Merge tag 'md/4.3' of git://neil.brown.name/md

Pull md updates from Neil Brown:

 - an assortment of little fixes, several for minor races only likely to
   be hit during testing

 - further cluster-md-raid1 development, not ready for real use yet.

 - new RAID6 syndrome code for ARM NEON

 - fix a race where a write can return before failure of one device is
   properly recorded in metadata, so an immediate crash might result in
   that write being lost.

* tag 'md/4.3' of git://neil.brown.name/md: (33 commits)
  md/raid5: ensure device failure recorded before write request returns.
  md/raid5: use bio_list for the list of bios to return.
  md/raid10: ensure device failure recorded before write request returns.
  md/raid1: ensure device failure recorded before write request returns.
  md-cluster: remove inappropriate try_module_get from join()
  md: extend spinlock protection in register_md_cluster_operations
  md-cluster: Read the disk bitmap sb and check if it needs recovery
  md-cluster: only call complete(&cinfo->completion) when node join cluster
  md-cluster: add missed lockres_free
  md-cluster: remove the unused sb_lock
  md-cluster: init suspend_list and suspend_lock early in join
  md-cluster: add the error check if failed to get dlm lock
  md-cluster: init completion within lockres_init
  md-cluster: fix deadlock issue on message lock
  md-cluster: transfer the resync ownership to another node
  md-cluster: split recover_slot for future code reuse
  md-cluster: use %pU to print UUIDs
  md: setup safemode_timer before it's being used
  md/raid5: handle possible race as reshape completes.
  md: sync sync_completed has correct value as recovery finishes.
  ...

Documentation/md-cluster.txt

@@ -91,7 +91,7 @@ The algorithm is:
     this message inappropriate or redundant.
 
  3. sender write LVB.
-    sender down-convert MESSAGE from EX to CR
+    sender down-convert MESSAGE from EX to CW
     sender try to get EX of ACK
     [ wait until all receiver has *processed* the MESSAGE ]
 
@@ -112,7 +112,7 @@ The algorithm is:
     sender down-convert ACK from EX to CR
     sender release MESSAGE
     sender release TOKEN
-                               receiver upconvert to EX of MESSAGE
+                               receiver upconvert to PR of MESSAGE
                                receiver get CR of ACK
                                receiver release MESSAGE
 

drivers/md/md.c

@@ -483,6 +483,8 @@ static void mddev_put(struct mddev *mddev)
 		bioset_free(bs);
 }
 
+static void md_safemode_timeout(unsigned long data);
+
 void mddev_init(struct mddev *mddev)
 {
 	mutex_init(&mddev->open_mutex);
@@ -490,7 +492,8 @@ void mddev_init(struct mddev *mddev)
 	mutex_init(&mddev->bitmap_info.mutex);
 	INIT_LIST_HEAD(&mddev->disks);
 	INIT_LIST_HEAD(&mddev->all_mddevs);
-	init_timer(&mddev->safemode_timer);
+	setup_timer(&mddev->safemode_timer, md_safemode_timeout,
+		    (unsigned long) mddev);
 	atomic_set(&mddev->active, 1);
 	atomic_set(&mddev->openers, 0);
 	atomic_set(&mddev->active_io, 0);
@@ -3255,8 +3258,6 @@ int strict_strtoul_scaled(const char *cp, unsigned long *res, int scale)
 	return 0;
 }
 
-static void md_safemode_timeout(unsigned long data);
-
 static ssize_t
 safe_delay_show(struct mddev *mddev, char *page)
 {
@@ -4189,6 +4190,8 @@ action_show(struct mddev *mddev, char *page)
 				type = "repair";
 		} else if (test_bit(MD_RECOVERY_RECOVER, &recovery))
 			type = "recover";
+		else if (mddev->reshape_position != MaxSector)
+			type = "reshape";
 	}
 	return sprintf(page, "%s\n", type);
 }
@@ -5180,8 +5183,6 @@ int md_run(struct mddev *mddev)
 	atomic_set(&mddev->max_corr_read_errors,
 		   MD_DEFAULT_MAX_CORRECTED_READ_ERRORS);
 	mddev->safemode = 0;
-	mddev->safemode_timer.function = md_safemode_timeout;
-	mddev->safemode_timer.data = (unsigned long) mddev;
 	mddev->safemode_delay = (200 * HZ)/1000 +1; /* 200 msec delay */
 	mddev->in_sync = 1;
 	smp_wmb();
@@ -5194,6 +5195,11 @@ int md_run(struct mddev *mddev)
 			if (sysfs_link_rdev(mddev, rdev))
 				/* failure here is OK */;
 
