I did a final remaster of a new Kde install yesterday, before changing the kernel in it. I first rebooted with the old kernel still installed, incase of problems.
At first boot, I got a segfault. So I thought I would reactivate dkms and hardrake in the Services. But dkms was activated.
It went so : I booted again on 2.6.26.tex3 kernel, reinstalled all dkms packages installed so that the modules would be there in the new kernel, then I checked dkms and harddrake in the Services (I don't know wether or not it should be done) so I finally saw that the new kernel did boot allright with hardrake activated and dkms not activated. I booted once or twice.
Then I removed the 2.6.26.tex3 from within Synaptic, and rebooted (with dkms and hardrake detection disabled in Services).
And now it does not boot anymore. /o\
I did some screenshots to show how it goes:
First boot on new kernel, with 2 kernels there:
Second boot of the new kernel, (after reinstalling all dkms packages installed and hardrake activated)
That worked !
Third boot, after I disabled hardrake detection in the Services:
doesn't work anymore.
I tried to reactivate the relevant services with drakxservices in chroot (VirtualBox allows easy chaining of the virtual machines).
I did :
where the kde vm was mounted, then as root:
mount -o bind /dev/ dev/
mount -t proc none proc/
Then I started the services and activated all that need to be:
Then I shutdown the Gnome pclos vm, put the Kde pclos vm back to master in the preferences, rebooted on it, but it's still looking like that, any entry I would try (normal or failsafe)
Did anyone try to switch kernel in virtual machine so far ?
How would you do that so that no such problem arises ?