[systemd-devel] systemd intentional behaviour? (signals)

[Lennart Poettering]

On Fri, 07.09.12 15:15, Gerardo Exequiel Pozzi (vmlinuz386 at yahoo.com.ar) wrote:

> >>Is this intentional? There is a way to restore systemd again without
> >>a forced reboot?
> >Yes, this is intentional. If we crash (or the user kills us) we print a
> >warning and freeze. If the user kills us again then we return control to
> >the kernel. Sounds like a really reasonable reaction to some really
> >pointless action by the user.
> >
> >I mean, the user issued the kill commands, hence he probably has a
> >reason to, even if that reason is "I want to shoot myself in the
> >foot" -- and hence we do what we do.
> >
> >Lennart
> >
> 
> OK, so in case of and internal crash, there is no way, at least this
> moment (maybe in a future?), to restore from "freeze" without doing
> a forced reboot?

If we are crashed we are crashed.

I looked into reexecing systemd in this case, but unfortunately the
kernel does not allow reexecing of PID 1 after it crashed. Not entirely
sure what's going on there. But even if that worked I am not sure this
would be too useful since most of its state would be lost...

Lennart

-- 
Lennart Poettering - Red Hat, Inc.