SUNRPC: Replace pool stats with per-CPU variables

Eliminate the use of bus-locked operations in svc_xprt_enqueue(),
which is a hot path. Replace them with per-cpu variables to reduce
cross-CPU memory bus traffic.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

include/linux/sunrpc/svc.h

@@ -21,14 +21,6 @@
 #include <linux/mm.h>
 #include <linux/pagevec.h>
 
-/* statistics for svc_pool structures */
-struct svc_pool_stats {
-	atomic_long_t	packets;
-	unsigned long	sockets_queued;
-	atomic_long_t	threads_woken;
-	atomic_long_t	threads_timedout;
-};
-
 /*
  *
  * RPC service thread pool.
@@ -45,7 +37,12 @@ struct svc_pool {
 	struct list_head	sp_sockets;	/* pending sockets */
 	unsigned int		sp_nrthreads;	/* # of threads in pool */
 	struct list_head	sp_all_threads;	/* all server threads */
-	struct svc_pool_stats	sp_stats;	/* statistics on pool operation */
+
+	/* statistics on pool operation */
+	struct percpu_counter	sp_sockets_queued;
+	struct percpu_counter	sp_threads_woken;
+	struct percpu_counter	sp_threads_timedout;
+
 #define	SP_TASK_PENDING		(0)		/* still work to do even if no
 						 * xprt is queued. */
 #define SP_CONGESTED		(1)

net/sunrpc/svc.c

@@ -512,6 +512,10 @@ __svc_create(struct svc_program *prog, unsigned int bufsize, int npools,
 		INIT_LIST_HEAD(&pool->sp_sockets);
 		INIT_LIST_HEAD(&pool->sp_all_threads);
 		spin_lock_init(&pool->sp_lock);
+
+		percpu_counter_init(&pool->sp_sockets_queued, 0, GFP_KERNEL);
+		percpu_counter_init(&pool->sp_threads_woken, 0, GFP_KERNEL);
+		percpu_counter_init(&pool->sp_threads_timedout, 0, GFP_KERNEL);
 	}
 
 	return serv;
@@ -565,6 +569,7 @@ void
 svc_destroy(struct kref *ref)
 {
 	struct svc_serv *serv = container_of(ref, struct svc_serv, sv_refcnt);
+	unsigned int i;
 
 	dprintk("svc: svc_destroy(%s)\n", serv->sv_program->pg_name);
 	timer_shutdown_sync(&serv->sv_temptimer);
@@ -580,6 +585,13 @@ svc_destroy(struct kref *ref)
 
 	svc_pool_map_put(serv->sv_nrpools);
 
+	for (i = 0; i < serv->sv_nrpools; i++) {
+		struct svc_pool *pool = &serv->sv_pools[i];
+
+		percpu_counter_destroy(&pool->sp_sockets_queued);
+		percpu_counter_destroy(&pool->sp_threads_woken);
+		percpu_counter_destroy(&pool->sp_threads_timedout);
+	}
 	kfree(serv->sv_pools);
 	kfree(serv);
 }

net/sunrpc/svc_xprt.c

@@ -462,11 +462,9 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 
 	pool = svc_pool_for_cpu(xprt->xpt_server);
 
-	atomic_long_inc(&pool->sp_stats.packets);
-
+	percpu_counter_inc(&pool->sp_sockets_queued);
 	spin_lock_bh(&pool->sp_lock);
 	list_add_tail(&xprt->xpt_ready, &pool->sp_sockets);
-	pool->sp_stats.sockets_queued++;
 	spin_unlock_bh(&pool->sp_lock);
 
 	/* find a thread for this xprt */
@@ -474,7 +472,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	list_for_each_entry_rcu(rqstp, &pool->sp_all_threads, rq_all) {
 		if (test_and_set_bit(RQ_BUSY, &rqstp->rq_flags))
 			continue;
-		atomic_long_inc(&pool->sp_stats.threads_woken);
+		percpu_counter_inc(&pool->sp_threads_woken);
 		rqstp->rq_qtime = ktime_get();
 		wake_up_process(rqstp->rq_task);
 		goto out_unlock;
@@ -769,7 +767,7 @@ static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp, long timeout)
 		goto out_found;
 
 	if (!time_left)
-		atomic_long_inc(&pool->sp_stats.threads_timedout);
+		percpu_counter_inc(&pool->sp_threads_timedout);
 
 	if (signalled() || kthread_should_stop())
 		return ERR_PTR(-EINTR);
@@ -1440,12 +1438,12 @@ static int svc_pool_stats_show(struct seq_file *m, void *p)
 		return 0;
 	}
 
-	seq_printf(m, "%u %lu %lu %lu %lu\n",
+	seq_printf(m, "%u %llu %llu %llu %llu\n",
 		pool->sp_id,
-		(unsigned long)atomic_long_read(&pool->sp_stats.packets),
-		pool->sp_stats.sockets_queued,
-		(unsigned long)atomic_long_read(&pool->sp_stats.threads_woken),
-		(unsigned long)atomic_long_read(&pool->sp_stats.threads_timedout));
+		percpu_counter_sum_positive(&pool->sp_sockets_queued),
+		percpu_counter_sum_positive(&pool->sp_sockets_queued),
+		percpu_counter_sum_positive(&pool->sp_threads_woken),
+		percpu_counter_sum_positive(&pool->sp_threads_timedout));
 
 	return 0;
 }