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

* 'for-linus' of git://neil.brown.name/md: (39 commits)
  md/raid5: correctly update sync_completed when we reach max_resync
  md/raid5: add missing call to schedule() after prepare_to_wait()
  md/linear: use call_rcu to free obsolete 'conf' structures.
  md linear: Protecting mddev with rcu locks to avoid races
  md: Move check for bitmap presence to personality code.
  md: remove chunksize rounding from common code.
  md: raid0/linear: ensure device sizes are rounded to chunk size.
  md: move assignment of ->utime so that it never gets skipped.
  md: Push down reconstruction log message to personality code.
  md: merge reconfig and check_reshape methods.
  md: remove unnecessary arguments from ->reconfig method.
  md: raid5: check stripe cache is large enough in start_reshape
  md: raid0: chunk_sectors cleanups.
  md: fix some comments.
  md/raid5: Use is_power_of_2() in raid5_reconfig()/raid6_reconfig().
  md: convert conf->chunk_size and conf->prev_chunk to sectors.
  md: Convert mddev->new_chunk to sectors.
  md: Make mddev->chunk_size sector-based.
  md: raid0 :Enables chunk size other than powers of 2.
  md: prepare for non-power-of-two chunk sizes
  ...

drivers/md/faulty.c

@@ -255,14 +255,14 @@ static void status(struct seq_file *seq, mddev_t *mddev)
 }
 
 
-static int reconfig(mddev_t *mddev, int layout, int chunk_size)
+static int reshape(mddev_t *mddev)
 {
-	int mode = layout & ModeMask;
-	int count = layout >> ModeShift;
+	int mode = mddev->new_layout & ModeMask;
+	int count = mddev->new_layout >> ModeShift;
 	conf_t *conf = mddev->private;
 
-	if (chunk_size != -1)
-		return -EINVAL;
+	if (mddev->new_layout < 0)
+		return 0;
 
 	/* new layout */
 	if (mode == ClearFaults)
@@ -279,6 +279,7 @@ static int reconfig(mddev_t *mddev, int layout, int chunk_size)
 		atomic_set(&conf->counters[mode], count);
 	} else
 		return -EINVAL;
+	mddev->new_layout = -1;
 	mddev->layout = -1; /* makes sure further changes come through */
 	return 0;
 }
@@ -298,8 +299,12 @@ static int run(mddev_t *mddev)
 {
 	mdk_rdev_t *rdev;
 	int i;
+	conf_t *conf;
+
+	if (md_check_no_bitmap(mddev))
+		return -EINVAL;
 
-	conf_t *conf = kmalloc(sizeof(*conf), GFP_KERNEL);
+	conf = kmalloc(sizeof(*conf), GFP_KERNEL);
 	if (!conf)
 		return -ENOMEM;
 
@@ -315,7 +320,7 @@ static int run(mddev_t *mddev)
 	md_set_array_sectors(mddev, faulty_size(mddev, 0, 0));
 	mddev->private = conf;
 
-	reconfig(mddev, mddev->layout, -1);
+	reshape(mddev);
 
 	return 0;
 }
@@ -338,7 +343,7 @@ static struct mdk_personality faulty_personality =
 	.run		= run,
 	.stop		= stop,
 	.status		= status,
-	.reconfig	= reconfig,
+	.check_reshape	= reshape,
 	.size		= faulty_size,
 };
 

drivers/md/linear.c

@@ -27,19 +27,27 @@
  */
 static inline dev_info_t *which_dev(mddev_t *mddev, sector_t sector)
 {
-	dev_info_t *hash;
-	linear_conf_t *conf = mddev_to_conf(mddev);
-	sector_t idx = sector >> conf->sector_shift;
+	int lo, mid, hi;
+	linear_conf_t *conf;
+
+	lo = 0;
+	hi = mddev->raid_disks - 1;
+	conf = rcu_dereference(mddev->private);
 
 	/*
-	 * sector_div(a,b) returns the remainer and sets a to a/b
+	 * Binary Search
 	 */
-	(void)sector_div(idx, conf->spacing);
-	hash = conf->hash_table[idx];
 
-	while (sector >= hash->num_sectors + hash->start_sector)
-		hash++;
-	return hash;
+	while (hi > lo) {
+
+		mid = (hi + lo) / 2;
+		if (sector < conf->disks[mid].end_sector)
+			hi = mid;
+		else
+			lo = mid + 1;
+	}
+
+	return conf->disks + lo;
 }
 
 /**
@@ -59,8 +67,10 @@ static int linear_mergeable_bvec(struct request_queue *q,
 	unsigned long maxsectors, bio_sectors = bvm->bi_size >> 9;
 	sector_t sector = bvm->bi_sector + get_start_sect(bvm->bi_bdev);
 
+	rcu_read_lock();
 	dev0 = which_dev(mddev, sector);
-	maxsectors = dev0->num_sectors - (sector - dev0->start_sector);
+	maxsectors = dev0->end_sector - sector;
+	rcu_read_unlock();
 
 	if (maxsectors < bio_sectors)
 		maxsectors = 0;
@@ -79,46 +89,57 @@ static int linear_mergeable_bvec(struct request_queue *q,
 static void linear_unplug(struct request_queue *q)
 {
 	mddev_t *mddev = q->queuedata;
-	linear_conf_t *conf = mddev_to_conf(mddev);
+	linear_conf_t *conf;
 	int i;
 
+	rcu_read_lock();
+	conf = rcu_dereference(mddev->private);
+
 	for (i=0; i < mddev->raid_disks; i++) {
 		struct request_queue *r_queue = bdev_get_queue(conf->disks[i].rdev->bdev);
 		blk_unplug(r_queue);
 	}
+	rcu_read_unlock();
 }
 
 static int linear_congested(void *data, int bits)
 {
 	mddev_t *mddev = data;
-	linear_conf_t *conf = mddev_to_conf(mddev);
+	linear_conf_t *conf;
 	int i, ret = 0;
 
+	rcu_read_lock();
+	conf = rcu_dereference(mddev->private);
+
 	for (i = 0; i < mddev->raid_disks && !ret ; i++) {
 		struct request_queue *q = bdev_get_queue(conf->disks[i].rdev->bdev);
 		ret |= bdi_congested(&q->backing_dev_info, bits);
 	}
+
+	rcu_read_unlock();
 	return ret;
 }
 
 static sector_t linear_size(mddev_t *mddev, sector_t sectors, int raid_disks)
 {
-	linear_conf_t *conf = mddev_to_conf(mddev);
+	linear_conf_t *conf;
+	sector_t array_sectors;
 
+	rcu_read_lock();
+	conf = rcu_dereference(mddev->private);
 	WARN_ONCE(sectors || raid_disks,
 		  "%s does not support generic reshape\n", __func__);
+	array_sectors = conf->array_sectors;
+	rcu_read_unlock();
 
-	return conf->array_sectors;
+	return array_sectors;
 }
 
 static linear_conf_t *linear_conf(mddev_t *mddev, int raid_disks)
 {
 	linear_conf_t *conf;
-	dev_info_t **table;
 	mdk_rdev_t *rdev;
-	int i, nb_zone, cnt;
-	sector_t min_sectors;
-	sector_t curr_sector;
+	int i, cnt;
 
 	conf = kzalloc (sizeof (*conf) + raid_disks*sizeof(dev_info_t),
 			GFP_KERNEL);
@@ -131,6 +152,7 @@ static linear_conf_t *linear_conf(mddev_t *mddev, int raid_disks)
 	list_for_each_entry(rdev, &mddev->disks, same_set) {
 		int j = rdev->raid_disk;
 		dev_info_t *disk = conf->disks + j;
+		sector_t sectors;
 
 		if (j < 0 || j >= raid_disks || disk->rdev) {
 			printk("linear: disk numbering problem. Aborting!\n");
@@ -138,6 +160,11 @@ static linear_conf_t *linear_conf(mddev_t *mddev, int raid_disks)
 		}
 
 		disk->rdev = rdev;
+		if (mddev->chunk_sectors) {
+			sectors = rdev->sectors;
+			sector_div(sectors, mddev->chunk_sectors);
+			rdev->sectors = sectors * mddev->chunk_sectors;
+		}
 
 		blk_queue_stack_limits(mddev->queue,
 				       rdev->bdev->bd_disk->queue);
@@ -149,102 +176,24 @@ static linear_conf_t *linear_conf(mddev_t *mddev, int raid_disks)
 		    queue_max_sectors(mddev->queue) > (PAGE_SIZE>>9))
 			blk_queue_max_sectors(mddev->queue, PAGE_SIZE>>9);
 
-		disk->num_sectors = rdev->sectors;
 		conf->array_sectors += rdev->sectors;
-
 		cnt++;
+
 	}
 	if (cnt != raid_disks) {
 		printk("linear: not enough drives present. Aborting!\n");
 		goto out;
 	}
 
-	min_sectors = conf->array_sectors;
-	sector_div(min_sectors, PAGE_SIZE/sizeof(struct dev_info *));
-	if (min_sectors == 0)
-		min_sectors = 1;
-
-	/* min_sectors is the minimum spacing that will fit the hash
-	 * table in one PAGE.  This may be much smaller than needed.
-	 * We find the smallest non-terminal set of consecutive devices
-	 * that is larger than min_sectors and use the size of that as
-	 * the actual spacing
-	 */
-	conf->spacing = conf->array_sectors;
-	for (i=0; i < cnt-1 ; i++) {
-		sector_t tmp = 0;
-		int j;
-		for (j = i; j < cnt - 1 && tmp < min_sectors; j++)
-			tmp += conf->disks[j].num_sectors;
-		if (tmp >= min_sectors && tmp < conf->spacing)
-			conf->spacing = tmp;
-	}
-
-	/* spacing may be too large for sector_div to work with,
-	 * so we might need to pre-shift
-	 */
-	conf->sector_shift = 0;
-	if (sizeof(sector_t) > sizeof(u32)) {
-		sector_t space = conf->spacing;
-		while (space > (sector_t)(~(u32)0)) {
-			space >>= 1;
-			conf->sector_shift++;
-		}
-	}
 	/*
-	 * This code was restructured to work around a gcc-2.95.3 internal
-	 * compiler error.  Alter it with care.
+	 * Here we calculate the device offsets.
 	 */
-	{
-		sector_t sz;
-		unsigned round;
-		unsigned long base;
-
-		sz = conf->array_sectors >> conf->sector_shift;
-		sz += 1; /* force round-up */
-		base = conf->spacing >> conf->sector_shift;
-		round = sector_div(sz, base);
-		nb_zone = sz + (round ? 1 : 0);
-	}
-	BUG_ON(nb_zone > PAGE_SIZE / sizeof(struct dev_info *));
-
-	conf->hash_table = kmalloc (sizeof (struct dev_info *) * nb_zone,
-					GFP_KERNEL);
-	if (!conf->hash_table)
-		goto out;
+	conf->disks[0].end_sector = conf->disks[0].rdev->sectors;
 
-	/*
-	 * Here we generate the linear hash table
-	 * First calculate the device offsets.
-	 */
-	conf->disks[0].start_sector = 0;
 	for (i = 1; i < raid_disks; i++)
-		conf->disks[i].start_sector =
-			conf->disks[i-1].start_sector +
-			conf->disks[i-1].num_sectors;
-
-	table = conf->hash_table;
-	i = 0;
-	for (curr_sector = 0;
-	     curr_sector < conf->array_sectors;
-	     curr_sector += conf->spacing) {
-
-		while (i < raid_disks-1 &&
-		       curr_sector >= conf->disks[i+1].start_sector)
-			i++;
-
-		*table ++ = conf->disks + i;
-	}
-
-	if (conf->sector_shift) {
-		conf->spacing >>= conf->sector_shift;
-		/* round spacing up so that when we divide by it,
-		 * we err on the side of "too-low", which is safest.
-		 */
-		conf->spacing++;
-	}
-
-	BUG_ON(table - conf->hash_table > nb_zone);
+		conf->disks[i].end_sector =
+			conf->disks[i-1].end_sector +
+			conf->disks[i].rdev->sectors;
 
 	return conf;
 
@@ -257,6 +206,8 @@ static int linear_run (mddev_t *mddev)
 {
 	linear_conf_t *conf;
 
+	if (md_check_no_bitmap(mddev))
+		return -EINVAL;
 	mddev->queue->queue_lock = &mddev->queue->__queue_lock;
 	conf = linear_conf(mddev, mddev->raid_disks);
 
@@ -272,6 +223,12 @@ static int linear_run (mddev_t *mddev)
 	return 0;
 }
 
+static void free_conf(struct rcu_head *head)
+{
+	linear_conf_t *conf = container_of(head, linear_conf_t, rcu);
+	kfree(conf);
+}
+
 static int linear_add(mddev_t *mddev, mdk_rdev_t *rdev)
 {
 	/* Adding a drive to a linear array allows the array to grow.
@@ -282,7 +239,7 @@ static int linear_add(mddev_t *mddev, mdk_rdev_t *rdev)
 	 * The current one is never freed until the array is stopped.
 	 * This avoids races.
 	 */
-	linear_conf_t *newconf;
+	linear_conf_t *newconf, *oldconf;
 
 	if (rdev->saved_raid_disk != mddev->raid_disks)
 		return -EINVAL;
@@ -294,25 +251,29 @@ static int linear_add(mddev_t *mddev, mdk_rdev_t *rdev)
 	if (!newconf)
 		return -ENOMEM;
 
-	newconf->prev = mddev_to_conf(mddev);
-	mddev->private = newconf;
+	oldconf = rcu_dereference(mddev->private);
 	mddev->raid_disks++;
+	rcu_assign_pointer(mddev->private, newconf);
 	md_set_array_sectors(mddev, linear_size(mddev, 0, 0));
 	set_capacity(mddev->gendisk, mddev->array_sectors);
+	call_rcu(&oldconf->rcu, free_conf);
 	return 0;
 }
 
 static int linear_stop (mddev_t *mddev)
 {
-	linear_conf_t *conf = mddev_to_conf(mddev);
-  
+	linear_conf_t *conf = mddev->private;
+
+	/*
+	 * We do not require rcu protection here since
+	 * we hold reconfig_mutex for both linear_add and
+	 * linear_stop, so they cannot race.
+	 * We should make sure any old 'conf's are properly
+	 * freed though.
+	 */
+	rcu_barrier();
 	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
-	do {
-		linear_conf_t *t = conf->prev;
-		kfree(conf->hash_table);
-		kfree(conf);
-		conf = t;
-	} while (conf);
+	kfree(conf);
 
 	return 0;
 }
@@ -322,6 +283,7 @@ static int linear_make_request (struct request_queue *q, struct bio *bio)
 	const int rw = bio_data_dir(bio);
 	mddev_t *mddev = q->queuedata;
 	dev_info_t *tmp_dev;
+	sector_t start_sector;
 	int cpu;
 
 	if (unlikely(bio_barrier(bio))) {
@@ -335,33 +297,36 @@ static int linear_make_request (struct request_queue *q, struct bio *bio)
 		      bio_sectors(bio));
 	part_stat_unlock();
 
