summaryrefslogtreecommitdiff
path: root/drivers/md/dm-log-writes.c
blob: 9de072b7782a53d5192087cb36f775d3630440a4 (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
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
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
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
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
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
/*
 * Copyright (C) 2014 Facebook. All rights reserved.
 *
 * This file is released under the GPL.
 */

#include <linux/device-mapper.h>

#include <linux/module.h>
#include <linux/init.h>
#include <linux/blkdev.h>
#include <linux/bio.h>
#include <linux/dax.h>
#include <linux/slab.h>
#include <linux/kthread.h>
#include <linux/freezer.h>
#include <linux/uio.h>

#define DM_MSG_PREFIX "log-writes"

/*
 * This target will sequentially log all writes to the target device onto the
 * log device.  This is helpful for replaying writes to check for fs consistency
 * at all times.  This target provides a mechanism to mark specific events to
 * check data at a later time.  So for example you would:
 *
 * write data
 * fsync
 * dmsetup message /dev/whatever mark mymark
 * unmount /mnt/test
 *
 * Then replay the log up to mymark and check the contents of the replay to
 * verify it matches what was written.
 *
 * We log writes only after they have been flushed, this makes the log describe
 * close to the order in which the data hits the actual disk, not its cache.  So
 * for example the following sequence (W means write, C means complete)
 *
 * Wa,Wb,Wc,Cc,Ca,FLUSH,FUAd,Cb,CFLUSH,CFUAd
 *
 * Would result in the log looking like this:
 *
 * c,a,flush,fuad,b,<other writes>,<next flush>
 *
 * This is meant to help expose problems where file systems do not properly wait
 * on data being written before invoking a FLUSH.  FUA bypasses cache so once it
 * completes it is added to the log as it should be on disk.
 *
 * We treat DISCARDs as if they don't bypass cache so that they are logged in
 * order of completion along with the normal writes.  If we didn't do it this
 * way we would process all the discards first and then write all the data, when
 * in fact we want to do the data and the discard in the order that they
 * completed.
 */
#define LOG_FLUSH_FLAG		(1 << 0)
#define LOG_FUA_FLAG		(1 << 1)
#define LOG_DISCARD_FLAG	(1 << 2)
#define LOG_MARK_FLAG		(1 << 3)
#define LOG_METADATA_FLAG	(1 << 4)

#define WRITE_LOG_VERSION 1ULL
#define WRITE_LOG_MAGIC 0x6a736677736872ULL

/*
 * The disk format for this is braindead simple.
 *
 * At byte 0 we have our super, followed by the following sequence for
 * nr_entries:
 *
 * [   1 sector    ][  entry->nr_sectors ]
 * [log_write_entry][    data written    ]
 *
 * The log_write_entry takes up a full sector so we can have arbitrary length
 * marks and it leaves us room for extra content in the future.
 */

/*
 * Basic info about the log for userspace.
 */
struct log_write_super {
	__le64 magic;
	__le64 version;
	__le64 nr_entries;
	__le32 sectorsize;
};

/*
 * sector - the sector we wrote.
 * nr_sectors - the number of sectors we wrote.
 * flags - flags for this log entry.
 * data_len - the size of the data in this log entry, this is for private log
 * entry stuff, the MARK data provided by userspace for example.
 */
struct log_write_entry {
	__le64 sector;
	__le64 nr_sectors;
	__le64 flags;
	__le64 data_len;
};

struct log_writes_c {
	struct dm_dev *dev;
	struct dm_dev *logdev;
	u64 logged_entries;
	u32 sectorsize;
	u32 sectorshift;
	atomic_t io_blocks;
	atomic_t pending_blocks;
	sector_t next_sector;
	sector_t end_sector;
	bool logging_enabled;
	bool device_supports_discard;
	spinlock_t blocks_lock;
	struct list_head unflushed_blocks;
	struct list_head logging_blocks;
	wait_queue_head_t wait;
	struct task_struct *log_kthread;
};

struct pending_block {
	int vec_cnt;
	u64 flags;
	sector_t sector;
	sector_t nr_sectors;
	char *data;
	u32 datalen;
	struct list_head list;
	struct bio_vec vecs[0];
};

