[libvirt] [libvirt-glib 6/6] Add gvir_domain_get_devices()

[Christophe Fergeau]

On Wed, Feb 29, 2012 at 04:01:40PM +0200, Zeeshan Ali (Khattak) wrote:
> On Wed, Feb 29, 2012 at 3:30 PM, Christophe Fergeau <cfergeau at redhat.com> wrote:
> > On Tue, Feb 28, 2012 at 08:25:07PM +0200, Zeeshan Ali (Khattak) wrote:
> >> From: "Zeeshan Ali (Khattak)" <zeeshanak at gnome.org>
> >>
> >> Currently we only support existing DomainDevice implementations:
> >> DomainDisk and DomainInterface.
> >> ---
> >>  .../libvirt-gobject-domain-device-private.h        |    2 +
> >>  libvirt-gobject/libvirt-gobject-domain-device.c    |   21 ++++++++++
> >>  libvirt-gobject/libvirt-gobject-domain.c           |   42 ++++++++++++++++++++
> >>  libvirt-gobject/libvirt-gobject-domain.h           |    3 +
> >>  libvirt-gobject/libvirt-gobject.sym                |    1 +
> >>  5 files changed, 69 insertions(+), 0 deletions(-)
> >>
> >> diff --git a/libvirt-gobject/libvirt-gobject-domain-device-private.h b/libvirt-gobject/libvirt-gobject-domain-device-private.h
> >> index 72c660e..292a2ac 100644
> >> --- a/libvirt-gobject/libvirt-gobject-domain-device-private.h
> >> +++ b/libvirt-gobject/libvirt-gobject-domain-device-private.h
> >> @@ -24,6 +24,8 @@
> >>
> >>  G_BEGIN_DECLS
> >>
> >> +G_GNUC_INTERNAL GVirDomainDevice *gvir_domain_device_new(GVirConfigDomainDevice *config,
> >> +                                                         GVirDomain *domain);
> >>  virDomainPtr gvir_domain_device_get_domain_handle(GVirDomainDevice *self);
> >>
> >>  G_END_DECLS
> >> diff --git a/libvirt-gobject/libvirt-gobject-domain-device.c b/libvirt-gobject/libvirt-gobject-domain-device.c
> >> index 85879d2..0ec5dad 100644
> >> --- a/libvirt-gobject/libvirt-gobject-domain-device.c
> >> +++ b/libvirt-gobject/libvirt-gobject-domain-device.c
> >> @@ -176,3 +176,24 @@ GVirDomain *gvir_domain_device_get_domain(GVirDomainDevice *device) {
> >>  GVirConfigDomainDevice *gvir_domain_device_get_config(GVirDomainDevice *device) {
> >>      return g_object_ref (device->priv->config);
> >>  }
> >> +
> >> +G_GNUC_INTERNAL GVirDomainDevice *gvir_domain_device_new(GVirConfigDomainDevice *config,
> >> +                                                         GVirDomain *domain)
> >
> > gvir_domain_new_device(GVirDomain *domain, GVirConfigDomainDevice *config)
> > maybe ?
> > Having the argument that is always non-NULL first feels better to me.
> 
> Who said config is nullable? :)

I assumed this because of the lack of g_return_val_if_fail(GVIR_CONFIG_IS_DOMAIN_DEVICE(config), ...)
It still kind of sounds better to me to say "let's create a device
associated with this GVirDomain", ie to have this method operating on a
GVirDomain.

> 
> >> +{
> >> +    GType type;
> >> +
> >> +    g_return_val_if_fail(GVIR_CONFIG_IS_DOMAIN_DEVICE(config), NULL);
> >> +
> >> +    if (GVIR_CONFIG_IS_DOMAIN_DISK(config))
> >> +        type = GVIR_TYPE_DOMAIN_DISK;
> >> +    else if (GVIR_CONFIG_IS_DOMAIN_INTERFACE(config))
> >> +        type = GVIR_TYPE_DOMAIN_INTERFACE;
> >> +    else {
> >> +        g_debug("Unknown device type: %s", G_OBJECT_TYPE_NAME(config));
> >> +        return NULL;
> >> +    }
> >
> > I'd have a slight preference for something like
> >
> > switch (G_OBJECT_TYPE(config)) {
> > case GVIR_CONFIG_TYPE_DOMAIN_DISK:
> >    type = GVIR_TYPE_DOMAIN_DISK;
> >    break;
> > ...
> >
> > but I'm fine with either version.
> 
> I would have prefered it too but if any of these classes are
> subclassed tomorrow and we forget to update this code, we'll start
> loosing config instances we can actually handle.

Ah good point.

> 
> >> +
> >> +        device_config = GVIR_CONFIG_DOMAIN_DEVICE(node->data);
> >> +        device = gvir_domain_device_new(device_config, domain);
> >> +        if (device != NULL)
> >> +             ret = g_list_append(ret, device);
> >
> > slight preference for g_list_prepend (much more efficient when the list is
> > big).
> 
> Good point but wouldn't you expect devices to be sorted in the order
> they are sorted in the XML?

You can always do a g_list_reverse after the fact, and I'd rather we avoid
to make such guarantees anyway. If gvir_config_domain_get_devices used
g_list_prepend (which I just noticed it's not doing), you'd need to prepend
to return things in the order of the XML :)
Not that important here, it's just a better practice to use g_list_prepend.

> 
> >> +
> >> +        g_object_unref (device_config);
> >> +    }
> >> +    g_list_free (config_devices);
> >
> > The individual elements need to be unref'ed too as well as config.
> 
> Thats what g_object_unref is doing just above? I think I should name

True, I didn't pay enough attention.

> this 'device_configs' to be consistent with 'device_config' and
> therefore clear..

I could pretend that's what confused me... :) This would be better indeed
now that you say it.

Christophe
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/libvir-list/attachments/20120229/0a5c583e/attachment-0001.sig>