md.c 138.9 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
/*
   md.c : Multiple Devices driver for Linux
	  Copyright (C) 1998, 1999, 2000 Ingo Molnar

     completely rewritten, based on the MD driver code from Marc Zyngier

   Changes:

   - RAID-1/RAID-5 extensions by Miguel de Icaza, Gadi Oxman, Ingo Molnar
   - RAID-6 extensions by H. Peter Anvin <hpa@zytor.com>
   - boot support for linear and striped mode by Harald Hoyer <HarryH@Royal.Net>
   - kerneld support by Boris Tobotras <boris@xtalk.msk.su>
   - kmod support by: Cyrus Durgin
   - RAID0 bugfixes: Mark Anthony Lisher <markal@iname.com>
   - Devfs support by Richard Gooch <rgooch@atnf.csiro.au>

   - lots of fixes and improvements to the RAID1/RAID5 and generic
     RAID code (such as request based resynchronization):

     Neil Brown <neilb@cse.unsw.edu.au>.

22 23 24
   - persistent bitmap code
     Copyright (C) 2003-2004, Paul Clements, SteelEye Technology, Inc.

L
Linus Torvalds 已提交
25 26 27 28 29 30 31 32 33 34 35
   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.
*/

#include <linux/module.h>
36
#include <linux/kthread.h>
L
Linus Torvalds 已提交
37 38
#include <linux/linkage.h>
#include <linux/raid/md.h>
39
#include <linux/raid/bitmap.h>
L
Linus Torvalds 已提交
40 41 42
#include <linux/sysctl.h>
#include <linux/buffer_head.h> /* for invalidate_bdev */
#include <linux/suspend.h>
43
#include <linux/poll.h>
A
Arjan van de Ven 已提交
44
#include <linux/mutex.h>
45
#include <linux/ctype.h>
L
Linus Torvalds 已提交
46 47 48

#include <linux/init.h>

49 50
#include <linux/file.h>

L
Linus Torvalds 已提交
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70
#ifdef CONFIG_KMOD
#include <linux/kmod.h>
#endif

#include <asm/unaligned.h>

#define MAJOR_NR MD_MAJOR
#define MD_DRIVER

/* 63 partitions with the alternate major number (mdp) */
#define MdpMinorShift 6

#define DEBUG 0
#define dprintk(x...) ((void)(DEBUG && printk(x)))


#ifndef MODULE
static void autostart_arrays (int part);
#endif

71
static LIST_HEAD(pers_list);
L
Linus Torvalds 已提交
72 73
static DEFINE_SPINLOCK(pers_lock);

74 75 76 77
static void md_print_devices(void);

#define MD_BUG(x...) { printk("md: bug in file %s, line %d\n", __FILE__, __LINE__); md_print_devices(); }

L
Linus Torvalds 已提交
78 79 80 81
/*
 * Current RAID-1,4,5 parallel reconstruction 'guaranteed speed limit'
 * is 1000 KB/sec, so the extra system load does not show up that much.
 * Increase it if you want to have more _guaranteed_ speed. Note that
82
 * the RAID driver will use the maximum available bandwidth if the IO
L
Linus Torvalds 已提交
83 84 85 86 87
 * subsystem is idle. There is also an 'absolute maximum' reconstruction
 * speed limit - in case reconstruction slows down your system despite
 * idle IO detection.
 *
 * you can change it via /proc/sys/dev/raid/speed_limit_min and _max.
88
 * or /sys/block/mdX/md/sync_speed_{min,max}
L
Linus Torvalds 已提交
89 90 91 92
 */

static int sysctl_speed_limit_min = 1000;
static int sysctl_speed_limit_max = 200000;
93 94 95 96 97 98 99 100 101 102 103
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;
}
L
Linus Torvalds 已提交
104 105 106 107 108 109 110 111 112

static struct ctl_table_header *raid_table_header;

static ctl_table raid_table[] = {
	{
		.ctl_name	= DEV_RAID_SPEED_LIMIT_MIN,
		.procname	= "speed_limit_min",
		.data		= &sysctl_speed_limit_min,
		.maxlen		= sizeof(int),
113
		.mode		= S_IRUGO|S_IWUSR,
L
Linus Torvalds 已提交
114 115 116 117 118 119 120
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= DEV_RAID_SPEED_LIMIT_MAX,
		.procname	= "speed_limit_max",
		.data		= &sysctl_speed_limit_max,
		.maxlen		= sizeof(int),
121
		.mode		= S_IRUGO|S_IWUSR,
L
Linus Torvalds 已提交
122 123 124 125 126 127 128 129 130 131
		.proc_handler	= &proc_dointvec,
	},
	{ .ctl_name = 0 }
};

static ctl_table raid_dir_table[] = {
	{
		.ctl_name	= DEV_RAID,
		.procname	= "raid",
		.maxlen		= 0,
132
		.mode		= S_IRUGO|S_IXUGO,
L
Linus Torvalds 已提交
133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
		.child		= raid_table,
	},
	{ .ctl_name = 0 }
};

static ctl_table raid_root_table[] = {
	{
		.ctl_name	= CTL_DEV,
		.procname	= "dev",
		.maxlen		= 0,
		.mode		= 0555,
		.child		= raid_dir_table,
	},
	{ .ctl_name = 0 }
};

static struct block_device_operations md_fops;

151 152
static int start_readonly;

153 154 155 156 157 158 159 160 161 162
/*
 * 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
 */
163
static DECLARE_WAIT_QUEUE_HEAD(md_event_waiters);
164
static atomic_t md_event_count;
165
void md_new_event(mddev_t *mddev)
166 167 168
{
	atomic_inc(&md_event_count);
	wake_up(&md_event_waiters);
169
	sysfs_notify(&mddev->kobj, NULL, "sync_action");
170
}
171
EXPORT_SYMBOL_GPL(md_new_event);
172

173 174 175
/* Alternate version that can be called from interrupts
 * when calling sysfs_notify isn't needed.
 */
176
static void md_new_event_inintr(mddev_t *mddev)
177 178 179 180 181
{
	atomic_inc(&md_event_count);
	wake_up(&md_event_waiters);
}

L
Linus Torvalds 已提交
182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230
/*
 * Enables to iterate over all existing md arrays
 * all_mddevs_lock protects this list.
 */
static LIST_HEAD(all_mddevs);
static DEFINE_SPINLOCK(all_mddevs_lock);


/*
 * iterates through all used mddevs in the system.
 * We take care to grab the all_mddevs_lock whenever navigating
 * the list, and to always hold a refcount when unlocked.
 * Any code which breaks out of this loop while own
 * a reference to the current mddev and must mddev_put it.
 */
#define ITERATE_MDDEV(mddev,tmp)					\
									\
	for (({ spin_lock(&all_mddevs_lock); 				\
		tmp = all_mddevs.next;					\
		mddev = NULL;});					\
	     ({ if (tmp != &all_mddevs)					\
			mddev_get(list_entry(tmp, mddev_t, all_mddevs));\
		spin_unlock(&all_mddevs_lock);				\
		if (mddev) mddev_put(mddev);				\
		mddev = list_entry(tmp, mddev_t, all_mddevs);		\
		tmp != &all_mddevs;});					\
	     ({ spin_lock(&all_mddevs_lock);				\
		tmp = tmp->next;})					\
		)


static int md_fail_request (request_queue_t *q, struct bio *bio)
{
	bio_io_error(bio, bio->bi_size);
	return 0;
}

static inline mddev_t *mddev_get(mddev_t *mddev)
{
	atomic_inc(&mddev->active);
	return mddev;
}

static void mddev_put(mddev_t *mddev)
{
	if (!atomic_dec_and_lock(&mddev->active, &all_mddevs_lock))
		return;
	if (!mddev->raid_disks && list_empty(&mddev->disks)) {
		list_del(&mddev->all_mddevs);
231
		spin_unlock(&all_mddevs_lock);
232
		blk_cleanup_queue(mddev->queue);
233
		kobject_unregister(&mddev->kobj);
234 235
	} else
		spin_unlock(&all_mddevs_lock);
L
Linus Torvalds 已提交
236 237 238 239 240 241 242 243 244 245 246 247
}

static mddev_t * mddev_find(dev_t unit)
{
	mddev_t *mddev, *new = NULL;

 retry:
	spin_lock(&all_mddevs_lock);
	list_for_each_entry(mddev, &all_mddevs, all_mddevs)
		if (mddev->unit == unit) {
			mddev_get(mddev);
			spin_unlock(&all_mddevs_lock);
248
			kfree(new);
L
Linus Torvalds 已提交
249 250 251 252 253 254 255 256 257 258
			return mddev;
		}

	if (new) {
		list_add(&new->all_mddevs, &all_mddevs);
		spin_unlock(&all_mddevs_lock);
		return new;
	}
	spin_unlock(&all_mddevs_lock);

259
	new = kzalloc(sizeof(*new), GFP_KERNEL);
L
Linus Torvalds 已提交
260 261 262 263 264 265 266 267 268
	if (!new)
		return NULL;

	new->unit = unit;
	if (MAJOR(unit) == MD_MAJOR)
		new->md_minor = MINOR(unit);
	else
		new->md_minor = MINOR(unit) >> MdpMinorShift;

269
	mutex_init(&new->reconfig_mutex);
L
Linus Torvalds 已提交
270 271 272 273
	INIT_LIST_HEAD(&new->disks);
	INIT_LIST_HEAD(&new->all_mddevs);
	init_timer(&new->safemode_timer);
	atomic_set(&new->active, 1);
274
	spin_lock_init(&new->write_lock);
275
	init_waitqueue_head(&new->sb_wait);
L
Linus Torvalds 已提交
276 277 278 279 280 281

	new->queue = blk_alloc_queue(GFP_KERNEL);
	if (!new->queue) {
		kfree(new);
		return NULL;
	}
282
	set_bit(QUEUE_FLAG_CLUSTER, &new->queue->queue_flags);
L
Linus Torvalds 已提交
283 284 285 286 287 288 289 290

	blk_queue_make_request(new->queue, md_fail_request);

	goto retry;
}

static inline int mddev_lock(mddev_t * mddev)
{
291
	return mutex_lock_interruptible(&mddev->reconfig_mutex);
L
Linus Torvalds 已提交
292 293 294 295
}

static inline int mddev_trylock(mddev_t * mddev)
{
296
	return mutex_trylock(&mddev->reconfig_mutex);
L
Linus Torvalds 已提交
297 298 299 300
}

static inline void mddev_unlock(mddev_t * mddev)
{
301
	mutex_unlock(&mddev->reconfig_mutex);
L
Linus Torvalds 已提交
302

303
	md_wakeup_thread(mddev->thread);
L
Linus Torvalds 已提交
304 305
}

306
static mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr)
L
Linus Torvalds 已提交
307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329
{
	mdk_rdev_t * rdev;
	struct list_head *tmp;

	ITERATE_RDEV(mddev,rdev,tmp) {
		if (rdev->desc_nr == nr)
			return rdev;
	}
	return NULL;
}

static mdk_rdev_t * find_rdev(mddev_t * mddev, dev_t dev)
{
	struct list_head *tmp;
	mdk_rdev_t *rdev;

	ITERATE_RDEV(mddev,rdev,tmp) {
		if (rdev->bdev->bd_dev == dev)
			return rdev;
	}
	return NULL;
}

330
static struct mdk_personality *find_pers(int level, char *clevel)
331 332
{
	struct mdk_personality *pers;
333 334
	list_for_each_entry(pers, &pers_list, list) {
		if (level != LEVEL_NONE && pers->level == level)
335
			return pers;
336 337 338
		if (strcmp(pers->name, clevel)==0)
			return pers;
	}
339 340 341
	return NULL;
}

342
static inline sector_t calc_dev_sboffset(struct block_device *bdev)
L
Linus Torvalds 已提交
343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375
{
	sector_t size = bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
	return MD_NEW_SIZE_BLOCKS(size);
}

static sector_t calc_dev_size(mdk_rdev_t *rdev, unsigned chunk_size)
{
	sector_t size;

	size = rdev->sb_offset;

	if (chunk_size)
		size &= ~((sector_t)chunk_size/1024 - 1);
	return size;
}

static int alloc_disk_sb(mdk_rdev_t * rdev)
{
	if (rdev->sb_page)
		MD_BUG();

	rdev->sb_page = alloc_page(GFP_KERNEL);
	if (!rdev->sb_page) {
		printk(KERN_ALERT "md: out of memory.\n");
		return -EINVAL;
	}

	return 0;
}

static void free_disk_sb(mdk_rdev_t * rdev)
{
	if (rdev->sb_page) {
376
		put_page(rdev->sb_page);
L
Linus Torvalds 已提交
377 378 379 380 381 382 383 384
		rdev->sb_loaded = 0;
		rdev->sb_page = NULL;
		rdev->sb_offset = 0;
		rdev->size = 0;
	}
}


385 386 387
static int super_written(struct bio *bio, unsigned int bytes_done, int error)
{
	mdk_rdev_t *rdev = bio->bi_private;
388
	mddev_t *mddev = rdev->mddev;
389 390 391
	if (bio->bi_size)
		return 1;

392 393 394 395
	if (error || !test_bit(BIO_UPTODATE, &bio->bi_flags)) {
		printk("md: super_written gets error=%d, uptodate=%d\n",
		       error, test_bit(BIO_UPTODATE, &bio->bi_flags));
		WARN_ON(test_bit(BIO_UPTODATE, &bio->bi_flags));
396
		md_error(mddev, rdev);
397
	}
398

399 400
	if (atomic_dec_and_test(&mddev->pending_writes))
		wake_up(&mddev->sb_wait);
N
Neil Brown 已提交
401
	bio_put(bio);
402 403 404
	return 0;
}

405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431
static int super_written_barrier(struct bio *bio, unsigned int bytes_done, int error)
{
	struct bio *bio2 = bio->bi_private;
	mdk_rdev_t *rdev = bio2->bi_private;
	mddev_t *mddev = rdev->mddev;
	if (bio->bi_size)
		return 1;

	if (!test_bit(BIO_UPTODATE, &bio->bi_flags) &&
	    error == -EOPNOTSUPP) {
		unsigned long flags;
		/* barriers don't appear to be supported :-( */
		set_bit(BarriersNotsupp, &rdev->flags);
		mddev->barriers_work = 0;
		spin_lock_irqsave(&mddev->write_lock, flags);
		bio2->bi_next = mddev->biolist;
		mddev->biolist = bio2;
		spin_unlock_irqrestore(&mddev->write_lock, flags);
		wake_up(&mddev->sb_wait);
		bio_put(bio);
		return 0;
	}
	bio_put(bio2);
	bio->bi_private = rdev;
	return super_written(bio, bytes_done, error);
}

432 433 434 435 436 437 438 439
void md_super_write(mddev_t *mddev, mdk_rdev_t *rdev,
		   sector_t sector, int size, struct page *page)
{
	/* write first size bytes of page to sector of rdev
	 * Increment mddev->pending_writes before returning
	 * and decrement it on completion, waking up sb_wait
	 * if zero is reached.
	 * If an error occurred, call md_error
440 441 442
	 *
	 * As we might need to resubmit the request if BIO_RW_BARRIER
	 * causes ENOTSUPP, we allocate a spare bio...
443 444
	 */
	struct bio *bio = bio_alloc(GFP_NOIO, 1);
445
	int rw = (1<<BIO_RW) | (1<<BIO_RW_SYNC);
446 447 448 449 450 451

	bio->bi_bdev = rdev->bdev;
	bio->bi_sector = sector;
	bio_add_page(bio, page, size, 0);
	bio->bi_private = rdev;
	bio->bi_end_io = super_written;
452 453
	bio->bi_rw = rw;

454
	atomic_inc(&mddev->pending_writes);
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487
	if (!test_bit(BarriersNotsupp, &rdev->flags)) {
		struct bio *rbio;
		rw |= (1<<BIO_RW_BARRIER);
		rbio = bio_clone(bio, GFP_NOIO);
		rbio->bi_private = bio;
		rbio->bi_end_io = super_written_barrier;
		submit_bio(rw, rbio);
	} else
		submit_bio(rw, bio);
}

void md_super_wait(mddev_t *mddev)
{
	/* wait for all superblock writes that were scheduled to complete.
	 * if any had to be retried (due to BARRIER problems), retry them
	 */
	DEFINE_WAIT(wq);
	for(;;) {
		prepare_to_wait(&mddev->sb_wait, &wq, TASK_UNINTERRUPTIBLE);
		if (atomic_read(&mddev->pending_writes)==0)
			break;
		while (mddev->biolist) {
			struct bio *bio;
			spin_lock_irq(&mddev->write_lock);
			bio = mddev->biolist;
			mddev->biolist = bio->bi_next ;
			bio->bi_next = NULL;
			spin_unlock_irq(&mddev->write_lock);
			submit_bio(bio->bi_rw, bio);
		}
		schedule();
	}
	finish_wait(&mddev->sb_wait, &wq);
488 489
}

L
Linus Torvalds 已提交
490 491 492 493 494 495 496 497 498
static int bi_complete(struct bio *bio, unsigned int bytes_done, int error)
{
	if (bio->bi_size)
		return 1;

	complete((struct completion*)bio->bi_private);
	return 0;
}

499
int sync_page_io(struct block_device *bdev, sector_t sector, int size,
L
Linus Torvalds 已提交
500 501
		   struct page *page, int rw)
{
502
	struct bio *bio = bio_alloc(GFP_NOIO, 1);
L
Linus Torvalds 已提交
503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520
	struct completion event;
	int ret;

	rw |= (1 << BIO_RW_SYNC);

	bio->bi_bdev = bdev;
	bio->bi_sector = sector;
	bio_add_page(bio, page, size, 0);
	init_completion(&event);
	bio->bi_private = &event;
	bio->bi_end_io = bi_complete;
	submit_bio(rw, bio);
	wait_for_completion(&event);

	ret = test_bit(BIO_UPTODATE, &bio->bi_flags);
	bio_put(bio);
	return ret;
}
521
EXPORT_SYMBOL_GPL(sync_page_io);
L
Linus Torvalds 已提交
522

523
static int read_disk_sb(mdk_rdev_t * rdev, int size)
L
Linus Torvalds 已提交
524 525 526 527 528 529 530 531 532 533
{
	char b[BDEVNAME_SIZE];
	if (!rdev->sb_page) {
		MD_BUG();
		return -EINVAL;
	}
	if (rdev->sb_loaded)
		return 0;


534
	if (!sync_page_io(rdev->bdev, rdev->sb_offset<<1, size, rdev->sb_page, READ))
L
Linus Torvalds 已提交
535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
		goto fail;
	rdev->sb_loaded = 1;
	return 0;

fail:
	printk(KERN_WARNING "md: disabled device %s, could not read superblock.\n",
		bdevname(rdev->bdev,b));
	return -EINVAL;
}

static int uuid_equal(mdp_super_t *sb1, mdp_super_t *sb2)
{
	if (	(sb1->set_uuid0 == sb2->set_uuid0) &&
		(sb1->set_uuid1 == sb2->set_uuid1) &&
		(sb1->set_uuid2 == sb2->set_uuid2) &&
		(sb1->set_uuid3 == sb2->set_uuid3))

		return 1;

	return 0;
}


static int sb_equal(mdp_super_t *sb1, mdp_super_t *sb2)
{
	int ret;
	mdp_super_t *tmp1, *tmp2;

	tmp1 = kmalloc(sizeof(*tmp1),GFP_KERNEL);
	tmp2 = kmalloc(sizeof(*tmp2),GFP_KERNEL);

	if (!tmp1 || !tmp2) {
		ret = 0;
		printk(KERN_INFO "md.c: sb1 is not equal to sb2!\n");
		goto abort;
	}

	*tmp1 = *sb1;
	*tmp2 = *sb2;

	/*
	 * nr_disks is not constant
	 */
	tmp1->nr_disks = 0;
	tmp2->nr_disks = 0;

	if (memcmp(tmp1, tmp2, MD_SB_GENERIC_CONSTANT_WORDS * 4))
		ret = 0;
	else
		ret = 1;

abort:
587 588
	kfree(tmp1);
	kfree(tmp2);
L
Linus Torvalds 已提交
589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
	return ret;
}

static unsigned int calc_sb_csum(mdp_super_t * sb)
{
	unsigned int disk_csum, csum;

	disk_csum = sb->sb_csum;
	sb->sb_csum = 0;
	csum = csum_partial((void *)sb, MD_SB_BYTES, 0);
	sb->sb_csum = disk_csum;
	return csum;
}


/*
 * Handle superblock details.
 * We want to be able to handle multiple superblock formats
 * so we have a common interface to them all, and an array of
 * different handlers.
 * We rely on user-space to write the initial superblock, and support
 * reading and updating of superblocks.
 * Interface methods are:
 *   int load_super(mdk_rdev_t *dev, mdk_rdev_t *refdev, int minor_version)
 *      loads and validates a superblock on dev.
 *      if refdev != NULL, compare superblocks on both devices
 *    Return:
 *      0 - dev has a superblock that is compatible with refdev
 *      1 - dev has a superblock that is compatible and newer than refdev
 *          so dev should be used as the refdev in future
 *     -EINVAL superblock incompatible or invalid
 *     -othererror e.g. -EIO
 *
 *   int validate_super(mddev_t *mddev, mdk_rdev_t *dev)
 *      Verify that dev is acceptable into mddev.
 *       The first time, mddev->raid_disks will be 0, and data from
 *       dev should be merged in.  Subsequent calls check that dev
 *       is new enough.  Return 0 or -EINVAL
 *
 *   void sync_super(mddev_t *mddev, mdk_rdev_t *dev)
 *     Update the superblock for rdev with data in mddev
 *     This does not write to disc.
 *
 */

struct super_type  {
	char 		*name;
	struct module	*owner;
	int		(*load_super)(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version);
	int		(*validate_super)(mddev_t *mddev, mdk_rdev_t *rdev);
	void		(*sync_super)(mddev_t *mddev, mdk_rdev_t *rdev);
};

/*
 * load_super for 0.90.0 
 */
static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version)
{
	char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
	mdp_super_t *sb;
	int ret;
	sector_t sb_offset;

	/*
	 * Calculate the position of the superblock,
	 * it's at the end of the disk.
	 *
	 * It also happens to be a multiple of 4Kb.
	 */
	sb_offset = calc_dev_sboffset(rdev->bdev);
	rdev->sb_offset = sb_offset;

661
	ret = read_disk_sb(rdev, MD_SB_BYTES);
L
Linus Torvalds 已提交
662 663 664 665 666 667 668 669 670 671 672 673 674 675
	if (ret) return ret;

	ret = -EINVAL;

	bdevname(rdev->bdev, b);
	sb = (mdp_super_t*)page_address(rdev->sb_page);

	if (sb->md_magic != MD_SB_MAGIC) {
		printk(KERN_ERR "md: invalid raid superblock magic on %s\n",
		       b);
		goto abort;
	}

	if (sb->major_version != 0 ||
676 677
	    sb->minor_version < 90 ||
	    sb->minor_version > 91) {
L
Linus Torvalds 已提交
678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694
		printk(KERN_WARNING "Bad version number %d.%d on %s\n",
			sb->major_version, sb->minor_version,
			b);
		goto abort;
	}

	if (sb->raid_disks <= 0)
		goto abort;

	if (csum_fold(calc_sb_csum(sb)) != csum_fold(sb->sb_csum)) {
		printk(KERN_WARNING "md: invalid superblock checksum on %s\n",
			b);
		goto abort;
	}

	rdev->preferred_minor = sb->md_minor;
	rdev->data_offset = 0;
695
	rdev->sb_size = MD_SB_BYTES;
L
Linus Torvalds 已提交
696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726

	if (sb->level == LEVEL_MULTIPATH)
		rdev->desc_nr = -1;
	else
		rdev->desc_nr = sb->this_disk.number;

	if (refdev == 0)
		ret = 1;
	else {
		__u64 ev1, ev2;
		mdp_super_t *refsb = (mdp_super_t*)page_address(refdev->sb_page);
		if (!uuid_equal(refsb, sb)) {
			printk(KERN_WARNING "md: %s has different UUID to %s\n",
				b, bdevname(refdev->bdev,b2));
			goto abort;
		}
		if (!sb_equal(refsb, sb)) {
			printk(KERN_WARNING "md: %s has same UUID"
			       " but different superblock to %s\n",
			       b, bdevname(refdev->bdev, b2));
			goto abort;
		}
		ev1 = md_event(sb);
		ev2 = md_event(refsb);
		if (ev1 > ev2)
			ret = 1;
		else 
			ret = 0;
	}
	rdev->size = calc_dev_size(rdev, sb->chunk_size);

