bitmap.c 54.8 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
/*
 * bitmap.c two-level bitmap (C) Peter T. Breuer (ptb@ot.uc3m.es) 2003
 *
 * bitmap_create  - sets up the bitmap structure
 * bitmap_destroy - destroys the bitmap structure
 *
 * additions, Copyright (C) 2003-2004, Paul Clements, SteelEye Technology, Inc.:
 * - added disk storage for bitmap
 * - changes to allow various bitmap chunk sizes
 */

/*
 * Still to do:
 *
 * flush after percent set rather than just time based. (maybe both).
 */

18
#include <linux/blkdev.h>
19 20 21 22 23 24 25 26 27 28
#include <linux/module.h>
#include <linux/errno.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/timer.h>
#include <linux/sched.h>
#include <linux/list.h>
#include <linux/file.h>
#include <linux/mount.h>
#include <linux/buffer_head.h>
29
#include "md.h"
30
#include "bitmap.h"
31

32
#include <linux/dm-dirty-log.h>
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
/* debug macros */

#define DEBUG 0

#if DEBUG
/* these are for debugging purposes only! */

/* define one and only one of these */
#define INJECT_FAULTS_1 0 /* cause bitmap_alloc_page to fail always */
#define INJECT_FAULTS_2 0 /* cause bitmap file to be kicked when first bit set*/
#define INJECT_FAULTS_3 0 /* treat bitmap file as kicked at init time */
#define INJECT_FAULTS_4 0 /* undef */
#define INJECT_FAULTS_5 0 /* undef */
#define INJECT_FAULTS_6 0

/* if these are defined, the driver will fail! debug only */
#define INJECT_FATAL_FAULT_1 0 /* fail kmalloc, causing bitmap_create to fail */
#define INJECT_FATAL_FAULT_2 0 /* undef */
#define INJECT_FATAL_FAULT_3 0 /* undef */
#endif

#ifndef PRINTK
#  if DEBUG > 0
#    define PRINTK(x...) printk(KERN_DEBUG x)
#  else
#    define PRINTK(x...)
#  endif
#endif

62
static inline char *bmname(struct bitmap *bitmap)
63 64 65 66 67 68 69 70 71 72 73
{
	return bitmap->mddev ? mdname(bitmap->mddev) : "mdX";
}

/*
 * just a placeholder - calls kmalloc for bitmap pages
 */
static unsigned char *bitmap_alloc_page(struct bitmap *bitmap)
{
	unsigned char *page;

74
#ifdef INJECT_FAULTS_1
75 76
	page = NULL;
#else
77
	page = kzalloc(PAGE_SIZE, GFP_NOIO);
78 79 80 81
#endif
	if (!page)
		printk("%s: bitmap_alloc_page FAILED\n", bmname(bitmap));
	else
82
		PRINTK("%s: bitmap_alloc_page: allocated page at %p\n",
83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105
			bmname(bitmap), page);
	return page;
}

/*
 * for now just a placeholder -- just calls kfree for bitmap pages
 */
static void bitmap_free_page(struct bitmap *bitmap, unsigned char *page)
{
	PRINTK("%s: bitmap_free_page: free page %p\n", bmname(bitmap), page);
	kfree(page);
}

/*
 * check a page and, if necessary, allocate it (or hijack it if the alloc fails)
 *
 * 1) check to see if this page is allocated, if it's not then try to alloc
 * 2) if the alloc fails, set the page's hijacked flag so we'll use the
 *    page pointer directly as a counter
 *
 * if we find our page, we increment the page's refcount so that it stays
 * allocated while we're using it
 */
106 107
static int bitmap_checkpage(struct bitmap *bitmap,
			    unsigned long page, int create)
108 109
__releases(bitmap->lock)
__acquires(bitmap->lock)
110 111 112 113
{
	unsigned char *mappage;

	if (page >= bitmap->pages) {
114 115 116 117
		/* This can happen if bitmap_start_sync goes beyond
		 * End-of-device while looking for a whole page.
		 * It is harmless.
		 */
118 119 120 121 122 123 124 125 126 127 128 129 130 131
		return -EINVAL;
	}

	if (bitmap->bp[page].hijacked) /* it's hijacked, don't try to alloc */
		return 0;

	if (bitmap->bp[page].map) /* page is already allocated, just return */
		return 0;

	if (!create)
		return -ENOENT;

	/* this page has not been allocated yet */

132 133 134 135 136
	spin_unlock_irq(&bitmap->lock);
	mappage = bitmap_alloc_page(bitmap);
	spin_lock_irq(&bitmap->lock);

	if (mappage == NULL) {
137 138 139 140 141 142
		PRINTK("%s: bitmap map page allocation failed, hijacking\n",
			bmname(bitmap));
		/* failed - set the hijacked flag so that we can use the
		 * pointer as a counter */
		if (!bitmap->bp[page].map)
			bitmap->bp[page].hijacked = 1;
143 144
	} else if (bitmap->bp[page].map ||
		   bitmap->bp[page].hijacked) {
145 146 147
		/* somebody beat us to getting the page */
		bitmap_free_page(bitmap, mappage);
		return 0;
148
	} else {
149

150
		/* no page was in place and we have one, so install it */
151

152 153 154
		bitmap->bp[page].map = mappage;
		bitmap->missing_pages--;
	}
155 156 157 158 159 160
	return 0;
}

/* if page is completely empty, put it back on the free list, or dealloc it */
/* if page was hijacked, unmark the flag so it might get alloced next time */
/* Note: lock should be held when calling this */
161
static void bitmap_checkfree(struct bitmap *bitmap, unsigned long page)
162 163 164 165 166 167 168 169 170 171 172
{
	char *ptr;

	if (bitmap->bp[page].count) /* page is still busy */
		return;

	/* page is no longer in use, it can be released */

	if (bitmap->bp[page].hijacked) { /* page was hijacked, undo this now */
		bitmap->bp[page].hijacked = 0;
		bitmap->bp[page].map = NULL;
173 174 175 176 177 178
	} else {
		/* normal case, free the page */
		ptr = bitmap->bp[page].map;
		bitmap->bp[page].map = NULL;
		bitmap->missing_pages++;
		bitmap_free_page(bitmap, ptr);
179 180 181 182 183 184 185 186 187 188 189
	}
}

/*
 * bitmap file handling - read and write the bitmap file and its superblock
 */

/*
 * basic page I/O operations
 */

190
/* IO operations when bitmap is stored near all superblocks */
191
static struct page *read_sb_page(mddev_t *mddev, loff_t offset,
192 193
				 struct page *page,
				 unsigned long index, int size)
194 195 196 197 198
{
	/* choose a good rdev and read the page from there */

	mdk_rdev_t *rdev;
	sector_t target;
199
	int did_alloc = 0;
200

201
	if (!page) {
202
		page = alloc_page(GFP_KERNEL);
203 204 205 206
		if (!page)
			return ERR_PTR(-ENOMEM);
		did_alloc = 1;
	}
207

208
	list_for_each_entry(rdev, &mddev->disks, same_set) {
209 210
		if (! test_bit(In_sync, &rdev->flags)
		    || test_bit(Faulty, &rdev->flags))
211 212
			continue;

J
Jonathan Brassow 已提交
213
		target = offset + index * (PAGE_SIZE/512);
214

215
		if (sync_page_io(rdev, target,
216
				 roundup(size, bdev_logical_block_size(rdev->bdev)),
J
Jonathan Brassow 已提交
217
				 page, READ, true)) {
218
			page->index = index;
219 220
			attach_page_buffers(page, NULL); /* so that free_buffer will
							  * quietly no-op */
221 222 223
			return page;
		}
	}
224 225
	if (did_alloc)
		put_page(page);
226
	return ERR_PTR(-EIO);
227 228 229

}

230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263
static mdk_rdev_t *next_active_rdev(mdk_rdev_t *rdev, mddev_t *mddev)
{
	/* Iterate the disks of an mddev, using rcu to protect access to the
	 * linked list, and raising the refcount of devices we return to ensure
	 * they don't disappear while in use.
	 * As devices are only added or removed when raid_disk is < 0 and
	 * nr_pending is 0 and In_sync is clear, the entries we return will
	 * still be in the same position on the list when we re-enter
	 * list_for_each_continue_rcu.
	 */
	struct list_head *pos;
	rcu_read_lock();
	if (rdev == NULL)
		/* start at the beginning */
		pos = &mddev->disks;
	else {
		/* release the previous rdev and start from there. */
		rdev_dec_pending(rdev, mddev);
		pos = &rdev->same_set;
	}
	list_for_each_continue_rcu(pos, &mddev->disks) {
		rdev = list_entry(pos, mdk_rdev_t, same_set);
		if (rdev->raid_disk >= 0 &&
		    !test_bit(Faulty, &rdev->flags)) {
			/* this is a usable devices */
			atomic_inc(&rdev->nr_pending);
			rcu_read_unlock();
			return rdev;
		}
	}
	rcu_read_unlock();
	return NULL;
}

264
static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait)
265
{
266
	mdk_rdev_t *rdev = NULL;
267
	struct block_device *bdev;
268
	mddev_t *mddev = bitmap->mddev;
269

270
	while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
271 272
		int size = PAGE_SIZE;
		loff_t offset = mddev->bitmap_info.offset;
273 274 275

		bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev;

