[libvirt] [PATCH 3/4] qemu: monitor: Add vcpu state information to monitor data

[Peter Krempa]

On Wed, Sep 14, 2016 at 10:26:51 +0800, Dou Liyang wrote:
> Hi Peter,
> 
> At 09/14/2016 12:27 AM, Peter Krempa wrote:
> > Return whether a vcpu entry is hotpluggable or online so that upper
> > layers don't have to infer the information from other data.
> >
> > Advantage is that this code can be tested by unit tests.
> > ---
> >  src/qemu/qemu_monitor.c                            | 11 +++++
> >  src/qemu/qemu_monitor.h                            |  4 ++
> >  .../qemumonitorjson-cpuinfo-ppc64-basic.data       | 48 ++++++++++++++++++++++
> >  .../qemumonitorjson-cpuinfo-ppc64-hotplug-1.data   | 48 ++++++++++++++++++++++
> >  .../qemumonitorjson-cpuinfo-ppc64-hotplug-2.data   | 48 ++++++++++++++++++++++
> >  .../qemumonitorjson-cpuinfo-ppc64-hotplug-4.data   | 48 ++++++++++++++++++++++
> >  .../qemumonitorjson-cpuinfo-ppc64-no-threads.data  | 32 +++++++++++++++
> >  ...emumonitorjson-cpuinfo-x86-basic-pluggable.data | 16 ++++++++
> >  .../qemumonitorjson-cpuinfo-x86-full.data          | 22 ++++++++++
> >  tests/qemumonitorjsontest.c                        |  3 ++
> >  10 files changed, 280 insertions(+)
> >
> > diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c
> > index 4489997..e700b15 100644
> > --- a/src/qemu/qemu_monitor.c
> > +++ b/src/qemu/qemu_monitor.c
> > @@ -1773,6 +1773,7 @@ qemuMonitorGetCPUInfoHotplug(struct qemuMonitorQueryHotpluggableCpusEntry *hotpl
> >      int order = 1;
> >      size_t totalvcpus = 0;
> >      size_t mastervcpu; /* this iterator is used for iterating hotpluggable entities */
> > +    size_t slavevcpu; /* this corresponds to subentries of a hotpluggable entry */
> >      size_t anyvcpu; /* this iterator is used for any vcpu entry in the result */
> >      size_t i;
> >      size_t j;
> > @@ -1816,6 +1817,10 @@ qemuMonitorGetCPUInfoHotplug(struct qemuMonitorQueryHotpluggableCpusEntry *hotpl
> >       * logical vcpus in the guest */
> >      mastervcpu = 0;
> >      for (i = 0; i < nhotplugvcpus; i++) {
> > +        vcpus[mastervcpu].online = !!hotplugvcpus[i].qom_path;
> > +        vcpus[mastervcpu].hotpluggable = !!hotplugvcpus[i].alias;
> > +        if (!vcpus[mastervcpu].online)
> > +            vcpus[mastervcpu].hotpluggable = true;
> 
> I think if the vcpu don't have an alias, we mark it as non-hotpluggable.
> so, the code above may not fit well.

The hotplugvcpus.alias field is NULL in the following two cases:
1) vcpu is online but was selected as non-hotpluggable
2) the vcpu is offline

The hotpluggable field in libvirt is covering both the hotplug and
hotunplug case, thus a vcpu shall be marked as hotpluggable if:

1) it's offline
2) has the correct alias.

> I guess you may mean that:
> 
> vcpus[mastervcpu].online = !!hotplugvcpus[i].qom_path;
> vcpus[mastervcpu].hotpluggable = false;
> if (!vcpus[mastervcpu].online && (!!hotplugvcpus[i].alias))
>      vcpus[mastervcpu].hotpluggable = true;

This is wrong as online vcpus would not be marked as hotpluggable (which
by the way in the libvirt context also means that it can be unplugged).

> 
> OR
> 
> vcpus[mastervcpu].online = !!hotplugvcpus[i].qom_path;
> 
> if (vcpus[mastervcpu].online)
>      vcpus[mastervcpu].hotpluggable = false;

This is wrong in the same regards.

> else
>      vcpus[mastervcpu].hotpluggable = !!hotplugvcpus[i].alias;

Also offline vcpu won't ever have an alias, since it's offline.

Peter