Sytem totally messsed up after kernel update

denijane · n00b Joined: 22 Jul 2014 Posts: 14

After my update to the latest revision for kernel 4.18.19, my laptop loads for 35 minutes. It's extremely slow and after logging in, the first attempts lasted about 5 minutes, ending with a system freeze and once with the message "Warning, system failed to mount to "/" sysroor" (or something like this I saw it just once). When trying to load another 10 times, I saw multiple dbus and irq errors.

On my last attempt, I did what this link suggested https://forums.gentoo.org/viewtopic-t-1090156.html, i.e.

Muso · Posted: Tue Dec 04, 2018 12:34 pm Post subject:

Have you checked this?
_________________
"You can lead a horticulture but you can't make her think" ~ Dorothy Parker
2021 is the year of the Linux Desktop!

denijane · n00b Joined: 22 Jul 2014 Posts: 14

Yes, I checked the Sabayon forum first. But there's nothing similar there, while here the topic I quoted had a similar issue and the dbus step actually helped. This is why I thought maybe here is the better place to post about the problem.

fedeliallalinea · Posted: Tue Dec 04, 2018 1:08 pm Post subject:

denijane · n00b Joined: 22 Jul 2014 Posts: 14

I really don't understand how could a distro with gentoo kernel be an "Unsupported software", but put my post wherever you like. I stayed up until 4:30 am to try to find out the problem, I'm utterly miserable because my laptop is pretty unusable, and I simply couldn't care less if you're in some war with Sabayon or not.

I think the Linux community spirit is supposed to mean people helping each other, regardless if they use Gentoo or Sabayon or whatever. But that's obviously just my opinion, so I couldn't care less. I just would really like if some kind soul here could give me a clue what the problem is, because this laptop is my workstation and I need it responsive and working. And I'm desperately needing help in debugging this.

xaviermiller · Posted: Tue Dec 04, 2018 1:37 pm Post subject:

Moved from Kernel & Hardware to Unsupported Software.

Hi,

Can you compare the .config from the 2 kernel versions?
Can you rollback to the previous working version?

It seems some module is missing. How did you configure and compile the kernel?
_________________
Kind regards,
Xavier Miller

denijane · n00b Joined: 22 Jul 2014 Posts: 14

Hi. I can't seem to find any .config in /boot. Tried

Muso · Posted: Tue Dec 04, 2018 3:29 pm Post subject:

Look in /usr/src

Find the old kernel you used and look at the .config in that directory.
_________________
"You can lead a horticulture but you can't make her think" ~ Dorothy Parker
2021 is the year of the Linux Desktop!

denijane · n00b Joined: 22 Jul 2014 Posts: 14

Well this is what I found from /proc/config.gz. In /usr/src I see the kernels, but there are no .config folders.

I'm backing up now, in case bad (or worst) things happen. What I find very troubling is that system takes forever to load, and it seems like it's constantly writing something on the hard. I have a hybrid SSD + SATA and I'm really worried about the hard. Until my update yesterday, I never ever have had problems with it. The laptop is about 7 months old, so that's certainly not normal.

https://bpaste.net/show/f649d8787c0e

fedeliallalinea · Posted: Tue Dec 04, 2018 4:35 pm Post subject:

denijane forum truncate too long message, use wgetpaste for post long message
_________________
Questions are guaranteed in life; Answers aren't.

denijane · n00b Joined: 22 Jul 2014 Posts: 14

Oh OK. I'm sorry about that. I edited the previous post with the bpaste link.

Jaglover · Posted: Tue Dec 04, 2018 5:02 pm Post subject:

I/O errors usually mean the hard drive is toast. Why don't you start with booting from an external media and diagnosing the drive.
_________________
My Gentoo installation notes.
Please learn how to denote units correctly!

denijane · n00b Joined: 22 Jul 2014 Posts: 14

So far tried to boot Mint and Kali linux from a bootable usb, unsuccessfully. It doesn't start, it stops at the same error:

denijane · n00b Joined: 22 Jul 2014 Posts: 14

Ah, just one more thing I just noticed. My CPU load seems to stay constantly on 100% which is very weird.