276 277
		if (page->index == bitmap->file_pages-1)
			size = roundup(bitmap->last_page_size,
278
				       bdev_logical_block_size(bdev));
279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
		/* Just make sure we aren't corrupting data or
		 * metadata
		 */
		if (mddev->external) {
			/* Bitmap could be anywhere. */
			if (rdev->sb_start + offset + (page->index
						       * (PAGE_SIZE/512))
			    > rdev->data_offset
			    &&
			    rdev->sb_start + offset
			    < (rdev->data_offset + mddev->dev_sectors
			     + (PAGE_SIZE/512)))
				goto bad_alignment;
		} else if (offset < 0) {
			/* DATA  BITMAP METADATA  */
			if (offset
			    + (long)(page->index * (PAGE_SIZE/512))
			    + size/512 > 0)
				/* bitmap runs in to metadata */
				goto bad_alignment;
			if (rdev->data_offset + mddev->dev_sectors
			    > rdev->sb_start + offset)
				/* data runs in to bitmap */
				goto bad_alignment;
		} else if (rdev->sb_start < rdev->data_offset) {
			/* METADATA BITMAP DATA */
			if (rdev->sb_start
			    + offset
			    + page->index*(PAGE_SIZE/512) + size/512
			    > rdev->data_offset)
				/* bitmap runs in to data */
				goto bad_alignment;
		} else {
			/* DATA METADATA BITMAP - no problems */
		}
		md_super_write(mddev, rdev,
			       rdev->sb_start + offset
			       + page->index * (PAGE_SIZE/512),
			       size,
			       page);
319
	}
320 321

	if (wait)
322
		md_super_wait(mddev);
323
	return 0;
324 325 326

 bad_alignment:
	return -EINVAL;
327 328
}

329
static void bitmap_file_kick(struct bitmap *bitmap);
330
/*
331
 * write out a page to a file
332
 */
333
static void write_page(struct bitmap *bitmap, struct page *page, int wait)
334
{
335
	struct buffer_head *bh;
336

337 338 339 340 341
	if (bitmap->file == NULL) {
		switch (write_sb_page(bitmap, page, wait)) {
		case -EINVAL:
			bitmap->flags |= BITMAP_WRITE_ERROR;
		}
342
	} else {
343

344
		bh = page_buffers(page);
345

346 347 348 349
		while (bh && bh->b_blocknr) {
			atomic_inc(&bitmap->pending_writes);
			set_buffer_locked(bh);
			set_buffer_mapped(bh);
J
Jens Axboe 已提交
350
			submit_bh(WRITE | REQ_SYNC, bh);
351 352
			bh = bh->b_this_page;
		}
353

354
		if (wait)
355 356
			wait_event(bitmap->write_wait,
				   atomic_read(&bitmap->pending_writes)==0);
357
	}
358 359
	if (bitmap->flags & BITMAP_WRITE_ERROR)
		bitmap_file_kick(bitmap);
360 361 362 363 364 365
}

static void end_bitmap_write(struct buffer_head *bh, int uptodate)
{
	struct bitmap *bitmap = bh->b_private;
	unsigned long flags;
366

367 368 369 370
	if (!uptodate) {
		spin_lock_irqsave(&bitmap->lock, flags);
		bitmap->flags |= BITMAP_WRITE_ERROR;
		spin_unlock_irqrestore(&bitmap->lock, flags);
371
	}
372 373 374
	if (atomic_dec_and_test(&bitmap->pending_writes))
		wake_up(&bitmap->write_wait);
}
375

376 377 378 379 380 381 382 383 384 385 386
/* copied from buffer.c */
static void
__clear_page_buffers(struct page *page)
{
	ClearPagePrivate(page);
	set_page_private(page, 0);
	page_cache_release(page);
}
static void free_buffers(struct page *page)
{
	struct buffer_head *bh = page_buffers(page);
387

388 389 390 391
	while (bh) {
		struct buffer_head *next = bh->b_this_page;
		free_buffer_head(bh);
		bh = next;
392
	}
393 394
	__clear_page_buffers(page);
	put_page(page);
395 396
}

397 398 399 400 401 402 403
/* read a page from a file.
 * We both read the page, and attach buffers to the page to record the
 * address of each block (using bmap).  These addresses will be used
 * to write the block later, completely bypassing the filesystem.
 * This usage is similar to how swap files are handled, and allows us
 * to write to a file with no concerns of memory allocation failing.
 */
404
static struct page *read_page(struct file *file, unsigned long index,
405 406
			      struct bitmap *bitmap,
			      unsigned long count)
407 408
{
	struct page *page = NULL;
J
Josef Sipek 已提交
409
	struct inode *inode = file->f_path.dentry->d_inode;
410 411
	struct buffer_head *bh;
	sector_t block;
412

413
	PRINTK("read bitmap file (%dB @ %llu)\n", (int)PAGE_SIZE,
414
			(unsigned long long)index << PAGE_SHIFT);
415

416 417 418
	page = alloc_page(GFP_KERNEL);
	if (!page)
		page = ERR_PTR(-ENOMEM);
419 420
	if (IS_ERR(page))
		goto out;
421 422 423

	bh = alloc_page_buffers(page, 1<<inode->i_blkbits, 0);
	if (!bh) {
424
		put_page(page);
425
		page = ERR_PTR(-ENOMEM);
426 427
		goto out;
	}
428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448
	attach_page_buffers(page, bh);
	block = index << (PAGE_SHIFT - inode->i_blkbits);
	while (bh) {
		if (count == 0)
			bh->b_blocknr = 0;
		else {
			bh->b_blocknr = bmap(inode, block);
			if (bh->b_blocknr == 0) {
				/* Cannot use this file! */
				free_buffers(page);
				page = ERR_PTR(-EINVAL);
				goto out;
			}
			bh->b_bdev = inode->i_sb->s_bdev;
			if (count < (1<<inode->i_blkbits))
				count = 0;
			else
				count -= (1<<inode->i_blkbits);

			bh->b_end_io = end_bitmap_write;
			bh->b_private = bitmap;
449 450 451 452
			atomic_inc(&bitmap->pending_writes);
			set_buffer_locked(bh);
			set_buffer_mapped(bh);
			submit_bh(READ, bh);
453 454 455 456 457
		}
		block++;
		bh = bh->b_this_page;
	}
	page->index = index;
458 459 460 461 462 463 464

	wait_event(bitmap->write_wait,
		   atomic_read(&bitmap->pending_writes)==0);
	if (bitmap->flags & BITMAP_WRITE_ERROR) {
		free_buffers(page);
		page = ERR_PTR(-EIO);
	}
465 466
out:
	if (IS_ERR(page))
467
		printk(KERN_ALERT "md: bitmap read error: (%dB @ %llu): %ld\n",
468 469
			(int)PAGE_SIZE,
			(unsigned long long)index << PAGE_SHIFT,
470 471 472 473 474 475 476 477 478
			PTR_ERR(page));
	return page;
}

/*
 * bitmap file superblock operations
 */

/* update the event counter and sync the superblock to disk */
479
void bitmap_update_sb(struct bitmap *bitmap)
480 481 482 483 484
{
	bitmap_super_t *sb;
	unsigned long flags;

	if (!bitmap || !bitmap->mddev) /* no bitmap for this array */
485
		return;
486 487
	if (bitmap->mddev->bitmap_info.external)
		return;
488 489 490
	spin_lock_irqsave(&bitmap->lock, flags);
	if (!bitmap->sb_page) { /* no superblock */
		spin_unlock_irqrestore(&bitmap->lock, flags);
491
		return;
492 493
	}
	spin_unlock_irqrestore(&bitmap->lock, flags);
494
	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
495
	sb->events = cpu_to_le64(bitmap->mddev->events);
496
	if (bitmap->mddev->events < bitmap->events_cleared)
497 498
		/* rocking back to read-only */
		bitmap->events_cleared = bitmap->mddev->events;
499 500
	sb->events_cleared = cpu_to_le64(bitmap->events_cleared);
	sb->state = cpu_to_le32(bitmap->flags);
501 502 503
	/* Just in case these have been changed via sysfs: */
	sb->daemon_sleep = cpu_to_le32(bitmap->mddev->bitmap_info.daemon_sleep/HZ);
	sb->write_behind = cpu_to_le32(bitmap->mddev->bitmap_info.max_write_behind);
504
	kunmap_atomic(sb, KM_USER0);
505
	write_page(bitmap, bitmap->sb_page, 1);
506 507 508 509 510 511 512 513 514
}

/* print out the bitmap file superblock */
void bitmap_print_sb(struct bitmap *bitmap)
{
	bitmap_super_t *sb;

	if (!bitmap || !bitmap->sb_page)
		return;
515
	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
516
	printk(KERN_DEBUG "%s: bitmap file superblock:\n", bmname(bitmap));
517 518 519
	printk(KERN_DEBUG "         magic: %08x\n", le32_to_cpu(sb->magic));
	printk(KERN_DEBUG "       version: %d\n", le32_to_cpu(sb->version));
	printk(KERN_DEBUG "          uuid: %08x.%08x.%08x.%08x\n",
520 521 522 523
					*(__u32 *)(sb->uuid+0),
					*(__u32 *)(sb->uuid+4),
					*(__u32 *)(sb->uuid+8),
					*(__u32 *)(sb->uuid+12));
524
	printk(KERN_DEBUG "        events: %llu\n",
525
			(unsigned long long) le64_to_cpu(sb->events));
526
	printk(KERN_DEBUG "events cleared: %llu\n",
527
			(unsigned long long) le64_to_cpu(sb->events_cleared));
528 529 530 531 532
	printk(KERN_DEBUG "         state: %08x\n", le32_to_cpu(sb->state));
	printk(KERN_DEBUG "     chunksize: %d B\n", le32_to_cpu(sb->chunksize));
	printk(KERN_DEBUG "  daemon sleep: %ds\n", le32_to_cpu(sb->daemon_sleep));
	printk(KERN_DEBUG "     sync size: %llu KB\n",
			(unsigned long long)le64_to_cpu(sb->sync_size)/2);
533
	printk(KERN_DEBUG "max write behind: %d\n", le32_to_cpu(sb->write_behind));
534
	kunmap_atomic(sb, KM_USER0);
535 536 537 538 539 540 541
}

/* read the superblock from the bitmap file and initialize some bitmap fields */
static int bitmap_read_sb(struct bitmap *bitmap)
{
	char *reason = NULL;
	bitmap_super_t *sb;
542
	unsigned long chunksize, daemon_sleep, write_behind;
543 544 545 546
	unsigned long long events;
	int err = -EINVAL;

