On Fri, May 13, 2016 at 01:28:34PM -0400, Cole Robinson wrote: > On 05/13/2016 01:19 PM, Pavel Hrdina wrote: > > Commit 510d28e3 refactored the seclabel code and one check was lost. > > Add the check back into the code. > > > > Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1298031 > > > > Signed-off-by: Pavel Hrdina <phrdina@xxxxxxxxxx> > > --- > > virtinst/seclabel.py | 5 +++++ > > 1 file changed, 5 insertions(+) > > > > diff --git a/virtinst/seclabel.py b/virtinst/seclabel.py > > index 10bf4e8..02221f0 100644 > > --- a/virtinst/seclabel.py > > +++ b/virtinst/seclabel.py > > @@ -87,6 +87,11 @@ class Seclabel(XMLBuilder): > > default_cb=_get_default_type, > > default_name=TYPE_DEFAULT) > > > > + def validate(self): > > + if self.type == self.TYPE_STATIC and not self.label: > > + raise RuntimeError(_("A label must be specified for 'static' " > > + "security type.")) > > + > > label = XMLProperty("./label") > > imagelabel = XMLProperty("./imagelabel") > > baselabel = XMLProperty("./baselabel") > > > > Hmm I think the bug is WONTFIX, since this just duplicates an explicit error > that libvirt will already throw for us. Yes it's slightly more annoying for > the user since virt-install will fully set up the VM before reporting the > error, but that's the case with hundreds of other validation checks. > > The only time bits like this are really important IMO is if a) it's via the > virt-manager UI and we want to use the validation check to block an action > early, or b) it's a really common misconfiguration. Neither cover this case > > Thanks, > Cole I'm OK with WONTFIX, it was really easy patch, so I sent it anyway. I'll close that bug. Thanks, Pavel _______________________________________________ virt-tools-list mailing list virt-tools-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/virt-tools-list
