On 08/27/2018 04:08 AM, Michal Privoznik wrote: > At the beginning of each dispatch function we check if owner > attributes were registered (these consist of ID, UUID, PID and > name). The check then consists of checking if ID is not zero. > This is not going to work with > VIR_LOCK_MANAGER_OBJECT_TYPE_DAEMON which doesn't set ID. Switch > to setting PID which is available for both cases. > > Signed-off-by: Michal Privoznik <mprivozn@xxxxxxxxxx> > --- > src/locking/lock_daemon_dispatch.c | 14 +++++++------- > 1 file changed, 7 insertions(+), 7 deletions(-) > BTW: My idea of setting id == -1 for deamon still works without any change required in/for this patch. So what would be concerning about using ownerPid would be that over the lifetime of the host the @pid can recycle; whereas, during the lifetime of the daemon don't we guarantee that the @id will be ever increasing? But right now I'm too lazy to go look and see if getting the next id is through libvirtd or virtlockd. Not against this, but I need to get feedback from earlier patches and of course your thoughts on the @id vs. @pid rotation. Plus I need see how this plays out in future patches. John -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list
