summaryrefslogtreecommitdiffstats
path: root/module/zfs/space_map.c
blob: 645a81387c780e1fdbfb3b5b213f0ec0c707dcb1 (plain)
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
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
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
170
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
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License (the "License").
 * You may not use this file except in compliance with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */
/*
 * Copyright 2009 Sun Microsystems, Inc.  All rights reserved.
 * Use is subject to license terms.
 */
/*
 * Copyright (c) 2013 by Delphix. All rights reserved.
 */

#include <sys/zfs_context.h>
#include <sys/spa.h>
#include <sys/dmu.h>
#include <sys/dmu_tx.h>
#include <sys/dnode.h>
#include <sys/dsl_pool.h>
#include <sys/zio.h>
#include <sys/space_map.h>
#include <sys/refcount.h>
#include <sys/zfeature.h>

/*
 * This value controls how the space map's block size is allowed to grow.
 * If the value is set to the same size as SPACE_MAP_INITIAL_BLOCKSIZE then
 * the space map block size will remain fixed. Setting this value to something
 * greater than SPACE_MAP_INITIAL_BLOCKSIZE will allow the space map to
 * increase its block size as needed. To maintain backwards compatibilty the
 * space map's block size must be a power of 2 and SPACE_MAP_INITIAL_BLOCKSIZE
 * or larger.
 */
int space_map_max_blksz = (1 << 12);

/*
 * Load the space map disk into the specified range tree. Segments of maptype
 * are added to the range tree, other segment types are removed.
 *
 * Note: space_map_load() will drop sm_lock across dmu_read() calls.
 * The caller must be OK with this.
 */
int
space_map_load(space_map_t *sm, range_tree_t *rt, maptype_t maptype)
{
	uint64_t *entry, *entry_map, *entry_map_end;
	uint64_t bufsize, size, offset, end, space;
	int error = 0;

	ASSERT(MUTEX_HELD(sm->sm_lock));

	end = space_map_length(sm);
	space = space_map_allocated(sm);

	VERIFY0(range_tree_space(rt));

	if (maptype == SM_FREE) {
		range_tree_add(rt, sm->sm_start, sm->sm_size);
		space = sm->sm_size - space;
	}

	bufsize = MAX(sm->sm_blksz, SPA_MINBLOCKSIZE);
	entry_map = zio_buf_alloc(bufsize);

	mutex_exit(sm->sm_lock);
	if (end > bufsize) {
		dmu_prefetch(sm->sm_os, space_map_object(sm), bufsize,
		    end - bufsize);
	}
	mutex_enter(sm->sm_lock);

	for (offset = 0; offset < end; offset += bufsize) {
		size = MIN(end - offset, bufsize);
		VERIFY(P2PHASE(size, sizeof (uint64_t)) == 0);
		VERIFY(size != 0);
		ASSERT3U(sm->sm_blksz, !=, 0);

		dprintf("object=%llu  offset=%llx  size=%llx\n",
		    space_map_object(sm), offset, size);

		mutex_exit(sm->sm_lock);
		error = dmu_read(sm->sm_os, space_map_object(sm), offset, size,
		    entry_map, DMU_READ_PREFETCH);
		mutex_enter(sm->sm_lock);
		if (error != 0)
			break;

		entry_map_end = entry_map + (size / sizeof (uint64_t));
		for (entry = entry_map; entry < entry_map_end; entry++) {
			uint64_t e = *entry;
			uint64_t offset, size;

			if (SM_DEBUG_DECODE(e))		/* Skip debug entries */
				continue;

			offset = (SM_OFFSET_DECODE(e) << sm->sm_shift) +
			    sm->sm_start;
			size = SM_RUN_DECODE(e) << sm->sm_shift;

			VERIFY0(P2PHASE(offset, 1ULL << sm->sm_shift));
			VERIFY0(P2PHASE(size, 1ULL << sm->sm_shift));
			VERIFY3U(offset, >=, sm->sm_start);
			VERIFY3U(offset + size, <=, sm->sm_start + sm->sm_size);
			if (SM_TYPE_DECODE(e) == maptype) {
				VERIFY3U(range_tree_space(rt) + size, <=,
				    sm->sm_size);
				range_tree_add(rt, offset, size);
			} else {
				range_tree_remove(rt, offset, size);
			}
		}
	}

	if (error == 0)
		VERIFY3U(range_tree_space(rt), ==, space);
	else
		range_tree_vacate(rt, NULL, NULL);

	zio_buf_free(entry_map, bufsize);
	return (error);
}

void
space_map_histogram_clear(space_map_t *sm)
{
	if (sm->sm_dbuf->db_size != sizeof (space_map_phys_t))
		return;

	bzero(sm->sm_phys->smp_histogram, sizeof (sm->sm_phys->smp_histogram));
}

boolean_t
space_map_histogram_verify(space_map_t *sm, range_tree_t *rt)
{
	int i;

	/*
	 * Verify that the in-core range tree does not have any
	 * ranges smaller than our sm_shift size.
	 */
	for (i = 0; i < sm->sm_shift; i++) {
		if (rt->rt_histogram[i] != 0)
			return (B_FALSE);
	}
	return (B_TRUE);
}

void
space_map_histogram_add(space_map_t *sm, range_tree_t *rt, dmu_tx_t *tx)
{
	int idx = 0;
	int i;

	ASSERT(MUTEX_HELD(rt->rt_lock));
	ASSERT(dmu_tx_is_syncing(tx));
	VERIFY3U(space_map_object(sm), !=, 0);

	if (sm->sm_dbuf->db_size != sizeof (space_map_phys_t))
		return;

	dmu_buf_will_dirty(sm->sm_dbuf, tx);

	ASSERT(space_map_histogram_verify(sm, rt));

	/*
	 * Transfer the content of the range tree histogram to the space
	 * map histogram. The space map histogram contains 32 buckets ranging
	 * between 2^sm_shift to 2^(32+sm_shift-1). The range tree,
	 * however, can represent ranges from 2^0 to 2^63. Since the space
	 * map only cares about allocatable blocks (minimum of sm_shift) we
	 * can safely ignore all ranges in the range tree smaller than sm_shift.
	 */
	for (i = sm->sm_shift; i < RANGE_TREE_HISTOGRAM_SIZE; i++) {

		/*
		 * Since the largest histogram bucket in the space map is
		 * 2^(32+sm_shift-1), we need to normalize the values in
		 * the range tree for any bucket larger than that size. For
		 * example given an sm_shift of 9, ranges larger than 2^40
		 * would get normalized as if they were 1TB ranges. Assume
		 * the range tree had a count of 5 in the 2^44 (16TB) bucket,
		 * the calculation below would normalize this to 5 * 2^4 (16).
		 */
		ASSERT3U(i, >=, idx + sm->sm_shift);
		sm->sm_phys->smp_histogram[idx] +=
		    rt->rt_histogram[i] << (i - idx - sm->sm_shift);

		/*
		 * Increment the space map's index as long as we haven't
		 * reached the maximum bucket size. Accumulate all ranges
		 * larger than the max bucket size into the last bucket.
		 */
		if (idx < SPACE_MAP_HISTOGRAM_SIZE(sm) - 1) {
			ASSERT3U(idx + sm->sm_shift, ==, i);
			idx++;
			ASSERT3U(idx, <, SPACE_MAP_HISTOGRAM_SIZE(sm));
		}
	}
}

uint64_t
space_map_entries(space_map_t *sm, range_tree_t *rt)
{
	avl_tree_t *t = &rt->rt_root;
	range_seg_t *rs;
	uint64_t size, entries;

	/*
	 * All space_maps always have a debug entry so account for it here.
	 */
	entries = 1;

	/*
	 * Traverse the range tree and calculate the number of space map
	 * entries that would be required to write out the range tree.
	 */
	for (rs = avl_first(t); rs != NULL; rs = AVL_NEXT(t, rs)) {
		size = (rs->rs_end - rs->rs_start) >> sm->sm_shift;
		entries += howmany(size, SM_RUN_MAX);
	}
	return (entries);
}

void
space_map_set_blocksize(space_map_t *sm, uint64_t size, dmu_tx_t *tx)
{
	uint32_t blksz;
	u_longlong_t blocks;

	ASSERT3U(sm->sm_blksz, !=, 0);
	ASSERT3U(space_map_object(sm), !=, 0);
	ASSERT(sm->sm_dbuf != NULL);
	VERIFY(ISP2(space_map_max_blksz));

	if (sm->sm_blksz >= space_map_max_blksz)
		return;

	/*
	 * The object contains more than one block so we can't adjust
	 * its size.
	 */
	if (sm->sm_phys->smp_objsize > sm->sm_blksz)
		return;

	if (size > sm->sm_blksz) {
		uint64_t newsz;

		/*
		 * Older software versions treat space map blocks as fixed
		 * entities. The DMU is capable of handling different block
		 * sizes making it possible for us to increase the
		 * block size and maintain backwards compatibility. The
		 * caveat is that the new block sizes must be a
		 * power of 2 so that old software can append to the file,
		 * adding more blocks. The block size can grow until it
		 * reaches space_map_max_blksz.
		 */
		newsz = ISP2(size) ? size : 1ULL << highbit(size);
		if (newsz > space_map_max_blksz)
			newsz = space_map_max_blksz;

		VERIFY0(dmu_object_set_blocksize(sm->sm_os,
		    space_map_object(sm), newsz, 0, tx));
		dmu_object_size_from_db(sm->sm_dbuf, &blksz, &blocks);

		zfs_dbgmsg("txg %llu, spa %s, increasing blksz from %d to %d",
		    dmu_tx_get_txg(tx), spa_name(dmu_objset_spa(sm->sm_os)),
		    sm->sm_blksz, blksz);

		VERIFY3U(newsz, ==, blksz);
		VERIFY3U(sm->sm_blksz, <, blksz);
		sm->sm_blksz = blksz;
	}
}

/*
 * Note: space_map_write() will drop sm_lock across dmu_write() calls.
 */
void
space_map_write(space_map_t *sm, range_tree_t *rt, maptype_t maptype,
    dmu_tx_t *tx)
{
	objset_t *os = sm->sm_os;
	spa_t *spa = dmu_objset_spa(os);
	avl_tree_t *t = &rt->rt_root;
	range_seg_t *rs;
	uint64_t size, total, rt_space, nodes;
	uint64_t *entry, *entry_map, *entry_map_end;
	uint64_t newsz, expected_entries, actual_entries = 1;

	ASSERT(MUTEX_HELD(rt->rt_lock));
	ASSERT(dsl_pool_sync_context(dmu_objset_pool(os)));
	VERIFY3U(space_map_object(sm), !=, 0);
	dmu_buf_will_dirty(sm->sm_dbuf, tx);

	/*
	 * This field is no longer necessary since the in-core space map
	 * now contains the object number but is maintained for backwards
	 * compatibility.
	 */
	sm->sm_phys->smp_object = sm->sm_object;

	if (range_tree_space(rt) == 0) {
		VERIFY3U(sm->sm_object, ==, sm->sm_phys->smp_object);
		return;
	}

	if (maptype == SM_ALLOC)
		sm->sm_phys->smp_alloc += range_tree_space(rt);
	else
		sm->sm_phys->smp_alloc -= range_tree_space(rt);

	expected_entries = space_map_entries(sm, rt);

	/*
	 * Calculate the new size for the space map on-disk and see if
	 * we can grow the block size to accommodate the new size.
	 */
	newsz = sm->sm_phys->smp_objsize + expected_entries * sizeof (uint64_t);
	space_map_set_blocksize(sm, newsz, tx);

	entry_map = zio_buf_alloc(sm->sm_blksz);
	entry_map_end = entry_map + (sm->sm_blksz / sizeof (uint64_t));
	entry = entry_map;

	*entry++ = SM_DEBUG_ENCODE(1) |
	    SM_DEBUG_ACTION_ENCODE(maptype) |
	    SM_DEBUG_SYNCPASS_ENCODE(spa_sync_pass(spa)) |
	    SM_DEBUG_TXG_ENCODE(dmu_tx_get_txg(tx));

	total = 0;
	nodes = avl_numnodes(&rt->rt_root);
	rt_space = range_tree_space(rt);
	for (rs = avl_first(t); rs != NULL; rs = AVL_NEXT(t, rs)) {
		uint64_t start;

		size = (rs->rs_end - rs->rs_start) >> sm->sm_shift;
		start = (rs->rs_start - sm->sm_start) >> sm->sm_shift;

		total += size << sm->sm_shift;

		while (size != 0) {
			uint64_t run_len;

			run_len = MIN(size, SM_RUN_MAX);

			if (entry == entry_map_end) {
				mutex_exit(rt->rt_lock);
				dmu_write(os, space_map_object(sm),
				    sm->sm_phys->smp_objsize, sm->sm_blksz,
				    entry_map, tx);
				mutex_enter(rt->rt_lock);
				sm->sm_phys->smp_objsize += sm->sm_blksz;
				entry = entry_map;
			}

			*entry++ = SM_OFFSET_ENCODE(start) |
			    SM_TYPE_ENCODE(maptype) |
			    SM_RUN_ENCODE(run_len);

			start += run_len;
			size -= run_len;
			actual_entries++;
		}
	}

	if (entry != entry_map) {
		size = (entry - entry_map) * sizeof (uint64_t);
		mutex_exit(rt->rt_lock);
		dmu_write(os, space_map_object(sm), sm->sm_phys->smp_objsize,
		    size, entry_map, tx);
		mutex_enter(rt->rt_lock);
		sm->sm_phys->smp_objsize += size;
	}
	ASSERT3U(expected_entries, ==, actual_entries);

	/*
	 * Ensure that the space_map's accounting wasn't changed
	 * while we were in the middle of writing it out.
	 */
	VERIFY3U(nodes, ==, avl_numnodes(&rt->rt_root));
	VERIFY3U(range_tree_space(rt), ==, rt_space);
	VERIFY3U(range_tree_space(rt), ==, total);

	zio_buf_free(entry_map, sm->sm_blksz);
}

static int
space_map_open_impl(space_map_t *sm)
{
	int error;
	u_longlong_t blocks;

	error = dmu_bonus_hold(sm->sm_os, sm->sm_object, sm, &sm->sm_dbuf);
	if (error)
		return (error);

	dmu_object_size_from_db(sm->sm_dbuf, &sm->sm_blksz, &blocks);
	sm->sm_phys = sm->sm_dbuf->db_data;
	return (0);
}

int
space_map_open(space_map_t **smp, objset_t *os, uint64_t object,
    uint64_t start, uint64_t size, uint8_t shift, kmutex_t *lp)
{
	space_map_t *sm;
	int error;

	ASSERT(*smp == NULL);
	ASSERT(os != NULL);
	ASSERT(object != 0);

	sm = kmem_alloc(sizeof (space_map_t), KM_PUSHPAGE);

	sm->sm_start = start;
	sm->sm_size = size;
	sm->sm_shift = shift;
	sm->sm_lock = lp;
	sm->sm_os = os;
	sm->sm_object = object;
	sm->sm_length = 0;
	sm->sm_alloc = 0;
	sm->sm_blksz = 0;
	sm->sm_dbuf = NULL;
	sm->sm_phys = NULL;

	error = space_map_open_impl(sm);
	if (error != 0) {
		space_map_close(sm);
		return (error);
	}

	*smp = sm;

	return (0);
}

void
space_map_close(space_map_t *sm)
{
	if (sm == NULL)
		return;

	if (sm->sm_dbuf != NULL)
		dmu_buf_rele(sm->sm_dbuf, sm);
	sm->sm_dbuf = NULL;
	sm->sm_phys = NULL;

	kmem_free(sm, sizeof (*sm));
}

static void
space_map_reallocate(space_map_t *sm, dmu_tx_t *tx)
{
	ASSERT(dmu_tx_is_syncing(tx));

	space_map_free(sm, tx);
	dmu_buf_rele(sm->sm_dbuf, sm);

	sm->sm_object = space_map_alloc(sm->sm_os, tx);
	VERIFY0(space_map_open_impl(sm));
}

void
space_map_truncate(space_map_t *sm, dmu_tx_t *tx)
{
	objset_t *os = sm->sm_os;
	spa_t *spa = dmu_objset_spa(os);
	dmu_object_info_t doi;
	int bonuslen;

	ASSERT(dsl_pool_sync_context(dmu_objset_pool(os)));
	ASSERT(dmu_tx_is_syncing(tx));

	VERIFY0(dmu_free_range(os, space_map_object(sm), 0, -1ULL, tx));
	dmu_object_info_from_db(sm->sm_dbuf, &doi);

	if (spa_feature_is_enabled(spa, SPA_FEATURE_SPACEMAP_HISTOGRAM)) {
		bonuslen = sizeof (space_map_phys_t);
		ASSERT3U(bonuslen, <=, dmu_bonus_max());
	} else {
		bonuslen = SPACE_MAP_SIZE_V0;
	}

	if (bonuslen != doi.doi_bonus_size ||
	    doi.doi_data_block_size != SPACE_MAP_INITIAL_BLOCKSIZE) {
		zfs_dbgmsg("txg %llu, spa %s, reallocating: "
		    "old bonus %u, old blocksz %u", dmu_tx_get_txg(tx),
		    spa_name(spa), doi.doi_bonus_size, doi.doi_data_block_size);
		space_map_reallocate(sm, tx);
		VERIFY3U(sm->sm_blksz, ==, SPACE_MAP_INITIAL_BLOCKSIZE);
	}

	dmu_buf_will_dirty(sm->sm_dbuf, tx);
	sm->sm_phys->smp_objsize = 0;
	sm->sm_phys->smp_alloc = 0;
}

/*
 * Update the in-core space_map allocation and length values.
 */
void
space_map_update(space_map_t *sm)
{
	if (sm == NULL)
		return;

	ASSERT(MUTEX_HELD(sm->sm_lock));

	sm->sm_alloc = sm->sm_phys->smp_alloc;
	sm->sm_length = sm->sm_phys->smp_objsize;
}

uint64_t
space_map_alloc(objset_t *os, dmu_tx_t *tx)
{
	spa_t *spa = dmu_objset_spa(os);
	uint64_t object;
	int bonuslen;

	if (spa_feature_is_enabled(spa, SPA_FEATURE_SPACEMAP_HISTOGRAM)) {
		spa_feature_incr(spa, SPA_FEATURE_SPACEMAP_HISTOGRAM, tx);
		bonuslen = sizeof (space_map_phys_t);
		ASSERT3U(bonuslen, <=, dmu_bonus_max());
	} else {
		bonuslen = SPACE_MAP_SIZE_V0;
	}

	object = dmu_object_alloc(os,
	    DMU_OT_SPACE_MAP, SPACE_MAP_INITIAL_BLOCKSIZE,
	    DMU_OT_SPACE_MAP_HEADER, bonuslen, tx);

	return (object);
}

void
space_map_free(space_map_t *sm, dmu_tx_t *tx)
{
	spa_t *spa;

	if (sm == NULL)
		return;

	spa = dmu_objset_spa(sm->sm_os);
	if (spa_feature_is_enabled(spa, SPA_FEATURE_SPACEMAP_HISTOGRAM)) {
		dmu_object_info_t doi;

		dmu_object_info_from_db(sm->sm_dbuf, &doi);
		if (doi.doi_bonus_size != SPACE_MAP_SIZE_V0) {
			VERIFY(spa_feature_is_active(spa,
			    SPA_FEATURE_SPACEMAP_HISTOGRAM));
			spa_feature_decr(spa,
			    SPA_FEATURE_SPACEMAP_HISTOGRAM, tx);
		}
	}

	VERIFY3U(dmu_object_free(sm->sm_os, space_map_object(sm), tx), ==, 0);
	sm->sm_object = 0;
}

uint64_t
space_map_object(space_map_t *sm)
{
	return (sm != NULL ? sm->sm_object : 0);
}

/*
 * Returns the already synced, on-disk allocated space.
 */
uint64_t
space_map_allocated(space_map_t *sm)
{
	return (sm != NULL ? sm->sm_alloc : 0);
}

/*
 * Returns the already synced, on-disk length;
 */
uint64_t
space_map_length(space_map_t *sm)
{
	return (sm != NULL ? sm->sm_length : 0);
}

/*
 * Returns the allocated space that is currently syncing.
 */
int64_t
space_map_alloc_delta(space_map_t *sm)
{
	if (sm == NULL)
		return (0);
	ASSERT(sm->sm_dbuf != NULL);
	return (sm->sm_phys->smp_alloc - space_map_allocated(sm));
}