forked from torvalds/linux
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathn2_core.c
2267 lines (1877 loc) · 52.7 KB
/
n2_core.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
/* n2_core.c: Niagara2 Stream Processing Unit (SPU) crypto support.
*
* Copyright (C) 2010 David S. Miller <[email protected]>
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/of.h>
#include <linux/of_device.h>
#include <linux/cpumask.h>
#include <linux/slab.h>
#include <linux/interrupt.h>
#include <linux/crypto.h>
#include <crypto/md5.h>
#include <crypto/sha.h>
#include <crypto/aes.h>
#include <crypto/des.h>
#include <linux/mutex.h>
#include <linux/delay.h>
#include <linux/sched.h>
#include <crypto/internal/hash.h>
#include <crypto/scatterwalk.h>
#include <crypto/algapi.h>
#include <asm/hypervisor.h>
#include <asm/mdesc.h>
#include "n2_core.h"
#define DRV_MODULE_NAME "n2_crypto"
#define DRV_MODULE_VERSION "0.1"
#define DRV_MODULE_RELDATE "April 29, 2010"
static char version[] __devinitdata =
DRV_MODULE_NAME ".c:v" DRV_MODULE_VERSION " (" DRV_MODULE_RELDATE ")\n";
MODULE_AUTHOR("David S. Miller ([email protected])");
MODULE_DESCRIPTION("Niagara2 Crypto driver");
MODULE_LICENSE("GPL");
MODULE_VERSION(DRV_MODULE_VERSION);
#define N2_CRA_PRIORITY 300
static DEFINE_MUTEX(spu_lock);
struct spu_queue {
cpumask_t sharing;
unsigned long qhandle;
spinlock_t lock;
u8 q_type;
void *q;
unsigned long head;
unsigned long tail;
struct list_head jobs;
unsigned long devino;
char irq_name[32];
unsigned int irq;
struct list_head list;
};
static struct spu_queue **cpu_to_cwq;
static struct spu_queue **cpu_to_mau;
static unsigned long spu_next_offset(struct spu_queue *q, unsigned long off)
{
if (q->q_type == HV_NCS_QTYPE_MAU) {
off += MAU_ENTRY_SIZE;
if (off == (MAU_ENTRY_SIZE * MAU_NUM_ENTRIES))
off = 0;
} else {
off += CWQ_ENTRY_SIZE;
if (off == (CWQ_ENTRY_SIZE * CWQ_NUM_ENTRIES))
off = 0;
}
return off;
}
struct n2_request_common {
struct list_head entry;
unsigned int offset;
};
#define OFFSET_NOT_RUNNING (~(unsigned int)0)
/* An async job request records the final tail value it used in
* n2_request_common->offset, test to see if that offset is in
* the range old_head, new_head, inclusive.
*/
static inline bool job_finished(struct spu_queue *q, unsigned int offset,
unsigned long old_head, unsigned long new_head)
{
if (old_head <= new_head) {
if (offset > old_head && offset <= new_head)
return true;
} else {
if (offset > old_head || offset <= new_head)
return true;
}
return false;
}
/* When the HEAD marker is unequal to the actual HEAD, we get
* a virtual device INO interrupt. We should process the
* completed CWQ entries and adjust the HEAD marker to clear
* the IRQ.
*/
static irqreturn_t cwq_intr(int irq, void *dev_id)
{
unsigned long off, new_head, hv_ret;
struct spu_queue *q = dev_id;
pr_err("CPU[%d]: Got CWQ interrupt for qhdl[%lx]\n",
smp_processor_id(), q->qhandle);
spin_lock(&q->lock);
hv_ret = sun4v_ncs_gethead(q->qhandle, &new_head);
pr_err("CPU[%d]: CWQ gethead[%lx] hv_ret[%lu]\n",
smp_processor_id(), new_head, hv_ret);
for (off = q->head; off != new_head; off = spu_next_offset(q, off)) {
/* XXX ... XXX */
}
hv_ret = sun4v_ncs_sethead_marker(q->qhandle, new_head);
if (hv_ret == HV_EOK)
q->head = new_head;
spin_unlock(&q->lock);
return IRQ_HANDLED;
}
static irqreturn_t mau_intr(int irq, void *dev_id)
{
struct spu_queue *q = dev_id;
unsigned long head, hv_ret;
spin_lock(&q->lock);
pr_err("CPU[%d]: Got MAU interrupt for qhdl[%lx]\n",
smp_processor_id(), q->qhandle);
hv_ret = sun4v_ncs_gethead(q->qhandle, &head);
pr_err("CPU[%d]: MAU gethead[%lx] hv_ret[%lu]\n",
