[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [libvirt] [PATCH v3] nwfilter:




libvir-list-bounces redhat com wrote on 08/13/2010 04:44:38 PM:


> Incomplete subject line?  It used to be
>  nwfilter: extend nwfilter reload support


Yes. Due to a race condition :-)

>
> On 08/13/2010 02:29 PM, Stefan Berger wrote:
> >  v3:
> >     - Fixed an indentation problem
> >     - added bool parameter to function terminating the IP address
> > learner threads to determine whether future threads may still run
> > (needed in case of driver reload) or all must terminate (need in case of
> > libvirtd termination)
> >
>
> > +    conn = virConnectOpen("qemu:///system");
> > +
> > +    if (conn) {
> > +        /* shut down all threads -- qemud for example will restart them */
> > +        virNWFilterLearnThreadsTerminate(true);
> > +
> > +        nwfilterDriverLock(driverState);
> > +        virNWFilterPoolLoadAllConfigs(conn,
> > + &driverState->pools,
>
> Indentation problem still looks like it is here.


The indentation problem doesn't exist in my local patch. Could this be another Thunderbird formatting issue?

>
> > @@ -1912,7 +1926,11 @@ qemudStartup(int privileged) {
> >                                  1, NULL, NULL) < 0)
> >          goto error;
> >
> > -    qemuReconnectDomains(qemu_driver);
> > +    conn = virConnectOpen(qemu_driver->privileged ?
> > +                          "qemu:///system" :
> > +                          "qemu:///session");
>
> How come this one is conditional, but the other one was hard-coded to
> system?


The same type of check exists several times in this file of the qemu driver. I ended up hard-coding it in the nwfilter driver case above since I don't have a flag in the nwfilter driver for whether it's privileged or not and I don't know what the difference would be.

Stefan
[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]