[net-next-2.6.git] / drivers / md / md.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_MD_H
#define _MD_MD_H

#include <linux/blkdev.h>
#include <linux/kobject.h>
#include <linux/list.h>
#include <linux/mm.h>
#include <linux/mutex.h>
#include <linux/timer.h>
#include <linux/wait.h>
#include <linux/workqueue.h>

#define MaxSector (~(sector_t)0)

typedef struct mddev_s mddev_t;
typedef struct mdk_rdev_s mdk_rdev_t;

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

	sector_t sectors;		/* Device size (in 512bytes sectors) */
	mddev_t *mddev;			/* RAID array if running */
	int last_events;		/* IO event timestamp */

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

	struct page	*sb_page;
	int		sb_loaded;
	__u64		sb_events;
	sector_t	data_offset;	/* start of data in array */
	sector_t 	sb_start;	/* offset of the super block (in 512byte sectors) */
	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 */
#define	AllReserved	6		/* If whole device is reserved for
					 * one array */
#define	AutoDetected	7		/* added by auto-detect */
#define Blocked		8		/* An error occured on an externally
					 * managed array, don't allow writes
					 * until it is cleared */
	wait_queue_head_t blocked_wait;

	int desc_nr;			/* descriptor index in the superblock */
	int raid_disk;			/* role of device in array */
	int new_raid_disk;		/* role that the device will have in
					 * the array after a level-change completes.
					 */
	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
					 */
	sector_t	recovery_offset;/* If this device has been partially
					 * recovered, this is where we were
					 * up to.
					 */

	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 timespec last_read_error;	/* monotonic time since our
						 * last read error
						 */
	atomic_t	corrected_errors; /* number of corrected read errors,
					   * for reporting to userspace and storing
					   * in superblock.
					   */
	struct work_struct del_work;	/* used for delayed sysfs removal */

	struct sysfs_dirent *sysfs_state; /* handle for 'state'
					   * sysfs entry */
};

struct mddev_s
{
	void				*private;
	struct mdk_personality		*pers;
	dev_t				unit;
	int				md_minor;
	struct list_head 		disks;
	unsigned long			flags;
#define MD_CHANGE_DEVS	0	/* Some device status has changed */
#define MD_CHANGE_CLEAN 1	/* transition to or from 'clean' */
#define MD_CHANGE_PENDING 2	/* superblock update in progress */

	int				suspended;
	atomic_t			active_io;
	int				ro;

	struct gendisk			*gendisk;

	struct kobject			kobj;
	int				hold_active;
#define	UNTIL_IOCTL	1
#define	UNTIL_STOP	2

	/* Superblock information */
	int				major_version,
					minor_version,
					patch_version;
	int				persistent;
	int 				external;	/* metadata is
							 * managed externally */
	char				metadata_type[17]; /* externally set*/
	int				chunk_sectors;
	time_t				ctime, utime;
	int				level, layout;
	char				clevel[16];
	int				raid_disks;
	int				max_disks;
	sector_t			dev_sectors; 	/* used size of
							 * component devices */
	sector_t			array_sectors; /* exported array size */
	int				external_size; /* size managed
							* externally */
	__u64				events;
	/* If the last 'event' was simply a clean->dirty transition, and
	 * we didn't write it to the spares, then it is safe and simple
	 * to just decrement the event count on a dirty->clean transition.
	 * So we record that possibility here.
	 */
	int				can_decrease_events;

	char				uuid[16];

	/* If the array is being reshaped, we need to record the
	 * new shape and an indication of where we are up to.
	 * This is written to the superblock.
	 * If reshape_position is MaxSector, then no reshape is happening (yet).
	 */
	sector_t			reshape_position;
	int				delta_disks, new_level, new_layout;
	int				new_chunk_sectors;

	struct mdk_thread_s		*thread;	/* management thread */
	struct mdk_thread_s		*sync_thread;	/* doing resync or reconstruct */
	sector_t			curr_resync;	/* last block scheduled */
	/* As resync requests can complete out of order, we cannot easily track
	 * how much resync has been completed.  So we occasionally pause until
	 * everything completes, then set curr_resync_completed to curr_resync.
	 * As such it may be well behind the real resync mark, but it is a value
	 * we are certain of.
	 */
	sector_t			curr_resync_completed;
	unsigned long			resync_mark;	/* a recent timestamp */
	sector_t			resync_mark_cnt;/* blocks written at resync_mark */
	sector_t			curr_mark_cnt; /* blocks scheduled now */