727 728 729 730
	if (rdev->size < sb->size && sb->level > 1)
		/* "this cannot possibly happen" ... */
		ret = -EINVAL;

L
Linus Torvalds 已提交
731 732 733 734 735 736 737 738 739 740 741
 abort:
	return ret;
}

/*
 * validate_super for 0.90.0
 */
static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
{
	mdp_disk_t *desc;
	mdp_super_t *sb = (mdp_super_t *)page_address(rdev->sb_page);
742
	__u64 ev1 = md_event(sb);
L
Linus Torvalds 已提交
743

744
	rdev->raid_disk = -1;
745
	rdev->flags = 0;
L
Linus Torvalds 已提交
746 747 748 749 750 751 752 753 754
	if (mddev->raid_disks == 0) {
		mddev->major_version = 0;
		mddev->minor_version = sb->minor_version;
		mddev->patch_version = sb->patch_version;
		mddev->persistent = ! sb->not_persistent;
		mddev->chunk_size = sb->chunk_size;
		mddev->ctime = sb->ctime;
		mddev->utime = sb->utime;
		mddev->level = sb->level;
755
		mddev->clevel[0] = 0;
L
Linus Torvalds 已提交
756 757 758
		mddev->layout = sb->layout;
		mddev->raid_disks = sb->raid_disks;
		mddev->size = sb->size;
759
		mddev->events = ev1;
760
		mddev->bitmap_offset = 0;
761
		mddev->default_bitmap_offset = MD_SB_BYTES >> 9;
L
Linus Torvalds 已提交
762

763 764 765 766 767 768 769 770 771 772 773 774 775 776
		if (mddev->minor_version >= 91) {
			mddev->reshape_position = sb->reshape_position;
			mddev->delta_disks = sb->delta_disks;
			mddev->new_level = sb->new_level;
			mddev->new_layout = sb->new_layout;
			mddev->new_chunk = sb->new_chunk;
		} else {
			mddev->reshape_position = MaxSector;
			mddev->delta_disks = 0;
			mddev->new_level = mddev->level;
			mddev->new_layout = mddev->layout;
			mddev->new_chunk = mddev->chunk_size;
		}

L
Linus Torvalds 已提交
777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792
		if (sb->state & (1<<MD_SB_CLEAN))
			mddev->recovery_cp = MaxSector;
		else {
			if (sb->events_hi == sb->cp_events_hi && 
				sb->events_lo == sb->cp_events_lo) {
				mddev->recovery_cp = sb->recovery_cp;
			} else
				mddev->recovery_cp = 0;
		}

		memcpy(mddev->uuid+0, &sb->set_uuid0, 4);
		memcpy(mddev->uuid+4, &sb->set_uuid1, 4);
		memcpy(mddev->uuid+8, &sb->set_uuid2, 4);
		memcpy(mddev->uuid+12,&sb->set_uuid3, 4);

		mddev->max_disks = MD_SB_DISKS;
793 794 795

		if (sb->state & (1<<MD_SB_BITMAP_PRESENT) &&
		    mddev->bitmap_file == NULL) {
796 797
			if (mddev->level != 1 && mddev->level != 4
			    && mddev->level != 5 && mddev->level != 6
798
			    && mddev->level != 10) {
799
				/* FIXME use a better test */
800
				printk(KERN_WARNING "md: bitmaps not supported for this level.\n");
801 802
				return -EINVAL;
			}
803
			mddev->bitmap_offset = mddev->default_bitmap_offset;
804 805
		}

806 807
	} else if (mddev->pers == NULL) {
		/* Insist on good event counter while assembling */
L
Linus Torvalds 已提交
808 809 810
		++ev1;
		if (ev1 < mddev->events) 
			return -EINVAL;
811 812 813 814 815 816
	} else if (mddev->bitmap) {
		/* if adding to array with a bitmap, then we can accept an
		 * older device ... but not too old.
		 */
		if (ev1 < mddev->bitmap->events_cleared)
			return 0;
817 818 819 820 821
	} else {
		if (ev1 < mddev->events)
			/* just a hot-add of a new device, leave raid_disk at -1 */
			return 0;
	}
822

L
Linus Torvalds 已提交
823 824 825 826
	if (mddev->level != LEVEL_MULTIPATH) {
		desc = sb->disks + rdev->desc_nr;

		if (desc->state & (1<<MD_DISK_FAULTY))
827
			set_bit(Faulty, &rdev->flags);
828 829
		else if (desc->state & (1<<MD_DISK_SYNC) /* &&
			    desc->raid_disk < mddev->raid_disks */) {
830
			set_bit(In_sync, &rdev->flags);
L
Linus Torvalds 已提交
831 832
			rdev->raid_disk = desc->raid_disk;
		}
833 834
		if (desc->state & (1<<MD_DISK_WRITEMOSTLY))
			set_bit(WriteMostly, &rdev->flags);
835
	} else /* MULTIPATH are always insync */
836
		set_bit(In_sync, &rdev->flags);
L
Linus Torvalds 已提交
837 838 839 840 841 842 843 844 845 846 847 848
	return 0;
}

/*
 * sync_super for 0.90.0
 */
static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
{
	mdp_super_t *sb;
	struct list_head *tmp;
	mdk_rdev_t *rdev2;
	int next_spare = mddev->raid_disks;
849

L
Linus Torvalds 已提交
850 851 852 853 854 855 856 857 858 859 860 861 862 863

	/* make rdev->sb match mddev data..
	 *
	 * 1/ zero out disks
	 * 2/ Add info for each disk, keeping track of highest desc_nr (next_spare);
	 * 3/ any empty disks < next_spare become removed
	 *
	 * disks[0] gets initialised to REMOVED because
	 * we cannot be sure from other fields if it has
	 * been initialised or not.
	 */
	int i;
	int active=0, working=0,failed=0,spare=0,nr_disks=0;

864 865
	rdev->sb_size = MD_SB_BYTES;

L
Linus Torvalds 已提交
866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889
	sb = (mdp_super_t*)page_address(rdev->sb_page);

	memset(sb, 0, sizeof(*sb));

	sb->md_magic = MD_SB_MAGIC;
	sb->major_version = mddev->major_version;
	sb->patch_version = mddev->patch_version;
	sb->gvalid_words  = 0; /* ignored */
	memcpy(&sb->set_uuid0, mddev->uuid+0, 4);
	memcpy(&sb->set_uuid1, mddev->uuid+4, 4);
	memcpy(&sb->set_uuid2, mddev->uuid+8, 4);
	memcpy(&sb->set_uuid3, mddev->uuid+12,4);

	sb->ctime = mddev->ctime;
	sb->level = mddev->level;
	sb->size  = mddev->size;
	sb->raid_disks = mddev->raid_disks;
	sb->md_minor = mddev->md_minor;
	sb->not_persistent = !mddev->persistent;
	sb->utime = mddev->utime;
	sb->state = 0;
	sb->events_hi = (mddev->events>>32);
	sb->events_lo = (u32)mddev->events;

890 891 892 893 894 895 896 897 898 899 900
	if (mddev->reshape_position == MaxSector)
		sb->minor_version = 90;
	else {
		sb->minor_version = 91;
		sb->reshape_position = mddev->reshape_position;
		sb->new_level = mddev->new_level;
		sb->delta_disks = mddev->delta_disks;
		sb->new_layout = mddev->new_layout;
		sb->new_chunk = mddev->new_chunk;
	}
	mddev->minor_version = sb->minor_version;
L
Linus Torvalds 已提交
901 902 903 904 905 906 907 908 909 910 911 912 913
	if (mddev->in_sync)
	{
		sb->recovery_cp = mddev->recovery_cp;
		sb->cp_events_hi = (mddev->events>>32);
		sb->cp_events_lo = (u32)mddev->events;
		if (mddev->recovery_cp == MaxSector)
			sb->state = (1<< MD_SB_CLEAN);
	} else
		sb->recovery_cp = 0;

	sb->layout = mddev->layout;
	sb->chunk_size = mddev->chunk_size;

914 915 916
	if (mddev->bitmap && mddev->bitmap_file == NULL)
		sb->state |= (1<<MD_SB_BITMAP_PRESENT);

L
Linus Torvalds 已提交
917 918 919
	sb->disks[0].state = (1<<MD_DISK_REMOVED);
	ITERATE_RDEV(mddev,rdev2,tmp) {
		mdp_disk_t *d;
920
		int desc_nr;
921 922
		if (rdev2->raid_disk >= 0 && test_bit(In_sync, &rdev2->flags)
		    && !test_bit(Faulty, &rdev2->flags))
923
			desc_nr = rdev2->raid_disk;
L
Linus Torvalds 已提交
924
		else
925
			desc_nr = next_spare++;
926
		rdev2->desc_nr = desc_nr;
L
Linus Torvalds 已提交
927 928 929 930 931
		d = &sb->disks[rdev2->desc_nr];
		nr_disks++;
		d->number = rdev2->desc_nr;
		d->major = MAJOR(rdev2->bdev->bd_dev);
		d->minor = MINOR(rdev2->bdev->bd_dev);
932 933
		if (rdev2->raid_disk >= 0 && test_bit(In_sync, &rdev2->flags)
		    && !test_bit(Faulty, &rdev2->flags))
L
Linus Torvalds 已提交
934 935 936
			d->raid_disk = rdev2->raid_disk;
		else
			d->raid_disk = rdev2->desc_nr; /* compatibility */
937
		if (test_bit(Faulty, &rdev2->flags))
L
Linus Torvalds 已提交
938
			d->state = (1<<MD_DISK_FAULTY);
939
		else if (test_bit(In_sync, &rdev2->flags)) {
L
Linus Torvalds 已提交
940 941 942 943 944 945 946 947 948
			d->state = (1<<MD_DISK_ACTIVE);
			d->state |= (1<<MD_DISK_SYNC);
			active++;
			working++;
		} else {
			d->state = 0;
			spare++;
			working++;
		}
949 950
		if (test_bit(WriteMostly, &rdev2->flags))
			d->state |= (1<<MD_DISK_WRITEMOSTLY);
L
Linus Torvalds 已提交
951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004
	}
	/* now set the "removed" and "faulty" bits on any missing devices */
	for (i=0 ; i < mddev->raid_disks ; i++) {
		mdp_disk_t *d = &sb->disks[i];
		if (d->state == 0 && d->number == 0) {
			d->number = i;
			d->raid_disk = i;
			d->state = (1<<MD_DISK_REMOVED);
			d->state |= (1<<MD_DISK_FAULTY);
			failed++;
		}
	}
	sb->nr_disks = nr_disks;
	sb->active_disks = active;
	sb->working_disks = working;
	sb->failed_disks = failed;
	sb->spare_disks = spare;

	sb->this_disk = sb->disks[rdev->desc_nr];
	sb->sb_csum = calc_sb_csum(sb);
}

/*
 * version 1 superblock
 */

static unsigned int calc_sb_1_csum(struct mdp_superblock_1 * sb)
{
	unsigned int disk_csum, csum;
	unsigned long long newcsum;
	int size = 256 + le32_to_cpu(sb->max_dev)*2;
	unsigned int *isuper = (unsigned int*)sb;
	int i;

	disk_csum = sb->sb_csum;
	sb->sb_csum = 0;
	newcsum = 0;
	for (i=0; size>=4; size -= 4 )
		newcsum += le32_to_cpu(*isuper++);

	if (size == 2)
		newcsum += le16_to_cpu(*(unsigned short*) isuper);

	csum = (newcsum & 0xffffffff) + (newcsum >> 32);
	sb->sb_csum = disk_csum;
	return cpu_to_le32(csum);
}

static int super_1_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version)
{
	struct mdp_superblock_1 *sb;
	int ret;
	sector_t sb_offset;
	char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
1005
	int bmask;
L
Linus Torvalds 已提交
1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018

	/*
	 * Calculate the position of the superblock.
	 * It is always aligned to a 4K boundary and
	 * depeding on minor_version, it can be:
	 * 0: At least 8K, but less than 12K, from end of device
	 * 1: At start of device
	 * 2: 4K from start of device.
	 */
	switch(minor_version) {
	case 0:
		sb_offset = rdev->bdev->bd_inode->i_size >> 9;
		sb_offset -= 8*2;
1019
		sb_offset &= ~(sector_t)(4*2-1);
L
Linus Torvalds 已提交
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033
		/* convert from sectors to K */
		sb_offset /= 2;
		break;
	case 1:
		sb_offset = 0;
		break;
	case 2:
		sb_offset = 4;
		break;
	default:
		return -EINVAL;
	}
	rdev->sb_offset = sb_offset;

1034 1035 1036 1037
	/* superblock is rarely larger than 1K, but it can be larger,
	 * and it is safe to read 4k, so we do that
	 */
	ret = read_disk_sb(rdev, 4096);
L
Linus Torvalds 已提交
1038 1039 1040 1041 1042 1043 1044 1045 1046
	if (ret) return ret;


	sb = (struct mdp_superblock_1*)page_address(rdev->sb_page);

	if (sb->magic != cpu_to_le32(MD_SB_MAGIC) ||
	    sb->major_version != cpu_to_le32(1) ||
	    le32_to_cpu(sb->max_dev) > (4096-256)/2 ||
	    le64_to_cpu(sb->super_offset) != (rdev->sb_offset<<1) ||
1047
	    (le32_to_cpu(sb->feature_map) & ~MD_FEATURE_ALL) != 0)
L
Linus Torvalds 已提交
1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061
		return -EINVAL;

	if (calc_sb_1_csum(sb) != sb->sb_csum) {
		printk("md: invalid superblock checksum on %s\n",
			bdevname(rdev->bdev,b));
		return -EINVAL;
	}
	if (le64_to_cpu(sb->data_size) < 10) {
		printk("md: data_size too small on %s\n",
		       bdevname(rdev->bdev,b));
		return -EINVAL;
	}
	rdev->preferred_minor = 0xffff;
	rdev->data_offset = le64_to_cpu(sb->data_offset);
1062
	atomic_set(&rdev->corrected_errors, le32_to_cpu(sb->cnt_corrected_read));
L
Linus Torvalds 已提交
1063

1064
	rdev->sb_size = le32_to_cpu(sb->max_dev) * 2 + 256;
1065
	bmask = queue_hardsect_size(rdev->bdev->bd_disk->queue)-1;
1066 1067 1068
	if (rdev->sb_size & bmask)
		rdev-> sb_size = (rdev->sb_size | bmask)+1;

1069 1070 1071 1072 1073
	if (sb->level == cpu_to_le32(LEVEL_MULTIPATH))
		rdev->desc_nr = -1;
	else
		rdev->desc_nr = le32_to_cpu(sb->dev_number);

L
Linus Torvalds 已提交
1074
	if (refdev == 0)
1075
		ret = 1;
L
Linus Torvalds 已提交
1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094
	else {
		__u64 ev1, ev2;
		struct mdp_superblock_1 *refsb = 
			(struct mdp_superblock_1*)page_address(refdev->sb_page);

		if (memcmp(sb->set_uuid, refsb->set_uuid, 16) != 0 ||
		    sb->level != refsb->level ||
		    sb->layout != refsb->layout ||
		    sb->chunksize != refsb->chunksize) {
			printk(KERN_WARNING "md: %s has strangely different"
				" superblock to %s\n",
				bdevname(rdev->bdev,b),
				bdevname(refdev->bdev,b2));
			return -EINVAL;
		}
		ev1 = le64_to_cpu(sb->events);
		ev2 = le64_to_cpu(refsb->events);

		if (ev1 > ev2)
1095 1096 1097
			ret = 1;
		else
			ret = 0;
L
Linus Torvalds 已提交
1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
	}
	if (minor_version) 
		rdev->size = ((rdev->bdev->bd_inode->i_size>>9) - le64_to_cpu(sb->data_offset)) / 2;
	else
		rdev->size = rdev->sb_offset;
	if (rdev->size < le64_to_cpu(sb->data_size)/2)
		return -EINVAL;
	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);
1108 1109 1110

	if (le32_to_cpu(sb->size) > rdev->size*2)
		return -EINVAL;
1111
	return ret;
L
Linus Torvalds 已提交
1112 1113 1114 1115 1116
}

static int super_1_validate(mddev_t *mddev, mdk_rdev_t *rdev)
{
	struct mdp_superblock_1 *sb = (struct mdp_superblock_1*)page_address(rdev->sb_page);
1117
	__u64 ev1 = le64_to_cpu(sb->events);
L
Linus Torvalds 已提交
1118

1119
	rdev->raid_disk = -1;
1120
	rdev->flags = 0;
L
Linus Torvalds 已提交
1121 1122 1123 1124 1125 1126 1127 1128
	if (mddev->raid_disks == 0) {
		mddev->major_version = 1;
		mddev->patch_version = 0;
		mddev->persistent = 1;
		mddev->chunk_size = le32_to_cpu(sb->chunksize) << 9;
		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);
1129
		mddev->clevel[0] = 0;
L
Linus Torvalds 已提交
1130 1131 1132
		mddev->layout = le32_to_cpu(sb->layout);
		mddev->raid_disks = le32_to_cpu(sb->raid_disks);
		mddev->size = le64_to_cpu(sb->size)/2;
1133
		mddev->events = ev1;
1134
		mddev->bitmap_offset = 0;
N
NeilBrown 已提交
1135
		mddev->default_bitmap_offset = 1024 >> 9;
L
Linus Torvalds 已提交
1136 1137 1138 1139 1140
		
		mddev->recovery_cp = le64_to_cpu(sb->resync_offset);
		memcpy(mddev->uuid, sb->set_uuid, 16);

		mddev->max_disks =  (4096-256)/2;
1141

1142
		if ((le32_to_cpu(sb->feature_map) & MD_FEATURE_BITMAP_OFFSET) &&
1143
		    mddev->bitmap_file == NULL ) {
1144 1145 1146
			if (mddev->level != 1 && mddev->level != 5 && mddev->level != 6
			    && mddev->level != 10) {
				printk(KERN_WARNING "md: bitmaps not supported for this level.\n");
1147 1148 1149 1150
				return -EINVAL;
			}
			mddev->bitmap_offset = (__s32)le32_to_cpu(sb->bitmap_offset);
		}
1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164
		if ((le32_to_cpu(sb->feature_map) & MD_FEATURE_RESHAPE_ACTIVE)) {
			mddev->reshape_position = le64_to_cpu(sb->reshape_position);
			mddev->delta_disks = le32_to_cpu(sb->delta_disks);
			mddev->new_level = le32_to_cpu(sb->new_level);
			mddev->new_layout = le32_to_cpu(sb->new_layout);
			mddev->new_chunk = le32_to_cpu(sb->new_chunk)<<9;
		} else {
			mddev->reshape_position = MaxSector;
			mddev->delta_disks = 0;
			mddev->new_level = mddev->level;
			mddev->new_layout = mddev->layout;
			mddev->new_chunk = mddev->chunk_size;
		}

1165 1166
	} else if (mddev->pers == NULL) {
		/* Insist of good event counter while assembling */
L
Linus Torvalds 已提交
1167 1168 1169
		++ev1;
		if (ev1 < mddev->events)
			return -EINVAL;
1170 1171 1172 1173 1174 1175
	} else if (mddev->bitmap) {
		/* If adding to array with a bitmap, then we can accept an
		 * older device, but not too old.
		 */
		if (ev1 < mddev->bitmap->events_cleared)
			return 0;
1176 1177 1178 1179 1180
	} else {
		if (ev1 < mddev->events)
			/* just a hot-add of a new device, leave raid_disk at -1 */
			return 0;
	}
L
Linus Torvalds 已提交
1181 1182 1183 1184 1185 1186 1187
	if (mddev->level != LEVEL_MULTIPATH) {
		int role;
		role = le16_to_cpu(sb->dev_roles[rdev->desc_nr]);
		switch(role) {
		case 0xffff: /* spare */
			break;
		case 0xfffe: /* faulty */
1188
			set_bit(Faulty, &rdev->flags);
L
Linus Torvalds 已提交
1189 1190
			break;
		default:
1191 1192 1193 1194 1195
			if ((le32_to_cpu(sb->feature_map) &
			     MD_FEATURE_RECOVERY_OFFSET))
				rdev->recovery_offset = le64_to_cpu(sb->recovery_offset);
			else
				set_bit(In_sync, &rdev->flags);
L
Linus Torvalds 已提交
1196 1197 1198
			rdev->raid_disk = role;
			break;
		}
1199 1200
		if (sb->devflags & WriteMostly1)
			set_bit(WriteMostly, &rdev->flags);
1201
	} else /* MULTIPATH are always insync */
1202
		set_bit(In_sync, &rdev->flags);
1203

L
Linus Torvalds 已提交
1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218
	return 0;
}

static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
{
	struct mdp_superblock_1 *sb;
	struct list_head *tmp;
	mdk_rdev_t *rdev2;
	int max_dev, i;
	/* make rdev->sb match mddev and rdev data. */

	sb = (struct mdp_superblock_1*)page_address(rdev->sb_page);

	sb->feature_map = 0;
	sb->pad0 = 0;
1219
	sb->recovery_offset = cpu_to_le64(0);
L
Linus Torvalds 已提交
1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230
	memset(sb->pad1, 0, sizeof(sb->pad1));
	memset(sb->pad2, 0, sizeof(sb->pad2));
	memset(sb->pad3, 0, sizeof(sb->pad3));

	sb->utime = cpu_to_le64((__u64)mddev->utime);
	sb->events = cpu_to_le64(mddev->events);
	if (mddev->in_sync)
		sb->resync_offset = cpu_to_le64(mddev->recovery_cp);
	else
		sb->resync_offset = cpu_to_le64(0);

1231 1232
	sb->cnt_corrected_read = atomic_read(&rdev->corrected_errors);

1233
	sb->raid_disks = cpu_to_le32(mddev->raid_disks);
N
NeilBrown 已提交
1234
	sb->size = cpu_to_le64(mddev->size<<1);
1235

1236 1237
	if (mddev->bitmap && mddev->bitmap_file == NULL) {
		sb->bitmap_offset = cpu_to_le32((__u32)mddev->bitmap_offset);
1238
		sb->feature_map = cpu_to_le32(MD_FEATURE_BITMAP_OFFSET);
1239
	}
1240 1241 1242 1243 1244 1245 1246 1247

	if (rdev->raid_disk >= 0 &&
	    !test_bit(In_sync, &rdev->flags) &&
	    rdev->recovery_offset > 0) {
		sb->feature_map |= cpu_to_le32(MD_FEATURE_RECOVERY_OFFSET);
		sb->recovery_offset = cpu_to_le64(rdev->recovery_offset);
	}

1248 1249 1250 1251 1252 1253 1254 1255
	if (mddev->reshape_position != MaxSector) {
		sb->feature_map |= cpu_to_le32(MD_FEATURE_RESHAPE_ACTIVE);
		sb->reshape_position = cpu_to_le64(mddev->reshape_position);
		sb->new_layout = cpu_to_le32(mddev->new_layout);
		sb->delta_disks = cpu_to_le32(mddev->delta_disks);
		sb->new_level = cpu_to_le32(mddev->new_level);
		sb->new_chunk = cpu_to_le32(mddev->new_chunk>>9);
	}
1256

L
Linus Torvalds 已提交
1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267
	max_dev = 0;
	ITERATE_RDEV(mddev,rdev2,tmp)
		if (rdev2->desc_nr+1 > max_dev)
			max_dev = rdev2->desc_nr+1;
	
	sb->max_dev = cpu_to_le32(max_dev);
	for (i=0; i<max_dev;i++)
		sb->dev_roles[i] = cpu_to_le16(0xfffe);
	
