[net-next-2.6.git] / include / linux / raid / md_k.h

/*
   md_k.h : kernel internal structure of the Linux MD driver
          Copyright (C) 1996-98 Ingo Molnar, Gadi Oxman
	  
   This program is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2, or (at your option)
   any later version.
   
   You should have received a copy of the GNU General Public License
   (for example /usr/src/linux/COPYING); if not, write to the Free
   Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  
*/

#ifndef _MD_K_H
#define _MD_K_H

/* and dm-bio-list.h is not under include/linux because.... ??? */
#include "../../../drivers/md/dm-bio-list.h"

#define	LEVEL_MULTIPATH		(-4)
#define	LEVEL_LINEAR		(-1)
#define	LEVEL_FAULTY		(-5)

/* we need a value for 'no level specified' and 0
 * means 'raid0', so we need something else.  This is
 * for internal use only
 */
#define	LEVEL_NONE		(-1000000)

#define MaxSector (~(sector_t)0)
#define MD_THREAD_NAME_MAX 14

typedef struct mddev_s mddev_t;
typedef struct mdk_rdev_s mdk_rdev_t;

#define MAX_MD_DEVS  256	/* Max number of md dev */

/*
 * options passed in raidrun:
 */

#define MAX_CHUNK_SIZE (4096*1024)

/*
 * MD's 'extended' device
 */
struct mdk_rdev_s
{
	struct list_head same_set;	/* RAID devices within the same set */

	sector_t size;			/* Device size (in blocks) */
	mddev_t *mddev;			/* RAID array if running */
	unsigned long last_events;	/* IO event timestamp */

	struct block_device *bdev;	/* block device handle */

	struct page	*sb_page;
	int		sb_loaded;
	sector_t	data_offset;	/* start of data in array */
	sector_t	sb_offset;
	int		sb_size;	/* bytes in the superblock */
	int		preferred_minor;	/* autorun support */

	struct kobject	kobj;

	/* A device can be in one of three states based on two flags:
	 * Not working:   faulty==1 in_sync==0
	 * Fully working: faulty==0 in_sync==1
	 * Working, but not
	 * in sync with array
	 *                faulty==0 in_sync==0
	 *
	 * It can never have faulty==1, in_sync==1
	 * This reduces the burden of testing multiple flags in many cases
	 */

	unsigned long	flags;
#define	Faulty		1		/* device is known to have a fault */
#define	In_sync		2		/* device is in_sync with rest of array */
#define	WriteMostly	4		/* Avoid reading if at all possible */
#define	BarriersNotsupp	5		/* BIO_RW_BARRIER is not supported */

	int desc_nr;			/* descriptor index in the superblock */
	int raid_disk;			/* role of device in array */
	int saved_raid_disk;		/* role that device used to have in the
					 * array and could again if we did a partial
					 * resync from the bitmap
					 */

	atomic_t	nr_pending;	/* number of pending requests.
					 * only maintained for arrays that
					 * support hot removal
					 */
	atomic_t	read_errors;	/* number of consecutive read errors that
					 * we have tried to ignore.
					 */
};

struct mddev_s
{
	void				*private;
	struct mdk_personality		*pers;
	dev_t				unit;
	int				md_minor;
	struct list_head 		disks;
	int				sb_dirty;
	int				ro;

	struct gendisk			*gendisk;

	struct kobject			kobj;

	/* Superblock information */
	int				major_version,
					minor_version,
					patch_version;
	int				persistent;
	int				chunk_size;
	time_t				ctime, utime;
	int				level, layout;
	int				raid_disks;
	int				max_disks;
	sector_t			size; /* used size of component devices */
	sector_t			array_size; /* exported array size */
	__u64				events;

	char				uuid[16];

	struct mdk_thread_s		*thread;	/* management thread */
	struct mdk_thread_s		*sync_thread;	/* doing resync or reconstruct */
	sector_t			curr_resync;	/* blocks scheduled */
	unsigned long			resync_mark;	/* a recent timestamp */
	sector_t			resync_mark_cnt;/* blocks written at resync_mark */

