rbd: don't retry watch reregistration if header object is gone

If the header object gets deleted (perhaps along with the entire pool), there is no point in attempting to reregister the watch. Treat this the same as blacklisting: fail all pending and new I/Os requiring the lock. Signed-off-by: Ilya Dryomov <[email protected]>
niw · Oct 15, 2016 · 4d73644 · 4d73644
1 parent 87c0fde
commit 4d73644
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
@@ -3944,7 +3944,7 @@ static void rbd_reregister_watch(struct work_struct *work)
 	ret = __rbd_register_watch(rbd_dev);
 	if (ret) {
 		rbd_warn(rbd_dev, "failed to reregister watch: %d", ret);
-		if (ret == -EBLACKLISTED) {
+		if (ret == -EBLACKLISTED || ret == -ENOENT) {
 			set_bit(RBD_DEV_FLAG_BLACKLISTED, &rbd_dev->flags);
 			need_to_wake = true;
 		} else {