( 4 / 35 )
Aug 20 14:42:06 omv-3241 sh[1676]: 14:42:06 [6]: virtual Calamares::JobResult CreatePartitionTableJob::exec()
Aug 20 14:42:06 omv-3241 sh[1676]:     Creating new partition table of type "gpt" , uncommitted partitions:
Aug 20 14:42:06 omv-3241 sh[1676]:     .. "New Partition"
Aug 20 14:42:06 omv-3241 sh[1676]:     .. Running QList("lsblk")
Aug 20 14:42:06 omv-3241 sh[1676]:     .. Finished. Exit code: 0 output:
Aug 20 14:42:06 omv-3241 sh[1676]: NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS
Aug 20 14:42:06 omv-3241 sh[1676]: loop0    7:0    0   2.7G  1 loop /run/initramfs/image
Aug 20 14:42:06 omv-3241 sh[1676]: sda      8:0    0 238.5G  0 disk
Aug 20 14:42:06 omv-3241 sh[1676]: `-sda1   8:1    0 238.5G  0 part
Aug 20 14:42:06 omv-3241 sh[1676]: sdb      8:16   1  14.6G  0 disk /run/initramfs/omdv
Aug 20 14:42:06 omv-3241 sh[1676]: |-sdb1   8:17   1   2.9G  0 part
Aug 20 14:42:06 omv-3241 sh[1676]: `-sdb2   8:18   1   6.4M  0 part
Aug 20 14:42:06 omv-3241 sh[1676]: zram0  253:0    0   7.6G  0 disk [SWAP]
Aug 20 14:42:06 omv-3241 sh[1676]:     .. lsblk output:
Aug 20 14:42:06 omv-3241 sh[1676]:  NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS
Aug 20 14:42:06 omv-3241 sh[1676]: loop0    7:0    0   2.7G  1 loop /run/initramfs/image
Aug 20 14:42:06 omv-3241 sh[1676]: sda      8:0    0 238.5G  0 disk
Aug 20 14:42:06 omv-3241 sh[1676]: `-sda1   8:1    0 238.5G  0 part
Aug 20 14:42:06 omv-3241 sh[1676]: sdb      8:16   1  14.6G  0 disk /run/initramfs/omdv
Aug 20 14:42:06 omv-3241 sh[1676]: |-sdb1   8:17   1   2.9G  0 part
Aug 20 14:42:06 omv-3241 sh[1676]: `-sdb2   8:18   1   6.4M  0 part
Aug 20 14:42:06 omv-3241 sh[1676]: zram0  253:0    0   7.6G  0 disk [SWAP]

 ==> pas de partition /boot/efi  créé , car initialement pas FULL EFI , la partie dm- provient de luks

Aug 20 14:42:14 omv-3241 kernel: Key type trusted registered
Aug 20 14:42:14 omv-3241 kernel: Key type encrypted registered
Aug 20 14:42:17 omv-3241 sh[1676]: 14:42:17 [6]: virtual void Calamares::JobThread::run()
Aug 20 14:42:17 omv-3241 sh[1676]:     Starting job "Définir des indicateurs (flags) sur la partition /dev/sda1" ( 6 / 35 )
Aug 20 14:42:17 omv-3241 sh[1676]: 14:42:17 [6]: virtual Calamares::JobResult SetPartFlagsJob::exec()
Aug 20 14:42:17 omv-3241 sh[1676]:     Setting flags on "/dev/sda" partition "/dev/sda1"
Aug 20 14:42:17 omv-3241 sh[1676]:      "boot"
Aug 20 14:42:18 omv-3241 sh[1676]: 14:42:18 [6]: virtual void Calamares::JobThread::run()
Aug 20 14:42:18 omv-3241 sh[1676]:     Starting job "Configurer les informations de la partition" ( 7 / 35 )
Aug 20 14:42:18 omv-3241 sh[1676]: 14:42:18 [6]: QVariantList FillGlobalStorageJob::createPartitionList() const
Aug 20 14:42:18 omv-3241 sh[1676]:     Building partition information map
Aug 20 14:42:18 omv-3241 sh[1676]:     .. partitions on "/dev/sda"
Aug 20 14:42:18 omv-3241 sh[1676]:     ..  mapping for "/dev/sda1" "/dev/sda1"
Aug 20 14:42:18 omv-3241 sh[1676]:      partlabel :   "root"
Aug 20 14:42:18 omv-3241 sh[1676]:      partition-uuid (partuuid) :   "PartUUID$40306c4f"
Aug 20 14:42:18 omv-3241 sh[1676]:      parttype :   ""
Aug 20 14:42:18 omv-3241 sh[1676]:      partattrs :   "0"
Aug 20 14:42:18 omv-3241 sh[1676]:      mountPoint: :   "/"
Aug 20 14:42:18 omv-3241 sh[1676]:      fs: :   "ext4"
Aug 20 14:42:18 omv-3241 sh[1676]:      fsName :   "luks"
Aug 20 14:42:18 omv-3241 sh[1676]:      filesystem-uuid (uuid) :   "FSUUID$6dc268d6"
Aug 20 14:42:18 omv-3241 sh[1676]:      claimed :   "true"
Aug 20 14:42:18 omv-3241 sh[1676]:      luksMapperName: :   "luks-412e5021-cc04-453b-b5c4-8b73075f54a7"

