[PATCH] md: RAID10 module

This patch adds a 'raid10' module which provides features similar to both
raid0 and raid1 in the one array.  Various combinations of layout are
supported.

This code is still "experimental", but appears to work.
Signed-off-by: Neil Brown <neilb@cse.unsw.edu.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

drivers/md/Kconfig

@@ -85,6 +85,24 @@ config MD_RAID1
 
 	  If unsure, say Y.
 
+config MD_RAID10
+	tristate "RAID-10 (mirrored striping) mode (EXPERIMENTAL)"
+	depends on BLK_DEV_MD && EXPERIMENTAL
+	---help---
+	  RAID-10 provides a combination of striping (RAID-0) and
+	  mirroring (RAID-1) with easier configuration and more flexable
+	  layout.
+	  Unlike RAID-0, but like RAID-1, RAID-10 requires all devices to
+	  be the same size (or atleast, only as much as the smallest device
+	  will be used).
+	  RAID-10 provides a variety of layouts that provide different levels
+	  of redundancy and performance.
+
+	  RAID-10 requires mdadm-1.7.0 or later, available at:
+
+	  ftp://ftp.kernel.org/pub/linux/utils/raid/mdadm/
+
+
 config MD_RAID5
 	tristate "RAID-4/RAID-5 mode"
 	depends on BLK_DEV_MD

drivers/md/Makefile

@@ -20,6 +20,7 @@ hostprogs-y	:= mktables
 obj-$(CONFIG_MD_LINEAR)		+= linear.o
 obj-$(CONFIG_MD_RAID0)		+= raid0.o
 obj-$(CONFIG_MD_RAID1)		+= raid1.o
+obj-$(CONFIG_MD_RAID10)		+= raid10.o
 obj-$(CONFIG_MD_RAID5)		+= raid5.o xor.o
 obj-$(CONFIG_MD_RAID6)		+= raid6.o xor.o
 obj-$(CONFIG_MD_MULTIPATH)	+= multipath.o

include/linux/raid/md_k.h

@@ -24,7 +24,8 @@
 #define HSM               6UL
 #define MULTIPATH         7UL
 #define RAID6		  8UL
-#define MAX_PERSONALITY   9UL
+#define	RAID10		  9UL
+#define MAX_PERSONALITY   10UL
 
 #define	LEVEL_MULTIPATH		(-4)
 #define	LEVEL_LINEAR		(-1)
@@ -43,6 +44,7 @@ static inline int pers_to_level (int pers)
 		case RAID1:		return 1;
 		case RAID5:		return 5;
 		case RAID6:		return 6;
+		case RAID10:		return 10;
 	}
 	BUG();
 	return MD_RESERVED;
@@ -60,6 +62,7 @@ static inline int level_to_pers (int level)
 		case 4:
 		case 5: return RAID5;
 		case 6: return RAID6;
+		case 10: return RAID10;
 	}
 	return MD_RESERVED;
 }

include/linux/raid/raid10.h

+#ifndef _RAID10_H
+#define _RAID10_H
+
+#include <linux/raid/md.h>
+
+typedef struct mirror_info mirror_info_t;
+
+struct mirror_info {
+	mdk_rdev_t	*rdev;
+	sector_t	head_position;
+};
+
+typedef struct r10bio_s r10bio_t;
+
+struct r10_private_data_s {
+	mddev_t			*mddev;
+	mirror_info_t		*mirrors;
+	int			raid_disks;
+	int			working_disks;
+	spinlock_t		device_lock;
+
+	/* geometry */
+	int			near_copies;  /* number of copies layed out raid0 style */
+	int 			far_copies;   /* number of copies layed out
+					       * at large strides across drives
+					       */
+	int			copies;	      /* near_copies * far_copies.
+					       * must be <= raid_disks
+					       */
+	sector_t		stride;	      /* distance between far copies.
+					       * This is size / far_copies
+					       */
+
+	int chunk_shift; /* shift from chunks to sectors */
+	sector_t chunk_mask;
+
+	struct list_head	retry_list;
+	/* for use when syncing mirrors: */
+
+	spinlock_t		resync_lock;
+	int nr_pending;
+	int barrier;
+	sector_t		next_resync;
+
+	wait_queue_head_t	wait_idle;
+	wait_queue_head_t	wait_resume;
+
+	mempool_t *r10bio_pool;
+	mempool_t *r10buf_pool;
+};
+
+typedef struct r10_private_data_s conf_t;
+
+/*
+ * this is the only point in the RAID code where we violate
+ * C type safety. mddev->private is an 'opaque' pointer.
+ */
+#define mddev_to_conf(mddev) ((conf_t *) mddev->private)
+
+/*
+ * this is our 'private' RAID10 bio.
+ *
+ * it contains information about what kind of IO operations were started
+ * for this RAID10 operation, and about their status:
+ */
+
+struct r10bio_s {
+	atomic_t		remaining; /* 'have we finished' count,
+					    * used from IRQ handlers
+					    */
+	sector_t		sector;	/* virtual sector number */
+	int			sectors;
+	unsigned long		state;
+	mddev_t			*mddev;
+	/*
+	 * original bio going to /dev/mdx
+	 */
+	struct bio		*master_bio;
+	/*
+	 * if the IO is in READ direction, then this is where we read
+	 */
+	int			read_slot;
+
+	struct list_head	retry_list;
+	/*
+	 * if the IO is in WRITE direction, then multiple bios are used,
+	 * one for each copy.
+	 * When resyncing we also use one for each copy.
+	 * When reconstructing, we use 2 bios, one for read, one for write.
+	 * We choose the number when they are allocated.
+	 */
+	struct {
+		struct bio		*bio;
+		sector_t addr;
+		int devnum;
+	} devs[0];
+};
+
+/* bits for r10bio.state */
+#define	R10BIO_Uptodate	0
+#define	R10BIO_IsSync	1
+#define	R10BIO_IsRecover 2
+#endif