On Mon, Dec 05, 2016 at 08:48:43AM +0100, Juergen Gross wrote: > Instead of calling add_change_node() at places where write_node() is > called, do that inside write_node(). > > Note that there is one case where add_change_node() is called now when > a later failure will prohibit the changed node to be written: in case > of a write_node failing due to an error in tdb_store(). As the only > visible change of behavior is a stale event fired for the node, while > the failing tdb_store() signals a corrupted xenstore database, the > stale event will be the least problem of this case. > > Signed-off-by: Juergen Gross <jgr...@suse.com>
Acked-by: Wei Liu <wei.l...@citrix.com> _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