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

	sector_t			resync_mismatches; /* count of sectors where
							    * parity/replica mismatch found
							    */

	/* allow user-space to request suspension of IO to regions of the array */
	sector_t			suspend_lo;
	sector_t			suspend_hi;
	/* if zero, use the system-wide default */
	int				sync_speed_min;
	int				sync_speed_max;

	/* resync even though the same disks are shared among md-devices */
	int				parallel_resync;

	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
	 * RECOVER:  doing recovery, or need to try it.
	 * INTR:     resync needs to be aborted for some reason
	 * 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 check-only, no repair
	 * RESHAPE:  A reshape is happening
	 *
	 * If neither SYNC or RESHAPE are set, then it is a recovery.
	 */
#define	MD_RECOVERY_RUNNING	0
#define	MD_RECOVERY_SYNC	1
#define	MD_RECOVERY_RECOVER	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
#define MD_RECOVERY_RESHAPE	8
#define	MD_RECOVERY_FROZEN	9

	unsigned long			recovery;
	int				recovery_disabled; /* if we detect that recovery
							    * will always fail, set this
							    * so we don't loop trying */

	int				in_sync;	/* know to not need resync */
	/* 'open_mutex' avoids races between 'md_open' and 'do_md_stop', so
	 * that we are never stopping an array while it is open.
	 * 'reconfig_mutex' protects all other reconfiguration.
	 * These locks are separate due to conflicting interactions
	 * with bdev->bd_mutex.
	 * Lock ordering is:
	 *  reconfig_mutex -> bd_mutex : e.g. do_md_run -> revalidate_disk
	 *  bd_mutex -> open_mutex:  e.g. __blkdev_get -> md_open
	 */
	struct mutex			open_mutex;
	struct mutex			reconfig_mutex;
	atomic_t			active;		/* general refcount */
	atomic_t			openers;	/* number of active opens */

	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;
	sector_t			resync_min;	/* user requested sync
							 * starts here */
	sector_t			resync_max;	/* resync should pause
							 * when it gets here */

	struct sysfs_dirent		*sysfs_state;	/* handle for 'array_state'
							 * file in sysfs.
							 */
	struct sysfs_dirent		*sysfs_action;  /* handle for 'sync_action' */

	struct work_struct del_work;	/* used for delayed sysfs removal */

	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; 
	struct request_queue		*queue;	/* for plugging ... */

	struct bitmap                   *bitmap; /* the bitmap for the device */
	struct {
		struct file		*file; /* the bitmap file */
		loff_t			offset; /* offset from superblock of
						 * start of bitmap. May be
						 * negative, but not '0'
						 * For external metadata, offset
						 * from start of device. 
						 */
		loff_t			default_offset; /* this is the offset to use when
							 * hot-adding a bitmap.  It should
							 * eventually be settable by sysfs.
							 */
		struct mutex		mutex;
		unsigned long		chunksize;
		unsigned long		daemon_sleep; /* how many seconds between updates? */
		unsigned long		max_write_behind; /* write-behind mode */
		int			external;
	} bitmap_info;

	atomic_t 			max_corr_read_errors; /* max read retries */
	struct list_head		all_mddevs;

	struct attribute_group		*to_remove;
	/* Generic barrier handling.
	 * If there is a pending barrier request, all other
	 * writes are blocked while the devices are flushed.
	 * The last to finish a flush schedules a worker to
	 * submit the barrier request (without the barrier flag),
	 * then submit more flush requests.
	 */
	struct bio *barrier;
	atomic_t flush_pending;
	struct work_struct barrier_work;
};


