bitmap.c 41.9 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 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 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76
/*
 * 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).
 * wait if count gets too high, wake when it drops to half.
 */

#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>
#include <linux/raid/md.h>
#include <linux/raid/bitmap.h>

/* 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

//#define DPRINTK PRINTK /* set this NULL to avoid verbose debug output */
#define DPRINTK(x...) do { } while(0)

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

static inline char * bmname(struct bitmap *bitmap)
{
	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;

77
#ifdef INJECT_FAULTS_1
78 79 80 81 82 83 84
	page = NULL;
#else
	page = kmalloc(PAGE_SIZE, GFP_NOIO);
#endif
	if (!page)
		printk("%s: bitmap_alloc_page FAILED\n", bmname(bitmap));
	else
85
		PRINTK("%s: bitmap_alloc_page: allocated page at %p\n",
86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169
			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
 */
static int bitmap_checkpage(struct bitmap *bitmap, unsigned long page, int create)
{
	unsigned char *mappage;

	if (page >= bitmap->pages) {
		printk(KERN_ALERT
			"%s: invalid bitmap page request: %lu (> %lu)\n",
			bmname(bitmap), page, bitmap->pages-1);
		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;

	spin_unlock_irq(&bitmap->lock);

	/* this page has not been allocated yet */

	if ((mappage = bitmap_alloc_page(bitmap)) == NULL) {
		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 */
		spin_lock_irq(&bitmap->lock);
		if (!bitmap->bp[page].map)
			bitmap->bp[page].hijacked = 1;
		goto out;
	}

	/* got a page */

	spin_lock_irq(&bitmap->lock);

	/* recheck the page */

	if (bitmap->bp[page].map || bitmap->bp[page].hijacked) {
		/* somebody beat us to getting the page */
		bitmap_free_page(bitmap, mappage);
		return 0;
	}

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

	memset(mappage, 0, PAGE_SIZE);
	bitmap->bp[page].map = mappage;
	bitmap->missing_pages--;
out:
	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 */
170
static void bitmap_checkfree(struct bitmap *bitmap, unsigned long page)
171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209
{
	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;
		return;
	}

	/* normal case, free the page */

#if 0
/* actually ... let's not.  We will probably need the page again exactly when
 * memory is tight and we are flusing to disk
 */
	return;
#else
	ptr = bitmap->bp[page].map;
	bitmap->bp[page].map = NULL;
	bitmap->missing_pages++;
	bitmap_free_page(bitmap, ptr);
	return;
#endif
}


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

/*
 * basic page I/O operations
 */

210 211 212 213 214 215 216 217 218 219 220 221 222
/* IO operations when bitmap is stored near all superblocks */
static struct page *read_sb_page(mddev_t *mddev, long offset, unsigned long index)
{
	/* choose a good rdev and read the page from there */

	mdk_rdev_t *rdev;
	struct list_head *tmp;
	struct page *page = alloc_page(GFP_KERNEL);
	sector_t target;

	if (!page)
		return ERR_PTR(-ENOMEM);

223
	rdev_for_each(rdev, tmp, mddev) {
224 225
		if (! test_bit(In_sync, &rdev->flags)
		    || test_bit(Faulty, &rdev->flags))
226 227
			continue;

228
		target = rdev->sb_start + offset + index * (PAGE_SIZE/512);
229

230 231
		if (sync_page_io(rdev->bdev, target, PAGE_SIZE, page, READ)) {
			page->index = index;
232 233
			attach_page_buffers(page, NULL); /* so that free_buffer will
							  * quietly no-op */
234 235 236 237
			return page;
		}
	}
	return ERR_PTR(-EIO);
238 239 240

}

241
static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait)
242 243
{
	mdk_rdev_t *rdev;
244
	mddev_t *mddev = bitmap->mddev;
245

246 247
	rcu_read_lock();
	rdev_for_each_rcu(rdev, mddev)
248
		if (test_bit(In_sync, &rdev->flags)
249 250 251 252 253
		    && !test_bit(Faulty, &rdev->flags)) {
			int size = PAGE_SIZE;
			if (page->index == bitmap->file_pages-1)
				size = roundup(bitmap->last_page_size,
					       bdev_hardsect_size(rdev->bdev));
254 255 256 257 258 259
			/* Just make sure we aren't corrupting data or
			 * metadata
			 */
			if (bitmap->offset < 0) {
				/* DATA  BITMAP METADATA  */
				if (bitmap->offset
260
				    + (long)(page->index * (PAGE_SIZE/512))
261 262
				    + size/512 > 0)
					/* bitmap runs in to metadata */
263
					goto bad_alignment;
264
				if (rdev->data_offset + mddev->size*2
265
				    > rdev->sb_start + bitmap->offset)
266
					/* data runs in to bitmap */
267
					goto bad_alignment;
268
			} else if (rdev->sb_start < rdev->data_offset) {
269
				/* METADATA BITMAP DATA */
270
				if (rdev->sb_start
271 272 273 274
				    + bitmap->offset
				    + page->index*(PAGE_SIZE/512) + size/512
				    > rdev->data_offset)
					/* bitmap runs in to data */
275
					goto bad_alignment;
276 277 278
			} else {
				/* DATA METADATA BITMAP - no problems */
			}
279
			md_super_write(mddev, rdev,
280
				       rdev->sb_start + bitmap->offset
281
				       + page->index * (PAGE_SIZE/512),
282
				       size,
283
				       page);
284
		}
285
	rcu_read_unlock();
286 287

	if (wait)
288
		md_super_wait(mddev);
289
	return 0;
290 291 292 293

 bad_alignment:
	rcu_read_unlock();
	return -EINVAL;
294 295
}

296
static void bitmap_file_kick(struct bitmap *bitmap);
297
/*
298
 * write out a page to a file
299
 */
300
static void write_page(struct bitmap *bitmap, struct page *page, int wait)
301
{
302
	struct buffer_head *bh;
303

