Re: [Qemu-devel] [PATCH RFC V2 6/9] vl.c: do not set 'type' property in obj_set_property

[Marcel Apfelbaum]

On Mon, 2014-03-03 at 11:11 +0100, Paolo Bonzini wrote:
> Il 02/03/2014 14:07, Marcel Apfelbaum ha scritto:
> > Filter out also 'type' property when setting
> > object's properties
> >
> > Signed-off-by: Marcel Apfelbaum <address@hidden>
> > ---
> >  vl.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/vl.c b/vl.c
> > index c4939ef..dc206e1 100644
> > --- a/vl.c
> > +++ b/vl.c
> > @@ -2766,7 +2766,8 @@ static int object_set_property(const char *name, 
> > const char *value, void *opaque
> >      StringInputVisitor *siv;
> >      Error *local_err = NULL;
> >
> > -    if (strcmp(name, "qom-type") == 0 || strcmp(name, "id") == 0) {
> > +    if (strcmp(name, "qom-type") == 0 || strcmp(name, "id") == 0 ||
> > +        strcmp(name, "type") == 0) {
> >          return 0;
> >      }
> >
> >
> 
> Unfortunately, it is quite possible to have a -object invocation where 
> the object has a "type" property.
Thanks, I was thinking that "type" is a built-in property of the Object,
derived directly from class's type, so no one can "set" this property.
> 
> I think you could change the -object implementation to use OptsVisitor, 
> similar to hmp_object_add in hmp.c.  Then the -object code can pre-parse 
> "qom-type" and "id", while the -machine code can pre-parse "type".
I'll look into it, thanks for the tip.

Thanks,
Marcel

> 
> Paolo