On Mon, Aug 11, 2014 at 12:15:41PM +0200, Giuseppe Scrivano wrote: > "Daniel P. Berrange" <berrange@xxxxxxxxxx> writes: > > > On Thu, Aug 07, 2014 at 04:10:31PM +0200, Giuseppe Scrivano wrote: > >> Signed-off-by: Giuseppe Scrivano <gscrivan@xxxxxxxxxx> > >> --- > >> src/conf/domain_conf.c | 34 +++++++++++++++++++++++++--------- > >> src/conf/domain_conf.h | 1 + > >> 2 files changed, 26 insertions(+), 9 deletions(-) > >> > >> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c > >> index c25c74b..3bdf46a 100644 > >> --- a/src/conf/domain_conf.c > >> +++ b/src/conf/domain_conf.c > >> @@ -342,7 +342,8 @@ VIR_ENUM_IMPL(virDomainFS, VIR_DOMAIN_FS_TYPE_LAST, > >> "file", > >> "template", > >> "ram", > >> - "bind") > >> + "bind", > >> + "mtp") > > > > I don't think this is the right way to represent it. > > > > The 'type' attribute on <filesystem> represents where the backing store > > for the filesystem comes from. > > > > The distinction of 9p vs mtp reflects the type of guest device to expose > > it as. > > > > We shouldn't try to overload these two concepts in the same attribute. > > We should instead try to add a <device> or <model> child element as we > > have for some other device types. > > I see, thanks for the clarification. > > Would you agree with something like this? > > <filesystem type='mount'> > <device name="mtp share">mtp</device> > ... What is the name="mtp share" bit trying to reflect ? It seems we're mostly biased towards <model> so I think we should aim for <model type='mtp|9p'/> Regards, 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 :| -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list