304 305 306 307 308
	if (bitmap->file == NULL) {
		switch (write_sb_page(bitmap, page, wait)) {
		case -EINVAL:
			bitmap->flags |= BITMAP_WRITE_ERROR;
		}
309
	} else {
310

311
		bh = page_buffers(page);
312

313 314 315 316 317 318 319
		while (bh && bh->b_blocknr) {
			atomic_inc(&bitmap->pending_writes);
			set_buffer_locked(bh);
			set_buffer_mapped(bh);
			submit_bh(WRITE, bh);
			bh = bh->b_this_page;
		}
320

321 322 323 324
		if (wait) {
			wait_event(bitmap->write_wait,
				   atomic_read(&bitmap->pending_writes)==0);
		}
325
	}
326 327
	if (bitmap->flags & BITMAP_WRITE_ERROR)
		bitmap_file_kick(bitmap);
328 329 330 331 332 333
}

static void end_bitmap_write(struct buffer_head *bh, int uptodate)
{
	struct bitmap *bitmap = bh->b_private;
	unsigned long flags;
334

335 336 337 338
	if (!uptodate) {
		spin_lock_irqsave(&bitmap->lock, flags);
		bitmap->flags |= BITMAP_WRITE_ERROR;
		spin_unlock_irqrestore(&bitmap->lock, flags);
339
	}
340 341 342
	if (atomic_dec_and_test(&bitmap->pending_writes))
		wake_up(&bitmap->write_wait);
}
343

344 345 346 347 348 349 350 351 352 353 354
/* 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);
355

356 357 358 359
	while (bh) {
		struct buffer_head *next = bh->b_this_page;
		free_buffer_head(bh);
		bh = next;
360
	}
361 362
	__clear_page_buffers(page);
	put_page(page);
363 364
}

365 366 367 368 369 370 371
/* 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.
 */
372
static struct page *read_page(struct file *file, unsigned long index,
373 374
			      struct bitmap *bitmap,
			      unsigned long count)
375 376
{
	struct page *page = NULL;
J
Josef Sipek 已提交
377
	struct inode *inode = file->f_path.dentry->d_inode;
378 379
	struct buffer_head *bh;
	sector_t block;
380

381 382
	PRINTK("read bitmap file (%dB @ %Lu)\n", (int)PAGE_SIZE,
			(unsigned long long)index << PAGE_SHIFT);
383

384 385 386
	page = alloc_page(GFP_KERNEL);
	if (!page)
		page = ERR_PTR(-ENOMEM);
387 388
	if (IS_ERR(page))
		goto out;
389 390 391

	bh = alloc_page_buffers(page, 1<<inode->i_blkbits, 0);
	if (!bh) {
392
		put_page(page);
393
		page = ERR_PTR(-ENOMEM);
394 395
		goto out;
	}
396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416
	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;
417 418 419 420
			atomic_inc(&bitmap->pending_writes);
			set_buffer_locked(bh);
			set_buffer_mapped(bh);
			submit_bh(READ, bh);
421 422 423 424 425
		}
		block++;
		bh = bh->b_this_page;
	}
	page->index = index;
426 427 428 429 430 431 432

	wait_event(bitmap->write_wait,
		   atomic_read(&bitmap->pending_writes)==0);
	if (bitmap->flags & BITMAP_WRITE_ERROR) {
		free_buffers(page);
		page = ERR_PTR(-EIO);
	}
433 434 435
out:
	if (IS_ERR(page))
		printk(KERN_ALERT "md: bitmap read error: (%dB @ %Lu): %ld\n",
436 437
			(int)PAGE_SIZE,
			(unsigned long long)index << PAGE_SHIFT,
438 439 440 441 442 443 444 445 446
			PTR_ERR(page));
	return page;
}

/*
 * bitmap file superblock operations
 */

/* update the event counter and sync the superblock to disk */
447
void bitmap_update_sb(struct bitmap *bitmap)
448 449 450 451 452
{
	bitmap_super_t *sb;
	unsigned long flags;

	if (!bitmap || !bitmap->mddev) /* no bitmap for this array */
453
		return;
454 455 456
	spin_lock_irqsave(&bitmap->lock, flags);
	if (!bitmap->sb_page) { /* no superblock */
		spin_unlock_irqrestore(&bitmap->lock, flags);
457
		return;
458 459
	}
	spin_unlock_irqrestore(&bitmap->lock, flags);
460
	sb = (bitmap_super_t *)kmap_atomic(bitmap->sb_page, KM_USER0);
461
	sb->events = cpu_to_le64(bitmap->mddev->events);
462 463 464 465 466
	if (bitmap->mddev->events < bitmap->events_cleared) {
		/* rocking back to read-only */
		bitmap->events_cleared = bitmap->mddev->events;
		sb->events_cleared = cpu_to_le64(bitmap->events_cleared);
	}
467
	kunmap_atomic(sb, KM_USER0);
468
	write_page(bitmap, bitmap->sb_page, 1);
469 470 471 472 473 474 475 476 477
}

