Auto-update from upstream

21 files changed, 2155 insertions, 893 deletions

diff --git a/drivers/md/.gitignore b/drivers/md/.gitignore
new file mode 100644
index 00000000000..a7afec6b19c
--- /dev/null
+++ b/drivers/md/.gitignore
@@ -0,0 +1,4 @@
+mktables
+raid6altivec*.c
+raid6int*.c
+raid6tables.c
diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c
index 252d55df964..76a189ceb52 100644
--- a/drivers/md/bitmap.c
+++ b/drivers/md/bitmap.c
@@ -315,6 +315,8 @@ static int write_page(struct bitmap *bitmap, struct page *page, int wait)
 	if (bitmap->file == NULL)
 		return write_sb_page(bitmap->mddev, bitmap->offset, page, wait);
 
+	flush_dcache_page(page); /* make sure visible to anyone reading the file */
+
 	if (wait)
 		lock_page(page);
 	else {
@@ -341,7 +343,7 @@ static int write_page(struct bitmap *bitmap, struct page *page, int wait)
 		/* add to list to be waited for by daemon */
 		struct page_list *item = mempool_alloc(bitmap->write_pool, GFP_NOIO);
 		item->page = page;
-		page_cache_get(page);
+		get_page(page);
 		spin_lock(&bitmap->write_lock);
 		list_add(&item->list, &bitmap->complete_pages);
 		spin_unlock(&bitmap->write_lock);
@@ -357,10 +359,10 @@ static struct page *read_page(struct file *file, unsigned long index,
 	struct inode *inode = file->f_mapping->host;
 	struct page *page = NULL;
 	loff_t isize = i_size_read(inode);
-	unsigned long end_index = isize >> PAGE_CACHE_SHIFT;
+	unsigned long end_index = isize >> PAGE_SHIFT;
 
-	PRINTK("read bitmap file (%dB @ %Lu)\n", (int)PAGE_CACHE_SIZE,
-			(unsigned long long)index << PAGE_CACHE_SHIFT);
+	PRINTK("read bitmap file (%dB @ %Lu)\n", (int)PAGE_SIZE,
+			(unsigned long long)index << PAGE_SHIFT);
 
 	page = read_cache_page(inode->i_mapping, index,
 			(filler_t *)inode->i_mapping->a_ops->readpage, file);
@@ -368,7 +370,7 @@ static struct page *read_page(struct file *file, unsigned long index,
 		goto out;
 	wait_on_page_locked(page);
 	if (!PageUptodate(page) || PageError(page)) {
-		page_cache_release(page);
+		put_page(page);
 		page = ERR_PTR(-EIO);
 		goto out;
 	}
@@ -376,14 +378,14 @@ static struct page *read_page(struct file *file, unsigned long index,
 	if (index > end_index) /* we have read beyond EOF */
 		*bytes_read = 0;
 	else if (index == end_index) /* possible short read */
-		*bytes_read = isize & ~PAGE_CACHE_MASK;
+		*bytes_read = isize & ~PAGE_MASK;
 	else
-		*bytes_read = PAGE_CACHE_SIZE; /* got a full page */
+		*bytes_read = PAGE_SIZE; /* got a full page */
 out:
 	if (IS_ERR(page))
 		printk(KERN_ALERT "md: bitmap read error: (%dB @ %Lu): %ld\n",
-			(int)PAGE_CACHE_SIZE,
-			(unsigned long long)index << PAGE_CACHE_SHIFT,
+			(int)PAGE_SIZE,
+			(unsigned long long)index << PAGE_SHIFT,
 			PTR_ERR(page));
 	return page;
 }
@@ -406,11 +408,11 @@ int bitmap_update_sb(struct bitmap *bitmap)
 		return 0;
 	}
 	spin_unlock_irqrestore(&bitmap->lock, flags);
-	sb = (bitmap_super_t *)kmap(bitmap->sb_page);
+	sb = (bitmap_super_t *)kmap_atomic(bitmap->sb_page, KM_USER0);
 	sb->events = cpu_to_le64(bitmap->mddev->events);
 	if (!bitmap->mddev->degraded)
 		sb->events_cleared = cpu_to_le64(bitmap->mddev->events);
-	kunmap(bitmap->sb_page);
+	kunmap_atomic(sb, KM_USER0);
 	return write_page(bitmap, bitmap->sb_page, 1);
 }
 
@@ -421,7 +423,7 @@ void bitmap_print_sb(struct bitmap *bitmap)
 
 	if (!bitmap || !bitmap->sb_page)
 		return;
-	sb = (bitmap_super_t *)kmap(bitmap->sb_page);
+	sb = (bitmap_super_t *)kmap_atomic(bitmap->sb_page, KM_USER0);
 	printk(KERN_DEBUG "%s: bitmap file superblock:\n", bmname(bitmap));
 	printk(KERN_DEBUG "         magic: %08x\n", le32_to_cpu(sb->magic));
 	printk(KERN_DEBUG "       version: %d\n", le32_to_cpu(sb->version));
@@ -440,7 +442,7 @@ void bitmap_print_sb(struct bitmap *bitmap)
 	printk(KERN_DEBUG "     sync size: %llu KB\n",
 			(unsigned long long)le64_to_cpu(sb->sync_size)/2);
 	printk(KERN_DEBUG "max write behind: %d\n", le32_to_cpu(sb->write_behind));
-	kunmap(bitmap->sb_page);
+	kunmap_atomic(sb, KM_USER0);
 }
 
 /* read the superblock from the bitmap file and initialize some bitmap fields */
@@ -466,7 +468,7 @@ static int bitmap_read_sb(struct bitmap *bitmap)
 		return err;
 	}
 
-	sb = (bitmap_super_t *)kmap(bitmap->sb_page);
+	sb = (bitmap_super_t *)kmap_atomic(bitmap->sb_page, KM_USER0);
 
 	if (bytes_read < sizeof(*sb)) { /* short read */
 		printk(KERN_INFO "%s: bitmap file superblock truncated\n",
@@ -485,12 +487,12 @@ static int bitmap_read_sb(struct bitmap *bitmap)
 	else if (le32_to_cpu(sb->version) < BITMAP_MAJOR_LO ||
 		 le32_to_cpu(sb->version) > BITMAP_MAJOR_HI)
 		reason = "unrecognized superblock version";
-	else if (chunksize < 512 || chunksize > (1024 * 1024 * 4))
-		reason = "bitmap chunksize out of range (512B - 4MB)";
+	else if (chunksize < PAGE_SIZE)
+		reason = "bitmap chunksize too small";
 	else if ((1 << ffz(~chunksize)) != chunksize)
 		reason = "bitmap chunksize not a power of 2";
-	else if (daemon_sleep < 1 || daemon_sleep > 15)
-		reason = "daemon sleep period out of range (1-15s)";
+	else if (daemon_sleep < 1 || daemon_sleep > MAX_SCHEDULE_TIMEOUT / HZ)
+		reason = "daemon sleep period out of range";
 	else if (write_behind > COUNTER_MAX)
 		reason = "write-behind limit out of range (0 - 16383)";
 	if (reason) {
@@ -535,7 +537,7 @@ success:
 		bitmap->events_cleared = bitmap->mddev->events;
 	err = 0;
 out:
-	kunmap(bitmap->sb_page);
+	kunmap_atomic(sb, KM_USER0);
 	if (err)
 		bitmap_print_sb(bitmap);
 	return err;
@@ -558,9 +560,9 @@ static void bitmap_mask_state(struct bitmap *bitmap, enum bitmap_state bits,
 		spin_unlock_irqrestore(&bitmap->lock, flags);
 		return;
 	}
-	page_cache_get(bitmap->sb_page);
+	get_page(bitmap->sb_page);
 	spin_unlock_irqrestore(&bitmap->lock, flags);
-	sb = (bitmap_super_t *)kmap(bitmap->sb_page);
+	sb = (bitmap_super_t *)kmap_atomic(bitmap->sb_page, KM_USER0);
 	switch (op) {
 		case MASK_SET: sb->state |= bits;
 				break;
@@ -568,8 +570,8 @@ static void bitmap_mask_state(struct bitmap *bitmap, enum bitmap_state bits,
 				break;
 		default: BUG();
 	}
-	kunmap(bitmap->sb_page);
-	page_cache_release(bitmap->sb_page);
+	kunmap_atomic(sb, KM_USER0);
+	put_page(bitmap->sb_page);
 }
 
 /*
@@ -622,12 +624,11 @@ static void bitmap_file_unmap(struct bitmap *bitmap)
 
 	while (pages--)
 		if (map[pages]->index != 0) /* 0 is sb_page, release it below */
-			page_cache_release(map[pages]);
+			put_page(map[pages]);
 	kfree(map);
 	kfree(attr);
 
-	if (sb_page)
-		page_cache_release(sb_page);
+	safe_put_page(sb_page);
 }
 
 static void bitmap_stop_daemon(struct bitmap *bitmap);
@@ -654,7 +655,7 @@ static void drain_write_queues(struct bitmap *bitmap)
 
 	while ((item = dequeue_page(bitmap))) {
 		/* don't bother to wait */
-		page_cache_release(item->page);
+		put_page(item->page);
 		mempool_free(item, bitmap->write_pool);
 	}
 
@@ -763,7 +764,7 @@ static void bitmap_file_set_bit(struct bitmap *bitmap, sector_t block)
 
 	/* make sure the page stays cached until it gets written out */
 	if (! (get_page_attr(bitmap, page) & BITMAP_PAGE_DIRTY))
-		page_cache_get(page);
+		get_page(page);
 
  	/* set the bit */
 	kaddr = kmap_atomic(page, KM_USER0);
@@ -854,6 +855,7 @@ static int bitmap_init_from_disk(struct bitmap *bitmap, sector_t start)
 	unsigned long bytes, offset, dummy;
 	int outofdate;
 	int ret = -ENOSPC;
+	void *paddr;
 
 	chunks = bitmap->chunks;
 	file = bitmap->file;
@@ -887,12 +889,10 @@ static int bitmap_init_from_disk(struct bitmap *bitmap, sector_t start)
 	if (!bitmap->filemap)
 		goto out;
 
-	bitmap->filemap_attr = kmalloc(sizeof(long) * num_pages, GFP_KERNEL);
+	bitmap->filemap_attr = kzalloc(sizeof(long) * num_pages, GFP_KERNEL);
 	if (!bitmap->filemap_attr)
 		goto out;
 
-	memset(bitmap->filemap_attr, 0, sizeof(long) * num_pages);
-
 	oldindex = ~0L;
 
 	for (i = 0; i < chunks; i++) {
@@ -901,8 +901,6 @@ static int bitmap_init_from_disk(struct bitmap *bitmap, sector_t start)
 		bit = file_page_offset(i);
 		if (index != oldindex) { /* this is a new page, read it in */
 			/* unmap the old page, we're done with it */
-			if (oldpage != NULL)
-				kunmap(oldpage);
 			if (index == 0) {
 				/*
 				 * if we're here then the superblock page
@@ -925,30 +923,32 @@ static int bitmap_init_from_disk(struct bitmap *bitmap, sector_t start)
 
 			oldindex = index;
 			oldpage = page;
-			kmap(page);
 
 			if (outofdate) {
 				/*
 				 * if bitmap is out of date, dirty the
 			 	 * whole page and write it out
 				 */
-				memset(page_address(page) + offset, 0xff,
+				paddr = kmap_atomic(page, KM_USER0);
+				memset(paddr + offset, 0xff,
 				       PAGE_SIZE - offset);
+				kunmap_atomic(paddr, KM_USER0);
 				ret = write_page(bitmap, page, 1);
 				if (ret) {
-					kunmap(page);
 					/* release, page not in filemap yet */
-					page_cache_release(page);
+					put_page(page);
 					goto out;
 				}
 			}
 
 			bitmap->filemap[bitmap->file_pages++] = page;
 		}
+		paddr = kmap_atomic(page, KM_USER0);
 		if (bitmap->flags & BITMAP_HOSTENDIAN)
-			b = test_bit(bit, page_address(page));
+			b = test_bit(bit, paddr);
 		else
-			b = ext2_test_bit(bit, page_address(page));
+			b = ext2_test_bit(bit, paddr);
+		kunmap_atomic(paddr, KM_USER0);
 		if (b) {
 			/* if the disk bit is set, set the memory bit */
 			bitmap_set_memory_bits(bitmap, i << CHUNK_BLOCK_SHIFT(bitmap),
@@ -963,9 +963,6 @@ static int bitmap_init_from_disk(struct bitmap *bitmap, sector_t start)
 	ret = 0;
 	bitmap_mask_state(bitmap, BITMAP_STALE, MASK_UNSET);
 
-	if (page) /* unmap the last page */
-		kunmap(page);
-
 	if (bit_cnt) { /* Kick recovery if any bits were set */
 		set_bit(MD_RECOVERY_NEEDED, &bitmap->mddev->recovery);
 		md_wakeup_thread(bitmap->mddev->thread);
@@ -1021,6 +1018,7 @@ int bitmap_daemon_work(struct bitmap *bitmap)
 	int err = 0;
 	int blocks;
 	int attr;
+	void *paddr;
 
 	if (bitmap == NULL)
 		return 0;
@@ -1043,7 +1041,7 @@ int bitmap_daemon_work(struct bitmap *bitmap)
 			/* skip this page unless it's marked as needing cleaning */
 			if (!((attr=get_page_attr(bitmap, page)) & BITMAP_PAGE_CLEAN)) {
 				if (attr & BITMAP_PAGE_NEEDWRITE) {
-					page_cache_get(page);
+					get_page(page);
 					clear_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
 				}
 				spin_unlock_irqrestore(&bitmap->lock, flags);
@@ -1057,13 +1055,13 @@ int bitmap_daemon_work(struct bitmap *bitmap)
 					default:
 						bitmap_file_kick(bitmap);
 					}
-					page_cache_release(page);
+					put_page(page);
 				}
 				continue;
 			}
 
 			/* grab the new page, sync and release the old */
-			page_cache_get(page);
+			get_page(page);
 			if (lastpage != NULL) {
 				if (get_page_attr(bitmap, lastpage) & BITMAP_PAGE_NEEDWRITE) {
 					clear_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
@@ -1077,14 +1075,12 @@ int bitmap_daemon_work(struct bitmap *bitmap)
 					set_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
 					spin_unlock_irqrestore(&bitmap->lock, flags);
 				}
-				kunmap(lastpage);
-				page_cache_release(lastpage);
+				put_page(lastpage);
 				if (err)
 					bitmap_file_kick(bitmap);
 			} else
 				spin_unlock_irqrestore(&bitmap->lock, flags);
 			lastpage = page;
-			kmap(page);
 /*
 			printk("bitmap clean at page %lu\n", j);
 */
@@ -1107,10 +1103,12 @@ int bitmap_daemon_work(struct bitmap *bitmap)
 						  -1);
 
 				/* clear the bit */
+				paddr = kmap_atomic(page, KM_USER0);
 				if (bitmap->flags & BITMAP_HOSTENDIAN)
-					clear_bit(file_page_offset(j), page_address(page));
+					clear_bit(file_page_offset(j), paddr);
 				else
-					ext2_clear_bit(file_page_offset(j), page_address(page));
+					ext2_clear_bit(file_page_offset(j), paddr);
+				kunmap_atomic(paddr, KM_USER0);
 			}
 		}
 		spin_unlock_irqrestore(&bitmap->lock, flags);
@@ -1118,7 +1116,6 @@ int bitmap_daemon_work(struct bitmap *bitmap)
 
 	/* now sync the final page */
 	if (lastpage != NULL) {
-		kunmap(lastpage);
 		spin_lock_irqsave(&bitmap->lock, flags);
 		if (get_page_attr(bitmap, lastpage) &BITMAP_PAGE_NEEDWRITE) {
 			clear_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
@@ -1133,7 +1130,7 @@ int bitmap_daemon_work(struct bitmap *bitmap)
 			spin_unlock_irqrestore(&bitmap->lock, flags);
 		}
 
-		page_cache_release(lastpage);
+		put_page(lastpage);
 	}
 
 	return err;
@@ -1184,7 +1181,7 @@ static void bitmap_writeback_daemon(mddev_t *mddev)
 		PRINTK("finished page writeback: %p\n", page);
 
 		err = PageError(page);
-		page_cache_release(page);
+		put_page(page);
 		if (err) {
 			printk(KERN_WARNING "%s: bitmap file writeback "
 			       "failed (page %lu): %d\n",
@@ -1530,6 +1527,8 @@ void bitmap_destroy(mddev_t *mddev)
 		return;
 
 	mddev->bitmap = NULL; /* disconnect from the md device */
+	if (mddev->thread)
+		mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
 
 	bitmap_free(bitmap);
 }
@@ -1555,12 +1554,10 @@ int bitmap_create(mddev_t *mddev)
 
 	BUG_ON(file && mddev->bitmap_offset);
 
-	bitmap = kmalloc(sizeof(*bitmap), GFP_KERNEL);
+	bitmap = kzalloc(sizeof(*bitmap), GFP_KERNEL);
 	if (!bitmap)
 		return -ENOMEM;
 
-	memset(bitmap, 0, sizeof(*bitmap));
-
 	spin_lock_init(&bitmap->lock);
 	bitmap->mddev = mddev;
 
@@ -1601,12 +1598,11 @@ int bitmap_create(mddev_t *mddev)
 #ifdef INJECT_FATAL_FAULT_1
 	bitmap->bp = NULL;
 #else
-	bitmap->bp = kmalloc(pages * sizeof(*bitmap->bp), GFP_KERNEL);
+	bitmap->bp = kzalloc(pages * sizeof(*bitmap->bp), GFP_KERNEL);
 #endif
 	err = -ENOMEM;
 	if (!bitmap->bp)
 		goto error;
-	memset(bitmap->bp, 0, pages * sizeof(*bitmap->bp));
 
 	bitmap->flags |= BITMAP_ACTIVE;
 
@@ -1636,6 +1632,8 @@ int bitmap_create(mddev_t *mddev)
 
 	if (IS_ERR(bitmap->writeback_daemon))
 		return PTR_ERR(bitmap->writeback_daemon);
+	mddev->thread->timeout = bitmap->daemon_sleep * HZ;
+
 	return bitmap_update_sb(bitmap);
 
  error:
diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index cf663105668..a601a427885 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -690,6 +690,8 @@ bad3:
 bad2:
 	crypto_free_tfm(tfm);
 bad1:
+	/* Must zero key material before freeing */
+	memset(cc, 0, sizeof(*cc) + cc->key_size * sizeof(u8));
 	kfree(cc);
 	return -EINVAL;
 }
@@ -706,6 +708,9 @@ static void crypt_dtr(struct dm_target *ti)
 		cc->iv_gen_ops->dtr(cc);
 	crypto_free_tfm(cc->tfm);
 	dm_put_device(ti, cc->dev);
+
+	/* Must zero key material before freeing */
+	memset(cc, 0, sizeof(*cc) + cc->key_size * sizeof(u8));
 	kfree(cc);
 }
 
diff --git a/drivers/md/dm-io.h b/drivers/md/dm-io.h
index 1a77f326570..f9035bfd1a9 100644
--- a/drivers/md/dm-io.h
+++ b/drivers/md/dm-io.h
@@ -9,9 +9,6 @@
 
 #include "dm.h"
 
-/* FIXME make this configurable */
-#define DM_MAX_IO_REGIONS 8
-
 struct io_region {
 	struct block_device *bdev;
 	sector_t sector;
diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index 07d44e19536..561bda5011e 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -270,6 +270,7 @@ static int dm_hash_rename(const char *old, const char *new)
 {
 	char *new_name, *old_name;
 	struct hash_cell *hc;
+	struct dm_table *table;
 
 	/*
 	 * duplicate new.
@@ -317,6 +318,15 @@ static int dm_hash_rename(const char *old, const char *new)
 	/* rename the device node in devfs */
 	register_with_devfs(hc);
 
+	/*
+	 * Wake up any dm event waiters.
+	 */
+	table = dm_get_table(hc->md);
+	if (table) {
+		dm_table_event(table);
+		dm_table_put(table);
+	}
+
 	up_write(&_hash_lock);
 	kfree(old_name);
 	return 0;
@@ -683,14 +693,18 @@ static int dev_rename(struct dm_ioctl *param, size_t param_size)
 static int do_suspend(struct dm_ioctl *param)
 {
 	int r = 0;
+	int do_lockfs = 1;
 	struct mapped_device *md;
 
 	md = find_device(param);
 	if (!md)
 		return -ENXIO;
 
+	if (param->flags & DM_SKIP_LOCKFS_FLAG)
+		do_lockfs = 0;
+
 	if (!dm_suspended(md))
-		r = dm_suspend(md);
+		r = dm_suspend(md, do_lockfs);
 
 	if (!r)
 		r = __dev_status(md, param);
@@ -702,6 +716,7 @@ static int do_suspend(struct dm_ioctl *param)
 static int do_resume(struct dm_ioctl *param)
 {
 	int r = 0;
+	int do_lockfs = 1;
 	struct hash_cell *hc;
 	struct mapped_device *md;
 	struct dm_table *new_map;
@@ -727,8 +742,10 @@ static int do_resume(struct dm_ioctl *param)
 	/* Do we need to load a new map ? */
 	if (new_map) {
 		/* Suspend if it isn't already suspended */
+		if (param->flags & DM_SKIP_LOCKFS_FLAG)
+			do_lockfs = 0;
 		if (!dm_suspended(md))
-			dm_suspend(md);
+			dm_suspend(md, do_lockfs);
 
 		r = dm_swap_table(md, new_map);
 		if (r) {
diff --git a/drivers/md/dm-log.c b/drivers/md/dm-log.c
index a76349cb10a..efe4adf7853 100644
--- a/drivers/md/dm-log.c
+++ b/drivers/md/dm-log.c
@@ -573,7 +573,7 @@ static int core_get_resync_work(struct dirty_log *log, region_t *region)
 					     lc->sync_search);
 		lc->sync_search = *region + 1;
 
-		if (*region == lc->region_count)
+		if (*region >= lc->region_count)
 			return 0;
 
 	} while (log_test_bit(lc->recovering_bits, *region));
diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c
index 6b0fc167092..6cfa8d435d5 100644
--- a/drivers/md/dm-raid1.c
+++ b/drivers/md/dm-raid1.c
@@ -562,6 +562,8 @@ struct mirror_set {
 	region_t nr_regions;
 	int in_sync;
 
+	struct mirror *default_mirror;	/* Default mirror */
+
 	unsigned int nr_mirrors;
 	struct mirror mirror[0];
 };
@@ -611,7 +613,7 @@ static int recover(struct mirror_set *ms, struct region *reg)
 	unsigned long flags = 0;
 
 	/* fill in the source */
-	m = ms->mirror + DEFAULT_MIRROR;
+	m = ms->default_mirror;
 	from.bdev = m->dev->bdev;
 	from.sector = m->offset + region_to_sector(reg->rh, reg->key);
 	if (reg->key == (ms->nr_regions - 1)) {
@@ -627,7 +629,7 @@ static int recover(struct mirror_set *ms, struct region *reg)
 
 	/* fill in the destinations */
 	for (i = 0, dest = to; i < ms->nr_mirrors; i++) {
-		if (i == DEFAULT_MIRROR)
+		if (&ms->mirror[i] == ms->default_mirror)
 			continue;
 
 		m = ms->mirror + i;
@@ -682,7 +684,7 @@ static void do_recovery(struct mirror_set *ms)
 static struct mirror *choose_mirror(struct mirror_set *ms, sector_t sector)
 {
 	/* FIXME: add read balancing */
-	return ms->mirror + DEFAULT_MIRROR;
+	return ms->default_mirror;
 }
 
 /*
@@ -709,7 +711,7 @@ static void do_reads(struct mirror_set *ms, struct bio_list *reads)
 		if (rh_in_sync(&ms->rh, region, 0))
 			m = choose_mirror(ms, bio->bi_sector);
 		else
-			m = ms->mirror + DEFAULT_MIRROR;
+			m = ms->default_mirror;
 
 		map_bio(ms, m, bio);
 		generic_make_request(bio);
@@ -833,7 +835,7 @@ static void do_writes(struct mirror_set *ms, struct bio_list *writes)
 		rh_delay(&ms->rh, bio);
 
 	while ((bio = bio_list_pop(&nosync))) {
-		map_bio(ms, ms->mirror + DEFAULT_MIRROR, bio);
+		map_bio(ms, ms->default_mirror, bio);
 		generic_make_request(bio);
 	}
 }
@@ -900,6 +902,7 @@ static struct mirror_set *alloc_context(unsigned int nr_mirrors,
 	ms->nr_mirrors = nr_mirrors;
 	ms->nr_regions = dm_sector_div_up(ti->len, region_size);
 	ms->in_sync = 0;
+	ms->default_mirror = &ms->mirror[DEFAULT_MIRROR];
 
 	if (rh_init(&ms->rh, ms, dl, region_size, ms->nr_regions)) {
 		ti->error = "dm-mirror: Error creating dirty region hash";
diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index ab54f99b7c3..4b9dd8fb1e5 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -371,6 +371,20 @@ static inline ulong round_up(ulong n, ulong size)
 	return (n + size) & ~size;
 }
 
+static void read_snapshot_metadata(struct dm_snapshot *s)
+{
+	if (s->have_metadata)
+		return;
+
+	if (s->store.read_metadata(&s->store)) {
+		down_write(&s->lock);
+		s->valid = 0;
+		up_write(&s->lock);
+	}
+
+	s->have_metadata = 1;
+}
+
 /*
  * Construct a snapshot mapping: <origin_dev> <COW-dev> <p/n> <chunk-size>
  */
@@ -848,16 +862,7 @@ static void snapshot_resume(struct dm_target *ti)
 {
 	struct dm_snapshot *s = (struct dm_snapshot *) ti->private;
 
-	if (s->have_metadata)
-		return;
-
-	if (s->store.read_metadata(&s->store)) {
-		down_write(&s->lock);
-		s->valid = 0;
-		up_write(&s->lock);
-	}
-
-	s->have_metadata = 1;
+	read_snapshot_metadata(s);
 }
 
 static int snapshot_status(struct dm_target *ti, status_type_t type,
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index a6d3baa46f6..a6f2dc66c3d 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -638,7 +638,7 @@ int dm_split_args(int *argc, char ***argvp, char *input)
 static void check_for_valid_limits(struct io_restrictions *rs)
 {
 	if (!rs->max_sectors)
-		rs->max_sectors = MAX_SECTORS;
+		rs->max_sectors = SAFE_MAX_SECTORS;
 	if (!rs->max_phys_segments)
 		rs->max_phys_segments = MAX_PHYS_SEGMENTS;
 	if (!rs->max_hw_segments)
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 930b9fc2795..0e481512f91 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -55,6 +55,7 @@ union map_info *dm_get_mapinfo(struct bio *bio)
  */
 #define DMF_BLOCK_IO 0
 #define DMF_SUSPENDED 1
+#define DMF_FROZEN 2
 
 struct mapped_device {
 	struct rw_semaphore io_lock;
@@ -97,7 +98,7 @@ struct mapped_device {
 	 * freeze/thaw support require holding onto a super block
 	 */
 	struct super_block *frozen_sb;
-	struct block_device *frozen_bdev;
+	struct block_device *suspended_bdev;
 };
 
 #define MIN_IOS 256
@@ -836,9 +837,9 @@ static void __set_size(struct mapped_device *md, sector_t size)
 {
 	set_capacity(md->disk, size);
 
