[systemd-devel] heads-up: chasing journal(?) related regression in 219 causing boot hang/fail

[Lennart Poettering]

On Sun, 22.02.15 17:55, Martin Pitt (martin.pitt at ubuntu.com) wrote:

> Lennart Poettering [2015-02-20 17:02 +0100]:
> > To me this appears as if dbus is hanging for some reason. Have you
> > checked what dbus is doing?
> 
> D-Bus itself seems to be fine. There are services on it, busctl works,
> etc.
> 
> Anyway, I now have a capable enough arsenal to reproduce that hang
> fully automatically and a git bisect run script, which after a few
> hours of grinding spat out that this is the culprit:
> 
>   http://cgit.freedesktop.org/systemd/systemd/commit/?id=13790add4bf64
>   (journald: allow restarting journald without losing stream connections)
> 
> and indeed reverting that on top of current git master (reverts
> cleanly) makes things work perfectly again.
> 
> I haven't drilled down into the patch itself yet, that's not something
> I want to start doing on a Sunday :-)

Any idea about the details of this? Is this reproducible with
unpatched 219 (in particular without the fsckd patches applied)

I have never seen this issue, can you tell me how to reproduce this on
my machine?

Does your boot process restart journald or so?

Lennart

-- 
Lennart Poettering, Red Hat