1 files changed, 105 insertions, 94 deletions

diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index e138922d51..750a802478 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -600,16 +600,13 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect
 	const sector_t this_sector = r1_bio->sector;
 	int sectors;
 	int best_good_sectors;
-	int best_disk, best_dist_disk, best_pending_disk;
-	int has_nonrot_disk;
+	int best_disk, best_dist_disk, best_pending_disk, sequential_disk;
 	int disk;
 	sector_t best_dist;
 	unsigned int min_pending;
 	struct md_rdev *rdev;
 	int choose_first;
-	int choose_next_idle;
 
-	rcu_read_lock();
 	/*
 	 * Check if we can balance. We can balance on the whole
 	 * device if no resync is going on, or below the resync window.
@@ -619,12 +616,11 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect
 	sectors = r1_bio->sectors;
 	best_disk = -1;
 	best_dist_disk = -1;
+	sequential_disk = -1;
 	best_dist = MaxSector;
 	best_pending_disk = -1;
 	min_pending = UINT_MAX;
 	best_good_sectors = 0;
-	has_nonrot_disk = 0;
-	choose_next_idle = 0;
 	clear_bit(R1BIO_FailFast, &r1_bio->state);
 
 	if ((conf->mddev->recovery_cp < this_sector + sectors) ||
@@ -640,9 +636,8 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect
 		sector_t first_bad;
 		int bad_sectors;
 		unsigned int pending;
-		bool nonrot;
 
-		rdev = rcu_dereference(conf->mirrors[disk].rdev);
+		rdev = conf->mirrors[disk].rdev;
 		if (r1_bio->bios[disk] == IO_BLOCKED
 		    || rdev == NULL
 		    || test_bit(Faulty, &rdev->flags))
@@ -706,8 +701,6 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect
 			/* At least two disks to choose from so failfast is OK */
 			set_bit(R1BIO_FailFast, &r1_bio->state);
 
-		nonrot = bdev_nonrot(rdev->bdev);
-		has_nonrot_disk |= nonrot;
 		pending = atomic_read(&rdev->nr_pending);
 		dist = abs(this_sector - conf->mirrors[disk].head_position);
 		if (choose_first) {
@@ -720,7 +713,6 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect
 			int opt_iosize = bdev_io_opt(rdev->bdev) >> 9;
 			struct raid1_info *mirror = &conf->mirrors[disk];
 
-			best_disk = disk;
 			/*
 			 * If buffered sequential IO size exceeds optimal
 			 * iosize, check if there is idle disk. If yes, choose
@@ -734,20 +726,27 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect
 			 * small, but not a big deal since when the second disk
 			 * starts IO, the first disk is likely still busy.
 			 */
-			if (nonrot && opt_iosize > 0 &&
+			if (test_bit(Nonrot, &rdev->flags) && opt_iosize > 0 &&
 			    mirror->seq_start != MaxSector &&
 			    mirror->next_seq_sect > opt_iosize &&
 			    mirror->next_seq_sect - opt_iosize >=
 			    mirror->seq_start) {
-				choose_next_idle = 1;
-				continue;
+				/*
+				 * Add 'pending' to avoid choosing this disk if
+				 * there is other idle disk.
+				 */
+				pending++;
+				/*
+				 * If there is no other idle disk, this disk
+				 * will be chosen.
+				 */
+				sequential_disk = disk;
+			} else {
+				best_disk = disk;
+				break;
 			}
-			break;
 		}
 
-		if (choose_next_idle)
-			continue;
-
 		if (min_pending > pending) {
 			min_pending = pending;
 			best_pending_disk = disk;
@@ -760,20 +759,27 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect
 	}
 
 	/*
+	 * sequential IO size exceeds optimal iosize, however, there is no other
+	 * idle disk, so choose the sequential disk.
+	 */
+	if (best_disk == -1 && min_pending != 0)
+		best_disk = sequential_disk;
+
+	/*
 	 * If all disks are rotational, choose the closest disk. If any disk is
 	 * non-rotational, choose the disk with less pending request even the
 	 * disk is rotational, which might/might not be optimal for raids with
 	 * mixed ratation/non-rotational disks depending on workload.
 	 */
 	if (best_disk == -1) {
-		if (has_nonrot_disk || min_pending == 0)
+		if (READ_ONCE(conf->nonrot_disks) || min_pending == 0)
 			best_disk = best_pending_disk;
 		else
 			best_disk = best_dist_disk;
 	}
 
 	if (best_disk >= 0) {
-		rdev = rcu_dereference(conf->mirrors[best_disk].rdev);
+		rdev = conf->mirrors[best_disk].rdev;
 		if (!rdev)
 			goto retry;
 		atomic_inc(&rdev->nr_pending);
@@ -784,7 +790,6 @@ static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sect
 
