Re: [Qemu-devel] [PATCH] qdev: obey no_user

[Luiz Capitulino <lcapitulino@redhat.com>]

On Mon, 07 Jan 2013 15:20:17 +0100
Andreas Färber <afaerber@suse.de> wrote:

> Am 17.12.2012 17:00, schrieb Christian Borntraeger:
> > since
> > 
> > commit 18b6dade8c0799c48f5c5e124b8c407cd5e22e96
> > qdev: refactor device creation to allow bus_info to be set only in class
> > 
> > A user can specify a device that is no_user.
> > For example on my i386 box, I can add a 2nd kvmvapic device.
> > 
> > This patch checks for no-user and rejects the device_add.
> > 
> > Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
> > ---
> >  hw/qdev-monitor.c | 5 +++++
> >  1 file changed, 5 insertions(+)
> > 
> > diff --git a/hw/qdev-monitor.c b/hw/qdev-monitor.c
> > index a1b4d6a..b2c34e7 100644
> > --- a/hw/qdev-monitor.c
> > +++ b/hw/qdev-monitor.c
> > @@ -426,6 +426,11 @@ DeviceState *qdev_device_add(QemuOpts *opts)
> >      }
> >  
> >      k = DEVICE_CLASS(obj);
> > +    if (k->no_user) {
> 
> Ack for this check...
> 
> > +        qerror_report(QERR_INVALID_PARAMETER_VALUE, "driver", "a driver name");
> > +        error_printf_unless_qmp("Try with argument 'help' for a list.\n");
> 
> ...but I'm not sure if this error reporting with QERR_... is still in
> the desired form, cc'ing Luiz.

The device_add command hasn't been converted to the qapi yet, so it's fine
to use qerror_report(). However, I'd drop the error_printf_unless_qmp() call,
as this has the potential of making the conversion harder.

> 
> Andreas
> 
> > +        return NULL;
> > +    }
> >  
> >      /* find bus */
> >      path = qemu_opt_get(opts, "bus");
>