-	down(&md->frozen_bdev->bd_inode->i_sem);
-	i_size_write(md->frozen_bdev->bd_inode, (loff_t)size << SECTOR_SHIFT);
-	up(&md->frozen_bdev->bd_inode->i_sem);
+	down(&md->suspended_bdev->bd_inode->i_sem);
+	i_size_write(md->suspended_bdev->bd_inode, (loff_t)size << SECTOR_SHIFT);
+	up(&md->suspended_bdev->bd_inode->i_sem);
 }
 
 static int __bind(struct mapped_device *md, struct dm_table *t)
@@ -902,10 +903,9 @@ int dm_create_with_minor(unsigned int minor, struct mapped_device **result)
 	return create_aux(minor, 1, result);
 }
 
-void *dm_get_mdptr(dev_t dev)
+static struct mapped_device *dm_find_md(dev_t dev)
 {
 	struct mapped_device *md;
-	void *mdptr = NULL;
 	unsigned minor = MINOR(dev);
 
 	if (MAJOR(dev) != _major || minor >= (1 << MINORBITS))
@@ -914,12 +914,32 @@ void *dm_get_mdptr(dev_t dev)
 	down(&_minor_lock);
 
 	md = idr_find(&_minor_idr, minor);
-
-	if (md && (dm_disk(md)->first_minor == minor))
-		mdptr = md->interface_ptr;
+	if (!md || (dm_disk(md)->first_minor != minor))
+		md = NULL;
 
 	up(&_minor_lock);
 
+	return md;
+}
+
+struct mapped_device *dm_get_md(dev_t dev)
+{
+	struct mapped_device *md = dm_find_md(dev);
+
+	if (md)
+		dm_get(md);
+
+	return md;
+}
+
+void *dm_get_mdptr(dev_t dev)
+{
+	struct mapped_device *md;
+	void *mdptr = NULL;
+
+	md = dm_find_md(dev);
+	if (md)
+		mdptr = md->interface_ptr;
 	return mdptr;
 }
 
@@ -991,43 +1011,33 @@ out:
  */
 static int lock_fs(struct mapped_device *md)
 {
-	int r = -ENOMEM;
-
-	md->frozen_bdev = bdget_disk(md->disk, 0);
-	if (!md->frozen_bdev) {
-		DMWARN("bdget failed in lock_fs");
-		goto out;
-	}
+	int r;
 
 	WARN_ON(md->frozen_sb);
 
-	md->frozen_sb = freeze_bdev(md->frozen_bdev);
+	md->frozen_sb = freeze_bdev(md->suspended_bdev);
 	if (IS_ERR(md->frozen_sb)) {
 		r = PTR_ERR(md->frozen_sb);
-		goto out_bdput;
+		md->frozen_sb = NULL;
+		return r;
 	}
 
+	set_bit(DMF_FROZEN, &md->flags);
+
 	/* don't bdput right now, we don't want the bdev
-	 * to go away while it is locked.  We'll bdput
-	 * in unlock_fs
+	 * to go away while it is locked.
 	 */
 	return 0;
-
-out_bdput:
-	bdput(md->frozen_bdev);
-	md->frozen_sb = NULL;
-	md->frozen_bdev = NULL;
-out:
-	return r;
 }
 
 static void unlock_fs(struct mapped_device *md)
 {
-	thaw_bdev(md->frozen_bdev, md->frozen_sb);
-	bdput(md->frozen_bdev);
+	if (!test_bit(DMF_FROZEN, &md->flags))
+		return;
 
+	thaw_bdev(md->suspended_bdev, md->frozen_sb);
 	md->frozen_sb = NULL;
-	md->frozen_bdev = NULL;
+	clear_bit(DMF_FROZEN, &md->flags);
 }
 
 /*
@@ -1037,7 +1047,7 @@ static void unlock_fs(struct mapped_device *md)
  * dm_bind_table, dm_suspend must be called to flush any in
  * flight bios and ensure that any further io gets deferred.
  */
-int dm_suspend(struct mapped_device *md)
+int dm_suspend(struct mapped_device *md, int do_lockfs)
 {
 	struct dm_table *map = NULL;
 	DECLARE_WAITQUEUE(wait, current);
@@ -1053,10 +1063,19 @@ int dm_suspend(struct mapped_device *md)
 	/* This does not get reverted if there's an error later. */
 	dm_table_presuspend_targets(map);
 
-	/* Flush I/O to the device. */
-	r = lock_fs(md);
-	if (r)
+	md->suspended_bdev = bdget_disk(md->disk, 0);
+	if (!md->suspended_bdev) {
+		DMWARN("bdget failed in dm_suspend");
+		r = -ENOMEM;
 		goto out;
+	}
+
+	/* Flush I/O to the device. */
+	if (do_lockfs) {
+		r = lock_fs(md);
+		if (r)
+			goto out;
+	}
 
 	/*
 	 * First we set the BLOCK_IO flag so no more ios will be mapped.
@@ -1105,6 +1124,11 @@ int dm_suspend(struct mapped_device *md)
 	r = 0;
 
 out:
+	if (r && md->suspended_bdev) {
+		bdput(md->suspended_bdev);
+		md->suspended_bdev = NULL;
+	}
+
 	dm_table_put(map);
 	up(&md->suspend_lock);
 	return r;
@@ -1135,6 +1159,9 @@ int dm_resume(struct mapped_device *md)
 
 	unlock_fs(md);
 
+	bdput(md->suspended_bdev);
+	md->suspended_bdev = NULL;
+
 	clear_bit(DMF_SUSPENDED, &md->flags);
 
 	dm_table_unplug_all(map);
diff --git a/drivers/md/dm.h b/drivers/md/dm.h
index e38c3fc1a1d..4eaf075da21 100644
--- a/drivers/md/dm.h
+++ b/drivers/md/dm.h
@@ -28,7 +28,7 @@
  * in types.h.
  */
 #ifdef CONFIG_LBD
-#define SECTOR_FORMAT "%Lu"
+#define SECTOR_FORMAT "%llu"
 #else
 #define SECTOR_FORMAT "%lu"
 #endif
@@ -58,6 +58,7 @@ int dm_create(struct mapped_device **md);
 int dm_create_with_minor(unsigned int minor, struct mapped_device **md);
 void dm_set_mdptr(struct mapped_device *md, void *ptr);
 void *dm_get_mdptr(dev_t dev);
+struct mapped_device *dm_get_md(dev_t dev);
 
 /*
  * Reference counting for md.
@@ -68,7 +69,7 @@ void dm_put(struct mapped_device *md);
 /*
  * A device can still be used while suspended, but I/O is deferred.
  */
-int dm_suspend(struct mapped_device *md);
+int dm_suspend(struct mapped_device *md, int with_lockfs);
 int dm_resume(struct mapped_device *md);
 
 /*
diff --git a/drivers/md/faulty.c b/drivers/md/faulty.c
index 0248f8e7eac..a7a5ab55433 100644
--- a/drivers/md/faulty.c
+++ b/drivers/md/faulty.c
@@ -316,9 +316,10 @@ static int stop(mddev_t *mddev)
 	return 0;
 }
 
-static mdk_personality_t faulty_personality =
+static struct mdk_personality faulty_personality =
 {
 	.name		= "faulty",
+	.level		= LEVEL_FAULTY,
 	.owner		= THIS_MODULE,
 	.make_request	= make_request,
 	.run		= run,
@@ -329,15 +330,17 @@ static mdk_personality_t faulty_personality =
 
 static int __init raid_init(void)
 {
-	return register_md_personality(FAULTY, &faulty_personality);
+	return register_md_personality(&faulty_personality);
 }
 
 static void raid_exit(void)
 {
-	unregister_md_personality(FAULTY);
+	unregister_md_personality(&faulty_personality);
 }
 
 module_init(raid_init);
 module_exit(raid_exit);
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("md-personality-10"); /* faulty */
+MODULE_ALIAS("md-faulty");
+MODULE_ALIAS("md-level--5");
diff --git a/drivers/md/kcopyd.c b/drivers/md/kcopyd.c
index eb703648597..ca99979c868 100644
--- a/drivers/md/kcopyd.c
+++ b/drivers/md/kcopyd.c
@@ -561,11 +561,13 @@ int kcopyd_copy(struct kcopyd_client *kc, struct io_region *from,
  * Cancels a kcopyd job, eg. someone might be deactivating a
  * mirror.
  */
+#if 0
 int kcopyd_cancel(struct kcopyd_job *job, int block)
 {
 	/* FIXME: finish */
 	return -1;
 }
+#endif  /*  0  */
 
 /*-----------------------------------------------------------------
  * Unit setup
@@ -684,4 +686,3 @@ void kcopyd_client_destroy(struct kcopyd_client *kc)
 EXPORT_SYMBOL(kcopyd_client_create);
 EXPORT_SYMBOL(kcopyd_client_destroy);
 EXPORT_SYMBOL(kcopyd_copy);
-EXPORT_SYMBOL(kcopyd_cancel);
diff --git a/drivers/md/linear.c b/drivers/md/linear.c
index 946efef3a8f..777585458c8 100644
--- a/drivers/md/linear.c
+++ b/drivers/md/linear.c
@@ -121,11 +121,10 @@ static int linear_run (mddev_t *mddev)
 	sector_t curr_offset;
 	struct list_head *tmp;
 
-	conf = kmalloc (sizeof (*conf) + mddev->raid_disks*sizeof(dev_info_t),
+	conf = kzalloc (sizeof (*conf) + mddev->raid_disks*sizeof(dev_info_t),
 			GFP_KERNEL);
 	if (!conf)
 		goto out;
-	memset(conf, 0, sizeof(*conf) + mddev->raid_disks*sizeof(dev_info_t));
 	mddev->private = conf;
 
 	cnt = 0;
@@ -352,9 +351,10 @@ static void linear_status (struct seq_file *seq, mddev_t *mddev)
 }
 
 
-static mdk_personality_t linear_personality=
+static struct mdk_personality linear_personality =
 {
 	.name		= "linear",
+	.level		= LEVEL_LINEAR,
 	.owner		= THIS_MODULE,
 	.make_request	= linear_make_request,
 	.run		= linear_run,
@@ -364,16 +364,18 @@ static mdk_personality_t linear_personality=
 
 static int __init linear_init (void)
 {
-	return register_md_personality (LINEAR, &linear_personality);
+	return register_md_personality (&linear_personality);
 }
 
 static void linear_exit (void)
 {
-	unregister_md_personality (LINEAR);
+	unregister_md_personality (&linear_personality);
 }
 
 
 module_init(linear_init);
 module_exit(linear_exit);
 MODULE_LICENSE("GPL");
-MODULE_ALIAS("md-personality-1"); /* LINEAR */
+MODULE_ALIAS("md-personality-1"); /* LINEAR - deprecated*/
+MODULE_ALIAS("md-linear");
+MODULE_ALIAS("md-level--1");
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 8175a2a222d..1b76fb29fb7 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -42,6 +42,7 @@
 #include <linux/devfs_fs_kernel.h>
 #include <linux/buffer_head.h> /* for invalidate_bdev */
 #include <linux/suspend.h>
+#include <linux/poll.h>
 
 #include <linux/init.h>
 
@@ -67,7 +68,7 @@
 static void autostart_arrays (int part);
 #endif
 
-static mdk_personality_t *pers[MAX_PERSONALITY];
+static LIST_HEAD(pers_list);
 static DEFINE_SPINLOCK(pers_lock);
 
 /*
@@ -80,10 +81,22 @@ static DEFINE_SPINLOCK(pers_lock);
  * idle IO detection.
  *
  * you can change it via /proc/sys/dev/raid/speed_limit_min and _max.
+ * or /sys/block/mdX/md/sync_speed_{min,max}
  */
 
 static int sysctl_speed_limit_min = 1000;
 static int sysctl_speed_limit_max = 200000;
+static inline int speed_min(mddev_t *mddev)
+{
+	return mddev->sync_speed_min ?
+		mddev->sync_speed_min : sysctl_speed_limit_min;
+}
+
+static inline int speed_max(mddev_t *mddev)
+{
+	return mddev->sync_speed_max ?
+		mddev->sync_speed_max : sysctl_speed_limit_max;
+}
 
 static struct ctl_table_header *raid_table_header;
 
@@ -134,6 +147,24 @@ static struct block_device_operations md_fops;
 static int start_readonly;
 
 /*
+ * We have a system wide 'event count' that is incremented
+ * on any 'interesting' event, and readers of /proc/mdstat
+ * can use 'poll' or 'select' to find out when the event
+ * count increases.
+ *
+ * Events are:
+ *  start array, stop array, error, add device, remove device,
+ *  start build, activate spare
+ */
+static DECLARE_WAIT_QUEUE_HEAD(md_event_waiters);
+static atomic_t md_event_count;
+static void md_new_event(mddev_t *mddev)
+{
+	atomic_inc(&md_event_count);
+	wake_up(&md_event_waiters);
+}
+
+/*
  * Enables to iterate over all existing md arrays
  * all_mddevs_lock protects this list.
  */
@@ -209,12 +240,10 @@ static mddev_t * mddev_find(dev_t unit)
 	}
 	spin_unlock(&all_mddevs_lock);
 
-	new = (mddev_t *) kmalloc(sizeof(*new), GFP_KERNEL);
+	new = kzalloc(sizeof(*new), GFP_KERNEL);
 	if (!new)
 		return NULL;
 
-	memset(new, 0, sizeof(*new));
-
 	new->unit = unit;
 	if (MAJOR(unit) == MD_MAJOR)
 		new->md_minor = MINOR(unit);
@@ -262,7 +291,7 @@ static inline void mddev_unlock(mddev_t * mddev)
 	md_wakeup_thread(mddev->thread);
 }
 
-mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr)
+static mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr)
 {
 	mdk_rdev_t * rdev;
 	struct list_head *tmp;
@@ -286,6 +315,18 @@ static mdk_rdev_t * find_rdev(mddev_t * mddev, dev_t dev)
 	return NULL;
 }
 
+static struct mdk_personality *find_pers(int level, char *clevel)
+{
+	struct mdk_personality *pers;
+	list_for_each_entry(pers, &pers_list, list) {
+		if (level != LEVEL_NONE && pers->level == level)
+			return pers;
+		if (strcmp(pers->name, clevel)==0)
+			return pers;
+	}
+	return NULL;
+}
+
 static inline sector_t calc_dev_sboffset(struct block_device *bdev)
 {
 	sector_t size = bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
@@ -320,7 +361,7 @@ static int alloc_disk_sb(mdk_rdev_t * rdev)
 static void free_disk_sb(mdk_rdev_t * rdev)
 {
 	if (rdev->sb_page) {
-		page_cache_release(rdev->sb_page);
+		put_page(rdev->sb_page);
 		rdev->sb_loaded = 0;
 		rdev->sb_page = NULL;
 		rdev->sb_offset = 0;
@@ -461,6 +502,7 @@ int sync_page_io(struct block_device *bdev, sector_t sector, int size,
 	bio_put(bio);
 	return ret;
 }
+EXPORT_SYMBOL_GPL(sync_page_io);
 
 static int read_disk_sb(mdk_rdev_t * rdev, int size)
 {
@@ -665,6 +707,10 @@ static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version
 	}
 	rdev->size = calc_dev_size(rdev, sb->chunk_size);
 
+	if (rdev->size < sb->size && sb->level > 1)
+		/* "this cannot possibly happen" ... */
+		ret = -EINVAL;
+
  abort:
 	return ret;
 }
@@ -688,6 +734,7 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 		mddev->ctime = sb->ctime;
 		mddev->utime = sb->utime;
 		mddev->level = sb->level;
+		mddev->clevel[0] = 0;
 		mddev->layout = sb->layout;
 		mddev->raid_disks = sb->raid_disks;
 		mddev->size = sb->size;
@@ -714,9 +761,10 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 
 		if (sb->state & (1<<MD_SB_BITMAP_PRESENT) &&
 		    mddev->bitmap_file == NULL) {
-			if (mddev->level != 1 && mddev->level != 5 && mddev->level != 6) {
+			if (mddev->level != 1 && mddev->level != 5 && mddev->level != 6
+			    && mddev->level != 10) {
 				/* FIXME use a better test */
-				printk(KERN_WARNING "md: bitmaps only support for raid1\n");
+				printk(KERN_WARNING "md: bitmaps not supported for this level.\n");
 				return -EINVAL;
 			}
 			mddev->bitmap_offset = mddev->default_bitmap_offset;
@@ -968,6 +1016,7 @@ static int super_1_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version)
 	}
 	rdev->preferred_minor = 0xffff;
 	rdev->data_offset = le64_to_cpu(sb->data_offset);
+	atomic_set(&rdev->corrected_errors, le32_to_cpu(sb->cnt_corrected_read));
 
 	rdev->sb_size = le32_to_cpu(sb->max_dev) * 2 + 256;
 	bmask = queue_hardsect_size(rdev->bdev->bd_disk->queue)-1;
@@ -1006,6 +1055,9 @@ static int super_1_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version)
 	rdev->size = le64_to_cpu(sb->data_size)/2;
 	if (le32_to_cpu(sb->chunksize))
 		rdev->size &= ~((sector_t)le32_to_cpu(sb->chunksize)/2 - 1);
+
+	if (le32_to_cpu(sb->size) > rdev->size*2)
+		return -EINVAL;
 	return 0;
 }
 
@@ -1023,6 +1075,7 @@ static int super_1_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 		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);
+		mddev->clevel[0] = 0;
 		mddev->layout = le32_to_cpu(sb->layout);
 		mddev->raid_disks = le32_to_cpu(sb->raid_disks);
 		mddev->size = le64_to_cpu(sb->size)/2;
@@ -1037,8 +1090,9 @@ static int super_1_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 
 		if ((le32_to_cpu(sb->feature_map) & MD_FEATURE_BITMAP_OFFSET) &&
 		    mddev->bitmap_file == NULL ) {
-			if (mddev->level != 1) {
-				printk(KERN_WARNING "md: bitmaps only supported for raid1\n");
+			if (mddev->level != 1 && mddev->level != 5 && mddev->level != 6
+			    && mddev->level != 10) {
+				printk(KERN_WARNING "md: bitmaps not supported for this level.\n");
 				return -EINVAL;
 			}
 			mddev->bitmap_offset = (__s32)le32_to_cpu(sb->bitmap_offset);
@@ -1105,6 +1159,8 @@ static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 	else
 		sb->resync_offset = cpu_to_le64(0);
 
+	sb->cnt_corrected_read = atomic_read(&rdev->corrected_errors);
+
 	if (mddev->bitmap && mddev->bitmap_file == NULL) {
 		sb->bitmap_offset = cpu_to_le32((__u32)mddev->bitmap_offset);
 		sb->feature_map = cpu_to_le32(MD_FEATURE_BITMAP_OFFSET);
@@ -1187,6 +1243,14 @@ static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 		MD_BUG();
 		return -EINVAL;
 	}
+	/* make sure rdev->size exceeds mddev->size */
+	if (rdev->size && (mddev->size == 0 || rdev->size < mddev->size)) {
+		if (mddev->pers)
+			/* Cannot change size, so fail */
+			return -ENOSPC;
+		else
+			mddev->size = rdev->size;
+	}
 	same_pdev = match_dev_unit(mddev, rdev);
 	if (same_pdev)
 		printk(KERN_WARNING
@@ -1496,6 +1560,26 @@ repeat:
 
 }
 
+/* words written to sysfs files may, or my not, be \n terminated.
+ * We want to accept with case. For this we use cmd_match.
+ */
+static int cmd_match(const char *cmd, const char *str)
+{
+	/* See if cmd, written into a sysfs file, matches
+	 * str.  They must either be the same, or cmd can
+	 * have a trailing newline
+	 */
+	while (*cmd && *str && *cmd == *str) {
+		cmd++;
+		str++;
+	}
+	if (*cmd == '\n')
+		cmd++;
+	if (*str || *cmd)
+		return 0;
+	return 1;
+}
+
 struct rdev_sysfs_entry {
 	struct attribute attr;
 	ssize_t (*show)(mdk_rdev_t *, char *);
@@ -1538,9 +1622,113 @@ super_show(mdk_rdev_t *rdev, char *page)
 }
 static struct rdev_sysfs_entry rdev_super = __ATTR_RO(super);
 
+static ssize_t
+errors_show(mdk_rdev_t *rdev, char *page)
+{
+	return sprintf(page, "%d\n", atomic_read(&rdev->corrected_errors));
+}
+
+static ssize_t
+errors_store(mdk_rdev_t *rdev, const char *buf, size_t len)
+{
+	char *e;
+	unsigned long n = simple_strtoul(buf, &e, 10);
+	if (*buf && (*e == 0 || *e == '\n')) {
+		atomic_set(&rdev->corrected_errors, n);
+		return len;
+	}
+	return -EINVAL;
+}
+static struct rdev_sysfs_entry rdev_errors =
+__ATTR(errors, 0644, errors_show, errors_store);
+
+static ssize_t
+slot_show(mdk_rdev_t *rdev, char *page)
+{
+	if (rdev->raid_disk < 0)
+		return sprintf(page, "none\n");
+	else
+		return sprintf(page, "%d\n", rdev->raid_disk);
+}
+
+static ssize_t
+slot_store(mdk_rdev_t *rdev, const char *buf, size_t len)
+{
+	char *e;
+	int slot = simple_strtoul(buf, &e, 10);
+	if (strncmp(buf, "none", 4)==0)
+		slot = -1;
+	else if (e==buf || (*e && *e!= '\n'))
+		return -EINVAL;
+	if (rdev->mddev->pers)
+		/* Cannot set slot in active array (yet) */
+		return -EBUSY;
+	if (slot >= rdev->mddev->raid_disks)
+		return -ENOSPC;
+	rdev->raid_disk = slot;
+	/* assume it is working */
+	rdev->flags = 0;
+	set_bit(In_sync, &rdev->flags);
+	return len;
+}
+
+
+static struct rdev_sysfs_entry rdev_slot =
+__ATTR(slot, 0644, slot_show, slot_store);
+
+static ssize_t
+offset_show(mdk_rdev_t *rdev, char *page)
+{
+	return sprintf(page, "%llu\n", (unsigned long long)rdev->data_offset);
+}
+
+static ssize_t
+offset_store(mdk_rdev_t *rdev, const char *buf, size_t len)
+{
+	char *e;
+	unsigned long long offset = simple_strtoull(buf, &e, 10);
+	if (e==buf || (*e && *e != '\n'))
+		return -EINVAL;
+	if (rdev->mddev->pers)
+		return -EBUSY;
+	rdev->data_offset = offset;
+	return len;
+}
+
+static struct rdev_sysfs_entry rdev_offset =
+__ATTR(offset, 0644, offset_show, offset_store);
+
+static ssize_t
+rdev_size_show(mdk_rdev_t *rdev, char *page)
+{
+	return sprintf(page, "%llu\n", (unsigned long long)rdev->size);
+}
+
+static ssize_t
+rdev_size_store(mdk_rdev_t *rdev, const char *buf, size_t len)
+{
+	char *e;
+	unsigned long long size = simple_strtoull(buf, &e, 10);
+	if (e==buf || (*e && *e != '\n'))
+		return -EINVAL;
+	if (rdev->mddev->pers)
+		return -EBUSY;
+	rdev->size = size;
+	if (size < rdev->mddev->size || rdev->mddev->size == 0)
+		rdev->mddev->size = size;
+	return len;
+}
+
+static struct rdev_sysfs_entry rdev_size =
+__ATTR(size, 0644, rdev_size_show, rdev_size_store);
+
 static struct attribute *rdev_default_attrs[] = {
 	&rdev_state.attr,
 	&rdev_super.attr,
+	&rdev_errors.attr,
+	&rdev_slot.attr,
+	&rdev_offset.attr,
+	&rdev_size.attr,
 	NULL,
 };
 static ssize_t
@@ -1598,12 +1786,11 @@ static mdk_rdev_t *md_import_device(dev_t newdev, int super_format, int super_mi
 	mdk_rdev_t *rdev;
 	sector_t size;
 
-	rdev = (mdk_rdev_t *) kmalloc(sizeof(*rdev), GFP_KERNEL);
+	rdev = kzalloc(sizeof(*rdev), GFP_KERNEL);
 	if (!rdev) {
 		printk(KERN_ERR "md: could not alloc mem for new device!\n");
 		return ERR_PTR(-ENOMEM);
 	}
-	memset(rdev, 0, sizeof(*rdev));
 
 	if ((err = alloc_disk_sb(rdev)))
 		goto abort_free;
@@ -1621,6 +1808,7 @@ static mdk_rdev_t *md_import_device(dev_t newdev, int super_format, int super_mi
 	rdev->data_offset = 0;
 	atomic_set(&rdev->nr_pending, 0);
 	atomic_set(&rdev->read_errors, 0);
+	atomic_set(&rdev->corrected_errors, 0);
 
 	size = rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
 	if (!size) {
@@ -1725,16 +1913,37 @@ static void analyze_sbs(mddev_t * mddev)
 static ssize_t
 level_show(mddev_t *mddev, char *page)
 {
-	mdk_personality_t *p = mddev->pers;
-	if (p == NULL && mddev->raid_disks == 0)
-		return 0;
-	if (mddev->level >= 0)
-		return sprintf(page, "raid%d\n", mddev->level);
-	else
+	struct mdk_personality *p = mddev->pers;
+	if (p)
 		return sprintf(page, "%s\n", p->name);
+	else if (mddev->clevel[0])
+		return sprintf(page, "%s\n", mddev->clevel);
+	else if (mddev->level != LEVEL_NONE)
+		return sprintf(page, "%d\n", mddev->level);
+	else
+		return 0;
+}
+
+static ssize_t
+level_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	int rv = len;
+	if (mddev->pers)
+		return -EBUSY;
+	if (len == 0)
+		return 0;
+	if (len >= sizeof(mddev->clevel))
+		return -ENOSPC;
+	strncpy(mddev->clevel, buf, len);
+	if (mddev->clevel[len-1] == '\n')
+		len--;
+	mddev->clevel[len] = 0;
+	mddev->level = LEVEL_NONE;
+	return rv;
 }
 
-static struct md_sysfs_entry md_level = __ATTR_RO(level);
+static struct md_sysfs_entry md_level =
+__ATTR(level, 0644, level_show, level_store);
 
 static ssize_t
 raid_disks_show(mddev_t *mddev, char *page)
@@ -1744,7 +1953,197 @@ raid_disks_show(mddev_t *mddev, char *page)
 	return sprintf(page, "%d\n", mddev->raid_disks);
 }
 