	ITERATE_RDEV(mddev,rdev2,tmp) {
		i = rdev2->desc_nr;
1268
		if (test_bit(Faulty, &rdev2->flags))
L
Linus Torvalds 已提交
1269
			sb->dev_roles[i] = cpu_to_le16(0xfffe);
1270
		else if (test_bit(In_sync, &rdev2->flags))
L
Linus Torvalds 已提交
1271
			sb->dev_roles[i] = cpu_to_le16(rdev2->raid_disk);
1272 1273
		else if (rdev2->raid_disk >= 0 && rdev2->recovery_offset > 0)
			sb->dev_roles[i] = cpu_to_le16(rdev2->raid_disk);
L
Linus Torvalds 已提交
1274 1275 1276 1277 1278 1279 1280 1281
		else
			sb->dev_roles[i] = cpu_to_le16(0xffff);
	}

	sb->sb_csum = calc_sb_1_csum(sb);
}


A
Adrian Bunk 已提交
1282
static struct super_type super_types[] = {
L
Linus Torvalds 已提交
1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328
	[0] = {
		.name	= "0.90.0",
		.owner	= THIS_MODULE,
		.load_super	= super_90_load,
		.validate_super	= super_90_validate,
		.sync_super	= super_90_sync,
	},
	[1] = {
		.name	= "md-1",
		.owner	= THIS_MODULE,
		.load_super	= super_1_load,
		.validate_super	= super_1_validate,
		.sync_super	= super_1_sync,
	},
};
	
static mdk_rdev_t * match_dev_unit(mddev_t *mddev, mdk_rdev_t *dev)
{
	struct list_head *tmp;
	mdk_rdev_t *rdev;

	ITERATE_RDEV(mddev,rdev,tmp)
		if (rdev->bdev->bd_contains == dev->bdev->bd_contains)
			return rdev;

	return NULL;
}

static int match_mddev_units(mddev_t *mddev1, mddev_t *mddev2)
{
	struct list_head *tmp;
	mdk_rdev_t *rdev;

	ITERATE_RDEV(mddev1,rdev,tmp)
		if (match_dev_unit(mddev2, rdev))
			return 1;

	return 0;
}

static LIST_HEAD(pending_raid_disks);

static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
{
	mdk_rdev_t *same_pdev;
	char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
1329
	struct kobject *ko;
1330
	char *s;
L
Linus Torvalds 已提交
1331 1332 1333 1334 1335

	if (rdev->mddev) {
		MD_BUG();
		return -EINVAL;
	}
1336 1337 1338 1339 1340 1341 1342 1343
	/* 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;
	}
L
Linus Torvalds 已提交
1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366
	same_pdev = match_dev_unit(mddev, rdev);
	if (same_pdev)
		printk(KERN_WARNING
			"%s: WARNING: %s appears to be on the same physical"
	 		" disk as %s. True\n     protection against single-disk"
			" failure might be compromised.\n",
			mdname(mddev), bdevname(rdev->bdev,b),
			bdevname(same_pdev->bdev,b2));

	/* Verify rdev->desc_nr is unique.
	 * If it is -1, assign a free number, else
	 * check number is not in use
	 */
	if (rdev->desc_nr < 0) {
		int choice = 0;
		if (mddev->pers) choice = mddev->raid_disks;
		while (find_rdev_nr(mddev, choice))
			choice++;
		rdev->desc_nr = choice;
	} else {
		if (find_rdev_nr(mddev, rdev->desc_nr))
			return -EBUSY;
	}
1367 1368 1369
	bdevname(rdev->bdev,b);
	if (kobject_set_name(&rdev->kobj, "dev-%s", b) < 0)
		return -ENOMEM;
1370 1371
	while ( (s=strchr(rdev->kobj.k_name, '/')) != NULL)
		*s = '!';
L
Linus Torvalds 已提交
1372 1373 1374
			
	list_add(&rdev->same_set, &mddev->disks);
	rdev->mddev = mddev;
1375
	printk(KERN_INFO "md: bind<%s>\n", b);
1376

1377
	rdev->kobj.parent = &mddev->kobj;
1378 1379
	kobject_add(&rdev->kobj);

1380 1381 1382 1383 1384
	if (rdev->bdev->bd_part)
		ko = &rdev->bdev->bd_part->kobj;
	else
		ko = &rdev->bdev->bd_disk->kobj;
	sysfs_create_link(&rdev->kobj, ko, "block");
1385
	bd_claim_by_disk(rdev->bdev, rdev, mddev->gendisk);
L
Linus Torvalds 已提交
1386 1387 1388 1389 1390 1391 1392 1393 1394 1395
	return 0;
}

static void unbind_rdev_from_array(mdk_rdev_t * rdev)
{
	char b[BDEVNAME_SIZE];
	if (!rdev->mddev) {
		MD_BUG();
		return;
	}
1396
	bd_release_from_disk(rdev->bdev, rdev->mddev->gendisk);
L
Linus Torvalds 已提交
1397 1398 1399
	list_del_init(&rdev->same_set);
	printk(KERN_INFO "md: unbind<%s>\n", bdevname(rdev->bdev,b));
	rdev->mddev = NULL;
1400 1401
	sysfs_remove_link(&rdev->kobj, "block");
	kobject_del(&rdev->kobj);
L
Linus Torvalds 已提交
1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414
}

/*
 * prevent the device from being mounted, repartitioned or
 * otherwise reused by a RAID array (or any other kernel
 * subsystem), by bd_claiming the device.
 */
static int lock_rdev(mdk_rdev_t *rdev, dev_t dev)
{
	int err = 0;
	struct block_device *bdev;
	char b[BDEVNAME_SIZE];

1415
	bdev = open_partition_by_devnum(dev, FMODE_READ|FMODE_WRITE);
L
Linus Torvalds 已提交
1416 1417 1418 1419 1420 1421 1422 1423 1424
	if (IS_ERR(bdev)) {
		printk(KERN_ERR "md: could not open %s.\n",
			__bdevname(dev, b));
		return PTR_ERR(bdev);
	}
	err = bd_claim(bdev, rdev);
	if (err) {
		printk(KERN_ERR "md: could not bd_claim %s.\n",
			bdevname(bdev, b));
1425
		blkdev_put_partition(bdev);
L
Linus Torvalds 已提交
1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438
		return err;
	}
	rdev->bdev = bdev;
	return err;
}

static void unlock_rdev(mdk_rdev_t *rdev)
{
	struct block_device *bdev = rdev->bdev;
	rdev->bdev = NULL;
	if (!bdev)
		MD_BUG();
	bd_release(bdev);
1439
	blkdev_put_partition(bdev);
L
Linus Torvalds 已提交
1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456
}

void md_autodetect_dev(dev_t dev);

static void export_rdev(mdk_rdev_t * rdev)
{
	char b[BDEVNAME_SIZE];
	printk(KERN_INFO "md: export_rdev(%s)\n",
		bdevname(rdev->bdev,b));
	if (rdev->mddev)
		MD_BUG();
	free_disk_sb(rdev);
	list_del_init(&rdev->same_set);
#ifndef MODULE
	md_autodetect_dev(rdev->bdev->bd_dev);
#endif
	unlock_rdev(rdev);
1457
	kobject_put(&rdev->kobj);
L
Linus Torvalds 已提交
1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528
}

static void kick_rdev_from_array(mdk_rdev_t * rdev)
{
	unbind_rdev_from_array(rdev);
	export_rdev(rdev);
}

static void export_array(mddev_t *mddev)
{
	struct list_head *tmp;
	mdk_rdev_t *rdev;

	ITERATE_RDEV(mddev,rdev,tmp) {
		if (!rdev->mddev) {
			MD_BUG();
			continue;
		}
		kick_rdev_from_array(rdev);
	}
	if (!list_empty(&mddev->disks))
		MD_BUG();
	mddev->raid_disks = 0;
	mddev->major_version = 0;
}

static void print_desc(mdp_disk_t *desc)
{
	printk(" DISK<N:%d,(%d,%d),R:%d,S:%d>\n", desc->number,
		desc->major,desc->minor,desc->raid_disk,desc->state);
}

static void print_sb(mdp_super_t *sb)
{
	int i;

	printk(KERN_INFO 
		"md:  SB: (V:%d.%d.%d) ID:<%08x.%08x.%08x.%08x> CT:%08x\n",
		sb->major_version, sb->minor_version, sb->patch_version,
		sb->set_uuid0, sb->set_uuid1, sb->set_uuid2, sb->set_uuid3,
		sb->ctime);
	printk(KERN_INFO "md:     L%d S%08d ND:%d RD:%d md%d LO:%d CS:%d\n",
		sb->level, sb->size, sb->nr_disks, sb->raid_disks,
		sb->md_minor, sb->layout, sb->chunk_size);
	printk(KERN_INFO "md:     UT:%08x ST:%d AD:%d WD:%d"
		" FD:%d SD:%d CSUM:%08x E:%08lx\n",
		sb->utime, sb->state, sb->active_disks, sb->working_disks,
		sb->failed_disks, sb->spare_disks,
		sb->sb_csum, (unsigned long)sb->events_lo);

	printk(KERN_INFO);
	for (i = 0; i < MD_SB_DISKS; i++) {
		mdp_disk_t *desc;

		desc = sb->disks + i;
		if (desc->number || desc->major || desc->minor ||
		    desc->raid_disk || (desc->state && (desc->state != 4))) {
			printk("     D %2d: ", i);
			print_desc(desc);
		}
	}
	printk(KERN_INFO "md:     THIS: ");
	print_desc(&sb->this_disk);

}

static void print_rdev(mdk_rdev_t *rdev)
{
	char b[BDEVNAME_SIZE];
	printk(KERN_INFO "md: rdev %s, SZ:%08llu F:%d S:%d DN:%u\n",
		bdevname(rdev->bdev,b), (unsigned long long)rdev->size,
1529 1530
	        test_bit(Faulty, &rdev->flags), test_bit(In_sync, &rdev->flags),
	        rdev->desc_nr);
L
Linus Torvalds 已提交
1531 1532 1533 1534 1535 1536 1537
	if (rdev->sb_loaded) {
		printk(KERN_INFO "md: rdev superblock:\n");
		print_sb((mdp_super_t*)page_address(rdev->sb_page));
	} else
		printk(KERN_INFO "md: no rdev superblock!\n");
}

1538
static void md_print_devices(void)
L
Linus Torvalds 已提交
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550
{
	struct list_head *tmp, *tmp2;
	mdk_rdev_t *rdev;
	mddev_t *mddev;
	char b[BDEVNAME_SIZE];

	printk("\n");
	printk("md:	**********************************\n");
	printk("md:	* <COMPLETE RAID STATE PRINTOUT> *\n");
	printk("md:	**********************************\n");
	ITERATE_MDDEV(mddev,tmp) {

1551 1552 1553 1554
		if (mddev->bitmap)
			bitmap_print_sb(mddev->bitmap);
		else
			printk("%s: ", mdname(mddev));
L
Linus Torvalds 已提交
1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566
		ITERATE_RDEV(mddev,rdev,tmp2)
			printk("<%s>", bdevname(rdev->bdev,b));
		printk("\n");

		ITERATE_RDEV(mddev,rdev,tmp2)
			print_rdev(rdev);
	}
	printk("md:	**********************************\n");
	printk("\n");
}


1567
static void sync_sbs(mddev_t * mddev, int nospares)
L
Linus Torvalds 已提交
1568
{
1569 1570 1571 1572 1573 1574
	/* Update each superblock (in-memory image), but
	 * if we are allowed to, skip spares which already
	 * have the right event counter, or have one earlier
	 * (which would mean they aren't being marked as dirty
	 * with the rest of the array)
	 */
L
Linus Torvalds 已提交
1575 1576 1577 1578
	mdk_rdev_t *rdev;
	struct list_head *tmp;

	ITERATE_RDEV(mddev,rdev,tmp) {
1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590
		if (rdev->sb_events == mddev->events ||
		    (nospares &&
		     rdev->raid_disk < 0 &&
		     (rdev->sb_events&1)==0 &&
		     rdev->sb_events+1 == mddev->events)) {
			/* Don't update this superblock */
			rdev->sb_loaded = 2;
		} else {
			super_types[mddev->major_version].
				sync_super(mddev, rdev);
			rdev->sb_loaded = 1;
		}
L
Linus Torvalds 已提交
1591 1592 1593
	}
}

1594
static void md_update_sb(mddev_t * mddev, int force_change)
L
Linus Torvalds 已提交
1595
{
1596
	int err;
L
Linus Torvalds 已提交
1597 1598
	struct list_head *tmp;
	mdk_rdev_t *rdev;
1599
	int sync_req;
1600
	int nospares = 0;
L
Linus Torvalds 已提交
1601 1602

repeat:
1603
	spin_lock_irq(&mddev->write_lock);
1604

1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616
	set_bit(MD_CHANGE_PENDING, &mddev->flags);
	if (test_and_clear_bit(MD_CHANGE_DEVS, &mddev->flags))
		force_change = 1;
	if (test_and_clear_bit(MD_CHANGE_CLEAN, &mddev->flags))
		/* just a clean<-> dirty transition, possibly leave spares alone,
		 * though if events isn't the right even/odd, we will have to do
		 * spares after all
		 */
		nospares = 1;
	if (force_change)
		nospares = 0;
	if (mddev->degraded)
1617 1618 1619 1620 1621 1622 1623 1624 1625
		/* If the array is degraded, then skipping spares is both
		 * dangerous and fairly pointless.
		 * Dangerous because a device that was removed from the array
		 * might have a event_count that still looks up-to-date,
		 * so it can be re-added without a resync.
		 * Pointless because if there are any spares to skip,
		 * then a recovery will happen and soon that array won't
		 * be degraded any more and the spare can go back to sleep then.
		 */
1626
		nospares = 0;
1627

1628
	sync_req = mddev->in_sync;
L
Linus Torvalds 已提交
1629
	mddev->utime = get_seconds();
1630 1631 1632

	/* If this is just a dirty<->clean transition, and the array is clean
	 * and 'events' is odd, we can roll back to the previous clean state */
1633
	if (nospares
1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653
	    && (mddev->in_sync && mddev->recovery_cp == MaxSector)
	    && (mddev->events & 1))
		mddev->events--;
	else {
		/* otherwise we have to go forward and ... */
		mddev->events ++;
		if (!mddev->in_sync || mddev->recovery_cp != MaxSector) { /* not clean */
			/* .. if the array isn't clean, insist on an odd 'events' */
			if ((mddev->events&1)==0) {
				mddev->events++;
				nospares = 0;
			}
		} else {
			/* otherwise insist on an even 'events' (for clean states) */
			if ((mddev->events&1)) {
				mddev->events++;
				nospares = 0;
			}
		}
	}
L
Linus Torvalds 已提交
1654 1655 1656 1657 1658 1659 1660 1661 1662 1663

	if (!mddev->events) {
		/*
		 * oops, this 64-bit counter should never wrap.
		 * Either we are in around ~1 trillion A.C., assuming
		 * 1 reboot per second, or we have a bug:
		 */
		MD_BUG();
		mddev->events --;
	}
1664
	sync_sbs(mddev, nospares);
L
Linus Torvalds 已提交
1665 1666 1667 1668 1669

	/*
	 * do not write anything to disk if using
	 * nonpersistent superblocks
	 */
1670
	if (!mddev->persistent) {
1671
		clear_bit(MD_CHANGE_PENDING, &mddev->flags);
1672
		spin_unlock_irq(&mddev->write_lock);
1673
		wake_up(&mddev->sb_wait);
L
Linus Torvalds 已提交
1674
		return;
1675
	}
1676
	spin_unlock_irq(&mddev->write_lock);
L
Linus Torvalds 已提交
1677 1678 1679 1680 1681

	dprintk(KERN_INFO 
		"md: updating %s RAID superblock on device (in sync %d)\n",
		mdname(mddev),mddev->in_sync);

1682
	err = bitmap_update_sb(mddev->bitmap);
L
Linus Torvalds 已提交
1683 1684 1685
	ITERATE_RDEV(mddev,rdev,tmp) {
		char b[BDEVNAME_SIZE];
		dprintk(KERN_INFO "md: ");
1686 1687
		if (rdev->sb_loaded != 1)
			continue; /* no noise on spare devices */
1688
		if (test_bit(Faulty, &rdev->flags))
L
Linus Torvalds 已提交
1689 1690 1691
			dprintk("(skipping faulty ");

		dprintk("%s ", bdevname(rdev->bdev,b));
1692
		if (!test_bit(Faulty, &rdev->flags)) {
1693
			md_super_write(mddev,rdev,
1694
				       rdev->sb_offset<<1, rdev->sb_size,
1695 1696 1697 1698
				       rdev->sb_page);
			dprintk(KERN_INFO "(write) %s's sb offset: %llu\n",
				bdevname(rdev->bdev,b),
				(unsigned long long)rdev->sb_offset);
1699
			rdev->sb_events = mddev->events;
1700

L
Linus Torvalds 已提交
1701 1702
		} else
			dprintk(")\n");
1703
		if (mddev->level == LEVEL_MULTIPATH)
L
Linus Torvalds 已提交
1704 1705 1706
			/* only need to write one superblock... */
			break;
	}
1707
	md_super_wait(mddev);
1708
	/* if there was a failure, MD_CHANGE_DEVS was set, and we re-write super */
1709

1710
	spin_lock_irq(&mddev->write_lock);
1711 1712
	if (mddev->in_sync != sync_req ||
	    test_bit(MD_CHANGE_DEVS, &mddev->flags)) {
1713
		/* have to write it out again */
1714
		spin_unlock_irq(&mddev->write_lock);
1715 1716
		goto repeat;
	}
1717
	clear_bit(MD_CHANGE_PENDING, &mddev->flags);
1718
	spin_unlock_irq(&mddev->write_lock);
1719
	wake_up(&mddev->sb_wait);
1720

L
Linus Torvalds 已提交
1721 1722
}

1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742
/* 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;
}

1743 1744 1745 1746 1747 1748 1749
struct rdev_sysfs_entry {
	struct attribute attr;
	ssize_t (*show)(mdk_rdev_t *, char *);
	ssize_t (*store)(mdk_rdev_t *, const char *, size_t);
};

static ssize_t
1750
state_show(mdk_rdev_t *rdev, char *page)
1751 1752 1753 1754
{
	char *sep = "";
	int len=0;

1755
	if (test_bit(Faulty, &rdev->flags)) {
1756 1757 1758
		len+= sprintf(page+len, "%sfaulty",sep);
		sep = ",";
	}
1759
	if (test_bit(In_sync, &rdev->flags)) {
1760 1761 1762
		len += sprintf(page+len, "%sin_sync",sep);
		sep = ",";
	}
1763 1764 1765 1766
	if (test_bit(WriteMostly, &rdev->flags)) {
		len += sprintf(page+len, "%swrite_mostly",sep);
		sep = ",";
	}
1767 1768
	if (!test_bit(Faulty, &rdev->flags) &&
	    !test_bit(In_sync, &rdev->flags)) {
1769 1770 1771 1772 1773 1774
		len += sprintf(page+len, "%sspare", sep);
		sep = ",";
	}
	return len+sprintf(page+len, "\n");
}

1775 1776 1777 1778 1779 1780
static ssize_t
state_store(mdk_rdev_t *rdev, const char *buf, size_t len)
{
	/* can write
	 *  faulty  - simulates and error
	 *  remove  - disconnects the device
1781 1782
	 *  writemostly - sets write_mostly
	 *  -writemostly - clears write_mostly
1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793
	 */
	int err = -EINVAL;
	if (cmd_match(buf, "faulty") && rdev->mddev->pers) {
		md_error(rdev->mddev, rdev);
		err = 0;
	} else if (cmd_match(buf, "remove")) {
		if (rdev->raid_disk >= 0)
			err = -EBUSY;
		else {
			mddev_t *mddev = rdev->mddev;
			kick_rdev_from_array(rdev);
1794
			md_update_sb(mddev, 1);
1795 1796 1797
			md_new_event(mddev);
			err = 0;
		}
1798 1799 1800 1801 1802 1803
	} else if (cmd_match(buf, "writemostly")) {
		set_bit(WriteMostly, &rdev->flags);
		err = 0;
	} else if (cmd_match(buf, "-writemostly")) {
		clear_bit(WriteMostly, &rdev->flags);
		err = 0;
1804 1805 1806
	}
	return err ? err : len;
}
1807 1808
static struct rdev_sysfs_entry rdev_state =
__ATTR(state, S_IRUGO|S_IWUSR, state_show, state_store);
1809 1810

static ssize_t
1811
super_show(mdk_rdev_t *rdev, char *page)
1812 1813 1814 1815 1816 1817 1818
{
	if (rdev->sb_loaded && rdev->sb_size) {
		memcpy(page, page_address(rdev->sb_page), rdev->sb_size);
		return rdev->sb_size;
	} else
		return 0;
}
1819 1820
static struct rdev_sysfs_entry rdev_super = __ATTR_RO(super);

1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838
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 =
1839
__ATTR(errors, S_IRUGO|S_IWUSR, errors_show, errors_store);
1840

1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872
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 =
1873
__ATTR(slot, S_IRUGO|S_IWUSR, slot_show, slot_store);
1874

1875 1876 1877
static ssize_t
offset_show(mdk_rdev_t *rdev, char *page)
{
1878
	return sprintf(page, "%llu\n", (unsigned long long)rdev->data_offset);
1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894
}

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 =
1895
__ATTR(offset, S_IRUGO|S_IWUSR, offset_show, offset_store);
1896

1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918
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 =
1919
__ATTR(size, S_IRUGO|S_IWUSR, rdev_size_show, rdev_size_store);
1920

1921 1922 1923
static struct attribute *rdev_default_attrs[] = {
	&rdev_state.attr,
	&rdev_super.attr,
1924
	&rdev_errors.attr,
1925
	&rdev_slot.attr,
1926
	&rdev_offset.attr,
1927
	&rdev_size.attr,
1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949
	NULL,
};
static ssize_t
rdev_attr_show(struct kobject *kobj, struct attribute *attr, char *page)
{
	struct rdev_sysfs_entry *entry = container_of(attr, struct rdev_sysfs_entry, attr);
	mdk_rdev_t *rdev = container_of(kobj, mdk_rdev_t, kobj);

	if (!entry->show)
		return -EIO;
	return entry->show(rdev, page);
}

static ssize_t
rdev_attr_store(struct kobject *kobj, struct attribute *attr,
	      const char *page, size_t length)
{
	struct rdev_sysfs_entry *entry = container_of(attr, struct rdev_sysfs_entry, attr);
	mdk_rdev_t *rdev = container_of(kobj, mdk_rdev_t, kobj);

	if (!entry->store)
		return -EIO;
1950 1951
	if (!capable(CAP_SYS_ADMIN))
		return -EACCES;
1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969
	return entry->store(rdev, page, length);
}

static void rdev_free(struct kobject *ko)
{
	mdk_rdev_t *rdev = container_of(ko, mdk_rdev_t, kobj);
	kfree(rdev);
}
static struct sysfs_ops rdev_sysfs_ops = {
	.show		= rdev_attr_show,
	.store		= rdev_attr_store,
};
static struct kobj_type rdev_ktype = {
	.release	= rdev_free,
	.sysfs_ops	= &rdev_sysfs_ops,
	.default_attrs	= rdev_default_attrs,
};

L
Linus Torvalds 已提交
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986
/*
 * Import a device. If 'super_format' >= 0, then sanity check the superblock
 *
 * mark the device faulty if:
 *
 *   - the device is nonexistent (zero size)
 *   - the device has no valid superblock
 *
 * a faulty rdev _never_ has rdev->sb set.
 */
static mdk_rdev_t *md_import_device(dev_t newdev, int super_format, int super_minor)
{
	char b[BDEVNAME_SIZE];
	int err;
	mdk_rdev_t *rdev;
	sector_t size;

