Only wakeup waiters if we've actually done work

AKAMAI: zfs: CR 3695072
Reviewed-by: Tim Chase <tim@chase2k.com>
Reviewed-by: Richard Yao <ryao@gentoo.org>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Debabrata Banerjee <dbanerje@akamai.com>
Issue #6035

module/zfs/arc.c

@@ -4434,16 +4434,16 @@ __arc_shrinker_func(struct shrinker *shrink, struct shrink_control *sc)
 #else
 		pages = btop(arc_evictable_memory());
 #endif
+		/*
+		 * We've shrunk what we can, wake up threads.
+		 */
+		cv_broadcast(&arc_reclaim_waiters_cv);
+
 	} else {
 		arc_kmem_reap_now();
 		pages = SHRINK_STOP;
 	}
 
-	/*
-	 * We've reaped what we can, wake up threads.
-	 */
-	cv_broadcast(&arc_reclaim_waiters_cv);
-
 	/*
 	 * When direct reclaim is observed it usually indicates a rapid
 	 * increase in memory pressure.  This occurs because the kswapd