struct per_bio_data {
	struct pending_block *block;
};

static inline sector_t bio_to_dev_sectors(struct log_writes_c *lc,
					  sector_t sectors)
{
	return sectors >> (lc->sectorshift - SECTOR_SHIFT);
}

static inline sector_t dev_to_bio_sectors(struct log_writes_c *lc,
					  sector_t sectors)
{
	return sectors << (lc->sectorshift - SECTOR_SHIFT);
}

static void put_pending_block(struct log_writes_c *lc)
{
	if (atomic_dec_and_test(&lc->pending_blocks)) {
		smp_mb__after_atomic();
		if (waitqueue_active(&lc->wait))
			wake_up(&lc->wait);
	}
}

static void put_io_block(struct log_writes_c *lc)
{
	if (atomic_dec_and_test(&lc->io_blocks)) {
		smp_mb__after_atomic();
		if (waitqueue_active(&lc->wait))
			wake_up(&lc->wait);
	}
}

static void log_end_io(struct bio *bio)
{
	struct log_writes_c *lc = bio->bi_private;

	if (bio->bi_status) {
		unsigned long flags;

		DMERR("Error writing log block, error=%d", bio->bi_status);
		spin_lock_irqsave(&lc->blocks_lock, flags);
		lc->logging_enabled = false;
		spin_unlock_irqrestore(&lc->blocks_lock, flags);
	}

	bio_free_pages(bio);
	put_io_block(lc);
	bio_put(bio);
}

/*
 * Meant to be called if there is an error, it will free all the pages
 * associated with the block.
 */
static void free_pending_block(struct log_writes_c *lc,
			       struct pending_block *block)
{
	int i;

	for (i = 0; i < block->vec_cnt; i++) {
		if (block->vecs[i].bv_page)
			__free_page(block->vecs[i].bv_page);
	}
	kfree(block->data);
	kfree(block);
	put_pending_block(lc);
}

static int write_metadata(struct log_writes_c *lc, void *entry,
			  size_t entrylen, void *data, size_t datalen,
			  sector_t sector)
{
	struct bio *bio;
	struct page *page;
	void *ptr;
	size_t ret;

	bio = bio_alloc(GFP_KERNEL, 1);
	if (!bio) {
		DMERR("Couldn't alloc log bio");
		goto error;
	}
	bio->bi_iter.bi_size = 0;
	bio->bi_iter.bi_sector = sector;
	bio_set_dev(bio, lc->logdev->bdev);
	bio->bi_end_io = log_end_io;
	bio->bi_private = lc;
	bio_set_op_attrs(bio, REQ_OP_WRITE, 0);

	page = alloc_page(GFP_KERNEL);
	if (!page) {
		DMERR("Couldn't alloc log page");
		bio_put(bio);
		goto error;
	}

	ptr = kmap_atomic(page);
	memcpy(ptr, entry, entrylen);
	if (datalen)
		memcpy(ptr + entrylen, data, datalen);
	memset(ptr + entrylen + datalen, 0,
	       lc->sectorsize - entrylen - datalen);
	kunmap_atomic(ptr);

	ret = bio_add_page(bio, page, lc->sectorsize, 0);
	if (ret != lc->sectorsize) {
		DMERR("Couldn't add page to the log block");
		goto error_bio;
	}
	submit_bio(bio);
	return 0;
error_bio:
	bio_put(bio);
	__free_page(page);
error:
	put_io_block(lc);
	return -1;
}

static int write_inline_data(struct log_writes_c *lc, void *entry,
			     size_t entrylen, void *data, size_t datalen,
			     sector_t sector)
{
	int num_pages, bio_pages, pg_datalen, pg_sectorlen, i;
	struct page *page;
	struct bio *bio;
	size_t ret;
	void *ptr;

	while (datalen) {
		num_pages = ALIGN(datalen, PAGE_SIZE) >> PAGE_SHIFT;
		bio_pages = min(num_pages, BIO_MAX_PAGES);

		atomic_inc(&lc->io_blocks);

		bio = bio_alloc(GFP_KERNEL, bio_pages);
		if (!bio) {
			DMERR("Couldn't alloc inline data bio");
			goto error;
		}