-static struct md_sysfs_entry md_raid_disks = __ATTR_RO(raid_disks);
+static int update_raid_disks(mddev_t *mddev, int raid_disks);
+
+static ssize_t
+raid_disks_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	/* can only set raid_disks if array is not yet active */
+	char *e;
+	int rv = 0;
+	unsigned long n = simple_strtoul(buf, &e, 10);
+
+	if (!*buf || (*e && *e != '\n'))
+		return -EINVAL;
+
+	if (mddev->pers)
+		rv = update_raid_disks(mddev, n);
+	else
+		mddev->raid_disks = n;
+	return rv ? rv : len;
+}
+static struct md_sysfs_entry md_raid_disks =
+__ATTR(raid_disks, 0644, raid_disks_show, raid_disks_store);
+
+static ssize_t
+chunk_size_show(mddev_t *mddev, char *page)
+{
+	return sprintf(page, "%d\n", mddev->chunk_size);
+}
+
+static ssize_t
+chunk_size_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	/* can only set chunk_size if array is not yet active */
+	char *e;
+	unsigned long n = simple_strtoul(buf, &e, 10);
+
+	if (mddev->pers)
+		return -EBUSY;
+	if (!*buf || (*e && *e != '\n'))
+		return -EINVAL;
+
+	mddev->chunk_size = n;
+	return len;
+}
+static struct md_sysfs_entry md_chunk_size =
+__ATTR(chunk_size, 0644, chunk_size_show, chunk_size_store);
+
+static ssize_t
+null_show(mddev_t *mddev, char *page)
+{
+	return -EINVAL;
+}
+
+static ssize_t
+new_dev_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	/* buf must be %d:%d\n? giving major and minor numbers */
+	/* The new device is added to the array.
+	 * If the array has a persistent superblock, we read the
+	 * superblock to initialise info and check validity.
+	 * Otherwise, only checking done is that in bind_rdev_to_array,
+	 * which mainly checks size.
+	 */
+	char *e;
+	int major = simple_strtoul(buf, &e, 10);
+	int minor;
+	dev_t dev;
+	mdk_rdev_t *rdev;
+	int err;
+
+	if (!*buf || *e != ':' || !e[1] || e[1] == '\n')
+		return -EINVAL;
+	minor = simple_strtoul(e+1, &e, 10);
+	if (*e && *e != '\n')
+		return -EINVAL;
+	dev = MKDEV(major, minor);
+	if (major != MAJOR(dev) ||
+	    minor != MINOR(dev))
+		return -EOVERFLOW;
+
+
+	if (mddev->persistent) {
+		rdev = md_import_device(dev, mddev->major_version,
+					mddev->minor_version);
+		if (!IS_ERR(rdev) && !list_empty(&mddev->disks)) {
+			mdk_rdev_t *rdev0 = list_entry(mddev->disks.next,
+						       mdk_rdev_t, same_set);
+			err = super_types[mddev->major_version]
+				.load_super(rdev, rdev0, mddev->minor_version);
+			if (err < 0)
+				goto out;
+		}
+	} else
+		rdev = md_import_device(dev, -1, -1);
+
+	if (IS_ERR(rdev))
+		return PTR_ERR(rdev);
+	err = bind_rdev_to_array(rdev, mddev);
+ out:
+	if (err)
+		export_rdev(rdev);
+	return err ? err : len;
+}
+
+static struct md_sysfs_entry md_new_device =
+__ATTR(new_dev, 0200, null_show, new_dev_store);
+
+static ssize_t
+size_show(mddev_t *mddev, char *page)
+{
+	return sprintf(page, "%llu\n", (unsigned long long)mddev->size);
+}
+
+static int update_size(mddev_t *mddev, unsigned long size);
+
+static ssize_t
+size_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	/* If array is inactive, we can reduce the component size, but
+	 * not increase it (except from 0).
+	 * If array is active, we can try an on-line resize
+	 */
+	char *e;
+	int err = 0;
+	unsigned long long size = simple_strtoull(buf, &e, 10);
+	if (!*buf || *buf == '\n' ||
+	    (*e && *e != '\n'))
+		return -EINVAL;
+
+	if (mddev->pers) {
+		err = update_size(mddev, size);
+		md_update_sb(mddev);
+	} else {
+		if (mddev->size == 0 ||
+		    mddev->size > size)
+			mddev->size = size;
+		else
+			err = -ENOSPC;
+	}
+	return err ? err : len;
+}
+
+static struct md_sysfs_entry md_size =
+__ATTR(component_size, 0644, size_show, size_store);
+
+
+/* Metdata version.
+ * This is either 'none' for arrays with externally managed metadata,
+ * or N.M for internally known formats
+ */
+static ssize_t
+metadata_show(mddev_t *mddev, char *page)
+{
+	if (mddev->persistent)
+		return sprintf(page, "%d.%d\n",
+			       mddev->major_version, mddev->minor_version);
+	else
+		return sprintf(page, "none\n");
+}
+
+static ssize_t
+metadata_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	int major, minor;
+	char *e;
+	if (!list_empty(&mddev->disks))
+		return -EBUSY;
+
+	if (cmd_match(buf, "none")) {
+		mddev->persistent = 0;
+		mddev->major_version = 0;
+		mddev->minor_version = 90;
+		return len;
+	}
+	major = simple_strtoul(buf, &e, 10);
+	if (e==buf || *e != '.')
+		return -EINVAL;
+	buf = e+1;
+	minor = simple_strtoul(buf, &e, 10);
+	if (e==buf || *e != '\n')
+		return -EINVAL;
+	if (major >= sizeof(super_types)/sizeof(super_types[0]) ||
+	    super_types[major].name == NULL)
+		return -ENOENT;
+	mddev->major_version = major;
+	mddev->minor_version = minor;
+	mddev->persistent = 1;
+	return len;
+}
+
+static struct md_sysfs_entry md_metadata =
+__ATTR(metadata_version, 0644, metadata_show, metadata_store);
 
 static ssize_t
 action_show(mddev_t *mddev, char *page)
@@ -1771,31 +2170,27 @@ action_store(mddev_t *mddev, const char *page, size_t len)
 	if (!mddev->pers || !mddev->pers->sync_request)
 		return -EINVAL;
 
-	if (strcmp(page, "idle")==0 || strcmp(page, "idle\n")==0) {
+	if (cmd_match(page, "idle")) {
 		if (mddev->sync_thread) {
 			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
 			md_unregister_thread(mddev->sync_thread);
 			mddev->sync_thread = NULL;
 			mddev->recovery = 0;
 		}
-		return len;
-	}
-
-	if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
-	    test_bit(MD_RECOVERY_NEEDED, &mddev->recovery))
+	} else if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
+		   test_bit(MD_RECOVERY_NEEDED, &mddev->recovery))
 		return -EBUSY;
-	if (strcmp(page, "resync")==0 || strcmp(page, "resync\n")==0 ||
-	    strcmp(page, "recover")==0 || strcmp(page, "recover\n")==0)
+	else if (cmd_match(page, "resync") || cmd_match(page, "recover"))
 		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	else {
-		if (strcmp(page, "check")==0 || strcmp(page, "check\n")==0)
+		if (cmd_match(page, "check"))
 			set_bit(MD_RECOVERY_CHECK, &mddev->recovery);
-		else if (strcmp(page, "repair")!=0 && strcmp(page, "repair\n")!=0)
+		else if (cmd_match(page, "repair"))
 			return -EINVAL;
 		set_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
 		set_bit(MD_RECOVERY_SYNC, &mddev->recovery);
-		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	}
+	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	md_wakeup_thread(mddev->thread);
 	return len;
 }
@@ -1814,15 +2209,107 @@ md_scan_mode = __ATTR(sync_action, S_IRUGO|S_IWUSR, action_show, action_store);
 static struct md_sysfs_entry
 md_mismatches = __ATTR_RO(mismatch_cnt);
 
+static ssize_t
+sync_min_show(mddev_t *mddev, char *page)
+{
+	return sprintf(page, "%d (%s)\n", speed_min(mddev),
+		       mddev->sync_speed_min ? "local": "system");
+}
+
+static ssize_t
+sync_min_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	int min;
+	char *e;
+	if (strncmp(buf, "system", 6)==0) {
+		mddev->sync_speed_min = 0;
+		return len;
+	}
+	min = simple_strtoul(buf, &e, 10);
+	if (buf == e || (*e && *e != '\n') || min <= 0)
+		return -EINVAL;
+	mddev->sync_speed_min = min;
+	return len;
+}
+
+static struct md_sysfs_entry md_sync_min =
+__ATTR(sync_speed_min, S_IRUGO|S_IWUSR, sync_min_show, sync_min_store);
+
+static ssize_t
+sync_max_show(mddev_t *mddev, char *page)
+{
+	return sprintf(page, "%d (%s)\n", speed_max(mddev),
+		       mddev->sync_speed_max ? "local": "system");
+}
+
+static ssize_t
+sync_max_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	int max;
+	char *e;
+	if (strncmp(buf, "system", 6)==0) {
+		mddev->sync_speed_max = 0;
+		return len;
+	}
+	max = simple_strtoul(buf, &e, 10);
+	if (buf == e || (*e && *e != '\n') || max <= 0)
+		return -EINVAL;
+	mddev->sync_speed_max = max;
+	return len;
+}
+
+static struct md_sysfs_entry md_sync_max =
+__ATTR(sync_speed_max, S_IRUGO|S_IWUSR, sync_max_show, sync_max_store);
+
+
+static ssize_t
+sync_speed_show(mddev_t *mddev, char *page)
+{
+	unsigned long resync, dt, db;
+	resync = (mddev->curr_resync - atomic_read(&mddev->recovery_active));
+	dt = ((jiffies - mddev->resync_mark) / HZ);
+	if (!dt) dt++;
+	db = resync - (mddev->resync_mark_cnt);
+	return sprintf(page, "%ld\n", db/dt/2); /* K/sec */
+}
+
+static struct md_sysfs_entry
+md_sync_speed = __ATTR_RO(sync_speed);
+
+static ssize_t
+sync_completed_show(mddev_t *mddev, char *page)
+{
+	unsigned long max_blocks, resync;
+
+	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
+		max_blocks = mddev->resync_max_sectors;
+	else
+		max_blocks = mddev->size << 1;
+
+	resync = (mddev->curr_resync - atomic_read(&mddev->recovery_active));
+	return sprintf(page, "%lu / %lu\n", resync, max_blocks);
+}
+
+static struct md_sysfs_entry
+md_sync_completed = __ATTR_RO(sync_completed);
+
 static struct attribute *md_default_attrs[] = {
 	&md_level.attr,
 	&md_raid_disks.attr,
+	&md_chunk_size.attr,
+	&md_size.attr,
+	&md_metadata.attr,
+	&md_new_device.attr,
 	NULL,
 };
 
 static struct attribute *md_redundancy_attrs[] = {
 	&md_scan_mode.attr,
 	&md_mismatches.attr,
+	&md_sync_min.attr,
+	&md_sync_max.attr,
+	&md_sync_speed.attr,
+	&md_sync_completed.attr,
 	NULL,
 };
 static struct attribute_group md_redundancy_group = {
@@ -1937,14 +2424,16 @@ static void md_safemode_timeout(unsigned long data)
 	md_wakeup_thread(mddev->thread);
 }
 
+static int start_dirty_degraded;
 
 static int do_md_run(mddev_t * mddev)
 {
-	int pnum, err;
+	int err;
 	int chunk_size;
 	struct list_head *tmp;
 	mdk_rdev_t *rdev;
 	struct gendisk *disk;
+	struct mdk_personality *pers;
 	char b[BDEVNAME_SIZE];
 
 	if (list_empty(&mddev->disks))
@@ -1961,20 +2450,8 @@ static int do_md_run(mddev_t * mddev)
 		analyze_sbs(mddev);
 
 	chunk_size = mddev->chunk_size;
-	pnum = level_to_pers(mddev->level);
 
-	if ((pnum != MULTIPATH) && (pnum != RAID1)) {
-		if (!chunk_size) {
-			/*
-			 * 'default chunksize' in the old md code used to
-			 * be PAGE_SIZE, baaad.
-			 * we abort here to be on the safe side. We don't
-			 * want to continue the bad practice.
-			 */
-			printk(KERN_ERR 
-				"no chunksize specified, see 'man raidtab'\n");
-			return -EINVAL;
-		}
+	if (chunk_size) {
 		if (chunk_size > MAX_CHUNK_SIZE) {
 			printk(KERN_ERR "too big chunk_size: %d > %d\n",
 				chunk_size, MAX_CHUNK_SIZE);
@@ -2010,10 +2487,10 @@ static int do_md_run(mddev_t * mddev)
 	}
 
 #ifdef CONFIG_KMOD
-	if (!pers[pnum])
-	{
-		request_module("md-personality-%d", pnum);
-	}
+	if (mddev->level != LEVEL_NONE)
+		request_module("md-level-%d", mddev->level);
+	else if (mddev->clevel[0])
+		request_module("md-%s", mddev->clevel);
 #endif
 
 	/*
@@ -2035,30 +2512,39 @@ static int do_md_run(mddev_t * mddev)
 		return -ENOMEM;
 
 	spin_lock(&pers_lock);
-	if (!pers[pnum] || !try_module_get(pers[pnum]->owner)) {
+	pers = find_pers(mddev->level, mddev->clevel);
+	if (!pers || !try_module_get(pers->owner)) {
 		spin_unlock(&pers_lock);
-		printk(KERN_WARNING "md: personality %d is not loaded!\n",
-		       pnum);
+		if (mddev->level != LEVEL_NONE)
+			printk(KERN_WARNING "md: personality for level %d is not loaded!\n",
+			       mddev->level);
+		else
+			printk(KERN_WARNING "md: personality for level %s is not loaded!\n",
+			       mddev->clevel);
 		return -EINVAL;
 	}
-
-	mddev->pers = pers[pnum];
+	mddev->pers = pers;
 	spin_unlock(&pers_lock);
+	mddev->level = pers->level;
+	strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
 
 	mddev->recovery = 0;
 	mddev->resync_max_sectors = mddev->size << 1; /* may be over-ridden by personality */
 	mddev->barriers_work = 1;
+	mddev->ok_start_degraded = start_dirty_degraded;
 
 	if (start_readonly)
 		mddev->ro = 2; /* read-only, but switch on first write */
 
-	/* before we start the array running, initialise the bitmap */
-	err = bitmap_create(mddev);
-	if (err)
-		printk(KERN_ERR "%s: failed to create bitmap (%d)\n",
-			mdname(mddev), err);
-	else
-		err = mddev->pers->run(mddev);
+	err = mddev->pers->run(mddev);
+	if (!err && mddev->pers->sync_request) {
+		err = bitmap_create(mddev);
+		if (err) {
+			printk(KERN_ERR "%s: failed to create bitmap (%d)\n",
+			       mdname(mddev), err);
+			mddev->pers->stop(mddev);
+		}
+	}
 	if (err) {
 		printk(KERN_ERR "md: pers->run() failed ...\n");
 		module_put(mddev->pers->owner);
@@ -2104,6 +2590,7 @@ static int do_md_run(mddev_t * mddev)
 	mddev->queue->make_request_fn = mddev->pers->make_request;
 
 	mddev->changed = 1;
+	md_new_event(mddev);
 	return 0;
 }
 
@@ -2231,6 +2718,7 @@ static int do_md_stop(mddev_t * mddev, int ro)
 		printk(KERN_INFO "md: %s switched to read-only mode.\n",
 			mdname(mddev));
 	err = 0;
+	md_new_event(mddev);
 out:
 	return err;
 }
@@ -2668,12 +3156,6 @@ static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 		if (info->state & (1<<MD_DISK_WRITEMOSTLY))
 			set_bit(WriteMostly, &rdev->flags);
 
-		err = bind_rdev_to_array(rdev, mddev);
-		if (err) {
-			export_rdev(rdev);
-			return err;
-		}
-
 		if (!mddev->persistent) {
 			printk(KERN_INFO "md: nonpersistent superblock ...\n");
 			rdev->sb_offset = rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
@@ -2681,8 +3163,11 @@ static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 			rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
 		rdev->size = calc_dev_size(rdev, mddev->chunk_size);
 
-		if (!mddev->size || (mddev->size > rdev->size))
-			mddev->size = rdev->size;
+		err = bind_rdev_to_array(rdev, mddev);
+		if (err) {
+			export_rdev(rdev);
+			return err;
+		}
 	}
 
 	return 0;
@@ -2705,6 +3190,7 @@ static int hot_remove_disk(mddev_t * mddev, dev_t dev)
 
 	kick_rdev_from_array(rdev);
 	md_update_sb(mddev);
+	md_new_event(mddev);
 
 	return 0;
 busy:
@@ -2753,15 +3239,6 @@ static int hot_add_disk(mddev_t * mddev, dev_t dev)
 	size = calc_dev_size(rdev, mddev->chunk_size);
 	rdev->size = size;
 
-	if (size < mddev->size) {
-		printk(KERN_WARNING 
-			"%s: disk size %llu blocks < array size %llu\n",
-			mdname(mddev), (unsigned long long)size,
-			(unsigned long long)mddev->size);
-		err = -ENOSPC;
-		goto abort_export;
-	}
-
 	if (test_bit(Faulty, &rdev->flags)) {
 		printk(KERN_WARNING 
 			"md: can not hot-add faulty %s disk to %s!\n",
@@ -2771,7 +3248,9 @@ static int hot_add_disk(mddev_t * mddev, dev_t dev)
 	}
 	clear_bit(In_sync, &rdev->flags);
 	rdev->desc_nr = -1;
-	bind_rdev_to_array(rdev, mddev);
+	err = bind_rdev_to_array(rdev, mddev);
+	if (err)
+		goto abort_export;
 
 	/*
 	 * The rest should better be atomic, we can have disk failures
@@ -2795,7 +3274,7 @@ static int hot_add_disk(mddev_t * mddev, dev_t dev)
 	 */
 	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	md_wakeup_thread(mddev->thread);
-
+	md_new_event(mddev);
 	return 0;
 
 abort_unbind_export:
@@ -2942,6 +3421,81 @@ static int set_array_info(mddev_t * mddev, mdu_array_info_t *info)
 	return 0;
 }
 
+static int update_size(mddev_t *mddev, unsigned long size)
+{
+	mdk_rdev_t * rdev;
+	int rv;
+	struct list_head *tmp;
+
+	if (mddev->pers->resize == NULL)
+		return -EINVAL;
+	/* The "size" is the amount of each device that is used.
+	 * This can only make sense for arrays with redundancy.
+	 * linear and raid0 always use whatever space is available
+	 * We can only consider changing the size if no resync
+	 * or reconstruction is happening, and if the new size
+	 * is acceptable. It must fit before the sb_offset or,
+	 * if that is <data_offset, it must fit before the
+	 * size of each device.
+	 * If size is zero, we find the largest size that fits.
+	 */
+	if (mddev->sync_thread)
+		return -EBUSY;
+	ITERATE_RDEV(mddev,rdev,tmp) {
+		sector_t avail;
+		int fit = (size == 0);
+		if (rdev->sb_offset > rdev->data_offset)
+			avail = (rdev->sb_offset*2) - rdev->data_offset;
+		else
+			avail = get_capacity(rdev->bdev->bd_disk)
+				- rdev->data_offset;
+		if (fit && (size == 0 || size > avail/2))
+			size = avail/2;
+		if (avail < ((sector_t)size << 1))
+			return -ENOSPC;
+	}
+	rv = mddev->pers->resize(mddev, (sector_t)size *2);
+	if (!rv) {
+		struct block_device *bdev;
+
+		bdev = bdget_disk(mddev->gendisk, 0);
+		if (bdev) {
+			down(&bdev->bd_inode->i_sem);
+			i_size_write(bdev->bd_inode, mddev->array_size << 10);
+			up(&bdev->bd_inode->i_sem);
+			bdput(bdev);
+		}
+	}
+	return rv;
+}
+
+static int update_raid_disks(mddev_t *mddev, int raid_disks)
+{
+	int rv;
+	/* change the number of raid disks */
+	if (mddev->pers->reshape == NULL)
+		return -EINVAL;
+	if (raid_disks <= 0 ||
+	    raid_disks >= mddev->max_disks)
+		return -EINVAL;
+	if (mddev->sync_thread)
+		return -EBUSY;
+	rv = mddev->pers->reshape(mddev, raid_disks);
+	if (!rv) {
+		struct block_device *bdev;
+
+		bdev = bdget_disk(mddev->gendisk, 0);
+		if (bdev) {
+			down(&bdev->bd_inode->i_sem);
+			i_size_write(bdev->bd_inode, mddev->array_size << 10);
+			up(&bdev->bd_inode->i_sem);
+			bdput(bdev);
+		}
+	}
+	return rv;
+}
+
+
 /*
  * update_array_info is used to change the configuration of an
  * on-line array.
@@ -2990,71 +3544,12 @@ static int update_array_info(mddev_t *mddev, mdu_array_info_t *info)
 		else
 			return mddev->pers->reconfig(mddev, info->layout, -1);
 	}
-	if (mddev->size != info->size) {
-		mdk_rdev_t * rdev;
-		struct list_head *tmp;
-		if (mddev->pers->resize == NULL)
-			return -EINVAL;
-		/* The "size" is the amount of each device that is used.
-		 * This can only make sense for arrays with redundancy.
-		 * linear and raid0 always use whatever space is available
-		 * We can only consider changing the size if no resync
-		 * or reconstruction is happening, and if the new size
-		 * is acceptable. It must fit before the sb_offset or,
-		 * if that is <data_offset, it must fit before the
-		 * size of each device.
-		 * If size is zero, we find the largest size that fits.
-		 */
-		if (mddev->sync_thread)
-			return -EBUSY;
-		ITERATE_RDEV(mddev,rdev,tmp) {
-			sector_t avail;
-			int fit = (info->size == 0);
-			if (rdev->sb_offset > rdev->data_offset)
-				avail = (rdev->sb_offset*2) - rdev->data_offset;
-			else
-				avail = get_capacity(rdev->bdev->bd_disk)
-					- rdev->data_offset;
-			if (fit && (info->size == 0 || info->size > avail/2))
-				info->size = avail/2;
-			if (avail < ((sector_t)info->size << 1))
-				return -ENOSPC;
-		}
-		rv = mddev->pers->resize(mddev, (sector_t)info->size *2);
-		if (!rv) {
-			struct block_device *bdev;
-
-			bdev = bdget_disk(mddev->gendisk, 0);
-			if (bdev) {
-				down(&bdev->bd_inode->i_sem);
-				i_size_write(bdev->bd_inode, mddev->array_size << 10);
-				up(&bdev->bd_inode->i_sem);
-				bdput(bdev);
-			}
-		}
-	}
-	if (mddev->raid_disks    != info->raid_disks) {
-		/* change the number of raid disks */
-		if (mddev->pers->reshape == NULL)
-			return -EINVAL;
-		if (info->raid_disks <= 0 ||
-		    info->raid_disks >= mddev->max_disks)
-			return -EINVAL;
-		if (mddev->sync_thread)
-			return -EBUSY;
-		rv = mddev->pers->reshape(mddev, info->raid_disks);
-		if (!rv) {
-			struct block_device *bdev;
-
-			bdev = bdget_disk(mddev->gendisk, 0);
-			if (bdev) {
-				down(&bdev->bd_inode->i_sem);
-				i_size_write(bdev->bd_inode, mddev->array_size << 10);
-				up(&bdev->bd_inode->i_sem);
-				bdput(bdev);
-			}
-		}
-	}
+	if (mddev->size != info->size)
+		rv = update_size(mddev, info->size);
+
+	if (mddev->raid_disks    != info->raid_disks)
+		rv = update_raid_disks(mddev, info->raid_disks);
+
 	if ((state ^ info->state) & (1<<MD_SB_BITMAP_PRESENT)) {
 		if (mddev->pers->quiesce == NULL)
 			return -EINVAL;
@@ -3476,11 +3971,10 @@ mdk_thread_t *md_register_thread(void (*run) (mddev_t *), mddev_t *mddev,
 {
 	mdk_thread_t *thread;
 
-	thread = kmalloc(sizeof(mdk_thread_t), GFP_KERNEL);
+	thread = kzalloc(sizeof(mdk_thread_t), GFP_KERNEL);
 	if (!thread)
 		return NULL;
 
-	memset(thread, 0, sizeof(mdk_thread_t));
 	init_waitqueue_head(&thread->wqueue);
 
 	thread->run = run;
@@ -3524,6 +4018,7 @@ void md_error(mddev_t *mddev, mdk_rdev_t *rdev)
 	set_bit(MD_RECOVERY_INTR, &mddev->recovery);
 	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	md_wakeup_thread(mddev->thread);
+	md_new_event(mddev);
 }
 
 /* seq_file implementation /proc/mdstat */
@@ -3664,24 +4159,29 @@ static void md_seq_stop(struct seq_file *seq, void *v)
 		mddev_put(mddev);
 }
 
+struct mdstat_info {
+	int event;
+};
+
 static int md_seq_show(struct seq_file *seq, void *v)
 {
 	mddev_t *mddev = v;
 	sector_t size;
 	struct list_head *tmp2;
 	mdk_rdev_t *rdev;
-	int i;
+	struct mdstat_info *mi = seq->private;
 	struct bitmap *bitmap;
 
 	if (v == (void*)1) {
+		struct mdk_personality *pers;
 		seq_printf(seq, "Personalities : ");
 		spin_lock(&pers_lock);
-		for (i = 0; i < MAX_PERSONALITY; i++)
-			if (pers[i])
-				seq_printf(seq, "[%s] ", pers[i]->name);
+		list_for_each_entry(pers, &pers_list, list)
+			seq_printf(seq, "[%s] ", pers->name);
 
 		spin_unlock(&pers_lock);
 		seq_printf(seq, "\n");
+		mi->event = atomic_read(&md_event_count);
 		return 0;
 	}
 	if (v == (void*)2) {
@@ -3790,47 +4290,68 @@ static struct seq_operations md_seq_ops = {
 static int md_seq_open(struct inode *inode, struct file *file)
 {
 	int error;
+	struct mdstat_info *mi = kmalloc(sizeof(*mi), GFP_KERNEL);
+	if (mi == NULL)
+		return -ENOMEM;
 
 	error = seq_open(file, &md_seq_ops);
+	if (error)
+		kfree(mi);
+	else {
+		struct seq_file *p = file->private_data;
+		p->private = mi;
+		mi->event = atomic_read(&md_event_count);
+	}
 	return error;
 }
 
+static int md_seq_release(struct inode *inode, struct file *file)
+{
+	struct seq_file *m = file->private_data;
+	struct mdstat_info *mi = m->private;
+	m->private = NULL;
+	kfree(mi);
+	return seq_release(inode, file);
+}
+
+static unsigned int mdstat_poll(struct file *filp, poll_table *wait)
+{
+	struct seq_file *m = filp->private_data;
+	struct mdstat_info *mi = m->private;
+	int mask;
+
+	poll_wait(filp, &md_event_waiters, wait);
+
+	/* always allow read */
+	mask = POLLIN | POLLRDNORM;
+
+	if (mi->event != atomic_read(&md_event_count))
+		mask |= POLLERR | POLLPRI;
+	return mask;
+}
+
 static struct file_operations md_seq_fops = {
 	.open           = md_seq_open,
 	.read           = seq_read,
 	.llseek         = seq_lseek,
-	.release	= seq_release,
+	.release	= md_seq_release,
+	.poll		= mdstat_poll,
 };
 
-int register_md_personality(int pnum, mdk_personality_t *p)
+int register_md_personality(struct mdk_personality *p)
 {
-	if (pnum >= MAX_PERSONALITY) {
-		printk(KERN_ERR
-		       "md: tried to install personality %s as nr %d, but max is %lu\n",
-		       p->name, pnum, MAX_PERSONALITY-1);
-		return -EINVAL;
-	}
-
 	spin_lock(&pers_lock);
-	if (pers[pnum]) {
-		spin_unlock(&pers_lock);
-		return -EBUSY;
-	}
-
-	pers[pnum] = p;
-	printk(KERN_INFO "md: %s personality registered as nr %d\n", p->name, pnum);
+	list_add_tail(&p->list, &pers_list);
+	printk(KERN_INFO "md: %s personality registered for level %d\n", p->name, p->level);
 	spin_unlock(&pers_lock);
 	return 0;
 }
 