+	if (mddev->degraded && !mddev->ro)
+		/* This ensures that recovering status is reported immediately
+		 * via sysfs - until a lack of spares is confirmed.
+		 */
+		set_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
 	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 
 	if (mddev->flags & MD_UPDATE_SB_FLAGS)
@@ -5741,16 +5747,16 @@ static int get_bitmap_file(struct mddev *mddev, void __user * arg)
 
 	err = 0;
 	spin_lock(&mddev->lock);
-	/* bitmap disabled, zero the first byte and copy out */
-	if (!mddev->bitmap_info.file)
-		file->pathname[0] = '\0';
-	else if ((ptr = file_path(mddev->bitmap_info.file,
-			       file->pathname, sizeof(file->pathname))),
-		 IS_ERR(ptr))
-		err = PTR_ERR(ptr);
-	else
-		memmove(file->pathname, ptr,
-			sizeof(file->pathname)-(ptr-file->pathname));
+	/* bitmap enabled */
+	if (mddev->bitmap_info.file) {
+		ptr = file_path(mddev->bitmap_info.file, file->pathname,
+				sizeof(file->pathname));
+		if (IS_ERR(ptr))
+			err = PTR_ERR(ptr);
+		else
+			memmove(file->pathname, ptr,
+				sizeof(file->pathname)-(ptr-file->pathname));
+	}
 	spin_unlock(&mddev->lock);
 
 	if (err == 0 &&
@@ -7069,7 +7075,7 @@ static void status_unused(struct seq_file *seq)
 	seq_printf(seq, "\n");
 }
 
-static void status_resync(struct seq_file *seq, struct mddev *mddev)
+static int status_resync(struct seq_file *seq, struct mddev *mddev)
 {
 	sector_t max_sectors, resync, res;
 	unsigned long dt, db;
@@ -7077,18 +7083,32 @@ static void status_resync(struct seq_file *seq, struct mddev *mddev)
 	int scale;
 	unsigned int per_milli;
 
-	if (mddev->curr_resync <= 3)
-		resync = 0;
-	else
-		resync = mddev->curr_resync
-			- atomic_read(&mddev->recovery_active);
-
 	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) ||
 	    test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
 		max_sectors = mddev->resync_max_sectors;
 	else
 		max_sectors = mddev->dev_sectors;
 
+	resync = mddev->curr_resync;
+	if (resync <= 3) {
+		if (test_bit(MD_RECOVERY_DONE, &mddev->recovery))
+			/* Still cleaning up */
+			resync = max_sectors;
+	} else
+		resync -= atomic_read(&mddev->recovery_active);
+
+	if (resync == 0) {
+		if (mddev->recovery_cp < MaxSector) {
+			seq_printf(seq, "\tresync=PENDING");
+			return 1;
+		}
+		return 0;
+	}
+	if (resync < 3) {
+		seq_printf(seq, "\tresync=DELAYED");
+		return 1;
+	}
+
 	WARN_ON(max_sectors == 0);
 	/* Pick 'scale' such that (resync>>scale)*1000 will fit
 	 * in a sector_t, and (max_sectors>>scale) will fit in a
@@ -7153,6 +7173,7 @@ static void status_resync(struct seq_file *seq, struct mddev *mddev)
 		   ((unsigned long)rt % 60)/6);
 