1987
	rdev = kzalloc(sizeof(*rdev), GFP_KERNEL);
L
Linus Torvalds 已提交
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
	if (!rdev) {
		printk(KERN_ERR "md: could not alloc mem for new device!\n");
		return ERR_PTR(-ENOMEM);
	}

	if ((err = alloc_disk_sb(rdev)))
		goto abort_free;

	err = lock_rdev(rdev, newdev);
	if (err)
		goto abort_free;

2000 2001 2002 2003
	rdev->kobj.parent = NULL;
	rdev->kobj.ktype = &rdev_ktype;
	kobject_init(&rdev->kobj);

L
Linus Torvalds 已提交
2004
	rdev->desc_nr = -1;
2005
	rdev->flags = 0;
L
Linus Torvalds 已提交
2006
	rdev->data_offset = 0;
2007
	rdev->sb_events = 0;
L
Linus Torvalds 已提交
2008
	atomic_set(&rdev->nr_pending, 0);
2009
	atomic_set(&rdev->read_errors, 0);
2010
	atomic_set(&rdev->corrected_errors, 0);
L
Linus Torvalds 已提交
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055

	size = rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
	if (!size) {
		printk(KERN_WARNING 
			"md: %s has zero or unknown size, marking faulty!\n",
			bdevname(rdev->bdev,b));
		err = -EINVAL;
		goto abort_free;
	}

	if (super_format >= 0) {
		err = super_types[super_format].
			load_super(rdev, NULL, super_minor);
		if (err == -EINVAL) {
			printk(KERN_WARNING 
				"md: %s has invalid sb, not importing!\n",
				bdevname(rdev->bdev,b));
			goto abort_free;
		}
		if (err < 0) {
			printk(KERN_WARNING 
				"md: could not read %s's sb, not importing!\n",
				bdevname(rdev->bdev,b));
			goto abort_free;
		}
	}
	INIT_LIST_HEAD(&rdev->same_set);

	return rdev;

abort_free:
	if (rdev->sb_page) {
		if (rdev->bdev)
			unlock_rdev(rdev);
		free_disk_sb(rdev);
	}
	kfree(rdev);
	return ERR_PTR(err);
}

/*
 * Check a full RAID array for plausibility
 */


2056
static void analyze_sbs(mddev_t * mddev)
L
Linus Torvalds 已提交
2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097
{
	int i;
	struct list_head *tmp;
	mdk_rdev_t *rdev, *freshest;
	char b[BDEVNAME_SIZE];

	freshest = NULL;
	ITERATE_RDEV(mddev,rdev,tmp)
		switch (super_types[mddev->major_version].
			load_super(rdev, freshest, mddev->minor_version)) {
		case 1:
			freshest = rdev;
			break;
		case 0:
			break;
		default:
			printk( KERN_ERR \
				"md: fatal superblock inconsistency in %s"
				" -- removing from array\n", 
				bdevname(rdev->bdev,b));
			kick_rdev_from_array(rdev);
		}


	super_types[mddev->major_version].
		validate_super(mddev, freshest);

	i = 0;
	ITERATE_RDEV(mddev,rdev,tmp) {
		if (rdev != freshest)
			if (super_types[mddev->major_version].
			    validate_super(mddev, rdev)) {
				printk(KERN_WARNING "md: kicking non-fresh %s"
					" from array!\n",
					bdevname(rdev->bdev,b));
				kick_rdev_from_array(rdev);
				continue;
			}
		if (mddev->level == LEVEL_MULTIPATH) {
			rdev->desc_nr = i++;
			rdev->raid_disk = rdev->desc_nr;
2098
			set_bit(In_sync, &rdev->flags);
L
Linus Torvalds 已提交
2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111
		}
	}



	if (mddev->recovery_cp != MaxSector &&
	    mddev->level >= 1)
		printk(KERN_ERR "md: %s: raid array is not clean"
		       " -- starting background reconstruction\n",
		       mdname(mddev));

}

2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157
static ssize_t
safe_delay_show(mddev_t *mddev, char *page)
{
	int msec = (mddev->safemode_delay*1000)/HZ;
	return sprintf(page, "%d.%03d\n", msec/1000, msec%1000);
}
static ssize_t
safe_delay_store(mddev_t *mddev, const char *cbuf, size_t len)
{
	int scale=1;
	int dot=0;
	int i;
	unsigned long msec;
	char buf[30];
	char *e;
	/* remove a period, and count digits after it */
	if (len >= sizeof(buf))
		return -EINVAL;
	strlcpy(buf, cbuf, len);
	buf[len] = 0;
	for (i=0; i<len; i++) {
		if (dot) {
			if (isdigit(buf[i])) {
				buf[i-1] = buf[i];
				scale *= 10;
			}
			buf[i] = 0;
		} else if (buf[i] == '.') {
			dot=1;
			buf[i] = 0;
		}
	}
	msec = simple_strtoul(buf, &e, 10);
	if (e == buf || (*e && *e != '\n'))
		return -EINVAL;
	msec = (msec * 1000) / scale;
	if (msec == 0)
		mddev->safemode_delay = 0;
	else {
		mddev->safemode_delay = (msec*HZ)/1000;
		if (mddev->safemode_delay == 0)
			mddev->safemode_delay = 1;
	}
	return len;
}
static struct md_sysfs_entry md_safe_delay =
2158
__ATTR(safe_mode_delay, S_IRUGO|S_IWUSR,safe_delay_show, safe_delay_store);
2159

2160
static ssize_t
2161
level_show(mddev_t *mddev, char *page)
2162
{
2163
	struct mdk_personality *p = mddev->pers;
2164
	if (p)
2165
		return sprintf(page, "%s\n", p->name);
2166 2167 2168 2169 2170 2171
	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;
2172 2173
}

2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192
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 =
2193
__ATTR(level, S_IRUGO|S_IWUSR, level_show, level_store);
2194

2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217

static ssize_t
layout_show(mddev_t *mddev, char *page)
{
	/* just a number, not meaningful for all levels */
	return sprintf(page, "%d\n", mddev->layout);
}

static ssize_t
layout_store(mddev_t *mddev, const char *buf, size_t len)
{
	char *e;
	unsigned long n = simple_strtoul(buf, &e, 10);
	if (mddev->pers)
		return -EBUSY;

	if (!*buf || (*e && *e != '\n'))
		return -EINVAL;

	mddev->layout = n;
	return len;
}
static struct md_sysfs_entry md_layout =
2218
__ATTR(layout, S_IRUGO|S_IWUSR, layout_show, layout_store);
2219 2220


2221
static ssize_t
2222
raid_disks_show(mddev_t *mddev, char *page)
2223
{
2224 2225
	if (mddev->raid_disks == 0)
		return 0;
2226 2227 2228
	return sprintf(page, "%d\n", mddev->raid_disks);
}

2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248
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 =
2249
__ATTR(raid_disks, S_IRUGO|S_IWUSR, raid_disks_show, raid_disks_store);
2250

2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272
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 =
2273
__ATTR(chunk_size, S_IRUGO|S_IWUSR, chunk_size_show, chunk_size_store);
2274

2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296
static ssize_t
resync_start_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%llu\n", (unsigned long long)mddev->recovery_cp);
}

static ssize_t
resync_start_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 long n = simple_strtoull(buf, &e, 10);

	if (mddev->pers)
		return -EBUSY;
	if (!*buf || (*e && *e != '\n'))
		return -EINVAL;

	mddev->recovery_cp = n;
	return len;
}
static struct md_sysfs_entry md_resync_start =
2297
__ATTR(resync_start, S_IRUGO|S_IWUSR, resync_start_show, resync_start_store);
2298

2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336
/*
 * The array state can be:
 *
 * clear
 *     No devices, no size, no level
 *     Equivalent to STOP_ARRAY ioctl
 * inactive
 *     May have some settings, but array is not active
 *        all IO results in error
 *     When written, doesn't tear down array, but just stops it
 * suspended (not supported yet)
 *     All IO requests will block. The array can be reconfigured.
 *     Writing this, if accepted, will block until array is quiessent
 * readonly
 *     no resync can happen.  no superblocks get written.
 *     write requests fail
 * read-auto
 *     like readonly, but behaves like 'clean' on a write request.
 *
 * clean - no pending writes, but otherwise active.
 *     When written to inactive array, starts without resync
 *     If a write request arrives then
 *       if metadata is known, mark 'dirty' and switch to 'active'.
 *       if not known, block and switch to write-pending
 *     If written to an active array that has pending writes, then fails.
 * active
 *     fully active: IO and resync can be happening.
 *     When written to inactive array, starts with resync
 *
 * write-pending
 *     clean, but writes are blocked waiting for 'active' to be written.
 *
 * active-idle
 *     like active, but no writes have been seen for a while (100msec).
 *
 */
enum array_state { clear, inactive, suspended, readonly, read_auto, clean, active,
		   write_pending, active_idle, bad_word};
2337
static char *array_states[] = {
2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436
	"clear", "inactive", "suspended", "readonly", "read-auto", "clean", "active",
	"write-pending", "active-idle", NULL };

static int match_word(const char *word, char **list)
{
	int n;
	for (n=0; list[n]; n++)
		if (cmd_match(word, list[n]))
			break;
	return n;
}

static ssize_t
array_state_show(mddev_t *mddev, char *page)
{
	enum array_state st = inactive;

	if (mddev->pers)
		switch(mddev->ro) {
		case 1:
			st = readonly;
			break;
		case 2:
			st = read_auto;
			break;
		case 0:
			if (mddev->in_sync)
				st = clean;
			else if (mddev->safemode)
				st = active_idle;
			else
				st = active;
		}
	else {
		if (list_empty(&mddev->disks) &&
		    mddev->raid_disks == 0 &&
		    mddev->size == 0)
			st = clear;
		else
			st = inactive;
	}
	return sprintf(page, "%s\n", array_states[st]);
}

static int do_md_stop(mddev_t * mddev, int ro);
static int do_md_run(mddev_t * mddev);
static int restart_array(mddev_t *mddev);

static ssize_t
array_state_store(mddev_t *mddev, const char *buf, size_t len)
{
	int err = -EINVAL;
	enum array_state st = match_word(buf, array_states);
	switch(st) {
	case bad_word:
		break;
	case clear:
		/* stopping an active array */
		if (mddev->pers) {
			if (atomic_read(&mddev->active) > 1)
				return -EBUSY;
			err = do_md_stop(mddev, 0);
		}
		break;
	case inactive:
		/* stopping an active array */
		if (mddev->pers) {
			if (atomic_read(&mddev->active) > 1)
				return -EBUSY;
			err = do_md_stop(mddev, 2);
		}
		break;
	case suspended:
		break; /* not supported yet */
	case readonly:
		if (mddev->pers)
			err = do_md_stop(mddev, 1);
		else {
			mddev->ro = 1;
			err = do_md_run(mddev);
		}
		break;
	case read_auto:
		/* stopping an active array */
		if (mddev->pers) {
			err = do_md_stop(mddev, 1);
			if (err == 0)
				mddev->ro = 2; /* FIXME mark devices writable */
		} else {
			mddev->ro = 2;
			err = do_md_run(mddev);
		}
		break;
	case clean:
		if (mddev->pers) {
			restart_array(mddev);
			spin_lock_irq(&mddev->write_lock);
			if (atomic_read(&mddev->writes_pending) == 0) {
				mddev->in_sync = 1;
2437
				set_bit(MD_CHANGE_CLEAN, &mddev->flags);
2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448
			}
			spin_unlock_irq(&mddev->write_lock);
		} else {
			mddev->ro = 0;
			mddev->recovery_cp = MaxSector;
			err = do_md_run(mddev);
		}
		break;
	case active:
		if (mddev->pers) {
			restart_array(mddev);
2449
			clear_bit(MD_CHANGE_CLEAN, &mddev->flags);
2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466
			wake_up(&mddev->sb_wait);
			err = 0;
		} else {
			mddev->ro = 0;
			err = do_md_run(mddev);
		}
		break;
	case write_pending:
	case active_idle:
		/* these cannot be set */
		break;
	}
	if (err)
		return err;
	else
		return len;
}
2467 2468
static struct md_sysfs_entry md_array_state =
__ATTR(array_state, S_IRUGO|S_IWUSR, array_state_show, array_state_store);
2469

2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527
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 =
2528
__ATTR(new_dev, S_IWUSR, null_show, new_dev_store);
2529

2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559
static ssize_t
bitmap_store(mddev_t *mddev, const char *buf, size_t len)
{
	char *end;
	unsigned long chunk, end_chunk;

	if (!mddev->bitmap)
		goto out;
	/* buf should be <chunk> <chunk> ... or <chunk>-<chunk> ... (range) */
	while (*buf) {
		chunk = end_chunk = simple_strtoul(buf, &end, 0);
		if (buf == end) break;
		if (*end == '-') { /* range */
			buf = end + 1;
			end_chunk = simple_strtoul(buf, &end, 0);
			if (buf == end) break;
		}
		if (*end && !isspace(*end)) break;
		bitmap_dirty_bits(mddev->bitmap, chunk, end_chunk);
		buf = end;
		while (isspace(*buf)) buf++;
	}
	bitmap_unplug(mddev->bitmap); /* flush the bits to disk */
out:
	return len;
}

static struct md_sysfs_entry md_bitmap =
__ATTR(bitmap_set_bits, S_IWUSR, null_show, bitmap_store);

2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583
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);
2584
		md_update_sb(mddev, 1);
2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595
	} 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 =
2596
__ATTR(component_size, S_IRUGO|S_IWUSR, size_show, size_store);
2597

2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643

/* 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 =
2644
__ATTR(metadata_version, S_IRUGO|S_IWUSR, metadata_show, metadata_store);
2645

2646
static ssize_t
2647
action_show(mddev_t *mddev, char *page)
2648
{
2649
	char *type = "idle";
N
NeilBrown 已提交
2650 2651
	if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
	    test_bit(MD_RECOVERY_NEEDED, &mddev->recovery)) {
2652 2653 2654
		if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
			type = "reshape";
		else if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667
			if (!test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery))
				type = "resync";
			else if (test_bit(MD_RECOVERY_CHECK, &mddev->recovery))
				type = "check";
			else
				type = "repair";
		} else
			type = "recover";
	}
	return sprintf(page, "%s\n", type);
}

static ssize_t
2668
action_store(mddev_t *mddev, const char *page, size_t len)
2669
{
2670 2671 2672
	if (!mddev->pers || !mddev->pers->sync_request)
		return -EINVAL;

2673
	if (cmd_match(page, "idle")) {
2674 2675 2676 2677 2678 2679
		if (mddev->sync_thread) {
			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
			md_unregister_thread(mddev->sync_thread);
			mddev->sync_thread = NULL;
			mddev->recovery = 0;
		}
2680 2681
	} else if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
		   test_bit(MD_RECOVERY_NEEDED, &mddev->recovery))
2682
		return -EBUSY;
2683
	else if (cmd_match(page, "resync") || cmd_match(page, "recover"))
2684
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
2685 2686 2687 2688 2689 2690 2691 2692
	else if (cmd_match(page, "reshape")) {
		int err;
		if (mddev->pers->start_reshape == NULL)
			return -EINVAL;
		err = mddev->pers->start_reshape(mddev);
		if (err)
			return err;
	} else {
2693
		if (cmd_match(page, "check"))
2694
			set_bit(MD_RECOVERY_CHECK, &mddev->recovery);
2695
		else if (!cmd_match(page, "repair"))
2696 2697 2698 2699
			return -EINVAL;
		set_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
		set_bit(MD_RECOVERY_SYNC, &mddev->recovery);
	}
2700
	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
2701 2702 2703 2704
	md_wakeup_thread(mddev->thread);
	return len;
}

2705
static ssize_t
2706
mismatch_cnt_show(mddev_t *mddev, char *page)
2707 2708 2709 2710 2711
{
	return sprintf(page, "%llu\n",
		       (unsigned long long) mddev->resync_mismatches);
}

2712 2713
static struct md_sysfs_entry md_scan_mode =
__ATTR(sync_action, S_IRUGO|S_IWUSR, action_show, action_store);
2714

2715

2716
static struct md_sysfs_entry md_mismatches = __ATTR_RO(mismatch_cnt);
2717

2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774
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;
2775
	resync = (mddev->curr_mark_cnt - atomic_read(&mddev->recovery_active));
2776 2777 2778 2779 2780 2781
	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 */
}

2782
static struct md_sysfs_entry md_sync_speed = __ATTR_RO(sync_speed);
2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797

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);
}

2798
static struct md_sysfs_entry md_sync_completed = __ATTR_RO(sync_completed);
2799

2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856
static ssize_t
suspend_lo_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%llu\n", (unsigned long long)mddev->suspend_lo);
}

static ssize_t
suspend_lo_store(mddev_t *mddev, const char *buf, size_t len)
{
	char *e;
	unsigned long long new = simple_strtoull(buf, &e, 10);

	if (mddev->pers->quiesce == NULL)
		return -EINVAL;
	if (buf == e || (*e && *e != '\n'))
		return -EINVAL;
	if (new >= mddev->suspend_hi ||
	    (new > mddev->suspend_lo && new < mddev->suspend_hi)) {
		mddev->suspend_lo = new;
		mddev->pers->quiesce(mddev, 2);
		return len;
	} else
		return -EINVAL;
}
static struct md_sysfs_entry md_suspend_lo =
__ATTR(suspend_lo, S_IRUGO|S_IWUSR, suspend_lo_show, suspend_lo_store);


static ssize_t
suspend_hi_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%llu\n", (unsigned long long)mddev->suspend_hi);
}

static ssize_t
suspend_hi_store(mddev_t *mddev, const char *buf, size_t len)
{
	char *e;
	unsigned long long new = simple_strtoull(buf, &e, 10);

	if (mddev->pers->quiesce == NULL)
		return -EINVAL;
	if (buf == e || (*e && *e != '\n'))
		return -EINVAL;
	if ((new <= mddev->suspend_lo && mddev->suspend_lo >= mddev->suspend_hi) ||
	    (new > mddev->suspend_lo && new > mddev->suspend_hi)) {
		mddev->suspend_hi = new;
		mddev->pers->quiesce(mddev, 1);
		mddev->pers->quiesce(mddev, 0);
		return len;
	} else
		return -EINVAL;
}
static struct md_sysfs_entry md_suspend_hi =
__ATTR(suspend_hi, S_IRUGO|S_IWUSR, suspend_hi_show, suspend_hi_store);


2857 2858
static struct attribute *md_default_attrs[] = {
	&md_level.attr,
2859
	&md_layout.attr,
2860
	&md_raid_disks.attr,
2861
	&md_chunk_size.attr,
2862
	&md_size.attr,
2863
	&md_resync_start.attr,
2864
	&md_metadata.attr,
2865
	&md_new_device.attr,
2866
	&md_safe_delay.attr,
2867
	&md_array_state.attr,
2868 2869 2870 2871
	NULL,
};

static struct attribute *md_redundancy_attrs[] = {
2872
	&md_scan_mode.attr,
2873
	&md_mismatches.attr,
2874 2875 2876 2877
	&md_sync_min.attr,
	&md_sync_max.attr,
	&md_sync_speed.attr,
	&md_sync_completed.attr,
2878 2879
	&md_suspend_lo.attr,
	&md_suspend_hi.attr,
2880
	&md_bitmap.attr,
2881 2882
	NULL,
};
2883 2884 2885 2886 2887
static struct attribute_group md_redundancy_group = {
	.name = NULL,
	.attrs = md_redundancy_attrs,
};

2888 2889 2890 2891 2892 2893

static ssize_t
md_attr_show(struct kobject *kobj, struct attribute *attr, char *page)
{
	struct md_sysfs_entry *entry = container_of(attr, struct md_sysfs_entry, attr);
	mddev_t *mddev = container_of(kobj, struct mddev_s, kobj);
2894
	ssize_t rv;
2895 2896 2897

	if (!entry->show)
		return -EIO;
I
Ingo Molnar 已提交
2898 2899 2900 2901 2902
	rv = mddev_lock(mddev);
	if (!rv) {
		rv = entry->show(mddev, page);
		mddev_unlock(mddev);
	}
2903
	return rv;
2904 2905 2906 2907 2908 2909 2910 2911
}

static ssize_t
md_attr_store(struct kobject *kobj, struct attribute *attr,
	      const char *page, size_t length)
{
	struct md_sysfs_entry *entry = container_of(attr, struct md_sysfs_entry, attr);
	mddev_t *mddev = container_of(kobj, struct mddev_s, kobj);
2912
	ssize_t rv;
2913 2914 2915

	if (!entry->store)
		return -EIO;
2916 2917
	if (!capable(CAP_SYS_ADMIN))
		return -EACCES;
I
Ingo Molnar 已提交
2918 2919 2920 2921 2922
	rv = mddev_lock(mddev);
	if (!rv) {
		rv = entry->store(mddev, page, length);
		mddev_unlock(mddev);
	}
2923
	return rv;
2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941
}

static void md_free(struct kobject *ko)
{
	mddev_t *mddev = container_of(ko, mddev_t, kobj);
	kfree(mddev);
}

static struct sysfs_ops md_sysfs_ops = {
	.show	= md_attr_show,
	.store	= md_attr_store,
};
static struct kobj_type md_ktype = {
	.release	= md_free,
	.sysfs_ops	= &md_sysfs_ops,
	.default_attrs	= md_default_attrs,
};

L
Linus Torvalds 已提交
2942 2943 2944 2945
int mdp_major = 0;

static struct kobject *md_probe(dev_t dev, int *part, void *data)
{
A
Arjan van de Ven 已提交
2946
	static DEFINE_MUTEX(disks_mutex);
L
Linus Torvalds 已提交
2947 2948 2949 2950 2951 2952 2953 2954 2955
	mddev_t *mddev = mddev_find(dev);
	struct gendisk *disk;
	int partitioned = (MAJOR(dev) != MD_MAJOR);
	int shift = partitioned ? MdpMinorShift : 0;
	int unit = MINOR(dev) >> shift;

	if (!mddev)
		return NULL;

A
Arjan van de Ven 已提交
2956
	mutex_lock(&disks_mutex);
L
Linus Torvalds 已提交
2957
	if (mddev->gendisk) {
A
Arjan van de Ven 已提交
2958
		mutex_unlock(&disks_mutex);
L
Linus Torvalds 已提交
2959 2960 2961 2962 2963
		mddev_put(mddev);
		return NULL;
	}
	disk = alloc_disk(1 << shift);
	if (!disk) {
A
Arjan van de Ven 已提交
2964
		mutex_unlock(&disks_mutex);
L
Linus Torvalds 已提交
2965 2966 2967 2968 2969
		mddev_put(mddev);
		return NULL;
	}
	disk->major = MAJOR(dev);
	disk->first_minor = unit << shift;
2970
	if (partitioned)
L
Linus Torvalds 已提交
2971
		sprintf(disk->disk_name, "md_d%d", unit);
2972
	else
L
Linus Torvalds 已提交
2973 2974 2975 2976 2977 2978
		sprintf(disk->disk_name, "md%d", unit);
	disk->fops = &md_fops;
	disk->private_data = mddev;
	disk->queue = mddev->queue;
	add_disk(disk);
	mddev->gendisk = disk;
A
Arjan van de Ven 已提交
2979
	mutex_unlock(&disks_mutex);
2980
	mddev->kobj.parent = &disk->kobj;
2981 2982 2983 2984
	mddev->kobj.k_name = NULL;
	snprintf(mddev->kobj.name, KOBJ_NAME_LEN, "%s", "md");
	mddev->kobj.ktype = &md_ktype;
	kobject_register(&mddev->kobj);
L
Linus Torvalds 已提交
2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995
	return NULL;
}

static void md_safemode_timeout(unsigned long data)
{
	mddev_t *mddev = (mddev_t *) data;

	mddev->safemode = 1;
	md_wakeup_thread(mddev->thread);
}

2996
static int start_dirty_degraded;
L
Linus Torvalds 已提交
2997 2998 2999

static int do_md_run(mddev_t * mddev)
{
3000
	int err;
L
Linus Torvalds 已提交
3001 3002 3003 3004
	int chunk_size;
	struct list_head *tmp;
	mdk_rdev_t *rdev;
	struct gendisk *disk;
3005
	struct mdk_personality *pers;
L
Linus Torvalds 已提交
3006 3007
	char b[BDEVNAME_SIZE];

