Merge branch 'md-next-raid10-optimize' into md-next

This patchset tries to avoid that two locks are held unconditionally
in hot path.

Test environment:

Architecture:
aarch64 Huawei KUNPENG 920
x86 Intel(R) Xeon(R) Platinum 8380

Raid10 initialize:
mdadm --create /dev/md0 --level 10 --bitmap none --raid-devices 4 \
    /dev/nvme0n1 /dev/nvme1n1 /dev/nvme2n1 /dev/nvme3n1

Test cmd:
(task set -c 0-15) fio -name=0 -ioengine=libaio -direct=1 -\
    group_reporting=1 -randseed=2022 -rwmixread=70 -refill_buffers \
    -filename=/dev/md0 -numjobs=16 -runtime=60s -bs=4k -iodepth=256 \
    -rw=randread

Test result:

aarch64:
before this patchset:           3.2 GiB/s
bind node before this patchset: 6.9 Gib/s
after this patchset:            7.9 Gib/s
bind node after this patchset:  8.0 Gib/s

x86:(bind node is not tested yet)
before this patchset: 7.0 GiB/s
after this patchset : 9.3 GiB/s

Please noted that in the test machine, memory access latency is very bad
across nodes compare to local node in aarch64, which is why bandwidth
while bind node is much better.

2 files changed, 96 insertions, 55 deletions

diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 077c7cdefcd4..58c711912875 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -79,6 +79,21 @@ static void end_reshape(struct r10conf *conf);
 
 #include "raid1-10.c"
 
+#define NULL_CMD
+#define cmd_before(conf, cmd) \
+	do { \
+		write_sequnlock_irq(&(conf)->resync_lock); \
+		cmd; \
+	} while (0)
+#define cmd_after(conf) write_seqlock_irq(&(conf)->resync_lock)
+
+#define wait_event_barrier_cmd(conf, cond, cmd) \
+	wait_event_cmd((conf)->wait_barrier, cond, cmd_before(conf, cmd), \
+		       cmd_after(conf))
+
+#define wait_event_barrier(conf, cond) \
+	wait_event_barrier_cmd(conf, cond, NULL_CMD)
+
 /*
  * for resync bio, r10bio pointer can be retrieved from the per-bio
  * 'struct resync_pages'.
@@ -274,6 +289,12 @@ static void put_buf(struct r10bio *r10_bio)
 	lower_barrier(conf);
 }
 
+static void wake_up_barrier(struct r10conf *conf)
+{
+	if (wq_has_sleeper(&conf->wait_barrier))
+		wake_up(&conf->wait_barrier);
+}
+
 static void reschedule_retry(struct r10bio *r10_bio)
 {
 	unsigned long flags;
@@ -930,78 +951,101 @@ static void flush_pending_writes(struct r10conf *conf)
 
 static void raise_barrier(struct r10conf *conf, int force)
 {
+	write_seqlock_irq(&conf->resync_lock);
 	BUG_ON(force && !conf->barrier);
-	spin_lock_irq(&conf->resync_lock);
 
 	/* Wait until no block IO is waiting (unless 'force') */
-	wait_event_lock_irq(conf->wait_barrier, force || !conf->nr_waiting,
-			    conf->resync_lock);
+	wait_event_barrier(conf, force || !conf->nr_waiting);
 
 	/* block any new IO from starting */
-	conf->barrier++;
+	WRITE_ONCE(conf->barrier, conf->barrier + 1);
 
 	/* Now wait for all pending IO to complete */
-	wait_event_lock_irq(conf->wait_barrier,
-			    !atomic_read(&conf->nr_pending) && conf->barrier < RESYNC_DEPTH,
-			    conf->resync_lock);
+	wait_event_barrier(conf, !atomic_read(&conf->nr_pending) &&
+				 conf->barrier < RESYNC_DEPTH);
 
