1

my laptop is failing to boot every time, so when I turn it off to go to the GRUB boot loader for help I'm seeing what looks like two images which are:

4.0.15-42 generic and recovery 4.0.15-20 generic and recovery

try as I might, 42 will not boot but 20 does. Is there any way that I am able to remove 42 so that my device will boot without me having to intervene with GRUB every time?

Thanks,

AJ

Big_Bad_Bob
  • 11
  • 3
  • Has this been happening for a long time? You probably used to have -40, did it boot? What about -38? -36? – user535733 Dec 11 '18 at 19:39
  • Hi, only just happened, I've not swtiched on in 6 months so had a massive update and now this happened. I've tried doing a clean install but GRUB still shows the two images. – Big_Bad_Bob Dec 11 '18 at 19:59
  • Open a terminal and try sudo apt update and sudo apt upgrade. If there ANY errors, then please edit your question and copy/paste in the complete output (not a summary). If there are NO errors, then don't paste anything, just say so. – user535733 Dec 11 '18 at 20:11
  • No errors there – Big_Bad_Bob Dec 11 '18 at 20:48
  • ok so used software manager and removed kernel -42 and laptop now boots fine using kernel -20 – Big_Bad_Bob Dec 11 '18 at 21:22
  • I was thinking of updating to -39 instead to see if that works instead, so that I'm at least somewhat more secure and up-to-date than -20. What would you recommend? – Big_Bad_Bob Dec 11 '18 at 21:24
  • Usually I recommend bisecting kernels so you can find exactly which one stopped working, and file a bug report. So try -30. If -30 works, try -36. The goal is to find the very first one that stops working so developers are looking for the problem in the right place. As a side benefit, you also know the highest kernel you can use before it breaks. – user535733 Dec 11 '18 at 22:57
  • It doesn't hurt to keep a mainline kernel around for fall-back purposes if you can spare the 400MB of space. I would go with 4.14.78 because I'm using it now but it is a month or two old: https://askubuntu.com/questions/119080/how-to-update-kernel-to-the-latest-mainline-version-without-any-distro-upgrade – WinEunuuchs2Unix Dec 12 '18 at 02:11
  • Appreciate the responses so far. What I've done is incremented one by one until today where I'm now at 4.15.0-30 and my laptop is booting up fine. I'll eventually reach the kernel where it stops working as user535733 stated. – Big_Bad_Bob Dec 13 '18 at 20:16

1 Answers1

0

I observed the same when upgrading from 4.15.0-39 to 4.15.0-42 on my Ubuntu 18.04.1 NAS (x86_64):

Failed to execute /init (error -2) second line: Kernel panic - not syncing: no working init found.

So apparently the init is broken. My solution was to remove 4.15.0-42 

sudo apt remove linux-image-4.15.0-42
sudo apt autoremove

and, thus, automatically go back to 4.15.0-39 which works fine. The autoremove removed amd64-microcode, intel-microcode iucode-tool and thermald which appear to be needed by -42 only.

(In order to be able to do the above, boot from a live CD/DVD/USB and chroot to the Ubuntu installation)

Robert
  • 163
  • thermald has been around for a long time and I would reinstall it along with tlp if it was removed as well. You can google these two packages for more information. – WinEunuuchs2Unix Dec 12 '18 at 02:09