3008 3009
	if (list_empty(&mddev->disks))
		/* cannot run an array with no devices.. */
L
Linus Torvalds 已提交
3010 3011 3012 3013 3014 3015 3016 3017
		return -EINVAL;

	if (mddev->pers)
		return -EBUSY;

	/*
	 * Analyze all RAID superblock(s)
	 */
3018 3019
	if (!mddev->raid_disks)
		analyze_sbs(mddev);
L
Linus Torvalds 已提交
3020 3021

	chunk_size = mddev->chunk_size;
3022 3023

	if (chunk_size) {
L
Linus Torvalds 已提交
3024 3025 3026 3027 3028 3029 3030 3031 3032
		if (chunk_size > MAX_CHUNK_SIZE) {
			printk(KERN_ERR "too big chunk_size: %d > %d\n",
				chunk_size, MAX_CHUNK_SIZE);
			return -EINVAL;
		}
		/*
		 * chunk-size has to be a power of 2 and multiples of PAGE_SIZE
		 */
		if ( (1 << ffz(~chunk_size)) != chunk_size) {
3033
			printk(KERN_ERR "chunk_size of %d not valid\n", chunk_size);
L
Linus Torvalds 已提交
3034 3035 3036 3037 3038 3039 3040 3041 3042 3043
			return -EINVAL;
		}
		if (chunk_size < PAGE_SIZE) {
			printk(KERN_ERR "too small chunk_size: %d < %ld\n",
				chunk_size, PAGE_SIZE);
			return -EINVAL;
		}

		/* devices must have minimum size of one chunk */
		ITERATE_RDEV(mddev,rdev,tmp) {
3044
			if (test_bit(Faulty, &rdev->flags))
L
Linus Torvalds 已提交
3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058
				continue;
			if (rdev->size < chunk_size / 1024) {
				printk(KERN_WARNING
					"md: Dev %s smaller than chunk_size:"
					" %lluk < %dk\n",
					bdevname(rdev->bdev,b),
					(unsigned long long)rdev->size,
					chunk_size / 1024);
				return -EINVAL;
			}
		}
	}

#ifdef CONFIG_KMOD
3059 3060 3061 3062
	if (mddev->level != LEVEL_NONE)
		request_module("md-level-%d", mddev->level);
	else if (mddev->clevel[0])
		request_module("md-%s", mddev->clevel);
L
Linus Torvalds 已提交
3063 3064 3065 3066 3067 3068 3069 3070 3071
#endif

	/*
	 * Drop all container device buffers, from now on
	 * the only valid external interface is through the md
	 * device.
	 * Also find largest hardsector size
	 */
	ITERATE_RDEV(mddev,rdev,tmp) {
3072
		if (test_bit(Faulty, &rdev->flags))
L
Linus Torvalds 已提交
3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083
			continue;
		sync_blockdev(rdev->bdev);
		invalidate_bdev(rdev->bdev, 0);
	}

	md_probe(mddev->unit, NULL, NULL);
	disk = mddev->gendisk;
	if (!disk)
		return -ENOMEM;

	spin_lock(&pers_lock);
3084
	pers = find_pers(mddev->level, mddev->clevel);
3085
	if (!pers || !try_module_get(pers->owner)) {
L
Linus Torvalds 已提交
3086
		spin_unlock(&pers_lock);
3087 3088 3089 3090 3091 3092
		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);
L
Linus Torvalds 已提交
3093 3094
		return -EINVAL;
	}
3095
	mddev->pers = pers;
L
Linus Torvalds 已提交
3096
	spin_unlock(&pers_lock);
3097 3098
	mddev->level = pers->level;
	strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
L
Linus Torvalds 已提交
3099

3100
	if (mddev->reshape_position != MaxSector &&
3101
	    pers->start_reshape == NULL) {
3102 3103 3104 3105 3106 3107
		/* This personality cannot handle reshaping... */
		mddev->pers = NULL;
		module_put(pers->owner);
		return -EINVAL;
	}

3108
	mddev->recovery = 0;
L
Linus Torvalds 已提交
3109
	mddev->resync_max_sectors = mddev->size << 1; /* may be over-ridden by personality */
3110
	mddev->barriers_work = 1;
3111
	mddev->ok_start_degraded = start_dirty_degraded;
L
Linus Torvalds 已提交
3112

3113 3114 3115
	if (start_readonly)
		mddev->ro = 2; /* read-only, but switch on first write */

3116 3117 3118 3119 3120 3121 3122 3123 3124
	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);
		}
	}
L
Linus Torvalds 已提交
3125 3126 3127 3128
	if (err) {
		printk(KERN_ERR "md: pers->run() failed ...\n");
		module_put(mddev->pers->owner);
		mddev->pers = NULL;
3129 3130
		bitmap_destroy(mddev);
		return err;
L
Linus Torvalds 已提交
3131
	}
3132 3133
	if (mddev->pers->sync_request)
		sysfs_create_group(&mddev->kobj, &md_redundancy_group);
3134 3135 3136
	else if (mddev->ro == 2) /* auto-readonly not meaningful */
		mddev->ro = 0;

L
Linus Torvalds 已提交
3137 3138 3139 3140
 	atomic_set(&mddev->writes_pending,0);
	mddev->safemode = 0;
	mddev->safemode_timer.function = md_safemode_timeout;
	mddev->safemode_timer.data = (unsigned long) mddev;
3141
	mddev->safemode_delay = (200 * HZ)/1000 +1; /* 200 msec delay */
L
Linus Torvalds 已提交
3142
	mddev->in_sync = 1;
3143 3144 3145 3146 3147 3148 3149

	ITERATE_RDEV(mddev,rdev,tmp)
		if (rdev->raid_disk >= 0) {
			char nm[20];
			sprintf(nm, "rd%d", rdev->raid_disk);
			sysfs_create_link(&mddev->kobj, &rdev->kobj, nm);
		}
L
Linus Torvalds 已提交
3150 3151 3152
	
	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	
3153 3154
	if (mddev->flags)
		md_update_sb(mddev, 0);
L
Linus Torvalds 已提交
3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167

	set_capacity(disk, mddev->array_size<<1);

	/* If we call blk_queue_make_request here, it will
	 * re-initialise max_sectors etc which may have been
	 * refined inside -> run.  So just set the bits we need to set.
	 * Most initialisation happended when we called
	 * blk_queue_make_request(..., md_fail_request)
	 * earlier.
	 */
	mddev->queue->queuedata = mddev;
	mddev->queue->make_request_fn = mddev->pers->make_request;

3168 3169 3170 3171
	/* If there is a partially-recovered drive we need to
	 * start recovery here.  If we leave it to md_check_recovery,
	 * it will remove the drives and not do the right thing
	 */
3172
	if (mddev->degraded && !mddev->sync_thread) {
3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192
		struct list_head *rtmp;
		int spares = 0;
		ITERATE_RDEV(mddev,rdev,rtmp)
			if (rdev->raid_disk >= 0 &&
			    !test_bit(In_sync, &rdev->flags) &&
			    !test_bit(Faulty, &rdev->flags))
				/* complete an interrupted recovery */
				spares++;
		if (spares && mddev->pers->sync_request) {
			mddev->recovery = 0;
			set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
			mddev->sync_thread = md_register_thread(md_do_sync,
								mddev,
								"%s_resync");
			if (!mddev->sync_thread) {
				printk(KERN_ERR "%s: could not start resync"
				       " thread...\n",
				       mdname(mddev));
				/* leave the spares where they are, it shouldn't hurt */
				mddev->recovery = 0;
3193
			}
3194 3195
		}
	}
3196 3197
	md_wakeup_thread(mddev->thread);
	md_wakeup_thread(mddev->sync_thread); /* possibly kick off a reshape */
3198

L
Linus Torvalds 已提交
3199
	mddev->changed = 1;
3200
	md_new_event(mddev);
L
Linus Torvalds 已提交
3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231
	return 0;
}

static int restart_array(mddev_t *mddev)
{
	struct gendisk *disk = mddev->gendisk;
	int err;

	/*
	 * Complain if it has no devices
	 */
	err = -ENXIO;
	if (list_empty(&mddev->disks))
		goto out;

	if (mddev->pers) {
		err = -EBUSY;
		if (!mddev->ro)
			goto out;

		mddev->safemode = 0;
		mddev->ro = 0;
		set_disk_ro(disk, 0);

		printk(KERN_INFO "md: %s switched to read-write mode.\n",
			mdname(mddev));
		/*
		 * Kick recovery or resync if necessary
		 */
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
		md_wakeup_thread(mddev->thread);
3232
		md_wakeup_thread(mddev->sync_thread);
L
Linus Torvalds 已提交
3233
		err = 0;
3234
	} else
L
Linus Torvalds 已提交
3235 3236 3237 3238 3239 3240
		err = -EINVAL;

out:
	return err;
}

3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266
/* similar to deny_write_access, but accounts for our holding a reference
 * to the file ourselves */
static int deny_bitmap_write_access(struct file * file)
{
	struct inode *inode = file->f_mapping->host;

	spin_lock(&inode->i_lock);
	if (atomic_read(&inode->i_writecount) > 1) {
		spin_unlock(&inode->i_lock);
		return -ETXTBSY;
	}
	atomic_set(&inode->i_writecount, -1);
	spin_unlock(&inode->i_lock);

	return 0;
}

static void restore_bitmap_write_access(struct file *file)
{
	struct inode *inode = file->f_mapping->host;

	spin_lock(&inode->i_lock);
	atomic_set(&inode->i_writecount, 1);
	spin_unlock(&inode->i_lock);
}

3267 3268 3269 3270 3271 3272
/* mode:
 *   0 - completely stop and dis-assemble array
 *   1 - switch to readonly
 *   2 - stop but do not disassemble array
 */
static int do_md_stop(mddev_t * mddev, int mode)
L
Linus Torvalds 已提交
3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283
{
	int err = 0;
	struct gendisk *disk = mddev->gendisk;

	if (mddev->pers) {
		if (atomic_read(&mddev->active)>2) {
			printk("md: %s still in use.\n",mdname(mddev));
			return -EBUSY;
		}

		if (mddev->sync_thread) {
3284
			set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
L
Linus Torvalds 已提交
3285 3286 3287 3288 3289 3290 3291 3292 3293
			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
			md_unregister_thread(mddev->sync_thread);
			mddev->sync_thread = NULL;
		}

		del_timer_sync(&mddev->safemode_timer);

		invalidate_partition(disk, 0);

3294 3295
		switch(mode) {
		case 1: /* readonly */
L
Linus Torvalds 已提交
3296
			err  = -ENXIO;
3297
			if (mddev->ro==1)
L
Linus Torvalds 已提交
3298 3299
				goto out;
			mddev->ro = 1;
3300 3301 3302
			break;
		case 0: /* disassemble */
		case 2: /* stop */
3303
			bitmap_flush(mddev);
3304
			md_super_wait(mddev);
L
Linus Torvalds 已提交
3305 3306 3307 3308
			if (mddev->ro)
				set_disk_ro(disk, 0);
			blk_queue_make_request(mddev->queue, md_fail_request);
			mddev->pers->stop(mddev);
3309 3310 3311
			if (mddev->pers->sync_request)
				sysfs_remove_group(&mddev->kobj, &md_redundancy_group);

L
Linus Torvalds 已提交
3312 3313 3314 3315 3316
			module_put(mddev->pers->owner);
			mddev->pers = NULL;
			if (mddev->ro)
				mddev->ro = 0;
		}
3317
		if (!mddev->in_sync || mddev->flags) {
L
Linus Torvalds 已提交
3318 3319
			/* mark array as shutdown cleanly */
			mddev->in_sync = 1;
3320
			md_update_sb(mddev, 1);
L
Linus Torvalds 已提交
3321
		}
3322
		if (mode == 1)
L
Linus Torvalds 已提交
3323
			set_disk_ro(disk, 1);
3324
		clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
L
Linus Torvalds 已提交
3325
	}
3326

L
Linus Torvalds 已提交
3327 3328 3329
	/*
	 * Free resources if final stop
	 */
3330
	if (mode == 0) {
3331 3332
		mdk_rdev_t *rdev;
		struct list_head *tmp;
L
Linus Torvalds 已提交
3333 3334 3335
		struct gendisk *disk;
		printk(KERN_INFO "md: %s stopped.\n", mdname(mddev));

3336 3337
		bitmap_destroy(mddev);
		if (mddev->bitmap_file) {
3338
			restore_bitmap_write_access(mddev->bitmap_file);
3339 3340 3341 3342 3343
			fput(mddev->bitmap_file);
			mddev->bitmap_file = NULL;
		}
		mddev->bitmap_offset = 0;

3344 3345 3346 3347 3348 3349 3350
		ITERATE_RDEV(mddev,rdev,tmp)
			if (rdev->raid_disk >= 0) {
				char nm[20];
				sprintf(nm, "rd%d", rdev->raid_disk);
				sysfs_remove_link(&mddev->kobj, nm);
			}

L
Linus Torvalds 已提交
3351 3352 3353
		export_array(mddev);

		mddev->array_size = 0;
3354 3355
		mddev->size = 0;
		mddev->raid_disks = 0;
3356
		mddev->recovery_cp = 0;
3357

L
Linus Torvalds 已提交
3358 3359 3360 3361
		disk = mddev->gendisk;
		if (disk)
			set_capacity(disk, 0);
		mddev->changed = 1;
3362
	} else if (mddev->pers)
L
Linus Torvalds 已提交
3363 3364 3365
		printk(KERN_INFO "md: %s switched to read-only mode.\n",
			mdname(mddev));
	err = 0;
3366
	md_new_event(mddev);
L
Linus Torvalds 已提交
3367 3368 3369 3370 3371 3372 3373 3374 3375 3376
out:
	return err;
}

static void autorun_array(mddev_t *mddev)
{
	mdk_rdev_t *rdev;
	struct list_head *tmp;
	int err;

3377
	if (list_empty(&mddev->disks))
L
Linus Torvalds 已提交
3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415
		return;

	printk(KERN_INFO "md: running: ");

	ITERATE_RDEV(mddev,rdev,tmp) {
		char b[BDEVNAME_SIZE];
		printk("<%s>", bdevname(rdev->bdev,b));
	}
	printk("\n");

	err = do_md_run (mddev);
	if (err) {
		printk(KERN_WARNING "md: do_md_run() returned %d\n", err);
		do_md_stop (mddev, 0);
	}
}

/*
 * lets try to run arrays based on all disks that have arrived
 * until now. (those are in pending_raid_disks)
 *
 * the method: pick the first pending disk, collect all disks with
 * the same UUID, remove all from the pending list and put them into
 * the 'same_array' list. Then order this list based on superblock
 * update time (freshest comes first), kick out 'old' disks and
 * compare superblocks. If everything's fine then run it.
 *
 * If "unit" is allocated, then bump its reference count
 */
static void autorun_devices(int part)
{
	struct list_head *tmp;
	mdk_rdev_t *rdev0, *rdev;
	mddev_t *mddev;
	char b[BDEVNAME_SIZE];

	printk(KERN_INFO "md: autorun ...\n");
	while (!list_empty(&pending_raid_disks)) {
3416
		int unit;
L
Linus Torvalds 已提交
3417
		dev_t dev;
3418
		LIST_HEAD(candidates);
L
Linus Torvalds 已提交
3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435
		rdev0 = list_entry(pending_raid_disks.next,
					 mdk_rdev_t, same_set);

		printk(KERN_INFO "md: considering %s ...\n",
			bdevname(rdev0->bdev,b));
		INIT_LIST_HEAD(&candidates);
		ITERATE_RDEV_PENDING(rdev,tmp)
			if (super_90_load(rdev, rdev0, 0) >= 0) {
				printk(KERN_INFO "md:  adding %s ...\n",
					bdevname(rdev->bdev,b));
				list_move(&rdev->same_set, &candidates);
			}
		/*
		 * now we have a set of devices, with all of them having
		 * mostly sane superblocks. It's time to allocate the
		 * mddev.
		 */
3436 3437 3438 3439 3440 3441 3442 3443 3444
		if (part) {
			dev = MKDEV(mdp_major,
				    rdev0->preferred_minor << MdpMinorShift);
			unit = MINOR(dev) >> MdpMinorShift;
		} else {
			dev = MKDEV(MD_MAJOR, rdev0->preferred_minor);
			unit = MINOR(dev);
		}
		if (rdev0->preferred_minor != unit) {
L
Linus Torvalds 已提交
3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509
			printk(KERN_INFO "md: unit number in %s is bad: %d\n",
			       bdevname(rdev0->bdev, b), rdev0->preferred_minor);
			break;
		}

		md_probe(dev, NULL, NULL);
		mddev = mddev_find(dev);
		if (!mddev) {
			printk(KERN_ERR 
				"md: cannot allocate memory for md drive.\n");
			break;
		}
		if (mddev_lock(mddev)) 
			printk(KERN_WARNING "md: %s locked, cannot run\n",
			       mdname(mddev));
		else if (mddev->raid_disks || mddev->major_version
			 || !list_empty(&mddev->disks)) {
			printk(KERN_WARNING 
				"md: %s already running, cannot run %s\n",
				mdname(mddev), bdevname(rdev0->bdev,b));
			mddev_unlock(mddev);
		} else {
			printk(KERN_INFO "md: created %s\n", mdname(mddev));
			ITERATE_RDEV_GENERIC(candidates,rdev,tmp) {
				list_del_init(&rdev->same_set);
				if (bind_rdev_to_array(rdev, mddev))
					export_rdev(rdev);
			}
			autorun_array(mddev);
			mddev_unlock(mddev);
		}
		/* on success, candidates will be empty, on error
		 * it won't...
		 */
		ITERATE_RDEV_GENERIC(candidates,rdev,tmp)
			export_rdev(rdev);
		mddev_put(mddev);
	}
	printk(KERN_INFO "md: ... autorun DONE.\n");
}

static int get_version(void __user * arg)
{
	mdu_version_t ver;

	ver.major = MD_MAJOR_VERSION;
	ver.minor = MD_MINOR_VERSION;
	ver.patchlevel = MD_PATCHLEVEL_VERSION;

	if (copy_to_user(arg, &ver, sizeof(ver)))
		return -EFAULT;

	return 0;
}

static int get_array_info(mddev_t * mddev, void __user * arg)
{
	mdu_array_info_t info;
	int nr,working,active,failed,spare;
	mdk_rdev_t *rdev;
	struct list_head *tmp;

	nr=working=active=failed=spare=0;
	ITERATE_RDEV(mddev,rdev,tmp) {
		nr++;
3510
		if (test_bit(Faulty, &rdev->flags))
L
Linus Torvalds 已提交
3511 3512 3513
			failed++;
		else {
			working++;
3514
			if (test_bit(In_sync, &rdev->flags))
L
Linus Torvalds 已提交
3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526
				active++;	
			else
				spare++;
		}
	}

	info.major_version = mddev->major_version;
	info.minor_version = mddev->minor_version;
	info.patch_version = MD_PATCHLEVEL_VERSION;
	info.ctime         = mddev->ctime;
	info.level         = mddev->level;
	info.size          = mddev->size;
3527 3528
	if (info.size != mddev->size) /* overflow */
		info.size = -1;
L
Linus Torvalds 已提交
3529 3530 3531 3532 3533 3534 3535 3536 3537
	info.nr_disks      = nr;
	info.raid_disks    = mddev->raid_disks;
	info.md_minor      = mddev->md_minor;
	info.not_persistent= !mddev->persistent;

	info.utime         = mddev->utime;
	info.state         = 0;
	if (mddev->in_sync)
		info.state = (1<<MD_SB_CLEAN);
3538 3539
	if (mddev->bitmap && mddev->bitmap_offset)
		info.state = (1<<MD_SB_BITMAP_PRESENT);
L
Linus Torvalds 已提交
3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553
	info.active_disks  = active;
	info.working_disks = working;
	info.failed_disks  = failed;
	info.spare_disks   = spare;

	info.layout        = mddev->layout;
	info.chunk_size    = mddev->chunk_size;

	if (copy_to_user(arg, &info, sizeof(info)))
		return -EFAULT;

	return 0;
}

3554
static int get_bitmap_file(mddev_t * mddev, void __user * arg)
3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589
{
	mdu_bitmap_file_t *file = NULL; /* too big for stack allocation */
	char *ptr, *buf = NULL;
	int err = -ENOMEM;

	file = kmalloc(sizeof(*file), GFP_KERNEL);
	if (!file)
		goto out;

	/* bitmap disabled, zero the first byte and copy out */
	if (!mddev->bitmap || !mddev->bitmap->file) {
		file->pathname[0] = '\0';
		goto copy_out;
	}

	buf = kmalloc(sizeof(file->pathname), GFP_KERNEL);
	if (!buf)
		goto out;

	ptr = file_path(mddev->bitmap->file, buf, sizeof(file->pathname));
	if (!ptr)
		goto out;

	strcpy(file->pathname, ptr);

copy_out:
	err = 0;
	if (copy_to_user(arg, file, sizeof(*file)))
		err = -EFAULT;
out:
	kfree(buf);
	kfree(file);
	return err;
}

L
Linus Torvalds 已提交
3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606
static int get_disk_info(mddev_t * mddev, void __user * arg)
{
	mdu_disk_info_t info;
	unsigned int nr;
	mdk_rdev_t *rdev;

	if (copy_from_user(&info, arg, sizeof(info)))
		return -EFAULT;

	nr = info.number;

	rdev = find_rdev_nr(mddev, nr);
	if (rdev) {
		info.major = MAJOR(rdev->bdev->bd_dev);
		info.minor = MINOR(rdev->bdev->bd_dev);
		info.raid_disk = rdev->raid_disk;
		info.state = 0;
3607
		if (test_bit(Faulty, &rdev->flags))
L
Linus Torvalds 已提交
3608
			info.state |= (1<<MD_DISK_FAULTY);
3609
		else if (test_bit(In_sync, &rdev->flags)) {
L
Linus Torvalds 已提交
3610 3611 3612
			info.state |= (1<<MD_DISK_ACTIVE);
			info.state |= (1<<MD_DISK_SYNC);
		}
3613 3614
		if (test_bit(WriteMostly, &rdev->flags))
			info.state |= (1<<MD_DISK_WRITEMOSTLY);
L
Linus Torvalds 已提交
3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678
	} else {
		info.major = info.minor = 0;
		info.raid_disk = -1;
		info.state = (1<<MD_DISK_REMOVED);
	}

	if (copy_to_user(arg, &info, sizeof(info)))
		return -EFAULT;

	return 0;
}

static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
{
	char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE];
	mdk_rdev_t *rdev;
	dev_t dev = MKDEV(info->major,info->minor);

	if (info->major != MAJOR(dev) || info->minor != MINOR(dev))
		return -EOVERFLOW;

	if (!mddev->raid_disks) {
		int err;
		/* expecting a device which has a superblock */
		rdev = md_import_device(dev, mddev->major_version, mddev->minor_version);
		if (IS_ERR(rdev)) {
			printk(KERN_WARNING 
				"md: md_import_device returned %ld\n",
				PTR_ERR(rdev));
			return PTR_ERR(rdev);
		}
		if (!list_empty(&mddev->disks)) {
			mdk_rdev_t *rdev0 = list_entry(mddev->disks.next,
							mdk_rdev_t, same_set);
			int err = super_types[mddev->major_version]
				.load_super(rdev, rdev0, mddev->minor_version);
			if (err < 0) {
				printk(KERN_WARNING 
					"md: %s has different UUID to %s\n",
					bdevname(rdev->bdev,b), 
					bdevname(rdev0->bdev,b2));
				export_rdev(rdev);
				return -EINVAL;
			}
		}
		err = bind_rdev_to_array(rdev, mddev);
		if (err)
			export_rdev(rdev);
		return err;
	}

