Просмотр исходного кода

rcu: Avoid having just-onlined CPU resched itself when RCU is idle

CPUs set rdp->qs_pending when coming online to resolve races with
grace-period start.  However, this means that if RCU is idle, the
just-onlined CPU might needlessly send itself resched IPIs.  Adjust
the online-CPU initialization to avoid this, and also to correctly
cause the CPU to respond to the current grace period if needed.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Josh Boyer <jwboyer@redhat.com>
Tested-by: Christian Hoffmann <email@christianhoffmann.info>
Paul E. McKenney 14 лет назад
Родитель
Сommit
06ae115a1d
1 измененных файлов с 8 добавлено и 3 удалено
  1. 8 3
      kernel/rcutree.c

+ 8 - 3
kernel/rcutree.c

@@ -1911,8 +1911,6 @@ rcu_init_percpu_data(int cpu, struct rcu_state *rsp, int preemptible)
 
 	/* Set up local state, ensuring consistent view of global state. */
 	raw_spin_lock_irqsave(&rnp->lock, flags);
-	rdp->passed_quiesce = 0;  /* We could be racing with new GP, */
-	rdp->qs_pending = 1;	 /*  so set up to respond to current GP. */
 	rdp->beenonline = 1;	 /* We have now been online. */
 	rdp->preemptible = preemptible;
 	rdp->qlen_last_fqs_check = 0;
@@ -1937,8 +1935,15 @@ rcu_init_percpu_data(int cpu, struct rcu_state *rsp, int preemptible)
 		rnp->qsmaskinit |= mask;
 		mask = rnp->grpmask;
 		if (rnp == rdp->mynode) {
-			rdp->gpnum = rnp->completed; /* if GP in progress... */
+			/*
+			 * If there is a grace period in progress, we will
+			 * set up to wait for it next time we run the
+			 * RCU core code.
+			 */
+			rdp->gpnum = rnp->completed;
 			rdp->completed = rnp->completed;
+			rdp->passed_quiesce = 0;
+			rdp->qs_pending = 0;
 			rdp->passed_quiesce_gpnum = rnp->gpnum - 1;
 			trace_rcu_grace_period(rsp->name, rdp->gpnum, "cpuonl");
 		}