-
Notifications
You must be signed in to change notification settings - Fork 0
/
dm-sadc.c
1075 lines (827 loc) · 27.9 KB
/
dm-sadc.c
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
#include <linux/device-mapper.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/bio.h>
#include <linux/completion.h>
#include <linux/string_helpers.h>
#define DM_MSG_PREFIX "sadc"
/* Disk state reset IOCTL command. */
#define RESET_DISK 0xDEADBEEF
#define pbas_in_bio bio_segments
#define bio_begin_lba(bio) ((bio)->bi_sector)
#define bio_end_lba bio_end_sector
#define bio_begin_pba(bio) (lba_to_pba(bio_begin_lba(bio)))
#define bio_end_pba(bio) (lba_to_pba(bio_end_lba(bio)))
#define MIN_DISK_SIZE (76LL << 10)
#define MAX_DISK_SIZE (10LL << 40)
#define LBA_SIZE 512
#define PBA_SIZE 4096
#define LBAS_IN_PBA (PBA_SIZE / LBA_SIZE)
#define lba_to_pba(lba) ((pba_t) (lba / LBAS_IN_PBA))
#define pba_to_lba(pba) (((lba_t) pba) * LBAS_IN_PBA)
typedef sector_t lba_t;
typedef int32_t pba_t;
#define MIN_IOS 16
#define MIN_POOL_PAGES 32
static struct kmem_cache *_io_pool;
struct io {
struct sadc_ctx *sc;
struct bio *bio;
struct work_struct work;
atomic_t pending;
};
struct cache_band {
int32_t nr;
pba_t begin_pba;
pba_t current_pba;
unsigned long *map;
};
struct sadc_ctx {
struct dm_dev *dev;
int64_t disk_size;
int32_t cache_percent;
int64_t cache_size;
int32_t nr_cache_bands;
int64_t usable_size;
int64_t wasted_size;
int32_t track_size;
int64_t band_size;
int32_t band_size_tracks;
int32_t band_size_pbas;
int32_t nr_valid_pbas;
int32_t nr_usable_pbas;
pba_t *pba_map;
struct cache_band *cache_bands;
int32_t nr_bands;
int32_t nr_usable_bands;
int32_t cache_assoc;
mempool_t *io_pool;
mempool_t *page_pool;
struct workqueue_struct *queue;
struct mutex lock;
struct completion io_completion;
struct bio_set *bs;
atomic_t error;
struct bio **tmp_bios;
struct bio **rmw_bios;
};
static char *readable(u64 size)
{
static char buf[10];
string_get_size(size, STRING_UNITS_2, buf, sizeof(buf));
return buf;
}
static inline void debug_bio(struct sadc_ctx *sc, struct bio *bio, const char *f)
{
int i;
unsigned long flags;
struct bio_vec *bv;
pr_debug("%10s: %c offset: %d size: %u\n",
f,
(bio_data_dir(bio) == READ ? 'R' : 'W'),
bio_begin_pba(bio),
bio->bi_size);
bio_for_each_segment(bv, bio, i) {
char *addr = bvec_kmap_irq(bv, &flags);
pr_debug("seg: %d, addr: %p, len: %u, offset: %u, char: [%d]\n",
i, addr, bv->bv_len, bv->bv_offset, *addr);
bvec_kunmap_irq(addr, &flags);
}
}
static bool unaligned_bio(struct bio *bio)
{
return bio_begin_lba(bio) & 0x7 || bio->bi_size & 0xfff;
}
static struct io *alloc_io(struct sadc_ctx *sc, struct bio *bio)
{
struct io *io = mempool_alloc(sc->io_pool, GFP_NOIO);
if (unlikely(!io)) {
DMERR("Could not allocate io from mempool!");
return NULL;
}
memset(io, 0, sizeof(*io));
io->sc = sc;
io->bio = bio;
atomic_set(&io->pending, 0);
return io;
}
static void sadcd(struct work_struct *work);
static void queue_io(struct io *io)
{
struct sadc_ctx *sc = io->sc;
INIT_WORK(&io->work, sadcd);
queue_work(sc->queue, &io->work);
}
static void release_io(struct io *io, int error)
{
struct sadc_ctx *sc = io->sc;
bool rmw_bio = io->bio == NULL;
WARN_ON(atomic_read(&io->pending));
mempool_free(io, sc->io_pool);
if (rmw_bio)
atomic_set(&sc->error, error);
else
bio_endio(io->bio, error);
}
static inline bool usable_pba(struct sadc_ctx *sc, pba_t pba)
{
return 0 <= pba && pba < sc->nr_usable_pbas;
}
static inline bool usable_band(struct sadc_ctx *sc, int32_t band)
{
return 0 <= band && band < sc->nr_usable_bands;
}
static inline pba_t band_begin_pba(struct sadc_ctx *sc, int32_t band)
{
WARN_ON(!usable_band(sc, band));
return band * sc->band_size_pbas;
}
static inline pba_t band_end_pba(struct sadc_ctx *sc, int32_t band)
{
return band_begin_pba(sc, band) + sc->band_size_pbas;
}
static inline int32_t pba_band(struct sadc_ctx *sc, pba_t pba)
{
WARN_ON(!usable_pba(sc, pba));
return pba / sc->band_size_pbas;
}
static inline int32_t bio_band(struct sadc_ctx *sc, struct bio *bio)
{
return pba_band(sc, bio_begin_pba(bio));
}
static inline int band_to_bit(struct sadc_ctx *sc, struct cache_band *cb,
int32_t band)
{
return (band - cb->nr) / sc->nr_cache_bands;
}
static inline int bit_to_band(struct sadc_ctx *sc, struct cache_band *cb,
int bit)
{
return bit * sc->nr_cache_bands + cb->nr;
}
static inline struct cache_band *cache_band(struct sadc_ctx *sc, int32_t band)
{
WARN_ON(!usable_band(sc, band));
return &sc->cache_bands[band % sc->nr_cache_bands];
}
static inline int32_t free_pbas_in_cache_band(struct sadc_ctx *sc,
struct cache_band *cb)
{
return sc->band_size_pbas - (cb->current_pba - cb->begin_pba);
}
static int32_t pbas_in_band(struct sadc_ctx *sc, struct bio *bio, int32_t band)
{
pba_t begin_pba = max(band_begin_pba(sc, band), bio_begin_pba(bio));
pba_t end_pba = min(band_end_pba(sc, band), bio_end_pba(bio));
return max(end_pba - begin_pba, 0);
}
static void unmap_pba_range(struct sadc_ctx *sc, pba_t begin, pba_t end)
{
int i;
WARN_ON(begin >= end);
WARN_ON(!usable_pba(sc, end - 1));
for (i = begin; i < end; ++i)
sc->pba_map[i] = -1;
}
static pba_t map_pba_range(struct sadc_ctx *sc, pba_t begin, pba_t end)
{
pba_t i;
int32_t b;
struct cache_band *cb;
WARN_ON(begin >= end);
WARN_ON(!usable_pba(sc, end - 1));
b = pba_band(sc, begin);
WARN_ON(b != pba_band(sc, end - 1));
cb = cache_band(sc, b);
WARN_ON(free_pbas_in_cache_band(sc, cb) < (end - begin));
for (i = begin; i < end; ++i)
sc->pba_map[i] = cb->current_pba++;
set_bit(band_to_bit(sc, cb, b), cb->map);
return sc->pba_map[begin];
}
static inline pba_t lookup_pba(struct sadc_ctx *sc, pba_t pba)
{
WARN_ON(!usable_pba(sc, pba));
return sc->pba_map[pba] == -1 ? pba : sc->pba_map[pba];
}
static inline lba_t lookup_lba(struct sadc_ctx *sc, lba_t lba)
{
return pba_to_lba(lookup_pba(sc, lba_to_pba(lba))) + lba % LBAS_IN_PBA;
}
static void do_free_bio_pages(struct sadc_ctx *sc, struct bio *bio)
{
int i;
struct bio_vec *bv;
bio_for_each_segment_all(bv, bio, i) {
WARN_ON(!bv->bv_page);
mempool_free(bv->bv_page, sc->page_pool);
bv->bv_page = NULL;
}
/* For now we should only have a single page per bio. */
WARN_ON(i != 1);
}
static void endio(struct bio *bio, int error)
{
struct io *io = bio->bi_private;
struct sadc_ctx *sc = io->sc;
bool rmw_bio = io->bio == NULL;
if (rmw_bio && bio_data_dir(bio) == WRITE)
do_free_bio_pages(sc, bio);
bio_put(bio);
if (atomic_dec_and_test(&io->pending)) {
release_io(io, error);
complete(&sc->io_completion);
}
}
static bool adjacent_pbas(struct sadc_ctx *sc, pba_t x, pba_t y)
{
return lookup_pba(sc, x) + 1 == lookup_pba(sc, y);
}
static struct bio *clone_remap_bio(struct io *io, struct bio *bio, int idx,
pba_t pba, int nr_pbas)
{
struct sadc_ctx *sc = io->sc;
struct bio *clone;
clone = bio_clone_bioset(bio, GFP_NOIO, sc->bs);
if (unlikely(!clone)) {
DMERR("Cannot clone a bio.");
return NULL;
}
if (bio_data_dir(bio) == READ)
pba = lookup_pba(sc, pba);
else
pba = map_pba_range(sc, pba, pba + nr_pbas);
clone->bi_sector = pba_to_lba(pba);
clone->bi_private = io;
clone->bi_end_io = endio;
clone->bi_bdev = sc->dev->bdev;
clone->bi_idx = idx;
clone->bi_vcnt = idx + nr_pbas;
clone->bi_size = nr_pbas * PBA_SIZE;
atomic_inc(&io->pending);
return clone;
}
static void release_bio(struct bio *bio)
{
struct io *io = bio->bi_private;
atomic_dec(&io->pending);
bio_put(bio);
}
static int handle_unaligned_io(struct sadc_ctx *sc, struct io *io)
{
struct bio *bio = io->bio;
struct bio *clone = bio_clone_bioset(bio, GFP_NOIO, sc->bs);
WARN_ON(bio_data_dir(bio) != READ);
WARN_ON(bio_end_lba(bio) > pba_to_lba(bio_begin_pba(bio) + 1));
if (unlikely(!clone)) {
DMERR("Cannot clone a bio.");
return -ENOMEM;
}
clone->bi_sector = lookup_lba(sc, bio_begin_lba(bio));
clone->bi_private = io;
clone->bi_end_io = endio;
clone->bi_bdev = sc->dev->bdev;
atomic_inc(&io->pending);
sc->tmp_bios[0] = clone;
return 1;
}
static int split_read_io(struct sadc_ctx *sc, struct io *io)
{
struct bio *bio = io->bio;
pba_t bp, p;
int i, n = 0, idx = 0;
if (unlikely(unaligned_bio(bio)))
return handle_unaligned_io(sc, io);
bp = bio_begin_pba(bio);
p = bp + 1;
for (i = 1; p < bio_end_pba(bio); ++i, ++p) {
if (adjacent_pbas(sc, p - 1, p))
continue;
sc->tmp_bios[n] = clone_remap_bio(io, bio, idx, bp, i - idx);
if (!sc->tmp_bios[n])
goto bad;
++n, idx = i, bp = p;
}
sc->tmp_bios[n] = clone_remap_bio(io, bio, idx, bp, i - idx);
if (sc->tmp_bios[n])
return n + 1;
bad:
while (n--)
release_bio(sc->tmp_bios[n]);
return -ENOMEM;
}
static int split_write_io(struct sadc_ctx *sc, struct io *io)
{
struct bio *bio = io->bio;
int32_t nr_pbas_bio1, nr_pbas_bio2;
int idx = 0;
pba_t p;
nr_pbas_bio1 = pbas_in_band(sc, bio, bio_band(sc, bio));
nr_pbas_bio2 = pbas_in_bio(bio) - nr_pbas_bio1;
p = bio_begin_pba(bio);
sc->tmp_bios[0] = clone_remap_bio(io, bio, idx, p, nr_pbas_bio1);
if (!sc->tmp_bios[0])
return -ENOMEM;
if (!nr_pbas_bio2)
return 1;
p += nr_pbas_bio1;
idx += nr_pbas_bio1;
sc->tmp_bios[1] =
clone_remap_bio(io, bio, idx, p, nr_pbas_bio2);
if (!sc->tmp_bios[1]) {
release_bio(sc->tmp_bios[0]);
return -ENOMEM;
}
return 2;
}
static int do_sync_io(struct sadc_ctx *sc, struct bio **bios, int n)
{
int i;
reinit_completion(&sc->io_completion);
for (i = 0; i < n; ++i)
generic_make_request(bios[i]);
wait_for_completion(&sc->io_completion);
return atomic_read(&sc->error);
}
typedef int (*split_t)(struct sadc_ctx *sc, struct io *io);
static void do_io(struct sadc_ctx *sc, struct io *io, split_t split)
{
int n = split(sc, io);
if (n < 0) {
release_io(io, n);
return;
}
WARN_ON(!n);
do_sync_io(sc, sc->tmp_bios, n);
}
static struct cache_band *cache_band_to_gc(struct sadc_ctx *sc, struct bio *bio)
{
int b = bio_band(sc, bio);
int nr_pbas = pbas_in_band(sc, bio, b);
struct cache_band *cb = cache_band(sc, b);
if (free_pbas_in_cache_band(sc, cb) < nr_pbas)
return cb;
if (!usable_band(sc, ++b))
return NULL;
cb = cache_band(sc, b);
nr_pbas = pbas_in_bio(bio) - nr_pbas;
return free_pbas_in_cache_band(sc, cb) < nr_pbas ? cb : NULL;
}
static struct bio *alloc_bio_with_page(struct sadc_ctx *sc, pba_t pba)
{
struct page *page = mempool_alloc(sc->page_pool, GFP_NOIO);
struct bio *bio = bio_alloc_bioset(GFP_NOIO, 1, sc->bs);
if (!bio || !page)
goto bad;
bio->bi_sector = pba_to_lba(pba);
bio->bi_bdev = sc->dev->bdev;
if (!bio_add_page(bio, page, PAGE_SIZE, 0))
goto bad;
return bio;
bad:
if (page)
mempool_free(page, sc->page_pool);
if (bio)
bio_put(bio);
return NULL;
}
static void free_rmw_bios(struct sadc_ctx *sc, int n)
{
int i;
for (i = 0; i < n; ++i) {
do_free_bio_pages(sc, sc->rmw_bios[i]);
bio_put(sc->rmw_bios[i]);
}
}
static bool alloc_rmw_bios(struct sadc_ctx *sc, int32_t band)
{
pba_t p = band_begin_pba(sc, band);
int i;
for (i = 0; i < sc->band_size_pbas; ++i) {
sc->rmw_bios[i] = alloc_bio_with_page(sc, p + i);
if (!sc->rmw_bios[i])
goto bad;
}
return true;
bad:
free_rmw_bios(sc, i);
return false;
}
static struct bio *clone_bio(struct io *io, struct bio *bio, pba_t pba)
{
struct sadc_ctx *sc = io->sc;
struct bio *clone = bio_clone_bioset(bio, GFP_NOIO, sc->bs);
if (unlikely(!clone)) {
DMERR("Cannot clone bio.");
return NULL;
}
clone->bi_private = io;
clone->bi_end_io = endio;
clone->bi_sector = pba_to_lba(pba);
atomic_inc(&io->pending);
return clone;
}
static int do_read_band(struct sadc_ctx *sc, int32_t band)
{
struct io *io = alloc_io(sc, NULL);
pba_t p = band_begin_pba(sc, band);
int i;
if (unlikely(!io))
return -ENOMEM;
for (i = 0; i < sc->band_size_pbas; ++i) {
sc->tmp_bios[i] = clone_bio(io, sc->rmw_bios[i], p + i);
if (!sc->tmp_bios[i])
goto bad;
}
return do_sync_io(sc, sc->tmp_bios, sc->band_size_pbas);
bad:
while (i--)
release_bio(sc->tmp_bios[i]);
return -ENOMEM;
}
static int do_modify_band(struct sadc_ctx *sc, int32_t band)
{
struct io *io = alloc_io(sc, NULL);
pba_t p = band_begin_pba(sc, band);
int i, j;
if (unlikely(!io))
return -ENOMEM;
for (i = j = 0; i < sc->band_size_pbas; ++i) {
pba_t pp = lookup_pba(sc, bio_begin_pba(sc->rmw_bios[i]));
if (pp == p + i)
continue;
sc->tmp_bios[j] = clone_bio(io, sc->rmw_bios[i], pp);
if (!sc->tmp_bios[j])
goto bad;
++j;
}
WARN_ON(!j);
return do_sync_io(sc, sc->tmp_bios, j);
bad:
while (j--)
release_bio(sc->tmp_bios[j]);
return -ENOMEM;
}
static int do_write_band(struct sadc_ctx *sc, int32_t band)
{
struct io *io = alloc_io(sc, NULL);
int i;
if (unlikely(!io))
return -ENOMEM;
for (i = 0; i < sc->band_size_pbas; ++i) {
sc->rmw_bios[i]->bi_private = io;
sc->rmw_bios[i]->bi_end_io = endio;
sc->rmw_bios[i]->bi_rw = WRITE;
}
atomic_set(&io->pending, sc->band_size_pbas);
return do_sync_io(sc, sc->rmw_bios, sc->band_size_pbas);
}
static int do_rmw_band(struct sadc_ctx *sc, int32_t band)
{
int r = 0;
if (!alloc_rmw_bios(sc, band))
return -ENOMEM;
pr_debug("Reading band %d\n", band);
r = do_read_band(sc, band);
if (r < 0)
goto bad;
pr_debug("Modifying band %d\n", band);
r = do_modify_band(sc, band);
if (r < 0)
goto bad;
pr_debug("Writing band %d\n", band);
return do_write_band(sc, band);
bad:
free_rmw_bios(sc, sc->band_size_pbas);
return r;
}
static void reset_cache_band(struct sadc_ctx *sc, struct cache_band *cb)
{
cb->current_pba = cb->begin_pba;
bitmap_zero(cb->map, sc->cache_assoc);
}
static int do_gc_cache_band(struct sadc_ctx *sc, struct cache_band *cb)
{
int i;
for_each_set_bit(i, cb->map, sc->cache_assoc) {
int b = bit_to_band(sc, cb, i);
int r = do_rmw_band(sc, b);
if (r < 0)
return r;
unmap_pba_range(sc, band_begin_pba(sc, b), band_end_pba(sc, b));
}
reset_cache_band(sc, cb);
return 0;
}
static int do_gc_if_required(struct sadc_ctx *sc, struct bio *bio)
{
struct cache_band *cb = cache_band_to_gc(sc, bio);
int r = 0;
if (!cb)
return r;
pr_debug("%d Starting GC.\n", current->pid);
do {
r = do_gc_cache_band(sc, cb);
if (r < 0)
return r;
cb = cache_band_to_gc(sc, bio);
} while (cb);
pr_debug("%d GC completed.\n", current->pid);
return r;
}
static void sadcd(struct work_struct *work)
{
struct io *io = container_of(work, struct io, work);
struct sadc_ctx *sc = io->sc;
struct bio *bio = io->bio;
mutex_lock(&sc->lock);
if (bio_data_dir(bio) == READ) {
do_io(sc, io, split_read_io);
} else {
int r;
WARN_ON(unaligned_bio(bio));
r = do_gc_if_required(sc, bio);
if (r < 0)
release_io(io, r);
else
do_io(sc, io, split_write_io);
}
mutex_unlock(&sc->lock);
}
static bool get_args(struct dm_target *ti, struct sadc_ctx *sc,
int argc, char **argv)
{
unsigned long long tmp;
char d;
if (argc != 5) {
ti->error = "dm-sadc: Invalid argument count.";
return false;
}
if (sscanf(argv[1], "%llu%c", &tmp, &d) != 1 || tmp & 0xfff ||
(tmp < 4 * 1024 || tmp > 2 * 1024 * 1024)) {
ti->error = "dm-sadc: Invalid track size.";
return false;
}
sc->track_size = tmp;
if (sscanf(argv[2], "%llu%c", &tmp, &d) != 1 || tmp < 1 || tmp > 200) {
ti->error = "dm-sadc: Invalid band size.";
return false;
}
sc->band_size_tracks = tmp;
if (sscanf(argv[3], "%llu%c", &tmp, &d) != 1 || tmp < 1 || tmp > 50) {
ti->error = "dm-sadc: Invalid cache percent.";
return false;
}
sc->cache_percent = tmp;
if (sscanf(argv[4], "%llu%c", &tmp, &d) != 1 ||
tmp < MIN_DISK_SIZE || tmp > MAX_DISK_SIZE) {
ti->error = "dm-sadc: Invalid disk size.";
return false;
}
sc->disk_size = tmp;
return true;
}
static void calc_params(struct sadc_ctx *sc)
{
sc->band_size = sc->band_size_tracks * sc->track_size;
sc->band_size_pbas = sc->band_size / PBA_SIZE;
sc->nr_bands = sc->disk_size / sc->band_size;
sc->nr_cache_bands = sc->nr_bands * sc->cache_percent / 100;
sc->cache_size = sc->nr_cache_bands * sc->band_size;
/*
* Make |nr_usable_bands| a multiple of |nr_cache_bands| so that all
* cache bands are equally loaded.
*/
sc->nr_usable_bands = (sc->nr_bands / sc->nr_cache_bands - 1) *
sc->nr_cache_bands;
sc->cache_assoc = sc->nr_usable_bands / sc->nr_cache_bands;
sc->usable_size = sc->nr_usable_bands * sc->band_size;
sc->wasted_size = sc->disk_size - sc->cache_size - sc->usable_size;
sc->nr_valid_pbas = (sc->usable_size + sc->cache_size) / PBA_SIZE;
sc->nr_usable_pbas = sc->usable_size / PBA_SIZE;
WARN_ON(sc->usable_size % PBA_SIZE);
}
static void print_params(struct sadc_ctx *sc)
{
DMINFO("Disk size: %s", readable(sc->disk_size));
DMINFO("Band size: %s", readable(sc->band_size));
DMINFO("Band size: %d pbas", sc->band_size_pbas);
DMINFO("Total number of bands: %d", sc->nr_bands);
DMINFO("Number of cache bands: %d", sc->nr_cache_bands);
DMINFO("Cache size: %s", readable(sc->cache_size));
DMINFO("Number of usable bands: %d", sc->nr_usable_bands);
DMINFO("Usable disk size: %s", readable(sc->usable_size));
DMINFO("Number of usable pbas: %d", sc->nr_usable_pbas);
DMINFO("Wasted disk size: %s", readable(sc->wasted_size));
}
static void sadc_dtr(struct dm_target *ti)
{
int i;
struct sadc_ctx *sc = (struct sadc_ctx *) ti->private;
DMINFO("Destructing...");
ti->private = NULL;
if (!sc)
return;
if (sc->tmp_bios)
vfree(sc->tmp_bios);
if (sc->rmw_bios)
vfree(sc->rmw_bios);
if (sc->pba_map)
vfree(sc->pba_map);
for (i = 0; i < sc->nr_cache_bands; ++i)
if (sc->cache_bands[i].map)
kfree(sc->cache_bands[i].map);
if (sc->cache_bands)
vfree(sc->cache_bands);
if (sc->io_pool)
mempool_destroy(sc->io_pool);
if (sc->queue)
destroy_workqueue(sc->queue);
if (sc->dev)
dm_put_device(ti, sc->dev);
kzfree(sc);
}
static bool alloc_structs(struct sadc_ctx *sc)
{
int32_t i, size, pba;
size = sizeof(int32_t) * sc->nr_usable_pbas;
sc->pba_map = vmalloc(size);
if (!sc->pba_map)
return false;
memset(sc->pba_map, -1, size);
size = sizeof(struct bio *) * sc->band_size_pbas;
sc->rmw_bios = vzalloc(size);
if (!sc->rmw_bios)
return false;
sc->tmp_bios = vzalloc(size);
if (!sc->tmp_bios)
return false;
size = sizeof(struct cache_band) * sc->nr_cache_bands;
sc->cache_bands = vmalloc(size);
if (!sc->cache_bands)
return false;
/* The cache region starts where the data region ends. */
pba = sc->nr_usable_pbas;
size = BITS_TO_LONGS(sc->cache_assoc) * sizeof(long);
for (i = 0; i < sc->nr_cache_bands; ++i, pba += sc->band_size_pbas) {
sc->cache_bands[i].nr = i;
sc->cache_bands[i].begin_pba = pba;
sc->cache_bands[i].map = kmalloc(size, GFP_KERNEL);
if (!sc->cache_bands[i].map)
return false;
reset_cache_band(sc, &sc->cache_bands[i]);
}
return true;
}
static int sadc_ctr(struct dm_target *ti, unsigned int argc, char **argv)
{
struct sadc_ctx *sc;
int32_t ret;
DMINFO("Constructing...");
sc = kzalloc(sizeof(*sc), GFP_KERNEL);
if (!sc) {
ti->error = "dm-sadc: Cannot allocate sadc context.";
return -ENOMEM;
}
ti->private = sc;
if (!get_args(ti, sc, argc, argv)) {
kzfree(sc);
return -EINVAL;
}
calc_params(sc);
print_params(sc);
ret = -ENOMEM;
if (!alloc_structs(sc)) {
ti->error = "Cannot allocate data structures.";
goto bad;
}
sc->io_pool = mempool_create_slab_pool(MIN_IOS, _io_pool);
if (!sc->io_pool) {
ti->error = "Cannot allocate mempool.";
goto bad;
}
sc->page_pool = mempool_create_page_pool(MIN_POOL_PAGES, 0);
if (!sc->page_pool) {
ti->error = "Cannot allocate page mempool.";
goto bad;
}
sc->bs = bioset_create(MIN_IOS, 0);
if (!sc->bs) {
ti->error = "Cannot allocate bioset.";
goto bad;
}
sc->queue = alloc_workqueue("sadcd",
WQ_NON_REENTRANT | WQ_MEM_RECLAIM, 1);
if (!sc->queue) {
ti->error = "Cannot allocate work queue.";
goto bad;
}
ret = -EINVAL;
if (dm_get_device(ti, argv[0], dm_table_get_mode(ti->table), &sc->dev)) {
ti->error = "dm-sadc: Device lookup failed.";
return -1;
}
mutex_init(&sc->lock);
init_completion(&sc->io_completion);
/* TODO: Reconsider proper values for these. */
ti->num_flush_bios = 1;
ti->num_discard_bios = 1;
ti->num_write_same_bios = 1;
return 0;
bad:
sadc_dtr(ti);
return ret;
}
static int sadc_map(struct dm_target *ti, struct bio *bio)
{
struct sadc_ctx *sc = ti->private;
struct io *io;
if (unlikely(bio->bi_rw & (REQ_FLUSH | REQ_DISCARD))) {
WARN_ON(bio_sectors(bio));
bio->bi_bdev = sc->dev->bdev;
return DM_MAPIO_REMAPPED;
}
io = alloc_io(sc, bio);
if (unlikely(!io))
return -EIO;
queue_io(io);
return DM_MAPIO_SUBMITTED;
}
static void sadc_status(struct dm_target *ti, status_type_t type,
unsigned status_flags, char *result, unsigned maxlen)
{
struct sadc_ctx *sc = (struct sadc_ctx *) ti->private;
switch (type) {
case STATUSTYPE_INFO:
result[0] = '\0';
break;
/* TODO: get string representation of device name.*/
case STATUSTYPE_TABLE:
snprintf(result, maxlen, "%s track size: %d, band size in tracks: %d, cache percent: %d%%",
sc->dev->name,
sc->track_size,
sc->band_size_tracks,
sc->cache_percent);
break;
}
}
static int reset_disk(struct sadc_ctx *sc)
{
int i;
DMINFO("Resetting disk...");