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

	if (!bitmap || !bitmap->sb_page)
		return;
478
	sb = (bitmap_super_t *)kmap_atomic(bitmap->sb_page, KM_USER0);
479
	printk(KERN_DEBUG "%s: bitmap file superblock:\n", bmname(bitmap));
480 481 482
	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",
483 484 485 486
					*(__u32 *)(sb->uuid+0),
					*(__u32 *)(sb->uuid+4),
					*(__u32 *)(sb->uuid+8),
					*(__u32 *)(sb->uuid+12));
487
	printk(KERN_DEBUG "        events: %llu\n",
488
			(unsigned long long) le64_to_cpu(sb->events));
489
	printk(KERN_DEBUG "events cleared: %llu\n",
490
			(unsigned long long) le64_to_cpu(sb->events_cleared));
491 492 493 494 495
	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);
496
	printk(KERN_DEBUG "max write behind: %d\n", le32_to_cpu(sb->write_behind));
497
	kunmap_atomic(sb, KM_USER0);
498 499 500 501 502 503 504
}

/* 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;
505
	unsigned long chunksize, daemon_sleep, write_behind;
506 507 508 509
	unsigned long long events;
	int err = -EINVAL;

	/* page 0 is the superblock, read it... */
510 511 512 513 514 515
	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 {
516 517
		bitmap->sb_page = read_sb_page(bitmap->mddev, bitmap->offset, 0);
	}
518 519 520 521 522 523
	if (IS_ERR(bitmap->sb_page)) {
		err = PTR_ERR(bitmap->sb_page);
		bitmap->sb_page = NULL;
		return err;
	}

524
	sb = (bitmap_super_t *)kmap_atomic(bitmap->sb_page, KM_USER0);
525 526 527

	chunksize = le32_to_cpu(sb->chunksize);
	daemon_sleep = le32_to_cpu(sb->daemon_sleep);
528
	write_behind = le32_to_cpu(sb->write_behind);
529 530 531 532

	/* verify that the bitmap-specific fields are valid */
	if (sb->magic != cpu_to_le32(BITMAP_MAGIC))
		reason = "bad magic";
533 534
	else if (le32_to_cpu(sb->version) < BITMAP_MAJOR_LO ||
		 le32_to_cpu(sb->version) > BITMAP_MAJOR_HI)
535
		reason = "unrecognized superblock version";
536 537
	else if (chunksize < PAGE_SIZE)
		reason = "bitmap chunksize too small";
538 539
	else if ((1 << ffz(~chunksize)) != chunksize)
		reason = "bitmap chunksize not a power of 2";
540 541
	else if (daemon_sleep < 1 || daemon_sleep > MAX_SCHEDULE_TIMEOUT / HZ)
		reason = "daemon sleep period out of range";
542 543
	else if (write_behind > COUNTER_MAX)
		reason = "write-behind limit out of range (0 - 16383)";
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
	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);
570
		sb->state |= cpu_to_le32(BITMAP_STALE);
571 572 573 574 575
	}
success:
	/* assign fields using values from superblock */
	bitmap->chunksize = chunksize;
	bitmap->daemon_sleep = daemon_sleep;
576
	bitmap->daemon_lastrun = jiffies;
577
	bitmap->max_write_behind = write_behind;
578
	bitmap->flags |= le32_to_cpu(sb->state);
579 580
	if (le32_to_cpu(sb->version) == BITMAP_MAJOR_HOSTENDIAN)
		bitmap->flags |= BITMAP_HOSTENDIAN;
581
	bitmap->events_cleared = le64_to_cpu(sb->events_cleared);
582
	if (sb->state & cpu_to_le32(BITMAP_STALE))
583
		bitmap->events_cleared = bitmap->mddev->events;
584 585
	err = 0;
out:
586
	kunmap_atomic(sb, KM_USER0);
587 588 589 590 591 592 593 594 595 596
	if (err)
		bitmap_print_sb(bitmap);
	return err;
}

enum bitmap_mask_op {
	MASK_SET,
	MASK_UNSET
};

597 598 599
/* 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)
600 601 602
{
	bitmap_super_t *sb;
	unsigned long flags;
603
	int old;
604 605

	spin_lock_irqsave(&bitmap->lock, flags);
606
	if (!bitmap->sb_page) { /* can't set the state */
607
		spin_unlock_irqrestore(&bitmap->lock, flags);
608
		return 0;
609 610
	}
	spin_unlock_irqrestore(&bitmap->lock, flags);
611
	sb = (bitmap_super_t *)kmap_atomic(bitmap->sb_page, KM_USER0);
612
	old = le32_to_cpu(sb->state) & bits;
613
	switch (op) {
614
		case MASK_SET: sb->state |= cpu_to_le32(bits);
615
				break;
616
		case MASK_UNSET: sb->state &= cpu_to_le32(~bits);
617 618 619
				break;
		default: BUG();
	}
620
	kunmap_atomic(sb, KM_USER0);
621
	return old;
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
}

/*
 * general bitmap file operations
 */

/* calculate the index of the page that contains this bit */
static inline unsigned long file_page_index(unsigned long chunk)
{
	return CHUNK_BIT_OFFSET(chunk) >> PAGE_BIT_SHIFT;
}

/* calculate the (bit) offset of this bit within a page */
static inline unsigned long file_page_offset(unsigned long chunk)
{
	return CHUNK_BIT_OFFSET(chunk) & (PAGE_BITS - 1);
}

/*
 * 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)
{
650
	if (file_page_index(chunk) >= bitmap->file_pages) return NULL;
651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674
	return bitmap->filemap[file_page_index(chunk) - file_page_index(0)];
}


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--)
		if (map[pages]->index != 0) /* 0 is sb_page, release it below */
675
			free_buffers(map[pages]);
676 677 678
	kfree(map);
	kfree(attr);

679 680
	if (sb_page)
		free_buffers(sb_page);
681 682 683 684 685 686 687 688 689 690 691 692
}