+	rcu_read_lock();
 	tmp_dev = which_dev(mddev, bio->bi_sector);
-    
-	if (unlikely(bio->bi_sector >= (tmp_dev->num_sectors +
-					tmp_dev->start_sector)
-		     || (bio->bi_sector <
-			 tmp_dev->start_sector))) {
+	start_sector = tmp_dev->end_sector - tmp_dev->rdev->sectors;
+
+
+	if (unlikely(bio->bi_sector >= (tmp_dev->end_sector)
+		     || (bio->bi_sector < start_sector))) {
 		char b[BDEVNAME_SIZE];
 
 		printk("linear_make_request: Sector %llu out of bounds on "
 			"dev %s: %llu sectors, offset %llu\n",
 			(unsigned long long)bio->bi_sector,
 			bdevname(tmp_dev->rdev->bdev, b),
-			(unsigned long long)tmp_dev->num_sectors,
-			(unsigned long long)tmp_dev->start_sector);
+			(unsigned long long)tmp_dev->rdev->sectors,
+			(unsigned long long)start_sector);
+		rcu_read_unlock();
 		bio_io_error(bio);
 		return 0;
 	}
 	if (unlikely(bio->bi_sector + (bio->bi_size >> 9) >
-		     tmp_dev->start_sector + tmp_dev->num_sectors)) {
+		     tmp_dev->end_sector)) {
 		/* This bio crosses a device boundary, so we have to
 		 * split it.
 		 */
 		struct bio_pair *bp;
+		sector_t end_sector = tmp_dev->end_sector;
+
+		rcu_read_unlock();
 
-		bp = bio_split(bio,
-			       tmp_dev->start_sector + tmp_dev->num_sectors
-			       - bio->bi_sector);
+		bp = bio_split(bio, end_sector - bio->bi_sector);
 
 		if (linear_make_request(q, &bp->bio1))
 			generic_make_request(&bp->bio1);
@@ -372,8 +337,9 @@ static int linear_make_request (struct request_queue *q, struct bio *bio)
 	}
 		    
 	bio->bi_bdev = tmp_dev->rdev->bdev;
-	bio->bi_sector = bio->bi_sector - tmp_dev->start_sector
+	bio->bi_sector = bio->bi_sector - start_sector
 		+ tmp_dev->rdev->data_offset;
+	rcu_read_unlock();
 
 	return 1;
 }
@@ -381,7 +347,7 @@ static int linear_make_request (struct request_queue *q, struct bio *bio)
 static void linear_status (struct seq_file *seq, mddev_t *mddev)
 {
 
-	seq_printf(seq, " %dk rounding", mddev->chunk_size/1024);
+	seq_printf(seq, " %dk rounding", mddev->chunk_sectors / 2);
 }
 
 

drivers/md/linear.h

@@ -3,27 +3,19 @@
 
 struct dev_info {
 	mdk_rdev_t	*rdev;
-	sector_t	num_sectors;
-	sector_t	start_sector;
+	sector_t	end_sector;
 };
 
 typedef struct dev_info dev_info_t;
 
 struct linear_private_data
 {
-	struct linear_private_data *prev;	/* earlier version */
-	dev_info_t		**hash_table;
-	sector_t		spacing;
 	sector_t		array_sectors;
-	int			sector_shift;	/* shift before dividing
-						 * by spacing
-						 */
 	dev_info_t		disks[0];
+	struct rcu_head		rcu;
 };
 
 
 typedef struct linear_private_data linear_conf_t;
 
-#define mddev_to_conf(mddev) ((linear_conf_t *) mddev->private)
-
 #endif

drivers/md/md.c

@@ -440,15 +440,6 @@ static inline sector_t calc_dev_sboffset(struct block_device *bdev)
 	return MD_NEW_SIZE_SECTORS(num_sectors);
 }
 
-static sector_t calc_num_sectors(mdk_rdev_t *rdev, unsigned chunk_size)
-{
-	sector_t num_sectors = rdev->sb_start;
-
-	if (chunk_size)
-		num_sectors &= ~((sector_t)chunk_size/512 - 1);
-	return num_sectors;
-}
-
 static int alloc_disk_sb(mdk_rdev_t * rdev)
 {
 	if (rdev->sb_page)
@@ -744,6 +735,24 @@ struct super_type  {
 						sector_t num_sectors);
 };
 
+/*
+ * Check that the given mddev has no bitmap.
+ *
+ * This function is called from the run method of all personalities that do not
+ * support bitmaps. It prints an error message and returns non-zero if mddev
+ * has a bitmap. Otherwise, it returns 0.
+ *
+ */
+int md_check_no_bitmap(mddev_t *mddev)
+{
+	if (!mddev->bitmap_file && !mddev->bitmap_offset)
+		return 0;
+	printk(KERN_ERR "%s: bitmaps are not supported for %s\n",
+		mdname(mddev), mddev->pers->name);
+	return 1;
+}
+EXPORT_SYMBOL(md_check_no_bitmap);
+
 /*
  * load_super for 0.90.0 
  */
@@ -797,17 +806,6 @@ static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version
 	rdev->data_offset = 0;
 	rdev->sb_size = MD_SB_BYTES;
 
-	if (sb->state & (1<<MD_SB_BITMAP_PRESENT)) {
-		if (sb->level != 1 && sb->level != 4
-		    && sb->level != 5 && sb->level != 6
-		    && sb->level != 10) {
-			/* FIXME use a better test */
-			printk(KERN_WARNING
-			       "md: bitmaps not supported for this level.\n");
-			goto abort;
-		}
-	}
-
 	if (sb->level == LEVEL_MULTIPATH)
 		rdev->desc_nr = -1;
 	else
@@ -836,7 +834,7 @@ static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version
 		else 
 			ret = 0;
 	}
-	rdev->sectors = calc_num_sectors(rdev, sb->chunk_size);
+	rdev->sectors = rdev->sb_start;
 
 	if (rdev->sectors < sb->size * 2 && sb->level > 1)
 		/* "this cannot possibly happen" ... */
@@ -866,7 +864,7 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 		mddev->minor_version = sb->minor_version;
 		mddev->patch_version = sb->patch_version;
 		mddev->external = 0;
-		mddev->chunk_size = sb->chunk_size;
+		mddev->chunk_sectors = sb->chunk_size >> 9;
 		mddev->ctime = sb->ctime;
 		mddev->utime = sb->utime;
 		mddev->level = sb->level;
@@ -883,13 +881,13 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 			mddev->delta_disks = sb->delta_disks;
 			mddev->new_level = sb->new_level;
 			mddev->new_layout = sb->new_layout;
-			mddev->new_chunk = sb->new_chunk;
+			mddev->new_chunk_sectors = sb->new_chunk >> 9;
 		} else {
 			mddev->reshape_position = MaxSector;
 			mddev->delta_disks = 0;
 			mddev->new_level = mddev->level;
 			mddev->new_layout = mddev->layout;
-			mddev->new_chunk = mddev->chunk_size;
+			mddev->new_chunk_sectors = mddev->chunk_sectors;
 		}
 
 		if (sb->state & (1<<MD_SB_CLEAN))
@@ -1004,7 +1002,7 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 		sb->new_level = mddev->new_level;
 		sb->delta_disks = mddev->delta_disks;
 		sb->new_layout = mddev->new_layout;
-		sb->new_chunk = mddev->new_chunk;
+		sb->new_chunk = mddev->new_chunk_sectors << 9;
 	}
 	mddev->minor_version = sb->minor_version;
 	if (mddev->in_sync)
@@ -1018,7 +1016,7 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 		sb->recovery_cp = 0;
 
 	sb->layout = mddev->layout;
-	sb->chunk_size = mddev->chunk_size;
+	sb->chunk_size = mddev->chunk_sectors << 9;
 
 	if (mddev->bitmap && mddev->bitmap_file == NULL)
 		sb->state |= (1<<MD_SB_BITMAP_PRESENT);
@@ -1185,17 +1183,6 @@ static int super_1_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version)
 		       bdevname(rdev->bdev,b));
 		return -EINVAL;
 	}
-	if ((le32_to_cpu(sb->feature_map) & MD_FEATURE_BITMAP_OFFSET)) {
-		if (sb->level != cpu_to_le32(1) &&
-		    sb->level != cpu_to_le32(4) &&
-		    sb->level != cpu_to_le32(5) &&
-		    sb->level != cpu_to_le32(6) &&
-		    sb->level != cpu_to_le32(10)) {
-			printk(KERN_WARNING
-			       "md: bitmaps not supported for this level.\n");
-			return -EINVAL;
-		}
-	}
 
 	rdev->preferred_minor = 0xffff;
 	rdev->data_offset = le64_to_cpu(sb->data_offset);
@@ -1248,9 +1235,6 @@ static int super_1_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version)
 	if (rdev->sectors < le64_to_cpu(sb->data_size))
 		return -EINVAL;
 	rdev->sectors = le64_to_cpu(sb->data_size);
-	if (le32_to_cpu(sb->chunksize))
-		rdev->sectors &= ~((sector_t)le32_to_cpu(sb->chunksize) - 1);
-
 	if (le64_to_cpu(sb->size) > rdev->sectors)
 		return -EINVAL;
 	return ret;
@@ -1271,7 +1255,7 @@ static int super_1_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 		mddev->major_version = 1;
 		mddev->patch_version = 0;
 		mddev->external = 0;
-		mddev->chunk_size = le32_to_cpu(sb->chunksize) << 9;
+		mddev->chunk_sectors = le32_to_cpu(sb->chunksize);
 		mddev->ctime = le64_to_cpu(sb->ctime) & ((1ULL << 32)-1);
 		mddev->utime = le64_to_cpu(sb->utime) & ((1ULL << 32)-1);
 		mddev->level = le32_to_cpu(sb->level);
@@ -1297,13 +1281,13 @@ static int super_1_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 			mddev->delta_disks = le32_to_cpu(sb->delta_disks);
 			mddev->new_level = le32_to_cpu(sb->new_level);
 			mddev->new_layout = le32_to_cpu(sb->new_layout);
-			mddev->new_chunk = le32_to_cpu(sb->new_chunk)<<9;
+			mddev->new_chunk_sectors = le32_to_cpu(sb->new_chunk);
 		} else {
 			mddev->reshape_position = MaxSector;
 			mddev->delta_disks = 0;
 			mddev->new_level = mddev->level;
 			mddev->new_layout = mddev->layout;
-			mddev->new_chunk = mddev->chunk_size;
+			mddev->new_chunk_sectors = mddev->chunk_sectors;
 		}
 
 	} else if (mddev->pers == NULL) {
@@ -1375,7 +1359,7 @@ static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 
 	sb->raid_disks = cpu_to_le32(mddev->raid_disks);
 	sb->size = cpu_to_le64(mddev->dev_sectors);
-	sb->chunksize = cpu_to_le32(mddev->chunk_size >> 9);
+	sb->chunksize = cpu_to_le32(mddev->chunk_sectors);
 	sb->level = cpu_to_le32(mddev->level);
 	sb->layout = cpu_to_le32(mddev->layout);
 
@@ -1402,7 +1386,7 @@ static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 		sb->new_layout = cpu_to_le32(mddev->new_layout);
 		sb->delta_disks = cpu_to_le32(mddev->delta_disks);
 		sb->new_level = cpu_to_le32(mddev->new_level);
-		sb->new_chunk = cpu_to_le32(mddev->new_chunk>>9);
+		sb->new_chunk = cpu_to_le32(mddev->new_chunk_sectors);
 	}
 
 	max_dev = 0;
@@ -1897,6 +1881,7 @@ static void md_update_sb(mddev_t * mddev, int force_change)
 	int sync_req;
 	int nospares = 0;
 
+	mddev->utime = get_seconds();
 	if (mddev->external)
 		return;
 repeat:
@@ -1926,7 +1911,6 @@ repeat:
 		nospares = 0;
 
 	sync_req = mddev->in_sync;
-	mddev->utime = get_seconds();
 
 	/* If this is just a dirty<->clean transition, and the array is clean
 	 * and 'events' is odd, we can roll back to the previous clean state */
@@ -2597,15 +2581,6 @@ static void analyze_sbs(mddev_t * mddev)
 			clear_bit(In_sync, &rdev->flags);
 		}
 	}
-
-
-
-	if (mddev->recovery_cp != MaxSector &&
-	    mddev->level >= 1)
-		printk(KERN_ERR "md: %s: raid array is not clean"
-		       " -- starting background reconstruction\n",
-		       mdname(mddev));
-
 }
 
 static void md_safemode_timeout(unsigned long data);
@@ -2746,7 +2721,7 @@ level_store(mddev_t *mddev, const char *buf, size_t len)
 	if (IS_ERR(priv)) {
 		mddev->new_level = mddev->level;
 		mddev->new_layout = mddev->layout;
-		mddev->new_chunk = mddev->chunk_size;
+		mddev->new_chunk_sectors = mddev->chunk_sectors;
 		mddev->raid_disks -= mddev->delta_disks;
 		mddev->delta_disks = 0;
 		module_put(pers->owner);
@@ -2764,7 +2739,7 @@ level_store(mddev_t *mddev, const char *buf, size_t len)
 	strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
 	mddev->level = mddev->new_level;
 	mddev->layout = mddev->new_layout;
-	mddev->chunk_size = mddev->new_chunk;
+	mddev->chunk_sectors = mddev->new_chunk_sectors;
 	mddev->delta_disks = 0;
 	pers->run(mddev);
 	mddev_resume(mddev);
@@ -2800,11 +2775,14 @@ layout_store(mddev_t *mddev, const char *buf, size_t len)
 
 	if (mddev->pers) {
 		int err;
-		if (mddev->pers->reconfig == NULL)
+		if (mddev->pers->check_reshape == NULL)
 			return -EBUSY;
-		err = mddev->pers->reconfig(mddev, n, -1);
-		if (err)
+		mddev->new_layout = n;
+		err = mddev->pers->check_reshape(mddev);
+		if (err) {
+			mddev->new_layout = mddev->layout;
 			return err;
+		}
 	} else {
 		mddev->new_layout = n;
 		if (mddev->reshape_position == MaxSector)
@@ -2857,10 +2835,11 @@ static ssize_t
 chunk_size_show(mddev_t *mddev, char *page)
 {
 	if (mddev->reshape_position != MaxSector &&
-	    mddev->chunk_size != mddev->new_chunk)
-		return sprintf(page, "%d (%d)\n", mddev->new_chunk,
-			       mddev->chunk_size);
-	return sprintf(page, "%d\n", mddev->chunk_size);
+	    mddev->chunk_sectors != mddev->new_chunk_sectors)
+		return sprintf(page, "%d (%d)\n",
+			       mddev->new_chunk_sectors << 9,
+			       mddev->chunk_sectors << 9);
+	return sprintf(page, "%d\n", mddev->chunk_sectors << 9);
 }
 
 static ssize_t
@@ -2874,15 +2853,18 @@ chunk_size_store(mddev_t *mddev, const char *buf, size_t len)
 
 	if (mddev->pers) {
 		int err;
-		if (mddev->pers->reconfig == NULL)
+		if (mddev->pers->check_reshape == NULL)
 			return -EBUSY;
-		err = mddev->pers->reconfig(mddev, -1, n);
-		if (err)
+		mddev->new_chunk_sectors = n >> 9;
+		err = mddev->pers->check_reshape(mddev);
+		if (err) {
+			mddev->new_chunk_sectors = mddev->chunk_sectors;
 			return err;
+		}
 	} else {
-		mddev->new_chunk = n;
+		mddev->new_chunk_sectors = n >> 9;
 		if (mddev->reshape_position == MaxSector)
-			mddev->chunk_size = n;
+			mddev->chunk_sectors = n >> 9;
 	}
 	return len;
 }
@@ -3527,8 +3509,9 @@ min_sync_store(mddev_t *mddev, const char *buf, size_t len)
 		return -EBUSY;
 