	/* page 0 is the superblock, read it... */
547 548 549 550 551 552
	if (bitmap->file) {
		loff_t isize = i_size_read(bitmap->file->f_mapping->host);
		int bytes = isize > PAGE_SIZE ? PAGE_SIZE : isize;

		bitmap->sb_page = read_page(bitmap->file, 0, bitmap, bytes);
	} else {
553 554
		bitmap->sb_page = read_sb_page(bitmap->mddev,
					       bitmap->mddev->bitmap_info.offset,
555 556
					       NULL,
					       0, sizeof(bitmap_super_t));
557
	}
558 559 560 561 562 563
	if (IS_ERR(bitmap->sb_page)) {
		err = PTR_ERR(bitmap->sb_page);
		bitmap->sb_page = NULL;
		return err;
	}

564
	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
565 566

	chunksize = le32_to_cpu(sb->chunksize);
567
	daemon_sleep = le32_to_cpu(sb->daemon_sleep) * HZ;
568
	write_behind = le32_to_cpu(sb->write_behind);
569 570 571 572

	/* verify that the bitmap-specific fields are valid */
	if (sb->magic != cpu_to_le32(BITMAP_MAGIC))
		reason = "bad magic";
573 574
	else if (le32_to_cpu(sb->version) < BITMAP_MAJOR_LO ||
		 le32_to_cpu(sb->version) > BITMAP_MAJOR_HI)
575
		reason = "unrecognized superblock version";
576
	else if (chunksize < 512)
577
		reason = "bitmap chunksize too small";
578 579
	else if ((1 << ffz(~chunksize)) != chunksize)
		reason = "bitmap chunksize not a power of 2";
580
	else if (daemon_sleep < 1 || daemon_sleep > MAX_SCHEDULE_TIMEOUT)
581
		reason = "daemon sleep period out of range";
582 583
	else if (write_behind > COUNTER_MAX)
		reason = "write-behind limit out of range (0 - 16383)";
584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609
	if (reason) {
		printk(KERN_INFO "%s: invalid bitmap file superblock: %s\n",
			bmname(bitmap), reason);
		goto out;
	}

	/* keep the array size field of the bitmap superblock up to date */
	sb->sync_size = cpu_to_le64(bitmap->mddev->resync_max_sectors);

	if (!bitmap->mddev->persistent)
		goto success;

	/*
	 * if we have a persistent array superblock, compare the
	 * bitmap's UUID and event counter to the mddev's
	 */
	if (memcmp(sb->uuid, bitmap->mddev->uuid, 16)) {
		printk(KERN_INFO "%s: bitmap superblock UUID mismatch\n",
			bmname(bitmap));
		goto out;
	}
	events = le64_to_cpu(sb->events);
	if (events < bitmap->mddev->events) {
		printk(KERN_INFO "%s: bitmap file is out of date (%llu < %llu) "
			"-- forcing full recovery\n", bmname(bitmap), events,
			(unsigned long long) bitmap->mddev->events);
610
		sb->state |= cpu_to_le32(BITMAP_STALE);
611 612 613
	}
success:
	/* assign fields using values from superblock */
614 615 616
	bitmap->mddev->bitmap_info.chunksize = chunksize;
	bitmap->mddev->bitmap_info.daemon_sleep = daemon_sleep;
	bitmap->mddev->bitmap_info.max_write_behind = write_behind;
617
	bitmap->flags |= le32_to_cpu(sb->state);
618 619
	if (le32_to_cpu(sb->version) == BITMAP_MAJOR_HOSTENDIAN)
		bitmap->flags |= BITMAP_HOSTENDIAN;
620
	bitmap->events_cleared = le64_to_cpu(sb->events_cleared);
621
	if (bitmap->flags & BITMAP_STALE)
622
		bitmap->events_cleared = bitmap->mddev->events;
623 624
	err = 0;
out:
625
	kunmap_atomic(sb, KM_USER0);
626 627 628 629 630 631 632 633 634 635
	if (err)
		bitmap_print_sb(bitmap);
	return err;
}

enum bitmap_mask_op {
	MASK_SET,
	MASK_UNSET
};

636 637 638
/* record the state of the bitmap in the superblock.  Return the old value */
static int bitmap_mask_state(struct bitmap *bitmap, enum bitmap_state bits,
			     enum bitmap_mask_op op)
639 640 641
{
	bitmap_super_t *sb;
	unsigned long flags;
642
	int old;
643 644

	spin_lock_irqsave(&bitmap->lock, flags);
645
	if (!bitmap->sb_page) { /* can't set the state */
646
		spin_unlock_irqrestore(&bitmap->lock, flags);
647
		return 0;
648 649
	}
	spin_unlock_irqrestore(&bitmap->lock, flags);
650
	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
651
	old = le32_to_cpu(sb->state) & bits;
652
	switch (op) {
653 654
	case MASK_SET:
		sb->state |= cpu_to_le32(bits);
655
		bitmap->flags |= bits;
656 657 658
		break;
	case MASK_UNSET:
		sb->state &= cpu_to_le32(~bits);
659
		bitmap->flags &= ~bits;
660 661 662
		break;
	default:
		BUG();
663
	}
664
	kunmap_atomic(sb, KM_USER0);
665
	return old;
666 667 668 669 670 671
}

/*
 * general bitmap file operations
 */

672 673 674 675 676 677
/*
 * on-disk bitmap:
 *
 * Use one bit per "chunk" (block set). We do the disk I/O on the bitmap
 * file a page at a time. There's a superblock at the start of the file.
 */
678
/* calculate the index of the page that contains this bit */
679
static inline unsigned long file_page_index(struct bitmap *bitmap, unsigned long chunk)
680
{
681 682 683
	if (!bitmap->mddev->bitmap_info.external)
		chunk += sizeof(bitmap_super_t) << 3;
	return chunk >> PAGE_BIT_SHIFT;
684 685 686
}

/* calculate the (bit) offset of this bit within a page */
687
static inline unsigned long file_page_offset(struct bitmap *bitmap, unsigned long chunk)
688
{
689 690 691
	if (!bitmap->mddev->bitmap_info.external)
		chunk += sizeof(bitmap_super_t) << 3;
	return chunk & (PAGE_BITS - 1);
692 693 694 695 696 697 698 699 700 701 702 703
}

/*
 * return a pointer to the page in the filemap that contains the given bit
 *
 * this lookup is complicated by the fact that the bitmap sb might be exactly
 * 1 page (e.g., x86) or less than 1 page -- so the bitmap might start on page
 * 0 or page 1
 */
static inline struct page *filemap_get_page(struct bitmap *bitmap,
					unsigned long chunk)
{
704 705
	if (bitmap->filemap == NULL)
		return NULL;
706 707
	if (file_page_index(bitmap, chunk) >= bitmap->file_pages)
		return NULL;
708 709
	return bitmap->filemap[file_page_index(bitmap, chunk)
			       - file_page_index(bitmap, 0)];
710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730
}

static void bitmap_file_unmap(struct bitmap *bitmap)
{
	struct page **map, *sb_page;
	unsigned long *attr;
	int pages;
	unsigned long flags;

	spin_lock_irqsave(&bitmap->lock, flags);
	map = bitmap->filemap;
	bitmap->filemap = NULL;
	attr = bitmap->filemap_attr;
	bitmap->filemap_attr = NULL;
	pages = bitmap->file_pages;
	bitmap->file_pages = 0;
	sb_page = bitmap->sb_page;
	bitmap->sb_page = NULL;
	spin_unlock_irqrestore(&bitmap->lock, flags);

	while (pages--)
731
		if (map[pages] != sb_page) /* 0 is sb_page, release it below */
732
			free_buffers(map[pages]);
733 734 735
	kfree(map);
	kfree(attr);

736 737
	if (sb_page)
		free_buffers(sb_page);
738 739 740 741 742 743 744 745 746 747 748 749
}

static void bitmap_file_put(struct bitmap *bitmap)
{
	struct file *file;
	unsigned long flags;

	spin_lock_irqsave(&bitmap->lock, flags);
	file = bitmap->file;
	bitmap->file = NULL;
	spin_unlock_irqrestore(&bitmap->lock, flags);

750 751 752
	if (file)
		wait_event(bitmap->write_wait,
			   atomic_read(&bitmap->pending_writes)==0);
753 754
	bitmap_file_unmap(bitmap);

755
	if (file) {
J
Josef Sipek 已提交
756
		struct inode *inode = file->f_path.dentry->d_inode;
757
		invalidate_mapping_pages(inode->i_mapping, 0, -1);
758
		fput(file);
759
	}
760 761 762 763 764 765 766 767 768 769 770
}

/*
 * bitmap_file_kick - if an error occurs while manipulating the bitmap file
 * then it is no longer reliable, so we stop using it and we mark the file
 * as failed in the superblock
 */
static void bitmap_file_kick(struct bitmap *bitmap)
{
	char *path, *ptr = NULL;

771 772
	if (bitmap_mask_state(bitmap, BITMAP_STALE, MASK_SET) == 0) {
		bitmap_update_sb(bitmap);
773

774 775 776
		if (bitmap->file) {
			path = kmalloc(PAGE_SIZE, GFP_KERNEL);
			if (path)
C
Christoph Hellwig 已提交
777 778 779
				ptr = d_path(&bitmap->file->f_path, path,
					     PAGE_SIZE);

780 781
			printk(KERN_ALERT
			      "%s: kicking failed bitmap file %s from array!\n",
C
Christoph Hellwig 已提交
782
			      bmname(bitmap), IS_ERR(ptr) ? "" : ptr);
783

784 785 786 787 788
			kfree(path);
		} else
			printk(KERN_ALERT
			       "%s: disabling internal bitmap due to errors\n",
			       bmname(bitmap));
789
	}
790 791 792 793 794 795 796

	bitmap_file_put(bitmap);

	return;
}

enum bitmap_page_attr {
797 798 799
	BITMAP_PAGE_DIRTY = 0,     /* there are set bits that need to be synced */
	BITMAP_PAGE_CLEAN = 1,     /* there are bits that might need to be cleared */
	BITMAP_PAGE_NEEDWRITE = 2, /* there are cleared bits that need to be synced */
800 801 802 803 804
};

static inline void set_page_attr(struct bitmap *bitmap, struct page *page,
				enum bitmap_page_attr attr)
{
805 806 807 808
	if (page)
		__set_bit((page->index<<2) + attr, bitmap->filemap_attr);
	else
		__set_bit(attr, &bitmap->logattrs);
809 810 811 812 813
}

static inline void clear_page_attr(struct bitmap *bitmap, struct page *page,
				enum bitmap_page_attr attr)
{
814 815 816 817
	if (page)
		__clear_bit((page->index<<2) + attr, bitmap->filemap_attr);
	else
		__clear_bit(attr, &bitmap->logattrs);
818 819
}