smp_processor_id(), head, hv_ret);
sun4v_ncs_sethead_marker(q->qhandle, head);
spin_unlock(&q->lock);
return IRQ_HANDLED;
}
static void *spu_queue_next(struct spu_queue *q, void *cur)
{
return q->q + spu_next_offset(q, cur - q->q);
}
static int spu_queue_num_free(struct spu_queue *q)
{
unsigned long head = q->head;
unsigned long tail = q->tail;
unsigned long end = (CWQ_ENTRY_SIZE * CWQ_NUM_ENTRIES);
unsigned long diff;
if (head > tail)
diff = head - tail;
else
diff = (end - tail) + head;
return (diff / CWQ_ENTRY_SIZE) - 1;
}
static void *spu_queue_alloc(struct spu_queue *q, int num_entries)
{
int avail = spu_queue_num_free(q);
if (avail >= num_entries)
return q->q + q->tail;
return NULL;
}
static unsigned long spu_queue_submit(struct spu_queue *q, void *last)
{
unsigned long hv_ret, new_tail;
new_tail = spu_next_offset(q, last - q->q);
hv_ret = sun4v_ncs_settail(q->qhandle, new_tail);
if (hv_ret == HV_EOK)
q->tail = new_tail;
return hv_ret;
}
static u64 control_word_base(unsigned int len, unsigned int hmac_key_len,
int enc_type, int auth_type,
unsigned int hash_len,
bool sfas, bool sob, bool eob, bool encrypt,
int opcode)
{
u64 word = (len - 1) & CONTROL_LEN;
word |= ((u64) opcode << CONTROL_OPCODE_SHIFT);
word |= ((u64) enc_type << CONTROL_ENC_TYPE_SHIFT);
word |= ((u64) auth_type << CONTROL_AUTH_TYPE_SHIFT);
if (sfas)
word |= CONTROL_STORE_FINAL_AUTH_STATE;
if (sob)
word |= CONTROL_START_OF_BLOCK;
if (eob)
word |= CONTROL_END_OF_BLOCK;
if (encrypt)
word |= CONTROL_ENCRYPT;
if (hmac_key_len)
word |= ((u64) (hmac_key_len - 1)) << CONTROL_HMAC_KEY_LEN_SHIFT;
if (hash_len)
word |= ((u64) (hash_len - 1)) << CONTROL_HASH_LEN_SHIFT;
return word;
}
#if 0
static inline bool n2_should_run_async(struct spu_queue *qp, int this_len)
{
if (this_len >= 64 ||
qp->head != qp->tail)
return true;
return false;
}
#endif
struct n2_ahash_alg {
struct list_head entry;
const char *hash_zero;
const u32 *hash_init;
u8 hw_op_hashsz;
u8 digest_size;
u8 auth_type;
u8 hmac_type;
struct ahash_alg alg;
};
static inline struct n2_ahash_alg *n2_ahash_alg(struct crypto_tfm *tfm)
{
struct crypto_alg *alg = tfm->__crt_alg;
struct ahash_alg *ahash_alg;
ahash_alg = container_of(alg, struct ahash_alg, halg.base);
return container_of(ahash_alg, struct n2_ahash_alg, alg);
}
struct n2_hmac_alg {
const char *child_alg;
struct n2_ahash_alg derived;
};
static inline struct n2_hmac_alg *n2_hmac_alg(struct crypto_tfm *tfm)
{
struct crypto_alg *alg = tfm->__crt_alg;
struct ahash_alg *ahash_alg;
ahash_alg = container_of(alg, struct ahash_alg, halg.base);
return container_of(ahash_alg, struct n2_hmac_alg, derived.alg);
}
struct n2_hash_ctx {
struct crypto_ahash *fallback_tfm;
};
#define N2_HASH_KEY_MAX 32 /* HW limit for all HMAC requests */
struct n2_hmac_ctx {
struct n2_hash_ctx base;
struct crypto_shash *child_shash;
int hash_key_len;
unsigned char hash_key[N2_HASH_KEY_MAX];
};
struct n2_hash_req_ctx {
union {
struct md5_state md5;
struct sha1_state sha1;
struct sha256_state sha256;
} u;
struct ahash_request fallback_req;
};
static int n2_hash_async_init(struct ahash_request *req)
{
struct n2_hash_req_ctx *rctx = ahash_request_ctx(req);
struct crypto_ahash *tfm = crypto_ahash_reqtfm(req);
struct n2_hash_ctx *ctx = crypto_ahash_ctx(tfm);
ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback_tfm);
rctx->fallback_req.base.flags = req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP;
return crypto_ahash_init(&rctx->fallback_req);
}
static int n2_hash_async_update(struct ahash_request *req)
{
struct n2_hash_req_ctx *rctx = ahash_request_ctx(req);
struct crypto_ahash *tfm = crypto_ahash_reqtfm(req);
