BOOT_IMAGE crashed message and occasional *possibly unrelated crashes?*
Machine I'm running:
I am currently running Fedora 30 on a recently purchased Laptop: Machine: Lenovo Model: P53s HD: 1TB SSD RAM: 24GB (8GB onboard) Video: Nvidia Quadro P520 w/2GB memory GDDR5 64Bit OS: Fedora 30 with Nvidia drivers installed. Problem: I get a message saying the following, every time I login to the KDE desktop: Code:
Oops! Immediate Reaction: My only recourse, after having tried ping and ssh, is to hard reset the machine (holding power button). Other similar threads: Unfortunately, this is similar to a question posted for Fedora 23, but, I am not sure it is relevant in this case and the link is below: https://www.linuxquestions.org/quest...54#post5453154 It talks about re-creating initrd - which I believe since migrated to initramfs (correct me if I'm wrong) and apparently this is created on the fly with the newer versions of fedora (I have fedora 30 running). What I have tried: I have used the abrt tool - It gives me such a long laundry list of problems - including ones that predated the install of Nvidia drivers for the machine. The output of the following command is possibly relevant: # abrt info -c kernel --pretty full |less Code:
Id da6d80c The other entries in the abort list look pretty much the same as this one - the note is "Not reportable" and the message that follows is the same. One entry that differed looks as follows: Code:
Id eb21e2c I also ran: Looked at stack trace "dmesg -H": (found the one line that may indicate an error). All the other lines seem to be ok. Code:
... What do you think is the cause of the problem/problems I am facing. Do you think the crashes are related to the messages shown in the code sections above? What should I resort to trying? Is there any other information that might be relevant to a search for answers, that I should chase down? |
Hi guys,
One more command that I used: journalctl -b Code:
-- Logs begin at Wed 2019-10-23 11:43:25 EDT, end at Wed 2019-11-13 15:47:15 EST. -- Every time I boot, prior to the KDE boot splash screen, or Logout of KDE - I see "Irq #31 disabled" |
Does this appear in a separate error window?
I don't know with which tool KDE reports errors and where they are logged. Perhaps you can find out. In any case, this doesn't look like a kernel problem. It rather seems that some KDE configuration file contains the string "BOOT_IMAGE=(hd0,gpt3)/vmlinuz-5.3.8-2000.fc30.x86_64", which KDE tries to execute and fails. Not knowing anything about KDE, I would perform a brute-force search for that string over all text documents on the PC. Or again, find out where KDE config and log files might be located and check them. |
Hi berndbausch,
It appears in a popup dialog box (nothing but an X in a circle in the right hand side). I am under the impression that it isn't a KDE issue - boot image suggests something to do with the booting of the system... Just that it manages to send an error to KDE - or it is possible that some interaction with KDE causes it to pop up. |
Quote:
Code:
it looks like BOOT_IMAGE=(hd0,gpt3)/ |
My Fedora 30 and 31 has been doing that for the last few months. You can send data to the kernel dev's if you want.
Once the kernel has been marked tainted it seems to stay that way for a while. There is a lot of information about tainted kernels online and how to discover more about the module(s) doing it. |
Looks like it might have come in with 5.3 - my old F30 system was still on 5.1, never saw the issue. Just updated and 5.3 came in, and so did the message. Also broke my wifi - dkms still builds ok, but the driver won't see my TP-Link dongle. Might have to go find a newer driver source.
|
Quote:
Perhaps you are right.... I'm not sure about the inexistent part though... What this looks like: This looks like something that isn't supposed to be still running to me - vmlinuz is the initial compressed kernel that is used in order to start the system... boot image is where it is supposed to be running from on the initrd (init ram disk), where this file is found - on hard disk 0 (hd0), with gpt3 (GUID Partition Table). This looks like a line from grub configuration.... New Thoughts on this: Not Sure Why a Line from Grub is getting put in a message... I had to install the new drivers for my Nvidia card - this is the only departure from a standard installation that I can think of - this involves something getting compiled and installed - the compilation may have led to a branch that executes or attempts to execute something that is a grub line on the machine - perhaps I need to check the install shell that causes the automated compile. - 6th - |
Quote:
Related: I needed to use dkms as well! (for compiling the Nvidia drivers) Perhaps there is something there... - 6th - |
Found Some More Evidence?:
Someone who was compiling something to do with Nvidia drivers came across the error too... He initially thought it was dkms - or so it would seem... He is showing this part here: Code:
... See this link: https://askubuntu.com/questions/6576...talling-kernel Random Hanging Explained? This might simultaneously explain the random hanging - plus the occasional screen garbage just prior to hanging.... - 6th - |
Quote:
I didn't understand why you thought this was relevant... Now, I do. - 6th - |
Hi guys,
Do I have things Partially Solved? Off to look to see discussions on Nvidia drivers now, and what I can do to fix the driver, or get one that works with the Fedora kernel - 6th - |
Part of the issue could be a loader issue. Run sudo efibootmgr
|
EfibootMgr Output:
Here is what efibootmgr gives: BootOrder: 0001,0000,001B,001C,001D,001E,001F,0020,0021,0022,0012,0011,0023,0024 Boot0000* Windows Boot Manager Boot0001* Fedora Boot0010 ThinkShield secure wipe Boot0011* LENOVO CLOUD Boot0012* HTTPS BOOT Boot0013 Setup Boot0014 Boot Menu Boot0015 Diagnostic Splash Screen Boot0016 Lenovo Diagnostics Boot0017 Regulatory Information Boot0018 Startup Interrupt Menu Boot0019 Rescue and Recovery Boot001A MEBx Hot Key Boot001B* USB CD Boot001C* USB FDD Boot001D* NVMe0 Boot001E* NVMe1 Boot001F* ATA HDD0 Boot0020* ATA HDD1 Boot0021* USB HDD Boot0022* PXE BOOT Boot0023 Other CD Boot0024 Other HDD Boot0025* IDER BOOT CDROM Boot0026* IDER BOOT Floppy Boot0027* ATA HDD Boot0028* ATAPI CD Note: Not sure why "Windows" is mentioned - as my machine is exclusively running on Linux at the moment. - 6th - |
Quote:
If it was mine, I'd cut down that list of entries. ISTR reading about some pretty flaky Lenovo UEFI BIOS behavior. Possibly Lenovo's BIOS gets overwhelmed by such a long list, and that passes through as corruption of the environment. Try removing BOOT_IMAGE as a potential problem by removing it from the kernel cmdline at the Grub menu, and if it helps, strike it from /etc/default/grub as well. When the kernel is tainted, it means that it is in a state that is not supported by the community, e.g. a proprietary NVidia driver has been installed. I find it hard to imagine how BOOT_IMAGE or the kernel's filename could be connected to anything KDE-related. KDE doesn't care what kernel you are running. NVidia GPUs don't necessarily require NVidia's tainting drivers. If nothing else solves the problem, try removing to narrow down the problem possibilities. The default FOSS driver works fine on F30 here, e.g. on a really old one currently booted: Code:
# inxi -V | head -n1 |
All times are GMT -5. The time now is 05:01 PM. |