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);

693 694 695
	if (file)
		wait_event(bitmap->write_wait,
			   atomic_read(&bitmap->pending_writes)==0);
696 697
	bitmap_file_unmap(bitmap);

698
	if (file) {
J
Josef Sipek 已提交
699
		struct inode *inode = file->f_path.dentry->d_inode;
700
		invalidate_mapping_pages(inode->i_mapping, 0, -1);
701
		fput(file);
702
	}
703 704 705 706 707 708 709 710 711 712 713 714
}


/*
 * 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;

715 716
	if (bitmap_mask_state(bitmap, BITMAP_STALE, MASK_SET) == 0) {
		bitmap_update_sb(bitmap);
717

718 719 720
		if (bitmap->file) {
			path = kmalloc(PAGE_SIZE, GFP_KERNEL);
			if (path)
C
Christoph Hellwig 已提交
721 722 723
				ptr = d_path(&bitmap->file->f_path, path,
					     PAGE_SIZE);

724

725 726
			printk(KERN_ALERT
			      "%s: kicking failed bitmap file %s from array!\n",
C
Christoph Hellwig 已提交
727
			      bmname(bitmap), IS_ERR(ptr) ? "" : ptr);
728

729 730 731 732 733
			kfree(path);
		} else
			printk(KERN_ALERT
			       "%s: disabling internal bitmap due to errors\n",
			       bmname(bitmap));
734
	}
735 736 737 738 739 740 741

	bitmap_file_put(bitmap);

	return;
}

enum bitmap_page_attr {
742 743 744
	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
745 746 747 748 749
};

static inline void set_page_attr(struct bitmap *bitmap, struct page *page,
				enum bitmap_page_attr attr)
{
750
	__set_bit((page->index<<2) + attr, bitmap->filemap_attr);
751 752 753 754 755
}

static inline void clear_page_attr(struct bitmap *bitmap, struct page *page,
				enum bitmap_page_attr attr)
{
756
	__clear_bit((page->index<<2) + attr, bitmap->filemap_attr);
757 758
}

759 760
static inline unsigned long test_page_attr(struct bitmap *bitmap, struct page *page,
					   enum bitmap_page_attr attr)
761
{
762
	return test_bit((page->index<<2) + attr, bitmap->filemap_attr);
763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778
}

/*
 * 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;
	struct page *page;
	void *kaddr;
	unsigned long chunk = block >> CHUNK_BLOCK_SHIFT(bitmap);

779
	if (!bitmap->filemap) {
780 781 782 783
		return;
	}

	page = filemap_get_page(bitmap, chunk);
784
	if (!page) return;
785 786 787 788
	bit = file_page_offset(chunk);

 	/* set the bit */
	kaddr = kmap_atomic(page, KM_USER0);
789 790 791 792
	if (bitmap->flags & BITMAP_HOSTENDIAN)
		set_bit(bit, kaddr);
	else
		ext2_set_bit(bit, kaddr);
793 794 795 796 797 798 799 800 801 802 803
	kunmap_atomic(kaddr, KM_USER0);
	PRINTK("set file bit %lu page %lu\n", bit, page->index);

	/* 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 */
804
void bitmap_unplug(struct bitmap *bitmap)
805
{
806 807
	unsigned long i, flags;
	int dirty, need_write;
808 809 810 811
	struct page *page;
	int wait = 0;

	if (!bitmap)
812
		return;
813 814 815 816 817

	/* 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);
818
		if (!bitmap->filemap) {
819
			spin_unlock_irqrestore(&bitmap->lock, flags);
820
			return;
821 822
		}
		page = bitmap->filemap[i];
823 824
		dirty = test_page_attr(bitmap, page, BITMAP_PAGE_DIRTY);
		need_write = test_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
825 826
		clear_page_attr(bitmap, page, BITMAP_PAGE_DIRTY);
		clear_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
827
		if (dirty)
828 829 830
			wait = 1;
		spin_unlock_irqrestore(&bitmap->lock, flags);

831
		if (dirty | need_write)
832
			write_page(bitmap, page, 0);
833 834
	}
	if (wait) { /* if any writes were performed, we need to wait on them */
835
		if (bitmap->file)
836 837
			wait_event(bitmap->write_wait,
				   atomic_read(&bitmap->pending_writes)==0);
838
		else
839
			md_super_wait(bitmap->mddev);
840
	}
841 842
	if (bitmap->flags & BITMAP_WRITE_ERROR)
		bitmap_file_kick(bitmap);
843 844
}

845
static void bitmap_set_memory_bits(struct bitmap *bitmap, sector_t offset, int needed);
846 847 848 849 850 851 852
/* * 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.
853 854 855
 *
 * We ignore all bits for sectors that end earlier than 'start'.
 * This is used when reading an out-of-date bitmap...
856
 */
857
static int bitmap_init_from_disk(struct bitmap *bitmap, sector_t start)
858 859 860 861 862
{
	unsigned long i, chunks, index, oldindex, bit;
	struct page *page = NULL, *oldpage = NULL;
	unsigned long num_pages, bit_cnt = 0;
	struct file *file;
863
	unsigned long bytes, offset;
864 865
	int outofdate;
	int ret = -ENOSPC;
866
	void *paddr;
867 868 869 870

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

871
	BUG_ON(!file && !bitmap->offset);
872

873
#ifdef INJECT_FAULTS_3
874 875 876 877 878 879 880 881 882
	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));

	bytes = (chunks + 7) / 8;
883

884
	num_pages = (bytes + sizeof(bitmap_super_t) + PAGE_SIZE - 1) / PAGE_SIZE;
885