 		conf->mirrors[best_disk].next_seq_sect = this_sector + sectors;
 	}
-	rcu_read_unlock();
 	*max_sectors = sectors;
 
 	return best_disk;
@@ -1235,14 +1240,12 @@ static void raid1_read_request(struct mddev *mddev, struct bio *bio,
 
 	if (r1bio_existed) {
 		/* Need to get the block device name carefully */
-		struct md_rdev *rdev;
-		rcu_read_lock();
-		rdev = rcu_dereference(conf->mirrors[r1_bio->read_disk].rdev);
+		struct md_rdev *rdev = conf->mirrors[r1_bio->read_disk].rdev;
+
 		if (rdev)
 			snprintf(b, sizeof(b), "%pg", rdev->bdev);
 		else
 			strcpy(b, "???");
-		rcu_read_unlock();
 	}
 
 	/*
@@ -1396,10 +1399,9 @@ static void raid1_write_request(struct mddev *mddev, struct bio *bio,
 
 	disks = conf->raid_disks * 2;
 	blocked_rdev = NULL;
-	rcu_read_lock();
 	max_sectors = r1_bio->sectors;
 	for (i = 0;  i < disks; i++) {
-		struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev);
+		struct md_rdev *rdev = conf->mirrors[i].rdev;
 
 		/*
 		 * The write-behind io is only attempted on drives marked as
@@ -1465,7 +1467,6 @@ static void raid1_write_request(struct mddev *mddev, struct bio *bio,
 		}
 		r1_bio->bios[i] = bio;
 	}
-	rcu_read_unlock();
 
 	if (unlikely(blocked_rdev)) {
 		/* Wait for this device to become unblocked */
@@ -1617,15 +1618,16 @@ static void raid1_status(struct seq_file *seq, struct mddev *mddev)
 	struct r1conf *conf = mddev->private;
 	int i;
 
+	lockdep_assert_held(&mddev->lock);
+
 	seq_printf(seq, " [%d/%d] [", conf->raid_disks,
 		   conf->raid_disks - mddev->degraded);
-	rcu_read_lock();
 	for (i = 0; i < conf->raid_disks; i++) {
-		struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev);
+		struct md_rdev *rdev = READ_ONCE(conf->mirrors[i].rdev);
+
 		seq_printf(seq, "%s",
 			   rdev && test_bit(In_sync, &rdev->flags) ? "U" : "_");
 	}
-	rcu_read_unlock();
 	seq_printf(seq, "]");
 }
 
@@ -1691,16 +1693,15 @@ static void print_conf(struct r1conf *conf)
 	pr_debug(" --- wd:%d rd:%d\n", conf->raid_disks - conf->mddev->degraded,
 		 conf->raid_disks);
 
-	rcu_read_lock();
+	lockdep_assert_held(&conf->mddev->reconfig_mutex);
 	for (i = 0; i < conf->raid_disks; i++) {
-		struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev);
+		struct md_rdev *rdev = conf->mirrors[i].rdev;
 		if (rdev)
 			pr_debug(" disk %d, wo:%d, o:%d, dev:%pg\n",
 				 i, !test_bit(In_sync, &rdev->flags),
 				 !test_bit(Faulty, &rdev->flags),
 				 rdev->bdev);
 	}
-	rcu_read_unlock();
 }
 
 static void close_sync(struct r1conf *conf)
@@ -1767,6 +1768,52 @@ static int raid1_spare_active(struct mddev *mddev)
 	return count;
 }
 
+static bool raid1_add_conf(struct r1conf *conf, struct md_rdev *rdev, int disk,
+			   bool replacement)
+{
+	struct raid1_info *info = conf->mirrors + disk;
+
+	if (replacement)
+		info += conf->raid_disks;
+
+	if (info->rdev)
+		return false;
+
+	if (bdev_nonrot(rdev->bdev)) {
+		set_bit(Nonrot, &rdev->flags);
+		WRITE_ONCE(conf->nonrot_disks, conf->nonrot_disks + 1);
+	}
+
+	rdev->raid_disk = disk;
+	info->head_position = 0;
+	info->seq_start = MaxSector;
+	WRITE_ONCE(info->rdev, rdev);
+
+	return true;
+}
+
+static bool raid1_remove_conf(struct r1conf *conf, int disk)
+{
+	struct raid1_info *info = conf->mirrors + disk;
+	struct md_rdev *rdev = info->rdev;
+
+	if (!rdev || test_bit(In_sync, &rdev->flags) ||
+	    atomic_read(&rdev->nr_pending))
+		return false;
+
+	/* Only remove non-faulty devices if recovery is not possible. */
+	if (!test_bit(Faulty, &rdev->flags) &&
+	    rdev->mddev->recovery_disabled != conf->recovery_disabled &&
+	    rdev->mddev->degraded < conf->raid_disks)
+		return false;
+
+	if (test_and_clear_bit(Nonrot, &rdev->flags))
+		WRITE_ONCE(conf->nonrot_disks, conf->nonrot_disks - 1);
+
+	WRITE_ONCE(info->rdev, NULL);
+	return true;
+}
+
 static int raid1_add_disk(struct mddev *mddev, struct md_rdev *rdev)
 {
 	struct r1conf *conf = mddev->private;
@@ -1802,15 +1849,13 @@ static int raid1_add_disk(struct mddev *mddev, struct md_rdev *rdev)
 				disk_stack_limits(mddev->gendisk, rdev->bdev,
 						  rdev->data_offset << 9);
 
-			p->head_position = 0;
-			rdev->raid_disk = mirror;
+			raid1_add_conf(conf, rdev, mirror, false);
 			err = 0;
 			/* As all devices are equivalent, we don't need a full recovery
 			 * if this was recently any drive of the array
 			 */
 			if (rdev->saved_raid_disk < 0)
 				conf->fullsync = 1;
-			rcu_assign_pointer(p->rdev, rdev);
 			break;
 		}
 		if (test_bit(WantReplacement, &p->rdev->flags) &&
@@ -1820,13 +1865,11 @@ static int raid1_add_disk(struct mddev *mddev, struct md_rdev *rdev)
 
 	if (err && repl_slot >= 0) {
 		/* Add this device as a replacement */
-		p = conf->mirrors + repl_slot;
 		clear_bit(In_sync, &rdev->flags);
 		set_bit(Replacement, &rdev->flags);
-		rdev->raid_disk = repl_slot;
+		raid1_add_conf(conf, rdev, repl_slot, true);
 		err = 0;
 		conf->fullsync = 1;
-		rcu_assign_pointer(p[conf->raid_disks].rdev, rdev);
 	}
 
 	print_conf(conf);
@@ -1843,36 +1886,20 @@ static int raid1_remove_disk(struct mddev *mddev, struct md_rdev *rdev)
 	if (unlikely(number >= conf->raid_disks))
 		goto abort;
 
-	if (rdev != p->rdev)
-		p = conf->mirrors + conf->raid_disks + number;
+	if (rdev != p->rdev) {
+		number += conf->raid_disks;
+		p = conf->mirrors + number;
+	}
 
 	print_conf(conf);
 	if (rdev == p->rdev) {
-		if (test_bit(In_sync, &rdev->flags) ||
-		    atomic_read(&rdev->nr_pending)) {
+		if (!raid1_remove_conf(conf, number)) {
 			err = -EBUSY;
 			goto abort;
 		}
-		/* Only remove non-faulty devices if recovery
-		 * is not possible.
-		 */
-		if (!test_bit(Faulty, &rdev->flags) &&
-		    mddev->recovery_disabled != conf->recovery_disabled &&
-		    mddev->degraded < conf->raid_disks) {
-			err = -EBUSY;
-			goto abort;
-		}
-		p->rdev = NULL;
-		if (!test_bit(RemoveSynchronized, &rdev->flags)) {
-			synchronize_rcu();
-			if (atomic_read(&rdev->nr_pending)) {
-				/* lost the race, try later */
-				err = -EBUSY;
-				p->rdev = rdev;
-				goto abort;
-			}
-		}
-		if (conf->mirrors[conf->raid_disks + number].rdev) {
+
+		if (number < conf->raid_disks &&
+		    conf->mirrors[conf->raid_disks + number].rdev) {
 			/* We just removed a device that is being replaced.
 			 * Move down the replacement.  We drain all IO before
 			 * doing this to avoid confusion.
@@ -1892,7 +1919,7 @@ static int raid1_remove_disk(struct mddev *mddev, struct md_rdev *rdev)
 				goto abort;
 			}
 			clear_bit(Replacement, &repl->flags);
-			p->rdev = repl;
+			WRITE_ONCE(p->rdev, repl);
 			conf->mirrors[conf->raid_disks + number].rdev = NULL;
 			unfreeze_array(conf);
 		}
@@ -2290,8 +2317,7 @@ static void fix_read_error(struct r1conf *conf, int read_disk,
 			sector_t first_bad;
 			int bad_sectors;
 
-			rcu_read_lock();
-			rdev = rcu_dereference(conf->mirrors[d].rdev);
+			rdev = conf->mirrors[d].rdev;
 			if (rdev &&
 			    (test_bit(In_sync, &rdev->flags) ||
 			     (!test_bit(Faulty, &rdev->flags) &&
@@ -2299,15 +2325,14 @@ static void fix_read_error(struct r1conf *conf, int read_disk,
 			    is_badblock(rdev, sect, s,
 					&first_bad, &bad_sectors) == 0) {
 				atomic_inc(&rdev->nr_pending);
-				rcu_read_unlock();
 				if (sync_page_io(rdev, sect, s<<9,
 					 conf->tmppage, REQ_OP_READ, false))
 					success = 1;
 				rdev_dec_pending(rdev, mddev);
 				if (success)
 					break;
-			} else
-				rcu_read_unlock();
+			}
+
 			d++;
 			if (d == conf->raid_disks * 2)
 				d = 0;
@@ -2326,29 +2351,24 @@ static void fix_read_error(struct r1conf *conf, int read_disk,
 			if (d==0)
 				d = conf->raid_disks * 2;
 			d--;
-			rcu_read_lock();
-			rdev = rcu_dereference(conf->mirrors[d].rdev);
+			rdev = conf->mirrors[d].rdev;
 			if (rdev &&
 			    !test_bit(Faulty, &rdev->flags)) {
 				atomic_inc(&rdev->nr_pending);
-				rcu_read_unlock();
 				r1_sync_page_io(rdev, sect, s,
 						conf->tmppage, REQ_OP_WRITE);
 				rdev_dec_pending(rdev, mddev);
-			} else
-				rcu_read_unlock();
+			}
 		}
 		d = start;
 		while (d != read_disk) {
 			if (d==0)
 				d = conf->raid_disks * 2;
 			d--;
-			rcu_read_lock();
-			rdev = rcu_dereference(conf->mirrors[d].rdev);
+			rdev = conf->mirrors[d].rdev;
 			if (rdev &&
 			    !test_bit(Faulty, &rdev->flags)) {
 				atomic_inc(&rdev->nr_pending);
-				rcu_read_unlock();
 				if (r1_sync_page_io(rdev, sect, s,
 						conf->tmppage, REQ_OP_READ)) {
 					atomic_add(s, &rdev->corrected_errors);
@@ -2359,8 +2379,7 @@ static void fix_read_error(struct r1conf *conf, int read_disk,
 						rdev->bdev);
 				}
 				rdev_dec_pending(rdev, mddev);
-			} else
-				rcu_read_unlock();
+			}
 		}
 		sectors -= s;
 		sect += s;
@@ -2741,7 +2760,6 @@ static sector_t raid1_sync_request(struct mddev *mddev, sector_t sector_nr,
 
 	r1_bio = raid1_alloc_init_r1buf(conf);
 
-	rcu_read_lock();
 	/*
 	 * If we get a correctably read error during resync or recovery,
 	 * we might want to read from a different device.  So we
@@ -2762,7 +2780,7 @@ static sector_t raid1_sync_request(struct mddev *mddev, sector_t sector_nr,
 		struct md_rdev *rdev;
 		bio = r1_bio->bios[i];
 
-		rdev = rcu_dereference(conf->mirrors[i].rdev);
+		rdev = conf->mirrors[i].rdev;
 		if (rdev == NULL ||
 		    test_bit(Faulty, &rdev->flags)) {
 			if (i < conf->raid_disks)
@@ -2820,7 +2838,6 @@ static sector_t raid1_sync_request(struct mddev *mddev, sector_t sector_nr,
 				bio->bi_opf |= MD_FAILFAST;
 		}
 	}
-	rcu_read_unlock();
 	if (disk < 0)
 		disk = wonly;
 	r1_bio->read_disk = disk;
@@ -3025,23 +3042,17 @@ static struct r1conf *setup_conf(struct mddev *mddev)
 
 	err = -EINVAL;
 	spin_lock_init(&conf->device_lock);
+	conf->raid_disks = mddev->raid_disks;
 	rdev_for_each(rdev, mddev) {
 		int disk_idx = rdev->raid_disk;
-		if (disk_idx >= mddev->raid_disks
-		    || disk_idx < 0)
+
+		if (disk_idx >= conf->raid_disks || disk_idx < 0)
 			continue;
-		if (test_bit(Replacement, &rdev->flags))
-			disk = conf->mirrors + mddev->raid_disks + disk_idx;
-		else
-			disk = conf->mirrors + disk_idx;
 
-		if (disk->rdev)
+		if (!raid1_add_conf(conf, rdev, disk_idx,
+				    test_bit(Replacement, &rdev->flags)))
 			goto abort;
-		disk->rdev = rdev;
-		disk->head_position = 0;
-		disk->seq_start = MaxSector;
 	}
-	conf->raid_disks = mddev->raid_disks;
 	conf->mddev = mddev;
 	INIT_LIST_HEAD(&conf->retry_list);
 	INIT_LIST_HEAD(&conf->bio_end_io_list);