-int unregister_md_personality(int pnum)
+int unregister_md_personality(struct mdk_personality *p)
 {
-	if (pnum >= MAX_PERSONALITY)
-		return -EINVAL;
-
-	printk(KERN_INFO "md: %s personality unregistered\n", pers[pnum]->name);
+	printk(KERN_INFO "md: %s personality unregistered\n", p->name);
 	spin_lock(&pers_lock);
-	pers[pnum] = NULL;
+	list_del_init(&p->list);
 	spin_unlock(&pers_lock);
 	return 0;
 }
@@ -4012,10 +4533,10 @@ static void md_do_sync(mddev_t *mddev)
 
 	printk(KERN_INFO "md: syncing RAID array %s\n", mdname(mddev));
 	printk(KERN_INFO "md: minimum _guaranteed_ reconstruction speed:"
-		" %d KB/sec/disc.\n", sysctl_speed_limit_min);
+		" %d KB/sec/disc.\n", speed_min(mddev));
 	printk(KERN_INFO "md: using maximum available idle IO bandwidth "
 	       "(but not more than %d KB/sec) for reconstruction.\n",
-	       sysctl_speed_limit_max);
+	       speed_max(mddev));
 
 	is_mddev_idle(mddev); /* this also initializes IO event counters */
 	/* we don't use the checkpoint if there's a bitmap */
@@ -4056,7 +4577,7 @@ static void md_do_sync(mddev_t *mddev)
 
 		skipped = 0;
 		sectors = mddev->pers->sync_request(mddev, j, &skipped,
-					    currspeed < sysctl_speed_limit_min);
+					    currspeed < speed_min(mddev));
 		if (sectors == 0) {
 			set_bit(MD_RECOVERY_ERR, &mddev->recovery);
 			goto out;
@@ -4069,7 +4590,11 @@ static void md_do_sync(mddev_t *mddev)
 
 		j += sectors;
 		if (j>1) mddev->curr_resync = j;
-
+		if (last_check == 0)
+			/* this is the earliers that rebuilt will be
+			 * visible in /proc/mdstat
+			 */
+			md_new_event(mddev);
 
 		if (last_check + window > io_sectors || j == max_sectors)
 			continue;
@@ -4117,8 +4642,8 @@ static void md_do_sync(mddev_t *mddev)
 		currspeed = ((unsigned long)(io_sectors-mddev->resync_mark_cnt))/2
 			/((jiffies-mddev->resync_mark)/HZ +1) +1;
 
-		if (currspeed > sysctl_speed_limit_min) {
-			if ((currspeed > sysctl_speed_limit_max) ||
+		if (currspeed > speed_min(mddev)) {
+			if ((currspeed > speed_max(mddev)) ||
 					!is_mddev_idle(mddev)) {
 				msleep(500);
 				goto repeat;
@@ -4255,6 +4780,7 @@ void md_check_recovery(mddev_t *mddev)
 			mddev->recovery = 0;
 			/* flag recovery needed just to double check */
 			set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+			md_new_event(mddev);
 			goto unlock;
 		}
 		/* Clear some bits that don't mean anything, but
@@ -4292,6 +4818,7 @@ void md_check_recovery(mddev_t *mddev)
 						sprintf(nm, "rd%d", rdev->raid_disk);
 						sysfs_create_link(&mddev->kobj, &rdev->kobj, nm);
 						spares++;
+						md_new_event(mddev);
 					} else
 						break;
 				}
@@ -4324,9 +4851,9 @@ void md_check_recovery(mddev_t *mddev)
 					mdname(mddev));
 				/* leave the spares where they are, it shouldn't hurt */
 				mddev->recovery = 0;
-			} else {
+			} else
 				md_wakeup_thread(mddev->sync_thread);
-			}
+			md_new_event(mddev);
 		}
 	unlock:
 		mddev_unlock(mddev);
@@ -4503,12 +5030,14 @@ static int set_ro(const char *val, struct kernel_param *kp)
 	int num = simple_strtoul(val, &e, 10);
 	if (*val && (*e == '\0' || *e == '\n')) {
 		start_readonly = num;
-		return 0;;
+		return 0;
 	}
 	return -EINVAL;
 }
 
 module_param_call(start_ro, set_ro, get_ro, NULL, 0600);
+module_param(start_dirty_degraded, int, 0644);
+
 
 EXPORT_SYMBOL(register_md_personality);
 EXPORT_SYMBOL(unregister_md_personality);
diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c
index 145cdc5ad00..e6aa309a66d 100644
--- a/drivers/md/multipath.c
+++ b/drivers/md/multipath.c
@@ -35,15 +35,10 @@
 #define	NR_RESERVED_BUFS	32
 
 
-static mdk_personality_t multipath_personality;
-
-
 static void *mp_pool_alloc(gfp_t gfp_flags, void *data)
 {
 	struct multipath_bh *mpb;
-	mpb = kmalloc(sizeof(*mpb), gfp_flags);
-	if (mpb) 
-		memset(mpb, 0, sizeof(*mpb));
+	mpb = kzalloc(sizeof(*mpb), gfp_flags);
 	return mpb;
 }
 
@@ -444,7 +439,7 @@ static int multipath_run (mddev_t *mddev)
 	 * should be freed in multipath_stop()]
 	 */
 
-	conf = kmalloc(sizeof(multipath_conf_t), GFP_KERNEL);
+	conf = kzalloc(sizeof(multipath_conf_t), GFP_KERNEL);
 	mddev->private = conf;
 	if (!conf) {
 		printk(KERN_ERR 
@@ -452,9 +447,8 @@ static int multipath_run (mddev_t *mddev)
 			mdname(mddev));
 		goto out;
 	}
-	memset(conf, 0, sizeof(*conf));
 
-	conf->multipaths = kmalloc(sizeof(struct multipath_info)*mddev->raid_disks,
+	conf->multipaths = kzalloc(sizeof(struct multipath_info)*mddev->raid_disks,
 				   GFP_KERNEL);
 	if (!conf->multipaths) {
 		printk(KERN_ERR 
@@ -462,7 +456,6 @@ static int multipath_run (mddev_t *mddev)
 			mdname(mddev));
 		goto out_free_conf;
 	}
-	memset(conf->multipaths, 0, sizeof(struct multipath_info)*mddev->raid_disks);
 
 	conf->working_disks = 0;
 	ITERATE_RDEV(mddev,rdev,tmp) {
@@ -557,9 +550,10 @@ static int multipath_stop (mddev_t *mddev)
 	return 0;
 }
 
-static mdk_personality_t multipath_personality=
+static struct mdk_personality multipath_personality =
 {
 	.name		= "multipath",
+	.level		= LEVEL_MULTIPATH,
 	.owner		= THIS_MODULE,
 	.make_request	= multipath_make_request,
 	.run		= multipath_run,
@@ -572,15 +566,17 @@ static mdk_personality_t multipath_personality=
 
 static int __init multipath_init (void)
 {
-	return register_md_personality (MULTIPATH, &multipath_personality);
+	return register_md_personality (&multipath_personality);
 }
 
 static void __exit multipath_exit (void)
 {
-	unregister_md_personality (MULTIPATH);
+	unregister_md_personality (&multipath_personality);
 }
 
 module_init(multipath_init);
 module_exit(multipath_exit);
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("md-personality-7"); /* MULTIPATH */
+MODULE_ALIAS("md-multipath");
+MODULE_ALIAS("md-level--4");
diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index fece3277c2a..abbca150202 100644
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -113,21 +113,16 @@ static int create_strip_zones (mddev_t *mddev)
 	}
 	printk("raid0: FINAL %d zones\n", conf->nr_strip_zones);
 
-	conf->strip_zone = kmalloc(sizeof(struct strip_zone)*
+	conf->strip_zone = kzalloc(sizeof(struct strip_zone)*
 				conf->nr_strip_zones, GFP_KERNEL);
 	if (!conf->strip_zone)
 		return 1;
-	conf->devlist = kmalloc(sizeof(mdk_rdev_t*)*
+	conf->devlist = kzalloc(sizeof(mdk_rdev_t*)*
 				conf->nr_strip_zones*mddev->raid_disks,
 				GFP_KERNEL);
 	if (!conf->devlist)
 		return 1;
 
-	memset(conf->strip_zone, 0,sizeof(struct strip_zone)*
-				   conf->nr_strip_zones);
-	memset(conf->devlist, 0,
-	       sizeof(mdk_rdev_t*) * conf->nr_strip_zones * mddev->raid_disks);
-
 	/* 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
 	 */
@@ -280,7 +275,11 @@ static int raid0_run (mddev_t *mddev)
 	mdk_rdev_t *rdev;
 	struct list_head *tmp;
 
-	printk("%s: setting max_sectors to %d, segment boundary to %d\n",
+	if (mddev->chunk_size == 0) {
+		printk(KERN_ERR "md/raid0: non-zero chunk size required.\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);
@@ -361,7 +360,7 @@ static int raid0_run (mddev_t *mddev)
 	 * chunksize should be used in that case.
 	 */
 	{
-		int stripe = mddev->raid_disks * mddev->chunk_size / PAGE_CACHE_SIZE;
+		int stripe = mddev->raid_disks * mddev->chunk_size / PAGE_SIZE;
 		if (mddev->queue->backing_dev_info.ra_pages < 2* stripe)
 			mddev->queue->backing_dev_info.ra_pages = 2* stripe;
 	}
@@ -512,9 +511,10 @@ static void raid0_status (struct seq_file *seq, mddev_t *mddev)
 	return;
 }
 
-static mdk_personality_t raid0_personality=
+static struct mdk_personality raid0_personality=
 {
 	.name		= "raid0",
+	.level		= 0,
 	.owner		= THIS_MODULE,
 	.make_request	= raid0_make_request,
 	.run		= raid0_run,
@@ -524,15 +524,17 @@ static mdk_personality_t raid0_personality=
 
 static int __init raid0_init (void)
 {
-	return register_md_personality (RAID0, &raid0_personality);
+	return register_md_personality (&raid0_personality);
 }
 
 static void raid0_exit (void)
 {
-	unregister_md_personality (RAID0);
+	unregister_md_personality (&raid0_personality);
 }
 
 module_init(raid0_init);
 module_exit(raid0_exit);
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("md-personality-2"); /* RAID0 */
+MODULE_ALIAS("md-raid0");
+MODULE_ALIAS("md-level-0");
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 229d7b20429..a06ff91f27e 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -47,10 +47,11 @@
  */
 #define	NR_RAID1_BIOS 256
 
-static mdk_personality_t raid1_personality;
 
 static void unplug_slaves(mddev_t *mddev);
 
+static void allow_barrier(conf_t *conf);
+static void lower_barrier(conf_t *conf);
 
 static void * r1bio_pool_alloc(gfp_t gfp_flags, void *data)
 {
@@ -59,10 +60,8 @@ static void * r1bio_pool_alloc(gfp_t gfp_flags, void *data)
 	int size = offsetof(r1bio_t, bios[pi->raid_disks]);
 
 	/* allocate a r1bio with room for raid_disks entries in the bios array */
-	r1_bio = kmalloc(size, gfp_flags);
-	if (r1_bio)
-		memset(r1_bio, 0, size);
-	else
+	r1_bio = kzalloc(size, gfp_flags);
+	if (!r1_bio)
 		unplug_slaves(pi->mddev);
 
 	return r1_bio;
@@ -104,15 +103,30 @@ static void * r1buf_pool_alloc(gfp_t gfp_flags, void *data)
 	}
 	/*
 	 * Allocate RESYNC_PAGES data pages and attach them to
-	 * the first bio;
+	 * the first bio.
+	 * If this is a user-requested check/repair, allocate
+	 * RESYNC_PAGES for each bio.
 	 */
-	bio = r1_bio->bios[0];
-	for (i = 0; i < RESYNC_PAGES; i++) {
-		page = alloc_page(gfp_flags);
-		if (unlikely(!page))
-			goto out_free_pages;
-
-		bio->bi_io_vec[i].bv_page = page;
+	if (test_bit(MD_RECOVERY_REQUESTED, &pi->mddev->recovery))
+		j = pi->raid_disks;
+	else
+		j = 1;
+	while(j--) {
+		bio = r1_bio->bios[j];
+		for (i = 0; i < RESYNC_PAGES; i++) {
+			page = alloc_page(gfp_flags);
+			if (unlikely(!page))
+				goto out_free_pages;
+
+			bio->bi_io_vec[i].bv_page = page;
+		}
+	}
+	/* If not user-requests, copy the page pointers to all bios */
+	if (!test_bit(MD_RECOVERY_REQUESTED, &pi->mddev->recovery)) {
+		for (i=0; i<RESYNC_PAGES ; i++)
+			for (j=1; j<pi->raid_disks; j++)
+				r1_bio->bios[j]->bi_io_vec[i].bv_page =
+					r1_bio->bios[0]->bi_io_vec[i].bv_page;
 	}
 
 	r1_bio->master_bio = NULL;
@@ -120,8 +134,10 @@ static void * r1buf_pool_alloc(gfp_t gfp_flags, void *data)
 	return r1_bio;
 
 out_free_pages:
-	for ( ; i > 0 ; i--)
-		__free_page(bio->bi_io_vec[i-1].bv_page);
+	for (i=0; i < RESYNC_PAGES ; i++)
+		for (j=0 ; j < pi->raid_disks; j++)
+			safe_put_page(r1_bio->bios[j]->bi_io_vec[i].bv_page);
+	j = -1;
 out_free_bio:
 	while ( ++j < pi->raid_disks )
 		bio_put(r1_bio->bios[j]);
@@ -132,14 +148,16 @@ out_free_bio:
 static void r1buf_pool_free(void *__r1_bio, void *data)
 {
 	struct pool_info *pi = data;
-	int i;
+	int i,j;
 	r1bio_t *r1bio = __r1_bio;
-	struct bio *bio = r1bio->bios[0];
 
-	for (i = 0; i < RESYNC_PAGES; i++) {
-		__free_page(bio->bi_io_vec[i].bv_page);
-		bio->bi_io_vec[i].bv_page = NULL;
-	}
+	for (i = 0; i < RESYNC_PAGES; i++)
+		for (j = pi->raid_disks; j-- ;) {
+			if (j == 0 ||
+			    r1bio->bios[j]->bi_io_vec[i].bv_page !=
+			    r1bio->bios[0]->bi_io_vec[i].bv_page)
+				safe_put_page(r1bio->bios[j]->bi_io_vec[i].bv_page);
+		}
 	for (i=0 ; i < pi->raid_disks; i++)
 		bio_put(r1bio->bios[i]);
 
@@ -152,7 +170,7 @@ static void put_all_bios(conf_t *conf, r1bio_t *r1_bio)
 
 	for (i = 0; i < conf->raid_disks; i++) {
 		struct bio **bio = r1_bio->bios + i;
-		if (*bio)
+		if (*bio && *bio != IO_BLOCKED)
 			bio_put(*bio);
 		*bio = NULL;
 	}
@@ -160,20 +178,13 @@ static void put_all_bios(conf_t *conf, r1bio_t *r1_bio)
 
 static inline void free_r1bio(r1bio_t *r1_bio)
 {
-	unsigned long flags;
-
 	conf_t *conf = mddev_to_conf(r1_bio->mddev);
 
 	/*
 	 * Wake up any possible resync thread that waits for the device
 	 * to go idle.
 	 */
-	spin_lock_irqsave(&conf->resync_lock, flags);
-	if (!--conf->nr_pending) {
-		wake_up(&conf->wait_idle);
-		wake_up(&conf->wait_resume);
-	}
-	spin_unlock_irqrestore(&conf->resync_lock, flags);
+	allow_barrier(conf);
 
 	put_all_bios(conf, r1_bio);
 	mempool_free(r1_bio, conf->r1bio_pool);
@@ -182,22 +193,17 @@ static inline void free_r1bio(r1bio_t *r1_bio)
 static inline void put_buf(r1bio_t *r1_bio)
 {
 	conf_t *conf = mddev_to_conf(r1_bio->mddev);
-	unsigned long flags;
+	int i;
 
-	mempool_free(r1_bio, conf->r1buf_pool);
+	for (i=0; i<conf->raid_disks; i++) {
+		struct bio *bio = r1_bio->bios[i];
+		if (bio->bi_end_io)
+			rdev_dec_pending(conf->mirrors[i].rdev, r1_bio->mddev);
+	}
 
-	spin_lock_irqsave(&conf->resync_lock, flags);
-	if (!conf->barrier)
-		BUG();
-	--conf->barrier;
-	wake_up(&conf->wait_resume);
-	wake_up(&conf->wait_idle);
+	mempool_free(r1_bio, conf->r1buf_pool);
 
-	if (!--conf->nr_pending) {
-		wake_up(&conf->wait_idle);
-		wake_up(&conf->wait_resume);
-	}
-	spin_unlock_irqrestore(&conf->resync_lock, flags);
+	lower_barrier(conf);
 }
 
 static void reschedule_retry(r1bio_t *r1_bio)
@@ -208,8 +214,10 @@ static void reschedule_retry(r1bio_t *r1_bio)
 
 	spin_lock_irqsave(&conf->device_lock, flags);
 	list_add(&r1_bio->retry_list, &conf->retry_list);
+	conf->nr_queued ++;
 	spin_unlock_irqrestore(&conf->device_lock, flags);
 
+	wake_up(&conf->wait_barrier);
 	md_wakeup_thread(mddev->thread);
 }
 
@@ -261,9 +269,9 @@ static int raid1_end_read_request(struct bio *bio, unsigned int bytes_done, int
 	/*
 	 * this branch is our 'one mirror IO has finished' event handler:
 	 */
-	if (!uptodate)
-		md_error(r1_bio->mddev, conf->mirrors[mirror].rdev);
-	else
+	update_head_pos(mirror, r1_bio);
+
+	if (uptodate || conf->working_disks <= 1) {
 		/*
 		 * Set R1BIO_Uptodate in our master bio, so that
 		 * we will return a good error code for to the higher
@@ -273,16 +281,11 @@ static int raid1_end_read_request(struct bio *bio, unsigned int bytes_done, int
 		 * user-side. So if something waits for IO, then it will
 		 * wait for the 'master' bio.
 		 */
-		set_bit(R1BIO_Uptodate, &r1_bio->state);
-
-	update_head_pos(mirror, r1_bio);
+		if (uptodate)
+			set_bit(R1BIO_Uptodate, &r1_bio->state);
 
-	/*
-	 * we have only one bio on the read side
-	 */
-	if (uptodate)
 		raid_end_bio_io(r1_bio);
-	else {
+	} else {
 		/*
 		 * oops, read error:
 		 */
@@ -378,7 +381,7 @@ static int raid1_end_write_request(struct bio *bio, unsigned int bytes_done, int
 			/* free extra copy of the data pages */
 			int i = bio->bi_vcnt;
 			while (i--)
-				__free_page(bio->bi_io_vec[i].bv_page);
+				safe_put_page(bio->bi_io_vec[i].bv_page);
 		}
 		/* clear the bitmap if all writes complete successfully */
 		bitmap_endwrite(r1_bio->mddev->bitmap, r1_bio->sector,
@@ -433,11 +436,13 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 		new_disk = 0;
 
 		for (rdev = rcu_dereference(conf->mirrors[new_disk].rdev);
+		     r1_bio->bios[new_disk] == IO_BLOCKED ||
 		     !rdev || !test_bit(In_sync, &rdev->flags)
 			     || test_bit(WriteMostly, &rdev->flags);
 		     rdev = rcu_dereference(conf->mirrors[++new_disk].rdev)) {
 
-			if (rdev && test_bit(In_sync, &rdev->flags))
+			if (rdev && test_bit(In_sync, &rdev->flags) &&
+				r1_bio->bios[new_disk] != IO_BLOCKED)
 				wonly_disk = new_disk;
 
 			if (new_disk == conf->raid_disks - 1) {
@@ -451,11 +456,13 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 
 	/* make sure the disk is operational */
 	for (rdev = rcu_dereference(conf->mirrors[new_disk].rdev);
+	     r1_bio->bios[new_disk] == IO_BLOCKED ||
 	     !rdev || !test_bit(In_sync, &rdev->flags) ||
 		     test_bit(WriteMostly, &rdev->flags);
 	     rdev = rcu_dereference(conf->mirrors[new_disk].rdev)) {
 
-		if (rdev && test_bit(In_sync, &rdev->flags))
+		if (rdev && test_bit(In_sync, &rdev->flags) &&
+		    r1_bio->bios[new_disk] != IO_BLOCKED)
 			wonly_disk = new_disk;
 
 		if (new_disk <= 0)
@@ -492,7 +499,7 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 
 		rdev = rcu_dereference(conf->mirrors[disk].rdev);
 
-		if (!rdev ||
+		if (!rdev || r1_bio->bios[disk] == IO_BLOCKED ||
 		    !test_bit(In_sync, &rdev->flags) ||
 		    test_bit(WriteMostly, &rdev->flags))
 			continue;
@@ -520,7 +527,7 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio)
 			/* cannot risk returning a device that failed
 			 * before we inc'ed nr_pending
 			 */
-			atomic_dec(&rdev->nr_pending);
+			rdev_dec_pending(rdev, conf->mddev);
 			goto retry;
 		}
 		conf->next_seq_sect = this_sector + sectors;
@@ -593,42 +600,119 @@ static int raid1_issue_flush(request_queue_t *q, struct gendisk *disk,
 	return ret;
 }
 
-/*
- * Throttle resync depth, so that we can both get proper overlapping of
- * requests, but are still able to handle normal requests quickly.
+/* Barriers....
+ * Sometimes we need to suspend IO while we do something else,
+ * either some resync/recovery, or reconfigure the array.
+ * To do this we raise a 'barrier'.
+ * The 'barrier' is a counter that can be raised multiple times
+ * to count how many activities are happening which preclude
+ * normal IO.
+ * We can only raise the barrier if there is no pending IO.
+ * i.e. if nr_pending == 0.
+ * We choose only to raise the barrier if no-one is waiting for the
+ * barrier to go down.  This means that as soon as an IO request
+ * is ready, no other operations which require a barrier will start
+ * until the IO request has had a chance.
+ *
+ * So: regular IO calls 'wait_barrier'.  When that returns there
+ *    is no backgroup IO happening,  It must arrange to call
+ *    allow_barrier when it has finished its IO.
+ * backgroup IO calls must call raise_barrier.  Once that returns
+ *    there is no normal IO happeing.  It must arrange to call
+ *    lower_barrier when the particular background IO completes.
  */
 #define RESYNC_DEPTH 32
 
-static void device_barrier(conf_t *conf, sector_t sect)
+static void raise_barrier(conf_t *conf)
 {
 	spin_lock_irq(&conf->resync_lock);
-	wait_event_lock_irq(conf->wait_idle, !waitqueue_active(&conf->wait_resume),
-			    conf->resync_lock, raid1_unplug(conf->mddev->queue));
-	
-	if (!conf->barrier++) {
-		wait_event_lock_irq(conf->wait_idle, !conf->nr_pending,
-				    conf->resync_lock, raid1_unplug(conf->mddev->queue));
-		if (conf->nr_pending)
-			BUG();
+
+	/* Wait until no block IO is waiting */
+	wait_event_lock_irq(conf->wait_barrier, !conf->nr_waiting,
+			    conf->resync_lock,
+			    raid1_unplug(conf->mddev->queue));
+
+	/* block any new IO from starting */
+	conf->barrier++;
+
+	/* No wait for all pending IO to complete */
+	wait_event_lock_irq(conf->wait_barrier,
+			    !conf->nr_pending && conf->barrier < RESYNC_DEPTH,
+			    conf->resync_lock,
+			    raid1_unplug(conf->mddev->queue));
+
+	spin_unlock_irq(&conf->resync_lock);
+}
+
+static void lower_barrier(conf_t *conf)
+{
+	unsigned long flags;
+	spin_lock_irqsave(&conf->resync_lock, flags);
+	conf->barrier--;
+	spin_unlock_irqrestore(&conf->resync_lock, flags);
+	wake_up(&conf->wait_barrier);
+}
+
+static void wait_barrier(conf_t *conf)
+{
+	spin_lock_irq(&conf->resync_lock);
+	if (conf->barrier) {
+		conf->nr_waiting++;
+		wait_event_lock_irq(conf->wait_barrier, !conf->barrier,
+				    conf->resync_lock,
+				    raid1_unplug(conf->mddev->queue));
+		conf->nr_waiting--;
 	}
-	wait_event_lock_irq(conf->wait_resume, conf->barrier < RESYNC_DEPTH,
-			    conf->resync_lock, raid1_unplug(conf->mddev->queue));
-	conf->next_resync = sect;
+	conf->nr_pending++;
+	spin_unlock_irq(&conf->resync_lock);
+}
+
+static void allow_barrier(conf_t *conf)
+{
+	unsigned long flags;
+	spin_lock_irqsave(&conf->resync_lock, flags);
+	conf->nr_pending--;
+	spin_unlock_irqrestore(&conf->resync_lock, flags);
+	wake_up(&conf->wait_barrier);
+}
+
+static void freeze_array(conf_t *conf)
+{
+	/* stop syncio and normal IO and wait for everything to
+	 * go quite.
+	 * We increment barrier and nr_waiting, and then
+	 * wait until barrier+nr_pending match nr_queued+2
+	 */
+	spin_lock_irq(&conf->resync_lock);
+	conf->barrier++;
+	conf->nr_waiting++;
+	wait_event_lock_irq(conf->wait_barrier,
+			    conf->barrier+conf->nr_pending == conf->nr_queued+2,
+			    conf->resync_lock,
+			    raid1_unplug(conf->mddev->queue));
+	spin_unlock_irq(&conf->resync_lock);
+}
+static void unfreeze_array(conf_t *conf)
+{
+	/* reverse the effect of the freeze */
+	spin_lock_irq(&conf->resync_lock);
+	conf->barrier--;
+	conf->nr_waiting--;
+	wake_up(&conf->wait_barrier);
 	spin_unlock_irq(&conf->resync_lock);
 }
 
+
 /* duplicate the data pages for behind I/O */
 static struct page **alloc_behind_pages(struct bio *bio)
 {
 	int i;
 	struct bio_vec *bvec;
-	struct page **pages = kmalloc(bio->bi_vcnt * sizeof(struct page *),
+	struct page **pages = kzalloc(bio->bi_vcnt * sizeof(struct page *),
 					GFP_NOIO);
 	if (unlikely(!pages))
 		goto do_sync_io;
 
-	memset(pages, 0, bio->bi_vcnt * sizeof(struct page *));
-
 	bio_for_each_segment(bvec, bio, i) {
 		pages[i] = alloc_page(GFP_NOIO);
 		if (unlikely(!pages[i]))
@@ -644,7 +728,7 @@ static struct page **alloc_behind_pages(struct bio *bio)
 do_sync_io:
 	if (pages)
 		for (i = 0; i < bio->bi_vcnt && pages[i]; i++)
-			__free_page(pages[i]);
+			put_page(pages[i]);
 	kfree(pages);
 	PRINTK("%dB behind alloc failed, doing sync I/O\n", bio->bi_size);
 	return NULL;
@@ -678,10 +762,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 	 */
 	md_write_start(mddev, bio); /* wait on superblock update early */
 
-	spin_lock_irq(&conf->resync_lock);
-	wait_event_lock_irq(conf->wait_resume, !conf->barrier, conf->resync_lock, );
-	conf->nr_pending++;
-	spin_unlock_irq(&conf->resync_lock);
+	wait_barrier(conf);
 
 	disk_stat_inc(mddev->gendisk, ios[rw]);
 	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
@@ -749,7 +830,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 		    !test_bit(Faulty, &rdev->flags)) {
 			atomic_inc(&rdev->nr_pending);
 			if (test_bit(Faulty, &rdev->flags)) {
-				atomic_dec(&rdev->nr_pending);
+				rdev_dec_pending(rdev, mddev);
 				r1_bio->bios[i] = NULL;
 			} else
 				r1_bio->bios[i] = bio;
@@ -909,13 +990,8 @@ static void print_conf(conf_t *conf)
 
 static void close_sync(conf_t *conf)
 {
-	spin_lock_irq(&conf->resync_lock);
-	wait_event_lock_irq(conf->wait_resume, !conf->barrier,
-			    conf->resync_lock, 	raid1_unplug(conf->mddev->queue));
-	spin_unlock_irq(&conf->resync_lock);
-
-	if (conf->barrier) BUG();
-	if (waitqueue_active(&conf->wait_idle)) BUG();
+	wait_barrier(conf);
+	allow_barrier(conf);
 
 	mempool_destroy(conf->r1buf_pool);
 	conf->r1buf_pool = NULL;
@@ -1015,28 +1091,27 @@ abort:
 
 static int end_sync_read(struct bio *bio, unsigned int bytes_done, int error)
 {
-	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	r1bio_t * r1_bio = (r1bio_t *)(bio->bi_private);
-	conf_t *conf = mddev_to_conf(r1_bio->mddev);
+	int i;
 
 	if (bio->bi_size)
 		return 1;
 
-	if (r1_bio->bios[r1_bio->read_disk] != bio)
-		BUG();
-	update_head_pos(r1_bio->read_disk, r1_bio);
+	for (i=r1_bio->mddev->raid_disks; i--; )
+		if (r1_bio->bios[i] == bio)
+			break;
+	BUG_ON(i < 0);
+	update_head_pos(i, r1_bio);
 	/*
 	 * we have read a block, now it needs to be re-written,
 	 * or re-read if the read failed.
 	 * We don't do much here, just schedule handling by raid1d
 	 */
-	if (!uptodate) {
-		md_error(r1_bio->mddev,
-			 conf->mirrors[r1_bio->read_disk].rdev);
-	} else
+	if (test_bit(BIO_UPTODATE, &bio->bi_flags))
 		set_bit(R1BIO_Uptodate, &r1_bio->state);
-	rdev_dec_pending(conf->mirrors[r1_bio->read_disk].rdev, conf->mddev);
-	reschedule_retry(r1_bio);
+
+	if (atomic_dec_and_test(&r1_bio->remaining))
+		reschedule_retry(r1_bio);
 	return 0;
 }
 
@@ -1066,7 +1141,6 @@ static int end_sync_write(struct bio *bio, unsigned int bytes_done, int error)
 		md_done_sync(mddev, r1_bio->sectors, uptodate);
 		put_buf(r1_bio);
 	}
-	rdev_dec_pending(conf->mirrors[mirror].rdev, mddev);
 	return 0;
 }
 
@@ -1079,34 +1153,173 @@ static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio)
 
 	bio = r1_bio->bios[r1_bio->read_disk];
 
-/*
-	if (r1_bio->sector == 0) printk("First sync write startss\n");
-*/
-	/*
-	 * schedule writes
-	 */
+
+	if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
+		/* We have read all readable devices.  If we haven't
+		 * got the block, then there is no hope left.
+		 * If we have, then we want to do a comparison
+		 * and skip the write if everything is the same.
+		 * If any blocks failed to read, then we need to
+		 * attempt an over-write
+		 */
+		int primary;
+		if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) {
+			for (i=0; i<mddev->raid_disks; i++)
+				if (r1_bio->bios[i]->bi_end_io == end_sync_read)
+					md_error(mddev, conf->mirrors[i].rdev);
+
+			md_done_sync(mddev, r1_bio->sectors, 1);
+			put_buf(r1_bio);
+			return;
+		}
+		for (primary=0; primary<mddev->raid_disks; primary++)
+			if (r1_bio->bios[primary]->bi_end_io == end_sync_read &&
+			    test_bit(BIO_UPTODATE, &r1_bio->bios[primary]->bi_flags)) {
+				r1_bio->bios[primary]->bi_end_io = NULL;
+				rdev_dec_pending(conf->mirrors[primary].rdev, mddev);
+				break;
+			}
+		r1_bio->read_disk = primary;
+		for (i=0; i<mddev->raid_disks; i++)
+			if (r1_bio->bios[i]->bi_end_io == end_sync_read &&
+			    test_bit(BIO_UPTODATE, &r1_bio->bios[i]->bi_flags)) {
+				int j;
+				int vcnt = r1_bio->sectors >> (PAGE_SHIFT- 9);
+				struct bio *pbio = r1_bio->bios[primary];
+				struct bio *sbio = r1_bio->bios[i];
+				for (j = vcnt; j-- ; )
+					if (memcmp(page_address(pbio->bi_io_vec[j].bv_page),
+						   page_address(sbio->bi_io_vec[j].bv_page),
+						   PAGE_SIZE))
+						break;
+				if (j >= 0)
+					mddev->resync_mismatches += r1_bio->sectors;
+				if (j < 0 || test_bit(MD_RECOVERY_CHECK, &mddev->recovery)) {
+					sbio->bi_end_io = NULL;
+					rdev_dec_pending(conf->mirrors[i].rdev, mddev);
+				} else {
+					/* fixup the bio for reuse */
+					sbio->bi_vcnt = vcnt;
+					sbio->bi_size = r1_bio->sectors << 9;
+					sbio->bi_idx = 0;
+					sbio->bi_phys_segments = 0;
+					sbio->bi_hw_segments = 0;
+					sbio->bi_hw_front_size = 0;
+					sbio->bi_hw_back_size = 0;
+					sbio->bi_flags &= ~(BIO_POOL_MASK - 1);
+					sbio->bi_flags |= 1 << BIO_UPTODATE;
+					sbio->bi_next = NULL;
+					sbio->bi_sector = r1_bio->sector +
+						conf->mirrors[i].rdev->data_offset;
+					sbio->bi_bdev = conf->mirrors[i].rdev->bdev;
+				}
+			}
+	}
 	if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) {
-		/*
-		 * There is no point trying a read-for-reconstruct as
-		 * reconstruct is about to be aborted
+		/* ouch - failed to read all of that.
+		 * Try some synchronous reads of other devices to get
+		 * good data, much like with normal read errors.  Only
+		 * read into the pages we already have so they we don't
+		 * need to re-issue the read request.
+		 * We don't need to freeze the array, because being in an
+		 * active sync request, there is no normal IO, and
+		 * no overlapping syncs.
 		 */
-		char b[BDEVNAME_SIZE];
-		printk(KERN_ALERT "raid1: %s: unrecoverable I/O read error"
-			" for block %llu\n",
-			bdevname(bio->bi_bdev,b), 
-			(unsigned long long)r1_bio->sector);
-		md_done_sync(mddev, r1_bio->sectors, 0);
-		put_buf(r1_bio);
-		return;
+		sector_t sect = r1_bio->sector;
+		int sectors = r1_bio->sectors;
+		int idx = 0;
+
+		while(sectors) {
+			int s = sectors;
+			int d = r1_bio->read_disk;
+			int success = 0;
+			mdk_rdev_t *rdev;
+
+			if (s > (PAGE_SIZE>>9))
+				s = PAGE_SIZE >> 9;
+			do {
+				if (r1_bio->bios[d]->bi_end_io == end_sync_read) {
+					rdev = conf->mirrors[d].rdev;
+					if (sync_page_io(rdev->bdev,
+							 sect + rdev->data_offset,
+							 s<<9,
+							 bio->bi_io_vec[idx].bv_page,
+							 READ)) {
+						success = 1;
+						break;
+					}
+				}
+				d++;
+				if (d == conf->raid_disks)
+					d = 0;
+			} while (!success && d != r1_bio->read_disk);
+
+			if (success) {
+				int start = d;
+				/* write it back and re-read */
+				set_bit(R1BIO_Uptodate, &r1_bio->state);
+				while (d != r1_bio->read_disk) {
+					if (d == 0)
+						d = conf->raid_disks;
+					d--;
+					if (r1_bio->bios[d]->bi_end_io != end_sync_read)
+						continue;
+					rdev = conf->mirrors[d].rdev;
+					atomic_add(s, &rdev->corrected_errors);
+					if (sync_page_io(rdev->bdev,
+							 sect + rdev->data_offset,
+							 s<<9,
+							 bio->bi_io_vec[idx].bv_page,
+							 WRITE) == 0)
+						md_error(mddev, rdev);
+				}
+				d = start;
+				while (d != r1_bio->read_disk) {
+					if (d == 0)
+						d = conf->raid_disks;
+					d--;
+					if (r1_bio->bios[d]->bi_end_io != end_sync_read)
+						continue;
+					rdev = conf->mirrors[d].rdev;
+					if (sync_page_io(rdev->bdev,
+							 sect + rdev->data_offset,
+							 s<<9,
+							 bio->bi_io_vec[idx].bv_page,
+							 READ) == 0)
+						md_error(mddev, rdev);
+				}
+			} else {
+				char b[BDEVNAME_SIZE];
+				/* Cannot read from anywhere, array is toast */
+				md_error(mddev, conf->mirrors[r1_bio->read_disk].rdev);
+				printk(KERN_ALERT "raid1: %s: unrecoverable I/O read error"
+				       " for block %llu\n",
+				       bdevname(bio->bi_bdev,b),
+				       (unsigned long long)r1_bio->sector);
+				md_done_sync(mddev, r1_bio->sectors, 0);
+				put_buf(r1_bio);
+				return;
+			}
+			sectors -= s;
+			sect += s;
+			idx ++;
+		}
 	}
 
+	/*
+	 * schedule writes
+	 */
 	atomic_set(&r1_bio->remaining, 1);
 	for (i = 0; i < disks ; i++) {
 		wbio = r1_bio->bios[i];
-		if (wbio->bi_end_io != end_sync_write)
+		if (wbio->bi_end_io == NULL ||
+		    (wbio->bi_end_io == end_sync_read &&
+		     (i == r1_bio->read_disk ||
+		      !test_bit(MD_RECOVERY_SYNC, &mddev->recovery))))
 			continue;
 
-		atomic_inc(&conf->mirrors[i].rdev->nr_pending);
+		wbio->bi_rw = WRITE;
+		wbio->bi_end_io = end_sync_write;
 		atomic_inc(&r1_bio->remaining);
 		md_sync_acct(conf->mirrors[i].rdev->bdev, wbio->bi_size >> 9);
 
@@ -1167,6 +1380,7 @@ static void raid1d(mddev_t *mddev)
 			break;
 		r1_bio = list_entry(head->prev, r1bio_t, retry_list);
 		list_del(head->prev);
+		conf->nr_queued--;
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 
 		mddev = r1_bio->mddev;
@@ -1206,6 +1420,86 @@ static void raid1d(mddev_t *mddev)
 				}
 		} else {
 			int disk;
+
+			/* we got a read error. Maybe the drive is bad.  Maybe just
+			 * the block and we can fix it.
+			 * We freeze all other IO, and try reading the block from
+			 * other devices.  When we find one, we re-write
+			 * and check it that fixes the read error.
+			 * This is all done synchronously while the array is
+			 * frozen
+			 */
+			sector_t sect = r1_bio->sector;
+			int sectors = r1_bio->sectors;
+			freeze_array(conf);
+			if (mddev->ro == 0) while(sectors) {
+				int s = sectors;
+				int d = r1_bio->read_disk;
+				int success = 0;
+
+				if (s > (PAGE_SIZE>>9))
+					s = PAGE_SIZE >> 9;
+
+				do {
+					rdev = conf->mirrors[d].rdev;
+					if (rdev &&
+					    test_bit(In_sync, &rdev->flags) &&
+					    sync_page_io(rdev->bdev,
+							 sect + rdev->data_offset,
+							 s<<9,
+							 conf->tmppage, READ))
+						success = 1;
+					else {
+						d++;
+						if (d == conf->raid_disks)
+							d = 0;
+					}
+				} while (!success && d != r1_bio->read_disk);
+
+				if (success) {
+					/* write it back and re-read */
+					int start = d;
+					while (d != r1_bio->read_disk) {
+						if (d==0)
+							d = conf->raid_disks;
+						d--;
+						rdev = conf->mirrors[d].rdev;
+						atomic_add(s, &rdev->corrected_errors);
+						if (rdev &&
+						    test_bit(In_sync, &rdev->flags)) {
+							if (sync_page_io(rdev->bdev,
+									 sect + rdev->data_offset,
+									 s<<9, conf->tmppage, WRITE) == 0)
+								/* Well, this device is dead */
+								md_error(mddev, rdev);
+						}
+					}
+					d = start;
+					while (d != r1_bio->read_disk) {
+						if (d==0)
+							d = conf->raid_disks;
+						d--;
+						rdev = conf->mirrors[d].rdev;
+						if (rdev &&
+						    test_bit(In_sync, &rdev->flags)) {
+							if (sync_page_io(rdev->bdev,
+									 sect + rdev->data_offset,
+									 s<<9, conf->tmppage, READ) == 0)
+								/* Well, this device is dead */
+								md_error(mddev, rdev);
+						}
+					}
+				} else {
+					/* Cannot read from anywhere -- bye bye array */
+					md_error(mddev, conf->mirrors[r1_bio->read_disk].rdev);
+					break;
+				}
+				sectors -= s;
+				sect += s;
+			}
+
+			unfreeze_array(conf);
+
 			bio = r1_bio->bios[r1_bio->read_disk];
 			if ((disk=read_balance(conf, r1_bio)) == -1) {
 				printk(KERN_ALERT "raid1: %s: unrecoverable I/O"
@@ -1214,7 +1508,8 @@ static void raid1d(mddev_t *mddev)
 				       (unsigned long long)r1_bio->sector);
 				raid_end_bio_io(r1_bio);
 			} else {
-				r1_bio->bios[r1_bio->read_disk] = NULL;
+				r1_bio->bios[r1_bio->read_disk] =
+					mddev->ro ? IO_BLOCKED : NULL;
 				r1_bio->read_disk = disk;
 				bio_put(bio);
 				bio = bio_clone(r1_bio->master_bio, GFP_NOIO);
@@ -1269,14 +1564,13 @@ static int init_resync(conf_t *conf)
 static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster)
 {
 	conf_t *conf = mddev_to_conf(mddev);
-	mirror_info_t *mirror;
 	r1bio_t *r1_bio;
 	struct bio *bio;
 	sector_t max_sector, nr_sectors;
-	int disk;
+	int disk = -1;
 	int i;
-	int wonly;
-	int write_targets = 0;
+	int wonly = -1;
+	int write_targets = 0, read_targets = 0;
 	int sync_blocks;
 	int still_degraded = 0;
 
@@ -1317,55 +1611,35 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 		return sync_blocks;
 	}
 	/*
-	 * If there is non-resync activity waiting for us then
-	 * put in a delay to throttle resync.
+	 * If there is non-resync activity waiting for a turn,
+	 * and resync is going fast enough,
+	 * then let it though before starting on this new sync request.
 	 */
-	if (!go_faster && waitqueue_active(&conf->wait_resume))
+	if (!go_faster && conf->nr_waiting)
 		msleep_interruptible(1000);
-	device_barrier(conf, sector_nr + RESYNC_SECTORS);
-
-	/*
-	 * If reconstructing, and >1 working disc,
-	 * could dedicate one to rebuild and others to
-	 * service read requests ..
-	 */
-	disk = conf->last_used;
-	/* make sure disk is operational */
-	wonly = disk;
-	while (conf->mirrors[disk].rdev == NULL ||
-	       !test_bit(In_sync, &conf->mirrors[disk].rdev->flags) ||
-	       test_bit(WriteMostly, &conf->mirrors[disk].rdev->flags)
-		) {
-		if (conf->mirrors[disk].rdev  &&
-		    test_bit(In_sync, &conf->mirrors[disk].rdev->flags))
-			wonly = disk;
-		if (disk <= 0)
-			disk = conf->raid_disks;
-		disk--;
-		if (disk == conf->last_used) {
-			disk = wonly;
-			break;
-		}
-	}
-	conf->last_used = disk;
-	atomic_inc(&conf->mirrors[disk].rdev->nr_pending);
 
+	raise_barrier(conf);
 
-	mirror = conf->mirrors + disk;
+	conf->next_resync = sector_nr;
 
 	r1_bio = mempool_alloc(conf->r1buf_pool, GFP_NOIO);
-
-	spin_lock_irq(&conf->resync_lock);
-	conf->nr_pending++;
-	spin_unlock_irq(&conf->resync_lock);
+	rcu_read_lock();
+	/*
+	 * If we get a correctably read error during resync or recovery,
+	 * we might want to read from a different device.  So we
+	 * flag all drives that could conceivably be read from for READ,
+	 * and any others (which will be non-In_sync devices) for WRITE.
+	 * If a read fails, we try reading from something else for which READ
+	 * is OK.
+	 */
 
 	r1_bio->mddev = mddev;
 	r1_bio->sector = sector_nr;
 	r1_bio->state = 0;
 	set_bit(R1BIO_IsSync, &r1_bio->state);
-	r1_bio->read_disk = disk;
 
 	for (i=0; i < conf->raid_disks; i++) {
+		mdk_rdev_t *rdev;
 		bio = r1_bio->bios[i];
 
 		/* take from bio_init */
@@ -1380,35 +1654,49 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 		bio->bi_end_io = NULL;
 		bio->bi_private = NULL;
 
-		if (i == disk) {
-			bio->bi_rw = READ;
-			bio->bi_end_io = end_sync_read;
-		} else if (conf->mirrors[i].rdev == NULL ||
-			   test_bit(Faulty, &conf->mirrors[i].rdev->flags)) {
+		rdev = rcu_dereference(conf->mirrors[i].rdev);
+		if (rdev == NULL ||
+			   test_bit(Faulty, &rdev->flags)) {
 			still_degraded = 1;
 			continue;
-		} else if (!test_bit(In_sync, &conf->mirrors[i].rdev->flags) ||
-			   sector_nr + RESYNC_SECTORS > mddev->recovery_cp   ||
-			   test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
+		} else if (!test_bit(In_sync, &rdev->flags)) {
 			bio->bi_rw = WRITE;
 			bio->bi_end_io = end_sync_write;
 			write_targets ++;
-		} else
-			/* no need to read or write here */
-			continue;
-		bio->bi_sector = sector_nr + conf->mirrors[i].rdev->data_offset;
-		bio->bi_bdev = conf->mirrors[i].rdev->bdev;
+		} else {
+			/* may need to read from here */
+			bio->bi_rw = READ;
+			bio->bi_end_io = end_sync_read;
+			if (test_bit(WriteMostly, &rdev->flags)) {
+				if (wonly < 0)
+					wonly = i;
+			} else {
+				if (disk < 0)
+					disk = i;
+			}
+			read_targets++;
+		}
+		atomic_inc(&rdev->nr_pending);
+		bio->bi_sector = sector_nr + rdev->data_offset;
+		bio->bi_bdev = rdev->bdev;
 		bio->bi_private = r1_bio;
 	}
+	rcu_read_unlock();
+	if (disk < 0)
+		disk = wonly;
+	r1_bio->read_disk = disk;
+
+	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) && read_targets > 0)
+		/* extra read targets are also write targets */
+		write_targets += read_targets-1;
 
-	if (write_targets == 0) {
+	if (write_targets == 0 || read_targets == 0) {
 		/* There is nowhere to write, so all non-sync
 		 * drives must be failed - so we are finished
 		 */
 		sector_t rv = max_sector - sector_nr;
 		*skipped = 1;
 		put_buf(r1_bio);
-		rdev_dec_pending(conf->mirrors[disk].rdev, mddev);
 		return rv;
 	}
 
@@ -1436,10 +1724,10 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 		for (i=0 ; i < conf->raid_disks; i++) {
 			bio = r1_bio->bios[i];
 			if (bio->bi_end_io) {
-				page = r1_bio->bios[0]->bi_io_vec[bio->bi_vcnt].bv_page;
+				page = bio->bi_io_vec[bio->bi_vcnt].bv_page;
 				if (bio_add_page(bio, page, len, 0) == 0) {
 					/* stop here */
-					r1_bio->bios[0]->bi_io_vec[bio->bi_vcnt].bv_page = page;
+					bio->bi_io_vec[bio->bi_vcnt].bv_page = page;
 					while (i > 0) {
 						i--;
 						bio = r1_bio->bios[i];
@@ -1459,12 +1747,28 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 		sync_blocks -= (len>>9);
 	} while (r1_bio->bios[disk]->bi_vcnt < RESYNC_PAGES);
  bio_full:
-	bio = r1_bio->bios[disk];
 	r1_bio->sectors = nr_sectors;
 
-	md_sync_acct(mirror->rdev->bdev, nr_sectors);
+	/* For a user-requested sync, we read all readable devices and do a
+	 * compare
+	 */
+	if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
+		atomic_set(&r1_bio->remaining, read_targets);
+		for (i=0; i<conf->raid_disks; i++) {
+			bio = r1_bio->bios[i];
+			if (bio->bi_end_io == end_sync_read) {
+				md_sync_acct(conf->mirrors[i].rdev->bdev, nr_sectors);
+				generic_make_request(bio);
+			}
+		}
+	} else {
+		atomic_set(&r1_bio->remaining, 1);
+		bio = r1_bio->bios[r1_bio->read_disk];
+		md_sync_acct(conf->mirrors[r1_bio->read_disk].rdev->bdev,
+			     nr_sectors);
+		generic_make_request(bio);
 
-	generic_make_request(bio);
+	}
 
 	return nr_sectors;
 }
@@ -1487,18 +1791,19 @@ static int run(mddev_t *mddev)
 	 * bookkeeping area. [whatever we allocate in run(),
 	 * should be freed in stop()]
 	 */
-	conf = kmalloc(sizeof(conf_t), GFP_KERNEL);
+	conf = kzalloc(sizeof(conf_t), GFP_KERNEL);
 	mddev->private = conf;
 	if (!conf)
 		goto out_no_mem;
 
-	memset(conf, 0, sizeof(*conf));
-	conf->mirrors = kmalloc(sizeof(struct mirror_info)*mddev->raid_disks, 
+	conf->mirrors = kzalloc(sizeof(struct mirror_info)*mddev->raid_disks,
 				 GFP_KERNEL);
 	if (!conf->mirrors)
 		goto out_no_mem;
 
-	memset(conf->mirrors, 0, sizeof(struct mirror_info)*mddev->raid_disks);
+	conf->tmppage = alloc_page(GFP_KERNEL);
+	if (!conf->tmppage)
+		goto out_no_mem;
 
 	conf->poolinfo = kmalloc(sizeof(*conf->poolinfo), GFP_KERNEL);
 	if (!conf->poolinfo)
@@ -1542,8 +1847,7 @@ static int run(mddev_t *mddev)
 		mddev->recovery_cp = MaxSector;
 
 	spin_lock_init(&conf->resync_lock);
-	init_waitqueue_head(&conf->wait_idle);
-	init_waitqueue_head(&conf->wait_resume);
+	init_waitqueue_head(&conf->wait_barrier);
 
 	bio_list_init(&conf->pending_bio_list);
 	bio_list_init(&conf->flushing_bio_list);
@@ -1583,7 +1887,6 @@ static int run(mddev_t *mddev)
 		       mdname(mddev));
 		goto out_free_conf;
 	}
-	if (mddev->bitmap) mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
 
 	printk(KERN_INFO 
 		"raid1: raid set %s active with %d out of %d mirrors\n",
@@ -1608,6 +1911,7 @@ out_free_conf:
 		if (conf->r1bio_pool)
 			mempool_destroy(conf->r1bio_pool);
 		kfree(conf->mirrors);
+		safe_put_page(conf->tmppage);
 		kfree(conf->poolinfo);
 		kfree(conf);
 		mddev->private = NULL;
@@ -1706,19 +2010,14 @@ static int raid1_reshape(mddev_t *mddev, int raid_disks)
 		kfree(newpoolinfo);
 		return -ENOMEM;
 	}
-	newmirrors = kmalloc(sizeof(struct mirror_info) * raid_disks, GFP_KERNEL);
+	newmirrors = kzalloc(sizeof(struct mirror_info) * raid_disks, GFP_KERNEL);
 	if (!newmirrors) {
 		kfree(newpoolinfo);
 		mempool_destroy(newpool);
 		return -ENOMEM;
 	}
-	memset(newmirrors, 0, sizeof(struct mirror_info)*raid_disks);
 
-	spin_lock_irq(&conf->resync_lock);
-	conf->barrier++;
-	wait_event_lock_irq(conf->wait_idle, !conf->nr_pending,
-			    conf->resync_lock, raid1_unplug(mddev->queue));
-	spin_unlock_irq(&conf->resync_lock);
+	raise_barrier(conf);
 
 	/* ok, everything is stopped */
 	oldpool = conf->r1bio_pool;
@@ -1738,12 +2037,7 @@ static int raid1_reshape(mddev_t *mddev, int raid_disks)
 	conf->raid_disks = mddev->raid_disks = raid_disks;
 
 	conf->last_used = 0; /* just make sure it is in-range */
-	spin_lock_irq(&conf->resync_lock);
-	conf->barrier--;
-	spin_unlock_irq(&conf->resync_lock);
-	wake_up(&conf->wait_resume);
-	wake_up(&conf->wait_idle);
-
+	lower_barrier(conf);
 
 	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	md_wakeup_thread(mddev->thread);
@@ -1758,33 +2052,19 @@ static void raid1_quiesce(mddev_t *mddev, int state)
 
 	switch(state) {
 	case 1:
-		spin_lock_irq(&conf->resync_lock);
-		conf->barrier++;
-		wait_event_lock_irq(conf->wait_idle, !conf->nr_pending,
-				    conf->resync_lock, raid1_unplug(mddev->queue));
-		spin_unlock_irq(&conf->resync_lock);
+		raise_barrier(conf);
 		break;
 	case 0:
-		spin_lock_irq(&conf->resync_lock);
-		conf->barrier--;
-		spin_unlock_irq(&conf->resync_lock);
-		wake_up(&conf->wait_resume);
-		wake_up(&conf->wait_idle);
+		lower_barrier(conf);
 		break;
 	}
-	if (mddev->thread) {
-		if (mddev->bitmap)
-			mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
-		else
-			mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
-		md_wakeup_thread(mddev->thread);
-	}
 }
 
 
-static mdk_personality_t raid1_personality =
+static struct mdk_personality raid1_personality =
 {
 	.name		= "raid1",
+	.level		= 1,
 	.owner		= THIS_MODULE,
 	.make_request	= make_request,
 	.run		= run,
@@ -1802,15 +2082,17 @@ static mdk_personality_t raid1_personality =
 
 static int __init raid_init(void)
 {
-	return register_md_personality(RAID1, &raid1_personality);
+	return register_md_personality(&raid1_personality);
 }
 
 static void raid_exit(void)
 {
-	unregister_md_personality(RAID1);
+	unregister_md_personality(&raid1_personality);
 }
 
 module_init(raid_init);
 module_exit(raid_exit);
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("md-personality-3"); /* RAID1 */
+MODULE_ALIAS("md-raid1");
+MODULE_ALIAS("md-level-1");
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 713dc9c2c73..9e658e519a2 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -18,7 +18,9 @@
  * Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
+#include "dm-bio-list.h"
 #include <linux/raid/raid10.h>
+#include <linux/raid/bitmap.h>
 
 /*
  * RAID10 provides a combination of RAID0 and RAID1 functionality.
@@ -47,6 +49,9 @@
 
 static void unplug_slaves(mddev_t *mddev);
 
+static void allow_barrier(conf_t *conf);
+static void lower_barrier(conf_t *conf);
+
 static void * r10bio_pool_alloc(gfp_t gfp_flags, void *data)
 {
 	conf_t *conf = data;
@@ -54,10 +59,8 @@ static void * r10bio_pool_alloc(gfp_t gfp_flags, void *data)
 	int size = offsetof(struct r10bio_s, devs[conf->copies]);
 