struct n2_hash_ctx *ctx = crypto_ahash_ctx(tfm);
ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback_tfm);
rctx->fallback_req.base.flags = req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP;
rctx->fallback_req.nbytes = req->nbytes;
rctx->fallback_req.src = req->src;
return crypto_ahash_update(&rctx->fallback_req);
}
static int n2_hash_async_final(struct ahash_request *req)
{
struct n2_hash_req_ctx *rctx = ahash_request_ctx(req);
struct crypto_ahash *tfm = crypto_ahash_reqtfm(req);
struct n2_hash_ctx *ctx = crypto_ahash_ctx(tfm);
ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback_tfm);
rctx->fallback_req.base.flags = req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP;
rctx->fallback_req.result = req->result;
return crypto_ahash_final(&rctx->fallback_req);
}
static int n2_hash_async_finup(struct ahash_request *req)
{
struct n2_hash_req_ctx *rctx = ahash_request_ctx(req);
struct crypto_ahash *tfm = crypto_ahash_reqtfm(req);
struct n2_hash_ctx *ctx = crypto_ahash_ctx(tfm);
ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback_tfm);
rctx->fallback_req.base.flags = req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP;
rctx->fallback_req.nbytes = req->nbytes;
rctx->fallback_req.src = req->src;
rctx->fallback_req.result = req->result;
return crypto_ahash_finup(&rctx->fallback_req);
}
static int n2_hash_cra_init(struct crypto_tfm *tfm)
{
const char *fallback_driver_name = tfm->__crt_alg->cra_name;
struct crypto_ahash *ahash = __crypto_ahash_cast(tfm);
struct n2_hash_ctx *ctx = crypto_ahash_ctx(ahash);
struct crypto_ahash *fallback_tfm;
int err;
fallback_tfm = crypto_alloc_ahash(fallback_driver_name, 0,
CRYPTO_ALG_NEED_FALLBACK);
if (IS_ERR(fallback_tfm)) {
pr_warning("Fallback driver '%s' could not be loaded!\n",
fallback_driver_name);
err = PTR_ERR(fallback_tfm);
goto out;
}
crypto_ahash_set_reqsize(ahash, (sizeof(struct n2_hash_req_ctx) +
crypto_ahash_reqsize(fallback_tfm)));
ctx->fallback_tfm = fallback_tfm;
return 0;
out:
return err;
}
static void n2_hash_cra_exit(struct crypto_tfm *tfm)
{
struct crypto_ahash *ahash = __crypto_ahash_cast(tfm);
struct n2_hash_ctx *ctx = crypto_ahash_ctx(ahash);
crypto_free_ahash(ctx->fallback_tfm);
}
static int n2_hmac_cra_init(struct crypto_tfm *tfm)
{
const char *fallback_driver_name = tfm->__crt_alg->cra_name;
struct crypto_ahash *ahash = __crypto_ahash_cast(tfm);
struct n2_hmac_ctx *ctx = crypto_ahash_ctx(ahash);
struct n2_hmac_alg *n2alg = n2_hmac_alg(tfm);
struct crypto_ahash *fallback_tfm;
struct crypto_shash *child_shash;
int err;
fallback_tfm = crypto_alloc_ahash(fallback_driver_name, 0,
CRYPTO_ALG_NEED_FALLBACK);
if (IS_ERR(fallback_tfm)) {
pr_warning("Fallback driver '%s' could not be loaded!\n",
fallback_driver_name);
err = PTR_ERR(fallback_tfm);
goto out;
}
child_shash = crypto_alloc_shash(n2alg->child_alg, 0, 0);
if (IS_ERR(child_shash)) {
pr_warning("Child shash '%s' could not be loaded!\n",
n2alg->child_alg);
err = PTR_ERR(child_shash);
goto out_free_fallback;
}
crypto_ahash_set_reqsize(ahash, (sizeof(struct n2_hash_req_ctx) +
crypto_ahash_reqsize(fallback_tfm)));
ctx->child_shash = child_shash;
ctx->base.fallback_tfm = fallback_tfm;
return 0;
out_free_fallback:
crypto_free_ahash(fallback_tfm);
out:
return err;
}
static void n2_hmac_cra_exit(struct crypto_tfm *tfm)
{
struct crypto_ahash *ahash = __crypto_ahash_cast(tfm);
struct n2_hmac_ctx *ctx = crypto_ahash_ctx(ahash);
crypto_free_ahash(ctx->base.fallback_tfm);
crypto_free_shash(ctx->child_shash);
}
static int n2_hmac_async_setkey(struct crypto_ahash *tfm, const u8 *key,
unsigned int keylen)
{
struct n2_hmac_ctx *ctx = crypto_ahash_ctx(tfm);
