when Fedora was recently installed, I got an error message after I booted it from my SSD.
Initramfs unpacking failed: XZ-compressed data is corrupt
I was able to fix this first error by manually setting the correct tact frequency of my RAM from 2133Mhz to 3200MHz (https://www.linuxquestions.org/questions/linux-software-2/xz-compression-data-is-corrupt-system-halted-error-on-boot-up-4175574040/).
Some days ago I noticed that my mainboard didn’t really turn off when I had shutdown my computer. Then I decided to completely turn off the computer’ electricity (BIOS battery is new). Always if I start the computer after I turned the electricity on again, a new initramfs error appears. It disappears if I start the computer via reset button (electricity wasn’t turned off). Since I read something about FCLK frequency in relation to the first error and informed myself further about the FCLK frequency (https://techgeekish.com/what-is-fclk-frequency/). My idea was: setting it manually might solve the problem “because there’s maybe a relation to the previous initramfs error”? I didn’t set it to auto anymore even if the correct FCLK frequency was automatically chosen (1600Mhz). I did set it to 1600Mhz. Nothing changed due to that. The new error consists. I then used Google. I found this (https://forums.linuxmint.com/viewtopic.php?t=323152) & this (https://bbs.archlinux.org/viewtopic.php?id=136540) but I didn’t change something so far…because I’m not sure what I’d actually doing there. So, the error is following:
Initramfs unpacking failed: uncompression error
How can I fix this? Are the solutions, which I’ve already found useful? Do I even have to fix it? Are there any big risks if I won’t fix it and continue as I did so far (shutdown, turn electricity off, turn electricity on, boot, reboot via reset button, select fedora to boot, boot)?
P.S. The error DOES NOT appear if I let the electricity on. That is somehow weird…(but somehow not weird because…RAM unloads…if the electricity is off…? And since the electricity is not off/on the RAM has something in itself which is needed so that the error doesn’t appear?)
This sounds like the frequency settings in BIOS may not be properly maintained and when you disconnect power they revert to default. You stated that you have a new battery for BIOS but have you confirmed it is 1) properly installed? and 2) the correct battery?
When power is retained to the machine it automatically maintains the bios settings. When power is disconnected it relies on the battery to retain those settings. I would check and affirm that everything I did was done properly, maybe even installing (another) new battery just in case. Sometimes the contacts in the battery holder may become damaged and not maintain good contact as another potential problem.
I bought the mainboard recently (less than a week ago). The battery was installed and the BIOS also “remembers” the time (it’s 2 hours less than my OS time but it’s saved).
I will exchange the battery as soon as possible (tomorrow or the day after tomorrow).
I will also check the contacts then.
Thanks for the help so far! I’m going to reply as soon as I’ve done all these steps!
Sincerely!
P.S. Is there maybe a setting in the BIOS which says “don’t use the battery’ power, always use the system’ power”?
I just wanted to write down a new appeared error (but I didn’t do the steps which I wanted to do tmw or day after tmw…so this is “just for the documentation”…for now) (the old error wasn’t there, only this new one Xd):
That does not seem to be related to bios, but the OS itself.
Please show us the output of dnf list installed xz* just in case this may be related to a recent malware found upstream from fedora. (apparently not an issue if the version shown is not 5.6.0)
Your description seems to imply that you may have purchased a second-hand motherboard. Is that correct? If so then this might be related to the bios version as well. Support for SSDs has changed with different motherboards, bios versions, etc.
The security vulnerability you’re talking about relates only to Fedora 40 Rawhide I believe (https://fedoramagazine.org/cve-2024-3094-security-alert-f40-rawhide/) .
With that board, new, there should have been no reason to replace the battery. (they usually last for many years on motherboards)
It was merely a guess and a hope that the malware was not a factor in the problems with uncompressing the initramfs image. I am glad to see that it is not and yes, I knew it should not have been a factor on f39, but sometimes better safe than sorry so I wanted confirmation.
The SSD should not be an issue.
Please show us the results from inxi -Fzxx
The only thing that initially sticks out to me is that you are using a GPU with HDMI, DP, and DVI-D ports and you are using the DVI-D port. Is there a reason you are not using one of the other ports?
This should not cause an issue with loading from the drive, but is just an observation.
