On 26.06.2013 04:51, Carlos E. R. wrote:
On Tuesday, 2013-06-25 at 21:13 +0400, Andrey Borzenkov wrote:
В Tue, 25 Jun 2013 13:02:23 -0400 Felix Miata <mrmazda@earthlink.net> пишет:
On 2013-06-25 18:12 (GMT+0200) Stephan Kulow composed:
It looks like I just broke my laptop by updating to factory, I can't get past the initrd with something that sounds like broken storage setup in systemd/mkinitrd/udev. So I advise everyone not to update before I found out it was my fault :)
This is the other thing that routinely annoys me about working initrds getting rebuilt every time any package that ever affects what makes up an initrd gets an update[1]. Why shouldn't working initrds be left alone, and new packages get put into them only either when a new kernel is installed, mkinitrd is called by a user, or an affirmative answer is given to a request to rebuild by an updated package installation? The way initrds get clobbered now, what point is there to enabling multiversion for kernels?
I dare to guess that the reason is to ensure that binaries in initrd match binaries on your system.
But it is an interesting point.
If we have multiversion, but initrd is remade for both old and current versions, a failure breaks all kernel versions. We can not be sure that the old version will keep working as a failsafe.
It would have saved my laptop, but then again people whos system is fixed by an update would love to have the initrd updated on updates - so where shall we stop? I guess we should leave old kernels with old initrds, but I don't think we should keep an old initrd for the new kernel. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org