	sector_t			resync_max_sectors; /* may be set by personality */

	sector_t			resync_mismatches; /* count of sectors where
							    * parity/replica mismatch found
							    */
	int				ok_start_degraded;
	/* recovery/resync flags 
	 * NEEDED:   we might need to start a resync/recover
	 * RUNNING:  a thread is running, or about to be started
	 * SYNC:     actually doing a resync, not a recovery
	 * ERR:      and IO error was detected - abort the resync/recovery
	 * INTR:     someone requested a (clean) early abort.
	 * DONE:     thread is done and is waiting to be reaped
	 * REQUEST:  user-space has requested a sync (used with SYNC)
	 * CHECK:    user-space request for for check-only, no repair
	 */
#define	MD_RECOVERY_RUNNING	0
#define	MD_RECOVERY_SYNC	1
#define	MD_RECOVERY_ERR		2
#define	MD_RECOVERY_INTR	3
#define	MD_RECOVERY_DONE	4
#define	MD_RECOVERY_NEEDED	5
#define	MD_RECOVERY_REQUESTED	6
#define	MD_RECOVERY_CHECK	7
	unsigned long			recovery;

	int				in_sync;	/* know to not need resync */
	struct semaphore		reconfig_sem;
	atomic_t			active;

	int				changed;	/* true if we might need to reread partition info */
	int				degraded;	/* whether md should consider
							 * adding a spare
							 */
	int				barriers_work;	/* initialised to true, cleared as soon
							 * as a barrier request to slave
							 * fails.  Only supported
							 */
	struct bio			*biolist; 	/* bios that need to be retried
							 * because BIO_RW_BARRIER is not supported
							 */

	atomic_t			recovery_active; /* blocks scheduled, but not written */
	wait_queue_head_t		recovery_wait;
	sector_t			recovery_cp;

	spinlock_t			write_lock;
	wait_queue_head_t		sb_wait;	/* for waiting on superblock updates */
	atomic_t			pending_writes;	/* number of active superblock writes */

	unsigned int			safemode;	/* if set, update "clean" superblock
							 * when no writes pending.
							 */ 
	unsigned int			safemode_delay;
	struct timer_list		safemode_timer;
	atomic_t			writes_pending; 
	request_queue_t			*queue;	/* for plugging ... */

	atomic_t                        write_behind; /* outstanding async IO */
	unsigned int                    max_write_behind; /* 0 = sync */

	struct bitmap                   *bitmap; /* the bitmap for the device */
	struct file			*bitmap_file; /* the bitmap file */
	long				bitmap_offset; /* offset from superblock of
							* start of bitmap. May be
							* negative, but not '0'
							*/
	long				default_bitmap_offset; /* this is the offset to use when
								* hot-adding a bitmap.  It should
								* eventually be settable by sysfs.
								*/

	struct list_head		all_mddevs;
};


static inline void rdev_dec_pending(mdk_rdev_t *rdev, mddev_t *mddev)
{
	int faulty = test_bit(Faulty, &rdev->flags);
	if (atomic_dec_and_test(&rdev->nr_pending) && faulty)
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
}

static inline void md_sync_acct(struct block_device *bdev, unsigned long nr_sectors)
{
        atomic_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io);
}

struct mdk_personality
{
	char *name;
	int level;
	struct list_head list;
	struct module *owner;
	int (*make_request)(request_queue_t *q, struct bio *bio);
	int (*run)(mddev_t *mddev);
	int (*stop)(mddev_t *mddev);
	void (*status)(struct seq_file *seq, mddev_t *mddev);
	/* error_handler must set ->faulty and clear ->in_sync
	 * if appropriate, and should abort recovery if needed 
	 */
	void (*error_handler)(mddev_t *mddev, mdk_rdev_t *rdev);
	int (*hot_add_disk) (mddev_t *mddev, mdk_rdev_t *rdev);
	int (*hot_remove_disk) (mddev_t *mddev, int number);
	int (*spare_active) (mddev_t *mddev);
	sector_t (*sync_request)(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster);
	int (*resize) (mddev_t *mddev, sector_t sectors);
	int (*reshape) (mddev_t *mddev, int raid_disks);
	int (*reconfig) (mddev_t *mddev, int layout, int chunk_size);
	/* quiesce moves between quiescence states
	 * 0 - fully active
	 * 1 - no new requests allowed
	 * others - reserved
	 */
	void (*quiesce) (mddev_t *mddev, int state);
};