 	seq_printf(seq, " speed=%ldK/sec", db/2/dt);
+	return 1;
 }
 
 static void *md_seq_start(struct seq_file *seq, loff_t *pos)
@@ -7298,13 +7319,8 @@ static int md_seq_show(struct seq_file *seq, void *v)
 			mddev->pers->status(seq, mddev);
 			seq_printf(seq, "\n      ");
 			if (mddev->pers->sync_request) {
-				if (mddev->curr_resync > 2) {
-					status_resync(seq, mddev);
+				if (status_resync(seq, mddev))
 					seq_printf(seq, "\n      ");
-				} else if (mddev->curr_resync >= 1)
-					seq_printf(seq, "\tresync=DELAYED\n      ");
-				else if (mddev->recovery_cp < MaxSector)
-					seq_printf(seq, "\tresync=PENDING\n      ");
 			}
 		} else
 			seq_printf(seq, "\n       ");
@@ -7387,15 +7403,19 @@ int unregister_md_personality(struct md_personality *p)
 }
 EXPORT_SYMBOL(unregister_md_personality);
 
-int register_md_cluster_operations(struct md_cluster_operations *ops, struct module *module)
+int register_md_cluster_operations(struct md_cluster_operations *ops,
+				   struct module *module)
 {
-	if (md_cluster_ops != NULL)
-		return -EALREADY;
+	int ret = 0;
 	spin_lock(&pers_lock);
-	md_cluster_ops = ops;
-	md_cluster_mod = module;
+	if (md_cluster_ops != NULL)
+		ret = -EALREADY;
+	else {
+		md_cluster_ops = ops;
+		md_cluster_mod = module;
+	}
 	spin_unlock(&pers_lock);
-	return 0;
+	return ret;
 }
 EXPORT_SYMBOL(register_md_cluster_operations);
 
@@ -7793,7 +7813,8 @@ void md_do_sync(struct md_thread *thread)
 		      > (max_sectors >> 4)) ||
 		     time_after_eq(jiffies, update_time + UPDATE_FREQUENCY) ||
 		     (j - mddev->curr_resync_completed)*2
-		     >= mddev->resync_max - mddev->curr_resync_completed
+		     >= mddev->resync_max - mddev->curr_resync_completed ||
+		     mddev->curr_resync_completed > mddev->resync_max
 			    )) {
 			/* time to update curr_resync_completed */
 			wait_event(mddev->recovery_wait,
@@ -7838,6 +7859,9 @@ void md_do_sync(struct md_thread *thread)
 			break;
 
 		j += sectors;
+		if (j > max_sectors)
+			/* when skipping, extra large numbers can be returned. */
+			j = max_sectors;
 		if (j > 2)
 			mddev->curr_resync = j;
 		if (mddev_is_clustered(mddev))
@@ -7906,12 +7930,15 @@ void md_do_sync(struct md_thread *thread)
 	blk_finish_plug(&plug);
 	wait_event(mddev->recovery_wait, !atomic_read(&mddev->recovery_active));
 
+	if (!test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery) &&
+	    !test_bit(MD_RECOVERY_INTR, &mddev->recovery) &&
+	    mddev->curr_resync > 2) {
+		mddev->curr_resync_completed = mddev->curr_resync;
+		sysfs_notify(&mddev->kobj, NULL, "sync_completed");
+	}
 	/* tell personality that we are finished */
 	mddev->pers->sync_request(mddev, max_sectors, &skipped);
 
-	if (mddev_is_clustered(mddev))
-		md_cluster_ops->resync_finish(mddev);
-
 	if (!test_bit(MD_RECOVERY_CHECK, &mddev->recovery) &&
 	    mddev->curr_resync > 2) {
 		if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
@@ -7945,6 +7972,9 @@ void md_do_sync(struct md_thread *thread)
 		}
 	}
  skip:
+	if (mddev_is_clustered(mddev))
+		md_cluster_ops->resync_finish(mddev);
+
 	set_bit(MD_CHANGE_DEVS, &mddev->flags);
 
 	spin_lock(&mddev->lock);
@@ -7955,11 +7985,11 @@ void md_do_sync(struct md_thread *thread)
 		mddev->resync_max = MaxSector;
 	} else if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery))
 		mddev->resync_min = mddev->curr_resync_completed;
+	set_bit(MD_RECOVERY_DONE, &mddev->recovery);
 	mddev->curr_resync = 0;
 	spin_unlock(&mddev->lock);
 