	/*
	 * add_new_disk can be used once the array is assembled
	 * to add "hot spares".  They must already have a superblock
	 * written
	 */
	if (mddev->pers) {
		int err;
		if (!mddev->pers->hot_add_disk) {
			printk(KERN_WARNING 
				"%s: personality does not support diskops!\n",
			       mdname(mddev));
			return -EINVAL;
		}
3679 3680 3681 3682 3683
		if (mddev->persistent)
			rdev = md_import_device(dev, mddev->major_version,
						mddev->minor_version);
		else
			rdev = md_import_device(dev, -1, -1);
L
Linus Torvalds 已提交
3684 3685 3686 3687 3688 3689
		if (IS_ERR(rdev)) {
			printk(KERN_WARNING 
				"md: md_import_device returned %ld\n",
				PTR_ERR(rdev));
			return PTR_ERR(rdev);
		}
3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701
		/* set save_raid_disk if appropriate */
		if (!mddev->persistent) {
			if (info->state & (1<<MD_DISK_SYNC)  &&
			    info->raid_disk < mddev->raid_disks)
				rdev->raid_disk = info->raid_disk;
			else
				rdev->raid_disk = -1;
		} else
			super_types[mddev->major_version].
				validate_super(mddev, rdev);
		rdev->saved_raid_disk = rdev->raid_disk;

3702
		clear_bit(In_sync, &rdev->flags); /* just to be sure */
3703 3704 3705
		if (info->state & (1<<MD_DISK_WRITEMOSTLY))
			set_bit(WriteMostly, &rdev->flags);

L
Linus Torvalds 已提交
3706 3707
		rdev->raid_disk = -1;
		err = bind_rdev_to_array(rdev, mddev);
3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718
		if (!err && !mddev->pers->hot_remove_disk) {
			/* If there is hot_add_disk but no hot_remove_disk
			 * then added disks for geometry changes,
			 * and should be added immediately.
			 */
			super_types[mddev->major_version].
				validate_super(mddev, rdev);
			err = mddev->pers->hot_add_disk(mddev, rdev);
			if (err)
				unbind_rdev_from_array(rdev);
		}
L
Linus Torvalds 已提交
3719 3720
		if (err)
			export_rdev(rdev);
3721 3722

		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
3723
		md_wakeup_thread(mddev->thread);
L
Linus Torvalds 已提交
3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750
		return err;
	}

	/* otherwise, add_new_disk is only allowed
	 * for major_version==0 superblocks
	 */
	if (mddev->major_version != 0) {
		printk(KERN_WARNING "%s: ADD_NEW_DISK not supported\n",
		       mdname(mddev));
		return -EINVAL;
	}

	if (!(info->state & (1<<MD_DISK_FAULTY))) {
		int err;
		rdev = md_import_device (dev, -1, 0);
		if (IS_ERR(rdev)) {
			printk(KERN_WARNING 
				"md: error, md_import_device() returned %ld\n",
				PTR_ERR(rdev));
			return PTR_ERR(rdev);
		}
		rdev->desc_nr = info->number;
		if (info->raid_disk < mddev->raid_disks)
			rdev->raid_disk = info->raid_disk;
		else
			rdev->raid_disk = -1;

3751 3752
		rdev->flags = 0;

L
Linus Torvalds 已提交
3753
		if (rdev->raid_disk < mddev->raid_disks)
3754 3755
			if (info->state & (1<<MD_DISK_SYNC))
				set_bit(In_sync, &rdev->flags);
L
Linus Torvalds 已提交
3756

3757 3758 3759
		if (info->state & (1<<MD_DISK_WRITEMOSTLY))
			set_bit(WriteMostly, &rdev->flags);

L
Linus Torvalds 已提交
3760 3761 3762 3763 3764 3765 3766
		if (!mddev->persistent) {
			printk(KERN_INFO "md: nonpersistent superblock ...\n");
			rdev->sb_offset = rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
		} else 
			rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
		rdev->size = calc_dev_size(rdev, mddev->chunk_size);

3767 3768 3769 3770 3771
		err = bind_rdev_to_array(rdev, mddev);
		if (err) {
			export_rdev(rdev);
			return err;
		}
L
Linus Torvalds 已提交
3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792
	}

	return 0;
}

static int hot_remove_disk(mddev_t * mddev, dev_t dev)
{
	char b[BDEVNAME_SIZE];
	mdk_rdev_t *rdev;

	if (!mddev->pers)
		return -ENODEV;

	rdev = find_rdev(mddev, dev);
	if (!rdev)
		return -ENXIO;

	if (rdev->raid_disk >= 0)
		goto busy;

	kick_rdev_from_array(rdev);
3793
	md_update_sb(mddev, 1);
3794
	md_new_event(mddev);
L
Linus Torvalds 已提交
3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842

	return 0;
busy:
	printk(KERN_WARNING "md: cannot remove active disk %s from %s ... \n",
		bdevname(rdev->bdev,b), mdname(mddev));
	return -EBUSY;
}

static int hot_add_disk(mddev_t * mddev, dev_t dev)
{
	char b[BDEVNAME_SIZE];
	int err;
	unsigned int size;
	mdk_rdev_t *rdev;

	if (!mddev->pers)
		return -ENODEV;

	if (mddev->major_version != 0) {
		printk(KERN_WARNING "%s: HOT_ADD may only be used with"
			" version-0 superblocks.\n",
			mdname(mddev));
		return -EINVAL;
	}
	if (!mddev->pers->hot_add_disk) {
		printk(KERN_WARNING 
			"%s: personality does not support diskops!\n",
			mdname(mddev));
		return -EINVAL;
	}

	rdev = md_import_device (dev, -1, 0);
	if (IS_ERR(rdev)) {
		printk(KERN_WARNING 
			"md: error, md_import_device() returned %ld\n",
			PTR_ERR(rdev));
		return -EINVAL;
	}

	if (mddev->persistent)
		rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
	else
		rdev->sb_offset =
			rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;

	size = calc_dev_size(rdev, mddev->chunk_size);
	rdev->size = size;

3843
	if (test_bit(Faulty, &rdev->flags)) {
L
Linus Torvalds 已提交
3844 3845 3846 3847 3848 3849
		printk(KERN_WARNING 
			"md: can not hot-add faulty %s disk to %s!\n",
			bdevname(rdev->bdev,b), mdname(mddev));
		err = -EINVAL;
		goto abort_export;
	}
3850
	clear_bit(In_sync, &rdev->flags);
L
Linus Torvalds 已提交
3851
	rdev->desc_nr = -1;
3852 3853 3854
	err = bind_rdev_to_array(rdev, mddev);
	if (err)
		goto abort_export;
L
Linus Torvalds 已提交
3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869

	/*
	 * The rest should better be atomic, we can have disk failures
	 * noticed in interrupt contexts ...
	 */

	if (rdev->desc_nr == mddev->max_disks) {
		printk(KERN_WARNING "%s: can not hot-add to full array!\n",
			mdname(mddev));
		err = -EBUSY;
		goto abort_unbind_export;
	}

	rdev->raid_disk = -1;

3870
	md_update_sb(mddev, 1);
L
Linus Torvalds 已提交
3871 3872 3873 3874 3875 3876 3877

	/*
	 * Kick recovery, maybe this spare has to be added to the
	 * array immediately.
	 */
	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	md_wakeup_thread(mddev->thread);
3878
	md_new_event(mddev);
L
Linus Torvalds 已提交
3879 3880 3881 3882 3883 3884 3885 3886 3887 3888
	return 0;

abort_unbind_export:
	unbind_rdev_from_array(rdev);

abort_export:
	export_rdev(rdev);
	return err;
}

3889 3890 3891 3892
static int set_bitmap_file(mddev_t *mddev, int fd)
{
	int err;

3893 3894 3895 3896 3897 3898 3899
	if (mddev->pers) {
		if (!mddev->pers->quiesce)
			return -EBUSY;
		if (mddev->recovery || mddev->sync_thread)
			return -EBUSY;
		/* we should be able to change the bitmap.. */
	}
3900 3901


3902 3903 3904 3905
	if (fd >= 0) {
		if (mddev->bitmap)
			return -EEXIST; /* cannot add when bitmap is present */
		mddev->bitmap_file = fget(fd);
3906

3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920
		if (mddev->bitmap_file == NULL) {
			printk(KERN_ERR "%s: error: failed to get bitmap file\n",
			       mdname(mddev));
			return -EBADF;
		}

		err = deny_bitmap_write_access(mddev->bitmap_file);
		if (err) {
			printk(KERN_ERR "%s: error: bitmap file is already in use\n",
			       mdname(mddev));
			fput(mddev->bitmap_file);
			mddev->bitmap_file = NULL;
			return err;
		}
3921
		mddev->bitmap_offset = 0; /* file overrides offset */
3922 3923 3924 3925 3926 3927 3928
	} else if (mddev->bitmap == NULL)
		return -ENOENT; /* cannot remove what isn't there */
	err = 0;
	if (mddev->pers) {
		mddev->pers->quiesce(mddev, 1);
		if (fd >= 0)
			err = bitmap_create(mddev);
3929
		if (fd < 0 || err) {
3930
			bitmap_destroy(mddev);
3931 3932
			fd = -1; /* make sure to put the file */
		}
3933
		mddev->pers->quiesce(mddev, 0);
3934 3935
	}
	if (fd < 0) {
3936 3937
		if (mddev->bitmap_file) {
			restore_bitmap_write_access(mddev->bitmap_file);
3938
			fput(mddev->bitmap_file);
3939
		}
3940 3941 3942
		mddev->bitmap_file = NULL;
	}

3943 3944 3945
	return err;
}

L
Linus Torvalds 已提交
3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983
/*
 * set_array_info is used two different ways
 * The original usage is when creating a new array.
 * In this usage, raid_disks is > 0 and it together with
 *  level, size, not_persistent,layout,chunksize determine the
 *  shape of the array.
 *  This will always create an array with a type-0.90.0 superblock.
 * The newer usage is when assembling an array.
 *  In this case raid_disks will be 0, and the major_version field is
 *  use to determine which style super-blocks are to be found on the devices.
 *  The minor and patch _version numbers are also kept incase the
 *  super_block handler wishes to interpret them.
 */
static int set_array_info(mddev_t * mddev, mdu_array_info_t *info)
{

	if (info->raid_disks == 0) {
		/* just setting version number for superblock loading */
		if (info->major_version < 0 ||
		    info->major_version >= sizeof(super_types)/sizeof(super_types[0]) ||
		    super_types[info->major_version].name == NULL) {
			/* maybe try to auto-load a module? */
			printk(KERN_INFO 
				"md: superblock version %d not known\n",
				info->major_version);
			return -EINVAL;
		}
		mddev->major_version = info->major_version;
		mddev->minor_version = info->minor_version;
		mddev->patch_version = info->patch_version;
		return 0;
	}
	mddev->major_version = MD_MAJOR_VERSION;
	mddev->minor_version = MD_MINOR_VERSION;
	mddev->patch_version = MD_PATCHLEVEL_VERSION;
	mddev->ctime         = get_seconds();

	mddev->level         = info->level;
3984
	mddev->clevel[0]     = 0;
L
Linus Torvalds 已提交
3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000
	mddev->size          = info->size;
	mddev->raid_disks    = info->raid_disks;
	/* don't set md_minor, it is determined by which /dev/md* was
	 * openned
	 */
	if (info->state & (1<<MD_SB_CLEAN))
		mddev->recovery_cp = MaxSector;
	else
		mddev->recovery_cp = 0;
	mddev->persistent    = ! info->not_persistent;

	mddev->layout        = info->layout;
	mddev->chunk_size    = info->chunk_size;

	mddev->max_disks     = MD_SB_DISKS;

4001 4002
	mddev->flags         = 0;
	set_bit(MD_CHANGE_DEVS, &mddev->flags);
L
Linus Torvalds 已提交
4003

4004 4005 4006
	mddev->default_bitmap_offset = MD_SB_BYTES >> 9;
	mddev->bitmap_offset = 0;

4007 4008
	mddev->reshape_position = MaxSector;

L
Linus Torvalds 已提交
4009 4010 4011 4012 4013
	/*
	 * Generate a 128 bit UUID
	 */
	get_random_bytes(mddev->uuid, 16);

4014 4015 4016 4017 4018
	mddev->new_level = mddev->level;
	mddev->new_chunk = mddev->chunk_size;
	mddev->new_layout = mddev->layout;
	mddev->delta_disks = 0;

L
Linus Torvalds 已提交
4019 4020 4021
	return 0;
}

4022 4023 4024 4025 4026
static int update_size(mddev_t *mddev, unsigned long size)
{
	mdk_rdev_t * rdev;
	int rv;
	struct list_head *tmp;
4027
	int fit = (size == 0);
4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060

	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;
		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) {
4061
			mutex_lock(&bdev->bd_inode->i_mutex);
4062
			i_size_write(bdev->bd_inode, (loff_t)mddev->array_size << 10);
4063
			mutex_unlock(&bdev->bd_inode->i_mutex);
4064 4065 4066 4067 4068 4069
			bdput(bdev);
		}
	}
	return rv;
}

4070 4071 4072 4073
static int update_raid_disks(mddev_t *mddev, int raid_disks)
{
	int rv;
	/* change the number of raid disks */
4074
	if (mddev->pers->check_reshape == NULL)
4075 4076 4077 4078
		return -EINVAL;
	if (raid_disks <= 0 ||
	    raid_disks >= mddev->max_disks)
		return -EINVAL;
4079
	if (mddev->sync_thread || mddev->reshape_position != MaxSector)
4080
		return -EBUSY;
4081 4082 4083
	mddev->delta_disks = raid_disks - mddev->raid_disks;

	rv = mddev->pers->check_reshape(mddev);
4084 4085 4086 4087
	return rv;
}


L
Linus Torvalds 已提交
4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099
/*
 * update_array_info is used to change the configuration of an
 * on-line array.
 * The version, ctime,level,size,raid_disks,not_persistent, layout,chunk_size
 * fields in the info are checked against the array.
 * Any differences that cannot be handled will cause an error.
 * Normally, only one change can be managed at a time.
 */
static int update_array_info(mddev_t *mddev, mdu_array_info_t *info)
{
	int rv = 0;
	int cnt = 0;
4100 4101 4102 4103 4104
	int state = 0;

	/* calculate expected state,ignoring low bits */
	if (mddev->bitmap && mddev->bitmap_offset)
		state |= (1 << MD_SB_BITMAP_PRESENT);
L
Linus Torvalds 已提交
4105 4106 4107 4108 4109 4110 4111 4112

	if (mddev->major_version != info->major_version ||
	    mddev->minor_version != info->minor_version ||
/*	    mddev->patch_version != info->patch_version || */
	    mddev->ctime         != info->ctime         ||
	    mddev->level         != info->level         ||
/*	    mddev->layout        != info->layout        || */
	    !mddev->persistent	 != info->not_persistent||
4113 4114 4115 4116
	    mddev->chunk_size    != info->chunk_size    ||
	    /* ignore bottom 8 bits of state, and allow SB_BITMAP_PRESENT to change */
	    ((state^info->state) & 0xfffffe00)
		)
L
Linus Torvalds 已提交
4117 4118
		return -EINVAL;
	/* Check there is only one change */
4119
	if (info->size >= 0 && mddev->size != info->size) cnt++;
L
Linus Torvalds 已提交
4120 4121
	if (mddev->raid_disks != info->raid_disks) cnt++;
	if (mddev->layout != info->layout) cnt++;
4122
	if ((state ^ info->state) & (1<<MD_SB_BITMAP_PRESENT)) cnt++;
L
Linus Torvalds 已提交
4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135
	if (cnt == 0) return 0;
	if (cnt > 1) return -EINVAL;

	if (mddev->layout != info->layout) {
		/* Change layout
		 * we don't need to do anything at the md level, the
		 * personality will take care of it all.
		 */
		if (mddev->pers->reconfig == NULL)
			return -EINVAL;
		else
			return mddev->pers->reconfig(mddev, info->layout, -1);
	}
4136
	if (info->size >= 0 && mddev->size != info->size)
4137 4138
		rv = update_size(mddev, info->size);

4139 4140 4141
	if (mddev->raid_disks    != info->raid_disks)
		rv = update_raid_disks(mddev, info->raid_disks);

4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170
	if ((state ^ info->state) & (1<<MD_SB_BITMAP_PRESENT)) {
		if (mddev->pers->quiesce == NULL)
			return -EINVAL;
		if (mddev->recovery || mddev->sync_thread)
			return -EBUSY;
		if (info->state & (1<<MD_SB_BITMAP_PRESENT)) {
			/* add the bitmap */
			if (mddev->bitmap)
				return -EEXIST;
			if (mddev->default_bitmap_offset == 0)
				return -EINVAL;
			mddev->bitmap_offset = mddev->default_bitmap_offset;
			mddev->pers->quiesce(mddev, 1);
			rv = bitmap_create(mddev);
			if (rv)
				bitmap_destroy(mddev);
			mddev->pers->quiesce(mddev, 0);
		} else {
			/* remove the bitmap */
			if (!mddev->bitmap)
				return -ENOENT;
			if (mddev->bitmap->file)
				return -EINVAL;
			mddev->pers->quiesce(mddev, 1);
			bitmap_destroy(mddev);
			mddev->pers->quiesce(mddev, 0);
			mddev->bitmap_offset = 0;
		}
	}
4171
	md_update_sb(mddev, 1);
L
Linus Torvalds 已提交
4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189
	return rv;
}

static int set_disk_faulty(mddev_t *mddev, dev_t dev)
{
	mdk_rdev_t *rdev;

	if (mddev->pers == NULL)
		return -ENODEV;

	rdev = find_rdev(mddev, dev);
	if (!rdev)
		return -ENODEV;

	md_error(mddev, rdev);
	return 0;
}

4190 4191 4192 4193 4194 4195 4196 4197 4198 4199
static int md_getgeo(struct block_device *bdev, struct hd_geometry *geo)
{
	mddev_t *mddev = bdev->bd_disk->private_data;

	geo->heads = 2;
	geo->sectors = 4;
	geo->cylinders = get_capacity(mddev->gendisk) / 8;
	return 0;
}

L
Linus Torvalds 已提交
4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301
static int md_ioctl(struct inode *inode, struct file *file,
			unsigned int cmd, unsigned long arg)
{
	int err = 0;
	void __user *argp = (void __user *)arg;
	mddev_t *mddev = NULL;

	if (!capable(CAP_SYS_ADMIN))
		return -EACCES;

	/*
	 * Commands dealing with the RAID driver but not any
	 * particular array:
	 */
	switch (cmd)
	{
		case RAID_VERSION:
			err = get_version(argp);
			goto done;

		case PRINT_RAID_DEBUG:
			err = 0;
			md_print_devices();
			goto done;

#ifndef MODULE
		case RAID_AUTORUN:
			err = 0;
			autostart_arrays(arg);
			goto done;
#endif
		default:;
	}

	/*
	 * Commands creating/starting a new array:
	 */

	mddev = inode->i_bdev->bd_disk->private_data;

	if (!mddev) {
		BUG();
		goto abort;
	}

	err = mddev_lock(mddev);
	if (err) {
		printk(KERN_INFO 
			"md: ioctl lock interrupted, reason %d, cmd %d\n",
			err, cmd);
		goto abort;
	}

	switch (cmd)
	{
		case SET_ARRAY_INFO:
			{
				mdu_array_info_t info;
				if (!arg)
					memset(&info, 0, sizeof(info));
				else if (copy_from_user(&info, argp, sizeof(info))) {
					err = -EFAULT;
					goto abort_unlock;
				}
				if (mddev->pers) {
					err = update_array_info(mddev, &info);
					if (err) {
						printk(KERN_WARNING "md: couldn't update"
						       " array info. %d\n", err);
						goto abort_unlock;
					}
					goto done_unlock;
				}
				if (!list_empty(&mddev->disks)) {
					printk(KERN_WARNING
					       "md: array %s already has disks!\n",
					       mdname(mddev));
					err = -EBUSY;
					goto abort_unlock;
				}
				if (mddev->raid_disks) {
					printk(KERN_WARNING
					       "md: array %s already initialised!\n",
					       mdname(mddev));
					err = -EBUSY;
					goto abort_unlock;
				}
				err = set_array_info(mddev, &info);
				if (err) {
					printk(KERN_WARNING "md: couldn't set"
					       " array info. %d\n", err);
					goto abort_unlock;
				}
			}
			goto done_unlock;

		default:;
	}

	/*
	 * Commands querying/configuring an existing array:
	 */
4302 4303 4304 4305
	/* if we are not initialised yet, only ADD_NEW_DISK, STOP_ARRAY,
	 * RUN_ARRAY, and SET_BITMAP_FILE are allowed */
	if (!mddev->raid_disks && cmd != ADD_NEW_DISK && cmd != STOP_ARRAY
			&& cmd != RUN_ARRAY && cmd != SET_BITMAP_FILE) {
L
Linus Torvalds 已提交
4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318
		err = -ENODEV;
		goto abort_unlock;
	}

	/*
	 * Commands even a read-only array can execute:
	 */
	switch (cmd)
	{
		case GET_ARRAY_INFO:
			err = get_array_info(mddev, argp);
			goto done_unlock;

4319
		case GET_BITMAP_FILE:
4320
			err = get_bitmap_file(mddev, argp);
4321 4322
			goto done_unlock;

L
Linus Torvalds 已提交
4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348
		case GET_DISK_INFO:
			err = get_disk_info(mddev, argp);
			goto done_unlock;

		case RESTART_ARRAY_RW:
			err = restart_array(mddev);
			goto done_unlock;

		case STOP_ARRAY:
			err = do_md_stop (mddev, 0);
			goto done_unlock;

		case STOP_ARRAY_RO:
			err = do_md_stop (mddev, 1);
			goto done_unlock;

	/*
	 * We have a problem here : there is no easy way to give a CHS
	 * virtual geometry. We currently pretend that we have a 2 heads
	 * 4 sectors (with a BIG number of cylinders...). This drives
	 * dosfs just mad... ;-)
	 */
	}

	/*
	 * The remaining ioctls are changing the state of the
4349 4350 4351 4352
	 * superblock, so we do not allow them on read-only arrays.
	 * However non-MD ioctls (e.g. get-size) will still come through
	 * here and hit the 'default' below, so only disallow
	 * 'md' ioctls, and switch to rw mode if started auto-readonly.
L
Linus Torvalds 已提交
4353
	 */
4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364
	if (_IOC_TYPE(cmd) == MD_MAJOR &&
	    mddev->ro && mddev->pers) {
		if (mddev->ro == 2) {
			mddev->ro = 0;
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
		md_wakeup_thread(mddev->thread);

		} else {
			err = -EROFS;
			goto abort_unlock;
		}
L
Linus Torvalds 已提交
4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394
	}

	switch (cmd)
	{
		case ADD_NEW_DISK:
		{
			mdu_disk_info_t info;
			if (copy_from_user(&info, argp, sizeof(info)))
				err = -EFAULT;
			else
				err = add_new_disk(mddev, &info);
			goto done_unlock;
		}

		case HOT_REMOVE_DISK:
			err = hot_remove_disk(mddev, new_decode_dev(arg));
			goto done_unlock;

		case HOT_ADD_DISK:
			err = hot_add_disk(mddev, new_decode_dev(arg));
			goto done_unlock;

		case SET_DISK_FAULTY:
			err = set_disk_faulty(mddev, new_decode_dev(arg));
			goto done_unlock;

		case RUN_ARRAY:
			err = do_md_run (mddev);
			goto done_unlock;

4395 4396 4397 4398
		case SET_BITMAP_FILE:
			err = set_bitmap_file(mddev, (int)arg);
			goto done_unlock;

