2

I'm running Lubuntu 14.04 on a HP Mini 5103 and after an upgrade yesterday I get a perfectly reproducable kernel panic.

  • Kernel that panics: 3.16.0-57
  • Kernel that doesn't panic: 3.16.0-56

Maybe related, but for the past half year or so, whenever I saw a notice for new updates and I tried to install them, sometimes (part of) the updates install fine and other times I get some error. When I try to submit the error, it asks for my password. Then this password dialog crashes when I confirm and spawns another report error dialog that asks for password again, which crashes again, etc...

My current workaround is to hold [shift] every time I turn on the computer and select the previous kernel, but I don't want to keep doing this. What I like is to:

  • Uninstall this offensive kernel, or
  • Select the previous kernel as default.

[edit]

ls -l /boot
total 216922
-rw-r--r-- 1 root root  1213454 aug 11 20:09 abi-3.16.0-46-generic
-rw-r--r-- 1 root root  1213779 aug 21 02:41 abi-3.16.0-48-generic
-rw-r--r-- 1 root root  1213632 sep  9 13:36 abi-3.16.0-49-generic
-rw-r--r-- 1 root root  1213747 okt 23 17:39 abi-3.16.0-52-generic
-rw-r--r-- 1 root root  1213747 nov  6 21:36 abi-3.16.0-53-generic
-rw-r--r-- 1 root root  1213747 nov 17 13:30 abi-3.16.0-55-generic
-rw-r--r-- 1 root root  1213821 dec  2 14:55 abi-3.16.0-56-generic
-rw-r--r-- 1 root root  1213821 jan 18 17:56 abi-3.16.0-59-generic
-rw-r--r-- 1 root root   176153 aug 11 20:09 config-3.16.0-46-generic
-rw-r--r-- 1 root root   176172 aug 21 02:41 config-3.16.0-48-generic
-rw-r--r-- 1 root root   176172 sep  9 13:36 config-3.16.0-49-generic
-rw-r--r-- 1 root root   176172 okt 23 17:39 config-3.16.0-52-generic
-rw-r--r-- 1 root root   176172 nov  6 21:36 config-3.16.0-53-generic
-rw-r--r-- 1 root root   176172 nov 17 13:30 config-3.16.0-55-generic
-rw-r--r-- 1 root root   176172 dec  2 14:55 config-3.16.0-56-generic
-rw-r--r-- 1 root root   176172 jan 18 17:56 config-3.16.0-59-generic
drwxr-xr-x 5 root root     1024 jan 20 16:36 grub
-rw-r--r-- 1 root root 19862602 sep  3 08:29 initrd.img-3.16.0-46-generic
-rw-r--r-- 1 root root 19861440 sep  4 13:03 initrd.img-3.16.0-48-generic
-rw-r--r-- 1 root root 19861112 sep 25 12:40 initrd.img-3.16.0-49-generic
-rw-r--r-- 1 root root 19861936 nov  5 08:57 initrd.img-3.16.0-52-generic
-rw-r--r-- 1 root root 19861924 nov 20 14:12 initrd.img-3.16.0-53-generic
-rw-r--r-- 1 root root 19862906 dec  1 09:21 initrd.img-3.16.0-55-generic
-rw-r--r-- 1 root root 19864377 dec 17 09:04 initrd.img-3.16.0-56-generic
drwx------ 2 root root    12288 aug 29 16:41 lost+found
-rw-r--r-- 1 root root   176500 mrt 12  2014 memtest86+.bin
-rw-r--r-- 1 root root   178176 mrt 12  2014 memtest86+.elf
-rw-r--r-- 1 root root   178680 mrt 12  2014 memtest86+_multiboot.bin
-rw------- 1 root root  2792155 aug 11 20:09 System.map-3.16.0-46-generic
-rw------- 1 root root  2791983 aug 21 02:41 System.map-3.16.0-48-generic
-rw------- 1 root root  2791983 sep  9 13:36 System.map-3.16.0-49-generic
-rw------- 1 root root  2792803 okt 23 17:39 System.map-3.16.0-52-generic
-rw------- 1 root root  2792803 nov  6 21:36 System.map-3.16.0-53-generic
-rw------- 1 root root  2792814 nov 17 13:30 System.map-3.16.0-55-generic
-rw------- 1 root root  2793015 dec  2 14:55 System.map-3.16.0-56-generic
-rw------- 1 root root  2793015 jan 18 17:56 System.map-3.16.0-59-generic
-rw------- 1 root root  6017280 aug 11 20:09 vmlinuz-3.16.0-46-generic
-rw------- 1 root root  6017696 aug 21 02:41 vmlinuz-3.16.0-48-generic
-rw------- 1 root root  6017504 sep  9 13:36 vmlinuz-3.16.0-49-generic
-rw------- 1 root root  6018848 okt 23 17:39 vmlinuz-3.16.0-52-generic
-rw------- 1 root root  6018848 nov  6 21:36 vmlinuz-3.16.0-53-generic
-rw------- 1 root root  6020480 nov 17 13:30 vmlinuz-3.16.0-55-generic
-rw------- 1 root root  6020960 dec  2 14:55 vmlinuz-3.16.0-56-generic
-rw------- 1 root root  6020800 jan 18 17:56 vmlinuz-3.16.0-59-generic
4