 	/* allocate a r10bio with room for raid_disks entries in the bios array */
-	r10_bio = kmalloc(size, gfp_flags);
-	if (r10_bio)
-		memset(r10_bio, 0, size);
-	else
+	r10_bio = kzalloc(size, gfp_flags);
+	if (!r10_bio)
 		unplug_slaves(conf->mddev);
 
 	return r10_bio;
@@ -129,10 +132,10 @@ static void * r10buf_pool_alloc(gfp_t gfp_flags, void *data)
 
 out_free_pages:
 	for ( ; i > 0 ; i--)
-		__free_page(bio->bi_io_vec[i-1].bv_page);
+		safe_put_page(bio->bi_io_vec[i-1].bv_page);
 	while (j--)
 		for (i = 0; i < RESYNC_PAGES ; i++)
-			__free_page(r10_bio->devs[j].bio->bi_io_vec[i].bv_page);
+			safe_put_page(r10_bio->devs[j].bio->bi_io_vec[i].bv_page);
 	j = -1;
 out_free_bio:
 	while ( ++j < nalloc )
@@ -152,7 +155,7 @@ static void r10buf_pool_free(void *__r10_bio, void *data)
 		struct bio *bio = r10bio->devs[j].bio;
 		if (bio) {
 			for (i = 0; i < RESYNC_PAGES; i++) {
-				__free_page(bio->bi_io_vec[i].bv_page);
+				safe_put_page(bio->bi_io_vec[i].bv_page);
 				bio->bi_io_vec[i].bv_page = NULL;
 			}
 			bio_put(bio);
@@ -167,7 +170,7 @@ static void put_all_bios(conf_t *conf, r10bio_t *r10_bio)
 
 	for (i = 0; i < conf->copies; i++) {
 		struct bio **bio = & r10_bio->devs[i].bio;
-		if (*bio)
+		if (*bio && *bio != IO_BLOCKED)
 			bio_put(*bio);
 		*bio = NULL;
 	}
@@ -175,20 +178,13 @@ static void put_all_bios(conf_t *conf, r10bio_t *r10_bio)
 
 static inline void free_r10bio(r10bio_t *r10_bio)
 {
-	unsigned long flags;
-
 	conf_t *conf = mddev_to_conf(r10_bio->mddev);
 
 	/*
 	 * Wake up any possible resync thread that waits for the device
 	 * to go idle.
 	 */
-	spin_lock_irqsave(&conf->resync_lock, flags);
-	if (!--conf->nr_pending) {
-		wake_up(&conf->wait_idle);
-		wake_up(&conf->wait_resume);
-	}
-	spin_unlock_irqrestore(&conf->resync_lock, flags);
+	allow_barrier(conf);
 
 	put_all_bios(conf, r10_bio);
 	mempool_free(r10_bio, conf->r10bio_pool);
@@ -197,22 +193,10 @@ static inline void free_r10bio(r10bio_t *r10_bio)
 static inline void put_buf(r10bio_t *r10_bio)
 {
 	conf_t *conf = mddev_to_conf(r10_bio->mddev);
-	unsigned long flags;
 
 	mempool_free(r10_bio, conf->r10buf_pool);
 
-	spin_lock_irqsave(&conf->resync_lock, flags);
-	if (!conf->barrier)
-		BUG();
-	--conf->barrier;
-	wake_up(&conf->wait_resume);
-	wake_up(&conf->wait_idle);
-
-	if (!--conf->nr_pending) {
-		wake_up(&conf->wait_idle);
-		wake_up(&conf->wait_resume);
-	}
-	spin_unlock_irqrestore(&conf->resync_lock, flags);
+	lower_barrier(conf);
 }
 
 static void reschedule_retry(r10bio_t *r10_bio)
@@ -223,6 +207,7 @@ static void reschedule_retry(r10bio_t *r10_bio)
 
 	spin_lock_irqsave(&conf->device_lock, flags);
 	list_add(&r10_bio->retry_list, &conf->retry_list);
+	conf->nr_queued ++;
 	spin_unlock_irqrestore(&conf->device_lock, flags);
 
 	md_wakeup_thread(mddev->thread);
@@ -268,9 +253,9 @@ static int raid10_end_read_request(struct bio *bio, unsigned int bytes_done, int
 	/*
 	 * this branch is our 'one mirror IO has finished' event handler:
 	 */
-	if (!uptodate)
-		md_error(r10_bio->mddev, conf->mirrors[dev].rdev);
-	else
+	update_head_pos(slot, r10_bio);
+
+	if (uptodate) {
 		/*
 		 * Set R10BIO_Uptodate in our master bio, so that
 		 * we will return a good error code to the higher
@@ -281,15 +266,8 @@ static int raid10_end_read_request(struct bio *bio, unsigned int bytes_done, int
 		 * wait for the 'master' bio.
 		 */
 		set_bit(R10BIO_Uptodate, &r10_bio->state);
-
-	update_head_pos(slot, r10_bio);
-
-	/*
-	 * we have only one bio on the read side
-	 */
-	if (uptodate)
 		raid_end_bio_io(r10_bio);
-	else {
+	} else {
 		/*
 		 * oops, read error:
 		 */
@@ -322,9 +300,11 @@ static int raid10_end_write_request(struct bio *bio, unsigned int bytes_done, in
 	/*
 	 * this branch is our 'one mirror IO has finished' event handler:
 	 */
-	if (!uptodate)
+	if (!uptodate) {
 		md_error(r10_bio->mddev, conf->mirrors[dev].rdev);
-	else
+		/* an I/O failed, we can't clear the bitmap */
+		set_bit(R10BIO_Degraded, &r10_bio->state);
+	} else
 		/*
 		 * Set R10BIO_Uptodate in our master bio, so that
 		 * we will return a good error code for to the higher
@@ -344,6 +324,11 @@ static int raid10_end_write_request(struct bio *bio, unsigned int bytes_done, in
 	 * already.
 	 */
 	if (atomic_dec_and_test(&r10_bio->remaining)) {
+		/* clear the bitmap if all writes complete successfully */
+		bitmap_endwrite(r10_bio->mddev->bitmap, r10_bio->sector,
+				r10_bio->sectors,
+				!test_bit(R10BIO_Degraded, &r10_bio->state),
+				0);
 		md_write_end(r10_bio->mddev);
 		raid_end_bio_io(r10_bio);
 	}
@@ -502,8 +487,9 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio)
 	rcu_read_lock();
 	/*
 	 * Check if we can balance. We can balance on the whole
-	 * device if no resync is going on, or below the resync window.
-	 * We take the first readable disk when above the resync window.
+	 * device if no resync is going on (recovery is ok), or below
+	 * the resync window. We take the first readable disk when
+	 * above the resync window.
 	 */
 	if (conf->mddev->recovery_cp < MaxSector
 	    && (this_sector + sectors >= conf->next_resync)) {
@@ -512,6 +498,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio)
 		disk = r10_bio->devs[slot].devnum;
 
 		while ((rdev = rcu_dereference(conf->mirrors[disk].rdev)) == NULL ||
+		       r10_bio->devs[slot].bio == IO_BLOCKED ||
 		       !test_bit(In_sync, &rdev->flags)) {
 			slot++;
 			if (slot == conf->copies) {
@@ -529,6 +516,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio)
 	slot = 0;
 	disk = r10_bio->devs[slot].devnum;
 	while ((rdev=rcu_dereference(conf->mirrors[disk].rdev)) == NULL ||
+	       r10_bio->devs[slot].bio == IO_BLOCKED ||
 	       !test_bit(In_sync, &rdev->flags)) {
 		slot ++;
 		if (slot == conf->copies) {
@@ -549,6 +537,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio)
 
 
 		if ((rdev=rcu_dereference(conf->mirrors[ndisk].rdev)) == NULL ||
+		    r10_bio->devs[nslot].bio == IO_BLOCKED ||
 		    !test_bit(In_sync, &rdev->flags))
 			continue;
 
@@ -607,7 +596,10 @@ static void unplug_slaves(mddev_t *mddev)
 
 static void raid10_unplug(request_queue_t *q)
 {
+	mddev_t *mddev = q->queuedata;
+
 	unplug_slaves(q->queuedata);
+	md_wakeup_thread(mddev->thread);
 }
 
 static int raid10_issue_flush(request_queue_t *q, struct gendisk *disk,
@@ -640,27 +632,107 @@ static int raid10_issue_flush(request_queue_t *q, struct gendisk *disk,
 	return ret;
 }
 
-/*
- * Throttle resync depth, so that we can both get proper overlapping of
- * requests, but are still able to handle normal requests quickly.
+/* Barriers....
+ * Sometimes we need to suspend IO while we do something else,
+ * either some resync/recovery, or reconfigure the array.
+ * To do this we raise a 'barrier'.
+ * The 'barrier' is a counter that can be raised multiple times
+ * to count how many activities are happening which preclude
+ * normal IO.
+ * We can only raise the barrier if there is no pending IO.
+ * i.e. if nr_pending == 0.
+ * We choose only to raise the barrier if no-one is waiting for the
+ * barrier to go down.  This means that as soon as an IO request
+ * is ready, no other operations which require a barrier will start
+ * until the IO request has had a chance.
+ *
+ * So: regular IO calls 'wait_barrier'.  When that returns there
+ *    is no backgroup IO happening,  It must arrange to call
+ *    allow_barrier when it has finished its IO.
+ * backgroup IO calls must call raise_barrier.  Once that returns
+ *    there is no normal IO happeing.  It must arrange to call
+ *    lower_barrier when the particular background IO completes.
  */
 #define RESYNC_DEPTH 32
 
-static void device_barrier(conf_t *conf, sector_t sect)
+static void raise_barrier(conf_t *conf, int force)
+{
+	BUG_ON(force && !conf->barrier);
+	spin_lock_irq(&conf->resync_lock);
+
+	/* Wait until no block IO is waiting (unless 'force') */
+	wait_event_lock_irq(conf->wait_barrier, force || !conf->nr_waiting,
+			    conf->resync_lock,
+			    raid10_unplug(conf->mddev->queue));
+
+	/* block any new IO from starting */
+	conf->barrier++;
+
+	/* No wait for all pending IO to complete */
+	wait_event_lock_irq(conf->wait_barrier,
+			    !conf->nr_pending && conf->barrier < RESYNC_DEPTH,
+			    conf->resync_lock,
+			    raid10_unplug(conf->mddev->queue));
+
+	spin_unlock_irq(&conf->resync_lock);
+}
+
+static void lower_barrier(conf_t *conf)
+{
+	unsigned long flags;
+	spin_lock_irqsave(&conf->resync_lock, flags);
+	conf->barrier--;
+	spin_unlock_irqrestore(&conf->resync_lock, flags);
+	wake_up(&conf->wait_barrier);
+}
+
+static void wait_barrier(conf_t *conf)
 {
 	spin_lock_irq(&conf->resync_lock);
-	wait_event_lock_irq(conf->wait_idle, !waitqueue_active(&conf->wait_resume),
-			    conf->resync_lock, unplug_slaves(conf->mddev));
-
-	if (!conf->barrier++) {
-		wait_event_lock_irq(conf->wait_idle, !conf->nr_pending,
-				    conf->resync_lock, unplug_slaves(conf->mddev));
-		if (conf->nr_pending)
-			BUG();
+	if (conf->barrier) {
+		conf->nr_waiting++;
+		wait_event_lock_irq(conf->wait_barrier, !conf->barrier,
+				    conf->resync_lock,
+				    raid10_unplug(conf->mddev->queue));
+		conf->nr_waiting--;
 	}
-	wait_event_lock_irq(conf->wait_resume, conf->barrier < RESYNC_DEPTH,
-			    conf->resync_lock, unplug_slaves(conf->mddev));
-	conf->next_resync = sect;
+	conf->nr_pending++;
+	spin_unlock_irq(&conf->resync_lock);
+}
+
+static void allow_barrier(conf_t *conf)
+{
+	unsigned long flags;
+	spin_lock_irqsave(&conf->resync_lock, flags);
+	conf->nr_pending--;
+	spin_unlock_irqrestore(&conf->resync_lock, flags);
+	wake_up(&conf->wait_barrier);
+}
+
+static void freeze_array(conf_t *conf)
+{
+	/* stop syncio and normal IO and wait for everything to
+	 * go quiet.
+	 * We increment barrier and nr_waiting, and then
+	 * wait until barrier+nr_pending match nr_queued+2
+	 */
+	spin_lock_irq(&conf->resync_lock);
+	conf->barrier++;
+	conf->nr_waiting++;
+	wait_event_lock_irq(conf->wait_barrier,
+			    conf->barrier+conf->nr_pending == conf->nr_queued+2,
+			    conf->resync_lock,
+			    raid10_unplug(conf->mddev->queue));
+	spin_unlock_irq(&conf->resync_lock);
+}
+
+static void unfreeze_array(conf_t *conf)
+{
+	/* reverse the effect of the freeze */
+	spin_lock_irq(&conf->resync_lock);
+	conf->barrier--;
+	conf->nr_waiting--;
+	wake_up(&conf->wait_barrier);
 	spin_unlock_irq(&conf->resync_lock);
 }
 
@@ -674,6 +746,8 @@ static int make_request(request_queue_t *q, struct bio * bio)
 	int i;
 	int chunk_sects = conf->chunk_mask + 1;
 	const int rw = bio_data_dir(bio);
+	struct bio_list bl;
+	unsigned long flags;
 
 	if (unlikely(bio_barrier(bio))) {
 		bio_endio(bio, bio->bi_size, -EOPNOTSUPP);
@@ -719,10 +793,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 	 * thread has put up a bar for new requests.
 	 * Continue immediately if no resync is active currently.
 	 */
-	spin_lock_irq(&conf->resync_lock);
-	wait_event_lock_irq(conf->wait_resume, !conf->barrier, conf->resync_lock, );
-	conf->nr_pending++;
-	spin_unlock_irq(&conf->resync_lock);
+	wait_barrier(conf);
 
 	disk_stat_inc(mddev->gendisk, ios[rw]);
 	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
@@ -734,6 +805,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 
 	r10_bio->mddev = mddev;
 	r10_bio->sector = bio->bi_sector;
+	r10_bio->state = 0;
 
 	if (rw == READ) {
 		/*
@@ -778,13 +850,16 @@ static int make_request(request_queue_t *q, struct bio * bio)
 		    !test_bit(Faulty, &rdev->flags)) {
 			atomic_inc(&rdev->nr_pending);
 			r10_bio->devs[i].bio = bio;
-		} else
+		} else {
 			r10_bio->devs[i].bio = NULL;
+			set_bit(R10BIO_Degraded, &r10_bio->state);
+		}
 	}
 	rcu_read_unlock();
 
-	atomic_set(&r10_bio->remaining, 1);
+	atomic_set(&r10_bio->remaining, 0);
 
+	bio_list_init(&bl);
 	for (i = 0; i < conf->copies; i++) {
 		struct bio *mbio;
 		int d = r10_bio->devs[i].devnum;
@@ -802,13 +877,14 @@ static int make_request(request_queue_t *q, struct bio * bio)
 		mbio->bi_private = r10_bio;
 
 		atomic_inc(&r10_bio->remaining);
-		generic_make_request(mbio);
+		bio_list_add(&bl, mbio);
 	}
 
-	if (atomic_dec_and_test(&r10_bio->remaining)) {
-		md_write_end(mddev);
-		raid_end_bio_io(r10_bio);
-	}
+	bitmap_startwrite(mddev->bitmap, bio->bi_sector, r10_bio->sectors, 0);
+	spin_lock_irqsave(&conf->device_lock, flags);
+	bio_list_merge(&conf->pending_bio_list, &bl);
+	blk_plug_device(mddev->queue);
+	spin_unlock_irqrestore(&conf->device_lock, flags);
 
 	return 0;
 }
@@ -897,13 +973,8 @@ static void print_conf(conf_t *conf)
 
 static void close_sync(conf_t *conf)
 {
-	spin_lock_irq(&conf->resync_lock);
-	wait_event_lock_irq(conf->wait_resume, !conf->barrier,
-			    conf->resync_lock, 	unplug_slaves(conf->mddev));
-	spin_unlock_irq(&conf->resync_lock);
-
-	if (conf->barrier) BUG();
-	if (waitqueue_active(&conf->wait_idle)) BUG();
+	wait_barrier(conf);
+	allow_barrier(conf);
 
 	mempool_destroy(conf->r10buf_pool);
 	conf->r10buf_pool = NULL;
@@ -971,7 +1042,12 @@ static int raid10_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 	if (!enough(conf))
 		return 0;
 
-	for (mirror=0; mirror < mddev->raid_disks; mirror++)
+	if (rdev->saved_raid_disk >= 0 &&
+	    conf->mirrors[rdev->saved_raid_disk].rdev == NULL)
+		mirror = rdev->saved_raid_disk;
+	else
+		mirror = 0;
+	for ( ; mirror < mddev->raid_disks; mirror++)
 		if ( !(p=conf->mirrors+mirror)->rdev) {
 
 			blk_queue_stack_limits(mddev->queue,
@@ -987,6 +1063,8 @@ static int raid10_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 			p->head_position = 0;
 			rdev->raid_disk = mirror;
 			found = 1;
+			if (rdev->saved_raid_disk != mirror)
+				conf->fullsync = 1;
 			rcu_assign_pointer(p->rdev, rdev);
 			break;
 		}
@@ -1027,7 +1105,6 @@ abort:
 
 static int end_sync_read(struct bio *bio, unsigned int bytes_done, int error)
 {
-	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	r10bio_t * r10_bio = (r10bio_t *)(bio->bi_private);
 	conf_t *conf = mddev_to_conf(r10_bio->mddev);
 	int i,d;
@@ -1042,9 +1119,16 @@ static int end_sync_read(struct bio *bio, unsigned int bytes_done, int error)
 		BUG();
 	update_head_pos(i, r10_bio);
 	d = r10_bio->devs[i].devnum;
-	if (!uptodate)
-		md_error(r10_bio->mddev,
-			 conf->mirrors[d].rdev);
+
+	if (test_bit(BIO_UPTODATE, &bio->bi_flags))
+		set_bit(R10BIO_Uptodate, &r10_bio->state);
+	else {
+		atomic_add(r10_bio->sectors,
+			   &conf->mirrors[d].rdev->corrected_errors);
+		if (!test_bit(MD_RECOVERY_SYNC, &conf->mddev->recovery))
+			md_error(r10_bio->mddev,
+				 conf->mirrors[d].rdev);
+	}
 
 	/* for reconstruct, we always reschedule after a read.
 	 * for resync, only after all reads
@@ -1132,23 +1216,32 @@ static void sync_request_write(mddev_t *mddev, r10bio_t *r10_bio)
 	fbio = r10_bio->devs[i].bio;
 
 	/* now find blocks with errors */
-	for (i=first+1 ; i < conf->copies ; i++) {
-		int vcnt, j, d;
+	for (i=0 ; i < conf->copies ; i++) {
+		int  j, d;
+		int vcnt = r10_bio->sectors >> (PAGE_SHIFT-9);
 
-		if (!test_bit(BIO_UPTODATE, &r10_bio->devs[i].bio->bi_flags))
-			continue;
-		/* We know that the bi_io_vec layout is the same for
-		 * both 'first' and 'i', so we just compare them.
-		 * All vec entries are PAGE_SIZE;
-		 */
 		tbio = r10_bio->devs[i].bio;
-		vcnt = r10_bio->sectors >> (PAGE_SHIFT-9);
-		for (j = 0; j < vcnt; j++)
-			if (memcmp(page_address(fbio->bi_io_vec[j].bv_page),
-				   page_address(tbio->bi_io_vec[j].bv_page),
-				   PAGE_SIZE))
-				break;
-		if (j == vcnt)
+
+		if (tbio->bi_end_io != end_sync_read)
+			continue;
+		if (i == first)
+			continue;
+		if (test_bit(BIO_UPTODATE, &r10_bio->devs[i].bio->bi_flags)) {
+			/* We know that the bi_io_vec layout is the same for
+			 * both 'first' and 'i', so we just compare them.
+			 * All vec entries are PAGE_SIZE;
+			 */
+			for (j = 0; j < vcnt; j++)
+				if (memcmp(page_address(fbio->bi_io_vec[j].bv_page),
+					   page_address(tbio->bi_io_vec[j].bv_page),
+					   PAGE_SIZE))
+					break;
+			if (j == vcnt)
+				continue;
+			mddev->resync_mismatches += r10_bio->sectors;
+		}
+		if (test_bit(MD_RECOVERY_CHECK, &mddev->recovery))
+			/* Don't fix anything. */
 			continue;
 		/* Ok, we need to write this bio
 		 * First we need to fixup bv_offset, bv_len and
@@ -1227,7 +1320,10 @@ static void recovery_request_write(mddev_t *mddev, r10bio_t *r10_bio)
 
 	atomic_inc(&conf->mirrors[d].rdev->nr_pending);
 	md_sync_acct(conf->mirrors[d].rdev->bdev, wbio->bi_size >> 9);
-	generic_make_request(wbio);
+	if (test_bit(R10BIO_Uptodate, &r10_bio->state))
+		generic_make_request(wbio);
+	else
+		bio_endio(wbio, wbio->bi_size, -EIO);
 }
 
 
@@ -1254,10 +1350,31 @@ static void raid10d(mddev_t *mddev)
 	for (;;) {
 		char b[BDEVNAME_SIZE];
 		spin_lock_irqsave(&conf->device_lock, flags);
+
+		if (conf->pending_bio_list.head) {
+			bio = bio_list_get(&conf->pending_bio_list);
+			blk_remove_plug(mddev->queue);
+			spin_unlock_irqrestore(&conf->device_lock, flags);
+			/* flush any pending bitmap writes to disk before proceeding w/ I/O */
+			if (bitmap_unplug(mddev->bitmap) != 0)
+				printk("%s: bitmap file write failed!\n", mdname(mddev));
+
+			while (bio) { /* submit pending writes */
+				struct bio *next = bio->bi_next;
+				bio->bi_next = NULL;
+				generic_make_request(bio);
+				bio = next;
+			}
+			unplug = 1;
+
+			continue;
+		}
+
 		if (list_empty(head))
 			break;
 		r10_bio = list_entry(head->prev, r10bio_t, retry_list);
 		list_del(head->prev);
+		conf->nr_queued--;
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 
 		mddev = r10_bio->mddev;
@@ -1270,8 +1387,96 @@ static void raid10d(mddev_t *mddev)
 			unplug = 1;
 		} else {
 			int mirror;
+			/* we got a read error. Maybe the drive is bad.  Maybe just
+			 * the block and we can fix it.
+			 * We freeze all other IO, and try reading the block from
+			 * other devices.  When we find one, we re-write
+			 * and check it that fixes the read error.
+			 * This is all done synchronously while the array is
+			 * frozen.
+			 */
+			int sect = 0; /* Offset from r10_bio->sector */
+			int sectors = r10_bio->sectors;
+			freeze_array(conf);
+			if (mddev->ro == 0) while(sectors) {
+				int s = sectors;
+				int sl = r10_bio->read_slot;
+				int success = 0;
+
+				if (s > (PAGE_SIZE>>9))
+					s = PAGE_SIZE >> 9;
+
+				do {
+					int d = r10_bio->devs[sl].devnum;
+					rdev = conf->mirrors[d].rdev;
+					if (rdev &&
+					    test_bit(In_sync, &rdev->flags) &&
+					    sync_page_io(rdev->bdev,
+							 r10_bio->devs[sl].addr +
+							 sect + rdev->data_offset,
+							 s<<9,
+							 conf->tmppage, READ))
+						success = 1;
+					else {
+						sl++;
+						if (sl == conf->copies)
+							sl = 0;
+					}
+				} while (!success && sl != r10_bio->read_slot);
+
+				if (success) {
+					int start = sl;
+					/* write it back and re-read */
+					while (sl != r10_bio->read_slot) {
+						int d;
+						if (sl==0)
+							sl = conf->copies;
+						sl--;
+						d = r10_bio->devs[sl].devnum;
+						rdev = conf->mirrors[d].rdev;
+						atomic_add(s, &rdev->corrected_errors);
+						if (rdev &&
+						    test_bit(In_sync, &rdev->flags)) {
+							if (sync_page_io(rdev->bdev,
+									 r10_bio->devs[sl].addr +
+									 sect + rdev->data_offset,
+									 s<<9, conf->tmppage, WRITE) == 0)
+								/* Well, this device is dead */
+								md_error(mddev, rdev);
+						}
+					}
+					sl = start;
+					while (sl != r10_bio->read_slot) {
+						int d;
+						if (sl==0)
+							sl = conf->copies;
+						sl--;
+						d = r10_bio->devs[sl].devnum;
+						rdev = conf->mirrors[d].rdev;
+						if (rdev &&
+						    test_bit(In_sync, &rdev->flags)) {
+							if (sync_page_io(rdev->bdev,
+									 r10_bio->devs[sl].addr +
+									 sect + rdev->data_offset,
+									 s<<9, conf->tmppage, READ) == 0)
+								/* Well, this device is dead */
+								md_error(mddev, rdev);
+						}
+					}
+				} else {
+					/* Cannot read from anywhere -- bye bye array */
+					md_error(mddev, conf->mirrors[r10_bio->devs[r10_bio->read_slot].devnum].rdev);
+					break;
+				}
+				sectors -= s;
+				sect += s;
+			}
+
+			unfreeze_array(conf);
+
 			bio = r10_bio->devs[r10_bio->read_slot].bio;
-			r10_bio->devs[r10_bio->read_slot].bio = NULL;
+			r10_bio->devs[r10_bio->read_slot].bio =
+				mddev->ro ? IO_BLOCKED : NULL;
 			bio_put(bio);
 			mirror = read_balance(conf, r10_bio);
 			if (mirror == -1) {
@@ -1360,6 +1565,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 	sector_t max_sector, nr_sectors;
 	int disk;
 	int i;
+	int max_sync;
+	int sync_blocks;
 
 	sector_t sectors_skipped = 0;
 	int chunks_skipped = 0;
@@ -1373,6 +1580,29 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
 		max_sector = mddev->resync_max_sectors;
 	if (sector_nr >= max_sector) {
+		/* If we aborted, we need to abort the
+		 * sync on the 'current' bitmap chucks (there can
+		 * be several when recovering multiple devices).
+		 * as we may have started syncing it but not finished.
+		 * We can find the current address in
+		 * mddev->curr_resync, but for recovery,
+		 * we need to convert that to several
+		 * virtual addresses.
+		 */
+		if (mddev->curr_resync < max_sector) { /* aborted */
+			if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
+				bitmap_end_sync(mddev->bitmap, mddev->curr_resync,
+						&sync_blocks, 1);
+			else for (i=0; i<conf->raid_disks; i++) {
+				sector_t sect =
+					raid10_find_virt(conf, mddev->curr_resync, i);
+				bitmap_end_sync(mddev->bitmap, sect,
+						&sync_blocks, 1);
+			}
+		} else /* completed sync */
+			conf->fullsync = 0;
+
+		bitmap_close_sync(mddev->bitmap);
 		close_sync(conf);
 		*skipped = 1;
 		return sectors_skipped;
@@ -1395,9 +1625,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 	 * If there is non-resync activity waiting for us then
 	 * put in a delay to throttle resync.
 	 */
-	if (!go_faster && waitqueue_active(&conf->wait_resume))
+	if (!go_faster && conf->nr_waiting)
 		msleep_interruptible(1000);
-	device_barrier(conf, sector_nr + RESYNC_SECTORS);
 
 	/* Again, very different code for resync and recovery.
 	 * Both must result in an r10bio with a list of bios that
@@ -1414,6 +1643,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 	 * end_sync_write if we will want to write.
 	 */
 