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)(mddev_t *mddev, 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);
	sector_t (*size) (mddev_t *mddev, sector_t sectors, int raid_disks);
	int (*check_reshape) (mddev_t *mddev);
	int (*start_reshape) (mddev_t *mddev);
	void (*finish_reshape) (mddev_t *mddev);
	/* quiesce moves between quiescence states
	 * 0 - fully active
	 * 1 - no new requests allowed
	 * others - reserved
	 */
	void (*quiesce) (mddev_t *mddev, int state);
	/* takeover is used to transition an array from one
	 * personality to another.  The new personality must be able
	 * to handle the data in the current layout.
	 * e.g. 2drive raid1 -> 2drive raid5
	 *      ndrive raid5 -> degraded n+1drive raid6 with special layout
	 * If the takeover succeeds, a new 'private' structure is returned.
	 * This needs to be installed and then ->run used to activate the
	 * array.
	 */
	void *(*takeover) (mddev_t *mddev);
};


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

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

/*
 * iterates through some rdev ringlist. It's safe to remove the
 * current 'rdev'. Dont touch 'tmp' though.
 */
#define rdev_for_each_list(rdev, tmp, head)				\
	list_for_each_entry_safe(rdev, tmp, head, same_set)

/*
 * iterates through the 'same array disks' ringlist
 */
#define rdev_for_each(rdev, tmp, mddev)				\
	list_for_each_entry_safe(rdev, tmp, &((mddev)->disks), same_set)

#define rdev_for_each_rcu(rdev, mddev)				\
	list_for_each_entry_rcu(rdev, &((mddev)->disks), same_set)

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)

static inline void safe_put_page(struct page *p)
{
	if (p) put_page(p);
}

extern int register_md_personality(struct mdk_personality *p);
extern int unregister_md_personality(struct mdk_personality *p);
extern mdk_thread_t * md_register_thread(void (*run) (mddev_t *mddev),
				mddev_t *mddev, const char *name);
extern void md_unregister_thread(mdk_thread_t *thread);
extern void md_wakeup_thread(mdk_thread_t *thread);
extern void md_check_recovery(mddev_t *mddev);
extern void md_write_start(mddev_t *mddev, struct bio *bi);
extern void md_write_end(mddev_t *mddev);
extern void md_done_sync(mddev_t *mddev, int blocks, int ok);
extern void md_error(mddev_t *mddev, mdk_rdev_t *rdev);

extern int mddev_congested(mddev_t *mddev, int bits);
extern void md_barrier_request(mddev_t *mddev, struct bio *bio);
extern void md_super_write(mddev_t *mddev, mdk_rdev_t *rdev,
			   sector_t sector, int size, struct page *page);
extern void md_super_wait(mddev_t *mddev);
extern int sync_page_io(struct block_device *bdev, sector_t sector, int size,
			struct page *page, int rw);
extern void md_do_sync(mddev_t *mddev);
extern void md_new_event(mddev_t *mddev);
extern int md_allow_write(mddev_t *mddev);
extern void md_wait_for_blocked_rdev(mdk_rdev_t *rdev, mddev_t *mddev);
extern void md_set_array_sectors(mddev_t *mddev, sector_t array_sectors);
extern int md_check_no_bitmap(mddev_t *mddev);
extern int md_integrity_register(mddev_t *mddev);
extern void md_integrity_add_rdev(mdk_rdev_t *rdev, mddev_t *mddev);
extern int strict_strtoul_scaled(const char *cp, unsigned long *res, int scale);
extern void restore_bitmap_write_access(struct file *file);