		bio->bi_iter.bi_size = 0;
		bio->bi_iter.bi_sector = sector;
		bio_set_dev(bio, lc->logdev->bdev);
		bio->bi_end_io = log_end_io;
		bio->bi_private = lc;
		bio_set_op_attrs(bio, REQ_OP_WRITE, 0);

		for (i = 0; i < bio_pages; i++) {
			pg_datalen = min_t(int, datalen, PAGE_SIZE);
			pg_sectorlen = ALIGN(pg_datalen, lc->sectorsize);

			page = alloc_page(GFP_KERNEL);
			if (!page) {
				DMERR("Couldn't alloc inline data page");
				goto error_bio;
			}

			ptr = kmap_atomic(page);
			memcpy(ptr, data, pg_datalen);
			if (pg_sectorlen > pg_datalen)
				memset(ptr + pg_datalen, 0, pg_sectorlen - pg_datalen);
			kunmap_atomic(ptr);

			ret = bio_add_page(bio, page, pg_sectorlen, 0);
			if (ret != pg_sectorlen) {
				DMERR("Couldn't add page of inline data");
				__free_page(page);
				goto error_bio;
			}

			datalen -= pg_datalen;
			data	+= pg_datalen;
		}
		submit_bio(bio);

		sector += bio_pages * PAGE_SECTORS;
	}
	return 0;
error_bio:
	bio_free_pages(bio);
	bio_put(bio);
error:
	put_io_block(lc);
	return -1;
}

static int log_one_block(struct log_writes_c *lc,
			 struct pending_block *block, sector_t sector)
{
	struct bio *bio;
	struct log_write_entry entry;
	size_t metadatalen, ret;
	int i;

	entry.sector = cpu_to_le64(block->sector);
	entry.nr_sectors = cpu_to_le64(block->nr_sectors);
	entry.flags = cpu_to_le64(block->flags);
	entry.data_len = cpu_to_le64(block->datalen);

	metadatalen = (block->flags & LOG_MARK_FLAG) ? block->datalen : 0;
	if (write_metadata(lc, &entry, sizeof(entry), block->data,
			   metadatalen, sector)) {
		free_pending_block(lc, block);
		return -1;
	}

	sector += dev_to_bio_sectors(lc, 1);

	if (block->datalen && metadatalen == 0) {
		if (write_inline_data(lc, &entry, sizeof(entry), block->data,
				      block->datalen, sector)) {
			free_pending_block(lc, block);
			return -1;
		}
		/* we don't support both inline data & bio data */
		goto out;
	}

	if (!block->vec_cnt)
		goto out;

	atomic_inc(&lc->io_blocks);
	bio = bio_alloc(GFP_KERNEL, min(block->vec_cnt, BIO_MAX_PAGES));
	if (!bio) {
		DMERR("Couldn't alloc log bio");
		goto error;
	}
	bio->bi_iter.bi_size = 0;
	bio->bi_iter.bi_sector = sector;
	bio_set_dev(bio, lc->logdev->bdev);
	bio->bi_end_io = log_end_io;
	bio->bi_private = lc;
	bio_set_op_attrs(bio, REQ_OP_WRITE, 0);

	for (i = 0; i < block->vec_cnt; i++) {
		/*
		 * The page offset is always 0 because we allocate a new page
		 * for every bvec in the original bio for simplicity sake.
		 */
		ret = bio_add_page(bio, block->vecs[i].bv_page,
				   block->vecs[i].bv_len, 0);
		if (ret != block->vecs[i].bv_len) {
			atomic_inc(&lc->io_blocks);
			submit_bio(bio);
			bio = bio_alloc(GFP_KERNEL, min(block->vec_cnt - i, BIO_MAX_PAGES));
			if (!bio) {
				DMERR("Couldn't alloc log bio");
				goto error;
			}
			bio->bi_iter.bi_size = 0;
			bio->bi_iter.bi_sector = sector;
			bio_set_dev(bio, lc->logdev->bdev);
			bio->bi_end_io = log_end_io;
			bio->bi_private = lc;
			bio_set_op_attrs(bio, REQ_OP_WRITE, 0);