 	/* Must be a multiple of chunk_size */
-	if (mddev->chunk_size) {
-		if (min & (sector_t)((mddev->chunk_size>>9)-1))
+	if (mddev->chunk_sectors) {
+		sector_t temp = min;
+		if (sector_div(temp, mddev->chunk_sectors))
 			return -EINVAL;
 	}
 	mddev->resync_min = min;
@@ -3564,8 +3547,9 @@ max_sync_store(mddev_t *mddev, const char *buf, size_t len)
 			return -EBUSY;
 
 		/* Must be a multiple of chunk_size */
-		if (mddev->chunk_size) {
-			if (max & (sector_t)((mddev->chunk_size>>9)-1))
+		if (mddev->chunk_sectors) {
+			sector_t temp = max;
+			if (sector_div(temp, mddev->chunk_sectors))
 				return -EINVAL;
 		}
 		mddev->resync_max = max;
@@ -3656,7 +3640,7 @@ reshape_position_store(mddev_t *mddev, const char *buf, size_t len)
 	mddev->delta_disks = 0;
 	mddev->new_level = mddev->level;
 	mddev->new_layout = mddev->layout;
-	mddev->new_chunk = mddev->chunk_size;
+	mddev->new_chunk_sectors = mddev->chunk_sectors;
 	return len;
 }
 
@@ -3976,11 +3960,9 @@ static int start_dirty_degraded;
 static int do_md_run(mddev_t * mddev)
 {
 	int err;
-	int chunk_size;
 	mdk_rdev_t *rdev;
 	struct gendisk *disk;
 	struct mdk_personality *pers;
-	char b[BDEVNAME_SIZE];
 
 	if (list_empty(&mddev->disks))
 		/* cannot run an array with no devices.. */
@@ -3998,38 +3980,6 @@ static int do_md_run(mddev_t * mddev)
 		analyze_sbs(mddev);
 	}
 
-	chunk_size = mddev->chunk_size;
-
-	if (chunk_size) {
-		if (chunk_size > MAX_CHUNK_SIZE) {
-			printk(KERN_ERR "too big chunk_size: %d > %d\n",
-				chunk_size, MAX_CHUNK_SIZE);
-			return -EINVAL;
-		}
-		/*
-		 * chunk-size has to be a power of 2
-		 */
-		if ( (1 << ffz(~chunk_size)) != chunk_size) {
-			printk(KERN_ERR "chunk_size of %d not valid\n", chunk_size);
-			return -EINVAL;
-		}
-
-		/* devices must have minimum size of one chunk */
-		list_for_each_entry(rdev, &mddev->disks, same_set) {
-			if (test_bit(Faulty, &rdev->flags))
-				continue;
-			if (rdev->sectors < chunk_size / 512) {
-				printk(KERN_WARNING
-					"md: Dev %s smaller than chunk_size:"
-					" %llu < %d\n",
-					bdevname(rdev->bdev,b),
-					(unsigned long long)rdev->sectors,
-					chunk_size / 512);
-				return -EINVAL;
-			}
-		}
-	}
-
 	if (mddev->level != LEVEL_NONE)
 		request_module("md-level-%d", mddev->level);
 	else if (mddev->clevel[0])
@@ -4405,7 +4355,7 @@ static int do_md_stop(mddev_t * mddev, int mode, int is_open)
 		mddev->flags = 0;
 		mddev->ro = 0;
 		mddev->metadata_type[0] = 0;
-		mddev->chunk_size = 0;
+		mddev->chunk_sectors = 0;
 		mddev->ctime = mddev->utime = 0;
 		mddev->layout = 0;
 		mddev->max_disks = 0;
@@ -4413,7 +4363,7 @@ static int do_md_stop(mddev_t * mddev, int mode, int is_open)
 		mddev->delta_disks = 0;
 		mddev->new_level = LEVEL_NONE;
 		mddev->new_layout = 0;
-		mddev->new_chunk = 0;
+		mddev->new_chunk_sectors = 0;
 		mddev->curr_resync = 0;
 		mddev->resync_mismatches = 0;
 		mddev->suspend_lo = mddev->suspend_hi = 0;
@@ -4618,7 +4568,7 @@ static int get_array_info(mddev_t * mddev, void __user * arg)
 	info.spare_disks   = spare;
 
 	info.layout        = mddev->layout;
-	info.chunk_size    = mddev->chunk_size;
+	info.chunk_size    = mddev->chunk_sectors << 9;
 
 	if (copy_to_user(arg, &info, sizeof(info)))
 		return -EFAULT;
@@ -4843,7 +4793,7 @@ static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 			rdev->sb_start = rdev->bdev->bd_inode->i_size / 512;
 		} else 
 			rdev->sb_start = calc_dev_sboffset(rdev->bdev);
-		rdev->sectors = calc_num_sectors(rdev, mddev->chunk_size);
+		rdev->sectors = rdev->sb_start;
 
 		err = bind_rdev_to_array(rdev, mddev);
 		if (err) {
@@ -4913,7 +4863,7 @@ static int hot_add_disk(mddev_t * mddev, dev_t dev)
 	else
 		rdev->sb_start = rdev->bdev->bd_inode->i_size / 512;
 
-	rdev->sectors = calc_num_sectors(rdev, mddev->chunk_size);
+	rdev->sectors = rdev->sb_start;
 
 	if (test_bit(Faulty, &rdev->flags)) {
 		printk(KERN_WARNING 
@@ -5062,7 +5012,7 @@ static int set_array_info(mddev_t * mddev, mdu_array_info_t *info)
 	mddev->external	     = 0;
 
 	mddev->layout        = info->layout;
-	mddev->chunk_size    = info->chunk_size;
+	mddev->chunk_sectors = info->chunk_size >> 9;
 
 	mddev->max_disks     = MD_SB_DISKS;
 
@@ -5081,7 +5031,7 @@ static int set_array_info(mddev_t * mddev, mdu_array_info_t *info)
 	get_random_bytes(mddev->uuid, 16);
 
 	mddev->new_level = mddev->level;
-	mddev->new_chunk = mddev->chunk_size;
+	mddev->new_chunk_sectors = mddev->chunk_sectors;
 	mddev->new_layout = mddev->layout;
 	mddev->delta_disks = 0;
 
@@ -5191,7 +5141,7 @@ static int update_array_info(mddev_t *mddev, mdu_array_info_t *info)
 	    mddev->level         != info->level         ||
 /*	    mddev->layout        != info->layout        || */
 	    !mddev->persistent	 != info->not_persistent||
-	    mddev->chunk_size    != info->chunk_size    ||
+	    mddev->chunk_sectors != info->chunk_size >> 9 ||
 	    /* ignore bottom 8 bits of state, and allow SB_BITMAP_PRESENT to change */
 	    ((state^info->state) & 0xfffffe00)
 		)
@@ -5215,10 +5165,15 @@ static int update_array_info(mddev_t *mddev, mdu_array_info_t *info)
 		 * we don't need to do anything at the md level, the
 		 * personality will take care of it all.
 		 */
-		if (mddev->pers->reconfig == NULL)
+		if (mddev->pers->check_reshape == NULL)
 			return -EINVAL;
-		else
-			return mddev->pers->reconfig(mddev, info->layout, -1);
+		else {
+			mddev->new_layout = info->layout;
+			rv = mddev->pers->check_reshape(mddev);
+			if (rv)
+				mddev->new_layout = mddev->layout;
+			return rv;
+		}
 	}
 	if (info->size >= 0 && mddev->dev_sectors / 2 != info->size)
 		rv = update_size(mddev, (sector_t)info->size * 2);
@@ -6717,7 +6672,8 @@ void md_check_recovery(mddev_t *mddev)
 		 */
 
 		if (mddev->reshape_position != MaxSector) {
-			if (mddev->pers->check_reshape(mddev) != 0)
+			if (mddev->pers->check_reshape == NULL ||
+			    mddev->pers->check_reshape(mddev) != 0)
 				/* Cannot proceed */
 				goto unlock;
 			set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);

drivers/md/md.h

@@ -29,13 +29,6 @@
 typedef struct mddev_s mddev_t;
 typedef struct mdk_rdev_s mdk_rdev_t;
 
-/*
- * options passed in raidrun:
- */
-
-/* Currently this must fit in an 'int' */
-#define MAX_CHUNK_SIZE (1<<30)
-
 /*
  * MD's 'extended' device
  */
@@ -145,7 +138,7 @@ struct mddev_s
 	int 				external;	/* metadata is
 							 * managed externally */
 	char				metadata_type[17]; /* externally set*/
-	int				chunk_size;
+	int				chunk_sectors;
 	time_t				ctime, utime;
 	int				level, layout;
 	char				clevel[16];
@@ -166,7 +159,8 @@ struct mddev_s
 	 * If reshape_position is MaxSector, then no reshape is happening (yet).
 	 */
 	sector_t			reshape_position;
-	int				delta_disks, new_level, new_layout, new_chunk;
+	int				delta_disks, new_level, new_layout;
+	int				new_chunk_sectors;
 
 	struct mdk_thread_s		*thread;	/* management thread */
 	struct mdk_thread_s		*sync_thread;	/* doing resync or reconstruct */
@@ -325,7 +319,6 @@ struct mdk_personality
 	int (*check_reshape) (mddev_t *mddev);
 	int (*start_reshape) (mddev_t *mddev);
 	void (*finish_reshape) (mddev_t *mddev);
-	int (*reconfig) (mddev_t *mddev, int layout, int chunk_size);
 	/* quiesce moves between quiescence states
 	 * 0 - fully active
 	 * 1 - no new requests allowed
@@ -437,5 +430,6 @@ extern void md_new_event(mddev_t *mddev);
 extern int md_allow_write(mddev_t *mddev);
 extern void md_wait_for_blocked_rdev(mdk_rdev_t *rdev, mddev_t *mddev);
 extern void md_set_array_sectors(mddev_t *mddev, sector_t array_sectors);
+extern int md_check_no_bitmap(mddev_t *mddev);
 
 #endif /* _MD_MD_H */

drivers/md/multipath.c

@@ -58,7 +58,7 @@ static void multipath_reschedule_retry (struct multipath_bh *mp_bh)
 {
 	unsigned long flags;
 	mddev_t *mddev = mp_bh->mddev;
-	multipath_conf_t *conf = mddev_to_conf(mddev);
+	multipath_conf_t *conf = mddev->private;
 
 	spin_lock_irqsave(&conf->device_lock, flags);
 	list_add(&mp_bh->retry_list, &conf->retry_list);
@@ -75,7 +75,7 @@ static void multipath_reschedule_retry (struct multipath_bh *mp_bh)
 static void multipath_end_bh_io (struct multipath_bh *mp_bh, int err)
 {
 	struct bio *bio = mp_bh->master_bio;
-	multipath_conf_t *conf = mddev_to_conf(mp_bh->mddev);
+	multipath_conf_t *conf = mp_bh->mddev->private;
 
 	bio_endio(bio, err);
 	mempool_free(mp_bh, conf->pool);
@@ -85,7 +85,7 @@ static void multipath_end_request(struct bio *bio, int error)
 {
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	struct multipath_bh * mp_bh = (struct multipath_bh *)(bio->bi_private);
-	multipath_conf_t *conf = mddev_to_conf(mp_bh->mddev);
+	multipath_conf_t *conf = mp_bh->mddev->private;
 	mdk_rdev_t *rdev = conf->multipaths[mp_bh->path].rdev;
 
 	if (uptodate)
@@ -107,7 +107,7 @@ static void multipath_end_request(struct bio *bio, int error)
 
 static void unplug_slaves(mddev_t *mddev)
 {
-	multipath_conf_t *conf = mddev_to_conf(mddev);
+	multipath_conf_t *conf = mddev->private;
 	int i;
 
 	rcu_read_lock();
@@ -138,7 +138,7 @@ static void multipath_unplug(struct request_queue *q)
 static int multipath_make_request (struct request_queue *q, struct bio * bio)
 {
 	mddev_t *mddev = q->queuedata;
-	multipath_conf_t *conf = mddev_to_conf(mddev);
+	multipath_conf_t *conf = mddev->private;
 	struct multipath_bh * mp_bh;
 	struct multipath_info *multipath;
 	const int rw = bio_data_dir(bio);
@@ -180,7 +180,7 @@ static int multipath_make_request (struct request_queue *q, struct bio * bio)
 
 static void multipath_status (struct seq_file *seq, mddev_t *mddev)
 {
-	multipath_conf_t *conf = mddev_to_conf(mddev);
+	multipath_conf_t *conf = mddev->private;
 	int i;
 	
 	seq_printf (seq, " [%d/%d] [", conf->raid_disks,
@@ -195,7 +195,7 @@ static void multipath_status (struct seq_file *seq, mddev_t *mddev)
 static int multipath_congested(void *data, int bits)
 {
 	mddev_t *mddev = data;
-	multipath_conf_t *conf = mddev_to_conf(mddev);
+	multipath_conf_t *conf = mddev->private;
 	int i, ret = 0;
 
 	rcu_read_lock();
@@ -220,7 +220,7 @@ static int multipath_congested(void *data, int bits)
  */
 static void multipath_error (mddev_t *mddev, mdk_rdev_t *rdev)
 {
-	multipath_conf_t *conf = mddev_to_conf(mddev);
+	multipath_conf_t *conf = mddev->private;
 
 	if (conf->working_disks <= 1) {
 		/*
@@ -367,7 +367,7 @@ static void multipathd (mddev_t *mddev)
 	struct multipath_bh *mp_bh;
 	struct bio *bio;
 	unsigned long flags;
-	multipath_conf_t *conf = mddev_to_conf(mddev);
+	multipath_conf_t *conf = mddev->private;
 	struct list_head *head = &conf->retry_list;
 
 	md_check_recovery(mddev);
@@ -421,6 +421,9 @@ static int multipath_run (mddev_t *mddev)
 	struct multipath_info *disk;
 	mdk_rdev_t *rdev;
 
