On Tue, 15 Oct 2019 at 11:56, Andrew Jones <drjo...@redhat.com> wrote:
>
> On Tue, Oct 15, 2019 at 10:59:16AM +0100, Beata Michalska wrote:
> > On Tue, 1 Oct 2019 at 14:04, Andrew Jones <drjo...@redhat.com> wrote:
> > > +
> > > + obj = object_new(object_class_get_name(oc));
> > > +
> > > + if (qdict_in) {
> > > + Visitor *visitor;
> > > + Error *err = NULL;
> > > +
> > > + visitor = qobject_input_visitor_new(model->props);
> > > + visit_start_struct(visitor, NULL, NULL, 0, &err);
> > > + if (err) {
> > > + object_unref(obj);
> >
> > Shouldn't we free the 'visitor' here as well ?
>
> Yes. Good catch. So we also need to fix
> target/s390x/cpu_models.c:cpu_model_from_info(), which has the same
> construction (the construction from which I derived this)
>
> >
> > > + error_propagate(errp, err);
> > > + return NULL;
> > > + }
> > > +
>
> What about the rest of the patch? With that fixed for v6 can I
> add your r-b?
>
I still got this feeling that we could optimize that a bit - which I'm
currently on, so hopefully I'll be able to add more comments soon if
that proves to be the case.
BR
Beata
> Thanks,
> drew