820 821
static inline unsigned long test_page_attr(struct bitmap *bitmap, struct page *page,
					   enum bitmap_page_attr attr)
822
{
823 824 825 826
	if (page)
		return test_bit((page->index<<2) + attr, bitmap->filemap_attr);
	else
		return test_bit(attr, &bitmap->logattrs);
827 828 829 830 831 832 833 834 835 836 837 838
}

/*
 * bitmap_file_set_bit -- called before performing a write to the md device
 * to set (and eventually sync) a particular bit in the bitmap file
 *
 * we set the bit immediately, then we record the page number so that
 * when an unplug occurs, we can flush the dirty pages out to disk
 */
static void bitmap_file_set_bit(struct bitmap *bitmap, sector_t block)
{
	unsigned long bit;
839
	struct page *page = NULL;
840 841 842
	void *kaddr;
	unsigned long chunk = block >> CHUNK_BLOCK_SHIFT(bitmap);

843 844 845 846 847
	if (!bitmap->filemap) {
		struct dm_dirty_log *log = bitmap->mddev->bitmap_info.log;
		if (log)
			log->type->mark_region(log, chunk);
	} else {
848

849 850 851 852
		page = filemap_get_page(bitmap, chunk);
		if (!page)
			return;
		bit = file_page_offset(bitmap, chunk);
853

854 855 856 857 858
		/* set the bit */
		kaddr = kmap_atomic(page, KM_USER0);
		if (bitmap->flags & BITMAP_HOSTENDIAN)
			set_bit(bit, kaddr);
		else
A
Akinobu Mita 已提交
859
			__test_and_set_bit_le(bit, kaddr);
860 861 862
		kunmap_atomic(kaddr, KM_USER0);
		PRINTK("set file bit %lu page %lu\n", bit, page->index);
	}
863 864 865 866 867 868 869
	/* record page number so it gets flushed to disk when unplug occurs */
	set_page_attr(bitmap, page, BITMAP_PAGE_DIRTY);
}

/* this gets called when the md device is ready to unplug its underlying
 * (slave) device queues -- before we let any writes go down, we need to
 * sync the dirty pages of the bitmap file to disk */
870
void bitmap_unplug(struct bitmap *bitmap)
871
{
872 873
	unsigned long i, flags;
	int dirty, need_write;
874 875 876 877
	struct page *page;
	int wait = 0;

	if (!bitmap)
878
		return;
879 880 881 882 883 884 885 886 887 888
	if (!bitmap->filemap) {
		/* Must be using a dirty_log */
		struct dm_dirty_log *log = bitmap->mddev->bitmap_info.log;
		dirty = test_and_clear_bit(BITMAP_PAGE_DIRTY, &bitmap->logattrs);
		need_write = test_and_clear_bit(BITMAP_PAGE_NEEDWRITE, &bitmap->logattrs);
		if (dirty || need_write)
			if (log->type->flush(log))
				bitmap->flags |= BITMAP_WRITE_ERROR;
		goto out;
	}
889 890 891 892 893

	/* look at each page to see if there are any set bits that need to be
	 * flushed out to disk */
	for (i = 0; i < bitmap->file_pages; i++) {
		spin_lock_irqsave(&bitmap->lock, flags);
894
		if (!bitmap->filemap) {
895
			spin_unlock_irqrestore(&bitmap->lock, flags);
896
			return;
897 898
		}
		page = bitmap->filemap[i];
899 900
		dirty = test_page_attr(bitmap, page, BITMAP_PAGE_DIRTY);
		need_write = test_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
901 902
		clear_page_attr(bitmap, page, BITMAP_PAGE_DIRTY);
		clear_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
903
		if (dirty)
904 905 906
			wait = 1;
		spin_unlock_irqrestore(&bitmap->lock, flags);

907
		if (dirty || need_write)
908
			write_page(bitmap, page, 0);
909 910
	}
	if (wait) { /* if any writes were performed, we need to wait on them */
911
		if (bitmap->file)
912 913
			wait_event(bitmap->write_wait,
				   atomic_read(&bitmap->pending_writes)==0);
914
		else
915
			md_super_wait(bitmap->mddev);
916
	}
917
out:
918 919
	if (bitmap->flags & BITMAP_WRITE_ERROR)
		bitmap_file_kick(bitmap);
920
}
921
EXPORT_SYMBOL(bitmap_unplug);
922

923
static void bitmap_set_memory_bits(struct bitmap *bitmap, sector_t offset, int needed);
924 925 926 927 928 929 930
/* * bitmap_init_from_disk -- called at bitmap_create time to initialize
 * the in-memory bitmap from the on-disk bitmap -- also, sets up the
 * memory mapping of the bitmap file
 * Special cases:
 *   if there's no bitmap file, or if the bitmap file had been
 *   previously kicked from the array, we mark all the bits as
 *   1's in order to cause a full resync.
931 932 933
 *
 * We ignore all bits for sectors that end earlier than 'start'.
 * This is used when reading an out-of-date bitmap...
934
 */
935
static int bitmap_init_from_disk(struct bitmap *bitmap, sector_t start)
936 937 938 939 940
{
	unsigned long i, chunks, index, oldindex, bit;
	struct page *page = NULL, *oldpage = NULL;
	unsigned long num_pages, bit_cnt = 0;
	struct file *file;
941
	unsigned long bytes, offset;
942 943
	int outofdate;
	int ret = -ENOSPC;
944
	void *paddr;
945 946 947 948

	chunks = bitmap->chunks;
	file = bitmap->file;

949
	BUG_ON(!file && !bitmap->mddev->bitmap_info.offset);
950

951
#ifdef INJECT_FAULTS_3
952 953 954 955 956 957 958 959
	outofdate = 1;
#else
	outofdate = bitmap->flags & BITMAP_STALE;
#endif
	if (outofdate)
		printk(KERN_INFO "%s: bitmap file is out of date, doing full "
			"recovery\n", bmname(bitmap));

960
	bytes = DIV_ROUND_UP(bitmap->chunks, 8);
961 962
	if (!bitmap->mddev->bitmap_info.external)
		bytes += sizeof(bitmap_super_t);
963

964
	num_pages = DIV_ROUND_UP(bytes, PAGE_SIZE);
965

966
	if (file && i_size_read(file->f_mapping->host) < bytes) {
967 968 969
		printk(KERN_INFO "%s: bitmap file too short %lu < %lu\n",
			bmname(bitmap),
			(unsigned long) i_size_read(file->f_mapping->host),
970
			bytes);
971
		goto err;
972
	}
973 974 975

	ret = -ENOMEM;

976
	bitmap->filemap = kmalloc(sizeof(struct page *) * num_pages, GFP_KERNEL);
977
	if (!bitmap->filemap)
978
		goto err;
979

980 981
	/* We need 4 bits per page, rounded up to a multiple of sizeof(unsigned long) */
	bitmap->filemap_attr = kzalloc(
982
		roundup(DIV_ROUND_UP(num_pages*4, 8), sizeof(unsigned long)),
983
		GFP_KERNEL);
984
	if (!bitmap->filemap_attr)
985
		goto err;
986 987 988 989

	oldindex = ~0L;

	for (i = 0; i < chunks; i++) {
990
		int b;
991 992
		index = file_page_index(bitmap, i);
		bit = file_page_offset(bitmap, i);
993
		if (index != oldindex) { /* this is a new page, read it in */
994
			int count;
995
			/* unmap the old page, we're done with it */
996
			if (index == num_pages-1)
997
				count = bytes - index * PAGE_SIZE;
998 999
			else
				count = PAGE_SIZE;
1000
			if (index == 0 && bitmap->sb_page) {
1001 1002 1003 1004 1005 1006 1007
				/*
				 * if we're here then the superblock page
				 * contains some bits (PAGE_SIZE != sizeof sb)
				 * we've already read it in, so just use it
				 */
				page = bitmap->sb_page;
				offset = sizeof(bitmap_super_t);
N
NeilBrown 已提交
1008
				if (!file)
1009 1010 1011 1012 1013
					page = read_sb_page(
						bitmap->mddev,
						bitmap->mddev->bitmap_info.offset,
						page,
						index, count);
1014
			} else if (file) {
1015
				page = read_page(file, index, bitmap, count);
1016 1017
				offset = 0;
			} else {
1018 1019
				page = read_sb_page(bitmap->mddev,
						    bitmap->mddev->bitmap_info.offset,
1020 1021
						    NULL,
						    index, count);
1022 1023
				offset = 0;
			}
1024 1025
			if (IS_ERR(page)) { /* read error */
				ret = PTR_ERR(page);
1026
				goto err;
1027 1028
			}

1029 1030 1031
			oldindex = index;
			oldpage = page;

1032 1033 1034
			bitmap->filemap[bitmap->file_pages++] = page;
			bitmap->last_page_size = count;

1035 1036 1037
			if (outofdate) {
				/*
				 * if bitmap is out of date, dirty the
1038
				 * whole page and write it out
1039
				 */
1040 1041
				paddr = kmap_atomic(page, KM_USER0);
				memset(paddr + offset, 0xff,
1042
				       PAGE_SIZE - offset);
1043
				kunmap_atomic(paddr, KM_USER0);
1044 1045 1046
				write_page(bitmap, page, 1);

				ret = -EIO;
1047
				if (bitmap->flags & BITMAP_WRITE_ERROR)
1048
					goto err;
1049 1050
			}
		}
1051
		paddr = kmap_atomic(page, KM_USER0);
1052
		if (bitmap->flags & BITMAP_HOSTENDIAN)
1053
			b = test_bit(bit, paddr);
1054
		else
A
Akinobu Mita 已提交
1055
			b = test_bit_le(bit, paddr);
1056
		kunmap_atomic(paddr, KM_USER0);
1057
		if (b) {
1058
			/* if the disk bit is set, set the memory bit */
1059 1060 1061 1062 1063
			int needed = ((sector_t)(i+1) << (CHUNK_BLOCK_SHIFT(bitmap))
				      >= start);
			bitmap_set_memory_bits(bitmap,
					       (sector_t)i << CHUNK_BLOCK_SHIFT(bitmap),
					       needed);
1064
			bit_cnt++;
1065
			set_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
1066 1067 1068
		}
	}