#endif /* _MD_MD_H */
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
63fe0817 CH	15	#ifndef _MD_MD_H
	16	#define _MD_MD_H
	17
	18	#include <linux/blkdev.h>
	19	#include <linux/kobject.h>
	20	#include <linux/list.h>
	21	#include <linux/mm.h>
	22	#include <linux/mutex.h>
	23	#include <linux/timer.h>
	24	#include <linux/wait.h>
	25	#include <linux/workqueue.h>
9361401e	26
1da177e4	27	#define MaxSector (~(sector_t)0)
1da177e4	28
1da177e4 LT	29	typedef struct mddev_s mddev_t;
	30	typedef struct mdk_rdev_s mdk_rdev_t;
	31
1da177e4 LT	32	/*
	33	* MD's 'extended' device
	34	*/
	35	struct mdk_rdev_s
	36	{
	37	struct list_head same_set; /* RAID devices within the same set */
	38
dd8ac336	39	sector_t sectors; /* Device size (in 512bytes sectors) */
1da177e4	40	mddev_t mddev; / RAID array if running */
eea1bf38	41	int last_events; /* IO event timestamp */
1da177e4 LT	42
	43	struct block_device bdev; / block device handle */
	44
	45	struct page *sb_page;
	46	int sb_loaded;
42543769	47	__u64 sb_events;
1da177e4	48	sector_t data_offset; /* start of data in array */
0f420358	49	sector_t sb_start; /* offset of the super block (in 512byte sectors) */
0002b271	50	int sb_size; /* bytes in the superblock */
1da177e4 LT	51	int preferred_minor; /* autorun support */
1da177e4 LT	52
86e6ffdd N	53	struct kobject kobj;
86e6ffdd N	54
1da177e4 LT	55	/* A device can be in one of three states based on two flags:
	56	* Not working: faulty==1 in_sync==0
	57	* Fully working: faulty==0 in_sync==1
	58	* Working, but not
	59	* in sync with array
	60	* faulty==0 in_sync==0
	61	*
	62	* It can never have faulty==1, in_sync==1
	63	* This reduces the burden of testing multiple flags in many cases
	64	*/
1da177e4	65
b2d444d7 N	66	unsigned long flags;
	67	#define Faulty 1 /* device is known to have a fault */
	68	#define In_sync 2 /* device is in_sync with rest of array */
8ddf9efe	69	#define WriteMostly 4 /* Avoid reading if at all possible */
a9701a30	70	#define BarriersNotsupp 5 /* BIO_RW_BARRIER is not supported */
c5d79adb N	71	#define AllReserved 6 /* If whole device is reserved for
c5d79adb N	72	* one array */
d0fae18f	73	#define AutoDetected 7 /* added by auto-detect */
6bfe0b49 DW	74	#define Blocked 8 /* An error occured on an externally
	75	* managed array, don't allow writes
	76	* until it is cleared */
	77	wait_queue_head_t blocked_wait;
8ddf9efe	78
1da177e4 LT	79	int desc_nr; /* descriptor index in the superblock */
1da177e4 LT	80	int raid_disk; /* role of device in array */
e93f68a1 N	81	int new_raid_disk; /* role that the device will have in
	82	* the array after a level-change completes.
	83	*/
41158c7e N	84	int saved_raid_disk; /* role that device used to have in the
	85	* array and could again if we did a partial
	86	* resync from the bitmap
	87	*/
5fd6c1dc N	88	sector_t recovery_offset;/* If this device has been partially
	89	* recovered, this is where we were
	90	* up to.
	91	*/
1da177e4 LT	92
	93	atomic_t nr_pending; /* number of pending requests.
	94	* only maintained for arrays that
	95	* support hot removal
	96	*/
ba22dcbf N	97	atomic_t read_errors; /* number of consecutive read errors that
	98	* we have tried to ignore.
	99	*/
1e50915f RB	100	struct timespec last_read_error; /* monotonic time since our
	101	* last read error
	102	*/
4dbcdc75 N	103	atomic_t corrected_errors; /* number of corrected read errors,
	104	* for reporting to userspace and storing
	105	* in superblock.
	106	*/
5792a285	107	struct work_struct del_work; /* used for delayed sysfs removal */
3c0ee63a N	108
	109	struct sysfs_dirent sysfs_state; / handle for 'state'
	110	* sysfs entry */
1da177e4 LT	111	};
1da177e4 LT	112
1da177e4 LT	113	struct mddev_s