It seems that lscpu doesn't know how many threads I have at all! How is this even possible?

Jaglover · Posted: Tue Dec 04, 2018 8:16 pm Post subject:

Hu · Moderator Joined: 06 Mar 2007 Posts: 21633

Muso · Posted: Wed Dec 05, 2018 3:11 am Post subject:

denijane · n00b Joined: 22 Jul 2014 Posts: 14

Guys, I don't want to fight with anyone. My presumption was that sabayon developers don't change that much in the kernel. It was based on the fact nobody there could (or would) help me with my touchpad/acpi problem with the same laptop before. Maybe I was wrong.

In any case after the update, a laptop which never showed any error or sign of wear (it's 7 months old), now is a toast. Since today it's no longer able to find my hard drive at all. I don't know if it's just a coincidence that it happened after the update or something about the original acpi problem lead to the hard destruction, but it's over. Needless to say I'm too sad to argue about anything.

NeddySeagoon · Posted: Wed Dec 05, 2018 12:25 pm Post subject:

denijane,

Rule 1 is Presume Nothing. I've learned that the hard way several times. Why presume its a kernel problem?

From your dmesg,

Yamakuzure · Posted: Wed Dec 05, 2018 12:42 pm Post subject:

Another idea:
Is /dev/sda your SSD?
If it is, how full is it? Did you fstrim it since you got your laptop?
_________________
Important German:

"Aha" - German reaction to pretend that you are really interested while giving no f*ck.
"Tja" - German reaction to the apocalypse, nuclear war, an alien invasion or no bread in the house.

denijane · n00b Joined: 22 Jul 2014 Posts: 14

Hi guys,
I'm sending my laptop to be repaired tomorrow, because today morning i found it unresponsive and i couldn't load the system anymore. Whatever the problem is, it escalated very quickly and killed it while i was backing up.

Yesterday I managed to run a smartmontools short test and it said the hard passed the test, but the first row (read_raw_something) said pre-failure state , all the other rows: "old age". It's my first ssd hard so i never knew i have to use fstrim or smartmontools. And I don't know if 7 months count as "old age" or there was something else.

So yeah, the hard is a toast. I'm thinking of switching to one M.2 SSD for the OS and regular SATA hard for everything else. It seems safer and easier to repair. But that's for when i can afford it.

Anyway, thank you all for trying to help me. At least I've learnt some things on how to check my SSD status, which will hopefully help me in the future to avoid such situations. Really, thanks a lot!

NeddySeagoon · Posted: Thu Dec 06, 2018 6:25 pm Post subject:

denijane,

You have just been unlucky.

smartmontools reads the drives internal error log' you don't need to run in as a matter of course but when all is not well, it can help determine if the problem is inside or outside the drive.
It works equally well on SSDs or rotating rust.

fstrim maintains the SSD write speed. The SSD write process can only change zeros into ones. That works provided that the area to be written has been erased' so that is is all zero.
Erase is a slow operation on SSDs so to maintain the write speed, needs a supply of erased blocks, ready to be written.
There are two ways to achieve that.

a) mount the filesystem with the discard option. This tells the kernel to send the trim command whenever it frees blocks on the SSD. The drive is than free to erase them whenever it wants to.
b) run fstrim in a cron job. This gives the drive a list of of all the unused blocks on the filesystem, including erased blocks. The drive works out what blocks need to be erased and adds them to its todo list.

Neither method is a command to the drive to perform any erases. Its up to the drive what it does and when. I'll gloss over why that is, its just a good thing for the drives life.
However, data recovery probably won't work after either of those options have been given.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

Jaglover · Posted: Thu Dec 06, 2018 6:42 pm Post subject:

NeddySeagoon,