"Daniel P. Berrange" <berra...@redhat.com> writes: > On Fri, Jul 27, 2012 at 12:49:53PM +0200, Paolo Bonzini wrote: >> Il 27/07/2012 12:33, Daniel P. Berrange ha scritto: >> > #qemu-img info -j /var/lib/libvirt/images/bar.qcow2 >> > { >> > "information": { >> > "actual_size": "139264", >> > "fmt": "qcow2", >> > "virtual_size": "10485760", >> > "filename": "/var/lib/libvirt/images/bar.qcow2", >> > "cluster_size": 65536, >> > "encrypted": 0, >> > "snapshot_list": [ >> > ], >> > "dirty_flag": 0, >> > "backing_filename": "/dev/sda1" >> > }, >> > "return": 0 >> > } >> > >> > IIUC,the 'return' element here is just duplicating the qemu-img >> > exit status. I think this is rather dubious, and would rather >> > just see the stuff in the 'information' sub-block be output >> > directly. It also seems to forget to mention the backing >> > file format. >> >> I wonder if we could add this also to the QEMU monitor ("info >> block-image foo"), so that the code would be shared between qemu-img.c >> and QEMU. > > It would certainly make sense to have the code & data format shared > between the two, so apps don't need to have 2 different JSON parsers
We should definitely describe the output in qapi-schema.json. Whether we also expose a QMP command--I'm not really convinced. It seems a bit unusual to me to want to get this info from an arbitrary image file from within QEMU. I'm not sure I see the use-case. I can definitely see it become exposed as part of query-block though as an image-info parameter. Regards, Anthony Liguori > > > Daniel > -- > |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| > |: http://libvirt.org -o- http://virt-manager.org :| > |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| > |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|