1069
	/* everything went OK */
1070 1071 1072 1073 1074 1075 1076 1077 1078
	ret = 0;
	bitmap_mask_state(bitmap, BITMAP_STALE, MASK_UNSET);

	if (bit_cnt) { /* Kick recovery if any bits were set */
		set_bit(MD_RECOVERY_NEEDED, &bitmap->mddev->recovery);
		md_wakeup_thread(bitmap->mddev->thread);
	}

	printk(KERN_INFO "%s: bitmap initialized from disk: "
1079 1080 1081 1082
		"read %lu/%lu pages, set %lu bits\n",
		bmname(bitmap), bitmap->file_pages, num_pages, bit_cnt);

	return 0;
1083

1084 1085 1086
 err:
	printk(KERN_INFO "%s: bitmap initialisation failed: %d\n",
	       bmname(bitmap), ret);
1087 1088 1089
	return ret;
}

1090 1091 1092 1093 1094
void bitmap_write_all(struct bitmap *bitmap)
{
	/* We don't actually write all bitmap blocks here,
	 * just flag them as needing to be written
	 */
1095
	int i;
1096

1097
	for (i = 0; i < bitmap->file_pages; i++)
1098 1099
		set_page_attr(bitmap, bitmap->filemap[i],
			      BITMAP_PAGE_NEEDWRITE);
1100 1101
}

1102 1103 1104 1105 1106 1107 1108 1109
static void bitmap_count_page(struct bitmap *bitmap, sector_t offset, int inc)
{
	sector_t chunk = offset >> CHUNK_BLOCK_SHIFT(bitmap);
	unsigned long page = chunk >> PAGE_COUNTER_SHIFT;
	bitmap->bp[page].count += inc;
	bitmap_checkfree(bitmap, page);
}
static bitmap_counter_t *bitmap_get_counter(struct bitmap *bitmap,
N
NeilBrown 已提交
1110
					    sector_t offset, sector_t *blocks,
1111 1112 1113 1114 1115 1116 1117
					    int create);

/*
 * bitmap daemon -- periodically wakes up to clean bits and flush pages
 *			out to disk
 */

1118
void bitmap_daemon_work(mddev_t *mddev)
1119
{
1120
	struct bitmap *bitmap;
1121
	unsigned long j;
1122 1123
	unsigned long flags;
	struct page *page = NULL, *lastpage = NULL;
N
NeilBrown 已提交
1124
	sector_t blocks;
1125
	void *paddr;
1126
	struct dm_dirty_log *log = mddev->bitmap_info.log;
1127

1128 1129 1130
	/* Use a mutex to guard daemon_work against
	 * bitmap_destroy.
	 */
1131
	mutex_lock(&mddev->bitmap_info.mutex);
1132 1133
	bitmap = mddev->bitmap;
	if (bitmap == NULL) {
1134
		mutex_unlock(&mddev->bitmap_info.mutex);
1135
		return;
1136
	}
1137
	if (time_before(jiffies, bitmap->daemon_lastrun
1138
			+ bitmap->mddev->bitmap_info.daemon_sleep))
1139 1140
		goto done;

1141
	bitmap->daemon_lastrun = jiffies;
1142 1143
	if (bitmap->allclean) {
		bitmap->mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
1144
		goto done;
1145 1146
	}
	bitmap->allclean = 1;
1147

1148
	spin_lock_irqsave(&bitmap->lock, flags);
1149 1150
	for (j = 0; j < bitmap->chunks; j++) {
		bitmap_counter_t *bmc;
1151 1152 1153 1154 1155 1156
		if (!bitmap->filemap) {
			if (!log)
				/* error or shutdown */
				break;
		} else
			page = filemap_get_page(bitmap, j);
1157 1158

		if (page != lastpage) {
1159
			/* skip this page unless it's marked as needing cleaning */
1160 1161 1162
			if (!test_page_attr(bitmap, page, BITMAP_PAGE_CLEAN)) {
				int need_write = test_page_attr(bitmap, page,
								BITMAP_PAGE_NEEDWRITE);
1163
				if (need_write)
1164
					clear_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
1165

1166
				spin_unlock_irqrestore(&bitmap->lock, flags);
1167
				if (need_write) {
1168
					write_page(bitmap, page, 0);
1169 1170
					bitmap->allclean = 0;
				}
1171 1172
				spin_lock_irqsave(&bitmap->lock, flags);
				j |= (PAGE_BITS - 1);
1173 1174 1175
				continue;
			}

1176 1177
			/* grab the new page, sync and release the old */
			if (lastpage != NULL) {
1178
				if (test_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE)) {
1179 1180
					clear_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
					spin_unlock_irqrestore(&bitmap->lock, flags);
1181
					write_page(bitmap, lastpage, 0);
1182 1183 1184 1185 1186 1187 1188
				} else {
					set_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
					spin_unlock_irqrestore(&bitmap->lock, flags);
				}
			} else
				spin_unlock_irqrestore(&bitmap->lock, flags);
			lastpage = page;
1189 1190 1191 1192

			/* We are possibly going to clear some bits, so make
			 * sure that events_cleared is up-to-date.
			 */
1193 1194
			if (bitmap->need_sync &&
			    bitmap->mddev->bitmap_info.external == 0) {
1195 1196 1197 1198 1199 1200 1201 1202
				bitmap_super_t *sb;
				bitmap->need_sync = 0;
				sb = kmap_atomic(bitmap->sb_page, KM_USER0);
				sb->events_cleared =
					cpu_to_le64(bitmap->events_cleared);
				kunmap_atomic(sb, KM_USER0);
				write_page(bitmap, bitmap->sb_page, 1);
			}
1203
			spin_lock_irqsave(&bitmap->lock, flags);
1204 1205
			if (!bitmap->need_sync)
				clear_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
1206
		}
1207 1208 1209
		bmc = bitmap_get_counter(bitmap,
					 (sector_t)j << CHUNK_BLOCK_SHIFT(bitmap),
					 &blocks, 0);
1210
		if (bmc) {
1211 1212 1213
			if (*bmc)
				bitmap->allclean = 0;

1214
			if (*bmc == 2) {
1215
				*bmc = 1; /* maybe clear the bit next time */
1216
				set_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
1217
			} else if (*bmc == 1 && !bitmap->need_sync) {
1218 1219
				/* we can clear the bit */
				*bmc = 0;
1220 1221
				bitmap_count_page(bitmap,
						  (sector_t)j << CHUNK_BLOCK_SHIFT(bitmap),
1222 1223 1224
						  -1);

				/* clear the bit */
1225 1226 1227 1228 1229 1230
				if (page) {
					paddr = kmap_atomic(page, KM_USER0);
					if (bitmap->flags & BITMAP_HOSTENDIAN)
						clear_bit(file_page_offset(bitmap, j),
							  paddr);
					else
A
Akinobu Mita 已提交
1231
						__test_and_clear_bit_le(file_page_offset(bitmap, j),
1232 1233 1234 1235
							       paddr);
					kunmap_atomic(paddr, KM_USER0);
				} else
					log->type->clear_region(log, j);
1236
			}
1237 1238
		} else
			j |= PAGE_COUNTER_MASK;
1239
	}
1240
	spin_unlock_irqrestore(&bitmap->lock, flags);
1241 1242

	/* now sync the final page */
1243
	if (lastpage != NULL || log != NULL) {
1244
		spin_lock_irqsave(&bitmap->lock, flags);
1245
		if (test_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE)) {
1246 1247
			clear_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
			spin_unlock_irqrestore(&bitmap->lock, flags);
1248 1249 1250 1251 1252
			if (lastpage)
				write_page(bitmap, lastpage, 0);
			else
				if (log->type->flush(log))
					bitmap->flags |= BITMAP_WRITE_ERROR;
1253 1254 1255 1256 1257 1258
		} else {
			set_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
			spin_unlock_irqrestore(&bitmap->lock, flags);
		}
	}

1259
 done:
1260
	if (bitmap->allclean == 0)
1261
		bitmap->mddev->thread->timeout =
1262
			bitmap->mddev->bitmap_info.daemon_sleep;
1263
	mutex_unlock(&mddev->bitmap_info.mutex);
1264 1265 1266
}

static bitmap_counter_t *bitmap_get_counter(struct bitmap *bitmap,
N
NeilBrown 已提交
1267
					    sector_t offset, sector_t *blocks,
1268
					    int create)
1269 1270
__releases(bitmap->lock)
__acquires(bitmap->lock)
1271 1272 1273 1274 1275 1276 1277 1278 1279
{
	/* If 'create', we might release the lock and reclaim it.
	 * The lock must have been taken with interrupts enabled.
	 * If !create, we don't release the lock.
	 */
	sector_t chunk = offset >> CHUNK_BLOCK_SHIFT(bitmap);
	unsigned long page = chunk >> PAGE_COUNTER_SHIFT;
	unsigned long pageoff = (chunk & PAGE_COUNTER_MASK) << COUNTER_BYTE_SHIFT;
	sector_t csize;
1280
	int err;
1281

1282 1283 1284 1285 1286 1287 1288
	err = bitmap_checkpage(bitmap, page, create);

	if (bitmap->bp[page].hijacked ||
	    bitmap->bp[page].map == NULL)
		csize = ((sector_t)1) << (CHUNK_BLOCK_SHIFT(bitmap) +
					  PAGE_COUNTER_SHIFT - 1);
	else
1289
		csize = ((sector_t)1) << (CHUNK_BLOCK_SHIFT(bitmap));
1290 1291 1292
	*blocks = csize - (offset & (csize - 1));

	if (err < 0)
1293
		return NULL;
1294

1295 1296 1297 1298 1299 1300 1301 1302
	/* now locked ... */

	if (bitmap->bp[page].hijacked) { /* hijacked pointer */
		/* should we use the first or second counter field
		 * of the hijacked pointer? */
		int hi = (pageoff > PAGE_COUNTER_MASK);
		return  &((bitmap_counter_t *)
			  &bitmap->bp[page].map)[hi];
1303
	} else /* page is allocated */
1304 1305 1306 1307
		return (bitmap_counter_t *)
			&(bitmap->bp[page].map[pageoff]);
}

1308
int bitmap_startwrite(struct bitmap *bitmap, sector_t offset, unsigned long sectors, int behind)
1309
{
1310 1311
	if (!bitmap)
		return 0;
1312 1313

	if (behind) {
1314
		int bw;
1315
		atomic_inc(&bitmap->behind_writes);
1316 1317 1318 1319
		bw = atomic_read(&bitmap->behind_writes);
		if (bw > bitmap->behind_writes_used)
			bitmap->behind_writes_used = bw;

1320
		PRINTK(KERN_DEBUG "inc write-behind count %d/%d\n",
1321
		       bw, bitmap->max_write_behind);
1322 1323
	}

1324
	while (sectors) {
N
NeilBrown 已提交
1325
		sector_t blocks;
1326 1327 1328 1329 1330 1331 1332 1333 1334
		bitmap_counter_t *bmc;

		spin_lock_irq(&bitmap->lock);
		bmc = bitmap_get_counter(bitmap, offset, &blocks, 1);
		if (!bmc) {
			spin_unlock_irq(&bitmap->lock);
			return 0;
		}

1335 1336 1337 1338 1339 1340 1341 1342 1343
		if (unlikely((*bmc & COUNTER_MAX) == COUNTER_MAX)) {
			DEFINE_WAIT(__wait);
			/* note that it is safe to do the prepare_to_wait
			 * after the test as long as we do it before dropping
			 * the spinlock.
			 */
			prepare_to_wait(&bitmap->overflow_wait, &__wait,
					TASK_UNINTERRUPTIBLE);
			spin_unlock_irq(&bitmap->lock);
J
Jens Axboe 已提交
1344
			io_schedule();
1345 1346 1347 1348
			finish_wait(&bitmap->overflow_wait, &__wait);
			continue;
		}

1349
		switch (*bmc) {
1350 1351
		case 0:
			bitmap_file_set_bit(bitmap, offset);
1352
			bitmap_count_page(bitmap, offset, 1);
1353 1354 1355 1356
			/* fall through */
		case 1:
			*bmc = 2;
		}
1357

1358 1359 1360 1361 1362 1363 1364
		(*bmc)++;

		spin_unlock_irq(&bitmap->lock);

		offset += blocks;
		if (sectors > blocks)
			sectors -= blocks;
1365 1366
		else
			sectors = 0;
1367
	}
1368
	bitmap->allclean = 0;
1369 1370
	return 0;
}
1371
EXPORT_SYMBOL(bitmap_startwrite);
1372 1373

void bitmap_endwrite(struct bitmap *bitmap, sector_t offset, unsigned long sectors,
1374
		     int success, int behind)
1375
{
1376 1377
	if (!bitmap)
		return;
1378
	if (behind) {
1379 1380
		if (atomic_dec_and_test(&bitmap->behind_writes))
			wake_up(&bitmap->behind_wait);
1381 1382 1383
		PRINTK(KERN_DEBUG "dec write-behind count %d/%d\n",
		  atomic_read(&bitmap->behind_writes), bitmap->max_write_behind);
	}
1384 1385 1386
	if (bitmap->mddev->degraded)
		/* Never clear bits or update events_cleared when degraded */
		success = 0;
1387

1388
	while (sectors) {
N
NeilBrown 已提交
1389
		sector_t blocks;
1390 1391 1392 1393 1394 1395 1396 1397 1398 1399
		unsigned long flags;
		bitmap_counter_t *bmc;

		spin_lock_irqsave(&bitmap->lock, flags);
		bmc = bitmap_get_counter(bitmap, offset, &blocks, 0);
		if (!bmc) {
			spin_unlock_irqrestore(&bitmap->lock, flags);
			return;
		}

1400 1401 1402 1403
		if (success &&
		    bitmap->events_cleared < bitmap->mddev->events) {
			bitmap->events_cleared = bitmap->mddev->events;
			bitmap->need_sync = 1;
1404
			sysfs_notify_dirent_safe(bitmap->sysfs_can_clear);
1405 1406
		}

1407 1408 1409
		if (!success && ! (*bmc & NEEDED_MASK))
			*bmc |= NEEDED_MASK;

1410 1411 1412
		if ((*bmc & COUNTER_MAX) == COUNTER_MAX)
			wake_up(&bitmap->overflow_wait);

1413
		(*bmc)--;
1414
		if (*bmc <= 2)
1415
			set_page_attr(bitmap,
1416 1417 1418
				      filemap_get_page(
					      bitmap,
					      offset >> CHUNK_BLOCK_SHIFT(bitmap)),
1419
				      BITMAP_PAGE_CLEAN);
1420

1421 1422 1423 1424
		spin_unlock_irqrestore(&bitmap->lock, flags);
		offset += blocks;
		if (sectors > blocks)
			sectors -= blocks;
1425 1426
		else
			sectors = 0;
1427 1428
	}
}
1429
EXPORT_SYMBOL(bitmap_endwrite);
1430

N
NeilBrown 已提交
1431
static int __bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks,
1432
			       int degraded)
1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448
{
	bitmap_counter_t *bmc;
	int rv;
	if (bitmap == NULL) {/* FIXME or bitmap set as 'failed' */
		*blocks = 1024;
		return 1; /* always resync if no bitmap */
	}
	spin_lock_irq(&bitmap->lock);
	bmc = bitmap_get_counter(bitmap, offset, blocks, 0);
	rv = 0;
	if (bmc) {
		/* locked */
		if (RESYNC(*bmc))
			rv = 1;
		else if (NEEDED(*bmc)) {
			rv = 1;
1449 1450 1451 1452
			if (!degraded) { /* don't set/clear bits if degraded */
				*bmc |= RESYNC_MASK;
				*bmc &= ~NEEDED_MASK;
			}
1453 1454 1455
		}
	}
	spin_unlock_irq(&bitmap->lock);
1456
	bitmap->allclean = 0;
1457 1458 1459
	return rv;
}

N
NeilBrown 已提交
1460
int bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks,
1461 1462 1463 1464 1465 1466 1467 1468 1469 1470
		      int degraded)
{
	/* bitmap_start_sync must always report on multiples of whole
	 * pages, otherwise resync (which is very PAGE_SIZE based) will
	 * get confused.
	 * So call __bitmap_start_sync repeatedly (if needed) until
	 * At least PAGE_SIZE>>9 blocks are covered.
	 * Return the 'or' of the result.
	 */
	int rv = 0;
N
NeilBrown 已提交
1471
	sector_t blocks1;
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481

	*blocks = 0;
	while (*blocks < (PAGE_SIZE>>9)) {
		rv |= __bitmap_start_sync(bitmap, offset,
					  &blocks1, degraded);
		offset += blocks1;
		*blocks += blocks1;
	}
	return rv;
}
1482
EXPORT_SYMBOL(bitmap_start_sync);
1483

N
NeilBrown 已提交
1484
void bitmap_end_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks, int aborted)
1485 1486 1487
{
	bitmap_counter_t *bmc;
	unsigned long flags;
1488 1489

	if (bitmap == NULL) {
1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503
		*blocks = 1024;
		return;
	}
	spin_lock_irqsave(&bitmap->lock, flags);
	bmc = bitmap_get_counter(bitmap, offset, blocks, 0);
	if (bmc == NULL)
		goto unlock;
	/* locked */
	if (RESYNC(*bmc)) {
		*bmc &= ~RESYNC_MASK;

		if (!NEEDED(*bmc) && aborted)
			*bmc |= NEEDED_MASK;
		else {
1504
			if (*bmc <= 2)
1505 1506 1507 1508 1509 1510 1511
				set_page_attr(bitmap,
					      filemap_get_page(bitmap, offset >> CHUNK_BLOCK_SHIFT(bitmap)),
					      BITMAP_PAGE_CLEAN);
		}
	}
 unlock:
	spin_unlock_irqrestore(&bitmap->lock, flags);
1512
	bitmap->allclean = 0;
1513
}
1514
EXPORT_SYMBOL(bitmap_end_sync);
1515 1516 1517 1518 1519 1520 1521 1522

void bitmap_close_sync(struct bitmap *bitmap)
{
	/* Sync has finished, and any bitmap chunks that weren't synced
	 * properly have been aborted.  It remains to us to clear the
	 * RESYNC bit wherever it is still on
	 */
	sector_t sector = 0;
N
NeilBrown 已提交
1523
	sector_t blocks;
N
NeilBrown 已提交
1524 1525
	if (!bitmap)
		return;
1526 1527
	while (sector < bitmap->mddev->resync_max_sectors) {
		bitmap_end_sync(bitmap, sector, &blocks, 0);
N
NeilBrown 已提交
1528 1529 1530
		sector += blocks;
	}
}
1531
EXPORT_SYMBOL(bitmap_close_sync);
N
NeilBrown 已提交
1532 1533 1534 1535

void bitmap_cond_end_sync(struct bitmap *bitmap, sector_t sector)
{
	sector_t s = 0;
N
NeilBrown 已提交
1536
	sector_t blocks;
N
NeilBrown 已提交
1537 1538 1539 1540 1541 1542 1543 1544

	if (!bitmap)
		return;
	if (sector == 0) {
		bitmap->last_end_sync = jiffies;
		return;
	}
	if (time_before(jiffies, (bitmap->last_end_sync
1545
				  + bitmap->mddev->bitmap_info.daemon_sleep)))
N
NeilBrown 已提交
1546 1547 1548 1549
		return;
	wait_event(bitmap->mddev->recovery_wait,
		   atomic_read(&bitmap->mddev->recovery_active) == 0);