886
	if (file && i_size_read(file->f_mapping->host) < bytes + sizeof(bitmap_super_t)) {
887 888 889 890
		printk(KERN_INFO "%s: bitmap file too short %lu < %lu\n",
			bmname(bitmap),
			(unsigned long) i_size_read(file->f_mapping->host),
			bytes + sizeof(bitmap_super_t));
891
		goto err;
892
	}
893 894 895

	ret = -ENOMEM;

896
	bitmap->filemap = kmalloc(sizeof(struct page *) * num_pages, GFP_KERNEL);
897
	if (!bitmap->filemap)
898
		goto err;
899

900 901
	/* We need 4 bits per page, rounded up to a multiple of sizeof(unsigned long) */
	bitmap->filemap_attr = kzalloc(
902
		roundup( DIV_ROUND_UP(num_pages*4, 8), sizeof(unsigned long)),
903
		GFP_KERNEL);
904
	if (!bitmap->filemap_attr)
905
		goto err;
906 907 908 909

	oldindex = ~0L;

	for (i = 0; i < chunks; i++) {
910
		int b;
911 912 913
		index = file_page_index(i);
		bit = file_page_offset(i);
		if (index != oldindex) { /* this is a new page, read it in */
914
			int count;
915
			/* unmap the old page, we're done with it */
916
			if (index == num_pages-1)
917 918
				count = bytes + sizeof(bitmap_super_t)
					- index * PAGE_SIZE;
919 920
			else
				count = PAGE_SIZE;
921 922 923 924 925 926 927 928
			if (index == 0) {
				/*
				 * 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);
929
			} else if (file) {
930
				page = read_page(file, index, bitmap, count);
931 932 933
				offset = 0;
			} else {
				page = read_sb_page(bitmap->mddev, bitmap->offset, index);
934 935
				offset = 0;
			}
936 937
			if (IS_ERR(page)) { /* read error */
				ret = PTR_ERR(page);
938
				goto err;
939 940
			}

941 942 943 944 945 946 947 948
			oldindex = index;
			oldpage = page;

			if (outofdate) {
				/*
				 * if bitmap is out of date, dirty the
			 	 * whole page and write it out
				 */
949 950
				paddr = kmap_atomic(page, KM_USER0);
				memset(paddr + offset, 0xff,
951
				       PAGE_SIZE - offset);
952
				kunmap_atomic(paddr, KM_USER0);
953 954 955 956
				write_page(bitmap, page, 1);

				ret = -EIO;
				if (bitmap->flags & BITMAP_WRITE_ERROR) {
957
					/* release, page not in filemap yet */
958
					put_page(page);
959
					goto err;
960 961 962 963
				}
			}

			bitmap->filemap[bitmap->file_pages++] = page;
964
			bitmap->last_page_size = count;
965
		}
966
		paddr = kmap_atomic(page, KM_USER0);
967
		if (bitmap->flags & BITMAP_HOSTENDIAN)
968
			b = test_bit(bit, paddr);
969
		else
970 971
			b = ext2_test_bit(bit, paddr);
		kunmap_atomic(paddr, KM_USER0);
972
		if (b) {
973
			/* if the disk bit is set, set the memory bit */
974 975 976
			bitmap_set_memory_bits(bitmap, i << CHUNK_BLOCK_SHIFT(bitmap),
					       ((i+1) << (CHUNK_BLOCK_SHIFT(bitmap)) >= start)
				);
977
			bit_cnt++;
978
			set_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
979 980 981 982 983 984 985 986 987 988 989 990 991
		}
	}

 	/* everything went OK */
	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: "
992 993 994 995
		"read %lu/%lu pages, set %lu bits\n",
		bmname(bitmap), bitmap->file_pages, num_pages, bit_cnt);

	return 0;
996

997 998 999
 err:
	printk(KERN_INFO "%s: bitmap initialisation failed: %d\n",
	       bmname(bitmap), ret);
1000 1001 1002
	return ret;
}

1003 1004 1005 1006 1007
void bitmap_write_all(struct bitmap *bitmap)
{
	/* We don't actually write all bitmap blocks here,
	 * just flag them as needing to be written
	 */
1008
	int i;
1009

1010 1011 1012
	for (i=0; i < bitmap->file_pages; i++)
		set_page_attr(bitmap, bitmap->filemap[i],
			      BITMAP_PAGE_NEEDWRITE);
1013 1014
}

1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035

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;
/*
	if (page == 0) printk("count page 0, offset %llu: %d gives %d\n",
			      (unsigned long long)offset, inc, bitmap->bp[page].count);
*/
	bitmap_checkfree(bitmap, page);
}
static bitmap_counter_t *bitmap_get_counter(struct bitmap *bitmap,
					    sector_t offset, int *blocks,
					    int create);

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

1036
void bitmap_daemon_work(struct bitmap *bitmap)
1037
{
1038
	unsigned long j;
1039 1040 1041
	unsigned long flags;
	struct page *page = NULL, *lastpage = NULL;
	int blocks;
1042
	void *paddr;
1043 1044

	if (bitmap == NULL)
1045
		return;
1046
	if (time_before(jiffies, bitmap->daemon_lastrun + bitmap->daemon_sleep*HZ))
1047 1048
		goto done;