+	max_sync = RESYNC_PAGES << (PAGE_SHIFT-9);
 	if (!test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
 		/* recovery... the complicated one */
 		int i, j, k;
@@ -1422,14 +1652,29 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 		for (i=0 ; i<conf->raid_disks; i++)
 			if (conf->mirrors[i].rdev &&
 			    !test_bit(In_sync, &conf->mirrors[i].rdev->flags)) {
+				int still_degraded = 0;
 				/* want to reconstruct this device */
 				r10bio_t *rb2 = r10_bio;
+				sector_t sect = raid10_find_virt(conf, sector_nr, i);
+				int must_sync;
+				/* Unless we are doing a full sync, we only need
+				 * to recover the block if it is set in the bitmap
+				 */
+				must_sync = bitmap_start_sync(mddev->bitmap, sect,
+							      &sync_blocks, 1);
+				if (sync_blocks < max_sync)
+					max_sync = sync_blocks;
+				if (!must_sync &&
+				    !conf->fullsync) {
+					/* yep, skip the sync_blocks here, but don't assume
+					 * that there will never be anything to do here
+					 */
+					chunks_skipped = -1;
+					continue;
+				}
 
 				r10_bio = mempool_alloc(conf->r10buf_pool, GFP_NOIO);
-				spin_lock_irq(&conf->resync_lock);
-				conf->nr_pending++;
-				if (rb2) conf->barrier++;
-				spin_unlock_irq(&conf->resync_lock);
+				raise_barrier(conf, rb2 != NULL);
 				atomic_set(&r10_bio->remaining, 0);
 
 				r10_bio->master_bio = (struct bio*)rb2;
@@ -1437,8 +1682,23 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 					atomic_inc(&rb2->remaining);
 				r10_bio->mddev = mddev;
 				set_bit(R10BIO_IsRecover, &r10_bio->state);
-				r10_bio->sector = raid10_find_virt(conf, sector_nr, i);
+				r10_bio->sector = sect;
+
 				raid10_find_phys(conf, r10_bio);
+				/* Need to check if this section will still be
+				 * degraded
+				 */
+				for (j=0; j<conf->copies;j++) {
+					int d = r10_bio->devs[j].devnum;
+					if (conf->mirrors[d].rdev == NULL ||
+					    test_bit(Faulty, &conf->mirrors[d].rdev->flags)) {
+						still_degraded = 1;
+						break;
+					}
+				}
+				must_sync = bitmap_start_sync(mddev->bitmap, sect,
+							      &sync_blocks, still_degraded);
+
 				for (j=0; j<conf->copies;j++) {
 					int d = r10_bio->devs[j].devnum;
 					if (conf->mirrors[d].rdev &&
@@ -1498,14 +1758,22 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 	} else {
 		/* resync. Schedule a read for every block at this virt offset */
 		int count = 0;
-		r10_bio = mempool_alloc(conf->r10buf_pool, GFP_NOIO);
 
-		spin_lock_irq(&conf->resync_lock);
-		conf->nr_pending++;
-		spin_unlock_irq(&conf->resync_lock);
+		if (!bitmap_start_sync(mddev->bitmap, sector_nr,
+				       &sync_blocks, mddev->degraded) &&
+		    !conf->fullsync && !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
+			/* We can skip this block */
+			*skipped = 1;
+			return sync_blocks + sectors_skipped;
+		}
+		if (sync_blocks < max_sync)
+			max_sync = sync_blocks;
+		r10_bio = mempool_alloc(conf->r10buf_pool, GFP_NOIO);
 
 		r10_bio->mddev = mddev;
 		atomic_set(&r10_bio->remaining, 0);
+		raise_barrier(conf, 0);
+		conf->next_resync = sector_nr;
 
 		r10_bio->master_bio = NULL;
 		r10_bio->sector = sector_nr;
@@ -1558,6 +1826,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 	}
 
 	nr_sectors = 0;
+	if (sector_nr + max_sync < max_sector)
+		max_sector = sector_nr + max_sync;
 	do {
 		struct page *page;
 		int len = PAGE_SIZE;
@@ -1632,11 +1902,11 @@ static int run(mddev_t *mddev)
 	int nc, fc;
 	sector_t stride, size;
 
-	if (mddev->level != 10) {
-		printk(KERN_ERR "raid10: %s: raid level not set correctly... (%d)\n",
-		       mdname(mddev), mddev->level);
-		goto out;
+	if (mddev->chunk_size == 0) {
+		printk(KERN_ERR "md/raid10: non-zero chunk size required.\n");
+		return -EINVAL;
 	}
+
 	nc = mddev->layout & 255;
 	fc = (mddev->layout >> 8) & 255;
 	if ((nc*fc) <2 || (nc*fc) > mddev->raid_disks ||
@@ -1650,22 +1920,24 @@ static int run(mddev_t *mddev)
 	 * bookkeeping area. [whatever we allocate in run(),
 	 * should be freed in stop()]
 	 */
-	conf = kmalloc(sizeof(conf_t), GFP_KERNEL);
+	conf = kzalloc(sizeof(conf_t), GFP_KERNEL);
 	mddev->private = conf;
 	if (!conf) {
 		printk(KERN_ERR "raid10: couldn't allocate memory for %s\n",
 			mdname(mddev));
 		goto out;
 	}
-	memset(conf, 0, sizeof(*conf));
-	conf->mirrors = kmalloc(sizeof(struct mirror_info)*mddev->raid_disks,
+	conf->mirrors = kzalloc(sizeof(struct mirror_info)*mddev->raid_disks,
 				 GFP_KERNEL);
 	if (!conf->mirrors) {
 		printk(KERN_ERR "raid10: couldn't allocate memory for %s\n",
 		       mdname(mddev));
 		goto out_free_conf;
 	}
-	memset(conf->mirrors, 0, sizeof(struct mirror_info)*mddev->raid_disks);
+
+	conf->tmppage = alloc_page(GFP_KERNEL);
+	if (!conf->tmppage)
+		goto out_free_conf;
 
 	conf->near_copies = nc;
 	conf->far_copies = fc;
@@ -1713,8 +1985,7 @@ static int run(mddev_t *mddev)
 	INIT_LIST_HEAD(&conf->retry_list);
 
 	spin_lock_init(&conf->resync_lock);
-	init_waitqueue_head(&conf->wait_idle);
-	init_waitqueue_head(&conf->wait_resume);
+	init_waitqueue_head(&conf->wait_barrier);
 
 	/* need to check that every block has at least one working mirror */
 	if (!enough(conf)) {
@@ -1763,7 +2034,7 @@ static int run(mddev_t *mddev)
 	 * maybe...
 	 */
 	{
-		int stripe = conf->raid_disks * mddev->chunk_size / PAGE_CACHE_SIZE;
+		int stripe = conf->raid_disks * mddev->chunk_size / 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;
@@ -1776,6 +2047,7 @@ static int run(mddev_t *mddev)
 out_free_conf:
 	if (conf->r10bio_pool)
 		mempool_destroy(conf->r10bio_pool);
+	safe_put_page(conf->tmppage);
 	kfree(conf->mirrors);
 	kfree(conf);
 	mddev->private = NULL;
@@ -1798,10 +2070,31 @@ static int stop(mddev_t *mddev)
 	return 0;
 }
 
+static void raid10_quiesce(mddev_t *mddev, int state)
+{
+	conf_t *conf = mddev_to_conf(mddev);
+
+	switch(state) {
+	case 1:
+		raise_barrier(conf, 0);
+		break;
+	case 0:
+		lower_barrier(conf);
+		break;
+	}
+	if (mddev->thread) {
+		if (mddev->bitmap)
+			mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
+		else
+			mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
+		md_wakeup_thread(mddev->thread);
+	}
+}
 
-static mdk_personality_t raid10_personality =
+static struct mdk_personality raid10_personality =
 {
 	.name		= "raid10",
+	.level		= 10,
 	.owner		= THIS_MODULE,
 	.make_request	= make_request,
 	.run		= run,
@@ -1812,19 +2105,22 @@ static mdk_personality_t raid10_personality =
 	.hot_remove_disk= raid10_remove_disk,
 	.spare_active	= raid10_spare_active,
 	.sync_request	= sync_request,
+	.quiesce	= raid10_quiesce,
 };
 
 static int __init raid_init(void)
 {
-	return register_md_personality(RAID10, &raid10_personality);
+	return register_md_personality(&raid10_personality);
 }
 
 static void raid_exit(void)
 {
-	unregister_md_personality(RAID10);
+	unregister_md_personality(&raid10_personality);
 }
 
 module_init(raid_init);
 module_exit(raid_exit);
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("md-personality-9"); /* RAID10 */
+MODULE_ALIAS("md-raid10");
+MODULE_ALIAS("md-level-10");
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index fafc4bc045f..54f4a9847e3 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -35,12 +35,10 @@
 #define STRIPE_SHIFT		(PAGE_SHIFT - 9)
 #define STRIPE_SECTORS		(STRIPE_SIZE>>9)
 #define	IO_THRESHOLD		1
-#define HASH_PAGES		1
-#define HASH_PAGES_ORDER	0
-#define NR_HASH			(HASH_PAGES * PAGE_SIZE / sizeof(struct stripe_head *))
+#define NR_HASH			(PAGE_SIZE / sizeof(struct hlist_head))
 #define HASH_MASK		(NR_HASH - 1)
 
-#define stripe_hash(conf, sect)	((conf)->stripe_hashtbl[((sect) >> STRIPE_SHIFT) & HASH_MASK])
+#define stripe_hash(conf, sect)	(&((conf)->stripe_hashtbl[((sect) >> STRIPE_SHIFT) & HASH_MASK]))
 
 /* bio's attached to a stripe+device for I/O are linked together in bi_sector
  * order without overlap.  There may be several bio's per stripe+device, and
@@ -113,29 +111,21 @@ static void release_stripe(struct stripe_head *sh)
 	spin_unlock_irqrestore(&conf->device_lock, flags);
 }
 
-static void remove_hash(struct stripe_head *sh)
+static inline void remove_hash(struct stripe_head *sh)
 {
 	PRINTK("remove_hash(), stripe %llu\n", (unsigned long long)sh->sector);
 
-	if (sh->hash_pprev) {
-		if (sh->hash_next)
-			sh->hash_next->hash_pprev = sh->hash_pprev;
-		*sh->hash_pprev = sh->hash_next;
-		sh->hash_pprev = NULL;
-	}
+	hlist_del_init(&sh->hash);
 }
 
-static __inline__ void insert_hash(raid5_conf_t *conf, struct stripe_head *sh)
+static inline void insert_hash(raid5_conf_t *conf, struct stripe_head *sh)
 {
-	struct stripe_head **shp = &stripe_hash(conf, sh->sector);
+	struct hlist_head *hp = stripe_hash(conf, sh->sector);
 
 	PRINTK("insert_hash(), stripe %llu\n", (unsigned long long)sh->sector);
 
 	CHECK_DEVLOCK();
-	if ((sh->hash_next = *shp) != NULL)
-		(*shp)->hash_pprev = &sh->hash_next;
-	*shp = sh;
-	sh->hash_pprev = shp;
+	hlist_add_head(&sh->hash, hp);
 }
 
 
@@ -167,7 +157,7 @@ static void shrink_buffers(struct stripe_head *sh, int num)
 		if (!p)
 			continue;
 		sh->dev[i].page = NULL;
-		page_cache_release(p);
+		put_page(p);
 	}
 }
 
@@ -228,10 +218,11 @@ static inline void init_stripe(struct stripe_head *sh, sector_t sector, int pd_i
 static struct stripe_head *__find_stripe(raid5_conf_t *conf, sector_t sector)
 {
 	struct stripe_head *sh;
+	struct hlist_node *hn;
 
 	CHECK_DEVLOCK();
 	PRINTK("__find_stripe, sector %llu\n", (unsigned long long)sector);
-	for (sh = stripe_hash(conf, sector); sh; sh = sh->hash_next)
+	hlist_for_each_entry(sh, hn, stripe_hash(conf, sector), hash)
 		if (sh->sector == sector)
 			return sh;
 	PRINTK("__stripe %llu not in cache\n", (unsigned long long)sector);
@@ -417,7 +408,7 @@ static int raid5_end_read_request(struct bio * bi, unsigned int bytes_done,
 		set_bit(R5_UPTODATE, &sh->dev[i].flags);
 #endif
 		if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
-			printk("R5: read error corrected!!\n");
+			printk(KERN_INFO "raid5: read error corrected!!\n");
 			clear_bit(R5_ReadError, &sh->dev[i].flags);
 			clear_bit(R5_ReWrite, &sh->dev[i].flags);
 		}
@@ -428,13 +419,14 @@ static int raid5_end_read_request(struct bio * bi, unsigned int bytes_done,
 		clear_bit(R5_UPTODATE, &sh->dev[i].flags);
 		atomic_inc(&conf->disks[i].rdev->read_errors);
 		if (conf->mddev->degraded)
-			printk("R5: read error not correctable.\n");
+			printk(KERN_WARNING "raid5: read error not correctable.\n");
 		else if (test_bit(R5_ReWrite, &sh->dev[i].flags))
 			/* Oh, no!!! */
-			printk("R5: read error NOT corrected!!\n");
+			printk(KERN_WARNING "raid5: read error NOT corrected!!\n");
 		else if (atomic_read(&conf->disks[i].rdev->read_errors)
 			 > conf->max_nr_stripes)
-			printk("raid5: Too many read errors, failing device.\n");
+			printk(KERN_WARNING
+			       "raid5: Too many read errors, failing device.\n");
 		else
 			retry = 1;
 		if (retry)
@@ -604,7 +596,7 @@ static sector_t raid5_compute_sector(sector_t r_sector, unsigned int raid_disks,
 			*dd_idx = (*pd_idx + 1 + *dd_idx) % raid_disks;
 			break;
 		default:
-			printk("raid5: unsupported algorithm %d\n",
+			printk(KERN_ERR "raid5: unsupported algorithm %d\n",
 				conf->algorithm);
 	}
 
@@ -645,7 +637,7 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i)
 			i -= (sh->pd_idx + 1);
 			break;
 		default:
-			printk("raid5: unsupported algorithm %d\n",
+			printk(KERN_ERR "raid5: unsupported algorithm %d\n",
 				conf->algorithm);
 	}
 
@@ -654,7 +646,7 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i)
 
 	check = raid5_compute_sector (r_sector, raid_disks, data_disks, &dummy1, &dummy2, conf);
 	if (check != sh->sector || dummy1 != dd_idx || dummy2 != sh->pd_idx) {
-		printk("compute_blocknr: map not correct\n");
+		printk(KERN_ERR "compute_blocknr: map not correct\n");
 		return 0;
 	}
 	return r_sector;
@@ -737,7 +729,7 @@ static void compute_block(struct stripe_head *sh, int dd_idx)
 		if (test_bit(R5_UPTODATE, &sh->dev[i].flags))
 			ptr[count++] = p;
 		else
-			printk("compute_block() %d, stripe %llu, %d"
+			printk(KERN_ERR "compute_block() %d, stripe %llu, %d"
 				" not present\n", dd_idx,
 				(unsigned long long)sh->sector, i);
 
@@ -960,11 +952,11 @@ static void handle_stripe(struct stripe_head *sh)
 	syncing = test_bit(STRIPE_SYNCING, &sh->state);
 	/* Now to look around and see what can be done */
 
+	rcu_read_lock();
 	for (i=disks; i--; ) {
 		mdk_rdev_t *rdev;
 		dev = &sh->dev[i];
 		clear_bit(R5_Insync, &dev->flags);
-		clear_bit(R5_Syncio, &dev->flags);
 
 		PRINTK("check %d: state 0x%lx read %p write %p written %p\n",
 			i, dev->flags, dev->toread, dev->towrite, dev->written);
@@ -1003,9 +995,9 @@ static void handle_stripe(struct stripe_head *sh)
 				non_overwrite++;
 		}
 		if (dev->written) written++;
-		rdev = conf->disks[i].rdev; /* FIXME, should I be looking rdev */
+		rdev = rcu_dereference(conf->disks[i].rdev);
 		if (!rdev || !test_bit(In_sync, &rdev->flags)) {
-			/* The ReadError flag wil just be confusing now */
+			/* The ReadError flag will just be confusing now */
 			clear_bit(R5_ReadError, &dev->flags);
 			clear_bit(R5_ReWrite, &dev->flags);
 		}
@@ -1016,6 +1008,7 @@ static void handle_stripe(struct stripe_head *sh)
 		} else
 			set_bit(R5_Insync, &dev->flags);
 	}
+	rcu_read_unlock();
 	PRINTK("locked=%d uptodate=%d to_read=%d"
 		" to_write=%d failed=%d failed_num=%d\n",
 		locked, uptodate, to_read, to_write, failed, failed_num);
@@ -1027,10 +1020,13 @@ static void handle_stripe(struct stripe_head *sh)
 			int bitmap_end = 0;
 
 			if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
-				mdk_rdev_t *rdev = conf->disks[i].rdev;
+				mdk_rdev_t *rdev;
+				rcu_read_lock();
+				rdev = rcu_dereference(conf->disks[i].rdev);
 				if (rdev && test_bit(In_sync, &rdev->flags))
 					/* multiple read failures in one stripe */
 					md_error(conf->mddev, rdev);
+				rcu_read_unlock();
 			}
 
 			spin_lock_irq(&conf->device_lock);
@@ -1179,9 +1175,6 @@ static void handle_stripe(struct stripe_head *sh)
 					locked++;
 					PRINTK("Reading block %d (sync=%d)\n", 
 						i, syncing);
-					if (syncing)
-						md_sync_acct(conf->disks[i].rdev->bdev,
-							     STRIPE_SECTORS);
 				}
 			}
 		}
@@ -1288,7 +1281,7 @@ static void handle_stripe(struct stripe_head *sh)
 	 * is available
 	 */
 	if (syncing && locked == 0 &&
-	    !test_bit(STRIPE_INSYNC, &sh->state) && failed <= 1) {
+	    !test_bit(STRIPE_INSYNC, &sh->state)) {
 		set_bit(STRIPE_HANDLE, &sh->state);
 		if (failed == 0) {
 			char *pagea;
@@ -1306,27 +1299,25 @@ static void handle_stripe(struct stripe_head *sh)
 				if (test_bit(MD_RECOVERY_CHECK, &conf->mddev->recovery))
 					/* don't try to repair!! */
 					set_bit(STRIPE_INSYNC, &sh->state);
+				else {
+					compute_block(sh, sh->pd_idx);
+					uptodate++;
+				}
 			}
 		}
 		if (!test_bit(STRIPE_INSYNC, &sh->state)) {
+			/* either failed parity check, or recovery is happening */
 			if (failed==0)
 				failed_num = sh->pd_idx;
-			/* should be able to compute the missing block and write it to spare */
-			if (!test_bit(R5_UPTODATE, &sh->dev[failed_num].flags)) {
-				if (uptodate+1 != disks)
-					BUG();
-				compute_block(sh, failed_num);
-				uptodate++;
-			}
-			if (uptodate != disks)
-				BUG();
 			dev = &sh->dev[failed_num];
+			BUG_ON(!test_bit(R5_UPTODATE, &dev->flags));
+			BUG_ON(uptodate != disks);
+
 			set_bit(R5_LOCKED, &dev->flags);
 			set_bit(R5_Wantwrite, &dev->flags);
 			clear_bit(STRIPE_DEGRADED, &sh->state);
 			locked++;
 			set_bit(STRIPE_INSYNC, &sh->state);
-			set_bit(R5_Syncio, &dev->flags);
 		}
 	}
 	if (syncing && locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
@@ -1392,7 +1383,7 @@ static void handle_stripe(struct stripe_head *sh)
 		rcu_read_unlock();
  
 		if (rdev) {
-			if (test_bit(R5_Syncio, &sh->dev[i].flags))
+			if (syncing)
 				md_sync_acct(rdev->bdev, STRIPE_SECTORS);
 
 			bi->bi_bdev = rdev->bdev;
@@ -1409,6 +1400,9 @@ static void handle_stripe(struct stripe_head *sh)
 			bi->bi_io_vec[0].bv_offset = 0;
 			bi->bi_size = STRIPE_SIZE;
 			bi->bi_next = NULL;
+			if (rw == WRITE &&
+			    test_bit(R5_ReWrite, &sh->dev[i].flags))
+				atomic_add(STRIPE_SECTORS, &rdev->corrected_errors);
 			generic_make_request(bi);
 		} else {
 			if (rw == 1)
@@ -1822,21 +1816,21 @@ static int run(mddev_t *mddev)
 	struct list_head *tmp;
 
 	if (mddev->level != 5 && mddev->level != 4) {
-		printk("raid5: %s: raid level not set to 4/5 (%d)\n", mdname(mddev), mddev->level);
+		printk(KERN_ERR "raid5: %s: raid level not set to 4/5 (%d)\n",
+		       mdname(mddev), mddev->level);
 		return -EIO;
 	}
 
-	mddev->private = kmalloc (sizeof (raid5_conf_t)
-				  + mddev->raid_disks * sizeof(struct disk_info),
-				  GFP_KERNEL);
+	mddev->private = kzalloc(sizeof (raid5_conf_t)
+				 + mddev->raid_disks * sizeof(struct disk_info),
+				 GFP_KERNEL);
 	if ((conf = mddev->private) == NULL)
 		goto abort;
-	memset (conf, 0, sizeof (*conf) + mddev->raid_disks * sizeof(struct disk_info) );
+
 	conf->mddev = mddev;
 
-	if ((conf->stripe_hashtbl = (struct stripe_head **) __get_free_pages(GFP_ATOMIC, HASH_PAGES_ORDER)) == NULL)
+	if ((conf->stripe_hashtbl = kzalloc(PAGE_SIZE, GFP_KERNEL)) == NULL)
 		goto abort;
-	memset(conf->stripe_hashtbl, 0, HASH_PAGES * PAGE_SIZE);
 
 	spin_lock_init(&conf->device_lock);
 	init_waitqueue_head(&conf->wait_for_stripe);
@@ -1903,10 +1897,17 @@ static int run(mddev_t *mddev)
 
 	if (mddev->degraded == 1 &&
 	    mddev->recovery_cp != MaxSector) {
-		printk(KERN_ERR 
-			"raid5: cannot start dirty degraded array for %s\n",
-			mdname(mddev));
-		goto abort;
+		if (mddev->ok_start_degraded)
+			printk(KERN_WARNING
+			       "raid5: starting dirty degraded array: %s"
+			       "- data corruption possible.\n",
+			       mdname(mddev));
+		else {
+			printk(KERN_ERR
+			       "raid5: cannot start dirty degraded array for %s\n",
+			       mdname(mddev));
+			goto abort;
+		}
 	}
 
 	{
@@ -1948,7 +1949,7 @@ static int run(mddev_t *mddev)
 	 */
 	{
 		int stripe = (mddev->raid_disks-1) * mddev->chunk_size
-			/ PAGE_CACHE_SIZE;
+			/ PAGE_SIZE;
 		if (mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
 			mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
 	}
@@ -1956,9 +1957,6 @@ static int run(mddev_t *mddev)
 	/* Ok, everything is just fine now */
 	sysfs_create_group(&mddev->kobj, &raid5_attrs_group);
 
-	if (mddev->bitmap)
-		mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
-
 	mddev->queue->unplug_fn = raid5_unplug_device;
 	mddev->queue->issue_flush_fn = raid5_issue_flush;
 
@@ -1967,9 +1965,7 @@ static int run(mddev_t *mddev)
 abort:
 	if (conf) {
 		print_raid5_conf(conf);
-		if (conf->stripe_hashtbl)
-			free_pages((unsigned long) conf->stripe_hashtbl,
-							HASH_PAGES_ORDER);
+		kfree(conf->stripe_hashtbl);
 		kfree(conf);
 	}
 	mddev->private = NULL;
@@ -1986,7 +1982,7 @@ static int stop(mddev_t *mddev)
 	md_unregister_thread(mddev->thread);
 	mddev->thread = NULL;
 	shrink_stripes(conf);
-	free_pages((unsigned long) conf->stripe_hashtbl, HASH_PAGES_ORDER);
+	kfree(conf->stripe_hashtbl);
 	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
 	sysfs_remove_group(&mddev->kobj, &raid5_attrs_group);
 	kfree(conf);
@@ -2014,12 +2010,12 @@ static void print_sh (struct stripe_head *sh)
 static void printall (raid5_conf_t *conf)
 {
 	struct stripe_head *sh;
+	struct hlist_node *hn;
 	int i;
 
 	spin_lock_irq(&conf->device_lock);
 	for (i = 0; i < NR_HASH; i++) {
-		sh = conf->stripe_hashtbl[i];
-		for (; sh; sh = sh->hash_next) {
+		hlist_for_each_entry(sh, hn, &conf->stripe_hashtbl[i], hash) {
 			if (sh->raid_conf != conf)
 				continue;
 			print_sh(sh);
@@ -2192,17 +2188,12 @@ static void raid5_quiesce(mddev_t *mddev, int state)
 		spin_unlock_irq(&conf->device_lock);
 		break;
 	}
-	if (mddev->thread) {
-		if (mddev->bitmap)
-			mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
-		else
-			mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
-		md_wakeup_thread(mddev->thread);
-	}
 }
-static mdk_personality_t raid5_personality=
+
+static struct mdk_personality raid5_personality =
 {
 	.name		= "raid5",
+	.level		= 5,
 	.owner		= THIS_MODULE,
 	.make_request	= make_request,
 	.run		= run,
@@ -2217,17 +2208,42 @@ static mdk_personality_t raid5_personality=
 	.quiesce	= raid5_quiesce,
 };
 
-static int __init raid5_init (void)
+static struct mdk_personality raid4_personality =
 {
-	return register_md_personality (RAID5, &raid5_personality);
+	.name		= "raid4",
+	.level		= 4,
+	.owner		= THIS_MODULE,
+	.make_request	= make_request,
+	.run		= run,
+	.stop		= stop,
+	.status		= status,
+	.error_handler	= error,
+	.hot_add_disk	= raid5_add_disk,
+	.hot_remove_disk= raid5_remove_disk,
+	.spare_active	= raid5_spare_active,
+	.sync_request	= sync_request,
+	.resize		= raid5_resize,
+	.quiesce	= raid5_quiesce,
+};
+
+static int __init raid5_init(void)
+{
+	register_md_personality(&raid5_personality);
+	register_md_personality(&raid4_personality);
+	return 0;
 }
 
-static void raid5_exit (void)
+static void raid5_exit(void)
 {
-	unregister_md_personality (RAID5);
+	unregister_md_personality(&raid5_personality);
+	unregister_md_personality(&raid4_personality);
 }
 
 module_init(raid5_init);
 module_exit(raid5_exit);
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("md-personality-4"); /* RAID5 */
+MODULE_ALIAS("md-raid5");
+MODULE_ALIAS("md-raid4");
+MODULE_ALIAS("md-level-5");
+MODULE_ALIAS("md-level-4");
diff --git a/drivers/md/raid6main.c b/drivers/md/raid6main.c
index 0000d162d19..8c823d686a6 100644
--- a/drivers/md/raid6main.c
+++ b/drivers/md/raid6main.c
@@ -40,12 +40,10 @@
 #define STRIPE_SHIFT		(PAGE_SHIFT - 9)
 #define STRIPE_SECTORS		(STRIPE_SIZE>>9)
 #define	IO_THRESHOLD		1
-#define HASH_PAGES		1
-#define HASH_PAGES_ORDER	0
-#define NR_HASH			(HASH_PAGES * PAGE_SIZE / sizeof(struct stripe_head *))
+#define NR_HASH			(PAGE_SIZE / sizeof(struct hlist_head))
 #define HASH_MASK		(NR_HASH - 1)
 
-#define stripe_hash(conf, sect)	((conf)->stripe_hashtbl[((sect) >> STRIPE_SHIFT) & HASH_MASK])
+#define stripe_hash(conf, sect)	(&((conf)->stripe_hashtbl[((sect) >> STRIPE_SHIFT) & HASH_MASK]))
 
 /* bio's attached to a stripe+device for I/O are linked together in bi_sector
  * order without overlap.  There may be several bio's per stripe+device, and
@@ -132,29 +130,21 @@ static void release_stripe(struct stripe_head *sh)
 	spin_unlock_irqrestore(&conf->device_lock, flags);
 }
 
-static void remove_hash(struct stripe_head *sh)
+static inline void remove_hash(struct stripe_head *sh)
 {
 	PRINTK("remove_hash(), stripe %llu\n", (unsigned long long)sh->sector);
 
-	if (sh->hash_pprev) {
-		if (sh->hash_next)
-			sh->hash_next->hash_pprev = sh->hash_pprev;
-		*sh->hash_pprev = sh->hash_next;
-		sh->hash_pprev = NULL;
-	}
+	hlist_del_init(&sh->hash);
 }
 
-static __inline__ void insert_hash(raid6_conf_t *conf, struct stripe_head *sh)
+static inline void insert_hash(raid6_conf_t *conf, struct stripe_head *sh)
 {
-	struct stripe_head **shp = &stripe_hash(conf, sh->sector);
+	struct hlist_head *hp = stripe_hash(conf, sh->sector);
 
 	PRINTK("insert_hash(), stripe %llu\n", (unsigned long long)sh->sector);
 
 	CHECK_DEVLOCK();
-	if ((sh->hash_next = *shp) != NULL)
-		(*shp)->hash_pprev = &sh->hash_next;
-	*shp = sh;
-	sh->hash_pprev = shp;
+	hlist_add_head(&sh->hash, hp);
 }
 
 
@@ -186,7 +176,7 @@ static void shrink_buffers(struct stripe_head *sh, int num)
 		if (!p)
 			continue;
 		sh->dev[i].page = NULL;
-		page_cache_release(p);
+		put_page(p);
 	}
 }
 
@@ -247,10 +237,11 @@ static inline void init_stripe(struct stripe_head *sh, sector_t sector, int pd_i
 static struct stripe_head *__find_stripe(raid6_conf_t *conf, sector_t sector)
 {
 	struct stripe_head *sh;
+	struct hlist_node *hn;
 
 	CHECK_DEVLOCK();
 	PRINTK("__find_stripe, sector %llu\n", (unsigned long long)sector);
-	for (sh = stripe_hash(conf, sector); sh; sh = sh->hash_next)
+	hlist_for_each_entry (sh, hn,  stripe_hash(conf, sector), hash)
 		if (sh->sector == sector)
 			return sh;
 	PRINTK("__stripe %llu not in cache\n", (unsigned long long)sector);
@@ -367,8 +358,8 @@ static void shrink_stripes(raid6_conf_t *conf)
 	conf->slab_cache = NULL;
 }
 
-static int raid6_end_read_request (struct bio * bi, unsigned int bytes_done,
-				   int error)
+static int raid6_end_read_request(struct bio * bi, unsigned int bytes_done,
+				  int error)
 {
  	struct stripe_head *sh = bi->bi_private;
 	raid6_conf_t *conf = sh->raid_conf;
@@ -420,9 +411,35 @@ static int raid6_end_read_request (struct bio * bi, unsigned int bytes_done,
 #else
 		set_bit(R5_UPTODATE, &sh->dev[i].flags);
 #endif
+		if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
+			printk(KERN_INFO "raid6: read error corrected!!\n");
+			clear_bit(R5_ReadError, &sh->dev[i].flags);
+			clear_bit(R5_ReWrite, &sh->dev[i].flags);
+		}
+		if (atomic_read(&conf->disks[i].rdev->read_errors))
+			atomic_set(&conf->disks[i].rdev->read_errors, 0);
 	} else {
-		md_error(conf->mddev, conf->disks[i].rdev);
+		int retry = 0;
 		clear_bit(R5_UPTODATE, &sh->dev[i].flags);
+		atomic_inc(&conf->disks[i].rdev->read_errors);
+		if (conf->mddev->degraded)
+			printk(KERN_WARNING "raid6: read error not correctable.\n");
+		else if (test_bit(R5_ReWrite, &sh->dev[i].flags))
+			/* Oh, no!!! */
+			printk(KERN_WARNING "raid6: read error NOT corrected!!\n");
+		else if (atomic_read(&conf->disks[i].rdev->read_errors)
+			 > conf->max_nr_stripes)
+			printk(KERN_WARNING
+			       "raid6: Too many read errors, failing device.\n");
+		else
+			retry = 1;
+		if (retry)
+			set_bit(R5_ReadError, &sh->dev[i].flags);
+		else {
+			clear_bit(R5_ReadError, &sh->dev[i].flags);
+			clear_bit(R5_ReWrite, &sh->dev[i].flags);
+			md_error(conf->mddev, conf->disks[i].rdev);
+		}
 	}
 	rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
 #if 0
@@ -805,7 +822,7 @@ static void compute_parity(struct stripe_head *sh, int method)
 }
 
 /* Compute one missing block */
-static void compute_block_1(struct stripe_head *sh, int dd_idx)
+static void compute_block_1(struct stripe_head *sh, int dd_idx, int nozero)
 {
 	raid6_conf_t *conf = sh->raid_conf;
 	int i, count, disks = conf->raid_disks;
@@ -821,7 +838,7 @@ static void compute_block_1(struct stripe_head *sh, int dd_idx)
 		compute_parity(sh, UPDATE_PARITY);
 	} else {
 		ptr[0] = page_address(sh->dev[dd_idx].page);
-		memset(ptr[0], 0, STRIPE_SIZE);
+		if (!nozero) memset(ptr[0], 0, STRIPE_SIZE);
 		count = 1;
 		for (i = disks ; i--; ) {
 			if (i == dd_idx || i == qd_idx)
@@ -838,7 +855,8 @@ static void compute_block_1(struct stripe_head *sh, int dd_idx)
 		}
 		if (count != 1)
 			xor_block(count, STRIPE_SIZE, ptr);
-		set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
+		if (!nozero) set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
+		else clear_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
 	}
 }
 
@@ -871,7 +889,7 @@ static void compute_block_2(struct stripe_head *sh, int dd_idx1, int dd_idx2)
 			return;
 		} else {
 			/* We're missing D+Q; recompute D from P */
-			compute_block_1(sh, (dd_idx1 == qd_idx) ? dd_idx2 : dd_idx1);
+			compute_block_1(sh, (dd_idx1 == qd_idx) ? dd_idx2 : dd_idx1, 0);
 			compute_parity(sh, UPDATE_PARITY); /* Is this necessary? */
 			return;
 		}
@@ -982,6 +1000,12 @@ static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, in
 }
 
 
+static int page_is_zero(struct page *p)
+{
+	char *a = page_address(p);
+	return ((*(u32*)a) == 0 &&
+		memcmp(a, a+4, STRIPE_SIZE-4)==0);
+}
 /*
  * handle_stripe - do things to a stripe.
  *
@@ -1000,7 +1024,7 @@ static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, in
  *
  */
 
-static void handle_stripe(struct stripe_head *sh)
+static void handle_stripe(struct stripe_head *sh, struct page *tmp_page)
 {
 	raid6_conf_t *conf = sh->raid_conf;
 	int disks = conf->raid_disks;
@@ -1027,11 +1051,11 @@ static void handle_stripe(struct stripe_head *sh)
 	syncing = test_bit(STRIPE_SYNCING, &sh->state);
 	/* Now to look around and see what can be done */
 
+	rcu_read_lock();
 	for (i=disks; i--; ) {
 		mdk_rdev_t *rdev;
 		dev = &sh->dev[i];
 		clear_bit(R5_Insync, &dev->flags);
-		clear_bit(R5_Syncio, &dev->flags);
 
 		PRINTK("check %d: state 0x%lx read %p write %p written %p\n",
 			i, dev->flags, dev->toread, dev->towrite, dev->written);
@@ -1070,14 +1094,21 @@ static void handle_stripe(struct stripe_head *sh)
 				non_overwrite++;
 		}
 		if (dev->written) written++;
-		rdev = conf->disks[i].rdev; /* FIXME, should I be looking rdev */
+		rdev = rcu_dereference(conf->disks[i].rdev);
 		if (!rdev || !test_bit(In_sync, &rdev->flags)) {
+			/* The ReadError flag will just be confusing now */
+			clear_bit(R5_ReadError, &dev->flags);
+			clear_bit(R5_ReWrite, &dev->flags);
+		}
+		if (!rdev || !test_bit(In_sync, &rdev->flags)
+		    || test_bit(R5_ReadError, &dev->flags)) {
 			if ( failed < 2 )
 				failed_num[failed] = i;
 			failed++;
 		} else
 			set_bit(R5_Insync, &dev->flags);
 	}
+	rcu_read_unlock();
 	PRINTK("locked=%d uptodate=%d to_read=%d"
 	       " to_write=%d failed=%d failed_num=%d,%d\n",
 	       locked, uptodate, to_read, to_write, failed,
@@ -1088,6 +1119,17 @@ static void handle_stripe(struct stripe_head *sh)
 	if (failed > 2 && to_read+to_write+written) {
 		for (i=disks; i--; ) {
 			int bitmap_end = 0;
+
+			if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
+				mdk_rdev_t *rdev;
+				rcu_read_lock();
+				rdev = rcu_dereference(conf->disks[i].rdev);
+				if (rdev && test_bit(In_sync, &rdev->flags))
+					/* multiple read failures in one stripe */
+					md_error(conf->mddev, rdev);
+				rcu_read_unlock();
+			}
+
 			spin_lock_irq(&conf->device_lock);
 			/* fail all writes first */
 			bi = sh->dev[i].towrite;
@@ -1123,7 +1165,8 @@ static void handle_stripe(struct stripe_head *sh)
 			}
 
 			/* fail any reads if this device is non-operational */
-			if (!test_bit(R5_Insync, &sh->dev[i].flags)) {
+			if (!test_bit(R5_Insync, &sh->dev[i].flags) ||
+			    test_bit(R5_ReadError, &sh->dev[i].flags)) {
 				bi = sh->dev[i].toread;
 				sh->dev[i].toread = NULL;
 				if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
@@ -1228,7 +1271,7 @@ static void handle_stripe(struct stripe_head *sh)
 				if (uptodate == disks-1) {
 					PRINTK("Computing stripe %llu block %d\n",
 					       (unsigned long long)sh->sector, i);
-					compute_block_1(sh, i);
+					compute_block_1(sh, i, 0);
 					uptodate++;
 				} else if ( uptodate == disks-2 && failed >= 2 ) {
 					/* Computing 2-failure is *very* expensive; only do it if failed >= 2 */
@@ -1259,9 +1302,6 @@ static void handle_stripe(struct stripe_head *sh)
 					locked++;
 					PRINTK("Reading block %d (sync=%d)\n",
 						i, syncing);
-					if (syncing)
-						md_sync_acct(conf->disks[i].rdev->bdev,
-							     STRIPE_SECTORS);
 				}
 			}
 		}
@@ -1323,7 +1363,7 @@ static void handle_stripe(struct stripe_head *sh)
 				/* We have failed blocks and need to compute them */
 				switch ( failed ) {
 				case 0:	BUG();
-				case 1: compute_block_1(sh, failed_num[0]); break;
+				case 1: compute_block_1(sh, failed_num[0], 0); break;
 				case 2: compute_block_2(sh, failed_num[0], failed_num[1]); break;
 				default: BUG();	/* This request should have been failed? */
 				}
@@ -1338,12 +1378,10 @@ static void handle_stripe(struct stripe_head *sh)
 					       (unsigned long long)sh->sector, i);
 					locked++;
 					set_bit(R5_Wantwrite, &sh->dev[i].flags);
-#if 0 /**** FIX: I don't understand the logic here... ****/
-					if (!test_bit(R5_Insync, &sh->dev[i].flags)
-					    || ((i==pd_idx || i==qd_idx) && failed == 0)) /* FIX? */
-						set_bit(STRIPE_INSYNC, &sh->state);
-#endif
 				}
+			/* after a RECONSTRUCT_WRITE, the stripe MUST be in-sync */
+			set_bit(STRIPE_INSYNC, &sh->state);
+
 			if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
 				atomic_dec(&conf->preread_active_stripes);
 				if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD)
@@ -1356,84 +1394,119 @@ static void handle_stripe(struct stripe_head *sh)
 	 * Any reads will already have been scheduled, so we just see if enough data
 	 * is available
 	 */
-	if (syncing && locked == 0 &&
-	    !test_bit(STRIPE_INSYNC, &sh->state) && failed <= 2) {
-		set_bit(STRIPE_HANDLE, &sh->state);
-#if 0 /* RAID-6: Don't support CHECK PARITY yet */
-		if (failed == 0) {
-			char *pagea;
-			if (uptodate != disks)
-				BUG();
-			compute_parity(sh, CHECK_PARITY);
-			uptodate--;
-			pagea = page_address(sh->dev[pd_idx].page);
-			if ((*(u32*)pagea) == 0 &&
-			    !memcmp(pagea, pagea+4, STRIPE_SIZE-4)) {
-				/* parity is correct (on disc, not in buffer any more) */
-				set_bit(STRIPE_INSYNC, &sh->state);
-			}
-		}
-#endif
-		if (!test_bit(STRIPE_INSYNC, &sh->state)) {
-			int failed_needupdate[2];
-			struct r5dev *adev, *bdev;
-
-			if ( failed < 1 )
-				failed_num[0] = pd_idx;
-			if ( failed < 2 )
-				failed_num[1] = (failed_num[0] == qd_idx) ? pd_idx : qd_idx;
+	if (syncing && locked == 0 && !test_bit(STRIPE_INSYNC, &sh->state)) {
+		int update_p = 0, update_q = 0;
+		struct r5dev *dev;
 
-			failed_needupdate[0] = !test_bit(R5_UPTODATE, &sh->dev[failed_num[0]].flags);
-			failed_needupdate[1] = !test_bit(R5_UPTODATE, &sh->dev[failed_num[1]].flags);
+		set_bit(STRIPE_HANDLE, &sh->state);
 
-			PRINTK("sync: failed=%d num=%d,%d fnu=%u%u\n",
-			       failed, failed_num[0], failed_num[1], failed_needupdate[0], failed_needupdate[1]);
+		BUG_ON(failed>2);
+		BUG_ON(uptodate < disks);
+		/* Want to check and possibly repair P and Q.
+		 * However there could be one 'failed' device, in which
+		 * case we can only check one of them, possibly using the
+		 * other to generate missing data
+		 */
 
-#if 0  /* RAID-6: This code seems to require that CHECK_PARITY destroys the uptodateness of the parity */
-			/* should be able to compute the missing block(s) and write to spare */
-			if ( failed_needupdate[0] ^ failed_needupdate[1] ) {
-				if (uptodate+1 != disks)
-					BUG();
-				compute_block_1(sh, failed_needupdate[0] ? failed_num[0] : failed_num[1]);
-				uptodate++;
-			} else if ( failed_needupdate[0] & failed_needupdate[1] ) {
-				if (uptodate+2 != disks)
-					BUG();
-				compute_block_2(sh, failed_num[0], failed_num[1]);
-				uptodate += 2;
+		/* If !tmp_page, we cannot do the calculations,
+		 * but as we have set STRIPE_HANDLE, we will soon be called
+		 * by stripe_handle with a tmp_page - just wait until then.
+		 */
+		if (tmp_page) {
+			if (failed == q_failed) {
+				/* The only possible failed device holds 'Q', so it makes
+				 * sense to check P (If anything else were failed, we would
+				 * have used P to recreate it).
+				 */
+				compute_block_1(sh, pd_idx, 1);
+				if (!page_is_zero(sh->dev[pd_idx].page)) {
+					compute_block_1(sh,pd_idx,0);
+					update_p = 1;
+				}
+			}
+			if (!q_failed && failed < 2) {
+				/* q is not failed, and we didn't use it to generate
+				 * anything, so it makes sense to check it
+				 */
+				memcpy(page_address(tmp_page),
+				       page_address(sh->dev[qd_idx].page),
+				       STRIPE_SIZE);
+				compute_parity(sh, UPDATE_PARITY);
+				if (memcmp(page_address(tmp_page),
+					   page_address(sh->dev[qd_idx].page),
+					   STRIPE_SIZE)!= 0) {
+					clear_bit(STRIPE_INSYNC, &sh->state);
+					update_q = 1;
+				}
+			}
+			if (update_p || update_q) {
+				conf->mddev->resync_mismatches += STRIPE_SECTORS;
+				if (test_bit(MD_RECOVERY_CHECK, &conf->mddev->recovery))
+					/* don't try to repair!! */
+					update_p = update_q = 0;
 			}
-#else
-			compute_block_2(sh, failed_num[0], failed_num[1]);
-			uptodate += failed_needupdate[0] + failed_needupdate[1];
-#endif
 
-			if (uptodate != disks)
-				BUG();
+			/* now write out any block on a failed drive,
+			 * or P or Q if they need it
+			 */
 
-			PRINTK("Marking for sync stripe %llu blocks %d,%d\n",
-			       (unsigned long long)sh->sector, failed_num[0], failed_num[1]);
+			if (failed == 2) {
+				dev = &sh->dev[failed_num[1]];
+				locked++;
+				set_bit(R5_LOCKED, &dev->flags);
+				set_bit(R5_Wantwrite, &dev->flags);
+			}
+			if (failed >= 1) {
+				dev = &sh->dev[failed_num[0]];
+				locked++;
+				set_bit(R5_LOCKED, &dev->flags);
+				set_bit(R5_Wantwrite, &dev->flags);
+			}
 
-			/**** FIX: Should we really do both of these unconditionally? ****/
-			adev = &sh->dev[failed_num[0]];
-			locked += !test_bit(R5_LOCKED, &adev->flags);
-			set_bit(R5_LOCKED, &adev->flags);
-			set_bit(R5_Wantwrite, &adev->flags);
-			bdev = &sh->dev[failed_num[1]];
-			locked += !test_bit(R5_LOCKED, &bdev->flags);
-			set_bit(R5_LOCKED, &bdev->flags);
+			if (update_p) {
+				dev = &sh->dev[pd_idx];
+				locked ++;
+				set_bit(R5_LOCKED, &dev->flags);
+				set_bit(R5_Wantwrite, &dev->flags);
+			}
+			if (update_q) {
+				dev = &sh->dev[qd_idx];
+				locked++;
+				set_bit(R5_LOCKED, &dev->flags);
+				set_bit(R5_Wantwrite, &dev->flags);
+			}
 			clear_bit(STRIPE_DEGRADED, &sh->state);
-			set_bit(R5_Wantwrite, &bdev->flags);
 
 			set_bit(STRIPE_INSYNC, &sh->state);
-			set_bit(R5_Syncio, &adev->flags);
-			set_bit(R5_Syncio, &bdev->flags);
 		}
 	}
+
 	if (syncing && locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
 		md_done_sync(conf->mddev, STRIPE_SECTORS,1);
 		clear_bit(STRIPE_SYNCING, &sh->state);
 	}
 
+	/* If the failed drives are just a ReadError, then we might need
+	 * to progress the repair/check process
+	 */
+	if (failed <= 2 && ! conf->mddev->ro)
+		for (i=0; i<failed;i++) {
+			dev = &sh->dev[failed_num[i]];
+			if (test_bit(R5_ReadError, &dev->flags)
+			    && !test_bit(R5_LOCKED, &dev->flags)
+			    && test_bit(R5_UPTODATE, &dev->flags)
+				) {
+				if (!test_bit(R5_ReWrite, &dev->flags)) {
+					set_bit(R5_Wantwrite, &dev->flags);
+					set_bit(R5_ReWrite, &dev->flags);
+					set_bit(R5_LOCKED, &dev->flags);
+				} else {
+					/* let's read it back */
+					set_bit(R5_Wantread, &dev->flags);
+					set_bit(R5_LOCKED, &dev->flags);
+				}
+			}
+		}
 	spin_unlock(&sh->lock);
 
 	while ((bi=return_bi)) {
@@ -1472,7 +1545,7 @@ static void handle_stripe(struct stripe_head *sh)
 		rcu_read_unlock();
 
 		if (rdev) {
-			if (test_bit(R5_Syncio, &sh->dev[i].flags))
+			if (syncing)
 				md_sync_acct(rdev->bdev, STRIPE_SECTORS);
 
 			bi->bi_bdev = rdev->bdev;
@@ -1489,6 +1562,9 @@ static void handle_stripe(struct stripe_head *sh)
 			bi->bi_io_vec[0].bv_offset = 0;
 			bi->bi_size = STRIPE_SIZE;
 			bi->bi_next = NULL;
+			if (rw == WRITE &&
+			    test_bit(R5_ReWrite, &sh->dev[i].flags))
+				atomic_add(STRIPE_SECTORS, &rdev->corrected_errors);
 			generic_make_request(bi);
 		} else {
 			if (rw == 1)
@@ -1664,7 +1740,7 @@ static int make_request (request_queue_t *q, struct bio * bi)
 			}
 			finish_wait(&conf->wait_for_overlap, &w);
 			raid6_plug_device(conf);
-			handle_stripe(sh);
+			handle_stripe(sh, NULL);
 			release_stripe(sh);
 		} else {
 			/* cannot get stripe for read-ahead, just give-up */
@@ -1728,6 +1804,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 		return rv;
 	}
 	if (!bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 1) &&
+	    !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery) &&
 	    !conf->fullsync && sync_blocks >= STRIPE_SECTORS) {
 		/* we can skip this block, and probably more */
 		sync_blocks /= STRIPE_SECTORS;
@@ -1765,7 +1842,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 	clear_bit(STRIPE_INSYNC, &sh->state);
 	spin_unlock(&sh->lock);
 
-	handle_stripe(sh);
+	handle_stripe(sh, NULL);
 	release_stripe(sh);
 
 	return STRIPE_SECTORS;
@@ -1821,7 +1898,7 @@ static void raid6d (mddev_t *mddev)
 		spin_unlock_irq(&conf->device_lock);
 
 		handled++;
-		handle_stripe(sh);
+		handle_stripe(sh, conf->spare_page);
 		release_stripe(sh);
 
 		spin_lock_irq(&conf->device_lock);
@@ -1848,17 +1925,19 @@ static int run(mddev_t *mddev)
 		return -EIO;
 	}
 
-	mddev->private = kmalloc (sizeof (raid6_conf_t)
-				  + mddev->raid_disks * sizeof(struct disk_info),
-				  GFP_KERNEL);
+	mddev->private = kzalloc(sizeof (raid6_conf_t)
+				 + mddev->raid_disks * sizeof(struct disk_info),
+				 GFP_KERNEL);
 	if ((conf = mddev->private) == NULL)
 		goto abort;
-	memset (conf, 0, sizeof (*conf) + mddev->raid_disks * sizeof(struct disk_info) );
 	conf->mddev = mddev;
 
-	if ((conf->stripe_hashtbl = (struct stripe_head **) __get_free_pages(GFP_ATOMIC, HASH_PAGES_ORDER)) == NULL)
+	if ((conf->stripe_hashtbl = kzalloc(PAGE_SIZE, GFP_KERNEL)) == NULL)
+		goto abort;
+
+	conf->spare_page = alloc_page(GFP_KERNEL);
+	if (!conf->spare_page)
 		goto abort;
-	memset(conf->stripe_hashtbl, 0, HASH_PAGES * PAGE_SIZE);
 
 	spin_lock_init(&conf->device_lock);
 	init_waitqueue_head(&conf->wait_for_stripe);
@@ -1929,13 +2008,18 @@ static int run(mddev_t *mddev)
 		goto abort;
 	}
 
-#if 0				/* FIX: For now */
 	if (mddev->degraded > 0 &&
 	    mddev->recovery_cp != MaxSector) {
-		printk(KERN_ERR "raid6: cannot start dirty degraded array for %s\n", mdname(mddev));
-		goto abort;
+		if (mddev->ok_start_degraded)
+			printk(KERN_WARNING "raid6: starting dirty degraded array:%s"
+			       "- data corruption possible.\n",
+			       mdname(mddev));
+		else {
+			printk(KERN_ERR "raid6: cannot start dirty degraded array"
+			       " for %s\n", mdname(mddev));
+			goto abort;
+		}
 	}
-#endif
 
 	{
 		mddev->thread = md_register_thread(raid6d, mddev, "%s_raid6");
@@ -1977,7 +2061,7 @@ static int run(mddev_t *mddev)
 	 */
 	{
 		int stripe = (mddev->raid_disks-2) * mddev->chunk_size
-			/ PAGE_CACHE_SIZE;
+			/ PAGE_SIZE;
 		if (mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
 			mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
 	}
@@ -1985,18 +2069,14 @@ static int run(mddev_t *mddev)
 	/* Ok, everything is just fine now */
 	mddev->array_size =  mddev->size * (mddev->raid_disks - 2);
 
-	if (mddev->bitmap)
-		mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
-
 	mddev->queue->unplug_fn = raid6_unplug_device;
 	mddev->queue->issue_flush_fn = raid6_issue_flush;
 	return 0;
 abort:
 	if (conf) {
 		print_raid6_conf(conf);
-		if (conf->stripe_hashtbl)
-			free_pages((unsigned long) conf->stripe_hashtbl,
-							HASH_PAGES_ORDER);
+		safe_put_page(conf->spare_page);
+		kfree(conf->stripe_hashtbl);
 		kfree(conf);
 	}
 	mddev->private = NULL;
@@ -2013,7 +2093,7 @@ static int stop (mddev_t *mddev)
 	md_unregister_thread(mddev->thread);
 	mddev->thread = NULL;
 	shrink_stripes(conf);
-	free_pages((unsigned long) conf->stripe_hashtbl, HASH_PAGES_ORDER);
+	kfree(conf->stripe_hashtbl);
 	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
 	kfree(conf);
 	mddev->private = NULL;
@@ -2040,12 +2120,13 @@ static void print_sh (struct seq_file *seq, struct stripe_head *sh)
 static void printall (struct seq_file *seq, raid6_conf_t *conf)
 {
 	struct stripe_head *sh;
+	struct hlist_node *hn;
 	int i;
 
 	spin_lock_irq(&conf->device_lock);
 	for (i = 0; i < NR_HASH; i++) {
 		sh = conf->stripe_hashtbl[i];
-		for (; sh; sh = sh->hash_next) {
+		hlist_for_each_entry(sh, hn, &conf->stripe_hashtbl[i], hash) {
 			if (sh->raid_conf != conf)
 				continue;
 			print_sh(seq, sh);
@@ -2223,17 +2304,12 @@ static void raid6_quiesce(mddev_t *mddev, int state)
 		spin_unlock_irq(&conf->device_lock);
 		break;
 	}
-	if (mddev->thread) {
-		if (mddev->bitmap)
-			mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
-		else
-			mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
-		md_wakeup_thread(mddev->thread);
-	}
 }
-static mdk_personality_t raid6_personality=
+
+static struct mdk_personality raid6_personality =
 {
 	.name		= "raid6",
+	.level		= 6,
 	.owner		= THIS_MODULE,
 	.make_request	= make_request,
 	.run		= run,
@@ -2248,7 +2324,7 @@ static mdk_personality_t raid6_personality=
 	.quiesce	= raid6_quiesce,
 };
 
-static int __init raid6_init (void)
+static int __init raid6_init(void)
 {
 	int e;
 
@@ -2256,15 +2332,17 @@ static int __init raid6_init (void)
 	if ( e )
 		return e;
 
-	return register_md_personality (RAID6, &raid6_personality);
+	return register_md_personality(&raid6_personality);
 }
 
 static void raid6_exit (void)
 {
-	unregister_md_personality (RAID6);
+	unregister_md_personality(&raid6_personality);
 }
 
 module_init(raid6_init);
 module_exit(raid6_exit);
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("md-personality-8"); /* RAID6 */
+MODULE_ALIAS("md-raid6");
+MODULE_ALIAS("md-level-6");