1550
	bitmap->mddev->curr_resync_completed = sector;
1551
	set_bit(MD_CHANGE_CLEAN, &bitmap->mddev->flags);
N
NeilBrown 已提交
1552 1553 1554 1555 1556
	sector &= ~((1ULL << CHUNK_BLOCK_SHIFT(bitmap)) - 1);
	s = 0;
	while (s < sector && s < bitmap->mddev->resync_max_sectors) {
		bitmap_end_sync(bitmap, s, &blocks, 0);
		s += blocks;
1557
	}
N
NeilBrown 已提交
1558
	bitmap->last_end_sync = jiffies;
1559
	sysfs_notify(&bitmap->mddev->kobj, NULL, "sync_completed");
1560
}
1561
EXPORT_SYMBOL(bitmap_cond_end_sync);
1562

1563
static void bitmap_set_memory_bits(struct bitmap *bitmap, sector_t offset, int needed)
1564 1565
{
	/* For each chunk covered by any of these sectors, set the
1566
	 * counter to 1 and set resync_needed.  They should all
1567 1568
	 * be 0 at this point
	 */
1569

N
NeilBrown 已提交
1570
	sector_t secs;
1571 1572 1573 1574
	bitmap_counter_t *bmc;
	spin_lock_irq(&bitmap->lock);
	bmc = bitmap_get_counter(bitmap, offset, &secs, 1);
	if (!bmc) {
1575
		spin_unlock_irq(&bitmap->lock);
1576
		return;
1577
	}
1578
	if (!*bmc) {
1579
		struct page *page;
1580
		*bmc = 1 | (needed ? NEEDED_MASK : 0);
1581 1582 1583 1584 1585
		bitmap_count_page(bitmap, offset, 1);
		page = filemap_get_page(bitmap, offset >> CHUNK_BLOCK_SHIFT(bitmap));
		set_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
	}
	spin_unlock_irq(&bitmap->lock);
1586
	bitmap->allclean = 0;
1587 1588
}

1589 1590 1591 1592 1593 1594
/* dirty the memory and file bits for bitmap chunks "s" to "e" */
void bitmap_dirty_bits(struct bitmap *bitmap, unsigned long s, unsigned long e)
{
	unsigned long chunk;

	for (chunk = s; chunk <= e; chunk++) {
1595
		sector_t sec = (sector_t)chunk << CHUNK_BLOCK_SHIFT(bitmap);
1596 1597
		bitmap_set_memory_bits(bitmap, sec, 1);
		bitmap_file_set_bit(bitmap, sec);
1598 1599 1600 1601 1602 1603
		if (sec < bitmap->mddev->recovery_cp)
			/* We are asserting that the array is dirty,
			 * so move the recovery_cp address back so
			 * that it is obvious that it is dirty
			 */
			bitmap->mddev->recovery_cp = sec;
1604 1605 1606
	}
}

1607 1608 1609 1610 1611 1612
/*
 * flush out any pending updates
 */
void bitmap_flush(mddev_t *mddev)
{
	struct bitmap *bitmap = mddev->bitmap;
1613
	long sleep;
1614 1615 1616 1617 1618 1619 1620

	if (!bitmap) /* there was no bitmap */
		return;

	/* run the daemon_work three time to ensure everything is flushed
	 * that can be
	 */
1621
	sleep = mddev->bitmap_info.daemon_sleep * 2;
1622
	bitmap->daemon_lastrun -= sleep;
1623
	bitmap_daemon_work(mddev);
1624
	bitmap->daemon_lastrun -= sleep;
1625
	bitmap_daemon_work(mddev);
1626
	bitmap->daemon_lastrun -= sleep;
1627
	bitmap_daemon_work(mddev);
1628 1629 1630
	bitmap_update_sb(bitmap);
}

1631 1632 1633
/*
 * free memory that was allocated
 */
1634
static void bitmap_free(struct bitmap *bitmap)
1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656
{
	unsigned long k, pages;
	struct bitmap_page *bp;

	if (!bitmap) /* there was no bitmap */
		return;

	/* release the bitmap file and kill the daemon */
	bitmap_file_put(bitmap);

	bp = bitmap->bp;
	pages = bitmap->pages;

	/* free all allocated memory */

	if (bp) /* deallocate the page memory */
		for (k = 0; k < pages; k++)
			if (bp[k].map && !bp[k].hijacked)
				kfree(bp[k].map);
	kfree(bp);
	kfree(bitmap);
}
1657

1658 1659 1660 1661 1662 1663 1664
void bitmap_destroy(mddev_t *mddev)
{
	struct bitmap *bitmap = mddev->bitmap;

	if (!bitmap) /* there was no bitmap */
		return;

1665
	mutex_lock(&mddev->bitmap_info.mutex);
1666
	mddev->bitmap = NULL; /* disconnect from the md device */
1667
	mutex_unlock(&mddev->bitmap_info.mutex);
1668 1669
	if (mddev->thread)
		mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
1670

1671 1672 1673
	if (bitmap->sysfs_can_clear)
		sysfs_put(bitmap->sysfs_can_clear);

1674 1675
	bitmap_free(bitmap);
}
1676 1677 1678 1679 1680 1681 1682 1683

/*
 * initialize the bitmap structure
 * if this returns an error, bitmap_destroy must be called to do clean up
 */
int bitmap_create(mddev_t *mddev)
{
	struct bitmap *bitmap;
1684
	sector_t blocks = mddev->resync_max_sectors;
1685 1686
	unsigned long chunks;
	unsigned long pages;
1687
	struct file *file = mddev->bitmap_info.file;
1688
	int err;
1689
	struct sysfs_dirent *bm = NULL;
1690

A
Alexey Dobriyan 已提交
1691
	BUILD_BUG_ON(sizeof(bitmap_super_t) != 256);
1692

1693 1694 1695
	if (!file
	    && !mddev->bitmap_info.offset
	    && !mddev->bitmap_info.log) /* bitmap disabled, nothing to do */
1696 1697
		return 0;

1698
	BUG_ON(file && mddev->bitmap_info.offset);
1699
	BUG_ON(mddev->bitmap_info.offset && mddev->bitmap_info.log);
1700

1701
	bitmap = kzalloc(sizeof(*bitmap), GFP_KERNEL);
1702 1703 1704 1705
	if (!bitmap)
		return -ENOMEM;

	spin_lock_init(&bitmap->lock);
1706 1707
	atomic_set(&bitmap->pending_writes, 0);
	init_waitqueue_head(&bitmap->write_wait);
1708
	init_waitqueue_head(&bitmap->overflow_wait);
1709
	init_waitqueue_head(&bitmap->behind_wait);
1710

1711 1712
	bitmap->mddev = mddev;

1713 1714
	if (mddev->kobj.sd)
		bm = sysfs_get_dirent(mddev->kobj.sd, NULL, "bitmap");
1715
	if (bm) {
1716
		bitmap->sysfs_can_clear = sysfs_get_dirent(bm, NULL, "can_clear");
1717 1718 1719 1720
		sysfs_put(bm);
	} else
		bitmap->sysfs_can_clear = NULL;

1721
	bitmap->file = file;
1722 1723
	if (file) {
		get_file(file);
1724 1725 1726 1727
		/* As future accesses to this file will use bmap,
		 * and bypass the page cache, we must sync the file
		 * first.
		 */
1728
		vfs_fsync(file, 1);
1729
	}
1730
	/* read superblock from bitmap file (this sets mddev->bitmap_info.chunksize) */
1731 1732 1733 1734 1735 1736 1737 1738 1739 1740
	if (!mddev->bitmap_info.external)
		err = bitmap_read_sb(bitmap);
	else {
		err = 0;
		if (mddev->bitmap_info.chunksize == 0 ||
		    mddev->bitmap_info.daemon_sleep == 0)
			/* chunksize and time_base need to be
			 * set first. */
			err = -EINVAL;
	}
1741
	if (err)
1742
		goto error;
1743

1744
	bitmap->daemon_lastrun = jiffies;
1745
	bitmap->chunkshift = ffz(~mddev->bitmap_info.chunksize);
1746 1747

	/* now that chunksize and chunkshift are set, we can use these macros */
1748
	chunks = (blocks + CHUNK_BLOCK_RATIO(bitmap) - 1) >>
1749
			CHUNK_BLOCK_SHIFT(bitmap);
1750
	pages = (chunks + PAGE_COUNTER_RATIO - 1) / PAGE_COUNTER_RATIO;
1751 1752 1753 1754 1755 1756 1757 1758 1759 1760

	BUG_ON(!pages);

	bitmap->chunks = chunks;
	bitmap->pages = pages;
	bitmap->missing_pages = pages;
	bitmap->counter_bits = COUNTER_BITS;

	bitmap->syncchunk = ~0UL;

1761
#ifdef INJECT_FATAL_FAULT_1
1762 1763
	bitmap->bp = NULL;
#else
1764
	bitmap->bp = kzalloc(pages * sizeof(*bitmap->bp), GFP_KERNEL);
1765
#endif
1766
	err = -ENOMEM;
1767
	if (!bitmap->bp)
1768
		goto error;
1769

1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797
	printk(KERN_INFO "created bitmap (%lu pages) for device %s\n",
		pages, bmname(bitmap));

	mddev->bitmap = bitmap;


	return (bitmap->flags & BITMAP_WRITE_ERROR) ? -EIO : 0;

 error:
	bitmap_free(bitmap);
	return err;
}

int bitmap_load(mddev_t *mddev)
{
	int err = 0;
	sector_t sector = 0;
	struct bitmap *bitmap = mddev->bitmap;

	if (!bitmap)
		goto out;

	/* Clear out old bitmap info first:  Either there is none, or we
	 * are resuming after someone else has possibly changed things,
	 * so we should forget old cached info.
	 * All chunks should be clean, but some might need_sync.
	 */
	while (sector < mddev->resync_max_sectors) {
N
NeilBrown 已提交
1798
		sector_t blocks;
1799 1800 1801 1802 1803
		bitmap_start_sync(bitmap, sector, &blocks, 0);
		sector += blocks;
	}
	bitmap_close_sync(bitmap);

