Le Wednesday 04 Dec 2013 à 16:56:05 (-0700), Eric Blake a écrit : > On 12/03/2013 06:26 AM, Benoît Canet wrote: > > Signed-off-by: Benoit Canet <ben...@irqsave.net> > > --- > > > > +BlockDriverState * bdrv_lookup_bs(bool has_device, const char * device, > > + bool has_node_name, const char * > > node_name, > > Style: no space after * (3 instances) > > > + Error **errp) > > +{ > > + BlockDriverState *bs = NULL; > > + > > + if ((has_device && has_node_name) || > > + (!has_device && !has_node_name)) { > > Could be shortened to: > > if (has_device == has_node_name) { > > > + error_setg(errp, "Use either device or node-name but not both."); > > We tend to avoid trailing '.' on error messages > > > > > -void qmp_block_passwd(const char *device, const char *password, Error > > **errp) > > +void qmp_block_passwd(bool has_device, const char * device, > > + bool has_node_name, const char * node_name, > > + const char * password, Error **errp) > > Again, no space after '*' > > > +++ b/include/block/block.h > > @@ -371,6 +371,9 @@ void bdrv_eject(BlockDriverState *bs, bool eject_flag); > > const char *bdrv_get_format_name(BlockDriverState *bs); > > BlockDriverState *bdrv_find(const char *name); > > BlockDriverState *bdrv_find_node(const char *node_name); > > +BlockDriverState * bdrv_lookup_bs(bool has_device, const char * device, > > + bool has_node_name, const char * > > node_name, > > + Error **errp); > > And again > > > +++ b/qapi-schema.json > > @@ -1675,7 +1675,11 @@ > > # determine which ones are encrypted, set the passwords with this command, > > and > > # then start the guest with the @cont command. > > # > > -# @device: the name of the device to set the password on > > +# Either @device or @node-name must be set but not both. > > +# > > +# @device: #optional the name of the block backend device to set the > > password on > > +# > > +# @node-name: #optional graph node name to set the password on (Since 1.8) > > 2.0 > > > # > > # @password: the password to use for the device > > # > > @@ -1689,7 +1693,8 @@ > > # > > # Since: 0.14.0 > > ## > > -{ 'command': 'block_passwd', 'data': {'device': 'str', 'password': 'str'} } > > +{ 'command': 'block_passwd', 'data': {'*device': 'str', > > + '*node-name': 'str', 'password': > > 'str'} } > > Seems like a reasonable addition; shouldn't cause any back-compat > problems (older management tools will always provide the now-optional > 'device'). > > Is it intentional that you are not exposing this new functionality in HMP?
Yes, I don't foresee any way to print the graph in HMP so I am limiting the changes to QMP. Best regards Benoît > > -- > Eric Blake eblake redhat com +1-919-301-3266 > Libvirt virtualization library http://libvirt.org >