In the strongest possible terms I would urge you to resist the cheap solution here (simply uninstalling 3.16.0-57). The Kernel isn't updated to just annoy people, it gets some really important security updates that could otherwise leave your system open to attack.

Instead I would either look at filing a bug against the kernel. If you have any development experience, knowing the exact break point allows you to see where the issue is, much more easily than anything else.

Alternatively, you're on the first rung of a series of LTS hardware enablement stacks (you're on the Utopic Kernel). You could hop to Vivid or Wily. You will need to change to Xenial's stack August 2016 because that'll be the only supported HWE stack.

The other alternative is uninstalling all the HWE packages, dropping back to Ubuntu 14.04.1's 3.13 kernel. This is supported for the full five years (into 2019). This is a perfectly sane option unless you have modern graphics requirements.

Your other issues may be related (especially if this is a graphics driver issue) so another stack version can

  • I don't understand any of the techno-babble. I just want keep this PC as low-maintenance as possible. I certainly don't want to write into my agenda that I have to intervene more in Aug 2016. How do I go back to Kernel 3.13? Graphics has never been an issue. – Mark Jeronimus Jan 20 '16 at 14:57
  • For going from 14.04.2 (Utopic HWE) to 14.04.1: askubuntu.com/a/589266/449 – Oli Jan 20 '16 at 15:15
  • It gives me dependency errors. It says I have to do sudo apt-get -f install to fix those but it gives the same errors: Errors encountered while processing: /var/cache/apt/archives/linux-3.13.0-76-generic_3.13.0-76.120_i386.deb dpkg returned an error code (1) – Mark Jeronimus Jan 20 '16 at 15:24
  • After some trying to fix my presumably broken package manager, it now says cannot copy extracted data for [...] No space left on device, but I have space left plenty df -h -> 117G 8.1G 103G 8% / <edit> Only other one that's more than 1% full is 236M 231M 0 100% /boot – Mark Jeronimus Jan 20 '16 at 15:29
  • Don't suppose you can run an apt-get autoremove now? – Oli Jan 20 '16 at 15:32
1

Answered by OP

It seems the root cause for my "maybe related" problem was a full /boot partition (size 236M), which also blocked my progress in solving my kernel panic.

  • Cleaned out kernels from /boot except the one you booted from,
  • Fixed broken package manager with apt-get -f install
  • Clean up with apt-get -f autoremove && apt-get update && apt-get upgrade

To fix my kernel panic:

  • Install an older kernel which has long term support (3.13 as of writing): apt install linux-generic libgl1-mesa-glx linux-libc-dev linux-tools-generic xserver-xorg libegl1-mesa-drivers
  • Reboot, enter GRUB, select this kernel
  • Remove other kernel from /boot
  • Remove other kernels from GRUB: update-grub

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.