On 07/19/2011 09:57 AM, Osier Yang wrote:
---
src/esx/esx_driver.c | 11 ++++++++++-
src/lxc/lxc_driver.c | 11 ++++++++++-
src/openvz/openvz_driver.c | 11 ++++++++++-
src/test/test_driver.c | 12 +++++++++++-
src/uml/uml_driver.c | 12 +++++++++++-
src/vmware/vmware_driver.c | 12 +++++++++++-
src/xen/xen_driver.c | 9 ++++++++-
7 files changed, 71 insertions(+), 7 deletions(-)
+++ b/src/test/test_driver.c
@@ -2535,12 +2535,16 @@ static int testDomainCreate(virDomainPtr domain) {
return testDomainCreateWithFlags(domain, 0);
}
-static int testDomainUndefine(virDomainPtr domain) {
+static int testDomainUndefineFlags(virDomainPtr domain,
+ unsigned int flags)
+{
testConnPtr privconn = domain->conn->privateData;
virDomainObjPtr privdom;
virDomainEventPtr event = NULL;
int ret = -1;
+ virCheckFlags(0, -1);
Huh, I just realized:
Since these hypervisors do not support managed save in the first place,
we could argue that they trivially support
VIR_DOMAIN_UNDEFINE_MANAGED_SAVE (that is, they properly guarantee that
all managed save files are safely deleted, since none exist in the first
place).
We could change all of these virCheckFlags(0, -1) into
virCheckFlags(VIR_DOMAIN_UNDEFINE_MANAGED_SAVE, -1), so that 'virsh
undefine dom --managed-save' would also work on these hypervisors.
I'm not sure whether to make that change though, so let's push this
patch as-is, and debate about that change as a separate item.
ACK.
--
Eric Blake eblake@xxxxxxxxxx +1-801-349-2682
Libvirt virtualization library http://libvirt.org
--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list