struct md_sysfs_entry {
	struct attribute attr;
	ssize_t (*show)(mddev_t *, char *);
	ssize_t (*store)(mddev_t *, const char *, size_t);
};


static inline char * mdname (mddev_t * mddev)
{
	return mddev->gendisk ? mddev->gendisk->disk_name : "mdX";
}

extern mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr);

/*
 * iterates through some rdev ringlist. It's safe to remove the
 * current 'rdev'. Dont touch 'tmp' though.
 */
#define ITERATE_RDEV_GENERIC(head,rdev,tmp)				\
									\
	for ((tmp) = (head).next;					\
		(rdev) = (list_entry((tmp), mdk_rdev_t, same_set)),	\
			(tmp) = (tmp)->next, (tmp)->prev != &(head)	\
		; )
/*
 * iterates through the 'same array disks' ringlist
 */
#define ITERATE_RDEV(mddev,rdev,tmp)					\
	ITERATE_RDEV_GENERIC((mddev)->disks,rdev,tmp)

/*
 * Iterates through 'pending RAID disks'
 */
#define ITERATE_RDEV_PENDING(rdev,tmp)					\
	ITERATE_RDEV_GENERIC(pending_raid_disks,rdev,tmp)

typedef struct mdk_thread_s {
	void			(*run) (mddev_t *mddev);
	mddev_t			*mddev;
	wait_queue_head_t	wqueue;
	unsigned long           flags;
	struct task_struct	*tsk;
	unsigned long		timeout;
} mdk_thread_t;

#define THREAD_WAKEUP  0

#define __wait_event_lock_irq(wq, condition, lock, cmd) 		\
do {									\
	wait_queue_t __wait;						\
	init_waitqueue_entry(&__wait, current);				\
									\
	add_wait_queue(&wq, &__wait);					\
	for (;;) {							\
		set_current_state(TASK_UNINTERRUPTIBLE);		\
		if (condition)						\
			break;						\
		spin_unlock_irq(&lock);					\
		cmd;							\
		schedule();						\
		spin_lock_irq(&lock);					\
	}								\
	current->state = TASK_RUNNING;					\
	remove_wait_queue(&wq, &__wait);				\
} while (0)

#define wait_event_lock_irq(wq, condition, lock, cmd) 			\
do {									\
	if (condition)	 						\
		break;							\
	__wait_event_lock_irq(wq, condition, lock, cmd);		\
} while (0)

#endif
Commit	Line	Data
1da177e4 LT	1	/*
	2	md_k.h : kernel internal structure of the Linux MD driver
	3	Copyright (C) 1996-98 Ingo Molnar, Gadi Oxman
	4
	5	This program is free software; you can redistribute it and/or modify
	6	it under the terms of the GNU General Public License as published by
	7	the Free Software Foundation; either version 2, or (at your option)
	8	any later version.
	9
	10	You should have received a copy of the GNU General Public License
	11	(for example /usr/src/linux/COPYING); if not, write to the Free
	12	Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
	13	*/
	14
	15	#ifndef _MD_K_H
	16	#define _MD_K_H
	17
06d91a5f N	18	/* and dm-bio-list.h is not under include/linux because.... ??? */
	19	#include "../../../drivers/md/dm-bio-list.h"
	20
1da177e4 LT	21	#define LEVEL_MULTIPATH (-4)
	22	#define LEVEL_LINEAR (-1)
	23	#define LEVEL_FAULTY (-5)
	24
2604b703 N	25	/* we need a value for 'no level specified' and 0
	26	* means 'raid0', so we need something else. This is
	27	* for internal use only
	28	*/
	29	#define LEVEL_NONE (-1000000)
	30
1da177e4 LT	31	#define MaxSector (~(sector_t)0)
	32	#define MD_THREAD_NAME_MAX 14
	33
1da177e4 LT	34	typedef struct mddev_s mddev_t;
	35	typedef struct mdk_rdev_s mdk_rdev_t;
	36
	37	#define MAX_MD_DEVS 256 /* Max number of md dev */
	38
	39	/*
	40	* options passed in raidrun:
	41	*/
	42
	43	#define MAX_CHUNK_SIZE (4096*1024)
	44
1da177e4 LT	45	/*
	46	* MD's 'extended' device
	47	*/
	48	struct mdk_rdev_s
	49	{
	50	struct list_head same_set; /* RAID devices within the same set */
	51
	52	sector_t size; /* Device size (in blocks) */
	53	mddev_t mddev; / RAID array if running */
	54	unsigned long last_events; /* IO event timestamp */
	55
	56	struct block_device bdev; / block device handle */
	57
	58	struct page *sb_page;
	59	int sb_loaded;
	60	sector_t data_offset; /* start of data in array */
	61	sector_t sb_offset;
0002b271	62	int sb_size; /* bytes in the superblock */
1da177e4 LT	63	int preferred_minor; /* autorun support */
1da177e4 LT	64
86e6ffdd N	65	struct kobject kobj;
86e6ffdd N	66
1da177e4 LT	67	/* A device can be in one of three states based on two flags:
	68	* Not working: faulty==1 in_sync==0
	69	* Fully working: faulty==0 in_sync==1
	70	* Working, but not
	71	* in sync with array
	72	* faulty==0 in_sync==0
	73	*
	74	* It can never have faulty==1, in_sync==1
	75	* This reduces the burden of testing multiple flags in many cases
	76	*/
1da177e4	77
b2d444d7 N	78	unsigned long flags;
	79	#define Faulty 1 /* device is known to have a fault */
	80	#define In_sync 2 /* device is in_sync with rest of array */
8ddf9efe	81	#define WriteMostly 4 /* Avoid reading if at all possible */
a9701a30	82	#define BarriersNotsupp 5 /* BIO_RW_BARRIER is not supported */
8ddf9efe	83
1da177e4 LT	84	int desc_nr; /* descriptor index in the superblock */
1da177e4 LT	85	int raid_disk; /* role of device in array */
41158c7e N	86	int saved_raid_disk; /* role that device used to have in the
	87	* array and could again if we did a partial
	88	* resync from the bitmap
	89	*/
1da177e4 LT	90
	91	atomic_t nr_pending; /* number of pending requests.
	92	* only maintained for arrays that
	93	* support hot removal
	94	*/
ba22dcbf N	95	atomic_t read_errors; /* number of consecutive read errors that
	96	* we have tried to ignore.
	97	*/
1da177e4 LT	98	};
1da177e4 LT	99
1da177e4 LT	100	struct mddev_s
	101	{
	102	void *private;
2604b703	103	struct mdk_personality *pers;
1da177e4 LT	104	dev_t unit;
	105	int md_minor;
	106	struct list_head disks;
	107	int sb_dirty;
	108	int ro;
	109
	110	struct gendisk *gendisk;
	111
eae1701f N	112	struct kobject kobj;
eae1701f N	113
1da177e4 LT	114	/* Superblock information */
	115	int major_version,
	116	minor_version,
	117	patch_version;
	118	int persistent;
	119	int chunk_size;
	120	time_t ctime, utime;
	121	int level, layout;
	122	int raid_disks;
	123	int max_disks;
	124	sector_t size; /* used size of component devices */
	125	sector_t array_size; /* exported array size */
	126	__u64 events;
	127
	128	char uuid[16];
	129
	130	struct mdk_thread_s thread; / management thread */
	131	struct mdk_thread_s sync_thread; / doing resync or reconstruct */
	132	sector_t curr_resync; /* blocks scheduled */
	133	unsigned long resync_mark; /* a recent timestamp */
	134	sector_t resync_mark_cnt;/* blocks written at resync_mark */
	135
	136	sector_t resync_max_sectors; /* may be set by personality */
9d88883e N	137
	138	sector_t resync_mismatches; /* count of sectors where
	139	* parity/replica mismatch found
	140	*/
6ff8d8ec	141	int ok_start_degraded;
1da177e4 LT	142	/* recovery/resync flags
	143	* NEEDED: we might need to start a resync/recover
	144	* RUNNING: a thread is running, or about to be started
	145	* SYNC: actually doing a resync, not a recovery
	146	* ERR: and IO error was detected - abort the resync/recovery
	147	* INTR: someone requested a (clean) early abort.
	148	* DONE: thread is done and is waiting to be reaped
24dd469d N	149	* REQUEST: user-space has requested a sync (used with SYNC)
24dd469d N	150	* CHECK: user-space request for for check-only, no repair
1da177e4 LT	151	*/
	152	#define MD_RECOVERY_RUNNING 0
	153	#define MD_RECOVERY_SYNC 1
	154	#define MD_RECOVERY_ERR 2
	155	#define MD_RECOVERY_INTR 3
	156	#define MD_RECOVERY_DONE 4
	157	#define MD_RECOVERY_NEEDED 5
24dd469d N	158	#define MD_RECOVERY_REQUESTED 6
24dd469d N	159	#define MD_RECOVERY_CHECK 7
1da177e4 LT	160	unsigned long recovery;
	161
	162	int in_sync; /* know to not need resync */
	163	struct semaphore reconfig_sem;
	164	atomic_t active;
	165
	166	int changed; /* true if we might need to reread partition info */
	167	int degraded; /* whether md should consider
	168	* adding a spare
	169	*/
a9701a30 N	170	int barriers_work; /* initialised to true, cleared as soon
	171	* as a barrier request to slave
	172	* fails. Only supported
	173	*/
	174	struct bio biolist; / bios that need to be retried
	175	* because BIO_RW_BARRIER is not supported
	176	*/
1da177e4 LT	177
	178	atomic_t recovery_active; /* blocks scheduled, but not written */
	179	wait_queue_head_t recovery_wait;
	180	sector_t recovery_cp;
06d91a5f N	181
06d91a5f N	182	spinlock_t write_lock;
3d310eb7	183	wait_queue_head_t sb_wait; /* for waiting on superblock updates */
7bfa19f2	184	atomic_t pending_writes; /* number of active superblock writes */
06d91a5f	185
1da177e4 LT	186	unsigned int safemode; /* if set, update "clean" superblock
	187	* when no writes pending.
	188	*/
	189	unsigned int safemode_delay;
	190	struct timer_list safemode_timer;
	191	atomic_t writes_pending;
	192	request_queue_t queue; / for plugging ... */
	193
4b6d287f N	194	atomic_t write_behind; /* outstanding async IO */
	195	unsigned int max_write_behind; /* 0 = sync */
	196
32a7627c N	197	struct bitmap bitmap; / the bitmap for the device */
32a7627c N	198	struct file bitmap_file; / the bitmap file */
a654b9d8 N	199	long bitmap_offset; /* offset from superblock of
	200	* start of bitmap. May be
	201	* negative, but not '0'
	202	*/
36fa3063 N	203	long default_bitmap_offset; /* this is the offset to use when
	204	* hot-adding a bitmap. It should
	205	* eventually be settable by sysfs.
	206	*/
32a7627c	207
1da177e4 LT	208	struct list_head all_mddevs;
	209	};
	210
	211
	212	static inline void rdev_dec_pending(mdk_rdev_t rdev, mddev_t mddev)
	213	{
b2d444d7	214	int faulty = test_bit(Faulty, &rdev->flags);
1da177e4 LT	215	if (atomic_dec_and_test(&rdev->nr_pending) && faulty)
	216	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	217	}
	218
	219	static inline void md_sync_acct(struct block_device *bdev, unsigned long nr_sectors)
	220	{
	221	atomic_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io);
	222	}
	223
2604b703	224	struct mdk_personality
1da177e4 LT	225	{
1da177e4 LT	226	char *name;
2604b703 N	227	int level;
2604b703 N	228	struct list_head list;
1da177e4 LT	229	struct module *owner;
	230	int (make_request)(request_queue_t q, struct bio *bio);
	231	int (run)(mddev_t mddev);
	232	int (stop)(mddev_t mddev);
	233	void (status)(struct seq_file seq, mddev_t *mddev);
	234	/* error_handler must set ->faulty and clear ->in_sync
	235	* if appropriate, and should abort recovery if needed
	236	*/
	237	void (error_handler)(mddev_t mddev, mdk_rdev_t *rdev);
	238	int (hot_add_disk) (mddev_t mddev, mdk_rdev_t *rdev);
	239	int (hot_remove_disk) (mddev_t mddev, int number);
	240	int (spare_active) (mddev_t mddev);
57afd89f	241	sector_t (sync_request)(mddev_t mddev, sector_t sector_nr, int *skipped, int go_faster);
1da177e4 LT	242	int (resize) (mddev_t mddev, sector_t sectors);
	243	int (reshape) (mddev_t mddev, int raid_disks);
	244	int (reconfig) (mddev_t mddev, int layout, int chunk_size);
36fa3063 N	245	/* quiesce moves between quiescence states
	246	* 0 - fully active
	247	* 1 - no new requests allowed
	248	* others - reserved
	249	*/
	250	void (quiesce) (mddev_t mddev, int state);
1da177e4 LT	251	};
	252
	253
007583c9 N	254	struct md_sysfs_entry {
	255	struct attribute attr;
	256	ssize_t (show)(mddev_t , char *);
	257	ssize_t (store)(mddev_t , const char *, size_t);
	258	};
	259
	260
1da177e4 LT	261	static inline char * mdname (mddev_t * mddev)
	262	{
	263	return mddev->gendisk ? mddev->gendisk->disk_name : "mdX";
	264	}
	265
	266	extern mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr);
	267
	268	/*
	269	* iterates through some rdev ringlist. It's safe to remove the
	270	* current 'rdev'. Dont touch 'tmp' though.
	271	*/
	272	#define ITERATE_RDEV_GENERIC(head,rdev,tmp) \
	273	\
	274	for ((tmp) = (head).next; \
	275	(rdev) = (list_entry((tmp), mdk_rdev_t, same_set)), \
	276	(tmp) = (tmp)->next, (tmp)->prev != &(head) \
	277	; )
	278	/*
	279	* iterates through the 'same array disks' ringlist
	280	*/
	281	#define ITERATE_RDEV(mddev,rdev,tmp) \
	282	ITERATE_RDEV_GENERIC((mddev)->disks,rdev,tmp)
	283
	284	/*
	285	* Iterates through 'pending RAID disks'
	286	*/
	287	#define ITERATE_RDEV_PENDING(rdev,tmp) \
	288	ITERATE_RDEV_GENERIC(pending_raid_disks,rdev,tmp)
	289
	290	typedef struct mdk_thread_s {
	291	void (run) (mddev_t mddev);
	292	mddev_t *mddev;
	293	wait_queue_head_t wqueue;
	294	unsigned long flags;
1da177e4	295	struct task_struct *tsk;
32a7627c	296	unsigned long timeout;
1da177e4 LT	297	} mdk_thread_t;
	298
	299	#define THREAD_WAKEUP 0
	300
	301	#define __wait_event_lock_irq(wq, condition, lock, cmd) \
	302	do { \
	303	wait_queue_t __wait; \
	304	init_waitqueue_entry(&__wait, current); \
	305	\
	306	add_wait_queue(&wq, &__wait); \
	307	for (;;) { \
	308	set_current_state(TASK_UNINTERRUPTIBLE); \
	309	if (condition) \
	310	break; \
	311	spin_unlock_irq(&lock); \
	312	cmd; \
	313	schedule(); \
	314	spin_lock_irq(&lock); \
	315	} \
	316	current->state = TASK_RUNNING; \
	317	remove_wait_queue(&wq, &__wait); \
	318	} while (0)
	319
	320	#define wait_event_lock_irq(wq, condition, lock, cmd) \
	321	do { \
	322	if (condition) \
	323	break; \
	324	__wait_event_lock_irq(wq, condition, lock, cmd); \
	325	} while (0)
	326
	327	#endif
	328