-	spin_unlock_irq(&conf->resync_lock);
+	write_sequnlock_irq(&conf->resync_lock);
 }
 
 static void lower_barrier(struct r10conf *conf)
 {
 	unsigned long flags;
-	spin_lock_irqsave(&conf->resync_lock, flags);
-	conf->barrier--;
-	spin_unlock_irqrestore(&conf->resync_lock, flags);
+
+	write_seqlock_irqsave(&conf->resync_lock, flags);
+	WRITE_ONCE(conf->barrier, conf->barrier - 1);
+	write_sequnlock_irqrestore(&conf->resync_lock, flags);
 	wake_up(&conf->wait_barrier);
 }
 
+static bool stop_waiting_barrier(struct r10conf *conf)
+{
+	struct bio_list *bio_list = current->bio_list;
+
+	/* barrier is dropped */
+	if (!conf->barrier)
+		return true;
+
+	/*
+	 * If there are already pending requests (preventing the barrier from
+	 * rising completely), and the pre-process bio queue isn't empty, then
+	 * don't wait, as we need to empty that queue to get the nr_pending
+	 * count down.
+	 */
+	if (atomic_read(&conf->nr_pending) && bio_list &&
+	    (!bio_list_empty(&bio_list[0]) || !bio_list_empty(&bio_list[1])))
+		return true;
+
+	/* move on if recovery thread is blocked by us */
+	if (conf->mddev->thread->tsk == current &&
+	    test_bit(MD_RECOVERY_RUNNING, &conf->mddev->recovery) &&
+	    conf->nr_queued > 0)
+		return true;
+
+	return false;
+}
+
+static bool wait_barrier_nolock(struct r10conf *conf)
+{
+	unsigned int seq = read_seqbegin(&conf->resync_lock);
+
+	if (READ_ONCE(conf->barrier))
+		return false;
+
+	atomic_inc(&conf->nr_pending);
+	if (!read_seqretry(&conf->resync_lock, seq))
+		return true;
+
+	if (atomic_dec_and_test(&conf->nr_pending))
+		wake_up_barrier(conf);
+
+	return false;
+}
+
 static bool wait_barrier(struct r10conf *conf, bool nowait)
 {
 	bool ret = true;
 
-	spin_lock_irq(&conf->resync_lock);
+	if (wait_barrier_nolock(conf))
+		return true;
+
+	write_seqlock_irq(&conf->resync_lock);
 	if (conf->barrier) {
-		struct bio_list *bio_list = current->bio_list;
-		conf->nr_waiting++;
-		/* Wait for the barrier to drop.
-		 * However if there are already pending
-		 * requests (preventing the barrier from
-		 * rising completely), and the
-		 * pre-process bio queue isn't empty,
-		 * then don't wait, as we need to empty
-		 * that queue to get the nr_pending
-		 * count down.
-		 */
 		/* Return false when nowait flag is set */
 		if (nowait) {
 			ret = false;
 		} else {
+			conf->nr_waiting++;
 			raid10_log(conf->mddev, "wait barrier");
-			wait_event_lock_irq(conf->wait_barrier,
-					    !conf->barrier ||
-					    (atomic_read(&conf->nr_pending) &&
-					     bio_list &&
-					     (!bio_list_empty(&bio_list[0]) ||
-					      !bio_list_empty(&bio_list[1]))) ||
-					     /* move on if recovery thread is
-					      * blocked by us
-					      */
-					     (conf->mddev->thread->tsk == current &&
-					      test_bit(MD_RECOVERY_RUNNING,
-						       &conf->mddev->recovery) &&
-					      conf->nr_queued > 0),
-					    conf->resync_lock);
+			wait_event_barrier(conf, stop_waiting_barrier(conf));
+			conf->nr_waiting--;
 		}
-		conf->nr_waiting--;
 		if (!conf->nr_waiting)
 			wake_up(&conf->wait_barrier);
 	}
 	/* Only increment nr_pending when we wait */
 	if (ret)
 		atomic_inc(&conf->nr_pending);
-	spin_unlock_irq(&conf->resync_lock);
+	write_sequnlock_irq(&conf->resync_lock);
 	return ret;
 }
 
