I'm running Ubuntu 18.04 x86_64, fully patched with dist-upgrade
. I have an old kernel and I am not sure if it safe to remove:
$ ls /boot/
config-4.15.0-64-generic memtest86+.elf
config-5.0.0-29-generic memtest86+_multiboot.bin
efi System.map-4.15.0-64-generic
grub System.map-5.0.0-29-generic
initrd.img-4.15.0-64-generic vmlinuz-4.15.0-64-generic
initrd.img-5.0.0-29-generic vmlinuz-5.0.0-29-generic
memtest86+.bin
I also have Nvidia and Virtual Box DKMS's in /usr/lib
for the kernels. I used dkms remove ...
to remove the 4.x modules, but they keep returning:
$ ls /lib/modules/
4.15.0-64-generic 5.0.0-29-generic
$ dkms status
nvidia, 430.26, 4.15.0-64-generic, x86_64: installed
nvidia, 430.26, 5.0.0-29-generic, x86_64: installed
virtualbox, 5.2.32, 4.15.0-64-generic, x86_64: installed
virtualbox, 5.2.32, 5.0.0-29-generic, x86_64: installed
I'm catching random crashes and I believe they are due to video drivers. Each boot and each login generates a Crash Report (not a panic). The crash happens with both Nouveau (generic X driver) and Nvidia (nvidia-driver-430). Unfortunately, Ubuntu does not provide the details. I am only offered the choice to Cancel or Send (and not View).
I'd like to remove the old kernel in hopes of clearing the issue. The rub is, the last time I tired to do it with SecureBoot on and signed kernels, the machine stooped booting.
My question is, is it safe to remove old Ubuntu 4.x kernel in this configuration?