==> le flag boot est ajouté sur sda1 , donc c'est un démarrage en mode MBR , pas EFI ( EFI demande le flag ESP )

et lorsque l'installation de Grub arrive 
Aug 20 14:46:19 omv-3241 sh[1676]: [PYTHON JOB]: "Bootloader: grub (bios)"
Aug 20 14:46:19 omv-3241 sh[1676]:     .. Running QList("grub2-install", "--target=i386-pc", "--recheck", "--force", "/dev/sda")
Aug 20 14:46:20 omv-3241 irqbalance[759]: 
                                          
                                          
                                          -----------------------------------------------------------------------------
Aug 20 14:46:20 omv-3241 sh[1676]:     .. Target cmd: QList("grub2-install", "--target=i386-pc", "--recheck", "--force", "/dev/sda") Exit code: 1 output:
Aug 20 14:46:20 omv-3241 sh[1676]: Installing for i386-pc platform.
Aug 20 14:46:20 omv-3241 sh[1676]: grub2-install: warning: this GPT partition label contains no BIOS Boot Partition; embedding won't be possible.
Aug 20 14:46:20 omv-3241 sh[1676]: grub2-install: error: embedding is not possible, but this is required for RAID and LVM install.
Aug 20 14:46:20 omv-3241 sh[1676]: WARNING: [PYTHON JOB]: "Command 'grub2-install --target=i386-pc --recheck --force /dev/sda' returned non-zero exit status 1."
==> Ko

rappel : lorsque vous démarrer sur une clé iso USB en mode UEFI[Legacy] , cela signifie que la carte-mère tente en UEFI et sinon se rabat sur un mode “bios Grub”, si le disque est formaté en GPT , alors il y doit y avoir la création par Calamares de la partition 8Mo en bios grub non formaté , et la commande installation grub est alors correcte

autre solution , bien revérifier tous les paramètres UEFI , à savoir

secureboot desactive
pas de Legacy , pas de CSM
tous les disques en AHCI

le test des commandes suivantes devraient vous fournir les info avant de lancer installation

inxi  -Fxza
test -d /sys/firmware/efi && echo efi || echo bios
sudo efibootmgr -v
sudo parted -l

olivier · August 24, 2024, 12:48pm

Thank you Stephane, a correct BIOS setting (inxi-F.txt (2.2 KB)) has effectively fixed this preliminary issue.
Now the installation has been successful but we are wondering why “mitigations=off” needs to be added to the boot command line to be able to see the sddm login screen. I will share the log in my next post.

olivier · August 24, 2024, 1:19pm

I have just done a normal reboot. No sddm login screen. Here are some logs:
dmesg.txt (72.2 KB)
journalctl-S2024-08-24.txt (379.9 KB)

Let me know if other log info would be needed.

I will also check the behavior of my older laptop (samsung i3) which seems to have the same problem and report.

ben79 · August 24, 2024, 1:21pm

I do not know how long it may take to learn why some Intel cpu’s need mitigations=off to work in Linux. The security consequences are not dire, it is unlikely that anything would happen. Many people with intel cpu’s disable this on purpose because on their hw the mitigations cause a performance drop of up to 25%. Also note that our ‘Fast but insecure’ boot option includes this parameter. OM devs would not do that if they thought there was a lot of danger for users. I would enjoy your ROME system, but if you need additional security you may want to reinstall with encryption now that you know the ‘workaround’ for your cpu. I am trying to figure out how to get LUKS or some encryption to work in OMLx systems myself.

You can do an internet search of this to learn more. One bit is this article, got to section 3.7.

I hope this post is making sense. This is a user decision.

Note: Our ‘Fast but insecure’ boot option includes ‘insecure’ because of mitigations=off.

ben79 · August 24, 2024, 1:28pm

This is what I was looking for:

[ 48.888268] traps: sddm-greeter-qt[1808] trap invalid opcode ip:7f5e184f5b7f sp:7ffc6c600460 error:0

This needs a bug report from someone that has this issue. The issue is why some security patches like Spectre/Meltdown and others trigger this invalid opcode in OMLx systems?

ben79 · August 24, 2024, 1:51pm

I wanted to split some posts out of this topic for Why some OMLx systems need mitigations=off? but that seems not so easy to do.

Please post anymore discussion about the need to use mitigations=off or sddm crashing because of:

traps: sddm-greeter-qt[1808] trap invalid opcode ip:7f5e184f5b7f sp:7ffc6c600460 error:0

or:

kernel: traps: sddm[1053] trap invalid opcode ip:7f6b0a398d65 sp:7fffa8bbc850 error:0

in that thread so we can track this issue and other users can find.

Edit: mitigations=off is a boot parameter but systems with this issue are booting the problem is sddm fails and that should not be happening. At least my user level knowledge thinks it shouldn’t.

olivier · September 8, 2024, 8:21am

I volunteer, just need to organize myself i.e. reproduce the issue on another intel laptop and open the bug report. I am working on it but given my reaction time please be patient

system · October 8, 2024, 8:21am

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.