	114	{
	115	void *private;
2604b703	116	struct mdk_personality *pers;
1da177e4 LT	117	dev_t unit;
	118	int md_minor;
	119	struct list_head disks;
850b2b42 N	120	unsigned long flags;
	121	#define MD_CHANGE_DEVS 0 /* Some device status has changed */
	122	#define MD_CHANGE_CLEAN 1 /* transition to or from 'clean' */
	123	#define MD_CHANGE_PENDING 2 /* superblock update in progress */
	124
409c57f3 N	125	int suspended;
409c57f3 N	126	atomic_t active_io;
1da177e4 LT	127	int ro;
	128
	129	struct gendisk *gendisk;
	130
eae1701f	131	struct kobject kobj;
d3374825 N	132	int hold_active;
d3374825 N	133	#define UNTIL_IOCTL 1
efeb53c0	134	#define UNTIL_STOP 2
eae1701f	135
1da177e4 LT	136	/* Superblock information */
	137	int major_version,
	138	minor_version,
	139	patch_version;
	140	int persistent;
e691063a N	141	int external; /* metadata is
	142	* managed externally */
	143	char metadata_type[17]; /* externally set*/
9d8f0363	144	int chunk_sectors;
1da177e4 LT	145	time_t ctime, utime;
1da177e4 LT	146	int level, layout;
d9d166c2	147	char clevel[16];
1da177e4 LT	148	int raid_disks;
1da177e4 LT	149	int max_disks;
58c0fed4 AN	150	sector_t dev_sectors; /* used size of
58c0fed4 AN	151	* component devices */
f233ea5c	152	sector_t array_sectors; /* exported array size */
b522adcd DW	153	int external_size; /* size managed
b522adcd DW	154	* externally */
1da177e4	155	__u64 events;
a8707c08 N	156	/* If the last 'event' was simply a clean->dirty transition, and
	157	* we didn't write it to the spares, then it is safe and simple
	158	* to just decrement the event count on a dirty->clean transition.
	159	* So we record that possibility here.
	160	*/
	161	int can_decrease_events;
1da177e4 LT	162
	163	char uuid[16];
	164
f6705578 N	165	/* If the array is being reshaped, we need to record the
	166	* new shape and an indication of where we are up to.
	167	* This is written to the superblock.
	168	* If reshape_position is MaxSector, then no reshape is happening (yet).
	169	*/
	170	sector_t reshape_position;
664e7c41 AN	171	int delta_disks, new_level, new_layout;
664e7c41 AN	172	int new_chunk_sectors;
f6705578	173
1da177e4 LT	174	struct mdk_thread_s thread; / management thread */
1da177e4 LT	175	struct mdk_thread_s sync_thread; / doing resync or reconstruct */
ff4e8d9a	176	sector_t curr_resync; /* last block scheduled */
97e4f42d N	177	/* As resync requests can complete out of order, we cannot easily track
	178	* how much resync has been completed. So we occasionally pause until
	179	* everything completes, then set curr_resync_completed to curr_resync.
	180	* As such it may be well behind the real resync mark, but it is a value
	181	* we are certain of.
	182	*/
	183	sector_t curr_resync_completed;
1da177e4 LT	184	unsigned long resync_mark; /* a recent timestamp */
1da177e4 LT	185	sector_t resync_mark_cnt;/* blocks written at resync_mark */
ff4e8d9a	186	sector_t curr_mark_cnt; /* blocks scheduled now */
1da177e4 LT	187
1da177e4 LT	188	sector_t resync_max_sectors; /* may be set by personality */
9d88883e N	189
	190	sector_t resync_mismatches; /* count of sectors where
	191	* parity/replica mismatch found
	192	*/
e464eafd N	193
	194	/* allow user-space to request suspension of IO to regions of the array */
	195	sector_t suspend_lo;
	196	sector_t suspend_hi;
88202a0c N	197	/* if zero, use the system-wide default */
	198	int sync_speed_min;
	199	int sync_speed_max;
	200
90b08710 BS	201	/* resync even though the same disks are shared among md-devices */
	202	int parallel_resync;
	203
6ff8d8ec	204	int ok_start_degraded;
1da177e4 LT	205	/* recovery/resync flags
	206	* NEEDED: we might need to start a resync/recover