L
Linus Torvalds 已提交
4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467
		default:
			err = -EINVAL;
			goto abort_unlock;
	}

done_unlock:
abort_unlock:
	mddev_unlock(mddev);

	return err;
done:
	if (err)
		MD_BUG();
abort:
	return err;
}

static int md_open(struct inode *inode, struct file *file)
{
	/*
	 * Succeed if we can lock the mddev, which confirms that
	 * it isn't being stopped right now.
	 */
	mddev_t *mddev = inode->i_bdev->bd_disk->private_data;
	int err;

	if ((err = mddev_lock(mddev)))
		goto out;

	err = 0;
	mddev_get(mddev);
	mddev_unlock(mddev);

	check_disk_change(inode->i_bdev);
 out:
	return err;
}

static int md_release(struct inode *inode, struct file * file)
{
 	mddev_t *mddev = inode->i_bdev->bd_disk->private_data;

	if (!mddev)
		BUG();
	mddev_put(mddev);

	return 0;
}

static int md_media_changed(struct gendisk *disk)
{
	mddev_t *mddev = disk->private_data;

	return mddev->changed;
}

static int md_revalidate(struct gendisk *disk)
{
	mddev_t *mddev = disk->private_data;

	mddev->changed = 0;
	return 0;
}
static struct block_device_operations md_fops =
{
	.owner		= THIS_MODULE,
	.open		= md_open,
	.release	= md_release,
	.ioctl		= md_ioctl,
4468
	.getgeo		= md_getgeo,
L
Linus Torvalds 已提交
4469 4470 4471 4472
	.media_changed	= md_media_changed,
	.revalidate_disk= md_revalidate,
};

A
Adrian Bunk 已提交
4473
static int md_thread(void * arg)
L
Linus Torvalds 已提交
4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488
{
	mdk_thread_t *thread = arg;

	/*
	 * md_thread is a 'system-thread', it's priority should be very
	 * high. We avoid resource deadlocks individually in each
	 * raid personality. (RAID5 does preallocation) We also use RR and
	 * the very same RT priority as kswapd, thus we will never get
	 * into a priority inversion deadlock.
	 *
	 * we definitely have to have equal or higher priority than
	 * bdflush, otherwise bdflush will deadlock if there are too
	 * many dirty RAID5 blocks.
	 */

N
NeilBrown 已提交
4489
	allow_signal(SIGKILL);
4490
	while (!kthread_should_stop()) {
L
Linus Torvalds 已提交
4491

4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504
		/* We need to wait INTERRUPTIBLE so that
		 * we don't add to the load-average.
		 * That means we need to be sure no signals are
		 * pending
		 */
		if (signal_pending(current))
			flush_signals(current);

		wait_event_interruptible_timeout
			(thread->wqueue,
			 test_bit(THREAD_WAKEUP, &thread->flags)
			 || kthread_should_stop(),
			 thread->timeout);
4505
		try_to_freeze();
L
Linus Torvalds 已提交
4506 4507 4508

		clear_bit(THREAD_WAKEUP, &thread->flags);

4509
		thread->run(thread->mddev);
L
Linus Torvalds 已提交
4510
	}
4511

L
Linus Torvalds 已提交
4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528
	return 0;
}

void md_wakeup_thread(mdk_thread_t *thread)
{
	if (thread) {
		dprintk("md: waking up MD thread %s.\n", thread->tsk->comm);
		set_bit(THREAD_WAKEUP, &thread->flags);
		wake_up(&thread->wqueue);
	}
}

mdk_thread_t *md_register_thread(void (*run) (mddev_t *), mddev_t *mddev,
				 const char *name)
{
	mdk_thread_t *thread;

4529
	thread = kzalloc(sizeof(mdk_thread_t), GFP_KERNEL);
L
Linus Torvalds 已提交
4530 4531 4532 4533 4534 4535 4536
	if (!thread)
		return NULL;

	init_waitqueue_head(&thread->wqueue);

	thread->run = run;
	thread->mddev = mddev;
4537
	thread->timeout = MAX_SCHEDULE_TIMEOUT;
N
NeilBrown 已提交
4538
	thread->tsk = kthread_run(md_thread, thread, name, mdname(thread->mddev));
4539
	if (IS_ERR(thread->tsk)) {
L
Linus Torvalds 已提交
4540 4541 4542 4543 4544 4545 4546 4547
		kfree(thread);
		return NULL;
	}
	return thread;
}

void md_unregister_thread(mdk_thread_t *thread)
{
4548
	dprintk("interrupting MD-thread pid %d\n", thread->tsk->pid);
4549 4550

	kthread_stop(thread->tsk);
L
Linus Torvalds 已提交
4551 4552 4553 4554 4555 4556 4557 4558 4559 4560
	kfree(thread);
}

void md_error(mddev_t *mddev, mdk_rdev_t *rdev)
{
	if (!mddev) {
		MD_BUG();
		return;
	}

4561
	if (!rdev || test_bit(Faulty, &rdev->flags))
L
Linus Torvalds 已提交
4562
		return;
4563
/*
L
Linus Torvalds 已提交
4564 4565 4566 4567 4568
	dprintk("md_error dev:%s, rdev:(%d:%d), (caller: %p,%p,%p,%p).\n",
		mdname(mddev),
		MAJOR(rdev->bdev->bd_dev), MINOR(rdev->bdev->bd_dev),
		__builtin_return_address(0),__builtin_return_address(1),
		__builtin_return_address(2),__builtin_return_address(3));
4569
*/
4570 4571
	if (!mddev->pers)
		return;
L
Linus Torvalds 已提交
4572 4573 4574 4575 4576 4577
	if (!mddev->pers->error_handler)
		return;
	mddev->pers->error_handler(mddev,rdev);
	set_bit(MD_RECOVERY_INTR, &mddev->recovery);
	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	md_wakeup_thread(mddev->thread);
4578
	md_new_event_inintr(mddev);
L
Linus Torvalds 已提交
4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605
}

/* seq_file implementation /proc/mdstat */

static void status_unused(struct seq_file *seq)
{
	int i = 0;
	mdk_rdev_t *rdev;
	struct list_head *tmp;

	seq_printf(seq, "unused devices: ");

	ITERATE_RDEV_PENDING(rdev,tmp) {
		char b[BDEVNAME_SIZE];
		i++;
		seq_printf(seq, "%s ",
			      bdevname(rdev->bdev,b));
	}
	if (!i)
		seq_printf(seq, "<none>");

	seq_printf(seq, "\n");
}


static void status_resync(struct seq_file *seq, mddev_t * mddev)
{
4606 4607 4608 4609
	sector_t max_blocks, resync, res;
	unsigned long dt, db, rt;
	int scale;
	unsigned int per_milli;
L
Linus Torvalds 已提交
4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624

	resync = (mddev->curr_resync - atomic_read(&mddev->recovery_active))/2;

	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
		max_blocks = mddev->resync_max_sectors >> 1;
	else
		max_blocks = mddev->size;

	/*
	 * Should not happen.
	 */
	if (!max_blocks) {
		MD_BUG();
		return;
	}
4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638
	/* Pick 'scale' such that (resync>>scale)*1000 will fit
	 * in a sector_t, and (max_blocks>>scale) will fit in a
	 * u32, as those are the requirements for sector_div.
	 * Thus 'scale' must be at least 10
	 */
	scale = 10;
	if (sizeof(sector_t) > sizeof(unsigned long)) {
		while ( max_blocks/2 > (1ULL<<(scale+32)))
			scale++;
	}
	res = (resync>>scale)*1000;
	sector_div(res, (u32)((max_blocks>>scale)+1));

	per_milli = res;
L
Linus Torvalds 已提交
4639
	{
4640
		int i, x = per_milli/50, y = 20-x;
L
Linus Torvalds 已提交
4641 4642 4643 4644 4645 4646 4647 4648
		seq_printf(seq, "[");
		for (i = 0; i < x; i++)
			seq_printf(seq, "=");
		seq_printf(seq, ">");
		for (i = 0; i < y; i++)
			seq_printf(seq, ".");
		seq_printf(seq, "] ");
	}
4649
	seq_printf(seq, " %s =%3u.%u%% (%llu/%llu)",
4650 4651
		   (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery)?
		    "reshape" :
4652 4653 4654 4655 4656
		    (test_bit(MD_RECOVERY_CHECK, &mddev->recovery)?
		     "check" :
		     (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) ?
		      "resync" : "recovery"))),
		   per_milli/10, per_milli % 10,
4657 4658
		   (unsigned long long) resync,
		   (unsigned long long) max_blocks);
L
Linus Torvalds 已提交
4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670

	/*
	 * We do not want to overflow, so the order of operands and
	 * the * 100 / 100 trick are important. We do a +1 to be
	 * safe against division by zero. We only estimate anyway.
	 *
	 * dt: time from mark until now
	 * db: blocks written from mark until now
	 * rt: remaining time
	 */
	dt = ((jiffies - mddev->resync_mark) / HZ);
	if (!dt) dt++;
4671 4672 4673
	db = (mddev->curr_mark_cnt - atomic_read(&mddev->recovery_active))
		- mddev->resync_mark_cnt;
	rt = (dt * ((unsigned long)(max_blocks-resync) / (db/2/100+1)))/100;
L
Linus Torvalds 已提交
4674 4675 4676

	seq_printf(seq, " finish=%lu.%lumin", rt / 60, (rt % 60)/6);

4677
	seq_printf(seq, " speed=%ldK/sec", db/2/dt);
L
Linus Torvalds 已提交
4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741
}

static void *md_seq_start(struct seq_file *seq, loff_t *pos)
{
	struct list_head *tmp;
	loff_t l = *pos;
	mddev_t *mddev;

	if (l >= 0x10000)
		return NULL;
	if (!l--)
		/* header */
		return (void*)1;

	spin_lock(&all_mddevs_lock);
	list_for_each(tmp,&all_mddevs)
		if (!l--) {
			mddev = list_entry(tmp, mddev_t, all_mddevs);
			mddev_get(mddev);
			spin_unlock(&all_mddevs_lock);
			return mddev;
		}
	spin_unlock(&all_mddevs_lock);
	if (!l--)
		return (void*)2;/* tail */
	return NULL;
}

static void *md_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
	struct list_head *tmp;
	mddev_t *next_mddev, *mddev = v;
	
	++*pos;
	if (v == (void*)2)
		return NULL;

	spin_lock(&all_mddevs_lock);
	if (v == (void*)1)
		tmp = all_mddevs.next;
	else
		tmp = mddev->all_mddevs.next;
	if (tmp != &all_mddevs)
		next_mddev = mddev_get(list_entry(tmp,mddev_t,all_mddevs));
	else {
		next_mddev = (void*)2;
		*pos = 0x10000;
	}		
	spin_unlock(&all_mddevs_lock);

	if (v != (void*)1)
		mddev_put(mddev);
	return next_mddev;

}

static void md_seq_stop(struct seq_file *seq, void *v)
{
	mddev_t *mddev = v;

	if (mddev && v != (void*)1 && v != (void*)2)
		mddev_put(mddev);
}

4742 4743 4744 4745
struct mdstat_info {
	int event;
};

L
Linus Torvalds 已提交
4746 4747 4748 4749 4750 4751
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;
4752
	struct mdstat_info *mi = seq->private;
4753
	struct bitmap *bitmap;
L
Linus Torvalds 已提交
4754 4755

	if (v == (void*)1) {
4756
		struct mdk_personality *pers;
L
Linus Torvalds 已提交
4757 4758
		seq_printf(seq, "Personalities : ");
		spin_lock(&pers_lock);
4759 4760
		list_for_each_entry(pers, &pers_list, list)
			seq_printf(seq, "[%s] ", pers->name);
L
Linus Torvalds 已提交
4761 4762 4763

		spin_unlock(&pers_lock);
		seq_printf(seq, "\n");
4764
		mi->event = atomic_read(&md_event_count);
L
Linus Torvalds 已提交
4765 4766 4767 4768 4769 4770 4771
		return 0;
	}
	if (v == (void*)2) {
		status_unused(seq);
		return 0;
	}

I
Ingo Molnar 已提交
4772
	if (mddev_lock(mddev) < 0)
L
Linus Torvalds 已提交
4773
		return -EINTR;
I
Ingo Molnar 已提交
4774

L
Linus Torvalds 已提交
4775 4776 4777 4778
	if (mddev->pers || mddev->raid_disks || !list_empty(&mddev->disks)) {
		seq_printf(seq, "%s : %sactive", mdname(mddev),
						mddev->pers ? "" : "in");
		if (mddev->pers) {
4779
			if (mddev->ro==1)
L
Linus Torvalds 已提交
4780
				seq_printf(seq, " (read-only)");
4781 4782
			if (mddev->ro==2)
				seq_printf(seq, "(auto-read-only)");
L
Linus Torvalds 已提交
4783 4784 4785 4786 4787 4788 4789 4790
			seq_printf(seq, " %s", mddev->pers->name);
		}

		size = 0;
		ITERATE_RDEV(mddev,rdev,tmp2) {
			char b[BDEVNAME_SIZE];
			seq_printf(seq, " %s[%d]",
				bdevname(rdev->bdev,b), rdev->desc_nr);
4791 4792
			if (test_bit(WriteMostly, &rdev->flags))
				seq_printf(seq, "(W)");
4793
			if (test_bit(Faulty, &rdev->flags)) {
L
Linus Torvalds 已提交
4794 4795
				seq_printf(seq, "(F)");
				continue;
4796 4797
			} else if (rdev->raid_disk < 0)
				seq_printf(seq, "(S)"); /* spare */
L
Linus Torvalds 已提交
4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808
			size += rdev->size;
		}

		if (!list_empty(&mddev->disks)) {
			if (mddev->pers)
				seq_printf(seq, "\n      %llu blocks",
					(unsigned long long)mddev->array_size);
			else
				seq_printf(seq, "\n      %llu blocks",
					(unsigned long long)size);
		}
4809 4810 4811 4812 4813 4814 4815 4816 4817
		if (mddev->persistent) {
			if (mddev->major_version != 0 ||
			    mddev->minor_version != 90) {
				seq_printf(seq," super %d.%d",
					   mddev->major_version,
					   mddev->minor_version);
			}
		} else
			seq_printf(seq, " super non-persistent");
L
Linus Torvalds 已提交
4818 4819 4820 4821

		if (mddev->pers) {
			mddev->pers->status (seq, mddev);
	 		seq_printf(seq, "\n      ");
4822 4823 4824 4825 4826 4827 4828 4829 4830
			if (mddev->pers->sync_request) {
				if (mddev->curr_resync > 2) {
					status_resync (seq, mddev);
					seq_printf(seq, "\n      ");
				} else if (mddev->curr_resync == 1 || mddev->curr_resync == 2)
					seq_printf(seq, "\tresync=DELAYED\n      ");
				else if (mddev->recovery_cp < MaxSector)
					seq_printf(seq, "\tresync=PENDING\n      ");
			}
4831 4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846
		} else
			seq_printf(seq, "\n       ");

		if ((bitmap = mddev->bitmap)) {
			unsigned long chunk_kb;
			unsigned long flags;
			spin_lock_irqsave(&bitmap->lock, flags);
			chunk_kb = bitmap->chunksize >> 10;
			seq_printf(seq, "bitmap: %lu/%lu pages [%luKB], "
				"%lu%s chunk",
				bitmap->pages - bitmap->missing_pages,
				bitmap->pages,
				(bitmap->pages - bitmap->missing_pages)
					<< (PAGE_SHIFT - 10),
				chunk_kb ? chunk_kb : bitmap->chunksize,
				chunk_kb ? "KB" : "B");
4847 4848 4849 4850
			if (bitmap->file) {
				seq_printf(seq, ", file: ");
				seq_path(seq, bitmap->file->f_vfsmnt,
					 bitmap->file->f_dentry," \t\n");
4851
			}
4852

4853 4854
			seq_printf(seq, "\n");
			spin_unlock_irqrestore(&bitmap->lock, flags);
L
Linus Torvalds 已提交
4855 4856 4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872 4873
		}

		seq_printf(seq, "\n");
	}
	mddev_unlock(mddev);
	
	return 0;
}

static struct seq_operations md_seq_ops = {
	.start  = md_seq_start,
	.next   = md_seq_next,
	.stop   = md_seq_stop,
	.show   = md_seq_show,
};

static int md_seq_open(struct inode *inode, struct file *file)
{
	int error;
4874 4875 4876
	struct mdstat_info *mi = kmalloc(sizeof(*mi), GFP_KERNEL);
	if (mi == NULL)
		return -ENOMEM;
L
Linus Torvalds 已提交
4877 4878

	error = seq_open(file, &md_seq_ops);
4879 4880 4881 4882 4883 4884 4885
	if (error)
		kfree(mi);
	else {
		struct seq_file *p = file->private_data;
		p->private = mi;
		mi->event = atomic_read(&md_event_count);
	}
L
Linus Torvalds 已提交
4886 4887 4888
	return error;
}

4889 4890 4891 4892 4893 4894 4895 4896 4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907 4908 4909 4910 4911 4912 4913
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;
}

L
Linus Torvalds 已提交
4914 4915 4916 4917
static struct file_operations md_seq_fops = {
	.open           = md_seq_open,
	.read           = seq_read,
	.llseek         = seq_lseek,
4918 4919
	.release	= md_seq_release,
	.poll		= mdstat_poll,
L
Linus Torvalds 已提交
4920 4921
};

4922
int register_md_personality(struct mdk_personality *p)
L
Linus Torvalds 已提交
4923 4924
{
	spin_lock(&pers_lock);
4925 4926
	list_add_tail(&p->list, &pers_list);
	printk(KERN_INFO "md: %s personality registered for level %d\n", p->name, p->level);
L
Linus Torvalds 已提交
4927 4928 4929 4930
	spin_unlock(&pers_lock);
	return 0;
}

4931
int unregister_md_personality(struct mdk_personality *p)
L
Linus Torvalds 已提交
4932
{
4933
	printk(KERN_INFO "md: %s personality unregistered\n", p->name);
L
Linus Torvalds 已提交
4934
	spin_lock(&pers_lock);
4935
	list_del_init(&p->list);
L
Linus Torvalds 已提交
4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949
	spin_unlock(&pers_lock);
	return 0;
}

static int is_mddev_idle(mddev_t *mddev)
{
	mdk_rdev_t * rdev;
	struct list_head *tmp;
	int idle;
	unsigned long curr_events;

	idle = 1;
	ITERATE_RDEV(mddev,rdev,tmp) {
		struct gendisk *disk = rdev->bdev->bd_contains->bd_disk;
4950 4951
		curr_events = disk_stat_read(disk, sectors[0]) + 
				disk_stat_read(disk, sectors[1]) - 
L
Linus Torvalds 已提交
4952
				atomic_read(&disk->sync_io);
4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963
		/* The difference between curr_events and last_events
		 * will be affected by any new non-sync IO (making
		 * curr_events bigger) and any difference in the amount of
		 * in-flight syncio (making current_events bigger or smaller)
		 * The amount in-flight is currently limited to
		 * 32*64K in raid1/10 and 256*PAGE_SIZE in raid5/6
		 * which is at most 4096 sectors.
		 * These numbers are fairly fragile and should be made
		 * more robust, probably by enforcing the
		 * 'window size' that md_do_sync sort-of uses.
		 *
L
Linus Torvalds 已提交
4964 4965
		 * Note: the following is an unsigned comparison.
		 */
4966
		if ((curr_events - rdev->last_events + 4096) > 8192) {
L
Linus Torvalds 已提交
4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986
			rdev->last_events = curr_events;
			idle = 0;
		}
	}
	return idle;
}

void md_done_sync(mddev_t *mddev, int blocks, int ok)
{
	/* another "blocks" (512byte) blocks have been synced */
	atomic_sub(blocks, &mddev->recovery_active);
	wake_up(&mddev->recovery_wait);
	if (!ok) {
		set_bit(MD_RECOVERY_ERR, &mddev->recovery);
		md_wakeup_thread(mddev->thread);
		// stop recovery, signal do_sync ....
	}
}


4987 4988
/* md_write_start(mddev, bi)
 * If we need to update some array metadata (e.g. 'active' flag
4989 4990
 * in superblock) before writing, schedule a superblock update
 * and wait for it to complete.
4991
 */
4992
void md_write_start(mddev_t *mddev, struct bio *bi)
L
Linus Torvalds 已提交
4993
{
4994
	if (bio_data_dir(bi) != WRITE)
4995
		return;
4996

4997 4998 4999 5000 5001 5002 5003
	BUG_ON(mddev->ro == 1);
	if (mddev->ro == 2) {
		/* need to switch to read/write */
		mddev->ro = 0;
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
		md_wakeup_thread(mddev->thread);
	}
5004 5005
	atomic_inc(&mddev->writes_pending);
	if (mddev->in_sync) {
5006
		spin_lock_irq(&mddev->write_lock);
5007 5008
		if (mddev->in_sync) {
			mddev->in_sync = 0;
5009
			set_bit(MD_CHANGE_CLEAN, &mddev->flags);
5010 5011
			md_wakeup_thread(mddev->thread);
		}
5012
		spin_unlock_irq(&mddev->write_lock);
5013
	}
5014
	wait_event(mddev->sb_wait, mddev->flags==0);
L
Linus Torvalds 已提交
5015 5016 5017 5018 5019 5020 5021
}

void md_write_end(mddev_t *mddev)
{
	if (atomic_dec_and_test(&mddev->writes_pending)) {
		if (mddev->safemode == 2)
			md_wakeup_thread(mddev->thread);
5022
		else if (mddev->safemode_delay)
L
Linus Torvalds 已提交
5023 5024 5025 5026
			mod_timer(&mddev->safemode_timer, jiffies + mddev->safemode_delay);
	}
}

A
Adrian Bunk 已提交
5027
static DECLARE_WAIT_QUEUE_HEAD(resync_wait);
L
Linus Torvalds 已提交
5028 5029 5030

#define SYNC_MARKS	10
#define	SYNC_MARK_STEP	(3*HZ)
5031
void md_do_sync(mddev_t *mddev)
L
Linus Torvalds 已提交
5032 5033 5034 5035
{
	mddev_t *mddev2;
	unsigned int currspeed = 0,
		 window;
5036
	sector_t max_sectors,j, io_sectors;
L
Linus Torvalds 已提交
5037 5038 5039 5040 5041
	unsigned long mark[SYNC_MARKS];
	sector_t mark_cnt[SYNC_MARKS];
	int last_mark,m;
	struct list_head *tmp;
	sector_t last_check;
5042
	int skipped = 0;
5043 5044
	struct list_head *rtmp;
	mdk_rdev_t *rdev;
5045
	char *desc;
L
Linus Torvalds 已提交
5046 5047 5048 5049

	/* just incase thread restarts... */
	if (test_bit(MD_RECOVERY_DONE, &mddev->recovery))
		return;
5050 5051
	if (mddev->ro) /* never try to sync a read-only array */
		return;
L
Linus Torvalds 已提交
5052

5053 5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064
	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
		if (test_bit(MD_RECOVERY_CHECK, &mddev->recovery))
			desc = "data-check";
		else if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery))
			desc = "requested-resync";
		else
			desc = "resync";
	} else if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
		desc = "reshape";
	else
		desc = "recovery";

