Le 05/02/17 à 16:56, Laurent Bigonville a écrit :
Le 05/02/17 à 11:48, Niels Thykier a écrit :
Hi,
Are there any news on this release critical bug affecting both jessie
and stretch?
Note that comment #35 might be interesting where Алексей Шилин writes:
I decided to try to postpone KDM startup until after plymouth quits,
so I
created /etc/systemd/system/kdm.service.d/wait4plymouth.conf with
the following
content:
-------------------------------------- 8<
--------------------------------------
[Unit]
Wants=plymouth-quit-wait.service
After=plymouth-quit-wait.service
-------------------------------------- >8
--------------------------------------
Indeed, after that I can't reproduce the issue anymore.
So, it seems to be a services' dependency problem.
I would say that this should be reassigned to the DM then.
Adding Wants=plymouth-quit-wait.service should be enough to ensure
that the splash-screen is stopped, it should be part of the default
dependency chain so the Wants= is not needed IMHO.
Wants=plymouth-quit.service might actually be enough if we look at other
DM, like the upstream .service file of sddm (the debian one seems boggus
to me, I opened #854260 for this).
But all of this actually need testing.
Also the bug from Carlos Prieto López at message #42 looks different to
me, plymouth already be killed when needed by GDM, so not too sure
what's wrong here.
Cheers,
Laurent Bigonville