1049
	bitmap->daemon_lastrun = jiffies;
1050 1051 1052 1053 1054
	if (bitmap->allclean) {
		bitmap->mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
		return;
	}
	bitmap->allclean = 1;
1055 1056 1057 1058

	for (j = 0; j < bitmap->chunks; j++) {
		bitmap_counter_t *bmc;
		spin_lock_irqsave(&bitmap->lock, flags);
1059
		if (!bitmap->filemap) {
1060 1061 1062 1063 1064 1065 1066 1067
			/* error or shutdown */
			spin_unlock_irqrestore(&bitmap->lock, flags);
			break;
		}

		page = filemap_get_page(bitmap, j);

		if (page != lastpage) {
1068
			/* skip this page unless it's marked as needing cleaning */
1069 1070 1071
			if (!test_page_attr(bitmap, page, BITMAP_PAGE_CLEAN)) {
				int need_write = test_page_attr(bitmap, page,
								BITMAP_PAGE_NEEDWRITE);
1072
				if (need_write)
1073
					clear_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
1074

1075
				spin_unlock_irqrestore(&bitmap->lock, flags);
1076
				if (need_write) {
1077
					write_page(bitmap, page, 0);
1078 1079
					bitmap->allclean = 0;
				}
1080 1081 1082
				continue;
			}

1083 1084
			/* grab the new page, sync and release the old */
			if (lastpage != NULL) {
1085
				if (test_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE)) {
1086 1087
					clear_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
					spin_unlock_irqrestore(&bitmap->lock, flags);
1088
					write_page(bitmap, lastpage, 0);
1089 1090 1091 1092 1093 1094 1095
				} else {
					set_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
					spin_unlock_irqrestore(&bitmap->lock, flags);
				}
			} else
				spin_unlock_irqrestore(&bitmap->lock, flags);
			lastpage = page;
1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108

			/* We are possibly going to clear some bits, so make
			 * sure that events_cleared is up-to-date.
			 */
			if (bitmap->need_sync) {
				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);
			}
1109 1110 1111 1112 1113 1114 1115 1116 1117
			spin_lock_irqsave(&bitmap->lock, flags);
			clear_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
		}
		bmc = bitmap_get_counter(bitmap, j << CHUNK_BLOCK_SHIFT(bitmap),
					&blocks, 0);
		if (bmc) {
/*
  if (j < 100) printk("bitmap: j=%lu, *bmc = 0x%x\n", j, *bmc);
*/
1118 1119 1120
			if (*bmc)
				bitmap->allclean = 0;

1121 1122 1123 1124 1125 1126 1127 1128 1129 1130
			if (*bmc == 2) {
				*bmc=1; /* maybe clear the bit next time */
				set_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
			} else if (*bmc == 1) {
				/* we can clear the bit */
				*bmc = 0;
				bitmap_count_page(bitmap, j << CHUNK_BLOCK_SHIFT(bitmap),
						  -1);

				/* clear the bit */
1131
				paddr = kmap_atomic(page, KM_USER0);
1132
				if (bitmap->flags & BITMAP_HOSTENDIAN)
1133
					clear_bit(file_page_offset(j), paddr);
1134
				else
1135 1136
					ext2_clear_bit(file_page_offset(j), paddr);
				kunmap_atomic(paddr, KM_USER0);
1137 1138 1139 1140 1141 1142 1143 1144
			}
		}
		spin_unlock_irqrestore(&bitmap->lock, flags);
	}

	/* now sync the final page */
	if (lastpage != NULL) {
		spin_lock_irqsave(&bitmap->lock, flags);
1145
		if (test_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE)) {
1146 1147
			clear_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
			spin_unlock_irqrestore(&bitmap->lock, flags);
1148
			write_page(bitmap, lastpage, 0);
1149 1150 1151 1152 1153 1154
		} else {
			set_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
			spin_unlock_irqrestore(&bitmap->lock, flags);
		}
	}

1155
 done:
1156 1157
	if (bitmap->allclean == 0)
		bitmap->mddev->thread->timeout = bitmap->daemon_sleep * HZ;
1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
}

static bitmap_counter_t *bitmap_get_counter(struct bitmap *bitmap,
					    sector_t offset, int *blocks,
					    int create)
{
	/* 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;

	if (bitmap_checkpage(bitmap, page, create) < 0) {
		csize = ((sector_t)1) << (CHUNK_BLOCK_SHIFT(bitmap));
		*blocks = csize - (offset & (csize- 1));
		return NULL;
	}
	/* 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);
		csize = ((sector_t)1) << (CHUNK_BLOCK_SHIFT(bitmap) +
					  PAGE_COUNTER_SHIFT - 1);
		*blocks = csize - (offset & (csize- 1));
		return  &((bitmap_counter_t *)
			  &bitmap->bp[page].map)[hi];
	} else { /* page is allocated */
		csize = ((sector_t)1) << (CHUNK_BLOCK_SHIFT(bitmap));
		*blocks = csize - (offset & (csize- 1));
		return (bitmap_counter_t *)
			&(bitmap->bp[page].map[pageoff]);
	}
}

1197
int bitmap_startwrite(struct bitmap *bitmap, sector_t offset, unsigned long sectors, int behind)
1198 1199
{
	if (!bitmap) return 0;
1200 1201 1202 1203 1204 1205 1206

	if (behind) {
		atomic_inc(&bitmap->behind_writes);
		PRINTK(KERN_DEBUG "inc write-behind count %d/%d\n",
		  atomic_read(&bitmap->behind_writes), bitmap->max_write_behind);
	}

1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217
	while (sectors) {
		int blocks;
		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;
		}

1218 1219 1220 1221 1222 1223 1224 1225 1226
		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);
1227
			blk_unplug(bitmap->mddev->queue);
1228 1229 1230 1231 1232
			schedule();
			finish_wait(&bitmap->overflow_wait, &__wait);
			continue;
		}