	207	* RUNNING: a thread is running, or about to be started
	208	* SYNC: actually doing a resync, not a recovery
72a23c21	209	* RECOVER: doing recovery, or need to try it.
dfc70645	210	* INTR: resync needs to be aborted for some reason
1da177e4	211	* DONE: thread is done and is waiting to be reaped
24dd469d	212	* REQUEST: user-space has requested a sync (used with SYNC)
411c9403	213	* CHECK: user-space request for check-only, no repair
ccfcc3c1 N	214	* RESHAPE: A reshape is happening
	215	*
	216	* If neither SYNC or RESHAPE are set, then it is a recovery.
1da177e4 LT	217	*/
	218	#define MD_RECOVERY_RUNNING 0
	219	#define MD_RECOVERY_SYNC 1
72a23c21	220	#define MD_RECOVERY_RECOVER 2
1da177e4 LT	221	#define MD_RECOVERY_INTR 3
	222	#define MD_RECOVERY_DONE 4
	223	#define MD_RECOVERY_NEEDED 5
24dd469d N	224	#define MD_RECOVERY_REQUESTED 6
24dd469d N	225	#define MD_RECOVERY_CHECK 7
ccfcc3c1	226	#define MD_RECOVERY_RESHAPE 8
5fd6c1dc N	227	#define MD_RECOVERY_FROZEN 9
5fd6c1dc N	228
1da177e4	229	unsigned long recovery;
4044ba58 N	230	int recovery_disabled; /* if we detect that recovery
	231	* will always fail, set this
	232	* so we don't loop trying */
1da177e4 LT	233
1da177e4 LT	234	int in_sync; /* know to not need resync */
c8c00a69 N	235	/* 'open_mutex' avoids races between 'md_open' and 'do_md_stop', so
	236	* that we are never stopping an array while it is open.
	237	* 'reconfig_mutex' protects all other reconfiguration.
	238	* These locks are separate due to conflicting interactions
	239	* with bdev->bd_mutex.
	240	* Lock ordering is:
	241	* reconfig_mutex -> bd_mutex : e.g. do_md_run -> revalidate_disk
	242	* bd_mutex -> open_mutex: e.g. __blkdev_get -> md_open
	243	*/
	244	struct mutex open_mutex;
df5b89b3	245	struct mutex reconfig_mutex;
f2ea68cf N	246	atomic_t active; /* general refcount */
f2ea68cf N	247	atomic_t openers; /* number of active opens */
1da177e4	248
1da177e4 LT	249	int degraded; /* whether md should consider
	250	* adding a spare
	251	*/
a9701a30 N	252	int barriers_work; /* initialised to true, cleared as soon
	253	* as a barrier request to slave
	254	* fails. Only supported
	255	*/
	256	struct bio biolist; / bios that need to be retried
	257	* because BIO_RW_BARRIER is not supported
	258	*/
1da177e4 LT	259
	260	atomic_t recovery_active; /* blocks scheduled, but not written */
	261	wait_queue_head_t recovery_wait;
	262	sector_t recovery_cp;
5e96ee65 NB	263	sector_t resync_min; /* user requested sync
5e96ee65 NB	264	* starts here */
c6207277 N	265	sector_t resync_max; /* resync should pause
c6207277 N	266	* when it gets here */
06d91a5f	267
b62b7590 N	268	struct sysfs_dirent sysfs_state; / handle for 'array_state'
	269	* file in sysfs.
	270	*/
0c3573f1	271	struct sysfs_dirent sysfs_action; / handle for 'sync_action' */
b62b7590	272
d3374825 N	273	struct work_struct del_work; /* used for delayed sysfs removal */
d3374825 N	274
06d91a5f	275	spinlock_t write_lock;
3d310eb7	276	wait_queue_head_t sb_wait; /* for waiting on superblock updates */
7bfa19f2	277	atomic_t pending_writes; /* number of active superblock writes */
06d91a5f	278
1da177e4 LT	279	unsigned int safemode; /* if set, update "clean" superblock
	280	* when no writes pending.
	281	*/
	282	unsigned int safemode_delay;
	283	struct timer_list safemode_timer;
	284	atomic_t writes_pending;
165125e1	285	struct request_queue queue; / for plugging ... */
1da177e4	286
32a7627c	287	struct bitmap bitmap; / the bitmap for the device */
c3d9714e N	288	struct {
c3d9714e N	289	struct file file; / the bitmap file */
f6af949c	290	loff_t offset; /* offset from superblock of
c3d9714e N	291	* start of bitmap. May be
c3d9714e N	292	* negative, but not '0'
f6af949c N	293	* For external metadata, offset
