dm-table.c 39.7 KB
Newer Older
L
Linus Torvalds 已提交
1 2
/*
 * Copyright (C) 2001 Sistina Software (UK) Limited.
3
 * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
L
Linus Torvalds 已提交
4 5 6 7 8 9 10 11 12 13 14
 *
 * This file is released under the GPL.
 */

#include "dm.h"

#include <linux/module.h>
#include <linux/vmalloc.h>
#include <linux/blkdev.h>
#include <linux/namei.h>
#include <linux/ctype.h>
15
#include <linux/string.h>
L
Linus Torvalds 已提交
16 17
#include <linux/slab.h>
#include <linux/interrupt.h>
A
Arjan van de Ven 已提交
18
#include <linux/mutex.h>
19
#include <linux/delay.h>
A
Arun Sharma 已提交
20
#include <linux/atomic.h>
L
Linus Torvalds 已提交
21

22 23
#define DM_MSG_PREFIX "table"

L
Linus Torvalds 已提交
24 25 26 27 28 29
#define MAX_DEPTH 16
#define NODE_SIZE L1_CACHE_BYTES
#define KEYS_PER_NODE (NODE_SIZE / sizeof(sector_t))
#define CHILDREN_PER_NODE (KEYS_PER_NODE + 1)

struct dm_table {
M
Mike Anderson 已提交
30
	struct mapped_device *md;
K
Kiyoshi Ueda 已提交
31
	unsigned type;
L
Linus Torvalds 已提交
32 33 34 35 36 37 38 39 40 41 42

	/* btree table */
	unsigned int depth;
	unsigned int counts[MAX_DEPTH];	/* in nodes */
	sector_t *index[MAX_DEPTH];

	unsigned int num_targets;
	unsigned int num_allocated;
	sector_t *highs;
	struct dm_target *targets;

43
	struct target_type *immutable_target_type;
44
	unsigned integrity_supported:1;
45
	unsigned singleton:1;
M
Mike Snitzer 已提交
46

L
Linus Torvalds 已提交
47 48 49 50 51
	/*
	 * Indicates the rw permissions for the new logical
	 * device.  This should be a combination of FMODE_READ
	 * and FMODE_WRITE.
	 */
52
	fmode_t mode;
L
Linus Torvalds 已提交
53 54 55 56 57 58 59

	/* a list of devices used by this table */
	struct list_head devices;

	/* events get handed up using this callback */
	void (*event_fn)(void *);
	void *event_context;
K
Kiyoshi Ueda 已提交
60 61

	struct dm_md_mempools *mempools;
62 63

	struct list_head target_callbacks;
L
Linus Torvalds 已提交
64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 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
};

/*
 * Similar to ceiling(log_size(n))
 */
static unsigned int int_log(unsigned int n, unsigned int base)
{
	int result = 0;

	while (n > 1) {
		n = dm_div_up(n, base);
		result++;
	}

	return result;
}

/*
 * Calculate the index of the child node of the n'th node k'th key.
 */
static inline unsigned int get_child(unsigned int n, unsigned int k)
{
	return (n * CHILDREN_PER_NODE) + k;
}

/*
 * Return the n'th node of level l from table t.
 */
static inline sector_t *get_node(struct dm_table *t,
				 unsigned int l, unsigned int n)
{
	return t->index[l] + (n * KEYS_PER_NODE);
}

/*
 * Return the highest key that you could lookup from the n'th
 * node on level l of the btree.
 */
static sector_t high(struct dm_table *t, unsigned int l, unsigned int n)
{
	for (; l < t->depth - 1; l++)
		n = get_child(n, CHILDREN_PER_NODE - 1);

	if (n >= t->counts[l])
		return (sector_t) - 1;

	return get_node(t, l, n)[KEYS_PER_NODE - 1];
}

/*
 * Fills in a level of the btree based on the highs of the level
 * below it.
 */
static int setup_btree_index(unsigned int l, struct dm_table *t)
{
	unsigned int n, k;
	sector_t *node;

	for (n = 0U; n < t->counts[l]; n++) {
		node = get_node(t, l, n);

		for (k = 0U; k < KEYS_PER_NODE; k++)
			node[k] = high(t, l + 1, get_child(n, k));
	}

	return 0;
}

void *dm_vcalloc(unsigned long nmemb, unsigned long elem_size)
{
	unsigned long size;
	void *addr;

	/*
	 * Check that we're not going to overflow.
	 */
	if (nmemb > (ULONG_MAX / elem_size))
		return NULL;

	size = nmemb * elem_size;
J
Joe Perches 已提交
144
	addr = vzalloc(size);
L
Linus Torvalds 已提交
145 146 147

	return addr;
}
148
EXPORT_SYMBOL(dm_vcalloc);
L
Linus Torvalds 已提交
149 150 151 152 153 154 155 156 157 158 159 160

/*
 * highs, and targets are managed as dynamic arrays during a
 * table load.
 */
static int alloc_targets(struct dm_table *t, unsigned int num)
{
	sector_t *n_highs;
	struct dm_target *n_targets;

	/*
	 * Allocate both the target array and offset array at once.
161 162
	 * Append an empty entry to catch sectors beyond the end of
	 * the device.
L
Linus Torvalds 已提交
163
	 */
164
	n_highs = (sector_t *) dm_vcalloc(num + 1, sizeof(struct dm_target) +
L
Linus Torvalds 已提交
165 166 167 168 169 170
					  sizeof(sector_t));
	if (!n_highs)
		return -ENOMEM;

	n_targets = (struct dm_target *) (n_highs + num);

171
	memset(n_highs, -1, sizeof(*n_highs) * num);
L
Linus Torvalds 已提交
172 173 174 175 176 177 178 179 180
	vfree(t->highs);

	t->num_allocated = num;
	t->highs = n_highs;
	t->targets = n_targets;

	return 0;
}

181
int dm_table_create(struct dm_table **result, fmode_t mode,
M
Mike Anderson 已提交
182
		    unsigned num_targets, struct mapped_device *md)
L
Linus Torvalds 已提交
183
{
D
Dmitry Monakhov 已提交
184
	struct dm_table *t = kzalloc(sizeof(*t), GFP_KERNEL);
L
Linus Torvalds 已提交
185 186 187 188 189

	if (!t)
		return -ENOMEM;

	INIT_LIST_HEAD(&t->devices);
190
	INIT_LIST_HEAD(&t->target_callbacks);
L
Linus Torvalds 已提交
191 192 193 194 195 196

	if (!num_targets)
		num_targets = KEYS_PER_NODE;

	num_targets = dm_round_up(num_targets, KEYS_PER_NODE);

197 198 199 200 201
	if (!num_targets) {
		kfree(t);
		return -ENOMEM;
	}

L
Linus Torvalds 已提交
202 203 204 205 206 207
	if (alloc_targets(t, num_targets)) {
		kfree(t);
		return -ENOMEM;
	}

	t->mode = mode;
M
Mike Anderson 已提交
208
	t->md = md;
L
Linus Torvalds 已提交
209 210 211 212
	*result = t;
	return 0;
}

213
static void free_devices(struct list_head *devices, struct mapped_device *md)
L
Linus Torvalds 已提交
214 215 216
{
	struct list_head *tmp, *next;

P
Paul Jimenez 已提交
217
	list_for_each_safe(tmp, next, devices) {
M
Mikulas Patocka 已提交
218 219
		struct dm_dev_internal *dd =
		    list_entry(tmp, struct dm_dev_internal, list);
220 221 222
		DMWARN("%s: dm_table_destroy: dm_put_device call missing for %s",
		       dm_device_name(md), dd->dm_dev->name);
		dm_put_table_device(md, dd->dm_dev);
L
Linus Torvalds 已提交
223 224 225 226
		kfree(dd);
	}
}

227
void dm_table_destroy(struct dm_table *t)
L
Linus Torvalds 已提交
228 229 230
{
	unsigned int i;

231 232 233
	if (!t)
		return;

234
	/* free the indexes */
L
Linus Torvalds 已提交
235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250
	if (t->depth >= 2)
		vfree(t->index[t->depth - 2]);

	/* free the targets */
	for (i = 0; i < t->num_targets; i++) {
		struct dm_target *tgt = t->targets + i;

		if (tgt->type->dtr)
			tgt->type->dtr(tgt);

		dm_put_target_type(tgt->type);
	}

	vfree(t->highs);

	/* free the device list */
251
	free_devices(&t->devices, t->md);
L
Linus Torvalds 已提交
252

K
Kiyoshi Ueda 已提交
253 254
	dm_free_md_mempools(t->mempools);

L
Linus Torvalds 已提交
255 256 257 258 259 260
	kfree(t);
}

/*
 * See if we've already got a device in the list.
 */
M
Mikulas Patocka 已提交
261
static struct dm_dev_internal *find_device(struct list_head *l, dev_t dev)
L
Linus Torvalds 已提交
262
{
M
Mikulas Patocka 已提交
263
	struct dm_dev_internal *dd;
L
Linus Torvalds 已提交
264 265

	list_for_each_entry (dd, l, list)
266
		if (dd->dm_dev->bdev->bd_dev == dev)
L
Linus Torvalds 已提交
267 268 269 270 271 272
			return dd;

	return NULL;
}

/*
273
 * If possible, this checks an area of a destination device is invalid.
L
Linus Torvalds 已提交
274
 */
275 276
static int device_area_is_invalid(struct dm_target *ti, struct dm_dev *dev,
				  sector_t start, sector_t len, void *data)
L
Linus Torvalds 已提交
277
{
278
	struct request_queue *q;
279 280 281 282
	struct queue_limits *limits = data;
	struct block_device *bdev = dev->bdev;
	sector_t dev_size =
		i_size_read(bdev->bd_inode) >> SECTOR_SHIFT;
283
	unsigned short logical_block_size_sectors =
284
		limits->logical_block_size >> SECTOR_SHIFT;
285
	char b[BDEVNAME_SIZE];
M
Mike Anderson 已提交
286

287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302
	/*
	 * Some devices exist without request functions,
	 * such as loop devices not yet bound to backing files.
	 * Forbid the use of such devices.
	 */
	q = bdev_get_queue(bdev);
	if (!q || !q->make_request_fn) {
		DMWARN("%s: %s is not yet initialised: "
		       "start=%llu, len=%llu, dev_size=%llu",
		       dm_device_name(ti->table->md), bdevname(bdev, b),
		       (unsigned long long)start,
		       (unsigned long long)len,
		       (unsigned long long)dev_size);
		return 1;
	}

M
Mike Anderson 已提交
303
	if (!dev_size)
304
		return 0;
M
Mike Anderson 已提交
305

306
	if ((start >= dev_size) || (start + len > dev_size)) {
307 308 309 310 311 312
		DMWARN("%s: %s too small for target: "
		       "start=%llu, len=%llu, dev_size=%llu",
		       dm_device_name(ti->table->md), bdevname(bdev, b),
		       (unsigned long long)start,
		       (unsigned long long)len,
		       (unsigned long long)dev_size);
313
		return 1;
314 315 316
	}

	if (logical_block_size_sectors <= 1)
317
		return 0;
318 319 320

	if (start & (logical_block_size_sectors - 1)) {
		DMWARN("%s: start=%llu not aligned to h/w "
321
		       "logical block size %u of %s",
322 323
		       dm_device_name(ti->table->md),
		       (unsigned long long)start,
324
		       limits->logical_block_size, bdevname(bdev, b));
325
		return 1;
326 327
	}

328
	if (len & (logical_block_size_sectors - 1)) {
329
		DMWARN("%s: len=%llu not aligned to h/w "
330
		       "logical block size %u of %s",
331
		       dm_device_name(ti->table->md),
332
		       (unsigned long long)len,
333
		       limits->logical_block_size, bdevname(bdev, b));
334
		return 1;
335 336
	}

337
	return 0;
L
Linus Torvalds 已提交
338 339 340
}

/*
341
 * This upgrades the mode on an already open dm_dev, being
L
Linus Torvalds 已提交
342
 * careful to leave things as they were if we fail to reopen the
343 344
 * device and not to touch the existing bdev field in case
 * it is accessed concurrently inside dm_table_any_congested().
L
Linus Torvalds 已提交
345
 */
346
static int upgrade_mode(struct dm_dev_internal *dd, fmode_t new_mode,
M
Mikulas Patocka 已提交
347
			struct mapped_device *md)
L
Linus Torvalds 已提交
348 349
{
	int r;
350
	struct dm_dev *old_dev, *new_dev;
L
Linus Torvalds 已提交
351

352
	old_dev = dd->dm_dev;
353

354 355
	r = dm_get_table_device(md, dd->dm_dev->bdev->bd_dev,
				dd->dm_dev->mode | new_mode, &new_dev);
356 357
	if (r)
		return r;
L
Linus Torvalds 已提交
358

359 360
	dd->dm_dev = new_dev;
	dm_put_table_device(md, old_dev);
L
Linus Torvalds 已提交
361

362
	return 0;
L
Linus Torvalds 已提交
363 364 365 366 367 368
}

/*
 * Add a device to the list, or just increment the usage count if
 * it's already present.
 */
369 370
int dm_get_device(struct dm_target *ti, const char *path, fmode_t mode,
		  struct dm_dev **result)
L
Linus Torvalds 已提交
371 372
{
	int r;
A
Andrew Morton 已提交
373
	dev_t uninitialized_var(dev);
M
Mikulas Patocka 已提交
374
	struct dm_dev_internal *dd;
L
Linus Torvalds 已提交
375
	unsigned int major, minor;
376
	struct dm_table *t = ti->table;
377
	char dummy;
L
Linus Torvalds 已提交
378

379
	BUG_ON(!t);
L
Linus Torvalds 已提交
380

381
	if (sscanf(path, "%u:%u%c", &major, &minor, &dummy) == 2) {
L
Linus Torvalds 已提交
382 383 384 385 386 387
		/* Extract the major/minor numbers */
		dev = MKDEV(major, minor);
		if (MAJOR(dev) != major || MINOR(dev) != minor)
			return -EOVERFLOW;
	} else {
		/* convert the path to a device */
388 389 390 391 392 393
		struct block_device *bdev = lookup_bdev(path);

		if (IS_ERR(bdev))
			return PTR_ERR(bdev);
		dev = bdev->bd_dev;
		bdput(bdev);
L
Linus Torvalds 已提交
394 395 396 397 398 399 400 401
	}

	dd = find_device(&t->devices, dev);
	if (!dd) {
		dd = kmalloc(sizeof(*dd), GFP_KERNEL);
		if (!dd)
			return -ENOMEM;

402
		if ((r = dm_get_table_device(t->md, dev, mode, &dd->dm_dev))) {
L
Linus Torvalds 已提交
403 404 405 406 407 408 409
			kfree(dd);
			return r;
		}

		atomic_set(&dd->count, 0);
		list_add(&dd->list, &t->devices);

410
	} else if (dd->dm_dev->mode != (mode | dd->dm_dev->mode)) {
411
		r = upgrade_mode(dd, mode, t->md);
L
Linus Torvalds 已提交
412 413 414 415 416
		if (r)
			return r;
	}
	atomic_inc(&dd->count);

417
	*result = dd->dm_dev;
L
Linus Torvalds 已提交
418 419
	return 0;
}
420
EXPORT_SYMBOL(dm_get_device);
L
Linus Torvalds 已提交
421

422 423
static int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev,
				sector_t start, sector_t len, void *data)
L
Linus Torvalds 已提交
424
{
425 426
	struct queue_limits *limits = data;
	struct block_device *bdev = dev->bdev;
427
	struct request_queue *q = bdev_get_queue(bdev);
A
Alasdair G Kergon 已提交
428 429 430 431 432
	char b[BDEVNAME_SIZE];

	if (unlikely(!q)) {
		DMWARN("%s: Cannot set limits for nonexistent device %s",
		       dm_device_name(ti->table->md), bdevname(bdev, b));
433
		return 0;
A
Alasdair G Kergon 已提交
434
	}
435

436 437
	if (bdev_stack_limits(limits, bdev, start) < 0)
		DMWARN("%s: adding target device %s caused an alignment inconsistency: "
438 439 440 441 442 443
		       "physical_block_size=%u, logical_block_size=%u, "
		       "alignment_offset=%u, start=%llu",
		       dm_device_name(ti->table->md), bdevname(bdev, b),
		       q->limits.physical_block_size,
		       q->limits.logical_block_size,
		       q->limits.alignment_offset,
444
		       (unsigned long long) start << SECTOR_SHIFT);
445

446 447 448 449
	/*
	 * Check if merge fn is supported.
	 * If not we'll force DM to use PAGE_SIZE or
	 * smaller I/O, just to be safe.
450
	 */
451
	if (dm_queue_merge_is_compulsory(q) && !ti->type->merge)
452 453
		blk_limits_max_hw_sectors(limits,
					  (unsigned int) (PAGE_SIZE >> 9));
454
	return 0;
455
}
456

L
Linus Torvalds 已提交
457
/*
458
 * Decrement a device's use count and remove it if necessary.
L
Linus Torvalds 已提交
459
 */
M
Mikulas Patocka 已提交
460
void dm_put_device(struct dm_target *ti, struct dm_dev *d)
L
Linus Torvalds 已提交
461
{
462 463 464
	int found = 0;
	struct list_head *devices = &ti->table->devices;
	struct dm_dev_internal *dd;
M
Mikulas Patocka 已提交
465

466 467 468 469 470 471 472 473 474 475 476
	list_for_each_entry(dd, devices, list) {
		if (dd->dm_dev == d) {
			found = 1;
			break;
		}
	}
	if (!found) {
		DMWARN("%s: device %s not in table devices list",
		       dm_device_name(ti->table->md), d->name);
		return;
	}
L
Linus Torvalds 已提交
477
	if (atomic_dec_and_test(&dd->count)) {
478
		dm_put_table_device(ti->table->md, d);
L
Linus Torvalds 已提交
479 480 481 482
		list_del(&dd->list);
		kfree(dd);
	}
}
483
EXPORT_SYMBOL(dm_put_device);
L
Linus Torvalds 已提交
484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500

/*
 * Checks to see if the target joins onto the end of the table.
 */
static int adjoin(struct dm_table *table, struct dm_target *ti)
{
	struct dm_target *prev;

	if (!table->num_targets)
		return !ti->begin;

	prev = &table->targets[table->num_targets - 1];
	return (ti->begin == (prev->begin + prev->len));
}

/*
 * Used to dynamically allocate the arg array.
501 502 503 504 505 506 507
 *
 * We do first allocation with GFP_NOIO because dm-mpath and dm-thin must
 * process messages even if some device is suspended. These messages have a
 * small fixed number of arguments.
 *
 * On the other hand, dm-switch needs to process bulk data using messages and
 * excessive use of GFP_NOIO could cause trouble.
L
Linus Torvalds 已提交
508 509 510 511 512
 */
static char **realloc_argv(unsigned *array_size, char **old_argv)
{
	char **argv;
	unsigned new_size;
513
	gfp_t gfp;
L
Linus Torvalds 已提交
514

515 516 517 518 519 520 521 522
	if (*array_size) {
		new_size = *array_size * 2;
		gfp = GFP_KERNEL;
	} else {
		new_size = 8;
		gfp = GFP_NOIO;
	}
	argv = kmalloc(new_size * sizeof(*argv), gfp);
L
Linus Torvalds 已提交
523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540
	if (argv) {
		memcpy(argv, old_argv, *array_size * sizeof(*argv));
		*array_size = new_size;
	}

	kfree(old_argv);
	return argv;
}

/*
 * Destructively splits up the argument list to pass to ctr.
 */
int dm_split_args(int *argc, char ***argvp, char *input)
{
	char *start, *end = input, *out, **argv = NULL;
	unsigned array_size = 0;

	*argc = 0;
541 542 543 544 545 546

	if (!input) {
		*argvp = NULL;
		return 0;
	}

L
Linus Torvalds 已提交
547 548 549 550 551 552
	argv = realloc_argv(&array_size, argv);
	if (!argv)
		return -ENOMEM;

	while (1) {
		/* Skip whitespace */
553
		start = skip_spaces(end);
L
Linus Torvalds 已提交
554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594

		if (!*start)
			break;	/* success, we hit the end */

		/* 'out' is used to remove any back-quotes */
		end = out = start;
		while (*end) {
			/* Everything apart from '\0' can be quoted */
			if (*end == '\\' && *(end + 1)) {
				*out++ = *(end + 1);
				end += 2;
				continue;
			}

			if (isspace(*end))
				break;	/* end of token */

			*out++ = *end++;
		}

		/* have we already filled the array ? */
		if ((*argc + 1) > array_size) {
			argv = realloc_argv(&array_size, argv);
			if (!argv)
				return -ENOMEM;
		}

		/* we know this is whitespace */
		if (*end)
			end++;

		/* terminate the string and put it in the array */
		*out = '\0';
		argv[*argc] = start;
		(*argc)++;
	}

	*argvp = argv;
	return 0;
}

595 596 597 598 599 600 601
/*
 * Impose necessary and sufficient conditions on a devices's table such
 * that any incoming bio which respects its logical_block_size can be
 * processed successfully.  If it falls across the boundary between
 * two or more targets, the size of each piece it gets split into must
 * be compatible with the logical_block_size of the target processing it.
 */
602 603
static int validate_hardware_logical_block_alignment(struct dm_table *table,
						 struct queue_limits *limits)
604 605 606 607 608 609
{
	/*
	 * This function uses arithmetic modulo the logical_block_size
	 * (in units of 512-byte sectors).
	 */
	unsigned short device_logical_block_size_sects =
610
		limits->logical_block_size >> SECTOR_SHIFT;
611 612 613 614 615 616 617 618 619 620 621 622 623

	/*
	 * Offset of the start of the next table entry, mod logical_block_size.
	 */
	unsigned short next_target_start = 0;

	/*
	 * Given an aligned bio that extends beyond the end of a
	 * target, how many sectors must the next target handle?
	 */
	unsigned short remaining = 0;

	struct dm_target *uninitialized_var(ti);
624
	struct queue_limits ti_limits;
625 626 627 628 629 630 631 632
	unsigned i = 0;

	/*
	 * Check each entry in the table in turn.
	 */
	while (i < dm_table_get_num_targets(table)) {
		ti = dm_table_get_target(table, i++);

633
		blk_set_stacking_limits(&ti_limits);
634 635 636 637 638 639

		/* combine all target devices' limits */
		if (ti->type->iterate_devices)
			ti->type->iterate_devices(ti, dm_set_device_limits,
						  &ti_limits);

640 641 642 643 644
		/*
		 * If the remaining sectors fall entirely within this
		 * table entry are they compatible with its logical_block_size?
		 */
		if (remaining < ti->len &&
645
		    remaining & ((ti_limits.logical_block_size >>
646 647 648 649 650 651 652 653 654 655 656 657
				  SECTOR_SHIFT) - 1))
			break;	/* Error */

		next_target_start =
		    (unsigned short) ((next_target_start + ti->len) &
				      (device_logical_block_size_sects - 1));
		remaining = next_target_start ?
		    device_logical_block_size_sects - next_target_start : 0;
	}

	if (remaining) {
		DMWARN("%s: table line %u (start sect %llu len %llu) "
658
		       "not aligned to h/w logical block size %u",
659 660 661
		       dm_device_name(table->md), i,
		       (unsigned long long) ti->begin,
		       (unsigned long long) ti->len,
662
		       limits->logical_block_size);
663 664 665 666 667 668
		return -EINVAL;
	}

	return 0;
}

L
Linus Torvalds 已提交
669 670 671 672 673 674 675
int dm_table_add_target(struct dm_table *t, const char *type,
			sector_t start, sector_t len, char *params)
{
	int r = -EINVAL, argc;
	char **argv;
	struct dm_target *tgt;

676 677 678 679 680 681
	if (t->singleton) {
		DMERR("%s: target type %s must appear alone in table",
		      dm_device_name(t->md), t->targets->type->name);
		return -EINVAL;
	}

682
	BUG_ON(t->num_targets >= t->num_allocated);
L
Linus Torvalds 已提交
683 684 685 686 687

	tgt = t->targets + t->num_targets;
	memset(tgt, 0, sizeof(*tgt));

	if (!len) {
688
		DMERR("%s: zero-length target", dm_device_name(t->md));
L
Linus Torvalds 已提交
689 690 691 692 693
		return -EINVAL;
	}

	tgt->type = dm_get_target_type(type);
	if (!tgt->type) {
694 695
		DMERR("%s: %s: unknown target type", dm_device_name(t->md),
		      type);
L
Linus Torvalds 已提交
696 697 698
		return -EINVAL;
	}

699 700 701 702 703 704 705 706 707
	if (dm_target_needs_singleton(tgt->type)) {
		if (t->num_targets) {
			DMERR("%s: target type %s must appear alone in table",
			      dm_device_name(t->md), type);
			return -EINVAL;
		}
		t->singleton = 1;
	}

708 709 710 711 712 713
	if (dm_target_always_writeable(tgt->type) && !(t->mode & FMODE_WRITE)) {
		DMERR("%s: target type %s may not be included in read-only tables",
		      dm_device_name(t->md), type);
		return -EINVAL;
	}

714 715 716 717 718 719 720 721 722 723 724 725 726 727 728
	if (t->immutable_target_type) {
		if (t->immutable_target_type != tgt->type) {
			DMERR("%s: immutable target type %s cannot be mixed with other target types",
			      dm_device_name(t->md), t->immutable_target_type->name);
			return -EINVAL;
		}
	} else if (dm_target_is_immutable(tgt->type)) {
		if (t->num_targets) {
			DMERR("%s: immutable target type %s cannot be mixed with other target types",
			      dm_device_name(t->md), tgt->type->name);
			return -EINVAL;
		}
		t->immutable_target_type = tgt->type;
	}

L
Linus Torvalds 已提交
729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755
	tgt->table = t;
	tgt->begin = start;
	tgt->len = len;
	tgt->error = "Unknown error";

	/*
	 * Does this target adjoin the previous one ?
	 */
	if (!adjoin(t, tgt)) {
		tgt->error = "Gap in table";
		r = -EINVAL;
		goto bad;
	}

	r = dm_split_args(&argc, &argv, params);
	if (r) {
		tgt->error = "couldn't split parameters (insufficient memory)";
		goto bad;
	}

	r = tgt->type->ctr(tgt, argc, argv);
	kfree(argv);
	if (r)
		goto bad;

	t->highs[t->num_targets++] = tgt->begin + tgt->len - 1;

756 757
	if (!tgt->num_discard_bios && tgt->discards_supported)
		DMWARN("%s: %s: ignoring discards_supported because num_discard_bios is zero.",
M
Mike Snitzer 已提交
758
		       dm_device_name(t->md), type);
M
Mike Snitzer 已提交
759

L
Linus Torvalds 已提交
760 761 762
	return 0;

 bad:
763
	DMERR("%s: %s: %s", dm_device_name(t->md), type, tgt->error);
L
Linus Torvalds 已提交
764 765 766 767
	dm_put_target_type(tgt->type);
	return r;
}

768 769 770 771 772 773 774
/*
 * Target argument parsing helpers.
 */
static int validate_next_arg(struct dm_arg *arg, struct dm_arg_set *arg_set,
			     unsigned *value, char **error, unsigned grouped)
{
	const char *arg_str = dm_shift_arg(arg_set);
775
	char dummy;
776 777

	if (!arg_str ||
778
	    (sscanf(arg_str, "%u%c", value, &dummy) != 1) ||
779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825
	    (*value < arg->min) ||
	    (*value > arg->max) ||
	    (grouped && arg_set->argc < *value)) {
		*error = arg->error;
		return -EINVAL;
	}

	return 0;
}

int dm_read_arg(struct dm_arg *arg, struct dm_arg_set *arg_set,
		unsigned *value, char **error)
{
	return validate_next_arg(arg, arg_set, value, error, 0);
}
EXPORT_SYMBOL(dm_read_arg);

int dm_read_arg_group(struct dm_arg *arg, struct dm_arg_set *arg_set,
		      unsigned *value, char **error)
{
	return validate_next_arg(arg, arg_set, value, error, 1);
}
EXPORT_SYMBOL(dm_read_arg_group);

const char *dm_shift_arg(struct dm_arg_set *as)
{
	char *r;

	if (as->argc) {
		as->argc--;
		r = *as->argv;
		as->argv++;
		return r;
	}

	return NULL;
}
EXPORT_SYMBOL(dm_shift_arg);

void dm_consume_args(struct dm_arg_set *as, unsigned num_args)
{
	BUG_ON(as->argc < num_args);
	as->argc -= num_args;
	as->argv += num_args;
}
EXPORT_SYMBOL(dm_consume_args);

826
static int dm_table_set_type(struct dm_table *t)
K
Kiyoshi Ueda 已提交
827 828
{
	unsigned i;
829
	unsigned bio_based = 0, request_based = 0, hybrid = 0;
830
	bool use_blk_mq = false;
K
Kiyoshi Ueda 已提交
831 832 833
	struct dm_target *tgt;
	struct dm_dev_internal *dd;
	struct list_head *devices;
834
	unsigned live_md_type = dm_get_md_type(t->md);
K
Kiyoshi Ueda 已提交
835 836 837

	for (i = 0; i < t->num_targets; i++) {
		tgt = t->targets + i;
838 839 840
		if (dm_target_hybrid(tgt))
			hybrid = 1;
		else if (dm_target_request_based(tgt))
K
Kiyoshi Ueda 已提交
841 842 843 844 845 846 847 848 849 850 851
			request_based = 1;
		else
			bio_based = 1;

		if (bio_based && request_based) {
			DMWARN("Inconsistent table: different target types"
			       " can't be mixed up");
			return -EINVAL;
		}
	}

852 853 854 855 856 857
	if (hybrid && !bio_based && !request_based) {
		/*
		 * The targets can work either way.
		 * Determine the type from the live device.
		 * Default to bio-based if device is new.
		 */
858 859
		if (live_md_type == DM_TYPE_REQUEST_BASED ||
		    live_md_type == DM_TYPE_MQ_REQUEST_BASED)
860 861 862 863 864
			request_based = 1;
		else
			bio_based = 1;
	}

K
Kiyoshi Ueda 已提交
865 866 867 868 869 870 871 872
	if (bio_based) {
		/* We must use this table as bio-based */
		t->type = DM_TYPE_BIO_BASED;
		return 0;
	}

	BUG_ON(!request_based); /* No targets in this table */

873 874 875 876 877 878 879 880 881 882 883
	/*
	 * Request-based dm supports only tables that have a single target now.
	 * To support multiple targets, request splitting support is needed,
	 * and that needs lots of changes in the block-layer.
	 * (e.g. request completion process for partial completion.)
	 */
	if (t->num_targets > 1) {
		DMWARN("Request-based dm doesn't support multiple targets yet");
		return -EINVAL;
	}

K
Kiyoshi Ueda 已提交
884 885 886
	/* Non-request-stackable devices can't be used for request-based dm */
	devices = dm_table_get_devices(t);
	list_for_each_entry(dd, devices, list) {
887 888 889 890 891
		struct request_queue *q = bdev_get_queue(dd->dm_dev->bdev);

		if (!blk_queue_stackable(q)) {
			DMERR("table load rejected: including"
			      " non-request-stackable devices");
K
Kiyoshi Ueda 已提交
892 893
			return -EINVAL;
		}
894 895 896 897 898 899 900 901 902 903 904 905 906

		if (q->mq_ops)
			use_blk_mq = true;
	}

	if (use_blk_mq) {
		/* verify _all_ devices in the table are blk-mq devices */
		list_for_each_entry(dd, devices, list)
			if (!bdev_get_queue(dd->dm_dev->bdev)->mq_ops) {
				DMERR("table load rejected: not all devices"
				      " are blk-mq request-stackable");
				return -EINVAL;
			}
907
		t->type = DM_TYPE_MQ_REQUEST_BASED;
K
Kiyoshi Ueda 已提交
908

909 910 911
	} else if (hybrid && list_empty(devices) && live_md_type != DM_TYPE_NONE) {
		/* inherit live MD type */
		t->type = live_md_type;
K
Kiyoshi Ueda 已提交
912

913 914
	} else
		t->type = DM_TYPE_REQUEST_BASED;
K
Kiyoshi Ueda 已提交
915 916 917 918 919 920 921 922 923

	return 0;
}

unsigned dm_table_get_type(struct dm_table *t)
{
	return t->type;
}

924 925 926 927 928
struct target_type *dm_table_get_immutable_target_type(struct dm_table *t)
{
	return t->immutable_target_type;
}

K
Kiyoshi Ueda 已提交
929 930
bool dm_table_request_based(struct dm_table *t)
{
931 932 933 934 935 936 937 938 939
	unsigned table_type = dm_table_get_type(t);

	return (table_type == DM_TYPE_REQUEST_BASED ||
		table_type == DM_TYPE_MQ_REQUEST_BASED);
}

bool dm_table_mq_request_based(struct dm_table *t)
{
	return dm_table_get_type(t) == DM_TYPE_MQ_REQUEST_BASED;
K
Kiyoshi Ueda 已提交
940 941
}

942
static int dm_table_alloc_md_mempools(struct dm_table *t)
K
Kiyoshi Ueda 已提交
943 944
{
	unsigned type = dm_table_get_type(t);
M
Mikulas Patocka 已提交
945 946 947
	unsigned per_bio_data_size = 0;
	struct dm_target *tgt;
	unsigned i;
K
Kiyoshi Ueda 已提交
948 949 950 951 952 953

	if (unlikely(type == DM_TYPE_NONE)) {
		DMWARN("no table type is set, can't allocate mempools");
		return -EINVAL;
	}

M
Mikulas Patocka 已提交
954 955 956 957 958 959 960
	if (type == DM_TYPE_BIO_BASED)
		for (i = 0; i < t->num_targets; i++) {
			tgt = t->targets + i;
			per_bio_data_size = max(per_bio_data_size, tgt->per_bio_data_size);
		}

	t->mempools = dm_alloc_md_mempools(type, t->integrity_supported, per_bio_data_size);
K
Kiyoshi Ueda 已提交
961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977
	if (!t->mempools)
		return -ENOMEM;

	return 0;
}

void dm_table_free_md_mempools(struct dm_table *t)
{
	dm_free_md_mempools(t->mempools);
	t->mempools = NULL;
}

struct dm_md_mempools *dm_table_get_md_mempools(struct dm_table *t)
{
	return t->mempools;
}

L
Linus Torvalds 已提交
978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994
static int setup_indexes(struct dm_table *t)
{
	int i;
	unsigned int total = 0;
	sector_t *indexes;

	/* allocate the space for *all* the indexes */
	for (i = t->depth - 2; i >= 0; i--) {
		t->counts[i] = dm_div_up(t->counts[i + 1], CHILDREN_PER_NODE);
		total += t->counts[i];
	}

	indexes = (sector_t *) dm_vcalloc(total, (unsigned long) NODE_SIZE);
	if (!indexes)
		return -ENOMEM;

	/* set up internal nodes, bottom-up */
J
Jun'ichi Nomura 已提交
995
	for (i = t->depth - 2; i >= 0; i--) {
L
Linus Torvalds 已提交
996 997 998 999 1000 1001 1002 1003 1004 1005 1006
		t->index[i] = indexes;
		indexes += (KEYS_PER_NODE * t->counts[i]);
		setup_btree_index(i, t);
	}

	return 0;
}

/*
 * Builds the btree to index the map.
 */
1007
static int dm_table_build_index(struct dm_table *t)
L
Linus Torvalds 已提交
1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025
{
	int r = 0;
	unsigned int leaf_nodes;

	/* how many indexes will the btree have ? */
	leaf_nodes = dm_div_up(t->num_targets, KEYS_PER_NODE);
	t->depth = 1 + int_log(leaf_nodes, CHILDREN_PER_NODE);

	/* leaf layer has already been set up */
	t->counts[t->depth - 1] = leaf_nodes;
	t->index[t->depth - 1] = t->highs;

	if (t->depth >= 2)
		r = setup_indexes(t);

	return r;
}

1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040
/*
 * Get a disk whose integrity profile reflects the table's profile.
 * If %match_all is true, all devices' profiles must match.
 * If %match_all is false, all devices must at least have an
 * allocated integrity profile; but uninitialized is ok.
 * Returns NULL if integrity support was inconsistent or unavailable.
 */
static struct gendisk * dm_table_get_integrity_disk(struct dm_table *t,
						    bool match_all)
{
	struct list_head *devices = dm_table_get_devices(t);
	struct dm_dev_internal *dd = NULL;
	struct gendisk *prev_disk = NULL, *template_disk = NULL;

	list_for_each_entry(dd, devices, list) {
1041
		template_disk = dd->dm_dev->bdev->bd_disk;
1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
		if (!blk_get_integrity(template_disk))
			goto no_integrity;
		if (!match_all && !blk_integrity_is_initialized(template_disk))
			continue; /* skip uninitialized profiles */
		else if (prev_disk &&
			 blk_integrity_compare(prev_disk, template_disk) < 0)
			goto no_integrity;
		prev_disk = template_disk;
	}

	return template_disk;

no_integrity:
	if (prev_disk)
		DMWARN("%s: integrity not set: %s and %s profile mismatch",
		       dm_device_name(t->md),
		       prev_disk->disk_name,
		       template_disk->disk_name);
	return NULL;
}

1063 1064
/*
 * Register the mapped device for blk_integrity support if
1065 1066 1067 1068 1069 1070 1071
 * the underlying devices have an integrity profile.  But all devices
 * may not have matching profiles (checking all devices isn't reliable
 * during table load because this table may use other DM device(s) which
 * must be resumed before they will have an initialized integity profile).
 * Stacked DM devices force a 2 stage integrity profile validation:
 * 1 - during load, validate all initialized integrity profiles match
 * 2 - during resume, validate all integrity profiles match
1072 1073 1074
 */
static int dm_table_prealloc_integrity(struct dm_table *t, struct mapped_device *md)
{
1075
	struct gendisk *template_disk = NULL;
1076

1077 1078 1079
	template_disk = dm_table_get_integrity_disk(t, false);
	if (!template_disk)
		return 0;
1080

1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100
	if (!blk_integrity_is_initialized(dm_disk(md))) {
		t->integrity_supported = 1;
		return blk_integrity_register(dm_disk(md), NULL);
	}

	/*
	 * If DM device already has an initalized integrity
	 * profile the new profile should not conflict.
	 */
	if (blk_integrity_is_initialized(template_disk) &&
	    blk_integrity_compare(dm_disk(md), template_disk) < 0) {
		DMWARN("%s: conflict with existing integrity profile: "
		       "%s profile mismatch",
		       dm_device_name(t->md),
		       template_disk->disk_name);
		return 1;
	}

	/* Preserve existing initialized integrity profile */
	t->integrity_supported = 1;
1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136
	return 0;
}

/*
 * Prepares the table for use by building the indices,
 * setting the type, and allocating mempools.
 */
int dm_table_complete(struct dm_table *t)
{
	int r;

	r = dm_table_set_type(t);
	if (r) {
		DMERR("unable to set table type");
		return r;
	}

	r = dm_table_build_index(t);
	if (r) {
		DMERR("unable to build btrees");
		return r;
	}

	r = dm_table_prealloc_integrity(t, t->md);
	if (r) {
		DMERR("could not register integrity profile.");
		return r;
	}

	r = dm_table_alloc_md_mempools(t);
	if (r)
		DMERR("unable to allocate mempools");

	return r;
}

A
Arjan van de Ven 已提交
1137
static DEFINE_MUTEX(_event_lock);
L
Linus Torvalds 已提交
1138 1139 1140
void dm_table_event_callback(struct dm_table *t,
			     void (*fn)(void *), void *context)
{
A
Arjan van de Ven 已提交
1141
	mutex_lock(&_event_lock);
L
Linus Torvalds 已提交
1142 1143
	t->event_fn = fn;
	t->event_context = context;
A
Arjan van de Ven 已提交
1144
	mutex_unlock(&_event_lock);
L
Linus Torvalds 已提交
1145 1146 1147 1148 1149 1150 1151 1152 1153 1154
}

void dm_table_event(struct dm_table *t)
{
	/*
	 * You can no longer call dm_table_event() from interrupt
	 * context, use a bottom half instead.
	 */
	BUG_ON(in_interrupt());

A
Arjan van de Ven 已提交
1155
	mutex_lock(&_event_lock);
L
Linus Torvalds 已提交
1156 1157
	if (t->event_fn)
		t->event_fn(t->event_context);
A
Arjan van de Ven 已提交
1158
	mutex_unlock(&_event_lock);
L
Linus Torvalds 已提交
1159
}
1160
EXPORT_SYMBOL(dm_table_event);
L
Linus Torvalds 已提交
1161 1162 1163 1164 1165

sector_t dm_table_get_size(struct dm_table *t)
{
	return t->num_targets ? (t->highs[t->num_targets - 1] + 1) : 0;
}
1166
EXPORT_SYMBOL(dm_table_get_size);
L
Linus Torvalds 已提交
1167 1168 1169

struct dm_target *dm_table_get_target(struct dm_table *t, unsigned int index)
{
1170
	if (index >= t->num_targets)
L
Linus Torvalds 已提交
1171 1172 1173 1174 1175 1176 1177
		return NULL;

	return t->targets + index;
}

/*
 * Search the btree for the correct target.
1178 1179 1180
 *
 * Caller should check returned pointer with dm_target_is_valid()
 * to trap I/O beyond end of device.
L
Linus Torvalds 已提交
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198
 */
struct dm_target *dm_table_find_target(struct dm_table *t, sector_t sector)
{
	unsigned int l, n = 0, k = 0;
	sector_t *node;

	for (l = 0; l < t->depth; l++) {
		n = get_child(n, k);
		node = get_node(t, l, n);

		for (k = 0; k < KEYS_PER_NODE; k++)
			if (node[k] >= sector)
				break;
	}

	return &t->targets[(KEYS_PER_NODE * n) + k];
}

1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233
static int count_device(struct dm_target *ti, struct dm_dev *dev,
			sector_t start, sector_t len, void *data)
{
	unsigned *num_devices = data;

	(*num_devices)++;

	return 0;
}

/*
 * Check whether a table has no data devices attached using each
 * target's iterate_devices method.
 * Returns false if the result is unknown because a target doesn't
 * support iterate_devices.
 */
bool dm_table_has_no_data_devices(struct dm_table *table)
{
	struct dm_target *uninitialized_var(ti);
	unsigned i = 0, num_devices = 0;

	while (i < dm_table_get_num_targets(table)) {
		ti = dm_table_get_target(table, i++);

		if (!ti->type->iterate_devices)
			return false;

		ti->type->iterate_devices(ti, count_device, &num_devices);
		if (num_devices)
			return false;
	}

	return true;
}

1234 1235 1236 1237 1238 1239 1240 1241 1242 1243
/*
 * Establish the new table's queue_limits and validate them.
 */
int dm_calculate_queue_limits(struct dm_table *table,
			      struct queue_limits *limits)
{
	struct dm_target *uninitialized_var(ti);
	struct queue_limits ti_limits;
	unsigned i = 0;

1244
	blk_set_stacking_limits(limits);
1245 1246

	while (i < dm_table_get_num_targets(table)) {
1247
		blk_set_stacking_limits(&ti_limits);
1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259

		ti = dm_table_get_target(table, i++);

		if (!ti->type->iterate_devices)
			goto combine_limits;

		/*
		 * Combine queue limits of all the devices this target uses.
		 */
		ti->type->iterate_devices(ti, dm_set_device_limits,
					  &ti_limits);

1260 1261 1262 1263
		/* Set I/O hints portion of queue limits */
		if (ti->type->io_hints)
			ti->type->io_hints(ti, &ti_limits);

1264 1265 1266 1267
		/*
		 * Check each device area is consistent with the target's
		 * overall queue limits.
		 */
1268 1269
		if (ti->type->iterate_devices(ti, device_area_is_invalid,
					      &ti_limits))
1270 1271 1272 1273 1274 1275 1276 1277
			return -EINVAL;

combine_limits:
		/*
		 * Merge this target's queue limits into the overall limits
		 * for the table.
		 */
		if (blk_stack_limits(limits, &ti_limits, 0) < 0)
1278
			DMWARN("%s: adding target device "
1279
			       "(start sect %llu len %llu) "
1280
			       "caused an alignment inconsistency",
1281 1282 1283 1284 1285 1286 1287 1288
			       dm_device_name(table->md),
			       (unsigned long long) ti->begin,
			       (unsigned long long) ti->len);
	}

	return validate_hardware_logical_block_alignment(table, limits);
}

M
Martin K. Petersen 已提交
1289 1290
/*
 * Set the integrity profile for this device if all devices used have
1291 1292 1293 1294
 * matching profiles.  We're quite deep in the resume path but still
 * don't know if all devices (particularly DM devices this device
 * may be stacked on) have matching profiles.  Even if the profiles
 * don't match we have no way to fail (to resume) at this point.
M
Martin K. Petersen 已提交
1295 1296 1297
 */
static void dm_table_set_integrity(struct dm_table *t)
{
1298
	struct gendisk *template_disk = NULL;
M
Martin K. Petersen 已提交
1299 1300 1301 1302

	if (!blk_get_integrity(dm_disk(t->md)))
		return;

1303
	template_disk = dm_table_get_integrity_disk(t, true);
1304 1305 1306 1307
	if (template_disk)
		blk_integrity_register(dm_disk(t->md),
				       blk_get_integrity(template_disk));
	else if (blk_integrity_is_initialized(dm_disk(t->md)))
1308 1309
		DMWARN("%s: device no longer has a valid integrity profile",
		       dm_device_name(t->md));
1310 1311 1312
	else
		DMWARN("%s: unable to establish an integrity profile",
		       dm_device_name(t->md));
M
Martin K. Petersen 已提交
1313 1314
}

1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337
static int device_flush_capable(struct dm_target *ti, struct dm_dev *dev,
				sector_t start, sector_t len, void *data)
{
	unsigned flush = (*(unsigned *)data);
	struct request_queue *q = bdev_get_queue(dev->bdev);

	return q && (q->flush_flags & flush);
}

static bool dm_table_supports_flush(struct dm_table *t, unsigned flush)
{
	struct dm_target *ti;
	unsigned i = 0;

	/*
	 * Require at least one underlying device to support flushes.
	 * t->devices includes internal dm devices such as mirror logs
	 * so we need to use iterate_devices here, which targets
	 * supporting flushes must provide.
	 */
	while (i < dm_table_get_num_targets(t)) {
		ti = dm_table_get_target(t, i++);

1338
		if (!ti->num_flush_bios)
1339 1340
			continue;

1341 1342 1343
		if (ti->flush_supported)
			return 1;

1344 1345 1346 1347 1348 1349 1350 1351
		if (ti->type->iterate_devices &&
		    ti->type->iterate_devices(ti, device_flush_capable, &flush))
			return 1;
	}

	return 0;
}

1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367
static bool dm_table_discard_zeroes_data(struct dm_table *t)
{
	struct dm_target *ti;
	unsigned i = 0;

	/* Ensure that all targets supports discard_zeroes_data. */
	while (i < dm_table_get_num_targets(t)) {
		ti = dm_table_get_target(t, i++);

		if (ti->discard_zeroes_data_unsupported)
			return 0;
	}

	return 1;
}

1368 1369 1370 1371 1372 1373 1374 1375
static int device_is_nonrot(struct dm_target *ti, struct dm_dev *dev,
			    sector_t start, sector_t len, void *data)
{
	struct request_queue *q = bdev_get_queue(dev->bdev);

	return q && blk_queue_nonrot(q);
}

1376 1377 1378 1379 1380 1381 1382 1383
static int device_is_not_random(struct dm_target *ti, struct dm_dev *dev,
			     sector_t start, sector_t len, void *data)
{
	struct request_queue *q = bdev_get_queue(dev->bdev);

	return q && !blk_queue_add_random(q);
}

1384 1385 1386 1387 1388 1389 1390 1391
static int queue_supports_sg_merge(struct dm_target *ti, struct dm_dev *dev,
				   sector_t start, sector_t len, void *data)
{
	struct request_queue *q = bdev_get_queue(dev->bdev);

	return q && !test_bit(QUEUE_FLAG_NO_SG_MERGE, &q->queue_flags);
}

1392 1393
static bool dm_table_all_devices_attribute(struct dm_table *t,
					   iterate_devices_callout_fn func)
1394 1395 1396 1397 1398 1399 1400 1401
{
	struct dm_target *ti;
	unsigned i = 0;

	while (i < dm_table_get_num_targets(t)) {
		ti = dm_table_get_target(t, i++);

		if (!ti->type->iterate_devices ||
1402
		    !ti->type->iterate_devices(ti, func, NULL))
1403 1404 1405 1406 1407 1408
			return 0;
	}

	return 1;
}

M
Mike Snitzer 已提交
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424
static int device_not_write_same_capable(struct dm_target *ti, struct dm_dev *dev,
					 sector_t start, sector_t len, void *data)
{
	struct request_queue *q = bdev_get_queue(dev->bdev);

	return q && !q->limits.max_write_same_sectors;
}

static bool dm_table_supports_write_same(struct dm_table *t)
{
	struct dm_target *ti;
	unsigned i = 0;

	while (i < dm_table_get_num_targets(t)) {
		ti = dm_table_get_target(t, i++);

1425
		if (!ti->num_write_same_bios)
M
Mike Snitzer 已提交
1426 1427 1428
			return false;

		if (!ti->type->iterate_devices ||
M
Mike Snitzer 已提交
1429
		    ti->type->iterate_devices(ti, device_not_write_same_capable, NULL))
M
Mike Snitzer 已提交
1430 1431 1432 1433 1434 1435
			return false;
	}

	return true;
}

1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472
static int device_discard_capable(struct dm_target *ti, struct dm_dev *dev,
				  sector_t start, sector_t len, void *data)
{
	struct request_queue *q = bdev_get_queue(dev->bdev);

	return q && blk_queue_discard(q);
}

static bool dm_table_supports_discards(struct dm_table *t)
{
	struct dm_target *ti;
	unsigned i = 0;

	/*
	 * Unless any target used by the table set discards_supported,
	 * require at least one underlying device to support discards.
	 * t->devices includes internal dm devices such as mirror logs
	 * so we need to use iterate_devices here, which targets
	 * supporting discard selectively must provide.
	 */
	while (i < dm_table_get_num_targets(t)) {
		ti = dm_table_get_target(t, i++);

		if (!ti->num_discard_bios)
			continue;

		if (ti->discards_supported)
			return 1;

		if (ti->type->iterate_devices &&
		    ti->type->iterate_devices(ti, device_discard_capable, NULL))
			return 1;
	}

	return 0;
}

1473 1474
void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
			       struct queue_limits *limits)
L
Linus Torvalds 已提交
1475
{
1476 1477
	unsigned flush = 0;

L
Linus Torvalds 已提交
1478
	/*
1479
	 * Copy table's limits to the DM device's request_queue
L
Linus Torvalds 已提交
1480
	 */
1481
	q->limits = *limits;
1482

M
Mike Snitzer 已提交
1483 1484 1485 1486 1487
	if (!dm_table_supports_discards(t))
		queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, q);
	else
		queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, q);

1488 1489 1490 1491 1492 1493 1494
	if (dm_table_supports_flush(t, REQ_FLUSH)) {
		flush |= REQ_FLUSH;
		if (dm_table_supports_flush(t, REQ_FUA))
			flush |= REQ_FUA;
	}
	blk_queue_flush(q, flush);

1495 1496 1497
	if (!dm_table_discard_zeroes_data(t))
		q->limits.discard_zeroes_data = 0;

1498 1499
	/* Ensure that all underlying devices are non-rotational. */
	if (dm_table_all_devices_attribute(t, device_is_nonrot))
1500 1501 1502 1503
		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, q);
	else
		queue_flag_clear_unlocked(QUEUE_FLAG_NONROT, q);

M
Mike Snitzer 已提交
1504 1505
	if (!dm_table_supports_write_same(t))
		q->limits.max_write_same_sectors = 0;
M
Mike Snitzer 已提交
1506

1507 1508 1509 1510 1511
	if (dm_table_all_devices_attribute(t, queue_supports_sg_merge))
		queue_flag_clear_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
	else
		queue_flag_set_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);

M
Martin K. Petersen 已提交
1512
	dm_table_set_integrity(t);
K
Kiyoshi Ueda 已提交
1513

1514 1515 1516 1517 1518 1519 1520 1521 1522
	/*
	 * Determine whether or not this queue's I/O timings contribute
	 * to the entropy pool, Only request-based targets use this.
	 * Clear QUEUE_FLAG_ADD_RANDOM if any underlying device does not
	 * have it set.
	 */
	if (blk_queue_add_random(q) && dm_table_all_devices_attribute(t, device_is_not_random))
		queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, q);

K
Kiyoshi Ueda 已提交
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534
	/*
	 * QUEUE_FLAG_STACKABLE must be set after all queue settings are
	 * visible to other CPUs because, once the flag is set, incoming bios
	 * are processed by request-based dm, which refers to the queue
	 * settings.
	 * Until the flag set, bios are passed to bio-based dm and queued to
	 * md->deferred where queue settings are not needed yet.
	 * Those bios are passed to request-based dm at the resume time.
	 */
	smp_mb();
	if (dm_table_request_based(t))
		queue_flag_set_unlocked(QUEUE_FLAG_STACKABLE, q);
L
Linus Torvalds 已提交
1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546
}

unsigned int dm_table_get_num_targets(struct dm_table *t)
{
	return t->num_targets;
}

struct list_head *dm_table_get_devices(struct dm_table *t)
{
	return &t->devices;
}

1547
fmode_t dm_table_get_mode(struct dm_table *t)
L
Linus Torvalds 已提交
1548 1549 1550
{
	return t->mode;
}
1551
EXPORT_SYMBOL(dm_table_get_mode);
L
Linus Torvalds 已提交
1552

1553 1554 1555 1556 1557 1558 1559
enum suspend_mode {
	PRESUSPEND,
	PRESUSPEND_UNDO,
	POSTSUSPEND,
};

static void suspend_targets(struct dm_table *t, enum suspend_mode mode)
L
Linus Torvalds 已提交
1560 1561 1562 1563 1564
{
	int i = t->num_targets;
	struct dm_target *ti = t->targets;

	while (i--) {
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574
		switch (mode) {
		case PRESUSPEND:
			if (ti->type->presuspend)
				ti->type->presuspend(ti);
			break;
		case PRESUSPEND_UNDO:
			if (ti->type->presuspend_undo)
				ti->type->presuspend_undo(ti);
			break;
		case POSTSUSPEND:
L
Linus Torvalds 已提交
1575 1576
			if (ti->type->postsuspend)
				ti->type->postsuspend(ti);
1577 1578
			break;
		}
L
Linus Torvalds 已提交
1579 1580 1581 1582 1583 1584
		ti++;
	}
}

void dm_table_presuspend_targets(struct dm_table *t)
{
1585 1586 1587
	if (!t)
		return;

1588 1589 1590 1591 1592 1593 1594 1595 1596
	suspend_targets(t, PRESUSPEND);
}

void dm_table_presuspend_undo_targets(struct dm_table *t)
{
	if (!t)
		return;

	suspend_targets(t, PRESUSPEND_UNDO);
L
Linus Torvalds 已提交
1597 1598 1599 1600
}

void dm_table_postsuspend_targets(struct dm_table *t)
{
1601 1602 1603
	if (!t)
		return;

1604
	suspend_targets(t, POSTSUSPEND);
L
Linus Torvalds 已提交
1605 1606
}

1607
int dm_table_resume_targets(struct dm_table *t)
L
Linus Torvalds 已提交
1608
{
1609 1610 1611 1612 1613 1614 1615 1616 1617
	int i, r = 0;

	for (i = 0; i < t->num_targets; i++) {
		struct dm_target *ti = t->targets + i;

		if (!ti->type->preresume)
			continue;

		r = ti->type->preresume(ti);
1618 1619 1620
		if (r) {
			DMERR("%s: %s: preresume failed, error = %d",
			      dm_device_name(t->md), ti->type->name, r);
1621
			return r;
1622
		}
1623
	}
L
Linus Torvalds 已提交
1624 1625 1626 1627 1628 1629 1630

	for (i = 0; i < t->num_targets; i++) {
		struct dm_target *ti = t->targets + i;

		if (ti->type->resume)
			ti->type->resume(ti);
	}
1631 1632

	return 0;
L
Linus Torvalds 已提交
1633 1634
}

1635 1636 1637 1638 1639 1640
void dm_table_add_target_callbacks(struct dm_table *t, struct dm_target_callbacks *cb)
{
	list_add(&cb->list, &t->target_callbacks);
}
EXPORT_SYMBOL_GPL(dm_table_add_target_callbacks);

L
Linus Torvalds 已提交
1641 1642
int dm_table_any_congested(struct dm_table *t, int bdi_bits)
{
M
Mikulas Patocka 已提交
1643
	struct dm_dev_internal *dd;
P
Paul Jimenez 已提交
1644
	struct list_head *devices = dm_table_get_devices(t);
1645
	struct dm_target_callbacks *cb;
L
Linus Torvalds 已提交
1646 1647
	int r = 0;

P
Paul Jimenez 已提交
1648
	list_for_each_entry(dd, devices, list) {
1649
		struct request_queue *q = bdev_get_queue(dd->dm_dev->bdev);
A
Alasdair G Kergon 已提交
1650 1651 1652 1653 1654 1655 1656
		char b[BDEVNAME_SIZE];

		if (likely(q))
			r |= bdi_congested(&q->backing_dev_info, bdi_bits);
		else
			DMWARN_LIMIT("%s: any_congested: nonexistent device %s",
				     dm_device_name(t->md),
1657
				     bdevname(dd->dm_dev->bdev, b));
L
Linus Torvalds 已提交
1658 1659
	}

1660 1661 1662 1663
	list_for_each_entry(cb, &t->target_callbacks, list)
		if (cb->congested_fn)
			r |= cb->congested_fn(cb, bdi_bits);

L
Linus Torvalds 已提交
1664 1665 1666
	return r;
}

1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680
int dm_table_any_busy_target(struct dm_table *t)
{
	unsigned i;
	struct dm_target *ti;

	for (i = 0; i < t->num_targets; i++) {
		ti = t->targets + i;
		if (ti->type->busy && ti->type->busy(ti))
			return 1;
	}

	return 0;
}

M
Mike Anderson 已提交
1681 1682 1683 1684
struct mapped_device *dm_table_get_md(struct dm_table *t)
{
	return t->md;
}
1685
EXPORT_SYMBOL(dm_table_get_md);
M
Mike Anderson 已提交
1686

1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705
void dm_table_run_md_queue_async(struct dm_table *t)
{
	struct mapped_device *md;
	struct request_queue *queue;
	unsigned long flags;

	if (!dm_table_request_based(t))
		return;

	md = dm_table_get_md(t);
	queue = dm_get_md_queue(md);
	if (queue) {
		spin_lock_irqsave(queue->queue_lock, flags);
		blk_run_queue_async(queue);
		spin_unlock_irqrestore(queue->queue_lock, flags);
	}
}
EXPORT_SYMBOL(dm_table_run_md_queue_async);