1233 1234 1235 1236 1237 1238 1239 1240 1241
		switch(*bmc) {
		case 0:
			bitmap_file_set_bit(bitmap, offset);
			bitmap_count_page(bitmap,offset, 1);
			blk_plug_device(bitmap->mddev->queue);
			/* fall through */
		case 1:
			*bmc = 2;
		}
1242

1243 1244 1245 1246 1247 1248 1249 1250 1251
		(*bmc)++;

		spin_unlock_irq(&bitmap->lock);

		offset += blocks;
		if (sectors > blocks)
			sectors -= blocks;
		else sectors = 0;
	}
1252
	bitmap->allclean = 0;
1253 1254 1255 1256
	return 0;
}

void bitmap_endwrite(struct bitmap *bitmap, sector_t offset, unsigned long sectors,
1257
		     int success, int behind)
1258 1259
{
	if (!bitmap) return;
1260 1261 1262 1263 1264 1265
	if (behind) {
		atomic_dec(&bitmap->behind_writes);
		PRINTK(KERN_DEBUG "dec write-behind count %d/%d\n",
		  atomic_read(&bitmap->behind_writes), bitmap->max_write_behind);
	}

1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277
	while (sectors) {
		int blocks;
		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;
		}

1278 1279 1280 1281 1282 1283
		if (success &&
		    bitmap->events_cleared < bitmap->mddev->events) {
			bitmap->events_cleared = bitmap->mddev->events;
			bitmap->need_sync = 1;
		}

1284 1285 1286
		if (!success && ! (*bmc & NEEDED_MASK))
			*bmc |= NEEDED_MASK;

1287 1288 1289
		if ((*bmc & COUNTER_MAX) == COUNTER_MAX)
			wake_up(&bitmap->overflow_wait);

1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303
		(*bmc)--;
		if (*bmc <= 2) {
			set_page_attr(bitmap,
				      filemap_get_page(bitmap, offset >> CHUNK_BLOCK_SHIFT(bitmap)),
				      BITMAP_PAGE_CLEAN);
		}
		spin_unlock_irqrestore(&bitmap->lock, flags);
		offset += blocks;
		if (sectors > blocks)
			sectors -= blocks;
		else sectors = 0;
	}
}

1304 1305
int bitmap_start_sync(struct bitmap *bitmap, sector_t offset, int *blocks,
			int degraded)
1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321
{
	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;
1322 1323 1324 1325
			if (!degraded) { /* don't set/clear bits if degraded */
				*bmc |= RESYNC_MASK;
				*bmc &= ~NEEDED_MASK;
			}
1326 1327 1328
		}
	}
	spin_unlock_irq(&bitmap->lock);
1329
	bitmap->allclean = 0;
1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365
	return rv;
}

void bitmap_end_sync(struct bitmap *bitmap, sector_t offset, int *blocks, int aborted)
{
	bitmap_counter_t *bmc;
	unsigned long flags;
/*
	if (offset == 0) printk("bitmap_end_sync 0 (%d)\n", aborted);
*/	if (bitmap == NULL) {
		*blocks = 1024;
		return;
	}
	spin_lock_irqsave(&bitmap->lock, flags);
	bmc = bitmap_get_counter(bitmap, offset, blocks, 0);
	if (bmc == NULL)
		goto unlock;
	/* locked */
/*
	if (offset == 0) printk("bitmap_end sync found 0x%x, blocks %d\n", *bmc, *blocks);
*/
	if (RESYNC(*bmc)) {
		*bmc &= ~RESYNC_MASK;

		if (!NEEDED(*bmc) && aborted)
			*bmc |= NEEDED_MASK;
		else {
			if (*bmc <= 2) {
				set_page_attr(bitmap,
					      filemap_get_page(bitmap, offset >> CHUNK_BLOCK_SHIFT(bitmap)),
					      BITMAP_PAGE_CLEAN);
			}
		}
	}
 unlock:
	spin_unlock_irqrestore(&bitmap->lock, flags);
1366
	bitmap->allclean = 0;
1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
}

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;
	int blocks;
N
NeilBrown 已提交
1377 1378
	if (!bitmap)
		return;
1379 1380
	while (sector < bitmap->mddev->resync_max_sectors) {
		bitmap_end_sync(bitmap, sector, &blocks, 0);
N
NeilBrown 已提交
1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406
		sector += blocks;
	}
}

void bitmap_cond_end_sync(struct bitmap *bitmap, sector_t sector)
{
	sector_t s = 0;
	int blocks;

	if (!bitmap)
		return;
	if (sector == 0) {
		bitmap->last_end_sync = jiffies;
		return;
	}
	if (time_before(jiffies, (bitmap->last_end_sync
				  + bitmap->daemon_sleep * HZ)))
		return;
	wait_event(bitmap->mddev->recovery_wait,
		   atomic_read(&bitmap->mddev->recovery_active) == 0);

	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;
1407
	}
N
NeilBrown 已提交
1408
	bitmap->last_end_sync = jiffies;
1409 1410
}

1411
static void bitmap_set_memory_bits(struct bitmap *bitmap, sector_t offset, int needed)
1412 1413
{
	/* For each chunk covered by any of these sectors, set the
1414
	 * counter to 1 and set resync_needed.  They should all
1415 1416
	 * be 0 at this point
	 */
1417 1418 1419 1420 1421 1422

	int secs;
	bitmap_counter_t *bmc;
	spin_lock_irq(&bitmap->lock);
	bmc = bitmap_get_counter(bitmap, offset, &secs, 1);
	if (!bmc) {
1423
		spin_unlock_irq(&bitmap->lock);
1424
		return;
1425
	}
1426 1427
	if (! *bmc) {
		struct page *page;
1428
		*bmc = 1 | (needed?NEEDED_MASK:0);
1429 1430 1431 1432 1433
		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);
1434
	bitmap->allclean = 0;
1435 1436
}

1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448
/* 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++) {
		sector_t sec = chunk << CHUNK_BLOCK_SHIFT(bitmap);
		bitmap_set_memory_bits(bitmap, sec, 1);
		bitmap_file_set_bit(bitmap, sec);
	}
}

1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471
/*
 * flush out any pending updates
 */
void bitmap_flush(mddev_t *mddev)
{
	struct bitmap *bitmap = mddev->bitmap;
	int sleep;

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

	/* run the daemon_work three time to ensure everything is flushed
	 * that can be
	 */
	sleep = bitmap->daemon_sleep;
	bitmap->daemon_sleep = 0;
	bitmap_daemon_work(bitmap);
	bitmap_daemon_work(bitmap);
	bitmap_daemon_work(bitmap);
	bitmap->daemon_sleep = sleep;
	bitmap_update_sb(bitmap);
}

1472 1473 1474
/*
 * free memory that was allocated
 */
1475
static void bitmap_free(struct bitmap *bitmap)
1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497
{
	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);
}
1498 1499 1500 1501 1502 1503 1504 1505
void bitmap_destroy(mddev_t *mddev)
{
	struct bitmap *bitmap = mddev->bitmap;

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

	mddev->bitmap = NULL; /* disconnect from the md device */
1506 1507
	if (mddev->thread)
		mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
1508 1509 1510

	bitmap_free(bitmap);
}
1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523

/*
 * 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;
	unsigned long blocks = mddev->resync_max_sectors;
	unsigned long chunks;
	unsigned long pages;
	struct file *file = mddev->bitmap_file;
	int err;
1524
	sector_t start;
1525

A
Alexey Dobriyan 已提交
1526
	BUILD_BUG_ON(sizeof(bitmap_super_t) != 256);
1527

1528
	if (!file && !mddev->bitmap_offset) /* bitmap disabled, nothing to do */
1529 1530
		return 0;

1531 1532
	BUG_ON(file && mddev->bitmap_offset);

1533
	bitmap = kzalloc(sizeof(*bitmap), GFP_KERNEL);
1534 1535 1536 1537
	if (!bitmap)
		return -ENOMEM;

	spin_lock_init(&bitmap->lock);
1538 1539
	atomic_set(&bitmap->pending_writes, 0);
	init_waitqueue_head(&bitmap->write_wait);
1540
	init_waitqueue_head(&bitmap->overflow_wait);
1541

1542 1543 1544
	bitmap->mddev = mddev;

	bitmap->file = file;
1545
	bitmap->offset = mddev->bitmap_offset;
1546 1547
	if (file) {
		get_file(file);
M
Mark Fasheh 已提交
1548 1549 1550 1551
		do_sync_mapping_range(file->f_mapping, 0, LLONG_MAX,
				      SYNC_FILE_RANGE_WAIT_BEFORE |
				      SYNC_FILE_RANGE_WRITE |
				      SYNC_FILE_RANGE_WAIT_AFTER);
1552
	}
1553 1554 1555
	/* read superblock from bitmap file (this sets bitmap->chunksize) */
	err = bitmap_read_sb(bitmap);
	if (err)
1556
		goto error;
1557

1558
	bitmap->chunkshift = ffz(~bitmap->chunksize);
1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573

	/* now that chunksize and chunkshift are set, we can use these macros */
 	chunks = (blocks + CHUNK_BLOCK_RATIO(bitmap) - 1) /
			CHUNK_BLOCK_RATIO(bitmap);
 	pages = (chunks + PAGE_COUNTER_RATIO - 1) / PAGE_COUNTER_RATIO;

	BUG_ON(!pages);

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

	bitmap->syncchunk = ~0UL;

1574
#ifdef INJECT_FATAL_FAULT_1
1575 1576
	bitmap->bp = NULL;
#else
1577
	bitmap->bp = kzalloc(pages * sizeof(*bitmap->bp), GFP_KERNEL);
1578
#endif
1579
	err = -ENOMEM;
1580
	if (!bitmap->bp)
1581
		goto error;
1582 1583 1584

	/* now that we have some pages available, initialize the in-memory
	 * bitmap from the on-disk bitmap */
1585 1586 1587 1588 1589 1590
	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;
	err = bitmap_init_from_disk(bitmap, start);
1591

1592
	if (err)
1593
		goto error;
1594 1595 1596 1597

	printk(KERN_INFO "created bitmap (%lu pages) for device %s\n",
		pages, bmname(bitmap));

1598 1599
	mddev->bitmap = bitmap;

1600 1601
	mddev->thread->timeout = bitmap->daemon_sleep * HZ;

1602 1603 1604
	bitmap_update_sb(bitmap);

	return (bitmap->flags & BITMAP_WRITE_ERROR) ? -EIO : 0;
1605 1606 1607 1608

 error:
	bitmap_free(bitmap);
	return err;
1609 1610 1611 1612 1613 1614 1615 1616 1617
}

/* the bitmap API -- for raid personalities */
EXPORT_SYMBOL(bitmap_startwrite);
EXPORT_SYMBOL(bitmap_endwrite);
EXPORT_SYMBOL(bitmap_start_sync);
EXPORT_SYMBOL(bitmap_end_sync);
EXPORT_SYMBOL(bitmap_unplug);
EXPORT_SYMBOL(bitmap_close_sync);
N
NeilBrown 已提交
1618
EXPORT_SYMBOL(bitmap_cond_end_sync);