Re: [Xen-devel] [PATCH v7 2/8] cxenstored: add support for systemd active sockets

[Ian Campbell <ian.campbell@xxxxxxxxxx>]

On Thu, 2015-08-06 at 10:20 +0100, Wei Liu wrote:
> On Thu, Aug 06, 2015 at 10:13:16AM +0100, Ian Campbell wrote:
> > On Wed, 2015-08-05 at 19:19 +0100, Wei Liu wrote:
> > > On Wed, Aug 05, 2015 at 06:24:37PM +0100, Wei Liu wrote:
> > > [...]
> > > > > 
> > > > 
> > > > Right. I misinterpreted sd_boot.
> > > > 
> > > > You patch, however, has the undesirable effect that it fails to 
> > > > report
> > > > error if xenstored is started by systemd but couldn't claim the
> > > > socket. I don't think this is the correct behaviour.
> > > > 
> > > > After consulting with systemd manual [0], I think we should check
> > > > sd_listen_fds return value to determine if it is started by 
> > > > systemd.
> > > > Currently it only checks for <= 0, which covers 1) not started by
> > > > systemd 2) an error occurs.
> > > > 
> > > > Hopefully I interpret the doc correctly this time. I will prepare a
> > > > patch shortly.
> > > > 
> > > > Wei.
> > > > 
> > > > [0] 
> > > > http://www.freedesktop.org/software/systemd/man/sd_listen_fds.html
> > > 
> > > Patch attached. I start cxenstored by hand and it seems to works fine 
> > > --
> > > now it fails with other errors.
> > 
> > What other errors? Are they blockers for accepting this patch?
> > 
> 
>  "FATAL: Failed to initialize dom0 state: Invalid argument"
> 
> Xenstored tries to initialise domain 0 state once again. That can be
> dealt with with a simple -D (--no-domain-init) option to stop it from
> doing that.

OK, is this just a symptom of having already started xenstored once from
the initscript and then trying to do it again?

If you disable the initscript/unit file then you should be able to start
xenstored manually (perhaps after having done some setup. e.g. mount
/proc/xen) without error I think.


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel