[mirror_ubuntu-kernels.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 MD_RESERVED       0UL
#define LINEAR            1UL
#define RAID0             2UL
#define RAID1             3UL
#define RAID5             4UL
#define TRANSLUCENT       5UL
#define HSM               6UL
#define MULTIPATH         7UL
#define RAID6		  8UL
#define	RAID10		  9UL
#define FAULTY		  10UL
#define MAX_PERSONALITY   11UL

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

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

static inline int pers_to_level (int pers)
{
	switch (pers) {
		case FAULTY:		return LEVEL_FAULTY;
		case MULTIPATH:		return LEVEL_MULTIPATH;
		case HSM:		return -3;
		case TRANSLUCENT:	return -2;
		case LINEAR:		return LEVEL_LINEAR;
		case RAID0:		return 0;
		case RAID1:		return 1;
		case RAID5:		return 5;
		case RAID6:		return 6;
		case RAID10:		return 10;
	}
	BUG();
	return MD_RESERVED;
}

static inline int level_to_pers (int level)
{
	switch (level) {
		case LEVEL_FAULTY: return FAULTY;
		case LEVEL_MULTIPATH: return MULTIPATH;
		case -3: return HSM;
		case -2: return TRANSLUCENT;
		case LEVEL_LINEAR: return LINEAR;
		case 0: return RAID0;
		case 1: return RAID1;
		case 4:
		case 5: return RAID5;
		case 6: return RAID6;
		case 10: return RAID10;
	}
	return MD_RESERVED;
}

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
	 */
	int faulty;			/* if faulty do not issue IO requests */
	int in_sync;			/* device is a full member of the array */

	unsigned long	flags;		/* Should include faulty and in_sync here. */
#define	WriteMostly	4		/* Avoid reading if at all possible */

	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
					 */
};

typedef struct mdk_personality_s mdk_personality_t;

struct mddev_s
{
	void				*private;
	mdk_personality_t		*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 */
	/* 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
	 */
#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
	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
							 */

	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 = rdev->faulty;
	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_s
{
	char *name;
	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);
};


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 completion	*event;
	struct task_struct	*tsk;
	unsigned long		timeout;
	const char		*name;
} 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 MD_RESERVED 0UL
	22	#define LINEAR 1UL
	23	#define RAID0 2UL
	24	#define RAID1 3UL
	25	#define RAID5 4UL
	26	#define TRANSLUCENT 5UL
	27	#define HSM 6UL
	28	#define MULTIPATH 7UL
	29	#define RAID6 8UL
	30	#define RAID10 9UL
	31	#define FAULTY 10UL
	32	#define MAX_PERSONALITY 11UL
	33
	34	#define LEVEL_MULTIPATH (-4)
	35	#define LEVEL_LINEAR (-1)
	36	#define LEVEL_FAULTY (-5)
	37
	38	#define MaxSector (~(sector_t)0)
	39	#define MD_THREAD_NAME_MAX 14
	40
	41	static inline int pers_to_level (int pers)
	42	{
	43	switch (pers) {
	44	case FAULTY: return LEVEL_FAULTY;
	45	case MULTIPATH: return LEVEL_MULTIPATH;
	46	case HSM: return -3;
	47	case TRANSLUCENT: return -2;
	48	case LINEAR: return LEVEL_LINEAR;
	49	case RAID0: return 0;
	50	case RAID1: return 1;
	51	case RAID5: return 5;
	52	case RAID6: return 6;
	53	case RAID10: return 10;
	54	}
	55	BUG();
	56	return MD_RESERVED;
	57	}
	58
	59	static inline int level_to_pers (int level)
	60	{
	61	switch (level) {
	62	case LEVEL_FAULTY: return FAULTY;
	63	case LEVEL_MULTIPATH: return MULTIPATH;
	64	case -3: return HSM;
	65	case -2: return TRANSLUCENT;
	66	case LEVEL_LINEAR: return LINEAR;
	67	case 0: return RAID0;
	68	case 1: return RAID1;
	69	case 4:
	70	case 5: return RAID5;
	71	case 6: return RAID6;
	72	case 10: return RAID10;
	73	}
	74	return MD_RESERVED;
	75	}
	76
	77	typedef struct mddev_s mddev_t;
	78	typedef struct mdk_rdev_s mdk_rdev_t;
	79
	80	#define MAX_MD_DEVS 256 /* Max number of md dev */
	81
	82	/*
	83	* options passed in raidrun:
	84	*/
85
86	#define MAX_CHUNK_SIZE (4096*1024)
87
1da177e4 LT	88	/*
	89	* MD's 'extended' device
	90	*/
	91	struct mdk_rdev_s
	92	{
	93	struct list_head same_set; /* RAID devices within the same set */
	94
	95	sector_t size; /* Device size (in blocks) */
	96	mddev_t mddev; / RAID array if running */
	97	unsigned long last_events; /* IO event timestamp */
	98
	99	struct block_device bdev; / block device handle */
	100
	101	struct page *sb_page;
	102	int sb_loaded;
	103	sector_t data_offset; /* start of data in array */
	104	sector_t sb_offset;
0002b271	105	int sb_size; /* bytes in the superblock */
1da177e4 LT	106	int preferred_minor; /* autorun support */
1da177e4 LT	107
86e6ffdd N	108	struct kobject kobj;
86e6ffdd N	109
1da177e4 LT	110	/* A device can be in one of three states based on two flags:
	111	* Not working: faulty==1 in_sync==0
	112	* Fully working: faulty==0 in_sync==1
	113	* Working, but not
	114	* in sync with array
	115	* faulty==0 in_sync==0
	116	*
	117	* It can never have faulty==1, in_sync==1
	118	* This reduces the burden of testing multiple flags in many cases
	119	*/
	120	int faulty; /* if faulty do not issue IO requests */
	121	int in_sync; /* device is a full member of the array */
	122
8ddf9efe N	123	unsigned long flags; /* Should include faulty and in_sync here. */
	124	#define WriteMostly 4 /* Avoid reading if at all possible */
	125
1da177e4 LT	126	int desc_nr; /* descriptor index in the superblock */
1da177e4 LT	127	int raid_disk; /* role of device in array */
41158c7e N	128	int saved_raid_disk; /* role that device used to have in the
	129	* array and could again if we did a partial
	130	* resync from the bitmap
	131	*/
1da177e4 LT	132
	133	atomic_t nr_pending; /* number of pending requests.
	134	* only maintained for arrays that
	135	* support hot removal
	136	*/
	137	};
	138
	139	typedef struct mdk_personality_s mdk_personality_t;
	140
	141	struct mddev_s
	142	{
	143	void *private;
	144	mdk_personality_t *pers;
	145	dev_t unit;
	146	int md_minor;
	147	struct list_head disks;
	148	int sb_dirty;
	149	int ro;
	150
	151	struct gendisk *gendisk;
	152
eae1701f N	153	struct kobject kobj;
eae1701f N	154
1da177e4 LT	155	/* Superblock information */
	156	int major_version,
	157	minor_version,
	158	patch_version;
	159	int persistent;
	160	int chunk_size;
	161	time_t ctime, utime;
	162	int level, layout;
	163	int raid_disks;
	164	int max_disks;
	165	sector_t size; /* used size of component devices */
	166	sector_t array_size; /* exported array size */
	167	__u64 events;
	168
	169	char uuid[16];
	170
	171	struct mdk_thread_s thread; / management thread */
	172	struct mdk_thread_s sync_thread; / doing resync or reconstruct */
	173	sector_t curr_resync; /* blocks scheduled */
	174	unsigned long resync_mark; /* a recent timestamp */
	175	sector_t resync_mark_cnt;/* blocks written at resync_mark */
	176
	177	sector_t resync_max_sectors; /* may be set by personality */
	178	/* recovery/resync flags
	179	* NEEDED: we might need to start a resync/recover
	180	* RUNNING: a thread is running, or about to be started
	181	* SYNC: actually doing a resync, not a recovery
	182	* ERR: and IO error was detected - abort the resync/recovery
	183	* INTR: someone requested a (clean) early abort.
	184	* DONE: thread is done and is waiting to be reaped
	185	*/
	186	#define MD_RECOVERY_RUNNING 0
	187	#define MD_RECOVERY_SYNC 1
	188	#define MD_RECOVERY_ERR 2
	189	#define MD_RECOVERY_INTR 3
	190	#define MD_RECOVERY_DONE 4
	191	#define MD_RECOVERY_NEEDED 5
	192	unsigned long recovery;
	193
	194	int in_sync; /* know to not need resync */
	195	struct semaphore reconfig_sem;
	196	atomic_t active;
	197
	198	int changed; /* true if we might need to reread partition info */
	199	int degraded; /* whether md should consider
	200	* adding a spare
	201	*/
	202
	203	atomic_t recovery_active; /* blocks scheduled, but not written */
	204	wait_queue_head_t recovery_wait;
	205	sector_t recovery_cp;
06d91a5f N	206
06d91a5f N	207	spinlock_t write_lock;
3d310eb7	208	wait_queue_head_t sb_wait; /* for waiting on superblock updates */
7bfa19f2	209	atomic_t pending_writes; /* number of active superblock writes */
06d91a5f	210
1da177e4 LT	211	unsigned int safemode; /* if set, update "clean" superblock
	212	* when no writes pending.
	213	*/
	214	unsigned int safemode_delay;
	215	struct timer_list safemode_timer;
	216	atomic_t writes_pending;
	217	request_queue_t queue; / for plugging ... */
	218
4b6d287f N	219	atomic_t write_behind; /* outstanding async IO */
	220	unsigned int max_write_behind; /* 0 = sync */
	221
32a7627c N	222	struct bitmap bitmap; / the bitmap for the device */
32a7627c N	223	struct file bitmap_file; / the bitmap file */
a654b9d8 N	224	long bitmap_offset; /* offset from superblock of
	225	* start of bitmap. May be
	226	* negative, but not '0'
	227	*/
36fa3063 N	228	long default_bitmap_offset; /* this is the offset to use when
	229	* hot-adding a bitmap. It should
	230	* eventually be settable by sysfs.
	231	*/
32a7627c	232
1da177e4 LT	233	struct list_head all_mddevs;
	234	};
	235
	236
	237	static inline void rdev_dec_pending(mdk_rdev_t rdev, mddev_t mddev)
	238	{
	239	int faulty = rdev->faulty;
	240	if (atomic_dec_and_test(&rdev->nr_pending) && faulty)
	241	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	242	}
	243
	244	static inline void md_sync_acct(struct block_device *bdev, unsigned long nr_sectors)
	245	{
	246	atomic_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io);
	247	}
	248
	249	struct mdk_personality_s
	250	{
	251	char *name;
	252	struct module *owner;
	253	int (make_request)(request_queue_t q, struct bio *bio);
	254	int (run)(mddev_t mddev);
	255	int (stop)(mddev_t mddev);
	256	void (status)(struct seq_file seq, mddev_t *mddev);
	257	/* error_handler must set ->faulty and clear ->in_sync
	258	* if appropriate, and should abort recovery if needed
	259	*/
	260	void (error_handler)(mddev_t mddev, mdk_rdev_t *rdev);
	261	int (hot_add_disk) (mddev_t mddev, mdk_rdev_t *rdev);
	262	int (hot_remove_disk) (mddev_t mddev, int number);
	263	int (spare_active) (mddev_t mddev);
57afd89f	264	sector_t (sync_request)(mddev_t mddev, sector_t sector_nr, int *skipped, int go_faster);
1da177e4 LT	265	int (resize) (mddev_t mddev, sector_t sectors);
	266	int (reshape) (mddev_t mddev, int raid_disks);
	267	int (reconfig) (mddev_t mddev, int layout, int chunk_size);
36fa3063 N	268	/* quiesce moves between quiescence states
	269	* 0 - fully active
	270	* 1 - no new requests allowed
	271	* others - reserved
	272	*/
	273	void (quiesce) (mddev_t mddev, int state);
1da177e4 LT	274	};
	275
	276
	277	static inline char * mdname (mddev_t * mddev)
	278	{
	279	return mddev->gendisk ? mddev->gendisk->disk_name : "mdX";
	280	}
	281
	282	extern mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr);
	283
	284	/*
	285	* iterates through some rdev ringlist. It's safe to remove the
	286	* current 'rdev'. Dont touch 'tmp' though.
	287	*/
	288	#define ITERATE_RDEV_GENERIC(head,rdev,tmp) \
	289	\
	290	for ((tmp) = (head).next; \
	291	(rdev) = (list_entry((tmp), mdk_rdev_t, same_set)), \
	292	(tmp) = (tmp)->next, (tmp)->prev != &(head) \
	293	; )
	294	/*
	295	* iterates through the 'same array disks' ringlist
	296	*/
	297	#define ITERATE_RDEV(mddev,rdev,tmp) \
	298	ITERATE_RDEV_GENERIC((mddev)->disks,rdev,tmp)
	299
	300	/*
	301	* Iterates through 'pending RAID disks'
	302	*/
	303	#define ITERATE_RDEV_PENDING(rdev,tmp) \
	304	ITERATE_RDEV_GENERIC(pending_raid_disks,rdev,tmp)
	305
	306	typedef struct mdk_thread_s {
	307	void (run) (mddev_t mddev);
	308	mddev_t *mddev;
	309	wait_queue_head_t wqueue;
	310	unsigned long flags;
	311	struct completion *event;
	312	struct task_struct *tsk;
32a7627c	313	unsigned long timeout;
1da177e4 LT	314	const char *name;
	315	} mdk_thread_t;
	316
	317	#define THREAD_WAKEUP 0
	318
	319	#define __wait_event_lock_irq(wq, condition, lock, cmd) \
	320	do { \
	321	wait_queue_t __wait; \
	322	init_waitqueue_entry(&__wait, current); \
	323	\
	324	add_wait_queue(&wq, &__wait); \
	325	for (;;) { \
	326	set_current_state(TASK_UNINTERRUPTIBLE); \
	327	if (condition) \
	328	break; \
	329	spin_unlock_irq(&lock); \
	330	cmd; \
	331	schedule(); \
	332	spin_lock_irq(&lock); \
	333	} \
	334	current->state = TASK_RUNNING; \
	335	remove_wait_queue(&wq, &__wait); \
	336	} while (0)
	337
	338	#define wait_event_lock_irq(wq, condition, lock, cmd) \
	339	do { \
	340	if (condition) \
	341	break; \
	342	__wait_event_lock_irq(wq, condition, lock, cmd); \
	343	} while (0)
	344
	345	#endif
	346