On Fri, May 03, 2013 at 06:34:38AM -0600, Eric Blake wrote:
On 05/03/2013 06:15 AM, Daniel P. Berrange wrote:
> From: "Daniel P. Berrange" <berrange(a)redhat.com>
>
> The lockd plugin for the lock manager was not correctly
> handling the release of resource locks. This meant that
> during migration, or when pausing a VM, the locks would
> not get released. This in turn made it impossible to
> resume the domain, or finish migration
> ---
> src/locking/lock_driver_lockd.c | 38 ++++++++++++++++++++++++++------------
> 1 file changed, 26 insertions(+), 12 deletions(-)
>
> diff --git a/src/locking/lock_driver_lockd.c b/src/locking/lock_driver_lockd.c
> -
> - if (virNetClientProgramCall(program,
> - client,
> - counter++,
> - VIR_LOCK_SPACE_PROTOCOL_PROC_RELEASE_RESOURCE,
> - 0, NULL, NULL, NULL,
> -
(xdrproc_t)xdr_virLockSpaceProtocolReleaseResourceArgs, &args,
> - (xdrproc_t)xdr_void, NULL) < 0)
So the old code released the first resource,
> - goto cleanup;
> + for (i = 0 ; i < priv->nresources ; i++) {
but the new code loops over all resources. Yeah, I could see how that
causes a problem if more than one resource is held.
Actually the old code didn't release any resources since it failed to
fill in the resname field.
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 :|