[pkg-cryptsetup-devel] Bug#768314: Bug#768314: cryptsetup: Passphrase prompt rolls by without stopping

Jonas Meurer jonas at freesources.org
Mon Dec 15 20:32:42 UTC 2014


clone 768314 -1
severity -1 normal
retitle -1 systemd without plymouth kills input prompts at boot process
reassign -1 release-notes
close 768314
thanks

Hi,

Am 15.12.2014 um 20:50 schrieb Kjetil Kjernsmo:
>> Am 2014-12-15 12:07, schrieb Kjetil Kjernsmo:
>> > I think it is systemd (no conscious decision from my side):
>>
>> as written earlier, systemd has it's own cryptsetup agent and
>> uses that one to unlock encrypted devices instead of the the
>> cryptdisks initscripts used by sysvinit.
>>
>> the problem is, that systemd doesn't wait for optional user
>> input after running cryptsetup. the boot logger plymouth is
>> responsible for this task.
> 
> Aha, OK!
> 
>> could you try to install plymouth on your system and see whether
>> that fixes your issues? you can configure plymouth to use a
>> text-mode theme if you don't like the graphical boot splash
>> screen that hides all the boot log messages.
> 
> Ah, yeah, that did work! In fact, I only had to write the passphrase
> once, which is kinda weird, what if I didn't use the same passphrase for
> both partitions? (Oh, I shouldn't say that I do ;-))

Good to know. So at least we know the root for this issue now :)

>> if missing plymouth is the reason for the issues you discovered,
>> then i fear that there's no easy solution to fix this bug. it seems
>> like systemd simply requires something like plymouth to properly
>> display user prompts at boot time :-/
> 
> Mmmm, yeah... And I can't use a different init at this point?

Sure, you can. Sysvinit should still be supported as well as openrc.
'apt-get install sysvinit-core systemd-shim' should do the trick, but I
didn't test that myself yet.

>> that's a separate issue. seems like systemd indeed doesn't support
>> separate /usr partitions.
> 
> Hmpf, yeah... :-( And plymouth didn't fix it... :-(

Sorry, but I cannot help here. I suggest that you discuss this issue on
debian-user or on IRC in #debian-systemd. But read this as well:
https://wiki.debian.org/systemd#Booting_with_lvm_.28especially_with_separate_.2Fusr.29_fails
https://lists.debian.org/debian-user/2014/10/msg02504.html
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=742048

> Oh well, I suppose you could say wontfix on this bug, then!

Honestly, I prefer to close this bugreport. It's not a bug in cryptsetup
at all, rather a shortcoming of systemd without plymouth that is intended.

Maybe the best would be to add a note to jessie release notes? I cloned
the bugreport and assigned it to release-notes with the bts commands above.

Cheers,
 jonas



More information about the pkg-cryptsetup-devel mailing list