rcu: Make rcu_gp_cleanup() more accurately predict need for new GP

Currently, rcu_gp_cleanup() scans the rcu_node tree in order to reset
state to reflect the end of the grace period.  It also checks to see
whether a new grace period is needed, but in a number of cases, rather
than directly cause the new grace period to be immediately started, it
instead leaves the grace-period-needed state where various fail-safes
can find it.  This works fine, but results in higher contention on the
root rcu_node structure's ->lock, which is undesirable, and contention
on that lock has recently become noticeable.

This commit therefore makes rcu_gp_cleanup() immediately start a new
grace period if there is any need for one.

It is quite possible that it will later be necessary to throttle the
grace-period rate, but that can be dealt with when and if.

Reported-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Nicholas Piggin <npiggin@gmail.com>

kernel/rcu/tree.c

@@ -1763,14 +1763,14 @@ out:
  * Clean up any old requests for the just-ended grace period.  Also return
  * whether any additional grace periods have been requested.
  */
-static int rcu_future_gp_cleanup(struct rcu_state *rsp, struct rcu_node *rnp)
+static bool rcu_future_gp_cleanup(struct rcu_state *rsp, struct rcu_node *rnp)
 {
 	int c = rnp->completed;
-	int needmore;
+	bool needmore;
 	struct rcu_data *rdp = this_cpu_ptr(rsp->rda);
 
 	need_future_gp_element(rnp, c) = 0;
-	needmore = need_future_gp_element(rnp, c + 1);
+	needmore = need_any_future_gp(rnp);
 	trace_rcu_future_gp(rnp, rdp, c,
 			    needmore ? TPS("CleanupMore") : TPS("Cleanup"));
 	return needmore;
@@ -2113,7 +2113,6 @@ static void rcu_gp_cleanup(struct rcu_state *rsp)
 {
 	unsigned long gp_duration;
 	bool needgp = false;
-	int nocb = 0;
 	struct rcu_data *rdp;
 	struct rcu_node *rnp = rcu_get_root(rsp);
 	struct swait_queue_head *sq;
@@ -2152,7 +2151,7 @@ static void rcu_gp_cleanup(struct rcu_state *rsp)
 		if (rnp == rdp->mynode)
 			needgp = __note_gp_changes(rsp, rnp, rdp) || needgp;
 		/* smp_mb() provided by prior unlock-lock pair. */
-		nocb += rcu_future_gp_cleanup(rsp, rnp);
+		needgp = rcu_future_gp_cleanup(rsp, rnp) || needgp;
 		sq = rcu_nocb_gp_get(rnp);
 		raw_spin_unlock_irq_rcu_node(rnp);
 		rcu_nocb_gp_cleanup(sq);
@@ -2162,13 +2161,18 @@ static void rcu_gp_cleanup(struct rcu_state *rsp)
 	}
 	rnp = rcu_get_root(rsp);
 	raw_spin_lock_irq_rcu_node(rnp); /* Order GP before ->completed update. */
-	rcu_nocb_gp_set(rnp, nocb);
 
 	/* Declare grace period done. */
 	WRITE_ONCE(rsp->completed, rsp->gpnum);
 	trace_rcu_grace_period(rsp->name, rsp->completed, TPS("end"));
 	rsp->gp_state = RCU_GP_IDLE;
+	/* Check for GP requests since above loop. */
 	rdp = this_cpu_ptr(rsp->rda);
+	if (need_any_future_gp(rnp)) {
+		trace_rcu_future_gp(rnp, rdp, rsp->completed - 1,
+				    TPS("CleanupMore"));
+		needgp = true;
+	}
 	/* Advance CBs to reduce false positives below. */
 	needgp = rcu_advance_cbs(rsp, rnp, rdp) || needgp;
 	if (needgp || cpu_needs_another_gp(rsp, rdp)) {

kernel/rcu/tree.h

@@ -469,7 +469,6 @@ static void print_cpu_stall_info_end(void);
 static void zero_cpu_stall_ticks(struct rcu_data *rdp);
 static void increment_cpu_stall_ticks(void);
 static bool rcu_nocb_cpu_needs_barrier(struct rcu_state *rsp, int cpu);
-static void rcu_nocb_gp_set(struct rcu_node *rnp, int nrq);
 static struct swait_queue_head *rcu_nocb_gp_get(struct rcu_node *rnp);
 static void rcu_nocb_gp_cleanup(struct swait_queue_head *sq);
 static void rcu_init_one_nocb(struct rcu_node *rnp);

kernel/rcu/tree_plugin.h

@@ -1780,19 +1780,6 @@ static void rcu_nocb_gp_cleanup(struct swait_queue_head *sq)
 	swake_up_all(sq);
 }
 
-/*
- * Set the root rcu_node structure's ->need_future_gp field
- * based on the sum of those of all rcu_node structures.  This does
- * double-count the root rcu_node structure's requests, but this
- * is necessary to handle the possibility of a rcu_nocb_kthread()
- * having awakened during the time that the rcu_node structures
- * were being updated for the end of the previous grace period.
- */
-static void rcu_nocb_gp_set(struct rcu_node *rnp, int nrq)
-{
-	need_future_gp_element(rnp, rnp->completed + 1) += nrq;
-}
-
 static struct swait_queue_head *rcu_nocb_gp_get(struct rcu_node *rnp)
 {
 	return &rnp->nocb_gp_wq[rnp->completed & 0x1];
@@ -2495,10 +2482,6 @@ static void rcu_nocb_gp_cleanup(struct swait_queue_head *sq)
 {
 }
 
-static void rcu_nocb_gp_set(struct rcu_node *rnp, int nrq)
-{
-}
-
 static struct swait_queue_head *rcu_nocb_gp_get(struct rcu_node *rnp)
 {
 	return NULL;