@@ -1009,7 +1053,7 @@ static void allow_barrier(struct r10conf *conf)
 {
 	if ((atomic_dec_and_test(&conf->nr_pending)) ||
 			(conf->array_freeze_pending))
-		wake_up(&conf->wait_barrier);
+		wake_up_barrier(conf);
 }
 
 static void freeze_array(struct r10conf *conf, int extra)
@@ -1026,27 +1070,24 @@ static void freeze_array(struct r10conf *conf, int extra)
 	 * must match the number of pending IOs (nr_pending) before
 	 * we continue.
 	 */
-	spin_lock_irq(&conf->resync_lock);
+	write_seqlock_irq(&conf->resync_lock);
 	conf->array_freeze_pending++;
-	conf->barrier++;
+	WRITE_ONCE(conf->barrier, conf->barrier + 1);
 	conf->nr_waiting++;
-	wait_event_lock_irq_cmd(conf->wait_barrier,
-				atomic_read(&conf->nr_pending) == conf->nr_queued+extra,
-				conf->resync_lock,
-				flush_pending_writes(conf));
-
+	wait_event_barrier_cmd(conf, atomic_read(&conf->nr_pending) ==
+			conf->nr_queued + extra, flush_pending_writes(conf));
 	conf->array_freeze_pending--;
-	spin_unlock_irq(&conf->resync_lock);
+	write_sequnlock_irq(&conf->resync_lock);
 }
 
 static void unfreeze_array(struct r10conf *conf)
 {
 	/* reverse the effect of the freeze */
-	spin_lock_irq(&conf->resync_lock);
-	conf->barrier--;
+	write_seqlock_irq(&conf->resync_lock);
+	WRITE_ONCE(conf->barrier, conf->barrier - 1);
 	conf->nr_waiting--;
 	wake_up(&conf->wait_barrier);
-	spin_unlock_irq(&conf->resync_lock);
+	write_sequnlock_irq(&conf->resync_lock);
 }
 
 static sector_t choose_data_offset(struct r10bio *r10_bio,
@@ -1885,7 +1926,7 @@ static bool raid10_make_request(struct mddev *mddev, struct bio *bio)
 	__make_request(mddev, bio, sectors);
 
 	/* In case raid10d snuck in to freeze_array */
-	wake_up(&conf->wait_barrier);
+	wake_up_barrier(conf);
 	return true;
 }
 
@@ -4033,7 +4074,7 @@ static struct r10conf *setup_conf(struct mddev *mddev)
 	INIT_LIST_HEAD(&conf->retry_list);
 	INIT_LIST_HEAD(&conf->bio_end_io_list);
 
-	spin_lock_init(&conf->resync_lock);
+	seqlock_init(&conf->resync_lock);
 	init_waitqueue_head(&conf->wait_barrier);
 	atomic_set(&conf->nr_pending, 0);
 
@@ -4352,7 +4393,7 @@ static void *raid10_takeover_raid0(struct mddev *mddev, sector_t size, int devs)
 				rdev->new_raid_disk = rdev->raid_disk * 2;
 				rdev->sectors = size;
 			}
-		conf->barrier = 1;
+		WRITE_ONCE(conf->barrier, 1);
 	}
 
 	return conf;
diff --git a/drivers/md/raid10.h b/drivers/md/raid10.h
index 5c0804d8bb1f..8c072ce0bc54 100644
--- a/drivers/md/raid10.h
+++ b/drivers/md/raid10.h
@@ -76,7 +76,7 @@ struct r10conf {
 	/* queue pending writes and submit them on unplug */
 	struct bio_list		pending_bio_list;
 
-	spinlock_t		resync_lock;
+	seqlock_t		resync_lock;
 	atomic_t		nr_pending;
 	int			nr_waiting;
 	int			nr_queued;