On Tue, 15.02.11 10:47, Andrey Borzenkov (arvidj...@gmail.com) wrote: > > On Mon, Feb 14, 2011 at 12:22 AM, Lennart Poettering > <lenn...@poettering.net> wrote: > > On Thu, 10.02.11 07:02, Andrey Borzenkov (arvidj...@gmail.com) wrote: > > > >> --- a/units/plymouth-start.service > >> +++ b/units/plymouth-start.service > >> @@ -12,7 +12,13 @@ Wants=systemd-ask-password-plymouth.path > >> After=systemd-vconsole-setup.service udev-settle.service > >> Before=systemd-ask-password-plymouth.service > >> > >> +# Dracut informs us with this flag file if plymouth is already running > >> +ConditionPathExists=!/dev/.systemd/plymouth > >> + > >> [Service] > >> ExecStart=/sbin/plymouthd --mode=boot > >> ExecStartPost=-/bin/plymouth --show-splash > >> > > We added this flag file mostly to make it unnecessary to spawn plymouth > > twice. > > > > Hmm ... how is systemd-ask-password-plymouth.service launched then? > Currently it pulled in by plymouth-start.service; but if > plymouth-start is skipped ...
ConditionXXXX= is a very very weak version of disabling. Basically all this does is that the actual execution of the binary is skipped, otherwise the service is processed as if it was to be started. That means all its dependencies are pulled in and they are executed in order. It's really just that the execution of the service binary itself is skipped and assumed to have succeeded. In this case this means that s-a-p-p.p is still pulled in. > Assuming we push patch to created flag file by plymouth (is patch > available anywhere BTW?); can systemd trigger unit startup when path > becomes available? The patch isn't available yet. Follow the bug I linked. Lennart -- Lennart Poettering - Red Hat, Inc. _______________________________________________ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel
