Merge branch 'for-linus' of git://neil.brown.name/md

* 'for-linus' of git://neil.brown.name/md:
  md: allow resync_start to be set while an array is active.
  md/raid10:  reformat some loops with less indenting.
  md/raid10: remove unused variable.
  md/raid10: make more use of 'slot' in raid10d.
  md/raid10: some tidying up in fix_read_error
  md/raid1: improve handling of pages allocated for write-behind.
  md/raid1: try fix_sync_read_error before process_checks.
  md/raid1: tidy up new functions: process_checks and fix_sync_read_error.
  md/raid1: split out two sub-functions from sync_request_write
  md: make error_handler functions more uniform and correct.
  md/multipath: discard ->working_disks in favour of ->degraded
  md/raid1: clean up read_balance.
  md: simplify raid10 read_balance
  md/bitmap: fix saving of events_cleared and other state.
  md: reject a re-add request that cannot be honoured.
  md: Fix race when creating a new md device.

drivers/md/bitmap.c

@@ -493,11 +493,11 @@ void bitmap_update_sb(struct bitmap *bitmap)
 	spin_unlock_irqrestore(&bitmap->lock, flags);
 	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
 	sb->events = cpu_to_le64(bitmap->mddev->events);
-	if (bitmap->mddev->events < bitmap->events_cleared) {
+	if (bitmap->mddev->events < bitmap->events_cleared)
 		/* rocking back to read-only */
 		bitmap->events_cleared = bitmap->mddev->events;
-		sb->events_cleared = cpu_to_le64(bitmap->events_cleared);
-	}
+	sb->events_cleared = cpu_to_le64(bitmap->events_cleared);
+	sb->state = cpu_to_le32(bitmap->flags);
 	/* Just in case these have been changed via sysfs: */
 	sb->daemon_sleep = cpu_to_le32(bitmap->mddev->bitmap_info.daemon_sleep/HZ);
 	sb->write_behind = cpu_to_le32(bitmap->mddev->bitmap_info.max_write_behind);
@@ -618,7 +618,7 @@ static int bitmap_read_sb(struct bitmap *bitmap)
 	if (le32_to_cpu(sb->version) == BITMAP_MAJOR_HOSTENDIAN)
 		bitmap->flags |= BITMAP_HOSTENDIAN;
 	bitmap->events_cleared = le64_to_cpu(sb->events_cleared);
-	if (sb->state & cpu_to_le32(BITMAP_STALE))
+	if (bitmap->flags & BITMAP_STALE)
 		bitmap->events_cleared = bitmap->mddev->events;
 	err = 0;
 out:
@@ -652,9 +652,11 @@ static int bitmap_mask_state(struct bitmap *bitmap, enum bitmap_state bits,
 	switch (op) {
 	case MASK_SET:
 		sb->state |= cpu_to_le32(bits);
+		bitmap->flags |= bits;
 		break;
 	case MASK_UNSET:
 		sb->state &= cpu_to_le32(~bits);
+		bitmap->flags &= ~bits;
 		break;
 	default:
 		BUG();

drivers/md/md.c

@@ -3324,7 +3324,7 @@ resync_start_store(mddev_t *mddev, const char *buf, size_t len)
 	char *e;
 	unsigned long long n = simple_strtoull(buf, &e, 10);
 
-	if (mddev->pers)
+	if (mddev->pers && !test_bit(MD_RECOVERY_FROZEN, &mddev->recovery))
 		return -EBUSY;
 	if (cmd_match(buf, "none"))
 		n = MaxSector;
@@ -4347,13 +4347,19 @@ static int md_alloc(dev_t dev, char *name)
 	disk->fops = &md_fops;
 	disk->private_data = mddev;
 	disk->queue = mddev->queue;
+	blk_queue_flush(mddev->queue, REQ_FLUSH | REQ_FUA);
 	/* Allow extended partitions.  This makes the
 	 * 'mdp' device redundant, but we can't really
 	 * remove it now.
 	 */
 	disk->flags |= GENHD_FL_EXT_DEVT;
-	add_disk(disk);
 	mddev->gendisk = disk;
+	/* As soon as we call add_disk(), another thread could get
+	 * through to md_open, so make sure it doesn't get too far
+	 */
+	mutex_lock(&mddev->open_mutex);
+	add_disk(disk);
+
 	error = kobject_init_and_add(&mddev->kobj, &md_ktype,
 				     &disk_to_dev(disk)->kobj, "%s", "md");
 	if (error) {
@@ -4367,8 +4373,7 @@ static int md_alloc(dev_t dev, char *name)
 	if (mddev->kobj.sd &&
 	    sysfs_create_group(&mddev->kobj, &md_bitmap_group))
 		printk(KERN_DEBUG "pointless warning\n");
-
-	blk_queue_flush(mddev->queue, REQ_FLUSH | REQ_FUA);
+	mutex_unlock(&mddev->open_mutex);
  abort:
 	mutex_unlock(&disks_mutex);
 	if (!error && mddev->kobj.sd) {
@@ -5211,6 +5216,16 @@ static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 		} else
 			super_types[mddev->major_version].
 				validate_super(mddev, rdev);
+		if ((info->state & (1<<MD_DISK_SYNC)) &&
+		    (!test_bit(In_sync, &rdev->flags) ||
+		     rdev->raid_disk != info->raid_disk)) {
+			/* This was a hot-add request, but events doesn't
+			 * match, so reject it.
+			 */
+			export_rdev(rdev);
+			return -EINVAL;
+		}
+
 		if (test_bit(In_sync, &rdev->flags))
 			rdev->saved_raid_disk = rdev->raid_disk;
 		else

drivers/md/multipath.c

@@ -146,7 +146,7 @@ static void multipath_status (struct seq_file *seq, mddev_t *mddev)
 	int i;
 	
 	seq_printf (seq, " [%d/%d] [", conf->raid_disks,
-						 conf->working_disks);
+		    conf->raid_disks - mddev->degraded);
 	for (i = 0; i < conf->raid_disks; i++)
 		seq_printf (seq, "%s",
 			       conf->multipaths[i].rdev && 
@@ -186,35 +186,36 @@ static int multipath_congested(void *data, int bits)
 static void multipath_error (mddev_t *mddev, mdk_rdev_t *rdev)
 {
 	multipath_conf_t *conf = mddev->private;
+	char b[BDEVNAME_SIZE];
 
-	if (conf->working_disks <= 1) {
+	if (conf->raid_disks - mddev->degraded <= 1) {
 		/*
 		 * Uh oh, we can do nothing if this is our last path, but
 		 * first check if this is a queued request for a device
 		 * which has just failed.
 		 */
 		printk(KERN_ALERT 
-			"multipath: only one IO path left and IO error.\n");
+		       "multipath: only one IO path left and IO error.\n");
 		/* leave it active... it's all we have */
-	} else {
-		/*
-		 * Mark disk as unusable
-		 */
-		if (!test_bit(Faulty, &rdev->flags)) {
-			char b[BDEVNAME_SIZE];
-			clear_bit(In_sync, &rdev->flags);
-			set_bit(Faulty, &rdev->flags);
-			set_bit(MD_CHANGE_DEVS, &mddev->flags);
-			conf->working_disks--;
-			mddev->degraded++;
-			printk(KERN_ALERT "multipath: IO failure on %s,"
-				" disabling IO path.\n"
-				"multipath: Operation continuing"
-				" on %d IO paths.\n",
-				bdevname (rdev->bdev,b),
-				conf->working_disks);
-		}
+		return;
+	}
+	/*
+	 * Mark disk as unusable
+	 */
+	if (test_and_clear_bit(In_sync, &rdev->flags)) {
+		unsigned long flags;
+		spin_lock_irqsave(&conf->device_lock, flags);
+		mddev->degraded++;
+		spin_unlock_irqrestore(&conf->device_lock, flags);
 	}
+	set_bit(Faulty, &rdev->flags);
+	set_bit(MD_CHANGE_DEVS, &mddev->flags);
+	printk(KERN_ALERT "multipath: IO failure on %s,"
+	       " disabling IO path.\n"
+	       "multipath: Operation continuing"
+	       " on %d IO paths.\n",
+	       bdevname(rdev->bdev, b),
+	       conf->raid_disks - mddev->degraded);
 }
 
 static void print_multipath_conf (multipath_conf_t *conf)
@@ -227,7 +228,7 @@ static void print_multipath_conf (multipath_conf_t *conf)
 		printk("(conf==NULL)\n");
 		return;
 	}
-	printk(" --- wd:%d rd:%d\n", conf->working_disks,
+	printk(" --- wd:%d rd:%d\n", conf->raid_disks - conf->mddev->degraded,
 			 conf->raid_disks);
 
 	for (i = 0; i < conf->raid_disks; i++) {
@@ -274,10 +275,11 @@ static int multipath_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 							   PAGE_CACHE_SIZE - 1);
 			}
 
-			conf->working_disks++;
+			spin_lock_irq(&conf->device_lock);
 			mddev->degraded--;
 			rdev->raid_disk = path;
 			set_bit(In_sync, &rdev->flags);
+			spin_unlock_irq(&conf->device_lock);
 			rcu_assign_pointer(p->rdev, rdev);
 			err = 0;
 			md_integrity_add_rdev(rdev, mddev);
@@ -391,6 +393,7 @@ static int multipath_run (mddev_t *mddev)
 	int disk_idx;
 	struct multipath_info *disk;
 	mdk_rdev_t *rdev;
+	int working_disks;
 
 	if (md_check_no_bitmap(mddev))
 		return -EINVAL;
@@ -424,7 +427,7 @@ static int multipath_run (mddev_t *mddev)
 		goto out_free_conf;
 	}
 
-	conf->working_disks = 0;
+	working_disks = 0;
 	list_for_each_entry(rdev, &mddev->disks, same_set) {
 		disk_idx = rdev->raid_disk;
 		if (disk_idx < 0 ||
@@ -446,7 +449,7 @@ static int multipath_run (mddev_t *mddev)
 		}
 
 		if (!test_bit(Faulty, &rdev->flags))
-			conf->working_disks++;
+			working_disks++;
 	}
 
 	conf->raid_disks = mddev->raid_disks;
@@ -454,12 +457,12 @@ static int multipath_run (mddev_t *mddev)
 	spin_lock_init(&conf->device_lock);
 	INIT_LIST_HEAD(&conf->retry_list);
 
-	if (!conf->working_disks) {
+	if (!working_disks) {
 		printk(KERN_ERR "multipath: no operational IO paths for %s\n",
 			mdname(mddev));
 		goto out_free_conf;
 	}
-	mddev->degraded = conf->raid_disks - conf->working_disks;
+	mddev->degraded = conf->raid_disks - working_disks;
 
 	conf->pool = mempool_create_kmalloc_pool(NR_RESERVED_BUFS,
 						 sizeof(struct multipath_bh));
@@ -481,7 +484,8 @@ static int multipath_run (mddev_t *mddev)
 
 	printk(KERN_INFO 
 		"multipath: array %s active with %d out of %d IO paths\n",
-		mdname(mddev), conf->working_disks, mddev->raid_disks);
+		mdname(mddev), conf->raid_disks - mddev->degraded,
+	       mddev->raid_disks);
 	/*
 	 * Ok, everything is just fine now
 	 */

drivers/md/multipath.h

@@ -9,7 +9,6 @@ struct multipath_private_data {
 	mddev_t			*mddev;
 	struct multipath_info	*multipaths;
 	int			raid_disks;
-	int			working_disks;
 	spinlock_t		device_lock;
 	struct list_head	retry_list;
 

drivers/md/raid1.c

@@ -297,23 +297,24 @@ static void raid1_end_read_request(struct bio *bio, int error)
 	rdev_dec_pending(conf->mirrors[mirror].rdev, conf->mddev);
 }
 
-static void r1_bio_write_done(r1bio_t *r1_bio, int vcnt, struct bio_vec *bv,
-			      int behind)
+static void r1_bio_write_done(r1bio_t *r1_bio)
 {
 	if (atomic_dec_and_test(&r1_bio->remaining))
 	{
 		/* it really is the end of this request */
 		if (test_bit(R1BIO_BehindIO, &r1_bio->state)) {
 			/* free extra copy of the data pages */
-			int i = vcnt;
+			int i = r1_bio->behind_page_count;
 			while (i--)
-				safe_put_page(bv[i].bv_page);
+				safe_put_page(r1_bio->behind_pages[i]);
+			kfree(r1_bio->behind_pages);
+			r1_bio->behind_pages = NULL;
 		}
 		/* clear the bitmap if all writes complete successfully */
 		bitmap_endwrite(r1_bio->mddev->bitmap, r1_bio->sector,
 				r1_bio->sectors,
 				!test_bit(R1BIO_Degraded, &r1_bio->state),
-				behind);
+				test_bit(R1BIO_BehindIO, &r1_bio->state));
 		md_write_end(r1_bio->mddev);
 		raid_end_bio_io(r1_bio);
 	}
@@ -386,7 +387,7 @@ static void raid1_end_write_request(struct bio *bio, int error)
 	 * Let's see if all mirrored write operations have finished
 	 * already.
 	 */
-	r1_bio_write_done(r1_bio, bio->bi_vcnt, bio->bi_io_vec, behind);
+	r1_bio_write_done(r1_bio);
 
 	if (to_put)
 		bio_put(to_put);
@@ -411,10 +412,10 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 {
 	const sector_t this_sector = r1_bio->sector;
 	const int sectors = r1_bio->sectors;
-	int new_disk = -1;
 	int start_disk;
+	int best_disk;
 	int i;
-	sector_t new_distance, current_distance;
+	sector_t best_dist;
 	mdk_rdev_t *rdev;
 	int choose_first;
 
@@ -425,6 +426,8 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 	 * We take the first readable disk when above the resync window.
 	 */
  retry:
+	best_disk = -1;
+	best_dist = MaxSector;
 	if (conf->mddev->recovery_cp < MaxSector &&
 	    (this_sector + sectors >= conf->next_resync)) {
 		choose_first = 1;
@@ -434,8 +437,8 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 		start_disk = conf->last_used;
 	}
 
-	/* make sure the disk is operational */
 	for (i = 0 ; i < conf->raid_disks ; i++) {
+		sector_t dist;
 		int disk = start_disk + i;
 		if (disk >= conf->raid_disks)
 			disk -= conf->raid_disks;
@@ -443,60 +446,43 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 		rdev = rcu_dereference(conf->mirrors[disk].rdev);
 		if (r1_bio->bios[disk] == IO_BLOCKED
 		    || rdev == NULL
-		    || !test_bit(In_sync, &rdev->flags))
+		    || test_bit(Faulty, &rdev->flags))
 			continue;
-
-		new_disk = disk;
-		if (!test_bit(WriteMostly, &rdev->flags))
-			break;
-	}
-
-	if (new_disk < 0 || choose_first)
-		goto rb_out;
-
-	/*
-	 * Don't change to another disk for sequential reads:
-	 */
-	if (conf->next_seq_sect == this_sector)
-		goto rb_out;
-	if (this_sector == conf->mirrors[new_disk].head_position)
-		goto rb_out;
-
-	current_distance = abs(this_sector 
-			       - conf->mirrors[new_disk].head_position);
-
-	/* look for a better disk - i.e. head is closer */
-	start_disk = new_disk;
-	for (i = 1; i < conf->raid_disks; i++) {
-		int disk = start_disk + 1;
-		if (disk >= conf->raid_disks)
-			disk -= conf->raid_disks;
-
-		rdev = rcu_dereference(conf->mirrors[disk].rdev);
-		if (r1_bio->bios[disk] == IO_BLOCKED
-		    || rdev == NULL
-		    || !test_bit(In_sync, &rdev->flags)
-		    || test_bit(WriteMostly, &rdev->flags))
+		if (!test_bit(In_sync, &rdev->flags) &&
+		    rdev->recovery_offset < this_sector + sectors)
 			continue;
-
-		if (!atomic_read(&rdev->nr_pending)) {
-			new_disk = disk;
+		if (test_bit(WriteMostly, &rdev->flags)) {
+			/* Don't balance among write-mostly, just
+			 * use the first as a last resort */
+			if (best_disk < 0)
+				best_disk = disk;
+			continue;
+		}
+		/* This is a reasonable device to use.  It might
+		 * even be best.
+		 */
+		dist = abs(this_sector - conf->mirrors[disk].head_position);
+		if (choose_first
+		    /* Don't change to another disk for sequential reads */
+		    || conf->next_seq_sect == this_sector
+		    || dist == 0
+		    /* If device is idle, use it */
+		    || atomic_read(&rdev->nr_pending) == 0) {
+			best_disk = disk;
 			break;
 		}
-		new_distance = abs(this_sector - conf->mirrors[disk].head_position);
-		if (new_distance < current_distance) {
-			current_distance = new_distance;
-			new_disk = disk;
+		if (dist < best_dist) {
+			best_dist = dist;
+			best_disk = disk;
 		}
 	}
 
- rb_out:
-	if (new_disk >= 0) {
-		rdev = rcu_dereference(conf->mirrors[new_disk].rdev);
+	if (best_disk >= 0) {
+		rdev = rcu_dereference(conf->mirrors[best_disk].rdev);
 		if (!rdev)
 			goto retry;
 		atomic_inc(&rdev->nr_pending);
-		if (!test_bit(In_sync, &rdev->flags)) {
+		if (test_bit(Faulty, &rdev->flags)) {
 			/* cannot risk returning a device that failed
 			 * before we inc'ed nr_pending
 			 */
@@ -504,11 +490,11 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 			goto retry;
 		}
 		conf->next_seq_sect = this_sector + sectors;
-		conf->last_used = new_disk;
+		conf->last_used = best_disk;
 	}
 	rcu_read_unlock();
 
-	return new_disk;
+	return best_disk;
 }
 
 static int raid1_congested(void *data, int bits)
@@ -675,37 +661,36 @@ static void unfreeze_array(conf_t *conf)
 
 
 /* duplicate the data pages for behind I/O 
- * We return a list of bio_vec rather than just page pointers
- * as it makes freeing easier
  */
-static struct bio_vec *alloc_behind_pages(struct bio *bio)
+static void alloc_behind_pages(struct bio *bio, r1bio_t *r1_bio)
 {
 	int i;
 	struct bio_vec *bvec;
-	struct bio_vec *pages = kzalloc(bio->bi_vcnt * sizeof(struct bio_vec),
+	struct page **pages = kzalloc(bio->bi_vcnt * sizeof(struct page*),
 					GFP_NOIO);
 	if (unlikely(!pages))
-		goto do_sync_io;
+		return;
 
 	bio_for_each_segment(bvec, bio, i) {
-		pages[i].bv_page = alloc_page(GFP_NOIO);
-		if (unlikely(!pages[i].bv_page))
+		pages[i] = alloc_page(GFP_NOIO);
+		if (unlikely(!pages[i]))
 			goto do_sync_io;
-		memcpy(kmap(pages[i].bv_page) + bvec->bv_offset,
+		memcpy(kmap(pages[i]) + bvec->bv_offset,
 			kmap(bvec->bv_page) + bvec->bv_offset, bvec->bv_len);
-		kunmap(pages[i].bv_page);
+		kunmap(pages[i]);
 		kunmap(bvec->bv_page);
 	}
-
-	return pages;
+	r1_bio->behind_pages = pages;
+	r1_bio->behind_page_count = bio->bi_vcnt;
+	set_bit(R1BIO_BehindIO, &r1_bio->state);
+	return;
 
 do_sync_io:
-	if (pages)
-		for (i = 0; i < bio->bi_vcnt && pages[i].bv_page; i++)
-			put_page(pages[i].bv_page);
+	for (i = 0; i < bio->bi_vcnt; i++)
+		if (pages[i])
+			put_page(pages[i]);
 	kfree(pages);
 	PRINTK("%dB behind alloc failed, doing sync I/O\n", bio->bi_size);
-	return NULL;
 }
 
 static int make_request(mddev_t *mddev, struct bio * bio)
@@ -717,7 +702,6 @@ static int make_request(mddev_t *mddev, struct bio * bio)
 	int i, targets = 0, disks;
 	struct bitmap *bitmap;
 	unsigned long flags;
-	struct bio_vec *behind_pages = NULL;
 	const int rw = bio_data_dir(bio);
 	const unsigned long do_sync = (bio->bi_rw & REQ_SYNC);
 	const unsigned long do_flush_fua = (bio->bi_rw & (REQ_FLUSH | REQ_FUA));
@@ -870,9 +854,8 @@ static int make_request(mddev_t *mddev, struct bio * bio)
 	if (bitmap &&
 	    (atomic_read(&bitmap->behind_writes)
 	     < mddev->bitmap_info.max_write_behind) &&
-	    !waitqueue_active(&bitmap->behind_wait) &&
-	    (behind_pages = alloc_behind_pages(bio)) != NULL)
-		set_bit(R1BIO_BehindIO, &r1_bio->state);
+	    !waitqueue_active(&bitmap->behind_wait))
+		alloc_behind_pages(bio, r1_bio);
 
 	atomic_set(&r1_bio->remaining, 1);
 	atomic_set(&r1_bio->behind_remaining, 0);
@@ -893,7 +876,7 @@ static int make_request(mddev_t *mddev, struct bio * bio)
 		mbio->bi_rw = WRITE | do_flush_fua | do_sync;
 		mbio->bi_private = r1_bio;
 
-		if (behind_pages) {
+		if (r1_bio->behind_pages) {
 			struct bio_vec *bvec;
 			int j;
 
@@ -905,7 +888,7 @@ static int make_request(mddev_t *mddev, struct bio * bio)
 			 * them all
 			 */
 			__bio_for_each_segment(bvec, mbio, j, 0)
-				bvec->bv_page = behind_pages[j].bv_page;
+				bvec->bv_page = r1_bio->behind_pages[j];
 			if (test_bit(WriteMostly, &conf->mirrors[i].rdev->flags))
 				atomic_inc(&r1_bio->behind_remaining);
 		}
@@ -915,8 +898,7 @@ static int make_request(mddev_t *mddev, struct bio * bio)
 		bio_list_add(&conf->pending_bio_list, mbio);
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 	}
-	r1_bio_write_done(r1_bio, bio->bi_vcnt, behind_pages, behind_pages != NULL);
-	kfree(behind_pages); /* the behind pages are attached to the bios now */
+	r1_bio_write_done(r1_bio);
 
 	/* In case raid1d snuck in to freeze_array */
 	wake_up(&conf->wait_barrier);
@@ -1196,194 +1178,210 @@ static void end_sync_write(struct bio *bio, int error)
 	}
 }
 
-static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio)
+static int fix_sync_read_error(r1bio_t *r1_bio)
 {
+	/* Try some synchronous reads of other devices to get
+	 * good data, much like with normal read errors.  Only
+	 * read into the pages we already have so we don't
+	 * need to re-issue the read request.
+	 * We don't need to freeze the array, because being in an
+	 * active sync request, there is no normal IO, and
+	 * no overlapping syncs.
+	 */
+	mddev_t *mddev = r1_bio->mddev;
 	conf_t *conf = mddev->private;
-	int i;
-	int disks = conf->raid_disks;
-	struct bio *bio, *wbio;
-
-	bio = r1_bio->bios[r1_bio->read_disk];
+	struct bio *bio = r1_bio->bios[r1_bio->read_disk];
+	sector_t sect = r1_bio->sector;
+	int sectors = r1_bio->sectors;
+	int idx = 0;
 
+	while(sectors) {
+		int s = sectors;
+		int d = r1_bio->read_disk;
+		int success = 0;
+		mdk_rdev_t *rdev;
+		int start;
 
-	if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
-		/* We have read all readable devices.  If we haven't
-		 * got the block, then there is no hope left.
-		 * If we have, then we want to do a comparison
-		 * and skip the write if everything is the same.
-		 * If any blocks failed to read, then we need to
-		 * attempt an over-write
-		 */
-		int primary;
-		if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) {
-			for (i=0; i<mddev->raid_disks; i++)
-				if (r1_bio->bios[i]->bi_end_io == end_sync_read)
-					md_error(mddev, conf->mirrors[i].rdev);
+		if (s > (PAGE_SIZE>>9))
+			s = PAGE_SIZE >> 9;
+		do {
+			if (r1_bio->bios[d]->bi_end_io == end_sync_read) {
+				/* No rcu protection needed here devices
+				 * can only be removed when no resync is
+				 * active, and resync is currently active
+				 */
+				rdev = conf->mirrors[d].rdev;
+				if (sync_page_io(rdev,
+						 sect,
+						 s<<9,
+						 bio->bi_io_vec[idx].bv_page,
+						 READ, false)) {
+					success = 1;
+					break;
+				}
+			}
+			d++;
+			if (d == conf->raid_disks)
+				d = 0;
+		} while (!success && d != r1_bio->read_disk);
 
-			md_done_sync(mddev, r1_bio->sectors, 1);
+		if (!success) {
+			char b[BDEVNAME_SIZE];
+			/* Cannot read from anywhere, array is toast */
+			md_error(mddev, conf->mirrors[r1_bio->read_disk].rdev);
+			printk(KERN_ALERT "md/raid1:%s: %s: unrecoverable I/O read error"
+			       " for block %llu\n",
+			       mdname(mddev),
+			       bdevname(bio->bi_bdev, b),
+			       (unsigned long long)r1_bio->sector);
+			md_done_sync(mddev, r1_bio->sectors, 0);
 			put_buf(r1_bio);
-			return;
+			return 0;
 		}
-		for (primary=0; primary<mddev->raid_disks; primary++)
-			if (r1_bio->bios[primary]->bi_end_io == end_sync_read &&
-			    test_bit(BIO_UPTODATE, &r1_bio->bios[primary]->bi_flags)) {
-				r1_bio->bios[primary]->bi_end_io = NULL;
-				rdev_dec_pending(conf->mirrors[primary].rdev, mddev);
-				break;
-			}
-		r1_bio->read_disk = primary;
-		for (i=0; i<mddev->raid_disks; i++)
-			if (r1_bio->bios[i]->bi_end_io == end_sync_read) {
-				int j;
-				int vcnt = r1_bio->sectors >> (PAGE_SHIFT- 9);
-				struct bio *pbio = r1_bio->bios[primary];
-				struct bio *sbio = r1_bio->bios[i];
-
-				if (test_bit(BIO_UPTODATE, &sbio->bi_flags)) {
-					for (j = vcnt; j-- ; ) {
-						struct page *p, *s;
-						p = pbio->bi_io_vec[j].bv_page;
-						s = sbio->bi_io_vec[j].bv_page;
-						if (memcmp(page_address(p),
-							   page_address(s),
-							   PAGE_SIZE))
-							break;
-					}
-				} else
-					j = 0;
-				if (j >= 0)
-					mddev->resync_mismatches += r1_bio->sectors;
-				if (j < 0 || (test_bit(MD_RECOVERY_CHECK, &mddev->recovery)
-					      && test_bit(BIO_UPTODATE, &sbio->bi_flags))) {
-					sbio->bi_end_io = NULL;
-					rdev_dec_pending(conf->mirrors[i].rdev, mddev);
-				} else {
-					/* fixup the bio for reuse */
-					int size;
-					sbio->bi_vcnt = vcnt;
-					sbio->bi_size = r1_bio->sectors << 9;
-					sbio->bi_idx = 0;
-					sbio->bi_phys_segments = 0;
-					sbio->bi_flags &= ~(BIO_POOL_MASK - 1);
-					sbio->bi_flags |= 1 << BIO_UPTODATE;
-					sbio->bi_next = NULL;
-					sbio->bi_sector = r1_bio->sector +
-						conf->mirrors[i].rdev->data_offset;
-					sbio->bi_bdev = conf->mirrors[i].rdev->bdev;
-					size = sbio->bi_size;
-					for (j = 0; j < vcnt ; j++) {
-						struct bio_vec *bi;
-						bi = &sbio->bi_io_vec[j];
-						bi->bv_offset = 0;
-						if (size > PAGE_SIZE)
-							bi->bv_len = PAGE_SIZE;
-						else
-							bi->bv_len = size;
-						size -= PAGE_SIZE;
-						memcpy(page_address(bi->bv_page),
-						       page_address(pbio->bi_io_vec[j].bv_page),
-						       PAGE_SIZE);
-					}
 
-				}
-			}
+		start = d;
+		/* write it back and re-read */
+		while (d != r1_bio->read_disk) {
+			if (d == 0)
+				d = conf->raid_disks;
+			d--;
+			if (r1_bio->bios[d]->bi_end_io != end_sync_read)
+				continue;
+			rdev = conf->mirrors[d].rdev;
+			if (sync_page_io(rdev,
+					 sect,
+					 s<<9,
+					 bio->bi_io_vec[idx].bv_page,
+					 WRITE, false) == 0) {
+				r1_bio->bios[d]->bi_end_io = NULL;
+				rdev_dec_pending(rdev, mddev);
+				md_error(mddev, rdev);
+			} else
+				atomic_add(s, &rdev->corrected_errors);
+		}
+		d = start;
+		while (d != r1_bio->read_disk) {
+			if (d == 0)
+				d = conf->raid_disks;
+			d--;
+			if (r1_bio->bios[d]->bi_end_io != end_sync_read)
+				continue;
+			rdev = conf->mirrors[d].rdev;
+			if (sync_page_io(rdev,
+					 sect,
+					 s<<9,
+					 bio->bi_io_vec[idx].bv_page,
+					 READ, false) == 0)
+				md_error(mddev, rdev);
+		}
+		sectors -= s;
+		sect += s;
+		idx ++;
 	}
-	if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) {
-		/* ouch - failed to read all of that.
-		 * Try some synchronous reads of other devices to get
-		 * good data, much like with normal read errors.  Only
-		 * read into the pages we already have so we don't
-		 * need to re-issue the read request.
-		 * We don't need to freeze the array, because being in an
-		 * active sync request, there is no normal IO, and
-		 * no overlapping syncs.
-		 */
-		sector_t sect = r1_bio->sector;
-		int sectors = r1_bio->sectors;
-		int idx = 0;
-
-		while(sectors) {
-			int s = sectors;
-			int d = r1_bio->read_disk;
-			int success = 0;
-			mdk_rdev_t *rdev;
-
-			if (s > (PAGE_SIZE>>9))
-				s = PAGE_SIZE >> 9;
-			do {
-				if (r1_bio->bios[d]->bi_end_io == end_sync_read) {
-					/* No rcu protection needed here devices
-					 * can only be removed when no resync is
-					 * active, and resync is currently active
-					 */
-					rdev = conf->mirrors[d].rdev;
-					if (sync_page_io(rdev,
-							 sect,
-							 s<<9,
-							 bio->bi_io_vec[idx].bv_page,
-							 READ, false)) {
-						success = 1;
-						break;
-					}
-				}
-				d++;
-				if (d == conf->raid_disks)
-					d = 0;
-			} while (!success && d != r1_bio->read_disk);
-
-			if (success) {
-				int start = d;
-				/* write it back and re-read */
-				set_bit(R1BIO_Uptodate, &r1_bio->state);
-				while (d != r1_bio->read_disk) {
-					if (d == 0)
-						d = conf->raid_disks;
-					d--;
-					if (r1_bio->bios[d]->bi_end_io != end_sync_read)
-						continue;
-					rdev = conf->mirrors[d].rdev;
-					atomic_add(s, &rdev->corrected_errors);
-					if (sync_page_io(rdev,
-							 sect,
-							 s<<9,
-							 bio->bi_io_vec[idx].bv_page,
-							 WRITE, false) == 0)
-						md_error(mddev, rdev);
-				}
-				d = start;
-				while (d != r1_bio->read_disk) {
-					if (d == 0)
-						d = conf->raid_disks;
-					d--;
-					if (r1_bio->bios[d]->bi_end_io != end_sync_read)
-						continue;
-					rdev = conf->mirrors[d].rdev;
-					if (sync_page_io(rdev,
-							 sect,
-							 s<<9,
-							 bio->bi_io_vec[idx].bv_page,
-							 READ, false) == 0)
-						md_error(mddev, rdev);
-				}
-			} else {
-				char b[BDEVNAME_SIZE];
-				/* Cannot read from anywhere, array is toast */
-				md_error(mddev, conf->mirrors[r1_bio->read_disk].rdev);
-				printk(KERN_ALERT "md/raid1:%s: %s: unrecoverable I/O read error"
-				       " for block %llu\n",
-				       mdname(mddev),
-				       bdevname(bio->bi_bdev, b),
-				       (unsigned long long)r1_bio->sector);
-				md_done_sync(mddev, r1_bio->sectors, 0);
-				put_buf(r1_bio);
-				return;
+	set_bit(R1BIO_Uptodate, &r1_bio->state);
+	set_bit(BIO_UPTODATE, &bio->bi_flags);
+	return 1;
+}
+
+static int process_checks(r1bio_t *r1_bio)
+{
+	/* We have read all readable devices.  If we haven't
+	 * got the block, then there is no hope left.
+	 * If we have, then we want to do a comparison
+	 * and skip the write if everything is the same.
+	 * If any blocks failed to read, then we need to
+	 * attempt an over-write
+	 */
+	mddev_t *mddev = r1_bio->mddev;
+	conf_t *conf = mddev->private;
+	int primary;
+	int i;
+
+	for (primary = 0; primary < conf->raid_disks; primary++)
+		if (r1_bio->bios[primary]->bi_end_io == end_sync_read &&
+		    test_bit(BIO_UPTODATE, &r1_bio->bios[primary]->bi_flags)) {
+			r1_bio->bios[primary]->bi_end_io = NULL;
+			rdev_dec_pending(conf->mirrors[primary].rdev, mddev);
+			break;
+		}
+	r1_bio->read_disk = primary;
+	for (i = 0; i < conf->raid_disks; i++) {
+		int j;
+		int vcnt = r1_bio->sectors >> (PAGE_SHIFT- 9);
+		struct bio *pbio = r1_bio->bios[primary];
+		struct bio *sbio = r1_bio->bios[i];
+		int size;
+
+		if (r1_bio->bios[i]->bi_end_io != end_sync_read)
+			continue;
+
+		if (test_bit(BIO_UPTODATE, &sbio->bi_flags)) {
+			for (j = vcnt; j-- ; ) {
+				struct page *p, *s;
+				p = pbio->bi_io_vec[j].bv_page;
+				s = sbio->bi_io_vec[j].bv_page;
+				if (memcmp(page_address(p),
+					   page_address(s),
+					   PAGE_SIZE))
+					break;
 			}
-			sectors -= s;
-			sect += s;
-			idx ++;
+		} else
+			j = 0;
+		if (j >= 0)
+			mddev->resync_mismatches += r1_bio->sectors;
+		if (j < 0 || (test_bit(MD_RECOVERY_CHECK, &mddev->recovery)
+			      && test_bit(BIO_UPTODATE, &sbio->bi_flags))) {
+			/* No need to write to this device. */
+			sbio->bi_end_io = NULL;
+			rdev_dec_pending(conf->mirrors[i].rdev, mddev);
+			continue;
+		}
+		/* fixup the bio for reuse */
+		sbio->bi_vcnt = vcnt;
+		sbio->bi_size = r1_bio->sectors << 9;
+		sbio->bi_idx = 0;
+		sbio->bi_phys_segments = 0;
+		sbio->bi_flags &= ~(BIO_POOL_MASK - 1);
+		sbio->bi_flags |= 1 << BIO_UPTODATE;
+		sbio->bi_next = NULL;
+		sbio->bi_sector = r1_bio->sector +
+			conf->mirrors[i].rdev->data_offset;
+		sbio->bi_bdev = conf->mirrors[i].rdev->bdev;
+		size = sbio->bi_size;
+		for (j = 0; j < vcnt ; j++) {
+			struct bio_vec *bi;
+			bi = &sbio->bi_io_vec[j];
+			bi->bv_offset = 0;
+			if (size > PAGE_SIZE)
+				bi->bv_len = PAGE_SIZE;
+			else
+				bi->bv_len = size;
+			size -= PAGE_SIZE;
+			memcpy(page_address(bi->bv_page),
+			       page_address(pbio->bi_io_vec[j].bv_page),
+			       PAGE_SIZE);
 		}
 	}
+	return 0;
+}
 
+static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio)
+{
+	conf_t *conf = mddev->private;
+	int i;
+	int disks = conf->raid_disks;
+	struct bio *bio, *wbio;
+
+	bio = r1_bio->bios[r1_bio->read_disk];
+
+	if (!test_bit(R1BIO_Uptodate, &r1_bio->state))
+		/* ouch - failed to read all of that. */
+		if (!fix_sync_read_error(r1_bio))
+			return;
+
+	if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery))
+		if (process_checks(r1_bio) < 0)
+			return;
 	/*
 	 * schedule writes
 	 */
@@ -2063,7 +2061,7 @@ static int raid1_resize(mddev_t *mddev, sector_t sectors)
 	set_capacity(mddev->gendisk, mddev->array_sectors);
 	revalidate_disk(mddev->gendisk);
 	if (sectors > mddev->dev_sectors &&
-	    mddev->recovery_cp == MaxSector) {
+	    mddev->recovery_cp > mddev->dev_sectors) {
 		mddev->recovery_cp = mddev->dev_sectors;
 		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	}

drivers/md/raid1.h

@@ -94,7 +94,9 @@ struct r1bio_s {
 	int			read_disk;
 
 	struct list_head	retry_list;
-	struct bitmap_update	*bitmap_update;
+	/* Next two are only valid when R1BIO_BehindIO is set */
+	struct page		**behind_pages;
+	int			behind_page_count;
 	/*
 	 * if the IO is in WRITE direction, then multiple bios are used.
 	 * We choose the number when they are allocated.

drivers/md/raid10.c

@@ -271,9 +271,10 @@ static void raid10_end_read_request(struct bio *bio, int error)
 		 */
 		set_bit(R10BIO_Uptodate, &r10_bio->state);
 		raid_end_bio_io(r10_bio);
+		rdev_dec_pending(conf->mirrors[dev].rdev, conf->mddev);
 	} else {
 		/*
-		 * oops, read error:
+		 * oops, read error - keep the refcount on the rdev
 		 */
 		char b[BDEVNAME_SIZE];
 		if (printk_ratelimit())
@@ -282,8 +283,6 @@ static void raid10_end_read_request(struct bio *bio, int error)
 			       bdevname(conf->mirrors[dev].rdev->bdev,b), (unsigned long long)r10_bio->sector);
 		reschedule_retry(r10_bio);
 	}
-
-	rdev_dec_pending(conf->mirrors[dev].rdev, conf->mddev);
 }
 
 static void raid10_end_write_request(struct bio *bio, int error)
@@ -488,13 +487,19 @@ static int raid10_mergeable_bvec(struct request_queue *q,
 static int read_balance(conf_t *conf, r10bio_t *r10_bio)
 {
 	const sector_t this_sector = r10_bio->sector;
-	int disk, slot, nslot;
+	int disk, slot;
 	const int sectors = r10_bio->sectors;
-	sector_t new_distance, current_distance;
+	sector_t new_distance, best_dist;
 	mdk_rdev_t *rdev;
+	int do_balance;
+	int best_slot;
 
 	raid10_find_phys(conf, r10_bio);
 	rcu_read_lock();
+retry:
+	best_slot = -1;
+	best_dist = MaxSector;
+	do_balance = 1;
 	/*
 	 * Check if we can balance. We can balance on the whole
 	 * device if no resync is going on (recovery is ok), or below
@@ -502,86 +507,58 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio)
 	 * above the resync window.
 	 */
 	if (conf->mddev->recovery_cp < MaxSector
-	    && (this_sector + sectors >= conf->next_resync)) {
-		/* make sure that disk is operational */
-		slot = 0;
-		disk = r10_bio->devs[slot].devnum;
-
-		while ((rdev = rcu_dereference(conf->mirrors[disk].rdev)) == NULL ||
-		       r10_bio->devs[slot].bio == IO_BLOCKED ||
-		       !test_bit(In_sync, &rdev->flags)) {
-			slot++;
-			if (slot == conf->copies) {
-				slot = 0;
-				disk = -1;
-				break;
-			}
-			disk = r10_bio->devs[slot].devnum;
-		}
-		goto rb_out;
-	}
-
+	    && (this_sector + sectors >= conf->next_resync))
+		do_balance = 0;
 
-	/* make sure the disk is operational */
-	slot = 0;
-	disk = r10_bio->devs[slot].devnum;
-	while ((rdev=rcu_dereference(conf->mirrors[disk].rdev)) == NULL ||
-	       r10_bio->devs[slot].bio == IO_BLOCKED ||
-	       !test_bit(In_sync, &rdev->flags)) {
-		slot ++;
-		if (slot == conf->copies) {
-			disk = -1;
-			goto rb_out;
-		}
+	for (slot = 0; slot < conf->copies ; slot++) {
+		if (r10_bio->devs[slot].bio == IO_BLOCKED)
+			continue;
 		disk = r10_bio->devs[slot].devnum;
-	}
-
-
-	current_distance = abs(r10_bio->devs[slot].addr -
-			       conf->mirrors[disk].head_position);
-
-	/* Find the disk whose head is closest,
-	 * or - for far > 1 - find the closest to partition beginning */
-
-	for (nslot = slot; nslot < conf->copies; nslot++) {
-		int ndisk = r10_bio->devs[nslot].devnum;
-
-
-		if ((rdev=rcu_dereference(conf->mirrors[ndisk].rdev)) == NULL ||
-		    r10_bio->devs[nslot].bio == IO_BLOCKED ||
-		    !test_bit(In_sync, &rdev->flags))
+		rdev = rcu_dereference(conf->mirrors[disk].rdev);
+		if (rdev == NULL)
 			continue;
+		if (!test_bit(In_sync, &rdev->flags))
+			continue;
+
+		if (!do_balance)
+			break;
 
 		/* This optimisation is debatable, and completely destroys
 		 * sequential read speed for 'far copies' arrays.  So only
 		 * keep it for 'near' arrays, and review those later.
 		 */
-		if (conf->near_copies > 1 && !atomic_read(&rdev->nr_pending)) {
-			disk = ndisk;
-			slot = nslot;
+		if (conf->near_copies > 1 && !atomic_read(&rdev->nr_pending))
 			break;
-		}
 
 		/* for far > 1 always use the lowest address */
 		if (conf->far_copies > 1)
-			new_distance = r10_bio->devs[nslot].addr;
+			new_distance = r10_bio->devs[slot].addr;
 		else
-			new_distance = abs(r10_bio->devs[nslot].addr -
-					   conf->mirrors[ndisk].head_position);
-		if (new_distance < current_distance) {
-			current_distance = new_distance;
-			disk = ndisk;
-			slot = nslot;
+			new_distance = abs(r10_bio->devs[slot].addr -
+					   conf->mirrors[disk].head_position);
+		if (new_distance < best_dist) {
+			best_dist = new_distance;
+			best_slot = slot;
 		}
 	}
+	if (slot == conf->copies)
+		slot = best_slot;
 
-rb_out:
-	r10_bio->read_slot = slot;
-/*	conf->next_seq_sect = this_sector + sectors;*/
-
-	if (disk >= 0 && (rdev=rcu_dereference(conf->mirrors[disk].rdev))!= NULL)
-		atomic_inc(&conf->mirrors[disk].rdev->nr_pending);
-	else
+	if (slot >= 0) {
+		disk = r10_bio->devs[slot].devnum;
+		rdev = rcu_dereference(conf->mirrors[disk].rdev);
+		if (!rdev)
+			goto retry;
+		atomic_inc(&rdev->nr_pending);
+		if (test_bit(Faulty, &rdev->flags)) {
+			/* Cannot risk returning a device that failed
+			 * before we inc'ed nr_pending
+			 */
+			rdev_dec_pending(rdev, conf->mddev);
+			goto retry;
+		}
+		r10_bio->read_slot = slot;
+	} else
 		disk = -1;
 	rcu_read_unlock();
 
@@ -1460,40 +1437,33 @@ static void fix_read_error(conf_t *conf, mddev_t *mddev, r10bio_t *r10_bio)
 	int max_read_errors = atomic_read(&mddev->max_corr_read_errors);
 	int d = r10_bio->devs[r10_bio->read_slot].devnum;
 
-	rcu_read_lock();
-	rdev = rcu_dereference(conf->mirrors[d].rdev);
-	if (rdev) { /* If rdev is not NULL */
-		char b[BDEVNAME_SIZE];
-		int cur_read_error_count = 0;
+	/* still own a reference to this rdev, so it cannot
+	 * have been cleared recently.
+	 */
+	rdev = conf->mirrors[d].rdev;
 
-		bdevname(rdev->bdev, b);
+	if (test_bit(Faulty, &rdev->flags))
+		/* drive has already been failed, just ignore any
+		   more fix_read_error() attempts */
+		return;
 
-		if (test_bit(Faulty, &rdev->flags)) {
-			rcu_read_unlock();
-			/* drive has already been failed, just ignore any
-			   more fix_read_error() attempts */
-			return;
-		}
+	check_decay_read_errors(mddev, rdev);
+	atomic_inc(&rdev->read_errors);
+	if (atomic_read(&rdev->read_errors) > max_read_errors) {
+		char b[BDEVNAME_SIZE];
+		bdevname(rdev->bdev, b);
 
-		check_decay_read_errors(mddev, rdev);
-		atomic_inc(&rdev->read_errors);
-		cur_read_error_count = atomic_read(&rdev->read_errors);
-		if (cur_read_error_count > max_read_errors) {
-			rcu_read_unlock();
-			printk(KERN_NOTICE
-			       "md/raid10:%s: %s: Raid device exceeded "
-			       "read_error threshold "
-			       "[cur %d:max %d]\n",
-			       mdname(mddev),
-			       b, cur_read_error_count, max_read_errors);
-			printk(KERN_NOTICE
-			       "md/raid10:%s: %s: Failing raid "
-			       "device\n", mdname(mddev), b);
-			md_error(mddev, conf->mirrors[d].rdev);
-			return;
-		}
+		printk(KERN_NOTICE
+		       "md/raid10:%s: %s: Raid device exceeded "
+		       "read_error threshold [cur %d:max %d]\n",
+		       mdname(mddev), b,
+		       atomic_read(&rdev->read_errors), max_read_errors);
+		printk(KERN_NOTICE
+		       "md/raid10:%s: %s: Failing raid device\n",
+		       mdname(mddev), b);
+		md_error(mddev, conf->mirrors[d].rdev);
+		return;
 	}
-	rcu_read_unlock();
 
 	while(sectors) {
 		int s = sectors;
@@ -1562,8 +1532,8 @@ static void fix_read_error(conf_t *conf, mddev_t *mddev, r10bio_t *r10_bio)
 					       "write failed"
 					       " (%d sectors at %llu on %s)\n",
 					       mdname(mddev), s,
-					       (unsigned long long)(sect+
-					       rdev->data_offset),
+					       (unsigned long long)(
+						       sect + rdev->data_offset),
 					       bdevname(rdev->bdev, b));
 					printk(KERN_NOTICE "md/raid10:%s: %s: failing "
 					       "drive\n",
@@ -1599,8 +1569,8 @@ static void fix_read_error(conf_t *conf, mddev_t *mddev, r10bio_t *r10_bio)
 					       "corrected sectors"
 					       " (%d sectors at %llu on %s)\n",
 					       mdname(mddev), s,
-					       (unsigned long long)(sect+
-						    rdev->data_offset),
+					       (unsigned long long)(
+						       sect + rdev->data_offset),
 					       bdevname(rdev->bdev, b));
 					printk(KERN_NOTICE "md/raid10:%s: %s: failing drive\n",
 					       mdname(mddev),
@@ -1612,8 +1582,8 @@ static void fix_read_error(conf_t *conf, mddev_t *mddev, r10bio_t *r10_bio)
 					       "md/raid10:%s: read error corrected"
 					       " (%d sectors at %llu on %s)\n",
 					       mdname(mddev), s,
-					       (unsigned long long)(sect+
-					            rdev->data_offset),
+					       (unsigned long long)(
+						       sect + rdev->data_offset),
 					       bdevname(rdev->bdev, b));
 				}
 
@@ -1663,7 +1633,8 @@ static void raid10d(mddev_t *mddev)
 		else if (test_bit(R10BIO_IsRecover, &r10_bio->state))
 			recovery_request_write(mddev, r10_bio);
 		else {
-			int mirror;
+			int slot = r10_bio->read_slot;
+			int mirror = r10_bio->devs[slot].devnum;
 			/* we got a read error. Maybe the drive is bad.  Maybe just
 			 * the block and we can fix it.
 			 * We freeze all other IO, and try reading the block from
@@ -1677,9 +1648,10 @@ static void raid10d(mddev_t *mddev)
 				fix_read_error(conf, mddev, r10_bio);
 				unfreeze_array(conf);
 			}
+			rdev_dec_pending(conf->mirrors[mirror].rdev, mddev);
 
-			bio = r10_bio->devs[r10_bio->read_slot].bio;
-			r10_bio->devs[r10_bio->read_slot].bio =
+			bio = r10_bio->devs[slot].bio;
+			r10_bio->devs[slot].bio =
 				mddev->ro ? IO_BLOCKED : NULL;
 			mirror = read_balance(conf, r10_bio);
 			if (mirror == -1) {
@@ -1693,6 +1665,7 @@ static void raid10d(mddev_t *mddev)
 			} else {
 				const unsigned long do_sync = (r10_bio->master_bio->bi_rw & REQ_SYNC);
 				bio_put(bio);
+				slot = r10_bio->read_slot;
 				rdev = conf->mirrors[mirror].rdev;
 				if (printk_ratelimit())
 					printk(KERN_ERR "md/raid10:%s: %s: redirecting sector %llu to"
@@ -1702,8 +1675,8 @@ static void raid10d(mddev_t *mddev)
 					       (unsigned long long)r10_bio->sector);
 				bio = bio_clone_mddev(r10_bio->master_bio,
 						      GFP_NOIO, mddev);
-				r10_bio->devs[r10_bio->read_slot].bio = bio;
-				bio->bi_sector = r10_bio->devs[r10_bio->read_slot].addr
+				r10_bio->devs[slot].bio = bio;
+				bio->bi_sector = r10_bio->devs[slot].addr
 					+ rdev->data_offset;
 				bio->bi_bdev = rdev->bdev;
 				bio->bi_rw = READ | do_sync;
@@ -1763,13 +1736,13 @@ static int init_resync(conf_t *conf)
  *
  */
 
-static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster)
+static sector_t sync_request(mddev_t *mddev, sector_t sector_nr,
+			     int *skipped, int go_faster)
 {
 	conf_t *conf = mddev->private;
 	r10bio_t *r10_bio;
 	struct bio *biolist = NULL, *bio;
 	sector_t max_sector, nr_sectors;
-	int disk;
 	int i;
 	int max_sync;
 	sector_t sync_blocks;
@@ -1858,108 +1831,114 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 		int j, k;
 		r10_bio = NULL;
 
-		for (i=0 ; i<conf->raid_disks; i++)
-			if (conf->mirrors[i].rdev &&
-			    !test_bit(In_sync, &conf->mirrors[i].rdev->flags)) {
-				int still_degraded = 0;
-				/* want to reconstruct this device */
-				r10bio_t *rb2 = r10_bio;
-				sector_t sect = raid10_find_virt(conf, sector_nr, i);
-				int must_sync;
-				/* Unless we are doing a full sync, we only need
-				 * to recover the block if it is set in the bitmap
-				 */
-				must_sync = bitmap_start_sync(mddev->bitmap, sect,
-							      &sync_blocks, 1);
-				if (sync_blocks < max_sync)
-					max_sync = sync_blocks;
-				if (!must_sync &&
-				    !conf->fullsync) {
-					/* yep, skip the sync_blocks here, but don't assume
-					 * that there will never be anything to do here
-					 */
-					chunks_skipped = -1;
-					continue;
-				}
+		for (i=0 ; i<conf->raid_disks; i++) {
+			int still_degraded;
+			r10bio_t *rb2;
+			sector_t sect;
+			int must_sync;
 
-				r10_bio = mempool_alloc(conf->r10buf_pool, GFP_NOIO);
-				raise_barrier(conf, rb2 != NULL);
-				atomic_set(&r10_bio->remaining, 0);
+			if (conf->mirrors[i].rdev == NULL ||
+			    test_bit(In_sync, &conf->mirrors[i].rdev->flags)) 
+				continue;
 
-				r10_bio->master_bio = (struct bio*)rb2;
-				if (rb2)
-					atomic_inc(&rb2->remaining);
-				r10_bio->mddev = mddev;
-				set_bit(R10BIO_IsRecover, &r10_bio->state);
-				r10_bio->sector = sect;
+			still_degraded = 0;
+			/* want to reconstruct this device */
+			rb2 = r10_bio;
+			sect = raid10_find_virt(conf, sector_nr, i);
+			/* Unless we are doing a full sync, we only need
+			 * to recover the block if it is set in the bitmap
+			 */
+			must_sync = bitmap_start_sync(mddev->bitmap, sect,
+						      &sync_blocks, 1);
+			if (sync_blocks < max_sync)
+				max_sync = sync_blocks;
+			if (!must_sync &&
+			    !conf->fullsync) {
+				/* yep, skip the sync_blocks here, but don't assume
+				 * that there will never be anything to do here
+				 */
+				chunks_skipped = -1;
+				continue;
+			}
 
-				raid10_find_phys(conf, r10_bio);
+			r10_bio = mempool_alloc(conf->r10buf_pool, GFP_NOIO);
+			raise_barrier(conf, rb2 != NULL);
+			atomic_set(&r10_bio->remaining, 0);
 
-				/* Need to check if the array will still be
-				 * degraded
-				 */
-				for (j=0; j<conf->raid_disks; j++)
-					if (conf->mirrors[j].rdev == NULL ||
-					    test_bit(Faulty, &conf->mirrors[j].rdev->flags)) {
-						still_degraded = 1;
-						break;
-					}
-
-				must_sync = bitmap_start_sync(mddev->bitmap, sect,
-							      &sync_blocks, still_degraded);
-
-				for (j=0; j<conf->copies;j++) {
-					int d = r10_bio->devs[j].devnum;
-					if (conf->mirrors[d].rdev &&
-					    test_bit(In_sync, &conf->mirrors[d].rdev->flags)) {
-						/* This is where we read from */
-						bio = r10_bio->devs[0].bio;
-						bio->bi_next = biolist;
-						biolist = bio;
-						bio->bi_private = r10_bio;
-						bio->bi_end_io = end_sync_read;
-						bio->bi_rw = READ;
-						bio->bi_sector = r10_bio->devs[j].addr +
-							conf->mirrors[d].rdev->data_offset;
-						bio->bi_bdev = conf->mirrors[d].rdev->bdev;
-						atomic_inc(&conf->mirrors[d].rdev->nr_pending);
-						atomic_inc(&r10_bio->remaining);
-						/* and we write to 'i' */
-
-						for (k=0; k<conf->copies; k++)
-							if (r10_bio->devs[k].devnum == i)
-								break;
-						BUG_ON(k == conf->copies);
-						bio = r10_bio->devs[1].bio;
-						bio->bi_next = biolist;
-						biolist = bio;
-						bio->bi_private = r10_bio;
-						bio->bi_end_io = end_sync_write;
-						bio->bi_rw = WRITE;
-						bio->bi_sector = r10_bio->devs[k].addr +
-							conf->mirrors[i].rdev->data_offset;
-						bio->bi_bdev = conf->mirrors[i].rdev->bdev;
-
-						r10_bio->devs[0].devnum = d;
-						r10_bio->devs[1].devnum = i;
+			r10_bio->master_bio = (struct bio*)rb2;
+			if (rb2)
+				atomic_inc(&rb2->remaining);
+			r10_bio->mddev = mddev;
+			set_bit(R10BIO_IsRecover, &r10_bio->state);
+			r10_bio->sector = sect;
 
-						break;
-					}
-				}
-				if (j == conf->copies) {
-					/* Cannot recover, so abort the recovery */
-					put_buf(r10_bio);
-					if (rb2)
-						atomic_dec(&rb2->remaining);
-					r10_bio = rb2;
-					if (!test_and_set_bit(MD_RECOVERY_INTR,
-							      &mddev->recovery))
-						printk(KERN_INFO "md/raid10:%s: insufficient "
-						       "working devices for recovery.\n",
-						       mdname(mddev));
+			raid10_find_phys(conf, r10_bio);
+
+			/* Need to check if the array will still be
+			 * degraded
+			 */
+			for (j=0; j<conf->raid_disks; j++)
+				if (conf->mirrors[j].rdev == NULL ||
+				    test_bit(Faulty, &conf->mirrors[j].rdev->flags)) {
+					still_degraded = 1;
 					break;
 				}
+
+			must_sync = bitmap_start_sync(mddev->bitmap, sect,
+						      &sync_blocks, still_degraded);
+
+			for (j=0; j<conf->copies;j++) {
+				int d = r10_bio->devs[j].devnum;
+				if (!conf->mirrors[d].rdev ||
+				    !test_bit(In_sync, &conf->mirrors[d].rdev->flags))
+					continue;
+				/* This is where we read from */
+				bio = r10_bio->devs[0].bio;
+				bio->bi_next = biolist;
+				biolist = bio;
+				bio->bi_private = r10_bio;
+				bio->bi_end_io = end_sync_read;
+				bio->bi_rw = READ;
+				bio->bi_sector = r10_bio->devs[j].addr +
+					conf->mirrors[d].rdev->data_offset;
+				bio->bi_bdev = conf->mirrors[d].rdev->bdev;
+				atomic_inc(&conf->mirrors[d].rdev->nr_pending);
+				atomic_inc(&r10_bio->remaining);
+				/* and we write to 'i' */
+
+				for (k=0; k<conf->copies; k++)
+					if (r10_bio->devs[k].devnum == i)
+						break;
+				BUG_ON(k == conf->copies);
+				bio = r10_bio->devs[1].bio;
+				bio->bi_next = biolist;
+				biolist = bio;
+				bio->bi_private = r10_bio;
+				bio->bi_end_io = end_sync_write;
+				bio->bi_rw = WRITE;
+				bio->bi_sector = r10_bio->devs[k].addr +
+					conf->mirrors[i].rdev->data_offset;
+				bio->bi_bdev = conf->mirrors[i].rdev->bdev;
+
+				r10_bio->devs[0].devnum = d;
+				r10_bio->devs[1].devnum = i;
+
+				break;
+			}
+			if (j == conf->copies) {
+				/* Cannot recover, so abort the recovery */
+				put_buf(r10_bio);
+				if (rb2)
+					atomic_dec(&rb2->remaining);
+				r10_bio = rb2;
+				if (!test_and_set_bit(MD_RECOVERY_INTR,
+						      &mddev->recovery))
+					printk(KERN_INFO "md/raid10:%s: insufficient "
+					       "working devices for recovery.\n",
+					       mdname(mddev));
+				break;
 			}
+		}
 		if (biolist == NULL) {
 			while (r10_bio) {
 				r10bio_t *rb2 = r10_bio;
@@ -1977,7 +1956,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 
 		if (!bitmap_start_sync(mddev->bitmap, sector_nr,
 				       &sync_blocks, mddev->degraded) &&
-		    !conf->fullsync && !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
+		    !conf->fullsync && !test_bit(MD_RECOVERY_REQUESTED,
+						 &mddev->recovery)) {
 			/* We can skip this block */
 			*skipped = 1;
 			return sync_blocks + sectors_skipped;
@@ -2022,7 +2002,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 			for (i=0; i<conf->copies; i++) {
 				int d = r10_bio->devs[i].devnum;
 				if (r10_bio->devs[i].bio->bi_end_io)
-					rdev_dec_pending(conf->mirrors[d].rdev, mddev);
+					rdev_dec_pending(conf->mirrors[d].rdev,
+							 mddev);
 			}
 			put_buf(r10_bio);
 			biolist = NULL;
@@ -2047,26 +2028,27 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 	do {
 		struct page *page;
 		int len = PAGE_SIZE;
-		disk = 0;
 		if (sector_nr + (len>>9) > max_sector)
 			len = (max_sector - sector_nr) << 9;
 		if (len == 0)
 			break;
 		for (bio= biolist ; bio ; bio=bio->bi_next) {
+			struct bio *bio2;
 			page = bio->bi_io_vec[bio->bi_vcnt].bv_page;
-			if (bio_add_page(bio, page, len, 0) == 0) {
-				/* stop here */
-				struct bio *bio2;
-				bio->bi_io_vec[bio->bi_vcnt].bv_page = page;
-				for (bio2 = biolist; bio2 && bio2 != bio; bio2 = bio2->bi_next) {
-					/* remove last page from this bio */
-					bio2->bi_vcnt--;
-					bio2->bi_size -= len;
-					bio2->bi_flags &= ~(1<< BIO_SEG_VALID);
-				}
-				goto bio_full;
+			if (bio_add_page(bio, page, len, 0))
+				continue;
+
+			/* stop here */
+			bio->bi_io_vec[bio->bi_vcnt].bv_page = page;
+			for (bio2 = biolist;
+			     bio2 && bio2 != bio;
+			     bio2 = bio2->bi_next) {
+				/* remove last page from this bio */
+				bio2->bi_vcnt--;
+				bio2->bi_size -= len;
+				bio2->bi_flags &= ~(1<< BIO_SEG_VALID);
 			}
-			disk = i;
+			goto bio_full;
 		}
 		nr_sectors += len>>9;
 		sector_nr += len>>9;

drivers/md/raid5.c

@@ -1700,27 +1700,25 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev)
 	raid5_conf_t *conf = mddev->private;
 	pr_debug("raid456: error called\n");
 
-	if (!test_bit(Faulty, &rdev->flags)) {
-		set_bit(MD_CHANGE_DEVS, &mddev->flags);
-		if (test_and_clear_bit(In_sync, &rdev->flags)) {
-			unsigned long flags;
-			spin_lock_irqsave(&conf->device_lock, flags);
-			mddev->degraded++;
-			spin_unlock_irqrestore(&conf->device_lock, flags);
-			/*
-			 * if recovery was running, make sure it aborts.
-			 */
-			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
-		}
-		set_bit(Faulty, &rdev->flags);
-		printk(KERN_ALERT
-		       "md/raid:%s: Disk failure on %s, disabling device.\n"
-		       "md/raid:%s: Operation continuing on %d devices.\n",
-		       mdname(mddev),
-		       bdevname(rdev->bdev, b),
-		       mdname(mddev),
-		       conf->raid_disks - mddev->degraded);
+	if (test_and_clear_bit(In_sync, &rdev->flags)) {
+		unsigned long flags;
+		spin_lock_irqsave(&conf->device_lock, flags);
+		mddev->degraded++;
+		spin_unlock_irqrestore(&conf->device_lock, flags);
+		/*
+		 * if recovery was running, make sure it aborts.
+		 */
+		set_bit(MD_RECOVERY_INTR, &mddev->recovery);
 	}
+	set_bit(Faulty, &rdev->flags);
+	set_bit(MD_CHANGE_DEVS, &mddev->flags);
+	printk(KERN_ALERT
+	       "md/raid:%s: Disk failure on %s, disabling device.\n"
+	       "md/raid:%s: Operation continuing on %d devices.\n",
+	       mdname(mddev),
+	       bdevname(rdev->bdev, b),
+	       mdname(mddev),
+	       conf->raid_disks - mddev->degraded);
 }
 
 /*
@@ -5391,7 +5389,8 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors)
 		return -EINVAL;
 	set_capacity(mddev->gendisk, mddev->array_sectors);
 	revalidate_disk(mddev->gendisk);
-	if (sectors > mddev->dev_sectors && mddev->recovery_cp == MaxSector) {
+	if (sectors > mddev->dev_sectors &&
+	    mddev->recovery_cp > mddev->dev_sectors) {
 		mddev->recovery_cp = mddev->dev_sectors;
 		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	}