f6af949c N	294	* from start of device.
c3d9714e	295	*/
f6af949c	296	loff_t default_offset; /* this is the offset to use when
c3d9714e N	297	* hot-adding a bitmap. It should
	298	* eventually be settable by sysfs.
	299	*/
	300	struct mutex mutex;
42a04b50 N	301	unsigned long chunksize;
	302	unsigned long daemon_sleep; /* how many seconds between updates? */
	303	unsigned long max_write_behind; /* write-behind mode */
ece5cff0	304	int external;
c3d9714e	305	} bitmap_info;
32a7627c	306
1e50915f	307	atomic_t max_corr_read_errors; /* max read retries */
1da177e4	308	struct list_head all_mddevs;
a2826aa9	309
a64c876f	310	struct attribute_group *to_remove;
a2826aa9 N	311	/* Generic barrier handling.
	312	* If there is a pending barrier request, all other
	313	* writes are blocked while the devices are flushed.
	314	* The last to finish a flush schedules a worker to
	315	* submit the barrier request (without the barrier flag),
	316	* then submit more flush requests.
	317	*/
	318	struct bio *barrier;
	319	atomic_t flush_pending;
	320	struct work_struct barrier_work;
1da177e4 LT	321	};
	322
	323
	324	static inline void rdev_dec_pending(mdk_rdev_t rdev, mddev_t mddev)
	325	{
b2d444d7	326	int faulty = test_bit(Faulty, &rdev->flags);
1da177e4 LT	327	if (atomic_dec_and_test(&rdev->nr_pending) && faulty)
	328	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	329	}
	330
	331	static inline void md_sync_acct(struct block_device *bdev, unsigned long nr_sectors)
	332	{
	333	atomic_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io);
	334	}
	335
2604b703	336	struct mdk_personality
1da177e4 LT	337	{
1da177e4 LT	338	char *name;
2604b703 N	339	int level;
2604b703 N	340	struct list_head list;
1da177e4	341	struct module *owner;
21a52c6d	342	int (make_request)(mddev_t mddev, struct bio *bio);
1da177e4 LT	343	int (run)(mddev_t mddev);
	344	int (stop)(mddev_t mddev);
	345	void (status)(struct seq_file seq, mddev_t *mddev);
	346	/* error_handler must set ->faulty and clear ->in_sync
	347	* if appropriate, and should abort recovery if needed
	348	*/
	349	void (error_handler)(mddev_t mddev, mdk_rdev_t *rdev);
	350	int (hot_add_disk) (mddev_t mddev, mdk_rdev_t *rdev);
	351	int (hot_remove_disk) (mddev_t mddev, int number);
	352	int (spare_active) (mddev_t mddev);
57afd89f	353	sector_t (sync_request)(mddev_t mddev, sector_t sector_nr, int *skipped, int go_faster);
1da177e4	354	int (resize) (mddev_t mddev, sector_t sectors);
80c3a6ce	355	sector_t (size) (mddev_t mddev, sector_t sectors, int raid_disks);
63c70c4f N	356	int (check_reshape) (mddev_t mddev);
63c70c4f N	357	int (start_reshape) (mddev_t mddev);
cea9c228	358	void (finish_reshape) (mddev_t mddev);
36fa3063 N	359	/* quiesce moves between quiescence states
	360	* 0 - fully active
	361	* 1 - no new requests allowed
	362	* others - reserved
	363	*/
	364	void (quiesce) (mddev_t mddev, int state);
245f46c2 N	365	/* takeover is used to transition an array from one
	366	* personality to another. The new personality must be able
	367	* to handle the data in the current layout.
	368	* e.g. 2drive raid1 -> 2drive raid5
	369	* ndrive raid5 -> degraded n+1drive raid6 with special layout
	370	* If the takeover succeeds, a new 'private' structure is returned.
	371	* This needs to be installed and then ->run used to activate the
	372	* array.
	373	*/
	374	void (takeover) (mddev_t *mddev);
1da177e4 LT	375	};
	376
	377
007583c9 N	378	struct md_sysfs_entry {
	379	struct attribute attr;
	380	ssize_t (show)(mddev_t , char *);
	381	ssize_t (store)(mddev_t , const char *, size_t);
	382	};
43a70507	383	extern struct attribute_group md_bitmap_group;
007583c9	384