L
Linus Torvalds 已提交
5065 5066 5067 5068 5069 5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084
	/* we overload curr_resync somewhat here.
	 * 0 == not engaged in resync at all
	 * 2 == checking that there is no conflict with another sync
	 * 1 == like 2, but have yielded to allow conflicting resync to
	 *		commense
	 * other == active in resync - this many blocks
	 *
	 * Before starting a resync we must have set curr_resync to
	 * 2, and then checked that every "conflicting" array has curr_resync
	 * less than ours.  When we find one that is the same or higher
	 * we wait on resync_wait.  To avoid deadlock, we reduce curr_resync
	 * to 1 if we choose to yield (based arbitrarily on address of mddev structure).
	 * This will mean we have to start checking from the beginning again.
	 *
	 */

	do {
		mddev->curr_resync = 2;

	try_again:
5085
		if (kthread_should_stop()) {
N
NeilBrown 已提交
5086
			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
L
Linus Torvalds 已提交
5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104
			goto skip;
		}
		ITERATE_MDDEV(mddev2,tmp) {
			if (mddev2 == mddev)
				continue;
			if (mddev2->curr_resync && 
			    match_mddev_units(mddev,mddev2)) {
				DEFINE_WAIT(wq);
				if (mddev < mddev2 && mddev->curr_resync == 2) {
					/* arbitrarily yield */
					mddev->curr_resync = 1;
					wake_up(&resync_wait);
				}
				if (mddev > mddev2 && mddev->curr_resync == 1)
					/* no need to wait here, we can wait the next
					 * time 'round when curr_resync == 2
					 */
					continue;
5105 5106
				prepare_to_wait(&resync_wait, &wq, TASK_UNINTERRUPTIBLE);
				if (!kthread_should_stop() &&
5107
				    mddev2->curr_resync >= mddev->curr_resync) {
5108 5109
					printk(KERN_INFO "md: delaying %s of %s"
					       " until %s has finished (they"
L
Linus Torvalds 已提交
5110
					       " share one or more physical units)\n",
5111
					       desc, mdname(mddev), mdname(mddev2));
L
Linus Torvalds 已提交
5112 5113 5114 5115 5116 5117 5118 5119 5120 5121
					mddev_put(mddev2);
					schedule();
					finish_wait(&resync_wait, &wq);
					goto try_again;
				}
				finish_wait(&resync_wait, &wq);
			}
		}
	} while (mddev->curr_resync < 2);

5122
	j = 0;
5123
	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
L
Linus Torvalds 已提交
5124
		/* resync follows the size requested by the personality,
5125
		 * which defaults to physical size, but can be virtual size
L
Linus Torvalds 已提交
5126 5127
		 */
		max_sectors = mddev->resync_max_sectors;
5128
		mddev->resync_mismatches = 0;
5129 5130 5131 5132
		/* we don't use the checkpoint if there's a bitmap */
		if (!mddev->bitmap &&
		    !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery))
			j = mddev->recovery_cp;
5133 5134
	} else if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
		max_sectors = mddev->size << 1;
5135
	else {
L
Linus Torvalds 已提交
5136 5137
		/* recovery follows the physical size of devices */
		max_sectors = mddev->size << 1;
5138 5139 5140 5141 5142 5143 5144 5145
		j = MaxSector;
		ITERATE_RDEV(mddev,rdev,rtmp)
			if (rdev->raid_disk >= 0 &&
			    !test_bit(Faulty, &rdev->flags) &&
			    !test_bit(In_sync, &rdev->flags) &&
			    rdev->recovery_offset < j)
				j = rdev->recovery_offset;
	}
L
Linus Torvalds 已提交
5146

5147 5148 5149
	printk(KERN_INFO "md: %s of RAID array %s\n", desc, mdname(mddev));
	printk(KERN_INFO "md: minimum _guaranteed_  speed:"
		" %d KB/sec/disk.\n", speed_min(mddev));
5150
	printk(KERN_INFO "md: using maximum available idle IO bandwidth "
5151 5152
	       "(but not more than %d KB/sec) for %s.\n",
	       speed_max(mddev), desc);
L
Linus Torvalds 已提交
5153 5154

	is_mddev_idle(mddev); /* this also initializes IO event counters */
5155

5156
	io_sectors = 0;
L
Linus Torvalds 已提交
5157 5158
	for (m = 0; m < SYNC_MARKS; m++) {
		mark[m] = jiffies;
5159
		mark_cnt[m] = io_sectors;
L
Linus Torvalds 已提交
5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177
	}
	last_mark = 0;
	mddev->resync_mark = mark[last_mark];
	mddev->resync_mark_cnt = mark_cnt[last_mark];

	/*
	 * Tune reconstruction:
	 */
	window = 32*(PAGE_SIZE/512);
	printk(KERN_INFO "md: using %dk window, over a total of %llu blocks.\n",
		window/2,(unsigned long long) max_sectors/2);

	atomic_set(&mddev->recovery_active, 0);
	init_waitqueue_head(&mddev->recovery_wait);
	last_check = 0;

	if (j>2) {
		printk(KERN_INFO 
5178 5179
		       "md: resuming %s of %s from checkpoint.\n",
		       desc, mdname(mddev));
L
Linus Torvalds 已提交
5180 5181 5182 5183
		mddev->curr_resync = j;
	}

	while (j < max_sectors) {
5184
		sector_t sectors;
L
Linus Torvalds 已提交
5185

5186 5187
		skipped = 0;
		sectors = mddev->pers->sync_request(mddev, j, &skipped,
5188
					    currspeed < speed_min(mddev));
5189
		if (sectors == 0) {
L
Linus Torvalds 已提交
5190 5191 5192
			set_bit(MD_RECOVERY_ERR, &mddev->recovery);
			goto out;
		}
5193 5194 5195 5196 5197 5198

		if (!skipped) { /* actual IO requested */
			io_sectors += sectors;
			atomic_add(sectors, &mddev->recovery_active);
		}

L
Linus Torvalds 已提交
5199 5200
		j += sectors;
		if (j>1) mddev->curr_resync = j;
5201
		mddev->curr_mark_cnt = io_sectors;
5202 5203 5204 5205 5206
		if (last_check == 0)
			/* this is the earliers that rebuilt will be
			 * visible in /proc/mdstat
			 */
			md_new_event(mddev);
5207 5208

		if (last_check + window > io_sectors || j == max_sectors)
L
Linus Torvalds 已提交
5209 5210
			continue;

5211
		last_check = io_sectors;
L
Linus Torvalds 已提交
5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224

		if (test_bit(MD_RECOVERY_INTR, &mddev->recovery) ||
		    test_bit(MD_RECOVERY_ERR, &mddev->recovery))
			break;

	repeat:
		if (time_after_eq(jiffies, mark[last_mark] + SYNC_MARK_STEP )) {
			/* step marks */
			int next = (last_mark+1) % SYNC_MARKS;

			mddev->resync_mark = mark[next];
			mddev->resync_mark_cnt = mark_cnt[next];
			mark[next] = jiffies;
5225
			mark_cnt[next] = io_sectors - atomic_read(&mddev->recovery_active);
L
Linus Torvalds 已提交
5226 5227 5228 5229
			last_mark = next;
		}


5230
		if (kthread_should_stop()) {
L
Linus Torvalds 已提交
5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249 5250
			/*
			 * got a signal, exit.
			 */
			printk(KERN_INFO 
				"md: md_do_sync() got signal ... exiting\n");
			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
			goto out;
		}

		/*
		 * this loop exits only if either when we are slower than
		 * the 'hard' speed limit, or the system was IO-idle for
		 * a jiffy.
		 * the system might be non-idle CPU-wise, but we only care
		 * about not overloading the IO subsystem. (things like an
		 * e2fsck being done on the RAID array should execute fast)
		 */
		mddev->queue->unplug_fn(mddev->queue);
		cond_resched();

5251 5252
		currspeed = ((unsigned long)(io_sectors-mddev->resync_mark_cnt))/2
			/((jiffies-mddev->resync_mark)/HZ +1) +1;
L
Linus Torvalds 已提交
5253

5254 5255
		if (currspeed > speed_min(mddev)) {
			if ((currspeed > speed_max(mddev)) ||
L
Linus Torvalds 已提交
5256
					!is_mddev_idle(mddev)) {
5257
				msleep(500);
L
Linus Torvalds 已提交
5258 5259 5260 5261
				goto repeat;
			}
		}
	}
5262
	printk(KERN_INFO "md: %s: %s done.\n",mdname(mddev), desc);
L
Linus Torvalds 已提交
5263 5264 5265 5266 5267 5268 5269 5270 5271
	/*
	 * this also signals 'finished resyncing' to md_stop
	 */
 out:
	mddev->queue->unplug_fn(mddev->queue);

	wait_event(mddev->recovery_wait, !atomic_read(&mddev->recovery_active));

	/* tell personality that we are finished */
5272
	mddev->pers->sync_request(mddev, max_sectors, &skipped, 1);
L
Linus Torvalds 已提交
5273 5274

	if (!test_bit(MD_RECOVERY_ERR, &mddev->recovery) &&
5275 5276
	    test_bit(MD_RECOVERY_SYNC, &mddev->recovery) &&
	    !test_bit(MD_RECOVERY_CHECK, &mddev->recovery) &&
5277 5278 5279 5280 5281
	    mddev->curr_resync > 2) {
		if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
			if (test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
				if (mddev->curr_resync >= mddev->recovery_cp) {
					printk(KERN_INFO
5282 5283
					       "md: checkpointing %s of %s.\n",
					       desc, mdname(mddev));
5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297
					mddev->recovery_cp = mddev->curr_resync;
				}
			} else
				mddev->recovery_cp = MaxSector;
		} else {
			if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery))
				mddev->curr_resync = MaxSector;
			ITERATE_RDEV(mddev,rdev,rtmp)
				if (rdev->raid_disk >= 0 &&
				    !test_bit(Faulty, &rdev->flags) &&
				    !test_bit(In_sync, &rdev->flags) &&
				    rdev->recovery_offset < mddev->curr_resync)
					rdev->recovery_offset = mddev->curr_resync;
		}
L
Linus Torvalds 已提交
5298 5299 5300 5301 5302 5303 5304 5305
	}

 skip:
	mddev->curr_resync = 0;
	wake_up(&resync_wait);
	set_bit(MD_RECOVERY_DONE, &mddev->recovery);
	md_wakeup_thread(mddev->thread);
}
5306
EXPORT_SYMBOL_GPL(md_do_sync);
L
Linus Torvalds 已提交
5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336


/*
 * This routine is regularly called by all per-raid-array threads to
 * deal with generic issues like resync and super-block update.
 * Raid personalities that don't have a thread (linear/raid0) do not
 * need this as they never do any recovery or update the superblock.
 *
 * It does not do any resync itself, but rather "forks" off other threads
 * to do that as needed.
 * When it is determined that resync is needed, we set MD_RECOVERY_RUNNING in
 * "->recovery" and create a thread at ->sync_thread.
 * When the thread finishes it sets MD_RECOVERY_DONE (and might set MD_RECOVERY_ERR)
 * and wakeups up this thread which will reap the thread and finish up.
 * This thread also removes any faulty devices (with nr_pending == 0).
 *
 * The overall approach is:
 *  1/ if the superblock needs updating, update it.
 *  2/ If a recovery thread is running, don't do anything else.
 *  3/ If recovery has finished, clean up, possibly marking spares active.
 *  4/ If there are any faulty devices, remove them.
 *  5/ If array is degraded, try to add spares devices
 *  6/ If array has spares or is not in-sync, start a resync thread.
 */
void md_check_recovery(mddev_t *mddev)
{
	mdk_rdev_t *rdev;
	struct list_head *rtmp;


5337 5338
	if (mddev->bitmap)
		bitmap_daemon_work(mddev->bitmap);
L
Linus Torvalds 已提交
5339 5340 5341

	if (mddev->ro)
		return;
5342 5343 5344 5345 5346 5347 5348 5349 5350 5351

	if (signal_pending(current)) {
		if (mddev->pers->sync_request) {
			printk(KERN_INFO "md: %s in immediate safe mode\n",
			       mdname(mddev));
			mddev->safemode = 2;
		}
		flush_signals(current);
	}

L
Linus Torvalds 已提交
5352
	if ( ! (
5353
		mddev->flags ||
L
Linus Torvalds 已提交
5354
		test_bit(MD_RECOVERY_NEEDED, &mddev->recovery) ||
5355 5356 5357 5358
		test_bit(MD_RECOVERY_DONE, &mddev->recovery) ||
		(mddev->safemode == 1) ||
		(mddev->safemode == 2 && ! atomic_read(&mddev->writes_pending)
		 && !mddev->in_sync && mddev->recovery_cp == MaxSector)
L
Linus Torvalds 已提交
5359 5360
		))
		return;
5361

5362
	if (mddev_trylock(mddev)) {
L
Linus Torvalds 已提交
5363
		int spares =0;
5364

5365
		spin_lock_irq(&mddev->write_lock);
5366 5367 5368
		if (mddev->safemode && !atomic_read(&mddev->writes_pending) &&
		    !mddev->in_sync && mddev->recovery_cp == MaxSector) {
			mddev->in_sync = 1;
5369
			set_bit(MD_CHANGE_CLEAN, &mddev->flags);
5370 5371 5372
		}
		if (mddev->safemode == 1)
			mddev->safemode = 0;
5373
		spin_unlock_irq(&mddev->write_lock);
5374

5375 5376
		if (mddev->flags)
			md_update_sb(mddev, 0);
5377 5378


L
Linus Torvalds 已提交
5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394
		if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) &&
		    !test_bit(MD_RECOVERY_DONE, &mddev->recovery)) {
			/* resync/recovery still happening */
			clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
			goto unlock;
		}
		if (mddev->sync_thread) {
			/* resync has finished, collect result */
			md_unregister_thread(mddev->sync_thread);
			mddev->sync_thread = NULL;
			if (!test_bit(MD_RECOVERY_ERR, &mddev->recovery) &&
			    !test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
				/* success...*/
				/* activate any spares */
				mddev->pers->spare_active(mddev);
			}
5395
			md_update_sb(mddev, 1);
5396 5397 5398 5399 5400 5401 5402 5403

			/* if array is no-longer degraded, then any saved_raid_disk
			 * information must be scrapped
			 */
			if (!mddev->degraded)
				ITERATE_RDEV(mddev,rdev,rtmp)
					rdev->saved_raid_disk = -1;

L
Linus Torvalds 已提交
5404 5405 5406
			mddev->recovery = 0;
			/* flag recovery needed just to double check */
			set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
5407
			md_new_event(mddev);
L
Linus Torvalds 已提交
5408 5409
			goto unlock;
		}
5410 5411 5412 5413 5414 5415 5416
		/* Clear some bits that don't mean anything, but
		 * might be left set
		 */
		clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
		clear_bit(MD_RECOVERY_ERR, &mddev->recovery);
		clear_bit(MD_RECOVERY_INTR, &mddev->recovery);
		clear_bit(MD_RECOVERY_DONE, &mddev->recovery);
L
Linus Torvalds 已提交
5417

5418 5419
		if (test_bit(MD_RECOVERY_FROZEN, &mddev->recovery))
			goto unlock;
L
Linus Torvalds 已提交
5420 5421 5422 5423 5424 5425 5426 5427
		/* no recovery is running.
		 * remove any failed drives, then
		 * add spares if possible.
		 * Spare are also removed and re-added, to allow
		 * the personality to fail the re-add.
		 */
		ITERATE_RDEV(mddev,rdev,rtmp)
			if (rdev->raid_disk >= 0 &&
5428
			    (test_bit(Faulty, &rdev->flags) || ! test_bit(In_sync, &rdev->flags)) &&
L
Linus Torvalds 已提交
5429
			    atomic_read(&rdev->nr_pending)==0) {
5430 5431 5432 5433
				if (mddev->pers->hot_remove_disk(mddev, rdev->raid_disk)==0) {
					char nm[20];
					sprintf(nm,"rd%d", rdev->raid_disk);
					sysfs_remove_link(&mddev->kobj, nm);
L
Linus Torvalds 已提交
5434
					rdev->raid_disk = -1;
5435
				}
L
Linus Torvalds 已提交
5436 5437 5438 5439 5440
			}

		if (mddev->degraded) {
			ITERATE_RDEV(mddev,rdev,rtmp)
				if (rdev->raid_disk < 0
5441
				    && !test_bit(Faulty, &rdev->flags)) {
5442
					rdev->recovery_offset = 0;
5443 5444 5445 5446
					if (mddev->pers->hot_add_disk(mddev,rdev)) {
						char nm[20];
						sprintf(nm, "rd%d", rdev->raid_disk);
						sysfs_create_link(&mddev->kobj, &rdev->kobj, nm);
L
Linus Torvalds 已提交
5447
						spares++;
5448
						md_new_event(mddev);
5449
					} else
L
Linus Torvalds 已提交
5450 5451 5452 5453
						break;
				}
		}

5454 5455 5456 5457 5458 5459 5460
		if (spares) {
			clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
			clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
		} else if (mddev->recovery_cp < MaxSector) {
			set_bit(MD_RECOVERY_SYNC, &mddev->recovery);
		} else if (!test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
			/* nothing to be done ... */
L
Linus Torvalds 已提交
5461
			goto unlock;
5462

L
Linus Torvalds 已提交
5463 5464
		if (mddev->pers->sync_request) {
			set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
5465 5466 5467 5468 5469 5470 5471
			if (spares && mddev->bitmap && ! mddev->bitmap->file) {
				/* We are adding a device or devices to an array
				 * which has the bitmap stored on all devices.
				 * So make sure all bitmap pages get written
				 */
				bitmap_write_all(mddev->bitmap);
			}
L
Linus Torvalds 已提交
5472 5473 5474 5475 5476 5477 5478 5479 5480
			mddev->sync_thread = md_register_thread(md_do_sync,
								mddev,
								"%s_resync");
			if (!mddev->sync_thread) {
				printk(KERN_ERR "%s: could not start resync"
					" thread...\n", 
					mdname(mddev));
				/* leave the spares where they are, it shouldn't hurt */
				mddev->recovery = 0;
5481
			} else
L
Linus Torvalds 已提交
5482
				md_wakeup_thread(mddev->sync_thread);
5483
			md_new_event(mddev);
L
Linus Torvalds 已提交
5484 5485 5486 5487 5488 5489
		}
	unlock:
		mddev_unlock(mddev);
	}
}

A
Adrian Bunk 已提交
5490 5491
static int md_notify_reboot(struct notifier_block *this,
			    unsigned long code, void *x)
L
Linus Torvalds 已提交
5492 5493 5494 5495 5496 5497 5498 5499 5500
{
	struct list_head *tmp;
	mddev_t *mddev;

	if ((code == SYS_DOWN) || (code == SYS_HALT) || (code == SYS_POWER_OFF)) {

		printk(KERN_INFO "md: stopping all md devices.\n");

		ITERATE_MDDEV(mddev,tmp)
5501
			if (mddev_trylock(mddev)) {
L
Linus Torvalds 已提交
5502
				do_md_stop (mddev, 1);
5503 5504
				mddev_unlock(mddev);
			}
L
Linus Torvalds 已提交
5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515
		/*
		 * certain more exotic SCSI devices are known to be
		 * volatile wrt too early system reboots. While the
		 * right place to handle this issue is the given
		 * driver, we do want to have a safe RAID driver ...
		 */
		mdelay(1000*1);
	}
	return NOTIFY_DONE;
}

A
Adrian Bunk 已提交
5516
static struct notifier_block md_notifier = {
L
Linus Torvalds 已提交
5517 5518 5519 5520 5521 5522 5523 5524 5525 5526 5527 5528 5529 5530 5531 5532
	.notifier_call	= md_notify_reboot,
	.next		= NULL,
	.priority	= INT_MAX, /* before any real devices */
};

static void md_geninit(void)
{
	struct proc_dir_entry *p;

	dprintk("md: sizeof(mdp_super_t) = %d\n", (int)sizeof(mdp_super_t));

	p = create_proc_entry("mdstat", S_IRUGO, NULL);
	if (p)
		p->proc_fops = &md_seq_fops;
}

A
Adrian Bunk 已提交
5533
static int __init md_init(void)
L
Linus Torvalds 已提交
5534 5535 5536 5537 5538 5539 5540
{
	if (register_blkdev(MAJOR_NR, "md"))
		return -1;
	if ((mdp_major=register_blkdev(0, "mdp"))<=0) {
		unregister_blkdev(MAJOR_NR, "md");
		return -1;
	}
5541 5542 5543
	blk_register_region(MKDEV(MAJOR_NR, 0), 1UL<<MINORBITS, THIS_MODULE,
			    md_probe, NULL, NULL);
	blk_register_region(MKDEV(mdp_major, 0), 1UL<<MINORBITS, THIS_MODULE,
L
Linus Torvalds 已提交
5544 5545 5546 5547 5548 5549 5550 5551 5552 5553 5554 5555 5556 5557 5558 5559 5560 5561 5562 5563 5564 5565 5566 5567 5568 5569 5570 5571 5572 5573 5574 5575 5576 5577 5578 5579 5580 5581 5582 5583
			    md_probe, NULL, NULL);

	register_reboot_notifier(&md_notifier);
	raid_table_header = register_sysctl_table(raid_root_table, 1);

	md_geninit();
	return (0);
}


#ifndef MODULE

/*
 * Searches all registered partitions for autorun RAID arrays
 * at boot time.
 */
static dev_t detected_devices[128];
static int dev_cnt;

void md_autodetect_dev(dev_t dev)
{
	if (dev_cnt >= 0 && dev_cnt < 127)
		detected_devices[dev_cnt++] = dev;
}


static void autostart_arrays(int part)
{
	mdk_rdev_t *rdev;
	int i;

	printk(KERN_INFO "md: Autodetecting RAID arrays.\n");

	for (i = 0; i < dev_cnt; i++) {
		dev_t dev = detected_devices[i];

		rdev = md_import_device(dev,0, 0);
		if (IS_ERR(rdev))
			continue;

5584
		if (test_bit(Faulty, &rdev->flags)) {
L
Linus Torvalds 已提交
5585 5586 5587 5588 5589 5590 5591 5592 5593 5594 5595 5596 5597 5598 5599 5600
			MD_BUG();
			continue;
		}
		list_add(&rdev->same_set, &pending_raid_disks);
	}
	dev_cnt = 0;

	autorun_devices(part);
}

#endif

static __exit void md_exit(void)
{
	mddev_t *mddev;
	struct list_head *tmp;
5601

5602 5603
	blk_unregister_region(MKDEV(MAJOR_NR,0), 1U << MINORBITS);
	blk_unregister_region(MKDEV(mdp_major,0), 1U << MINORBITS);
L
Linus Torvalds 已提交
5604 5605 5606 5607 5608 5609 5610 5611 5612 5613 5614 5615 5616 5617 5618 5619 5620 5621 5622 5623 5624

	unregister_blkdev(MAJOR_NR,"md");
	unregister_blkdev(mdp_major, "mdp");
	unregister_reboot_notifier(&md_notifier);
	unregister_sysctl_table(raid_table_header);
	remove_proc_entry("mdstat", NULL);
	ITERATE_MDDEV(mddev,tmp) {
		struct gendisk *disk = mddev->gendisk;
		if (!disk)
			continue;
		export_array(mddev);
		del_gendisk(disk);
		put_disk(disk);
		mddev->gendisk = NULL;
		mddev_put(mddev);
	}
}

module_init(md_init)
module_exit(md_exit)

5625 5626 5627 5628 5629 5630 5631 5632 5633 5634
static int get_ro(char *buffer, struct kernel_param *kp)
{
	return sprintf(buffer, "%d", start_readonly);
}
static int set_ro(const char *val, struct kernel_param *kp)
{
	char *e;
	int num = simple_strtoul(val, &e, 10);
	if (*val && (*e == '\0' || *e == '\n')) {
		start_readonly = num;
5635
		return 0;
5636 5637 5638 5639
	}
	return -EINVAL;
}

5640 5641
module_param_call(start_ro, set_ro, get_ro, NULL, S_IRUSR|S_IWUSR);
module_param(start_dirty_degraded, int, S_IRUGO|S_IWUSR);
5642

5643

L
Linus Torvalds 已提交
5644 5645 5646 5647 5648 5649 5650 5651 5652 5653 5654
EXPORT_SYMBOL(register_md_personality);
EXPORT_SYMBOL(unregister_md_personality);
EXPORT_SYMBOL(md_error);
EXPORT_SYMBOL(md_done_sync);
EXPORT_SYMBOL(md_write_start);
EXPORT_SYMBOL(md_write_end);
EXPORT_SYMBOL(md_register_thread);
EXPORT_SYMBOL(md_unregister_thread);
EXPORT_SYMBOL(md_wakeup_thread);
EXPORT_SYMBOL(md_check_recovery);
MODULE_LICENSE("GPL");
5655
MODULE_ALIAS("md");
5656
MODULE_ALIAS_BLOCKDEV_MAJOR(MD_MAJOR);