rcu: Allow RCU grace-period cleanup to be preempted (c856bafa) · Commits · 戴 / test

RCU grace-period cleanup is currently carried out with interrupts disabled, which can result in excessive latency spikes on large systems (many hundreds or thousands of CPUs). This patch therefore makes the RCU grace-period cleanup be preemptible, including voluntary preemption points, which should eliminate those latency spikes. Similar spikes from forcing of quiescent states will be dealt with similarly by later patches. Updated to replace uses of spin_lock_irqsave() with spin_lock_irq(), as suggested by Peter Zijlstra. Reported-by:

Mike Galbraith <mgalbraith@suse.de> Reported-by:

Dimitri Sivanich <sivanich@sgi.com> Signed-off-by:

Paul E. McKenney <paulmck@linux.vnet.ibm.com> Reviewed-by:

Josh Triplett <josh@joshtriplett.org>

kernel/rcutree.c

+5 −6

Original line number	Diff line number	Diff line
		@@ -1151,7 +1151,7 @@ static int __noreturn rcu_gp_kthread(void *arg)
		* completed.
		*/
		if (*rdp->nxttail[RCU_WAIT_TAIL] == NULL) {
		raw_spin_unlock(&rnp->lock); /* irqs remain disabled. */
		raw_spin_unlock_irq(&rnp->lock);

		/*
		* Propagate new ->completed value to rcu_node
		@@ -1160,14 +1160,13 @@ static int __noreturn rcu_gp_kthread(void *arg)
		* to process their callbacks.
		*/
		rcu_for_each_node_breadth_first(rsp, rnp) {
		/* irqs already disabled. */
		raw_spin_lock(&rnp->lock);
		raw_spin_lock_irq(&rnp->lock);
		rnp->completed = rsp->gpnum;
		/* irqs remain disabled. */
		raw_spin_unlock(&rnp->lock);
		raw_spin_unlock_irq(&rnp->lock);
		cond_resched();
		}
		rnp = rcu_get_root(rsp);
		raw_spin_lock(&rnp->lock); /* irqs already disabled. */
		raw_spin_lock_irq(&rnp->lock);
		}

		rsp->completed = rsp->gpnum; /* Declare grace period done. */

Admin message