+	if (md_check_no_bitmap(mddev))
+		return -EINVAL;
+
 	if (mddev->level != LEVEL_MULTIPATH) {
 		printk("multipath: %s: raid level not set to multipath IO (%d)\n",
 		       mdname(mddev), mddev->level);
@@ -531,7 +534,7 @@ out:
 
 static int multipath_stop (mddev_t *mddev)
 {
-	multipath_conf_t *conf = mddev_to_conf(mddev);
+	multipath_conf_t *conf = mddev->private;
 
 	md_unregister_thread(mddev->thread);
 	mddev->thread = NULL;

drivers/md/multipath.h

@@ -18,12 +18,6 @@ struct multipath_private_data {
 
 typedef struct multipath_private_data multipath_conf_t;
 
-/*
- * this is the only point in the RAID code where we violate
- * C type safety. mddev->private is an 'opaque' pointer.
- */
-#define mddev_to_conf(mddev) ((multipath_conf_t *) mddev->private)
-
 /*
  * this is our 'private' 'collective' MULTIPATH buffer head.
  * it contains information about what kind of IO operations were started

drivers/md/raid0.c

@@ -26,8 +26,8 @@
 static void raid0_unplug(struct request_queue *q)
 {
 	mddev_t *mddev = q->queuedata;
-	raid0_conf_t *conf = mddev_to_conf(mddev);
-	mdk_rdev_t **devlist = conf->strip_zone[0].dev;
+	raid0_conf_t *conf = mddev->private;
+	mdk_rdev_t **devlist = conf->devlist;
 	int i;
 
 	for (i=0; i<mddev->raid_disks; i++) {
@@ -40,8 +40,8 @@ static void raid0_unplug(struct request_queue *q)
 static int raid0_congested(void *data, int bits)
 {
 	mddev_t *mddev = data;
-	raid0_conf_t *conf = mddev_to_conf(mddev);
-	mdk_rdev_t **devlist = conf->strip_zone[0].dev;
+	raid0_conf_t *conf = mddev->private;
+	mdk_rdev_t **devlist = conf->devlist;
 	int i, ret = 0;
 
 	for (i = 0; i < mddev->raid_disks && !ret ; i++) {
@@ -52,27 +52,60 @@ static int raid0_congested(void *data, int bits)
 	return ret;
 }
 
+/*
+ * inform the user of the raid configuration
+*/
+static void dump_zones(mddev_t *mddev)
+{
+	int j, k, h;
+	sector_t zone_size = 0;
+	sector_t zone_start = 0;
+	char b[BDEVNAME_SIZE];
+	raid0_conf_t *conf = mddev->private;
+	printk(KERN_INFO "******* %s configuration *********\n",
+		mdname(mddev));
+	h = 0;
+	for (j = 0; j < conf->nr_strip_zones; j++) {
+		printk(KERN_INFO "zone%d=[", j);
+		for (k = 0; k < conf->strip_zone[j].nb_dev; k++)
+			printk("%s/",
+			bdevname(conf->devlist[j*mddev->raid_disks
+						+ k]->bdev, b));
+		printk("]\n");
+
+		zone_size  = conf->strip_zone[j].zone_end - zone_start;
+		printk(KERN_INFO "        zone offset=%llukb "
+				"device offset=%llukb size=%llukb\n",
+			(unsigned long long)zone_start>>1,
+			(unsigned long long)conf->strip_zone[j].dev_start>>1,
+			(unsigned long long)zone_size>>1);
+		zone_start = conf->strip_zone[j].zone_end;
+	}
+	printk(KERN_INFO "**********************************\n\n");
+}
 
-static int create_strip_zones (mddev_t *mddev)
+static int create_strip_zones(mddev_t *mddev)
 {
-	int i, c, j;
-	sector_t current_start, curr_zone_start;
-	sector_t min_spacing;
-	raid0_conf_t *conf = mddev_to_conf(mddev);
-	mdk_rdev_t *smallest, *rdev1, *rdev2, *rdev;
+	int i, c, j, err;
+	sector_t curr_zone_end, sectors;
+	mdk_rdev_t *smallest, *rdev1, *rdev2, *rdev, **dev;
 	struct strip_zone *zone;
 	int cnt;
 	char b[BDEVNAME_SIZE];
- 
-	/*
-	 * The number of 'same size groups'
-	 */
-	conf->nr_strip_zones = 0;
- 
+	raid0_conf_t *conf = kzalloc(sizeof(*conf), GFP_KERNEL);
+
+	if (!conf)
+		return -ENOMEM;
 	list_for_each_entry(rdev1, &mddev->disks, same_set) {
 		printk(KERN_INFO "raid0: looking at %s\n",
 			bdevname(rdev1->bdev,b));
 		c = 0;
+
+		/* round size to chunk_size */
+		sectors = rdev1->sectors;
+		sector_div(sectors, mddev->chunk_sectors);
+		rdev1->sectors = sectors * mddev->chunk_sectors;
+
 		list_for_each_entry(rdev2, &mddev->disks, same_set) {
 			printk(KERN_INFO "raid0:   comparing %s(%llu)",
 			       bdevname(rdev1->bdev,b),
@@ -103,16 +136,16 @@ static int create_strip_zones (mddev_t *mddev)
 		}
 	}
 	printk(KERN_INFO "raid0: FINAL %d zones\n", conf->nr_strip_zones);
-
+	err = -ENOMEM;
 	conf->strip_zone = kzalloc(sizeof(struct strip_zone)*
 				conf->nr_strip_zones, GFP_KERNEL);
 	if (!conf->strip_zone)
-		return 1;
+		goto abort;
 	conf->devlist = kzalloc(sizeof(mdk_rdev_t*)*
 				conf->nr_strip_zones*mddev->raid_disks,
 				GFP_KERNEL);
 	if (!conf->devlist)
-		return 1;
+		goto abort;
 
 	/* The first zone must contain all devices, so here we check that
 	 * there is a proper alignment of slots to devices and find them all
@@ -120,7 +153,8 @@ static int create_strip_zones (mddev_t *mddev)
 	zone = &conf->strip_zone[0];
 	cnt = 0;
 	smallest = NULL;
-	zone->dev = conf->devlist;
+	dev = conf->devlist;
+	err = -EINVAL;
 	list_for_each_entry(rdev1, &mddev->disks, same_set) {
 		int j = rdev1->raid_disk;
 
@@ -129,12 +163,12 @@ static int create_strip_zones (mddev_t *mddev)
 				"aborting!\n", j);
 			goto abort;
 		}
-		if (zone->dev[j]) {
+		if (dev[j]) {
 			printk(KERN_ERR "raid0: multiple devices for %d - "
 				"aborting!\n", j);
 			goto abort;
 		}
-		zone->dev[j] = rdev1;
+		dev[j] = rdev1;
 
 		blk_queue_stack_limits(mddev->queue,
 				       rdev1->bdev->bd_disk->queue);
@@ -157,34 +191,32 @@ static int create_strip_zones (mddev_t *mddev)
 		goto abort;
 	}
 	zone->nb_dev = cnt;
-	zone->sectors = smallest->sectors * cnt;
-	zone->zone_start = 0;
+	zone->zone_end = smallest->sectors * cnt;
 
-	current_start = smallest->sectors;
-	curr_zone_start = zone->sectors;
+	curr_zone_end = zone->zone_end;
 
 	/* now do the other zones */
 	for (i = 1; i < conf->nr_strip_zones; i++)
 	{
 		zone = conf->strip_zone + i;
-		zone->dev = conf->strip_zone[i-1].dev + mddev->raid_disks;
+		dev = conf->devlist + i * mddev->raid_disks;
 
 		printk(KERN_INFO "raid0: zone %d\n", i);
-		zone->dev_start = current_start;
+		zone->dev_start = smallest->sectors;
 		smallest = NULL;
 		c = 0;
 
 		for (j=0; j<cnt; j++) {
 			char b[BDEVNAME_SIZE];
-			rdev = conf->strip_zone[0].dev[j];
+			rdev = conf->devlist[j];
 			printk(KERN_INFO "raid0: checking %s ...",
 				bdevname(rdev->bdev, b));
-			if (rdev->sectors <= current_start) {
+			if (rdev->sectors <= zone->dev_start) {
 				printk(KERN_INFO " nope.\n");
 				continue;
 			}
 			printk(KERN_INFO " contained as device %d\n", c);
-			zone->dev[c] = rdev;
+			dev[c] = rdev;
 			c++;
 			if (!smallest || rdev->sectors < smallest->sectors) {
 				smallest = rdev;
@@ -194,47 +226,39 @@ static int create_strip_zones (mddev_t *mddev)
 		}
 
 		zone->nb_dev = c;
-		zone->sectors = (smallest->sectors - current_start) * c;
+		sectors = (smallest->sectors - zone->dev_start) * c;
 		printk(KERN_INFO "raid0: zone->nb_dev: %d, sectors: %llu\n",
-			zone->nb_dev, (unsigned long long)zone->sectors);
+			zone->nb_dev, (unsigned long long)sectors);
 
-		zone->zone_start = curr_zone_start;
-		curr_zone_start += zone->sectors;
+		curr_zone_end += sectors;
+		zone->zone_end = curr_zone_end;
 
-		current_start = smallest->sectors;
 		printk(KERN_INFO "raid0: current zone start: %llu\n",
-			(unsigned long long)current_start);
-	}
-
-	/* Now find appropriate hash spacing.
-	 * We want a number which causes most hash entries to cover
-	 * at most two strips, but the hash table must be at most
-	 * 1 PAGE.  We choose the smallest strip, or contiguous collection
-	 * of strips, that has big enough size.  We never consider the last
-	 * strip though as it's size has no bearing on the efficacy of the hash
-	 * table.
-	 */
-	conf->spacing = curr_zone_start;
-	min_spacing = curr_zone_start;
-	sector_div(min_spacing, PAGE_SIZE/sizeof(struct strip_zone*));
-	for (i=0; i < conf->nr_strip_zones-1; i++) {
-		sector_t s = 0;
-		for (j = i; j < conf->nr_strip_zones - 1 &&
-				s < min_spacing; j++)
-			s += conf->strip_zone[j].sectors;
-		if (s >= min_spacing && s < conf->spacing)
-			conf->spacing = s;
+			(unsigned long long)smallest->sectors);
 	}
-
 	mddev->queue->unplug_fn = raid0_unplug;
-
 	mddev->queue->backing_dev_info.congested_fn = raid0_congested;
 	mddev->queue->backing_dev_info.congested_data = mddev;
 
+	/*
+	 * 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 "%s chunk_size of %d not valid\n",
+		       mdname(mddev),
+		       mddev->chunk_sectors << 9);
+		goto abort;
+	}
 	printk(KERN_INFO "raid0: done.\n");
+	mddev->private = conf;
 	return 0;
- abort:
-	return 1;
+abort:
+	kfree(conf->strip_zone);
+	kfree(conf->devlist);
+	kfree(conf);
+	mddev->private = NULL;
+	return err;
 }
 
 /**
@@ -252,10 +276,15 @@ static int raid0_mergeable_bvec(struct request_queue *q,
 	mddev_t *mddev = q->queuedata;
 	sector_t sector = bvm->bi_sector + get_start_sect(bvm->bi_bdev);
 	int max;
-	unsigned int chunk_sectors = mddev->chunk_size >> 9;
+	unsigned int chunk_sectors = mddev->chunk_sectors;
 	unsigned int bio_sectors = bvm->bi_size >> 9;
 
-	max =  (chunk_sectors - ((sector & (chunk_sectors - 1)) + bio_sectors)) << 9;
+	if (is_power_of_2(chunk_sectors))
+		max =  (chunk_sectors - ((sector & (chunk_sectors-1))
+						+ bio_sectors)) << 9;
+	else
+		max =  (chunk_sectors - (sector_div(sector, chunk_sectors)
+						+ bio_sectors)) << 9;
 	if (max < 0) max = 0; /* bio_add cannot handle a negative return */
 	if (max <= biovec->bv_len && bio_sectors == 0)
 		return biovec->bv_len;
@@ -277,84 +306,28 @@ static sector_t raid0_size(mddev_t *mddev, sector_t sectors, int raid_disks)
 	return array_sectors;
 }
 
-static int raid0_run (mddev_t *mddev)
+static int raid0_run(mddev_t *mddev)
 {
-	unsigned  cur=0, i=0, nb_zone;
-	s64 sectors;
-	raid0_conf_t *conf;
+	int ret;
 
-	if (mddev->chunk_size == 0) {
-		printk(KERN_ERR "md/raid0: non-zero chunk size required.\n");
+	if (mddev->chunk_sectors == 0) {
+		printk(KERN_ERR "md/raid0: chunk size must be set.\n");
 		return -EINVAL;
 	}
-	printk(KERN_INFO "%s: setting max_sectors to %d, segment boundary to %d\n",
-	       mdname(mddev),
-	       mddev->chunk_size >> 9,
-	       (mddev->chunk_size>>1)-1);
-	blk_queue_max_sectors(mddev->queue, mddev->chunk_size >> 9);
-	blk_queue_segment_boundary(mddev->queue, (mddev->chunk_size>>1) - 1);
+	if (md_check_no_bitmap(mddev))
+		return -EINVAL;
+	blk_queue_max_sectors(mddev->queue, mddev->chunk_sectors);
 	mddev->queue->queue_lock = &mddev->queue->__queue_lock;
 