			ret = bio_add_page(bio, block->vecs[i].bv_page,
					   block->vecs[i].bv_len, 0);
			if (ret != block->vecs[i].bv_len) {
				DMERR("Couldn't add page on new bio?");
				bio_put(bio);
				goto error;
			}
		}
		sector += block->vecs[i].bv_len >> SECTOR_SHIFT;
	}
	submit_bio(bio);
out:
	kfree(block->data);
	kfree(block);
	put_pending_block(lc);
	return 0;
error:
	free_pending_block(lc, block);
	put_io_block(lc);
	return -1;
}

static int log_super(struct log_writes_c *lc)
{
	struct log_write_super super;

	super.magic = cpu_to_le64(WRITE_LOG_MAGIC);
	super.version = cpu_to_le64(WRITE_LOG_VERSION);
	super.nr_entries = cpu_to_le64(lc->logged_entries);
	super.sectorsize = cpu_to_le32(lc->sectorsize);

	if (write_metadata(lc, &super, sizeof(super), NULL, 0, 0)) {
		DMERR("Couldn't write super");
		return -1;
	}

	return 0;
}

static inline sector_t logdev_last_sector(struct log_writes_c *lc)
{
	return i_size_read(lc->logdev->bdev->bd_inode) >> SECTOR_SHIFT;
}

static int log_writes_kthread(void *arg)
{
	struct log_writes_c *lc = (struct log_writes_c *)arg;
	sector_t sector = 0;

	while (!kthread_should_stop()) {
		bool super = false;
		bool logging_enabled;
		struct pending_block *block = NULL;
		int ret;

		spin_lock_irq(&lc->blocks_lock);
		if (!list_empty(&lc->logging_blocks)) {
			block = list_first_entry(&lc->logging_blocks,
						 struct pending_block, list);
			list_del_init(&block->list);
			if (!lc->logging_enabled)
				goto next;

			sector = lc->next_sector;
			if (!(block->flags & LOG_DISCARD_FLAG))
				lc->next_sector += dev_to_bio_sectors(lc, block->nr_sectors);
			lc->next_sector += dev_to_bio_sectors(lc, 1);

			/*
			 * Apparently the size of the device may not be known
			 * right away, so handle this properly.
			 */
			if (!lc->end_sector)
				lc->end_sector = logdev_last_sector(lc);
			if (lc->end_sector &&
			    lc->next_sector >= lc->end_sector) {
				DMERR("Ran out of space on the logdev");
				lc->logging_enabled = false;
				goto next;
			}
			lc->logged_entries++;
			atomic_inc(&lc->io_blocks);

			super = (block->flags & (LOG_FUA_FLAG | LOG_MARK_FLAG));
			if (super)
				atomic_inc(&lc->io_blocks);
		}
next:
		logging_enabled = lc->logging_enabled;
		spin_unlock_irq(&lc->blocks_lock);
		if (block) {
			if (logging_enabled) {
				ret = log_one_block(lc, block, sector);
				if (!ret && super)
					ret = log_super(lc);
				if (ret) {
					spin_lock_irq(&lc->blocks_lock);
					lc->logging_enabled = false;
					spin_unlock_irq(&lc->blocks_lock);
				}
			} else
				free_pending_block(lc, block);
			continue;
		}

		if (!try_to_freeze()) {
			set_current_state(TASK_INTERRUPTIBLE);
			if (!kthread_should_stop() &&
			    list_empty(&lc->logging_blocks))
				schedule();
			__set_current_state(TASK_RUNNING);
		}
	}
	return 0;
}