1da177e4 LT	385	static inline char * mdname (mddev_t * mddev)
	386	{
	387	return mddev->gendisk ? mddev->gendisk->disk_name : "mdX";
	388	}
	389
1da177e4 LT	390	/*
	391	* iterates through some rdev ringlist. It's safe to remove the
	392	* current 'rdev'. Dont touch 'tmp' though.
	393	*/
159ec1fc CR	394	#define rdev_for_each_list(rdev, tmp, head) \
	395	list_for_each_entry_safe(rdev, tmp, head, same_set)
	396
1da177e4 LT	397	/*
	398	* iterates through the 'same array disks' ringlist
	399	*/
d089c6af	400	#define rdev_for_each(rdev, tmp, mddev) \
159ec1fc	401	list_for_each_entry_safe(rdev, tmp, &((mddev)->disks), same_set)
1da177e4	402
4b80991c N	403	#define rdev_for_each_rcu(rdev, mddev) \
	404	list_for_each_entry_rcu(rdev, &((mddev)->disks), same_set)
	405
1da177e4 LT	406	typedef struct mdk_thread_s {
	407	void (run) (mddev_t mddev);
	408	mddev_t *mddev;
	409	wait_queue_head_t wqueue;
	410	unsigned long flags;
1da177e4	411	struct task_struct *tsk;
32a7627c	412	unsigned long timeout;
1da177e4 LT	413	} mdk_thread_t;
	414
	415	#define THREAD_WAKEUP 0
	416
	417	#define __wait_event_lock_irq(wq, condition, lock, cmd) \
	418	do { \
	419	wait_queue_t __wait; \
	420	init_waitqueue_entry(&__wait, current); \
	421	\
	422	add_wait_queue(&wq, &__wait); \
	423	for (;;) { \
	424	set_current_state(TASK_UNINTERRUPTIBLE); \
	425	if (condition) \
	426	break; \
	427	spin_unlock_irq(&lock); \
	428	cmd; \
	429	schedule(); \
	430	spin_lock_irq(&lock); \
	431	} \
	432	current->state = TASK_RUNNING; \
	433	remove_wait_queue(&wq, &__wait); \
	434	} while (0)
	435
	436	#define wait_event_lock_irq(wq, condition, lock, cmd) \
	437	do { \
	438	if (condition) \
	439	break; \
	440	__wait_event_lock_irq(wq, condition, lock, cmd); \
	441	} while (0)
	442
1345b1d8 N	443	static inline void safe_put_page(struct page *p)
	444	{
	445	if (p) put_page(p);
	446	}
	447
92022950 N	448	extern int register_md_personality(struct mdk_personality *p);
	449	extern int unregister_md_personality(struct mdk_personality *p);
	450	extern mdk_thread_t * md_register_thread(void (run) (mddev_t mddev),
	451	mddev_t mddev, const char name);
	452	extern void md_unregister_thread(mdk_thread_t *thread);
	453	extern void md_wakeup_thread(mdk_thread_t *thread);
	454	extern void md_check_recovery(mddev_t *mddev);
	455	extern void md_write_start(mddev_t mddev, struct bio bi);
	456	extern void md_write_end(mddev_t *mddev);
	457	extern void md_done_sync(mddev_t *mddev, int blocks, int ok);
	458	extern void md_error(mddev_t mddev, mdk_rdev_t rdev);
	459
3fa841d7	460	extern int mddev_congested(mddev_t *mddev, int bits);
a2826aa9	461	extern void md_barrier_request(mddev_t mddev, struct bio bio);
92022950 N	462	extern void md_super_write(mddev_t mddev, mdk_rdev_t rdev,
	463	sector_t sector, int size, struct page *page);
	464	extern void md_super_wait(mddev_t *mddev);
	465	extern int sync_page_io(struct block_device *bdev, sector_t sector, int size,
	466	struct page *page, int rw);
	467	extern void md_do_sync(mddev_t *mddev);
	468	extern void md_new_event(mddev_t *mddev);
	469	extern int md_allow_write(mddev_t *mddev);
	470	extern void md_wait_for_blocked_rdev(mdk_rdev_t rdev, mddev_t mddev);
1f403624	471	extern void md_set_array_sectors(mddev_t *mddev, sector_t array_sectors);
0894cc30	472	extern int md_check_no_bitmap(mddev_t *mddev);
ac5e7113	473	extern int md_integrity_register(mddev_t *mddev);
72e02075 N	474	extern void md_integrity_add_rdev(mdk_rdev_t rdev, mddev_t mddev);
72e02075 N	475	extern int strict_strtoul_scaled(const char cp, unsigned long res, int scale);
43a70507	476	extern void restore_bitmap_write_access(struct file *file);
63fe0817 CH	477
63fe0817 CH	478	#endif /* _MD_MD_H */