-	conf = kmalloc(sizeof (raid0_conf_t), GFP_KERNEL);
-	if (!conf)
-		goto out;
-	mddev->private = (void *)conf;
- 
-	conf->strip_zone = NULL;
-	conf->devlist = NULL;
-	if (create_strip_zones (mddev)) 
-		goto out_free_conf;
+	ret = create_strip_zones(mddev);
+	if (ret < 0)
+		return ret;
 
 	/* calculate array device size */
 	md_set_array_sectors(mddev, raid0_size(mddev, 0, 0));
 
 	printk(KERN_INFO "raid0 : md_size is %llu sectors.\n",
 		(unsigned long long)mddev->array_sectors);
-	printk(KERN_INFO "raid0 : conf->spacing is %llu sectors.\n",
-		(unsigned long long)conf->spacing);
-	{
-		sector_t s = raid0_size(mddev, 0, 0);
-		sector_t space = conf->spacing;
-		int round;
-		conf->sector_shift = 0;
-		if (sizeof(sector_t) > sizeof(u32)) {
-			/*shift down space and s so that sector_div will work */
-			while (space > (sector_t) (~(u32)0)) {
-				s >>= 1;
-				space >>= 1;
-				s += 1; /* force round-up */
-				conf->sector_shift++;
-			}
-		}
-		round = sector_div(s, (u32)space) ? 1 : 0;
-		nb_zone = s + round;
-	}
-	printk(KERN_INFO "raid0 : nb_zone is %d.\n", nb_zone);
-
-	printk(KERN_INFO "raid0 : Allocating %zu bytes for hash.\n",
-				nb_zone*sizeof(struct strip_zone*));
-	conf->hash_table = kmalloc (sizeof (struct strip_zone *)*nb_zone, GFP_KERNEL);
-	if (!conf->hash_table)
-		goto out_free_conf;
-	sectors = conf->strip_zone[cur].sectors;
-
-	conf->hash_table[0] = conf->strip_zone + cur;
-	for (i=1; i< nb_zone; i++) {
-		while (sectors <= conf->spacing) {
-			cur++;
-			sectors += conf->strip_zone[cur].sectors;
-		}
-		sectors -= conf->spacing;
-		conf->hash_table[i] = conf->strip_zone + cur;
-	}
-	if (conf->sector_shift) {
-		conf->spacing >>= conf->sector_shift;
-		/* round spacing up so when we divide by it, we
-		 * err on the side of too-low, which is safest
-		 */
-		conf->spacing++;
-	}
-
 	/* calculate the max read-ahead size.
 	 * For read-ahead of large files to be effective, we need to
 	 * readahead at least twice a whole stripe. i.e. number of devices
@@ -365,48 +338,107 @@ static int raid0_run (mddev_t *mddev)
 	 * chunksize should be used in that case.
 	 */
 	{
-		int stripe = mddev->raid_disks * mddev->chunk_size / PAGE_SIZE;
+		int stripe = mddev->raid_disks *
+			(mddev->chunk_sectors << 9) / PAGE_SIZE;
 		if (mddev->queue->backing_dev_info.ra_pages < 2* stripe)
 			mddev->queue->backing_dev_info.ra_pages = 2* stripe;
 	}
 
-
 	blk_queue_merge_bvec(mddev->queue, raid0_mergeable_bvec);
+	dump_zones(mddev);
 	return 0;
+}
 
-out_free_conf:
+static int raid0_stop(mddev_t *mddev)
+{
+	raid0_conf_t *conf = mddev->private;
+
+	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
 	kfree(conf->strip_zone);
 	kfree(conf->devlist);
 	kfree(conf);
 	mddev->private = NULL;
-out:
-	return -ENOMEM;
+	return 0;
 }
 
-static int raid0_stop (mddev_t *mddev)
+/* Find the zone which holds a particular offset
+ * Update *sectorp to be an offset in that zone
+ */
+static struct strip_zone *find_zone(struct raid0_private_data *conf,
+				    sector_t *sectorp)
 {
-	raid0_conf_t *conf = mddev_to_conf(mddev);
+	int i;
+	struct strip_zone *z = conf->strip_zone;
+	sector_t sector = *sectorp;
+
+	for (i = 0; i < conf->nr_strip_zones; i++)
+		if (sector < z[i].zone_end) {
+			if (i)
+				*sectorp = sector - z[i-1].zone_end;
+			return z + i;
+		}
+	BUG();
+}
 
-	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
-	kfree(conf->hash_table);
-	conf->hash_table = NULL;
-	kfree(conf->strip_zone);
-	conf->strip_zone = NULL;
-	kfree(conf);
-	mddev->private = NULL;
+/*
+ * remaps the bio to the target device. we separate two flows.
+ * power 2 flow and a general flow for the sake of perfromance
+*/
+static mdk_rdev_t *map_sector(mddev_t *mddev, struct strip_zone *zone,
+				sector_t sector, sector_t *sector_offset)
+{
+	unsigned int sect_in_chunk;
+	sector_t chunk;
+	raid0_conf_t *conf = mddev->private;
+	unsigned int chunk_sects = mddev->chunk_sectors;
+
+	if (is_power_of_2(chunk_sects)) {
+		int chunksect_bits = ffz(~chunk_sects);
+		/* find the sector offset inside the chunk */
+		sect_in_chunk  = sector & (chunk_sects - 1);
+		sector >>= chunksect_bits;
+		/* chunk in zone */
+		chunk = *sector_offset;
+		/* quotient is the chunk in real device*/
+		sector_div(chunk, zone->nb_dev << chunksect_bits);
+	} else{
+		sect_in_chunk = sector_div(sector, chunk_sects);
+		chunk = *sector_offset;
+		sector_div(chunk, chunk_sects * zone->nb_dev);
+	}
+	/*
+	*  position the bio over the real device
+	*  real sector = chunk in device + starting of zone
+	*	+ the position in the chunk
+	*/
+	*sector_offset = (chunk * chunk_sects) + sect_in_chunk;
+	return conf->devlist[(zone - conf->strip_zone)*mddev->raid_disks
+			     + sector_div(sector, zone->nb_dev)];
+}
 
-	return 0;
+/*
+ * Is io distribute over 1 or more chunks ?
+*/
+static inline int is_io_in_chunk_boundary(mddev_t *mddev,
+			unsigned int chunk_sects, struct bio *bio)
+{
+	if (likely(is_power_of_2(chunk_sects))) {
+		return chunk_sects >= ((bio->bi_sector & (chunk_sects-1))
+					+ (bio->bi_size >> 9));
+	} else{
+		sector_t sector = bio->bi_sector;
+		return chunk_sects >= (sector_div(sector, chunk_sects)
+						+ (bio->bi_size >> 9));
+	}
 }
 
-static int raid0_make_request (struct request_queue *q, struct bio *bio)
+static int raid0_make_request(struct request_queue *q, struct bio *bio)
 {
 	mddev_t *mddev = q->queuedata;
-	unsigned int sect_in_chunk, chunksect_bits, chunk_sects;
-	raid0_conf_t *conf = mddev_to_conf(mddev);
+	unsigned int chunk_sects;
+	sector_t sector_offset;
 	struct strip_zone *zone;
 	mdk_rdev_t *tmp_dev;
-	sector_t chunk;
-	sector_t sector, rsect;
 	const int rw = bio_data_dir(bio);
 	int cpu;
 
@@ -421,11 +453,9 @@ static int raid0_make_request (struct request_queue *q, struct bio *bio)
 		      bio_sectors(bio));
 	part_stat_unlock();
 
-	chunk_sects = mddev->chunk_size >> 9;
-	chunksect_bits = ffz(~chunk_sects);
-	sector = bio->bi_sector;
-
-	if (unlikely(chunk_sects < (bio->bi_sector & (chunk_sects - 1)) + (bio->bi_size >> 9))) {
+	chunk_sects = mddev->chunk_sectors;
+	if (unlikely(!is_io_in_chunk_boundary(mddev, chunk_sects, bio))) {
+		sector_t sector = bio->bi_sector;
 		struct bio_pair *bp;
 		/* Sanity check -- queue functions should prevent this happening */
 		if (bio->bi_vcnt != 1 ||
@@ -434,7 +464,12 @@ static int raid0_make_request (struct request_queue *q, struct bio *bio)
 		/* This is a one page bio that upper layers
 		 * refuse to split for us, so we need to split it.
 		 */
-		bp = bio_split(bio, chunk_sects - (bio->bi_sector & (chunk_sects - 1)));
+		if (likely(is_power_of_2(chunk_sects)))
+			bp = bio_split(bio, chunk_sects - (sector &
+							   (chunk_sects-1)));
+		else
+			bp = bio_split(bio, chunk_sects -
+				       sector_div(sector, chunk_sects));
 		if (raid0_make_request(q, &bp->bio1))
 			generic_make_request(&bp->bio1);
 		if (raid0_make_request(q, &bp->bio2))
@@ -443,34 +478,14 @@ static int raid0_make_request (struct request_queue *q, struct bio *bio)
 		bio_pair_release(bp);
 		return 0;
 	}
- 
-
-	{
-		sector_t x = sector >> conf->sector_shift;
-		sector_div(x, (u32)conf->spacing);
-		zone = conf->hash_table[x];
-	}
 
-	while (sector >= zone->zone_start + zone->sectors)
-		zone++;
-
-	sect_in_chunk = bio->bi_sector & (chunk_sects - 1);
-
-
-	{
-		sector_t x = (sector - zone->zone_start) >> chunksect_bits;
-
-		sector_div(x, zone->nb_dev);
-		chunk = x;
-
-		x = sector >> chunksect_bits;
-		tmp_dev = zone->dev[sector_div(x, zone->nb_dev)];
-	}
-	rsect = (chunk << chunksect_bits) + zone->dev_start + sect_in_chunk;
- 
+	sector_offset = bio->bi_sector;
+	zone =  find_zone(mddev->private, &sector_offset);
+	tmp_dev = map_sector(mddev, zone, bio->bi_sector,
+			     &sector_offset);
 	bio->bi_bdev = tmp_dev->bdev;
-	bio->bi_sector = rsect + tmp_dev->data_offset;
-
+	bio->bi_sector = sector_offset + zone->dev_start +
+		tmp_dev->data_offset;
 	/*
 	 * Let the main block layer submit the IO and resolve recursion:
 	 */
@@ -485,31 +500,35 @@ bad_map:
 	return 0;
 }
 
-static void raid0_status (struct seq_file *seq, mddev_t *mddev)
+static void raid0_status(struct seq_file *seq, mddev_t *mddev)
 {
 #undef MD_DEBUG
 #ifdef MD_DEBUG
 	int j, k, h;
 	char b[BDEVNAME_SIZE];
-	raid0_conf_t *conf = mddev_to_conf(mddev);
+	raid0_conf_t *conf = mddev->private;
 
+	sector_t zone_size;
+	sector_t zone_start = 0;
 	h = 0;
+
 	for (j = 0; j < conf->nr_strip_zones; j++) {
 		seq_printf(seq, "      z%d", j);
-		if (conf->hash_table[h] == conf->strip_zone+j)
-			seq_printf(seq, "(h%d)", h++);
 		seq_printf(seq, "=[");
 		for (k = 0; k < conf->strip_zone[j].nb_dev; k++)
 			seq_printf(seq, "%s/", bdevname(
-				conf->strip_zone[j].dev[k]->bdev,b));
-
-		seq_printf(seq, "] zs=%d ds=%d s=%d\n",
-				conf->strip_zone[j].zone_start,
-				conf->strip_zone[j].dev_start,
-				conf->strip_zone[j].sectors);
+				conf->devlist[j*mddev->raid_disks + k]
+						->bdev, b));
+
+		zone_size  = conf->strip_zone[j].zone_end - zone_start;
+		seq_printf(seq, "] ze=%lld ds=%lld s=%lld\n",
+			(unsigned long long)zone_start>>1,
+			(unsigned long long)conf->strip_zone[j].dev_start>>1,
+			(unsigned long long)zone_size>>1);
+		zone_start = conf->strip_zone[j].zone_end;
 	}
 #endif
-	seq_printf(seq, " %dk chunks", mddev->chunk_size/1024);
+	seq_printf(seq, " %dk chunks", mddev->chunk_sectors / 2);
 	return;
 }
 

drivers/md/raid0.h

@@ -3,26 +3,18 @@
 
 struct strip_zone
 {
-	sector_t zone_start;	/* Zone offset in md_dev (in sectors) */
+	sector_t zone_end;	/* Start of the next zone (in sectors) */
 	sector_t dev_start;	/* Zone offset in real dev (in sectors) */
-	sector_t sectors;	/* Zone size in sectors */
 	int nb_dev;		/* # of devices attached to the zone */
-	mdk_rdev_t **dev;	/* Devices attached to the zone */
 };
 
 struct raid0_private_data
 {
-	struct strip_zone **hash_table; /* Table of indexes into strip_zone */
 	struct strip_zone *strip_zone;
 	mdk_rdev_t **devlist; /* lists of rdevs, pointed to by strip_zone->dev */
 	int nr_strip_zones;
-
-	sector_t spacing;
-	int sector_shift; /* shift this before divide by spacing */
 };
 
 typedef struct raid0_private_data raid0_conf_t;
 
-#define mddev_to_conf(mddev) ((raid0_conf_t *) mddev->private)
-
 #endif

drivers/md/raid1.c

@@ -182,7 +182,7 @@ static void put_all_bios(conf_t *conf, r1bio_t *r1_bio)
 
 static void free_r1bio(r1bio_t *r1_bio)
 {
-	conf_t *conf = mddev_to_conf(r1_bio->mddev);
+	conf_t *conf = r1_bio->mddev->private;
 
 	/*
 	 * Wake up any possible resync thread that waits for the device
@@ -196,7 +196,7 @@ static void free_r1bio(r1bio_t *r1_bio)
 
 static void put_buf(r1bio_t *r1_bio)
 {
-	conf_t *conf = mddev_to_conf(r1_bio->mddev);
+	conf_t *conf = r1_bio->mddev->private;
 	int i;
 
 	for (i=0; i<conf->raid_disks; i++) {
@@ -214,7 +214,7 @@ static void reschedule_retry(r1bio_t *r1_bio)
 {
 	unsigned long flags;
 	mddev_t *mddev = r1_bio->mddev;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 
 	spin_lock_irqsave(&conf->device_lock, flags);
 	list_add(&r1_bio->retry_list, &conf->retry_list);
@@ -253,7 +253,7 @@ static void raid_end_bio_io(r1bio_t *r1_bio)
  */
 static inline void update_head_pos(int disk, r1bio_t *r1_bio)
 {
-	conf_t *conf = mddev_to_conf(r1_bio->mddev);
+	conf_t *conf = r1_bio->mddev->private;
 
 	conf->mirrors[disk].head_position =
 		r1_bio->sector + (r1_bio->sectors);
@@ -264,7 +264,7 @@ static void raid1_end_read_request(struct bio *bio, int error)
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	r1bio_t * r1_bio = (r1bio_t *)(bio->bi_private);
 	int mirror;
-	conf_t *conf = mddev_to_conf(r1_bio->mddev);
+	conf_t *conf = r1_bio->mddev->private;
 
 	mirror = r1_bio->read_disk;
 	/*
@@ -309,7 +309,7 @@ static void raid1_end_write_request(struct bio *bio, int error)
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	r1bio_t * r1_bio = (r1bio_t *)(bio->bi_private);
 	int mirror, behind = test_bit(R1BIO_BehindIO, &r1_bio->state);
-	conf_t *conf = mddev_to_conf(r1_bio->mddev);
+	conf_t *conf = r1_bio->mddev->private;
 	struct bio *to_put = NULL;
 
 
@@ -541,7 +541,7 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 
 static void unplug_slaves(mddev_t *mddev)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i;
 
 	rcu_read_lock();
@@ -573,7 +573,7 @@ static void raid1_unplug(struct request_queue *q)
 static int raid1_congested(void *data, int bits)
 {
 	mddev_t *mddev = data;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i, ret = 0;
 
 	rcu_read_lock();
@@ -772,7 +772,7 @@ do_sync_io:
 static int make_request(struct request_queue *q, struct bio * bio)
 {
 	mddev_t *mddev = q->queuedata;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	mirror_info_t *mirror;
 	r1bio_t *r1_bio;
 	struct bio *read_bio;
@@ -991,7 +991,7 @@ static int make_request(struct request_queue *q, struct bio * bio)
 
 static void status(struct seq_file *seq, mddev_t *mddev)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i;
 
 	seq_printf(seq, " [%d/%d] [", conf->raid_disks,
@@ -1010,7 +1010,7 @@ static void status(struct seq_file *seq, mddev_t *mddev)
 static void error(mddev_t *mddev, mdk_rdev_t *rdev)
 {
 	char b[BDEVNAME_SIZE];
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 
 	/*
 	 * If it is not operational, then we have already marked it as dead
@@ -1214,7 +1214,7 @@ static void end_sync_write(struct bio *bio, int error)
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	r1bio_t * r1_bio = (r1bio_t *)(bio->bi_private);
 	mddev_t *mddev = r1_bio->mddev;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i;
 	int mirror=0;
 
@@ -1248,7 +1248,7 @@ static void end_sync_write(struct bio *bio, int error)
 
 static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i;
 	int disks = conf->raid_disks;
 	struct bio *bio, *wbio;
@@ -1562,7 +1562,7 @@ static void raid1d(mddev_t *mddev)
 	r1bio_t *r1_bio;
 	struct bio *bio;
 	unsigned long flags;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	struct list_head *head = &conf->retry_list;
 	int unplug=0;
 	mdk_rdev_t *rdev;
@@ -1585,7 +1585,7 @@ static void raid1d(mddev_t *mddev)
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 
 		mddev = r1_bio->mddev;
-		conf = mddev_to_conf(mddev);
+		conf = mddev->private;
 		if (test_bit(R1BIO_IsSync, &r1_bio->state)) {
 			sync_request_write(mddev, r1_bio);
 			unplug = 1;
@@ -1706,7 +1706,7 @@ static int init_resync(conf_t *conf)
 
 static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	r1bio_t *r1_bio;
 	struct bio *bio;
 	sector_t max_sector, nr_sectors;
@@ -2052,6 +2052,10 @@ static int run(mddev_t *mddev)
 		goto out_free_conf;
 	}
 
+	if (mddev->recovery_cp != MaxSector)
+		printk(KERN_NOTICE "raid1: %s is not clean"
+		       " -- starting background reconstruction\n",
+		       mdname(mddev));
 	printk(KERN_INFO 
 		"raid1: raid set %s active with %d out of %d mirrors\n",
 		mdname(mddev), mddev->raid_disks - mddev->degraded, 
@@ -2087,7 +2091,7 @@ out:
 
 static int stop(mddev_t *mddev)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	struct bitmap *bitmap = mddev->bitmap;
 	int behind_wait = 0;
 
@@ -2155,16 +2159,16 @@ static int raid1_reshape(mddev_t *mddev)
 	mempool_t *newpool, *oldpool;
 	struct pool_info *newpoolinfo;
 	mirror_info_t *newmirrors;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int cnt, raid_disks;
 	unsigned long flags;
 	int d, d2, err;
 
 	/* Cannot change chunk_size, layout, or level */
-	if (mddev->chunk_size != mddev->new_chunk ||
+	if (mddev->chunk_sectors != mddev->new_chunk_sectors ||
 	    mddev->layout != mddev->new_layout ||
 	    mddev->level != mddev->new_level) {
-		mddev->new_chunk = mddev->chunk_size;
+		mddev->new_chunk_sectors = mddev->chunk_sectors;
 		mddev->new_layout = mddev->layout;
 		mddev->new_level = mddev->level;
 		return -EINVAL;
@@ -2252,7 +2256,7 @@ static int raid1_reshape(mddev_t *mddev)
 
 static void raid1_quiesce(mddev_t *mddev, int state)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 
 	switch(state) {
 	case 1:

drivers/md/raid1.h

@@ -63,12 +63,6 @@ struct r1_private_data_s {
 
 typedef struct r1_private_data_s conf_t;
 
-/*
- * this is the only point in the RAID code where we violate
- * C type safety. mddev->private is an 'opaque' pointer.
- */
-#define mddev_to_conf(mddev) ((conf_t *) mddev->private)
-
 /*
  * this is our 'private' RAID1 bio.
  *

drivers/md/raid10.c

@@ -188,7 +188,7 @@ static void put_all_bios(conf_t *conf, r10bio_t *r10_bio)
 
 static void free_r10bio(r10bio_t *r10_bio)
 {
-	conf_t *conf = mddev_to_conf(r10_bio->mddev);
+	conf_t *conf = r10_bio->mddev->private;
 
 	/*
 	 * Wake up any possible resync thread that waits for the device
@@ -202,7 +202,7 @@ static void free_r10bio(r10bio_t *r10_bio)
 
 static void put_buf(r10bio_t *r10_bio)
 {
-	conf_t *conf = mddev_to_conf(r10_bio->mddev);
+	conf_t *conf = r10_bio->mddev->private;
 
 	mempool_free(r10_bio, conf->r10buf_pool);
 
@@ -213,7 +213,7 @@ static void reschedule_retry(r10bio_t *r10_bio)
 {
 	unsigned long flags;
 	mddev_t *mddev = r10_bio->mddev;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 
 	spin_lock_irqsave(&conf->device_lock, flags);
 	list_add(&r10_bio->retry_list, &conf->retry_list);
@@ -245,7 +245,7 @@ static void raid_end_bio_io(r10bio_t *r10_bio)
  */
 static inline void update_head_pos(int slot, r10bio_t *r10_bio)
 {
-	conf_t *conf = mddev_to_conf(r10_bio->mddev);
+	conf_t *conf = r10_bio->mddev->private;
 
 	conf->mirrors[r10_bio->devs[slot].devnum].head_position =
 		r10_bio->devs[slot].addr + (r10_bio->sectors);
@@ -256,7 +256,7 @@ static void raid10_end_read_request(struct bio *bio, int error)
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	r10bio_t * r10_bio = (r10bio_t *)(bio->bi_private);
 	int slot, dev;
-	conf_t *conf = mddev_to_conf(r10_bio->mddev);
+	conf_t *conf = r10_bio->mddev->private;
 
 
 	slot = r10_bio->read_slot;
@@ -297,7 +297,7 @@ static void raid10_end_write_request(struct bio *bio, int error)
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	r10bio_t * r10_bio = (r10bio_t *)(bio->bi_private);
 	int slot, dev;
-	conf_t *conf = mddev_to_conf(r10_bio->mddev);
+	conf_t *conf = r10_bio->mddev->private;
 
 	for (slot = 0; slot < conf->copies; slot++)
 		if (r10_bio->devs[slot].bio == bio)
@@ -461,7 +461,7 @@ static int raid10_mergeable_bvec(struct request_queue *q,
 	mddev_t *mddev = q->queuedata;
 	sector_t sector = bvm->bi_sector + get_start_sect(bvm->bi_bdev);
 	int max;
-	unsigned int chunk_sectors = mddev->chunk_size >> 9;
+	unsigned int chunk_sectors = mddev->chunk_sectors;
 	unsigned int bio_sectors = bvm->bi_size >> 9;
 
 	max =  (chunk_sectors - ((sector & (chunk_sectors - 1)) + bio_sectors)) << 9;
@@ -596,7 +596,7 @@ rb_out:
 
 static void unplug_slaves(mddev_t *mddev)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i;
 
 	rcu_read_lock();
@@ -628,7 +628,7 @@ static void raid10_unplug(struct request_queue *q)
 static int raid10_congested(void *data, int bits)
 {
 	mddev_t *mddev = data;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i, ret = 0;
 
 	rcu_read_lock();
@@ -788,7 +788,7 @@ static void unfreeze_array(conf_t *conf)
 static int make_request(struct request_queue *q, struct bio * bio)
 {
 	mddev_t *mddev = q->queuedata;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	mirror_info_t *mirror;
 	r10bio_t *r10_bio;
 	struct bio *read_bio;
@@ -981,11 +981,11 @@ static int make_request(struct request_queue *q, struct bio * bio)
 
 static void status(struct seq_file *seq, mddev_t *mddev)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i;
 
 	if (conf->near_copies < conf->raid_disks)
-		seq_printf(seq, " %dK chunks", mddev->chunk_size/1024);
+		seq_printf(seq, " %dK chunks", mddev->chunk_sectors / 2);
 	if (conf->near_copies > 1)
 		seq_printf(seq, " %d near-copies", conf->near_copies);
 	if (conf->far_copies > 1) {
@@ -1006,7 +1006,7 @@ static void status(struct seq_file *seq, mddev_t *mddev)
 static void error(mddev_t *mddev, mdk_rdev_t *rdev)
 {
 	char b[BDEVNAME_SIZE];
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 
 	/*
 	 * If it is not operational, then we have already marked it as dead
@@ -1215,7 +1215,7 @@ abort:
 static void end_sync_read(struct bio *bio, int error)
 {
 	r10bio_t * r10_bio = (r10bio_t *)(bio->bi_private);
-	conf_t *conf = mddev_to_conf(r10_bio->mddev);
+	conf_t *conf = r10_bio->mddev->private;
 	int i,d;
 
 	for (i=0; i<conf->copies; i++)
@@ -1253,7 +1253,7 @@ static void end_sync_write(struct bio *bio, int error)
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	r10bio_t * r10_bio = (r10bio_t *)(bio->bi_private);
 	mddev_t *mddev = r10_bio->mddev;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i,d;
 
 	for (i = 0; i < conf->copies; i++)
@@ -1300,7 +1300,7 @@ static void end_sync_write(struct bio *bio, int error)
  */
 static void sync_request_write(mddev_t *mddev, r10bio_t *r10_bio)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i, first;
 	struct bio *tbio, *fbio;
 
@@ -1400,7 +1400,7 @@ done:
 
 static void recovery_request_write(mddev_t *mddev, r10bio_t *r10_bio)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	int i, d;
 	struct bio *bio, *wbio;
 
@@ -1549,7 +1549,7 @@ static void raid10d(mddev_t *mddev)
 	r10bio_t *r10_bio;
 	struct bio *bio;
 	unsigned long flags;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	struct list_head *head = &conf->retry_list;
 	int unplug=0;
 	mdk_rdev_t *rdev;
@@ -1572,7 +1572,7 @@ static void raid10d(mddev_t *mddev)
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 
 		mddev = r10_bio->mddev;
-		conf = mddev_to_conf(mddev);
+		conf = mddev->private;
 		if (test_bit(R10BIO_IsSync, &r10_bio->state)) {
 			sync_request_write(mddev, r10_bio);
 			unplug = 1;
@@ -1680,7 +1680,7 @@ static int init_resync(conf_t *conf)
 
 static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 	r10bio_t *r10_bio;
 	struct bio *biolist = NULL, *bio;
 	sector_t max_sector, nr_sectors;
@@ -2026,7 +2026,7 @@ static sector_t
 raid10_size(mddev_t *mddev, sector_t sectors, int raid_disks)
 {
 	sector_t size;
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 
 	if (!raid_disks)
 		raid_disks = mddev->raid_disks;
@@ -2050,9 +2050,10 @@ static int run(mddev_t *mddev)
 	int nc, fc, fo;
 	sector_t stride, size;
 
-	if (mddev->chunk_size < PAGE_SIZE) {
+	if (mddev->chunk_sectors < (PAGE_SIZE >> 9) ||
+	    !is_power_of_2(mddev->chunk_sectors)) {
 		printk(KERN_ERR "md/raid10: chunk size must be "
-		       "at least PAGE_SIZE(%ld).\n", PAGE_SIZE);
+		       "at least PAGE_SIZE(%ld) and be a power of 2.\n", PAGE_SIZE);
 		return -EINVAL;
 	}
 
@@ -2095,8 +2096,8 @@ static int run(mddev_t *mddev)
 	conf->far_copies = fc;
 	conf->copies = nc*fc;
 	conf->far_offset = fo;
-	conf->chunk_mask = (sector_t)(mddev->chunk_size>>9)-1;
-	conf->chunk_shift = ffz(~mddev->chunk_size) - 9;
+	conf->chunk_mask = mddev->chunk_sectors - 1;
+	conf->chunk_shift = ffz(~mddev->chunk_sectors);
 	size = mddev->dev_sectors >> conf->chunk_shift;
 	sector_div(size, fc);
 	size = size * conf->raid_disks;
@@ -2185,6 +2186,10 @@ static int run(mddev_t *mddev)
 		goto out_free_conf;
 	}
 
+	if (mddev->recovery_cp != MaxSector)
+		printk(KERN_NOTICE "raid10: %s is not clean"
+		       " -- starting background reconstruction\n",
+		       mdname(mddev));
 	printk(KERN_INFO
 		"raid10: raid set %s active with %d out of %d devices\n",
 		mdname(mddev), mddev->raid_disks - mddev->degraded,
@@ -2204,7 +2209,8 @@ static int run(mddev_t *mddev)
 	 * maybe...
 	 */
 	{
-		int stripe = conf->raid_disks * (mddev->chunk_size / PAGE_SIZE);
+		int stripe = conf->raid_disks *
+			((mddev->chunk_sectors << 9) / PAGE_SIZE);
 		stripe /= conf->near_copies;
 		if (mddev->queue->backing_dev_info.ra_pages < 2* stripe)
 			mddev->queue->backing_dev_info.ra_pages = 2* stripe;
@@ -2227,7 +2233,7 @@ out:
 
 static int stop(mddev_t *mddev)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 
 	raise_barrier(conf, 0);
 	lower_barrier(conf);
@@ -2245,7 +2251,7 @@ static int stop(mddev_t *mddev)
 
 static void raid10_quiesce(mddev_t *mddev, int state)
 {
-	conf_t *conf = mddev_to_conf(mddev);
+	conf_t *conf = mddev->private;
 
 	switch(state) {
 	case 1:

drivers/md/raid10.h

@@ -61,12 +61,6 @@ struct r10_private_data_s {
 
 typedef struct r10_private_data_s conf_t;
 
-/*
- * this is the only point in the RAID code where we violate
- * C type safety. mddev->private is an 'opaque' pointer.
- */
-#define mddev_to_conf(mddev) ((conf_t *) mddev->private)
-
 /*
  * this is our 'private' RAID10 bio.
  *

drivers/md/raid5.c

@@ -1274,8 +1274,8 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector,
 	sector_t new_sector;
 	int algorithm = previous ? conf->prev_algo
 				 : conf->algorithm;
-	int sectors_per_chunk = previous ? (conf->prev_chunk >> 9)
-					 : (conf->chunk_size >> 9);
+	int sectors_per_chunk = previous ? conf->prev_chunk_sectors
+					 : conf->chunk_sectors;
 	int raid_disks = previous ? conf->previous_raid_disks
 				  : conf->raid_disks;
 	int data_disks = raid_disks - conf->max_degraded;
@@ -1480,8 +1480,8 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i, int previous)
 	int raid_disks = sh->disks;
 	int data_disks = raid_disks - conf->max_degraded;
 	sector_t new_sector = sh->sector, check;
-	int sectors_per_chunk = previous ? (conf->prev_chunk >> 9)
-					 : (conf->chunk_size >> 9);
+	int sectors_per_chunk = previous ? conf->prev_chunk_sectors
+					 : conf->chunk_sectors;
 	int algorithm = previous ? conf->prev_algo
 				 : conf->algorithm;
 	sector_t stripe;
@@ -1997,8 +1997,7 @@ static void stripe_set_idx(sector_t stripe, raid5_conf_t *conf, int previous,
 			    struct stripe_head *sh)
 {
 	int sectors_per_chunk =
-		previous ? (conf->prev_chunk >> 9)
-			 : (conf->chunk_size >> 9);
+		previous ? conf->prev_chunk_sectors : conf->chunk_sectors;
 	int dd_idx;
 	int chunk_offset = sector_div(stripe, sectors_per_chunk);
 	int disks = previous ? conf->previous_raid_disks : conf->raid_disks;
@@ -3284,7 +3283,7 @@ static void activate_bit_delay(raid5_conf_t *conf)
 
 static void unplug_slaves(mddev_t *mddev)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	int i;
 
 	rcu_read_lock();
@@ -3308,7 +3307,7 @@ static void unplug_slaves(mddev_t *mddev)
 static void raid5_unplug_device(struct request_queue *q)
 {
 	mddev_t *mddev = q->queuedata;
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	unsigned long flags;
 
 	spin_lock_irqsave(&conf->device_lock, flags);
@@ -3327,7 +3326,7 @@ static void raid5_unplug_device(struct request_queue *q)
 static int raid5_congested(void *data, int bits)
 {
 	mddev_t *mddev = data;
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 
 	/* No difference between reads and writes.  Just check
 	 * how busy the stripe_cache is
@@ -3352,14 +3351,14 @@ static int raid5_mergeable_bvec(struct request_queue *q,
 	mddev_t *mddev = q->queuedata;
 	sector_t sector = bvm->bi_sector + get_start_sect(bvm->bi_bdev);
 	int max;
-	unsigned int chunk_sectors = mddev->chunk_size >> 9;
+	unsigned int chunk_sectors = mddev->chunk_sectors;
 	unsigned int bio_sectors = bvm->bi_size >> 9;
 
 	if ((bvm->bi_rw & 1) == WRITE)
 		return biovec->bv_len; /* always allow writes to be mergeable */
 
-	if (mddev->new_chunk < mddev->chunk_size)
-		chunk_sectors = mddev->new_chunk >> 9;
+	if (mddev->new_chunk_sectors < mddev->chunk_sectors)
+		chunk_sectors = mddev->new_chunk_sectors;
 	max =  (chunk_sectors - ((sector & (chunk_sectors - 1)) + bio_sectors)) << 9;
 	if (max < 0) max = 0;
 	if (max <= biovec->bv_len && bio_sectors == 0)
@@ -3372,11 +3371,11 @@ static int raid5_mergeable_bvec(struct request_queue *q,
 static int in_chunk_boundary(mddev_t *mddev, struct bio *bio)
 {
 	sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev);
-	unsigned int chunk_sectors = mddev->chunk_size >> 9;
+	unsigned int chunk_sectors = mddev->chunk_sectors;
 	unsigned int bio_sectors = bio->bi_size >> 9;
 
-	if (mddev->new_chunk < mddev->chunk_size)
-		chunk_sectors = mddev->new_chunk >> 9;
+	if (mddev->new_chunk_sectors < mddev->chunk_sectors)
+		chunk_sectors = mddev->new_chunk_sectors;
 	return  chunk_sectors >=
 		((sector & (chunk_sectors - 1)) + bio_sectors);
 }
@@ -3440,7 +3439,7 @@ static void raid5_align_endio(struct bio *bi, int error)
 	bio_put(bi);
 
 	mddev = raid_bi->bi_bdev->bd_disk->queue->queuedata;
-	conf = mddev_to_conf(mddev);
+	conf = mddev->private;
 	rdev = (void*)raid_bi->bi_next;
 	raid_bi->bi_next = NULL;
 
@@ -3482,7 +3481,7 @@ static int bio_fits_rdev(struct bio *bi)
 static int chunk_aligned_read(struct request_queue *q, struct bio * raid_bio)
 {
 	mddev_t *mddev = q->queuedata;
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	unsigned int dd_idx;
 	struct bio* align_bi;
 	mdk_rdev_t *rdev;
@@ -3599,7 +3598,7 @@ static struct stripe_head *__get_priority_stripe(raid5_conf_t *conf)
 static int make_request(struct request_queue *q, struct bio * bi)
 {
 	mddev_t *mddev = q->queuedata;
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	int dd_idx;
 	sector_t new_sector;
 	sector_t logical_sector, last_sector;
@@ -3696,6 +3695,7 @@ static int make_request(struct request_queue *q, struct bio * bi)
 				spin_unlock_irq(&conf->device_lock);
 				if (must_retry) {
 					release_stripe(sh);
+					schedule();
 					goto retry;
 				}
 			}
@@ -3791,10 +3791,10 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped
 	 * If old and new chunk sizes differ, we need to process the
 	 * largest of these
 	 */
-	if (mddev->new_chunk > mddev->chunk_size)
-		reshape_sectors = mddev->new_chunk / 512;
+	if (mddev->new_chunk_sectors > mddev->chunk_sectors)
+		reshape_sectors = mddev->new_chunk_sectors;
 	else
-		reshape_sectors = mddev->chunk_size / 512;
+		reshape_sectors = mddev->chunk_sectors;
 
 	/* we update the metadata when there is more than 3Meg
 	 * in the block range (that is rather arbitrary, should
@@ -3917,7 +3917,7 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped
 				     1, &dd_idx, NULL);
 	last_sector =
 		raid5_compute_sector(conf, ((stripe_addr+reshape_sectors)
-					    *(new_data_disks) - 1),
+					    * new_data_disks - 1),
 				     1, &dd_idx, NULL);
 	if (last_sector >= mddev->dev_sectors)
 		last_sector = mddev->dev_sectors - 1;
@@ -3946,7 +3946,7 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped
 		wait_event(conf->wait_for_overlap,
 			   atomic_read(&conf->reshape_stripes) == 0);
 		mddev->reshape_position = conf->reshape_progress;
-		mddev->curr_resync_completed = mddev->curr_resync;
+		mddev->curr_resync_completed = mddev->curr_resync + reshape_sectors;
 		conf->reshape_checkpoint = jiffies;
 		set_bit(MD_CHANGE_DEVS, &mddev->flags);
 		md_wakeup_thread(mddev->thread);
@@ -4129,7 +4129,7 @@ static int  retry_aligned_read(raid5_conf_t *conf, struct bio *raid_bio)
 static void raid5d(mddev_t *mddev)
 {
 	struct stripe_head *sh;
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	int handled;
 
 	pr_debug("+++ raid5d active\n");
@@ -4185,7 +4185,7 @@ static void raid5d(mddev_t *mddev)
 static ssize_t
 raid5_show_stripe_cache_size(mddev_t *mddev, char *page)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	if (conf)
 		return sprintf(page, "%d\n", conf->max_nr_stripes);
 	else
@@ -4195,7 +4195,7 @@ raid5_show_stripe_cache_size(mddev_t *mddev, char *page)
 static ssize_t
 raid5_store_stripe_cache_size(mddev_t *mddev, const char *page, size_t len)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	unsigned long new;
 	int err;
 
@@ -4233,7 +4233,7 @@ raid5_stripecache_size = __ATTR(stripe_cache_size, S_IRUGO | S_IWUSR,
 static ssize_t
 raid5_show_preread_threshold(mddev_t *mddev, char *page)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	if (conf)
 		return sprintf(page, "%d\n", conf->bypass_threshold);
 	else
@@ -4243,7 +4243,7 @@ raid5_show_preread_threshold(mddev_t *mddev, char *page)
 static ssize_t
 raid5_store_preread_threshold(mddev_t *mddev, const char *page, size_t len)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	unsigned long new;
 	if (len >= PAGE_SIZE)
 		return -EINVAL;
@@ -4267,7 +4267,7 @@ raid5_preread_bypass_threshold = __ATTR(preread_bypass_threshold,
 static ssize_t
 stripe_cache_active_show(mddev_t *mddev, char *page)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	if (conf)
 		return sprintf(page, "%d\n", atomic_read(&conf->active_stripes));
 	else
@@ -4291,7 +4291,7 @@ static struct attribute_group raid5_attrs_group = {
 static sector_t
 raid5_size(mddev_t *mddev, sector_t sectors, int raid_disks)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 
 	if (!sectors)
 		sectors = mddev->dev_sectors;
@@ -4303,8 +4303,8 @@ raid5_size(mddev_t *mddev, sector_t sectors, int raid_disks)
 			raid_disks = conf->previous_raid_disks;
 	}
 
-	sectors &= ~((sector_t)mddev->chunk_size/512 - 1);
-	sectors &= ~((sector_t)mddev->new_chunk/512 - 1);
+	sectors &= ~((sector_t)mddev->chunk_sectors - 1);
+	sectors &= ~((sector_t)mddev->new_chunk_sectors - 1);
 	return sectors * (raid_disks - conf->max_degraded);
 }
 
@@ -4336,9 +4336,11 @@ static raid5_conf_t *setup_conf(mddev_t *mddev)
 		return ERR_PTR(-EINVAL);
 	}
 
-	if (!mddev->new_chunk || mddev->new_chunk % PAGE_SIZE) {
+	if (!mddev->new_chunk_sectors ||
+	    (mddev->new_chunk_sectors << 9) % PAGE_SIZE ||
+	    !is_power_of_2(mddev->new_chunk_sectors)) {
 		printk(KERN_ERR "raid5: invalid chunk size %d for %s\n",
-			mddev->new_chunk, mdname(mddev));
+		       mddev->new_chunk_sectors << 9, mdname(mddev));
 		return ERR_PTR(-EINVAL);
 	}
 
@@ -4401,7 +4403,7 @@ static raid5_conf_t *setup_conf(mddev_t *mddev)
 			conf->fullsync = 1;
 	}
 
-	conf->chunk_size = mddev->new_chunk;
+	conf->chunk_sectors = mddev->new_chunk_sectors;
 	conf->level = mddev->new_level;
 	if (conf->level == 6)
 		conf->max_degraded = 2;
@@ -4411,7 +4413,7 @@ static raid5_conf_t *setup_conf(mddev_t *mddev)
 	conf->max_nr_stripes = NR_STRIPES;
 	conf->reshape_progress = mddev->reshape_position;
 	if (conf->reshape_progress != MaxSector) {
-		conf->prev_chunk = mddev->chunk_size;
+		conf->prev_chunk_sectors = mddev->chunk_sectors;
 		conf->prev_algo = mddev->layout;
 	}
 
@@ -4453,6 +4455,10 @@ static int run(mddev_t *mddev)
 	int working_disks = 0;
 	mdk_rdev_t *rdev;
 
+	if (mddev->recovery_cp != MaxSector)
+		printk(KERN_NOTICE "raid5: %s is not clean"
+		       " -- starting background reconstruction\n",
+		       mdname(mddev));
 	if (mddev->reshape_position != MaxSector) {
 		/* Check that we can continue the reshape.
 		 * Currently only disks can change, it must
@@ -4475,7 +4481,7 @@ static int run(mddev_t *mddev)
 		 * geometry.
 		 */
 		here_new = mddev->reshape_position;
-		if (sector_div(here_new, (mddev->new_chunk>>9)*
+		if (sector_div(here_new, mddev->new_chunk_sectors *
 			       (mddev->raid_disks - max_degraded))) {
 			printk(KERN_ERR "raid5: reshape_position not "
 			       "on a stripe boundary\n");
@@ -4483,7 +4489,7 @@ static int run(mddev_t *mddev)
 		}
 		/* here_new is the stripe we will write to */
 		here_old = mddev->reshape_position;
-		sector_div(here_old, (mddev->chunk_size>>9)*
+		sector_div(here_old, mddev->chunk_sectors *
 			   (old_disks-max_degraded));
 		/* here_old is the first stripe that we might need to read
 		 * from */
@@ -4498,7 +4504,7 @@ static int run(mddev_t *mddev)
 	} else {
 		BUG_ON(mddev->level != mddev->new_level);
 		BUG_ON(mddev->layout != mddev->new_layout);
-		BUG_ON(mddev->chunk_size != mddev->new_chunk);
+		BUG_ON(mddev->chunk_sectors != mddev->new_chunk_sectors);
 		BUG_ON(mddev->delta_disks != 0);
 	}
 
@@ -4532,7 +4538,7 @@ static int run(mddev_t *mddev)
 	}
 
 	/* device size must be a multiple of chunk size */
-	mddev->dev_sectors &= ~(mddev->chunk_size / 512 - 1);
+	mddev->dev_sectors &= ~(mddev->chunk_sectors - 1);
 	mddev->resync_max_sectors = mddev->dev_sectors;
 
 	if (mddev->degraded > 0 &&
@@ -4581,7 +4587,7 @@ static int run(mddev_t *mddev)
 	{
 		int data_disks = conf->previous_raid_disks - conf->max_degraded;
 		int stripe = data_disks *
-			(mddev->chunk_size / PAGE_SIZE);
+			((mddev->chunk_sectors << 9) / PAGE_SIZE);
 		if (mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
 			mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
 	}
@@ -4678,7 +4684,8 @@ static void status(struct seq_file *seq, mddev_t *mddev)
 	raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
 	int i;
 
-	seq_printf (seq, " level %d, %dk chunk, algorithm %d", mddev->level, mddev->chunk_size >> 10, mddev->layout);
+	seq_printf(seq, " level %d, %dk chunk, algorithm %d", mddev->level,
+		mddev->chunk_sectors / 2, mddev->layout);
 	seq_printf (seq, " [%d/%d] [", conf->raid_disks, conf->raid_disks - mddev->degraded);
 	for (i = 0; i < conf->raid_disks; i++)
 		seq_printf (seq, "%s",
@@ -4826,7 +4833,7 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors)
 	 * any io in the removed space completes, but it hardly seems
 	 * worth it.
 	 */
-	sectors &= ~((sector_t)mddev->chunk_size/512 - 1);
+	sectors &= ~((sector_t)mddev->chunk_sectors - 1);
 	md_set_array_sectors(mddev, raid5_size(mddev, sectors,
 					       mddev->raid_disks));
 	if (mddev->array_sectors >
@@ -4843,14 +4850,37 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors)
 	return 0;
 }
 
-static int raid5_check_reshape(mddev_t *mddev)
+static int check_stripe_cache(mddev_t *mddev)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	/* Can only proceed if there are plenty of stripe_heads.
+	 * We need a minimum of one full stripe,, and for sensible progress
+	 * it is best to have about 4 times that.
+	 * If we require 4 times, then the default 256 4K stripe_heads will
+	 * allow for chunk sizes up to 256K, which is probably OK.
+	 * If the chunk size is greater, user-space should request more
+	 * stripe_heads first.
+	 */
+	raid5_conf_t *conf = mddev->private;
+	if (((mddev->chunk_sectors << 9) / STRIPE_SIZE) * 4
+	    > conf->max_nr_stripes ||
+	    ((mddev->new_chunk_sectors << 9) / STRIPE_SIZE) * 4
+	    > conf->max_nr_stripes) {
+		printk(KERN_WARNING "raid5: reshape: not enough stripes.  Needed %lu\n",
+		       ((max(mddev->chunk_sectors, mddev->new_chunk_sectors) << 9)
+			/ STRIPE_SIZE)*4);
+		return 0;
+	}
+	return 1;
+}
+
+static int check_reshape(mddev_t *mddev)
+{
+	raid5_conf_t *conf = mddev->private;
 
 	if (mddev->delta_disks == 0 &&
 	    mddev->new_layout == mddev->layout &&
-	    mddev->new_chunk == mddev->chunk_size)
-		return -EINVAL; /* nothing to do */
+	    mddev->new_chunk_sectors == mddev->chunk_sectors)
+		return 0; /* nothing to do */
 	if (mddev->bitmap)
 		/* Cannot grow a bitmap yet */
 		return -EBUSY;
@@ -4869,28 +4899,15 @@ static int raid5_check_reshape(mddev_t *mddev)
 			return -EINVAL;
 	}
 
-	/* Can only proceed if there are plenty of stripe_heads.
-	 * We need a minimum of one full stripe,, and for sensible progress
-	 * it is best to have about 4 times that.
-	 * If we require 4 times, then the default 256 4K stripe_heads will
-	 * allow for chunk sizes up to 256K, which is probably OK.
-	 * If the chunk size is greater, user-space should request more
-	 * stripe_heads first.
-	 */
-	if ((mddev->chunk_size / STRIPE_SIZE) * 4 > conf->max_nr_stripes ||
-	    (mddev->new_chunk / STRIPE_SIZE) * 4 > conf->max_nr_stripes) {
-		printk(KERN_WARNING "raid5: reshape: not enough stripes.  Needed %lu\n",
-		       (max(mddev->chunk_size, mddev->new_chunk)
-			/ STRIPE_SIZE)*4);
+	if (!check_stripe_cache(mddev))
 		return -ENOSPC;
-	}
 
 	return resize_stripes(conf, conf->raid_disks + mddev->delta_disks);
 }
 
 static int raid5_start_reshape(mddev_t *mddev)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 	mdk_rdev_t *rdev;
 	int spares = 0;
 	int added_devices = 0;
@@ -4899,6 +4916,9 @@ static int raid5_start_reshape(mddev_t *mddev)
 	if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
 		return -EBUSY;
 
+	if (!check_stripe_cache(mddev))
+		return -ENOSPC;
+
 	list_for_each_entry(rdev, &mddev->disks, same_set)
 		if (rdev->raid_disk < 0 &&
 		    !test_bit(Faulty, &rdev->flags))
@@ -4925,8 +4945,8 @@ static int raid5_start_reshape(mddev_t *mddev)
 	spin_lock_irq(&conf->device_lock);
 	conf->previous_raid_disks = conf->raid_disks;
 	conf->raid_disks += mddev->delta_disks;
-	conf->prev_chunk = conf->chunk_size;
-	conf->chunk_size = mddev->new_chunk;
+	conf->prev_chunk_sectors = conf->chunk_sectors;
+	conf->chunk_sectors = mddev->new_chunk_sectors;
 	conf->prev_algo = conf->algorithm;
 	conf->algorithm = mddev->new_layout;
 	if (mddev->delta_disks < 0)
@@ -5008,7 +5028,7 @@ static void end_reshape(raid5_conf_t *conf)
 		 */
 		{
 			int data_disks = conf->raid_disks - conf->max_degraded;
-			int stripe = data_disks * (conf->chunk_size
+			int stripe = data_disks * ((conf->chunk_sectors << 9)
 						   / PAGE_SIZE);
 			if (conf->mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
 				conf->mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
@@ -5022,7 +5042,7 @@ static void end_reshape(raid5_conf_t *conf)
 static void raid5_finish_reshape(mddev_t *mddev)
 {
 	struct block_device *bdev;
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 
 	if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
 
@@ -5053,7 +5073,7 @@ static void raid5_finish_reshape(mddev_t *mddev)
 				raid5_remove_disk(mddev, d);
 		}
 		mddev->layout = conf->algorithm;
-		mddev->chunk_size = conf->chunk_size;
+		mddev->chunk_sectors = conf->chunk_sectors;
 		mddev->reshape_position = MaxSector;
 		mddev->delta_disks = 0;
 	}
@@ -5061,7 +5081,7 @@ static void raid5_finish_reshape(mddev_t *mddev)
 
 static void raid5_quiesce(mddev_t *mddev, int state)
 {
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
 
 	switch(state) {
 	case 2: /* resume for a suspend */
@@ -5111,7 +5131,7 @@ static void *raid5_takeover_raid1(mddev_t *mddev)
 
 	mddev->new_level = 5;
 	mddev->new_layout = ALGORITHM_LEFT_SYMMETRIC;
-	mddev->new_chunk = chunksect << 9;
+	mddev->new_chunk_sectors = chunksect;
 
 	return setup_conf(mddev);
 }
@@ -5150,24 +5170,24 @@ static void *raid5_takeover_raid6(mddev_t *mddev)
 }
 
 
-static int raid5_reconfig(mddev_t *mddev, int new_layout, int new_chunk)
+static int raid5_check_reshape(mddev_t *mddev)
 {
 	/* For a 2-drive array, the layout and chunk size can be changed
 	 * immediately as not restriping is needed.
 	 * For larger arrays we record the new value - after validation
 	 * to be used by a reshape pass.
 	 */
-	raid5_conf_t *conf = mddev_to_conf(mddev);
+	raid5_conf_t *conf = mddev->private;
+	int new_chunk = mddev->new_chunk_sectors;
 
-	if (new_layout >= 0 && !algorithm_valid_raid5(new_layout))
+	if (mddev->new_layout >= 0 && !algorithm_valid_raid5(mddev->new_layout))
 		return -EINVAL;
 	if (new_chunk > 0) {
-		if (new_chunk & (new_chunk-1))
-			/* not a power of 2 */
+		if (!is_power_of_2(new_chunk))
 			return -EINVAL;
-		if (new_chunk < PAGE_SIZE)
+		if (new_chunk < (PAGE_SIZE>>9))
 			return -EINVAL;
-		if (mddev->array_sectors & ((new_chunk>>9)-1))
+		if (mddev->array_sectors & (new_chunk-1))
 			/* not factor of array size */
 			return -EINVAL;
 	}
@@ -5175,49 +5195,39 @@ static int raid5_reconfig(mddev_t *mddev, int new_layout, int new_chunk)
 	/* They look valid */
 
 	if (mddev->raid_disks == 2) {
-
-		if (new_layout >= 0) {
-			conf->algorithm = new_layout;
-			mddev->layout = mddev->new_layout = new_layout;
+		/* can make the change immediately */
+		if (mddev->new_layout >= 0) {
+			conf->algorithm = mddev->new_layout;
+			mddev->layout = mddev->new_layout;
 		}
 		if (new_chunk > 0) {
-			conf->chunk_size = new_chunk;
-			mddev->chunk_size = mddev->new_chunk = new_chunk;
+			conf->chunk_sectors = new_chunk ;
+			mddev->chunk_sectors = new_chunk;
 		}
 		set_bit(MD_CHANGE_DEVS, &mddev->flags);
 		md_wakeup_thread(mddev->thread);
-	} else {
-		if (new_layout >= 0)
-			mddev->new_layout = new_layout;
-		if (new_chunk > 0)
-			mddev->new_chunk = new_chunk;
 	}
-	return 0;
+	return check_reshape(mddev);
 }
 
-static int raid6_reconfig(mddev_t *mddev, int new_layout, int new_chunk)
+static int raid6_check_reshape(mddev_t *mddev)
 {
-	if (new_layout >= 0 && !algorithm_valid_raid6(new_layout))
+	int new_chunk = mddev->new_chunk_sectors;
+
+	if (mddev->new_layout >= 0 && !algorithm_valid_raid6(mddev->new_layout))
 		return -EINVAL;
 	if (new_chunk > 0) {
-		if (new_chunk & (new_chunk-1))
-			/* not a power of 2 */
+		if (!is_power_of_2(new_chunk))
 			return -EINVAL;
-		if (new_chunk < PAGE_SIZE)
+		if (new_chunk < (PAGE_SIZE >> 9))
 			return -EINVAL;
-		if (mddev->array_sectors & ((new_chunk>>9)-1))
+		if (mddev->array_sectors & (new_chunk-1))
 			/* not factor of array size */
 			return -EINVAL;
 	}
 
 	/* They look valid */
-
-	if (new_layout >= 0)
-		mddev->new_layout = new_layout;
-	if (new_chunk > 0)
-		mddev->new_chunk = new_chunk;
-
-	return 0;
+	return check_reshape(mddev);
 }
 
 static void *raid5_takeover(mddev_t *mddev)
@@ -5227,8 +5237,6 @@ static void *raid5_takeover(mddev_t *mddev)
 	 *  raid1 - if there are two drives.  We need to know the chunk size
 	 *  raid4 - trivial - just use a raid4 layout.
 	 *  raid6 - Providing it is a *_6 layout
-	 *
-	 * For now, just do raid1
 	 */
 
 	if (mddev->level == 1)
@@ -5310,12 +5318,11 @@ static struct mdk_personality raid6_personality =
 	.sync_request	= sync_request,
 	.resize		= raid5_resize,
 	.size		= raid5_size,
-	.check_reshape	= raid5_check_reshape,
+	.check_reshape	= raid6_check_reshape,
 	.start_reshape  = raid5_start_reshape,
 	.finish_reshape = raid5_finish_reshape,
 	.quiesce	= raid5_quiesce,
 	.takeover	= raid6_takeover,
-	.reconfig	= raid6_reconfig,
 };
 static struct mdk_personality raid5_personality =
 {
@@ -5338,7 +5345,6 @@ static struct mdk_personality raid5_personality =
 	.finish_reshape = raid5_finish_reshape,
 	.quiesce	= raid5_quiesce,
 	.takeover	= raid5_takeover,
-	.reconfig	= raid5_reconfig,
 };
 
 static struct mdk_personality raid4_personality =

drivers/md/raid5.h

@@ -334,7 +334,8 @@ struct raid5_private_data {
 	struct hlist_head	*stripe_hashtbl;
 	mddev_t			*mddev;
 	struct disk_info	*spare;
-	int			chunk_size, level, algorithm;
+	int			chunk_sectors;
+	int			level, algorithm;
 	int			max_degraded;
 	int			raid_disks;
 	int			max_nr_stripes;
@@ -350,7 +351,8 @@ struct raid5_private_data {
 	 */
 	sector_t		reshape_safe;
 	int			previous_raid_disks;
-	int			prev_chunk, prev_algo;
+	int			prev_chunk_sectors;
+	int			prev_algo;
 	short			generation; /* increments with every reshape */
 	unsigned long		reshape_checkpoint; /* Time we last updated
 						     * metadata */
@@ -408,8 +410,6 @@ struct raid5_private_data {
 
 typedef struct raid5_private_data raid5_conf_t;
 
-#define mddev_to_conf(mddev) ((raid5_conf_t *) mddev->private)
-
 /*
  * Our supported algorithms
  */

include/linux/raid/md_p.h

@@ -232,7 +232,7 @@ struct mdp_superblock_1 {
 	__le64	reshape_position;	/* next address in array-space for reshape */
 	__le32	delta_disks;	/* change in number of raid_disks		*/
 	__le32	new_layout;	/* new layout					*/
-	__le32	new_chunk;	/* new chunk size (bytes)			*/
+	__le32	new_chunk;	/* new chunk size (512byte sectors)		*/
 	__u8	pad1[128-124];	/* set to 0 when written */
 
 	/* constant this-device information - 64 bytes */