and then you showed the pic which clearly has it. That means at least one upgrade had to be taken. Also in my first reply I mentioned BE you could choose as a fallback boot.
cd /.zfs and actually list the snapshots. This way you can diff the loader.conf from this BE to one you can't boot.I recently stumbled into this. It was hanging on last good process. Not showing the bad process.
Yea I had confused it with PWC web cam viewer. The App.
But you correctly identify it as a webcam, close enough here.
/boot/loader.d directory, worth checking out if something is there or not.bectl list
bectl mount 13.1-RELEASE-p2.... /mnt
cp -r /mnt/boot/kernel /some/other/dir/
rm -fr /mnt/boot
cp -r /boot /mnt
bectl umount 13.1-RELEASE-p2....
reboot
freebsd-update fetch
freebsd-update install
Phishfry Yeah, sometimes it can be you see the last good one, sometimes you may hit a problem during init of something that is a problem. But it does narrow it down a bit then.
Here OP said though:
But you correctly identify it as a webcam, close enough here.
Also (loaded) kernel doesn't bootstrap or we don't see it yet. So it has to be something very early. That's why my suggestion was to forget all (unload) and just go with bare kernel. I forgot completely about the/boot/loader.ddirectory, worth checking out if something is there or not.
OP shared the pics where loading old kernel didn't help. So I'd go with the /boot check next (maybe stale loader on efi partition?). It's worth checking out what is there compared to new system. That caused headaches before.
loader.conf I showed you earlier is the one I can't boot with, the one you see right there in this picture.
I'll be back to you whit that once martin tell me what he wants to see in my boot directory, thanks for your help..
Incorrect. All the modules that are listed in the loader.conf, even the non-existent pwc were loaded (and attempted to be loaded) by the loader. Hung is happening most likely when loader is leaving and attempting to jump to kernel. That's why it's most likely a problem in /boot, efi loader even more.
gpart show was to see if you maybe have a legacy boot partition too. I keep them around, they are handy (and old habits die hard, I still legacy boot all my boxes).mount -t msdosfs /dev/ada0p1 /tmp/efils -la /tmp/mydata/boot/loader.efi /mnt/efi/efi/boot/bootx64.efi and check if they are the same size. If they are use cksum to compare if they are really the same file. Actually do the ls -la /tmp/mydata/boot/*.efi and compare it to the one on efi partition. If neither file fits you could try it with the other BE (one you can boot) to see if any of that matches. I'm assuming you booted the 2nd BE and not the recovery media. If that's the case then compare the efi file to /boot/*.efi.cp /tmp/efi/efi/boot/bootx64.efi /tmp/efi/efi/boot/keep-bootx64.efi
cp /tmp/mydata/boot/loader.efi /tmp/efi/efi/boot/bootx64.efiYes anything I post on the forum awaits moderator approval, don't know why.
Everybody is on moderation for the first 10 posts and 10 days.
I am pretty sure they're not. I've try to boot with only loading the kernel and it hangs on mount. Trying to mount zfs says unknown filesystem.. Even if load zfs(4) and crypto(4). I still run into some issues later. The nvidia module doesn't load and some late filesystem for a linux jail doesn't mount for some reason.
and I understood it as that boot that was shown by OP. Which is not true as when boot continues modules get loaded (as shown in pic).
When you hit the loader prompt modules are not yet loaded, that is correct. If you just do boot from prompt (i.e. hit 3, type boot) modules from loader.conf will get loaded. If you do unload and start manually loading stuff nothing else is loaded from loader.conf. That's why I suggested that - to bypass any configuration to see if vanilla kernel starts bootstrapping.
Yes, please do that.
load /boot/kernel.old/kernel is shown. I am not sure I understand.
joshbsd read again the above post
We didn't talk about the same boot process. I told joshbsd to do a manual boot of old kernel ; I did expect that other modules are not loaded as I stated above.
Result:
# ls -la /tmp/mydata/boot/loader.efi
-r-xr-xr-x 2 root wheel 892928 Sep 23 20:01 /tmp/mydata/boot/loader.efi # ls -la /tmp/efi/efi/boot/bootx64.efi
-rwxr-xr-x 1 root wheel 8292928 Sep 23 15:51 /tmp/efi/efi/boot/bootx64.efiWe did try that by loading the old kernel. ( loader.conf not loaded)We didn't talk about the same boot process. I told joshbsd to do a manual boot of old kernel ; I did expect that other modules are not loaded as I stated above.
Also there's this facepalm moment, I missed that .. Phishfry mentioned it and VladiBG highlighted it now - why do you use syscons driver? You mentioned you didn't do any upgrade but I have a feeling you did a lot there.
Before anything comment out that kern.vty line, there's probably no need for you to use this legacy driver.
I still think this will fix.
I missed it too and now I don't know were it came from, I commented out the all line and rebooted now I go to the entire boot process and it bring me to the login screen sddm but I'm stuck there in a loop where I have to enter my credentials again and again, same in the CLIWe didn't talk about the same boot process. I told joshbsd to do a manual boot of old kernel ; I did expect that other modules are not loaded as I stated above.
Also there's this facepalm moment, I missed that .. Phishfry mentioned it and VladiBG highlighted it now - why do you use syscons driver? You mentioned you didn't do any upgrade but I have a feeling you did a lot there.
Before anything comment out that kern.vty line, there's probably no need for you to use this legacy driver.