 	wake_up(&resync_wait);
-	set_bit(MD_RECOVERY_DONE, &mddev->recovery);
 	md_wakeup_thread(mddev->thread);
 	return;
 }
@@ -8128,6 +8158,7 @@ void md_check_recovery(struct mddev *mddev)
 			 */
 			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
 			md_reap_sync_thread(mddev);
+			clear_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
 			clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 			goto unlock;
 		}
@@ -8574,6 +8605,7 @@ int rdev_set_badblocks(struct md_rdev *rdev, sector_t s, int sectors,
 		/* Make sure they get written out promptly */
 		sysfs_notify_dirent_safe(rdev->sysfs_state);
 		set_bit(MD_CHANGE_CLEAN, &rdev->mddev->flags);
+		set_bit(MD_CHANGE_PENDING, &rdev->mddev->flags);
 		md_wakeup_thread(rdev->mddev->thread);
 	}
 	return rv;

drivers/md/raid0.c

@@ -83,7 +83,7 @@ static int create_strip_zones(struct mddev *mddev, struct r0conf **private_conf)
 	char b[BDEVNAME_SIZE];
 	char b2[BDEVNAME_SIZE];
 	struct r0conf *conf = kzalloc(sizeof(*conf), GFP_KERNEL);
-	bool discard_supported = false;
+	unsigned short blksize = 512;
 
 	if (!conf)
 		return -ENOMEM;
@@ -98,6 +98,9 @@ static int create_strip_zones(struct mddev *mddev, struct r0conf **private_conf)
 		sector_div(sectors, mddev->chunk_sectors);
 		rdev1->sectors = sectors * mddev->chunk_sectors;
 
+		blksize = max(blksize, queue_logical_block_size(
+				      rdev1->bdev->bd_disk->queue));
+
 		rdev_for_each(rdev2, mddev) {
 			pr_debug("md/raid0:%s:   comparing %s(%llu)"
 				 " with %s(%llu)\n",
@@ -134,6 +137,18 @@ static int create_strip_zones(struct mddev *mddev, struct r0conf **private_conf)
 	}
 	pr_debug("md/raid0:%s: FINAL %d zones\n",
 		 mdname(mddev), conf->nr_strip_zones);
+	/*
+	 * now since we have the hard sector sizes, we can make sure
+	 * chunk size is a multiple of that sector size
+	 */
+	if ((mddev->chunk_sectors << 9) % blksize) {
+		printk(KERN_ERR "md/raid0:%s: chunk_size of %d not multiple of block size %d\n",
+		       mdname(mddev),
+		       mddev->chunk_sectors << 9, blksize);
+		err = -EINVAL;
+		goto abort;
+	}
+
 	err = -ENOMEM;
 	conf->strip_zone = kzalloc(sizeof(struct strip_zone)*
 				conf->nr_strip_zones, GFP_KERNEL);
@@ -188,16 +203,9 @@ static int create_strip_zones(struct mddev *mddev, struct r0conf **private_conf)
 		}
 		dev[j] = rdev1;
 
-		if (mddev->queue)
-			disk_stack_limits(mddev->gendisk, rdev1->bdev,
-					  rdev1->data_offset << 9);
-
 		if (!smallest || (rdev1->sectors < smallest->sectors))
 			smallest = rdev1;
 		cnt++;
-
-		if (blk_queue_discard(bdev_get_queue(rdev1->bdev)))
-			discard_supported = true;
 	}
 	if (cnt != mddev->raid_disks) {
 		printk(KERN_ERR "md/raid0:%s: too few disks (%d of %d) - "
@@ -258,28 +266,6 @@ static int create_strip_zones(struct mddev *mddev, struct r0conf **private_conf)
 			 (unsigned long long)smallest->sectors);
 	}
 
-	/*
-	 * now since we have the hard sector sizes, we can make sure
-	 * chunk size is a multiple of that sector size
-	 */
-	if ((mddev->chunk_sectors << 9) % queue_logical_block_size(mddev->queue)) {
-		printk(KERN_ERR "md/raid0:%s: chunk_size of %d not valid\n",
-		       mdname(mddev),
-		       mddev->chunk_sectors << 9);
-		goto abort;
-	}
-
-	if (mddev->queue) {
-		blk_queue_io_min(mddev->queue, mddev->chunk_sectors << 9);
-		blk_queue_io_opt(mddev->queue,
-				 (mddev->chunk_sectors << 9) * mddev->raid_disks);
-
-		if (!discard_supported)
-			queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, mddev->queue);
-		else
-			queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, mddev->queue);
-	}
-
 	pr_debug("md/raid0:%s: done.\n", mdname(mddev));
 	*private_conf = conf;
 
@@ -378,12 +364,6 @@ static int raid0_run(struct mddev *mddev)
 	if (md_check_no_bitmap(mddev))
 		return -EINVAL;
 
-	if (mddev->queue) {
-		blk_queue_max_hw_sectors(mddev->queue, mddev->chunk_sectors);
-		blk_queue_max_write_same_sectors(mddev->queue, mddev->chunk_sectors);
-		blk_queue_max_discard_sectors(mddev->queue, mddev->chunk_sectors);
-	}
-
 	/* if private is not null, we are here after takeover */
 	if (mddev->private == NULL) {
 		ret = create_strip_zones(mddev, &conf);
@@ -392,6 +372,29 @@ static int raid0_run(struct mddev *mddev)
 		mddev->private = conf;
 	}
 	conf = mddev->private;
+	if (mddev->queue) {
+		struct md_rdev *rdev;
+		bool discard_supported = false;
+
+		rdev_for_each(rdev, mddev) {
+			disk_stack_limits(mddev->gendisk, rdev->bdev,
+					  rdev->data_offset << 9);
+			if (blk_queue_discard(bdev_get_queue(rdev->bdev)))
+				discard_supported = true;
+		}
+		blk_queue_max_hw_sectors(mddev->queue, mddev->chunk_sectors);
+		blk_queue_max_write_same_sectors(mddev->queue, mddev->chunk_sectors);
+		blk_queue_max_discard_sectors(mddev->queue, mddev->chunk_sectors);
+
+		blk_queue_io_min(mddev->queue, mddev->chunk_sectors << 9);
+		blk_queue_io_opt(mddev->queue,
+				 (mddev->chunk_sectors << 9) * mddev->raid_disks);
+
+		if (!discard_supported)
+			queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, mddev->queue);
+		else
+			queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, mddev->queue);
+	}
 
 	/* calculate array device size */
 	md_set_array_sectors(mddev, raid0_size(mddev, 0, 0));

drivers/md/raid1.c

@@ -1474,6 +1474,7 @@ static void error(struct mddev *mddev, struct md_rdev *rdev)
 	 */
 	set_bit(MD_RECOVERY_INTR, &mddev->recovery);
 	set_bit(MD_CHANGE_DEVS, &mddev->flags);
+	set_bit(MD_CHANGE_PENDING, &mddev->flags);
 	printk(KERN_ALERT
 	       "md/raid1:%s: Disk failure on %s, disabling device.\n"
 	       "md/raid1:%s: Operation continuing on %d devices.\n",
@@ -2235,6 +2236,7 @@ static void handle_sync_write_finished(struct r1conf *conf, struct r1bio *r1_bio
 static void handle_write_finished(struct r1conf *conf, struct r1bio *r1_bio)
 {
 	int m;
+	bool fail = false;
 	for (m = 0; m < conf->raid_disks * 2 ; m++)
 		if (r1_bio->bios[m] == IO_MADE_GOOD) {
 			struct md_rdev *rdev = conf->mirrors[m].rdev;
@@ -2247,6 +2249,7 @@ static void handle_write_finished(struct r1conf *conf, struct r1bio *r1_bio)
 			 * narrow down and record precise write
 			 * errors.
 			 */
+			fail = true;
 			if (!narrow_write_error(r1_bio, m)) {
 				md_error(conf->mddev,
 					 conf->mirrors[m].rdev);
@@ -2258,7 +2261,13 @@ static void handle_write_finished(struct r1conf *conf, struct r1bio *r1_bio)
 		}
 	if (test_bit(R1BIO_WriteError, &r1_bio->state))
 		close_write(r1_bio);
-	raid_end_bio_io(r1_bio);
+	if (fail) {
+		spin_lock_irq(&conf->device_lock);
+		list_add(&r1_bio->retry_list, &conf->bio_end_io_list);
+		spin_unlock_irq(&conf->device_lock);
+		md_wakeup_thread(conf->mddev->thread);
+	} else
+		raid_end_bio_io(r1_bio);
 }
 
 static void handle_read_error(struct r1conf *conf, struct r1bio *r1_bio)
@@ -2364,6 +2373,23 @@ static void raid1d(struct md_thread *thread)
 
 	md_check_recovery(mddev);
 
+	if (!list_empty_careful(&conf->bio_end_io_list) &&
+	    !test_bit(MD_CHANGE_PENDING, &mddev->flags)) {
+		LIST_HEAD(tmp);
+		spin_lock_irqsave(&conf->device_lock, flags);
+		if (!test_bit(MD_CHANGE_PENDING, &mddev->flags)) {
+			list_add(&tmp, &conf->bio_end_io_list);
+			list_del_init(&conf->bio_end_io_list);
+		}
+		spin_unlock_irqrestore(&conf->device_lock, flags);
+		while (!list_empty(&tmp)) {
+			r1_bio = list_first_entry(&conf->bio_end_io_list,
+						  struct r1bio, retry_list);
+			list_del(&r1_bio->retry_list);
+			raid_end_bio_io(r1_bio);
+		}
+	}
+
 	blk_start_plug(&plug);
 	for (;;) {
 
@@ -2763,6 +2789,7 @@ static struct r1conf *setup_conf(struct mddev *mddev)
 	conf->raid_disks = mddev->raid_disks;
 	conf->mddev = mddev;
 	INIT_LIST_HEAD(&conf->retry_list);
+	INIT_LIST_HEAD(&conf->bio_end_io_list);
 
 	spin_lock_init(&conf->resync_lock);
 	init_waitqueue_head(&conf->wait_barrier);
@@ -3057,6 +3084,7 @@ static int raid1_reshape(struct mddev *mddev)
 
 	unfreeze_array(conf);
 
+	set_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
 	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	md_wakeup_thread(mddev->thread);
 

drivers/md/raid1.h

@@ -61,6 +61,11 @@ struct r1conf {
 	 * block, or anything else.
 	 */
 	struct list_head	retry_list;
+	/* A separate list of r1bio which just need raid_end_bio_io called.
+	 * This mustn't happen for writes which had any errors if the superblock
+	 * needs to be written.
+	 */
+	struct list_head	bio_end_io_list;
 
 	/* queue pending writes to be submitted on unplug */
 	struct bio_list		pending_bio_list;

drivers/md/raid10.c

@@ -1589,6 +1589,7 @@ static void error(struct mddev *mddev, struct md_rdev *rdev)
 	set_bit(Blocked, &rdev->flags);
 	set_bit(Faulty, &rdev->flags);
 	set_bit(MD_CHANGE_DEVS, &mddev->flags);
+	set_bit(MD_CHANGE_PENDING, &mddev->flags);
 	spin_unlock_irqrestore(&conf->device_lock, flags);
 	printk(KERN_ALERT
 	       "md/raid10:%s: Disk failure on %s, disabling device.\n"
@@ -2623,6 +2624,7 @@ static void handle_write_completed(struct r10conf *conf, struct r10bio *r10_bio)
 		}
 		put_buf(r10_bio);
 	} else {
+		bool fail = false;
 		for (m = 0; m < conf->copies; m++) {
 			int dev = r10_bio->devs[m].devnum;
 			struct bio *bio = r10_bio->devs[m].bio;
@@ -2634,6 +2636,7 @@ static void handle_write_completed(struct r10conf *conf, struct r10bio *r10_bio)
 					r10_bio->sectors, 0);
 				rdev_dec_pending(rdev, conf->mddev);
 			} else if (bio != NULL && bio->bi_error) {
+				fail = true;
 				if (!narrow_write_error(r10_bio, m)) {
 					md_error(conf->mddev, rdev);
 					set_bit(R10BIO_Degraded,
@@ -2654,7 +2657,13 @@ static void handle_write_completed(struct r10conf *conf, struct r10bio *r10_bio)
 		if (test_bit(R10BIO_WriteError,
 			     &r10_bio->state))
 			close_write(r10_bio);
-		raid_end_bio_io(r10_bio);
+		if (fail) {
+			spin_lock_irq(&conf->device_lock);
+			list_add(&r10_bio->retry_list, &conf->bio_end_io_list);
+			spin_unlock_irq(&conf->device_lock);
+			md_wakeup_thread(conf->mddev->thread);
+		} else
+			raid_end_bio_io(r10_bio);
 	}
 }
 
@@ -2669,6 +2678,23 @@ static void raid10d(struct md_thread *thread)
 
 	md_check_recovery(mddev);
 
+	if (!list_empty_careful(&conf->bio_end_io_list) &&
+	    !test_bit(MD_CHANGE_PENDING, &mddev->flags)) {
+		LIST_HEAD(tmp);
+		spin_lock_irqsave(&conf->device_lock, flags);
+		if (!test_bit(MD_CHANGE_PENDING, &mddev->flags)) {
+			list_add(&tmp, &conf->bio_end_io_list);
+			list_del_init(&conf->bio_end_io_list);
+		}
+		spin_unlock_irqrestore(&conf->device_lock, flags);
+		while (!list_empty(&tmp)) {
+			r10_bio = list_first_entry(&conf->bio_end_io_list,
+						  struct r10bio, retry_list);
+			list_del(&r10_bio->retry_list);
+			raid_end_bio_io(r10_bio);
+		}
+	}
+
 	blk_start_plug(&plug);
 	for (;;) {
 
@@ -3443,6 +3469,7 @@ static struct r10conf *setup_conf(struct mddev *mddev)
 	conf->reshape_safe = conf->reshape_progress;
 	spin_lock_init(&conf->device_lock);
 	INIT_LIST_HEAD(&conf->retry_list);
+	INIT_LIST_HEAD(&conf->bio_end_io_list);
 
 	spin_lock_init(&conf->resync_lock);
 	init_waitqueue_head(&conf->wait_barrier);
@@ -4097,7 +4124,7 @@ static sector_t reshape_request(struct mddev *mddev, sector_t sector_nr,
 	 * at a time, possibly less if that exceeds RESYNC_PAGES,
 	 * or we hit a bad block or something.
 	 * This might mean we pause for normal IO in the middle of
-	 * a chunk, but that is not a problem was mddev->reshape_position
+	 * a chunk, but that is not a problem as mddev->reshape_position
 	 * can record any location.
 	 *
 	 * If we will want to write to a location that isn't
@@ -4121,7 +4148,7 @@ static sector_t reshape_request(struct mddev *mddev, sector_t sector_nr,
 	 *
 	 * In all this the minimum difference in data offsets
 	 * (conf->offset_diff - always positive) allows a bit of slack,
-	 * so next can be after 'safe', but not by more than offset_disk
+	 * so next can be after 'safe', but not by more than offset_diff
 	 *
 	 * We need to prepare all the bios here before we start any IO
 	 * to ensure the size we choose is acceptable to all devices.

drivers/md/raid10.h

@@ -53,6 +53,12 @@ struct r10conf {
 	sector_t		offset_diff;
 
 	struct list_head	retry_list;
+	/* A separate list of r1bio which just need raid_end_bio_io called.
+	 * This mustn't happen for writes which had any errors if the superblock
+	 * needs to be written.
+	 */
+	struct list_head	bio_end_io_list;
+
 	/* queue pending writes and submit them on unplug */
 	struct bio_list		pending_bio_list;
 	int			pending_count;

drivers/md/raid5.h

@@ -265,7 +265,7 @@ struct stripe_head_state {
 	int dec_preread_active;
 	unsigned long ops_request;
 
-	struct bio *return_bi;
+	struct bio_list return_bi;
 	struct md_rdev *blocked_rdev;
 	int handle_bad_blocks;
 };
@@ -476,6 +476,9 @@ struct r5conf {
 	int			skip_copy; /* Don't copy data from bio to stripe cache */
 	struct list_head	*last_hold; /* detect hold_list promotions */
 
+	/* bios to have bi_end_io called after metadata is synced */
+	struct bio_list		return_bi;
+
 	atomic_t		reshape_stripes; /* stripes with pending writes for reshape */
 	/* unfortunately we need two cache names as we temporarily have
 	 * two caches.

lib/raid6/neon.c

@@ -40,9 +40,20 @@
 					(unsigned long)bytes, ptrs);	\
 		kernel_neon_end();					\
 	}								\
+	static void raid6_neon ## _n ## _xor_syndrome(int disks,	\
+					int start, int stop, 		\
+					size_t bytes, void **ptrs)	\
+	{								\
+		void raid6_neon ## _n  ## _xor_syndrome_real(int,	\
+				int, int, unsigned long, void**);	\
+		kernel_neon_begin();					\
+		raid6_neon ## _n ## _xor_syndrome_real(disks,		\
+			start, stop, (unsigned long)bytes, ptrs);	\
+		kernel_neon_end();					\
+	}								\
 	struct raid6_calls const raid6_neonx ## _n = {			\
 		raid6_neon ## _n ## _gen_syndrome,			\
-		NULL,		/* XOR not yet implemented */		\
+		raid6_neon ## _n ## _xor_syndrome,			\
 		raid6_have_neon,					\
 		"neonx" #_n,						\
 		0							\

lib/raid6/neon.uc

@@ -3,6 +3,7 @@
  *   neon.uc - RAID-6 syndrome calculation using ARM NEON instructions
  *
  *   Copyright (C) 2012 Rob Herring
+ *   Copyright (C) 2015 Linaro Ltd. <ard.biesheuvel@linaro.org>
  *
  *   Based on altivec.uc:
  *     Copyright 2002-2004 H. Peter Anvin - All Rights Reserved
@@ -78,3 +79,48 @@ void raid6_neon$#_gen_syndrome_real(int disks, unsigned long bytes, void **ptrs)
 		vst1q_u8(&q[d+NSIZE*$$], wq$$);
 	}
 }
+
+void raid6_neon$#_xor_syndrome_real(int disks, int start, int stop,
+				    unsigned long bytes, void **ptrs)
+{
+	uint8_t **dptr = (uint8_t **)ptrs;
+	uint8_t *p, *q;
+	int d, z, z0;
+
+	register unative_t wd$$, wq$$, wp$$, w1$$, w2$$;
+	const unative_t x1d = NBYTES(0x1d);
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	for ( d = 0 ; d < bytes ; d += NSIZE*$# ) {
+		wq$$ = vld1q_u8(&dptr[z0][d+$$*NSIZE]);
+		wp$$ = veorq_u8(vld1q_u8(&p[d+$$*NSIZE]), wq$$);
+
+		/* P/Q data pages */
+		for ( z = z0-1 ; z >= start ; z-- ) {
+			wd$$ = vld1q_u8(&dptr[z][d+$$*NSIZE]);
+			wp$$ = veorq_u8(wp$$, wd$$);
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+
+			w2$$ = vandq_u8(w2$$, x1d);
+			w1$$ = veorq_u8(w1$$, w2$$);
+			wq$$ = veorq_u8(w1$$, wd$$);
+		}
+		/* P/Q left side optimization */
+		for ( z = start-1 ; z >= 0 ; z-- ) {
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+
+			w2$$ = vandq_u8(w2$$, x1d);
+			wq$$ = veorq_u8(w1$$, w2$$);
+		}
+		w1$$ = vld1q_u8(&q[d+NSIZE*$$]);
+		wq$$ = veorq_u8(wq$$, w1$$);
+
+		vst1q_u8(&p[d+NSIZE*$$], wp$$);
+		vst1q_u8(&q[d+NSIZE*$$], wq$$);
+	}
+}