Re: [Qemu-devel] [PATCH 6/8] block: add eject request callback

[Paolo Bonzini]

On 10/28/2011 07:21 PM, Kevin Wolf wrote:
> > >  -    if (!force&&  !bdrv_dev_is_tray_open(bs)
> > >  -&&  bdrv_dev_is_medium_locked(bs)) {
> > >  -        qerror_report(QERR_DEVICE_LOCKED, bdrv_get_device_name(bs));
> > >  +    if (bdrv_dev_is_medium_locked(bs)&&  !bdrv_dev_is_tray_open(bs)) {
> > >  +        bdrv_dev_eject_request(bs, force);
> > >  +        if (!force) {
> > >  +            qerror_report(QERR_DEVICE_LOCKED, bdrv_get_device_name(bs));
> > >  +        }
> > >            return -1;
> > >        }
> > >        bdrv_close(bs);
> Now force doesn't force any more. It avoids the error message, but
> doesn't forcefully close the BlockDriverState any more. Intentional? If
> so, why is it a good idea?

In theory the guest OS should eject the disk itself.  However, force 
does unlock the disk so that: 1) two ejects will have the desired 
effect; 2) force eject followed by change will work even with the tray 
locked, unlike before this series.

Paolo