1804 1805 1806 1807 1808 1809 1810 1811
	if (mddev->bitmap_info.log) {
		unsigned long i;
		struct dm_dirty_log *log = mddev->bitmap_info.log;
		for (i = 0; i < bitmap->chunks; i++)
			if (!log->type->in_sync(log, i, 1))
				bitmap_set_memory_bits(bitmap,
						       (sector_t)i << CHUNK_BLOCK_SHIFT(bitmap),
						       1);
1812 1813 1814 1815 1816 1817 1818
	} else {
		sector_t start = 0;
		if (mddev->degraded == 0
		    || bitmap->events_cleared == mddev->events)
			/* no need to keep dirty bits to optimise a
			 * re-add of a missing device */
			start = mddev->recovery_cp;
1819

1820 1821
		err = bitmap_init_from_disk(bitmap, start);
	}
1822
	if (err)
1823
		goto out;
1824

1825
	mddev->thread->timeout = mddev->bitmap_info.daemon_sleep;
1826
	md_wakeup_thread(mddev->thread);
1827

1828 1829
	bitmap_update_sb(bitmap);

1830 1831 1832
	if (bitmap->flags & BITMAP_WRITE_ERROR)
		err = -EIO;
out:
1833
	return err;
1834
}
1835
EXPORT_SYMBOL_GPL(bitmap_load);
1836

1837 1838 1839 1840
static ssize_t
location_show(mddev_t *mddev, char *page)
{
	ssize_t len;
1841
	if (mddev->bitmap_info.file)
1842
		len = sprintf(page, "file");
1843
	else if (mddev->bitmap_info.offset)
1844
		len = sprintf(page, "%+lld", (long long)mddev->bitmap_info.offset);
1845
	else
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 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896
		len = sprintf(page, "none");
	len += sprintf(page+len, "\n");
	return len;
}

static ssize_t
location_store(mddev_t *mddev, const char *buf, size_t len)
{

	if (mddev->pers) {
		if (!mddev->pers->quiesce)
			return -EBUSY;
		if (mddev->recovery || mddev->sync_thread)
			return -EBUSY;
	}

	if (mddev->bitmap || mddev->bitmap_info.file ||
	    mddev->bitmap_info.offset) {
		/* bitmap already configured.  Only option is to clear it */
		if (strncmp(buf, "none", 4) != 0)
			return -EBUSY;
		if (mddev->pers) {
			mddev->pers->quiesce(mddev, 1);
			bitmap_destroy(mddev);
			mddev->pers->quiesce(mddev, 0);
		}
		mddev->bitmap_info.offset = 0;
		if (mddev->bitmap_info.file) {
			struct file *f = mddev->bitmap_info.file;
			mddev->bitmap_info.file = NULL;
			restore_bitmap_write_access(f);
			fput(f);
		}
	} else {
		/* No bitmap, OK to set a location */
		long long offset;
		if (strncmp(buf, "none", 4) == 0)
			/* nothing to be done */;
		else if (strncmp(buf, "file:", 5) == 0) {
			/* Not supported yet */
			return -EINVAL;
		} else {
			int rv;
			if (buf[0] == '+')
				rv = strict_strtoll(buf+1, 10, &offset);
			else
				rv = strict_strtoll(buf, 10, &offset);
			if (rv)
				return rv;
			if (offset == 0)
				return -EINVAL;
1897 1898
			if (mddev->bitmap_info.external == 0 &&
			    mddev->major_version == 0 &&
1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933
			    offset != mddev->bitmap_info.default_offset)
				return -EINVAL;
			mddev->bitmap_info.offset = offset;
			if (mddev->pers) {
				mddev->pers->quiesce(mddev, 1);
				rv = bitmap_create(mddev);
				if (rv) {
					bitmap_destroy(mddev);
					mddev->bitmap_info.offset = 0;
				}
				mddev->pers->quiesce(mddev, 0);
				if (rv)
					return rv;
			}
		}
	}
	if (!mddev->external) {
		/* Ensure new bitmap info is stored in
		 * metadata promptly.
		 */
		set_bit(MD_CHANGE_DEVS, &mddev->flags);
		md_wakeup_thread(mddev->thread);
	}
	return len;
}

static struct md_sysfs_entry bitmap_location =
__ATTR(location, S_IRUGO|S_IWUSR, location_show, location_store);

static ssize_t
timeout_show(mddev_t *mddev, char *page)
{
	ssize_t len;
	unsigned long secs = mddev->bitmap_info.daemon_sleep / HZ;
	unsigned long jifs = mddev->bitmap_info.daemon_sleep % HZ;
1934

1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026
	len = sprintf(page, "%lu", secs);
	if (jifs)
		len += sprintf(page+len, ".%03u", jiffies_to_msecs(jifs));
	len += sprintf(page+len, "\n");
	return len;
}

static ssize_t
timeout_store(mddev_t *mddev, const char *buf, size_t len)
{
	/* timeout can be set at any time */
	unsigned long timeout;
	int rv = strict_strtoul_scaled(buf, &timeout, 4);
	if (rv)
		return rv;

	/* just to make sure we don't overflow... */
	if (timeout >= LONG_MAX / HZ)
		return -EINVAL;

	timeout = timeout * HZ / 10000;

	if (timeout >= MAX_SCHEDULE_TIMEOUT)
		timeout = MAX_SCHEDULE_TIMEOUT-1;
	if (timeout < 1)
		timeout = 1;
	mddev->bitmap_info.daemon_sleep = timeout;
	if (mddev->thread) {
		/* if thread->timeout is MAX_SCHEDULE_TIMEOUT, then
		 * the bitmap is all clean and we don't need to
		 * adjust the timeout right now
		 */
		if (mddev->thread->timeout < MAX_SCHEDULE_TIMEOUT) {
			mddev->thread->timeout = timeout;
			md_wakeup_thread(mddev->thread);
		}
	}
	return len;
}

static struct md_sysfs_entry bitmap_timeout =
__ATTR(time_base, S_IRUGO|S_IWUSR, timeout_show, timeout_store);

static ssize_t
backlog_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%lu\n", mddev->bitmap_info.max_write_behind);
}

static ssize_t
backlog_store(mddev_t *mddev, const char *buf, size_t len)
{
	unsigned long backlog;
	int rv = strict_strtoul(buf, 10, &backlog);
	if (rv)
		return rv;
	if (backlog > COUNTER_MAX)
		return -EINVAL;
	mddev->bitmap_info.max_write_behind = backlog;
	return len;
}

static struct md_sysfs_entry bitmap_backlog =
__ATTR(backlog, S_IRUGO|S_IWUSR, backlog_show, backlog_store);

static ssize_t
chunksize_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%lu\n", mddev->bitmap_info.chunksize);
}

static ssize_t
chunksize_store(mddev_t *mddev, const char *buf, size_t len)
{
	/* Can only be changed when no bitmap is active */
	int rv;
	unsigned long csize;
	if (mddev->bitmap)
		return -EBUSY;
	rv = strict_strtoul(buf, 10, &csize);
	if (rv)
		return rv;
	if (csize < 512 ||
	    !is_power_of_2(csize))
		return -EINVAL;
	mddev->bitmap_info.chunksize = csize;
	return len;
}

static struct md_sysfs_entry bitmap_chunksize =
__ATTR(chunksize, S_IRUGO|S_IWUSR, chunksize_show, chunksize_store);

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 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079
static ssize_t metadata_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%s\n", (mddev->bitmap_info.external
				      ? "external" : "internal"));
}

static ssize_t metadata_store(mddev_t *mddev, const char *buf, size_t len)
{
	if (mddev->bitmap ||
	    mddev->bitmap_info.file ||
	    mddev->bitmap_info.offset)
		return -EBUSY;
	if (strncmp(buf, "external", 8) == 0)
		mddev->bitmap_info.external = 1;
	else if (strncmp(buf, "internal", 8) == 0)
		mddev->bitmap_info.external = 0;
	else
		return -EINVAL;
	return len;
}

static struct md_sysfs_entry bitmap_metadata =
__ATTR(metadata, S_IRUGO|S_IWUSR, metadata_show, metadata_store);

static ssize_t can_clear_show(mddev_t *mddev, char *page)
{
	int len;
	if (mddev->bitmap)
		len = sprintf(page, "%s\n", (mddev->bitmap->need_sync ?
					     "false" : "true"));
	else
		len = sprintf(page, "\n");
	return len;
}

static ssize_t can_clear_store(mddev_t *mddev, const char *buf, size_t len)
{
	if (mddev->bitmap == NULL)
		return -ENOENT;
	if (strncmp(buf, "false", 5) == 0)
		mddev->bitmap->need_sync = 1;
	else if (strncmp(buf, "true", 4) == 0) {
		if (mddev->degraded)
			return -EBUSY;
		mddev->bitmap->need_sync = 0;
	} else
		return -EINVAL;
	return len;
}

static struct md_sysfs_entry bitmap_can_clear =
__ATTR(can_clear, S_IRUGO|S_IWUSR, can_clear_show, can_clear_store);

2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100
static ssize_t
behind_writes_used_show(mddev_t *mddev, char *page)
{
	if (mddev->bitmap == NULL)
		return sprintf(page, "0\n");
	return sprintf(page, "%lu\n",
		       mddev->bitmap->behind_writes_used);
}

static ssize_t
behind_writes_used_reset(mddev_t *mddev, const char *buf, size_t len)
{
	if (mddev->bitmap)
		mddev->bitmap->behind_writes_used = 0;
	return len;
}

static struct md_sysfs_entry max_backlog_used =
__ATTR(max_backlog_used, S_IRUGO | S_IWUSR,
       behind_writes_used_show, behind_writes_used_reset);

2101 2102 2103 2104 2105
static struct attribute *md_bitmap_attrs[] = {
	&bitmap_location.attr,
	&bitmap_timeout.attr,
	&bitmap_backlog.attr,
	&bitmap_chunksize.attr,
2106 2107
	&bitmap_metadata.attr,
	&bitmap_can_clear.attr,
2108
	&max_backlog_used.attr,
2109 2110 2111 2112 2113 2114 2115
	NULL
};
struct attribute_group md_bitmap_group = {
	.name = "bitmap",
	.attrs = md_bitmap_attrs,
};