/*
 * Construct a log-writes mapping:
 * log-writes <dev_path> <log_dev_path>
 */
static int log_writes_ctr(struct dm_target *ti, unsigned int argc, char **argv)
{
	struct log_writes_c *lc;
	struct dm_arg_set as;
	const char *devname, *logdevname;
	int ret;

	as.argc = argc;
	as.argv = argv;

	if (argc < 2) {
		ti->error = "Invalid argument count";
		return -EINVAL;
	}

	lc = kzalloc(sizeof(struct log_writes_c), GFP_KERNEL);
	if (!lc) {
		ti->error = "Cannot allocate context";
		return -ENOMEM;
	}
	spin_lock_init(&lc->blocks_lock);
	INIT_LIST_HEAD(&lc->unflushed_blocks);
	INIT_LIST_HEAD(&lc->logging_blocks);
	init_waitqueue_head(&lc->wait);
	atomic_set(&lc->io_blocks, 0);
	atomic_set(&lc->pending_blocks, 0);

	devname = dm_shift_arg(&as);
	ret = dm_get_device(ti, devname, dm_table_get_mode(ti->table), &lc->dev);
	if (ret) {
		ti->error = "Device lookup failed";
		goto bad;
	}

	logdevname = dm_shift_arg(&as);
	ret = dm_get_device(ti, logdevname, dm_table_get_mode(ti->table),
			    &lc->logdev);
	if (ret) {
		ti->error = "Log device lookup failed";
		dm_put_device(ti, lc->dev);
		goto bad;
	}

	lc->sectorsize = bdev_logical_block_size(lc->dev->bdev);
	lc->sectorshift = ilog2(lc->sectorsize);
	lc->log_kthread = kthread_run(log_writes_kthread, lc, "log-write");
	if (IS_ERR(lc->log_kthread)) {
		ret = PTR_ERR(lc->log_kthread);
		ti->error = "Couldn't alloc kthread";
		dm_put_device(ti, lc->dev);
		dm_put_device(ti, lc->logdev);
		goto bad;
	}

	/*
	 * next_sector is in 512b sectors to correspond to what bi_sector expects.
	 * The super starts at sector 0, and the next_sector is the next logical
	 * one based on the sectorsize of the device.
	 */
	lc->next_sector = lc->sectorsize >> SECTOR_SHIFT;
	lc->logging_enabled = true;
	lc->end_sector = logdev_last_sector(lc);
	lc->device_supports_discard = true;

	ti->num_flush_bios = 1;
	ti->flush_supported = true;
	ti->num_discard_bios = 1;
	ti->discards_supported = true;
	ti->per_io_data_size = sizeof(struct per_bio_data);
	ti->private = lc;
	return 0;

bad:
	kfree(lc);
	return ret;
}

static int log_mark(struct log_writes_c *lc, char *data)
{
	struct pending_block *block;
	size_t maxsize = lc->sectorsize - sizeof(struct log_write_entry);

	block = kzalloc(sizeof(struct pending_block), GFP_KERNEL);
	if (!block) {
		DMERR("Error allocating pending block");
		return -ENOMEM;
	}

	block->data = kstrndup(data, maxsize - 1, GFP_KERNEL);
	if (!block->data) {
		DMERR("Error copying mark data");
		kfree(block);
		return -ENOMEM;
	}
	atomic_inc(&lc->pending_blocks);
	block->datalen = strlen(block->data);
	block->flags |= LOG_MARK_FLAG;
	spin_lock_irq(&lc->blocks_lock);
	list_add_tail(&block->list, &lc->logging_blocks);
	spin_unlock_irq(&lc->blocks_lock);
	wake_up_process(lc->log_kthread);
	return 0;
}

static int log_dax(struct log_writes_c *lc, sector_t sector, size_t bytes,
		   struct iov_iter *i)
{
	struct pending_block *block;

	if (!bytes)
		return 0;

	block = kzalloc(sizeof(struct pending_block), GFP_KERNEL);
	if (!block) {
		DMERR("Error allocating dax pending block");
		return -ENOMEM;
	}

	block->data = kzalloc(bytes, GFP_KERNEL);
	if (!block->data) {
		DMERR("Error allocating dax data space");
		kfree(block);
		return -ENOMEM;
	}

	/* write data provided via the iterator */
	if (!copy_from_iter(block->data, bytes, i)) {
		DMERR("Error copying dax data");
		kfree(block->data);
		kfree(block);
		return -EIO;
	}