The next thing is that you have 4 drives, all apparently sata (sda, sdb, sdc, & sdd), the last 2 seemingly connected by USB. The system is installed on /dev/sdb.
As far as I can tell there really should be no problem unless you may have something miss-configured in the bios such as memory speed, overclocking, or similar; or potentially something is corrupted with the way the OS was installed.
Have you verified all the bios settings to match the RAM installed, the CPU installed, and the power supply is adequate for the included hardware? Have you considered doing a reinstall just in case the os itself may be the problem?
The time should also be displayed with hostnamectl just to rule out other issues caused by time configs.
I use DVI-D because then I can get 144 FPS (because I can use 144 Hz).
Ok.
I had to adjust my memory speed (to 3200 MHz) because the BIOS recognized the wrong memory speed (2133 MHz). I did not overclock anything. Ok but what could it be?
I have a Corsair CS750M.
I’ve read that the initramfs can be regenerated so that a reinstallation of the OS isn’t needed.
I think there’s nothing about time there…?
annatar@fedora:~$ hostnamectl
Static hostname: (unset)
Transient hostname: fedora.fritz.box
Icon name: computer-desktop
Chassis: desktop 🖥️
Machine ID: <I removed this text, hope that's alright>
Boot ID: <I removed this text, hope that's alright>
Operating System: Fedora Linux 39 (Workstation Edition)
CPE OS Name: cpe:/o:fedoraproject:fedora:39
OS Support End: Tue 2024-11-12
OS Support Remaining: 6month 3w 6d
Kernel: Linux 6.8.5-201.fc39.x86_64
Architecture: x86-64
Hardware Vendor: ASUS
Hardware Model: PRIME B550-PLUS
Firmware Version: 3607
Firmware Date: Fri 2024-03-22
Firmware Age: 3w 3d
annatar@fedora:~$
You mentioned time. The BIOS time is -2 than my OS time. Is that a problem? On my OS it’s ~11 PM, in my BIOS it’d be ~9 PM.
The BIOS may be using a lower memory speed due to limitations of other components. Did you use memory from an ASUS Qualified vendor? Try running the system with 2133 to see if the problems go away. Also check that the power supply has ample capacity – a marginal supply may give poor quality power under load.
It’s not going to work on 2133MHz as the reason had been stated out right at the beginning of this thread. The PSU isn’t a bad quality; the PSU’ name also had been stated in this thread. And my hardware isn’t really limitated as this fact can be filtered out by reading the already given data about my specs, which had been given already (I guessed this is the data: https://discussion.fedoraproject.org/t/initramfs-several-errors/113160/9).
If I’d doubt something about my hardware, it’d be maybe - but I’m not sure so far since we’re still trying to find the problem/s - that my RAM somehow could have problems. Further details, which could underline this guess, in a seperate reply (https://discussion.fedoraproject.org/t/initramfs-several-errors/113160/17).
no problem, nothing has damaged. This is the console log:
annatar@fedora:~$ timedatectl
Local time: Tue 2024-04-16 11:19:14 CEST
Universal time: Tue 2024-04-16 09:19:14 UTC
RTC time: Tue 2024-04-16 09:19:14
Time zone: Europe/Berlin (CEST, +0200)
System clock synchronized: yes
NTP service: active
RTC in local TZ: no
annatar@fedora:~$
P.S. I’ve got a new error: https://discussion.fedoraproject.org/t/initramfs-several-errors/113160/17
The “broken padding” thing would quite likely be to say you have physical memory issues. We can’t in that case advise much more than e.g. running memtest86 to diagnose such.
Yes, I will try that. I just began to read the Github guide (https://github.com/memtest86plus/memtest86plus/?tab=readme-ov-file#build-and-installation); looks very complicated but somehow I will be able to run this.
So I have now used Memtest. I had completely forgotten that Memtest just continues unless you cancel the program yourself; it ran 5 hours . I also forgot to save the result somehow - I installed the Fedora package but I think I still have to save the results separately somehow. Anyway, there was an error. This is bad. I’m going to run Memtest again tomorrow. After that I’m going to post a screenshot/photo here. Cya!
.
. I also forgot to save the result somehow - I installed the Fedora package but I think I still have to save the results separately somehow. Anyway, there was an error. This is bad. I’m going to run Memtest again tomorrow. After that I’m going to post a screenshot/photo here. Cya!