[systemd-devel] systemd --user sometimes is not stopped by systemd-exit.service

[Andrey Borzenkov]

В Tue, 1 Oct 2013 04:34:36 +0200
Lennart Poettering <lennart at poettering.net> пишет:

> On Thu, 26.09.13 09:41, Andrey Borzenkov (arvidjaar at gmail.com) wrote:
> 
> > I'm still observing timeouts stopping user at 0.service on shutdown. It
> > appears to be some race condition, because enabling debugging of either
> > main systemd or user systemd, or even changing console from vt to
> > serial would hide it. Sometimes it also works using vt console as well.
> > 
> > Here is difference between good case and bad case.
> 
> Hmm, if you patch systemd to generate a message when RTMAX-6 is received
> (and only then, leaving debug mode off), do you see this message
> generated? i.e. I'd really like to know whether the shutdown signal
> tthat is sent from the user systemd's systemd-exit.service to the user
> systemd process is actually received...
> 

systemd does not receive SIGRTMIN+24

> 
> Race condition? Which one?
> 

Well, between sending and receiving ... if I new better ... :)

> The workaround patch is nothing we can merge. We need the ability for
> people to install user units that are invoked before the session goes
> down, i.e. units that are ordered against exit.target and
> systemd-exit.service... This is the whole reason we do this indirection
> via signal sending at all...
> 

I realized that patch was completely wrong after that.