On 09/06/2014 12:47 PM, Felix Miata wrote:
On 2014-09-06 13:14 (GMT+0200) Carlos E. R. composed:
On 2014-09-06 07:39, Felix Miata wrote:
On 2014-09-05 20:32 (GMT+0200) David Haller composed:
Press 'e' on the grub entry, add
init=/bin/bash
Tested with 13.1 in a vm. Rebooting is a problem though, shutdown, reboot, halt --reboot don't work (because systemd is not running! WHAT A STUPID setup!), exit/ctrl-d, ends in a kernel-panic in the vm. Haven't checked it out more now.
Fails in same manner on Rawhide, even using fullpath commands, though error from latter is different: "Failed to talk to init daemon". C-A-D doesn't work either. Great thing that systemd. :-(
If you use "init=/bin/bash" there is no systemd in use. It is completely out of the picture, so don't blame it when it ain't done nothing.
You mean besides subsuming everything it can into the monolith that it is, and forcing changes to other things that make them incompatible with sysvinit?
You'd have exactly the same problem with systemv, because it is not running. There is just one process running: bash. And of course there is not init daemon, that's bash.
In 11.4 booted with init=/bin/bash on cmdline:
/sbin/reboot reboots the box /sbin/halt -p turns the box off C-A-D reboots the box
None do more than produce error message with Lennart's wonderfully evolved sysvinit replacement on Rawhide. :-(
I think you are missing something here. First, this is a decidedly unnatural situation that has been contrived. You are not starting things with systemd so of course they can't be stopped by systemd. What the old 'halt/reboot' did was essentially pull the carpet of the OS from under the processes. Kaboom! Gone! If you ran though each of the sysvinint "K" series and all that was set up properly, many functions got shut down, but it was still possible to bypass all this. At the end of it, well, you have the 11.4 system, go look at what the final instruction was. You could run that and anyway and Kaboom! Gone! A point of systemd was to get around various problems the sysvinit had with communicating state to various processes. Its there in the design discussions and it clearly explains he problems surrounding the sysvinit style of shutdown and how many things can be left in an indeterminate state. I can't say if a fringe case like this entered into the scope of the design of systemd, but rather than just bemoaning and besmirching systemd, what not ask Lennart?
Interestingly, C-A-D does produce reboot on 13.1, but not on 13.2. :-p
Yes that is interesting. -- /"\ \ / ASCII Ribbon Campaign X Against HTML Mail / \ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org