	/* rewind the iterator so that the block driver can use it */
	iov_iter_revert(i, bytes);

	block->datalen = bytes;
	block->sector = bio_to_dev_sectors(lc, sector);
	block->nr_sectors = ALIGN(bytes, lc->sectorsize) >> lc->sectorshift;

	atomic_inc(&lc->pending_blocks);
	spin_lock_irq(&lc->blocks_lock);
	list_add_tail(&block->list, &lc->unflushed_blocks);
	spin_unlock_irq(&lc->blocks_lock);
	wake_up_process(lc->log_kthread);

	return 0;
}

static void log_writes_dtr(struct dm_target *ti)
{
	struct log_writes_c *lc = ti->private;

	spin_lock_irq(&lc->blocks_lock);
	list_splice_init(&lc->unflushed_blocks, &lc->logging_blocks);
	spin_unlock_irq(&lc->blocks_lock);

	/*
	 * This is just nice to have since it'll update the super to include the
	 * unflushed blocks, if it fails we don't really care.
	 */
	log_mark(lc, "dm-log-writes-end");
	wake_up_process(lc->log_kthread);
	wait_event(lc->wait, !atomic_read(&lc->io_blocks) &&
		   !atomic_read(&lc->pending_blocks));
	kthread_stop(lc->log_kthread);

	WARN_ON(!list_empty(&lc->logging_blocks));
	WARN_ON(!list_empty(&lc->unflushed_blocks));
	dm_put_device(ti, lc->dev);
	dm_put_device(ti, lc->logdev);
	kfree(lc);
}

static void normal_map_bio(struct dm_target *ti, struct bio *bio)
{
	struct log_writes_c *lc = ti->private;

	bio_set_dev(bio, lc->dev->bdev);
}

static int log_writes_map(struct dm_target *ti, struct bio *bio)
{
	struct log_writes_c *lc = ti->private;
	struct per_bio_data *pb = dm_per_bio_data(bio, sizeof(struct per_bio_data));
	struct pending_block *block;
	struct bvec_iter iter;
	struct bio_vec bv;
	size_t alloc_size;
	int i = 0;
	bool flush_bio = (bio->bi_opf & REQ_PREFLUSH);
	bool fua_bio = (bio->bi_opf & REQ_FUA);
	bool discard_bio = (bio_op(bio) == REQ_OP_DISCARD);
	bool meta_bio = (bio->bi_opf & REQ_META);

	pb->block = NULL;

	/* Don't bother doing anything if logging has been disabled */
	if (!lc->logging_enabled)
		goto map_bio;

	/*
	 * Map reads as normal.
	 */
	if (bio_data_dir(bio) == READ)
		goto map_bio;

	/* No sectors and not a flush?  Don't care */
	if (!bio_sectors(bio) && !flush_bio)
		goto map_bio;

	/*
	 * Discards will have bi_size set but there's no actual data, so just
	 * allocate the size of the pending block.
	 */
	if (discard_bio)
		alloc_size = sizeof(struct pending_block);
	else
		alloc_size = sizeof(struct pending_block) + sizeof(struct bio_vec) * bio_segments(bio);

	block = kzalloc(alloc_size, GFP_NOIO);
	if (!block) {
		DMERR("Error allocating pending block");
		spin_lock_irq(&lc->blocks_lock);
		lc->logging_enabled = false;
		spin_unlock_irq(&lc->blocks_lock);
		return DM_MAPIO_KILL;
	}
	INIT_LIST_HEAD(&block->list);
	pb->block = block;
	atomic_inc(&lc->pending_blocks);

	if (flush_bio)
		block->flags |= LOG_FLUSH_FLAG;
	if (fua_bio)
		block->flags |= LOG_FUA_FLAG;
	if (discard_bio)
		block->flags |= LOG_DISCARD_FLAG;
	if (meta_bio)
		block->flags |= LOG_METADATA_FLAG;

	block->sector = bio_to_dev_sectors(lc, bio->bi_iter.bi_sector);
	block->nr_sectors = bio_to_dev_sectors(lc, bio_sectors(bio));

	/* We don't need the data, just submit */
	if (discard_bio) {
		WARN_ON(flush_bio || fua_bio);
		if (lc->device_supports_discard)
			goto map_bio;
		bio_endio(bio);
		return DM_MAPIO_SUBMITTED;
	}

	/* Flush bio, splice the unflushed blocks onto this list and submit */
	if (flush_bio && !bio_sectors(bio)) {
		spin_lock_irq(&lc->blocks_lock);
		list_splice_init(&lc->unflushed_blocks, &block->list);
		spin_unlock_irq(&lc->blocks_lock);
		goto map_bio;
	}

	/*
	 * We will write this bio somewhere else way later so we need to copy
	 * the actual contents into new pages so we know the data will always be
	 * there.
	 *
	 * We do this because this could be a bio from O_DIRECT in which case we
	 * can't just hold onto the page until some later point, we have to
	 * manually copy the contents.
	 */
	bio_for_each_segment(bv, bio, iter) {
		struct page *page;
		void *src, *dst;

		page = alloc_page(GFP_NOIO);
		if (!page) {
			DMERR("Error allocing page");
			free_pending_block(lc, block);
			spin_lock_irq(&lc->blocks_lock);
			lc->logging_enabled = false;
			spin_unlock_irq(&lc->blocks_lock);
			return DM_MAPIO_KILL;
		}

		src = kmap_atomic(bv.bv_page);
		dst = kmap_atomic(page);
		memcpy(dst, src + bv.bv_offset, bv.bv_len);
		kunmap_atomic(dst);
		kunmap_atomic(src);
		block->vecs[i].bv_page = page;
		block->vecs[i].bv_len = bv.bv_len;
		block->vec_cnt++;
		i++;
	}

	/* Had a flush with data in it, weird */
	if (flush_bio) {
		spin_lock_irq(&lc->blocks_lock);
		list_splice_init(&lc->unflushed_blocks, &block->list);
		spin_unlock_irq(&lc->blocks_lock);
	}
map_bio:
	normal_map_bio(ti, bio);
	return DM_MAPIO_REMAPPED;
}

static int normal_end_io(struct dm_target *ti, struct bio *bio,
		blk_status_t *error)
{
	struct log_writes_c *lc = ti->private;
	struct per_bio_data *pb = dm_per_bio_data(bio, sizeof(struct per_bio_data));

	if (bio_data_dir(bio) == WRITE && pb->block) {
		struct pending_block *block = pb->block;
		unsigned long flags;

		spin_lock_irqsave(&lc->blocks_lock, flags);
		if (block->flags & LOG_FLUSH_FLAG) {
			list_splice_tail_init(&block->list, &lc->logging_blocks);
			list_add_tail(&block->list, &lc->logging_blocks);
			wake_up_process(lc->log_kthread);
		} else if (block->flags & LOG_FUA_FLAG) {
			list_add_tail(&block->list, &lc->logging_blocks);
			wake_up_process(lc->log_kthread);
		} else
			list_add_tail(&block->list, &lc->unflushed_blocks);
		spin_unlock_irqrestore(&lc->blocks_lock, flags);
	}

	return DM_ENDIO_DONE;
}

/*
 * INFO format: <logged entries> <highest allocated sector>
 */
static void log_writes_status(struct dm_target *ti, status_type_t type,
			      unsigned status_flags, char *result,
			      unsigned maxlen)
{
	unsigned sz = 0;
	struct log_writes_c *lc = ti->private;

	switch (type) {
	case STATUSTYPE_INFO:
		DMEMIT("%llu %llu", lc->logged_entries,
		       (unsigned long long)lc->next_sector - 1);
		if (!lc->logging_enabled)
			DMEMIT(" logging_disabled");
		break;

	case STATUSTYPE_TABLE:
		DMEMIT("%s %s", lc->dev->name, lc->logdev->name);
		break;
	}
}

static int log_writes_prepare_ioctl(struct dm_target *ti,
				    struct block_device **bdev)
{
	struct log_writes_c *lc = ti->private;
	struct dm_dev *dev = lc->dev;

	*bdev = dev->bdev;
	/*
	 * Only pass ioctls through if the device sizes match exactly.
	 */
	if (ti->len != i_size_read(dev->bdev->bd_inode) >> SECTOR_SHIFT)
		return 1;
	return 0;
}

static int log_writes_iterate_devices(struct dm_target *ti,
				      iterate_devices_callout_fn fn,
				      void *data)
{
	struct log_writes_c *lc = ti->private;

	return fn(ti, lc->dev, 0, ti->len, data);
}

/*
 * Messages supported:
 *   mark <mark data> - specify the marked data.
 */
static int log_writes_message(struct dm_target *ti, unsigned argc, char **argv,
			      char *result, unsigned maxlen)
{
	int r = -EINVAL;
	struct log_writes_c *lc = ti->private;

	if (argc != 2) {
		DMWARN("Invalid log-writes message arguments, expect 2 arguments, got %d", argc);
		return r;
	}

	if (!strcasecmp(argv[0], "mark"))
		r = log_mark(lc, argv[1]);
	else
		DMWARN("Unrecognised log writes target message received: %s", argv[0]);

	return r;
}

static void log_writes_io_hints(struct dm_target *ti, struct queue_limits *limits)
{
	struct log_writes_c *lc = ti->private;
	struct request_queue *q = bdev_get_queue(lc->dev->bdev);

	if (!q || !blk_queue_discard(q)) {
		lc->device_supports_discard = false;
		limits->discard_granularity = lc->sectorsize;
		limits->max_discard_sectors = (UINT_MAX >> SECTOR_SHIFT);
	}
	limits->logical_block_size = bdev_logical_block_size(lc->dev->bdev);
	limits->physical_block_size = bdev_physical_block_size(lc->dev->bdev);
	limits->io_min = limits->physical_block_size;
}

static long log_writes_dax_direct_access(struct dm_target *ti, pgoff_t pgoff,
					 long nr_pages, void **kaddr, pfn_t *pfn)
{
	struct log_writes_c *lc = ti->private;
	sector_t sector = pgoff * PAGE_SECTORS;
	int ret;

	ret = bdev_dax_pgoff(lc->dev->bdev, sector, nr_pages * PAGE_SIZE, &pgoff);
	if (ret)
		return ret;
	return dax_direct_access(lc->dev->dax_dev, pgoff, nr_pages, kaddr, pfn);
}

static size_t log_writes_dax_copy_from_iter(struct dm_target *ti,
					    pgoff_t pgoff, void *addr, size_t bytes,
					    struct iov_iter *i)
{
	struct log_writes_c *lc = ti->private;
	sector_t sector = pgoff * PAGE_SECTORS;
	int err;

	if (bdev_dax_pgoff(lc->dev->bdev, sector, ALIGN(bytes, PAGE_SIZE), &pgoff))
		return 0;

	/* Don't bother doing anything if logging has been disabled */
	if (!lc->logging_enabled)
		goto dax_copy;

	err = log_dax(lc, sector, bytes, i);
	if (err) {
		DMWARN("Error %d logging DAX write", err);
		return 0;
	}
dax_copy:
	return dax_copy_from_iter(lc->dev->dax_dev, pgoff, addr, bytes, i);
}

static struct target_type log_writes_target = {
	.name   = "log-writes",
	.version = {1, 1, 0},
	.module = THIS_MODULE,
	.ctr    = log_writes_ctr,
	.dtr    = log_writes_dtr,
	.map    = log_writes_map,
	.end_io = normal_end_io,
	.status = log_writes_status,
	.prepare_ioctl = log_writes_prepare_ioctl,
	.message = log_writes_message,
	.iterate_devices = log_writes_iterate_devices,
	.io_hints = log_writes_io_hints,
	.direct_access = log_writes_dax_direct_access,
	.dax_copy_from_iter = log_writes_dax_copy_from_iter,
};

static int __init dm_log_writes_init(void)
{
	int r = dm_register_target(&log_writes_target);

	if (r < 0)
		DMERR("register failed %d", r);

	return r;
}

static void __exit dm_log_writes_exit(void)
{
	dm_unregister_target(&log_writes_target);
}

module_init(dm_log_writes_init);
module_exit(dm_log_writes_exit);

MODULE_DESCRIPTION(DM_NAME " log writes target");
MODULE_AUTHOR("Josef Bacik <jbacik@fb.com>");
MODULE_LICENSE("GPL");