struct crypto_shash *child_shash = ctx->child_shash;
struct crypto_ahash *fallback_tfm;
struct {
struct shash_desc shash;
char ctx[crypto_shash_descsize(child_shash)];
} desc;
int err, bs, ds;
fallback_tfm = ctx->base.fallback_tfm;
err = crypto_ahash_setkey(fallback_tfm, key, keylen);
if (err)
return err;
desc.shash.tfm = child_shash;
desc.shash.flags = crypto_ahash_get_flags(tfm) &
CRYPTO_TFM_REQ_MAY_SLEEP;
bs = crypto_shash_blocksize(child_shash);
ds = crypto_shash_digestsize(child_shash);
BUG_ON(ds > N2_HASH_KEY_MAX);
if (keylen > bs) {
err = crypto_shash_digest(&desc.shash, key, keylen,
ctx->hash_key);
if (err)
return err;
keylen = ds;
} else if (keylen <= N2_HASH_KEY_MAX)
memcpy(ctx->hash_key, key, keylen);
ctx->hash_key_len = keylen;
return err;
}
static unsigned long wait_for_tail(struct spu_queue *qp)
{
unsigned long head, hv_ret;
do {
hv_ret = sun4v_ncs_gethead(qp->qhandle, &head);
if (hv_ret != HV_EOK) {
pr_err("Hypervisor error on gethead\n");
break;
}
if (head == qp->tail) {
qp->head = head;
break;
}
} while (1);
return hv_ret;
}
static unsigned long submit_and_wait_for_tail(struct spu_queue *qp,
struct cwq_initial_entry *ent)
{
unsigned long hv_ret = spu_queue_submit(qp, ent);
if (hv_ret == HV_EOK)
hv_ret = wait_for_tail(qp);
return hv_ret;
}
static int n2_do_async_digest(struct ahash_request *req,
unsigned int auth_type, unsigned int digest_size,
unsigned int result_size, void *hash_loc,
unsigned long auth_key, unsigned int auth_key_len)
{
struct crypto_ahash *tfm = crypto_ahash_reqtfm(req);
struct cwq_initial_entry *ent;
struct crypto_hash_walk walk;
struct spu_queue *qp;
unsigned long flags;
int err = -ENODEV;
int nbytes, cpu;
/* The total effective length of the operation may not
* exceed 2^16.
*/
if (unlikely(req->nbytes > (1 << 16))) {
struct n2_hash_req_ctx *rctx = ahash_request_ctx(req);
struct n2_hash_ctx *ctx = crypto_ahash_ctx(tfm);
ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback_tfm);
rctx->fallback_req.base.flags =
req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP;
rctx->fallback_req.nbytes = req->nbytes;
rctx->fallback_req.src = req->src;
rctx->fallback_req.result = req->result;
return crypto_ahash_digest(&rctx->fallback_req);
}
nbytes = crypto_hash_walk_first(req, &walk);
cpu = get_cpu();
qp = cpu_to_cwq[cpu];
if (!qp)
goto out;
spin_lock_irqsave(&qp->lock, flags);
/* XXX can do better, improve this later by doing a by-hand scatterlist
* XXX walk, etc.
*/
ent = qp->q + qp->tail;
ent->control = control_word_base(nbytes, auth_key_len, 0,
auth_type, digest_size,
false, true, false, false,
OPCODE_INPLACE_BIT |
OPCODE_AUTH_MAC);
ent->src_addr = __pa(walk.data);
ent->auth_key_addr = auth_key;
ent->auth_iv_addr = __pa(hash_loc);
ent->final_auth_state_addr = 0UL;
ent->enc_key_addr = 0UL;
ent->enc_iv_addr = 0UL;
ent->dest_addr = __pa(hash_loc);
nbytes = crypto_hash_walk_done(&walk, 0);
while (nbytes > 0) {
ent = spu_queue_next(qp, ent);
ent->control = (nbytes - 1);
ent->src_addr = __pa(walk.data);
ent->auth_key_addr = 0UL;
ent->auth_iv_addr = 0UL;
ent->final_auth_state_addr = 0UL;
ent->enc_key_addr = 0UL;
ent->enc_iv_addr = 0UL;
ent->dest_addr = 0UL;
nbytes = crypto_hash_walk_done(&walk, 0);
}
ent->control |= CONTROL_END_OF_BLOCK;
if (submit_and_wait_for_tail(qp, ent) != HV_EOK)
err = -EINVAL;
else
err = 0;
spin_unlock_irqrestore(&qp->lock, flags);
if (!err)
memcpy(req->result, hash_loc, result_size);
out:
put_cpu();
return err;
}
static int n2_hash_async_digest(struct ahash_request *req)
{
struct n2_ahash_alg *n2alg = n2_ahash_alg(req->base.tfm);
struct n2_hash_req_ctx *rctx = ahash_request_ctx(req);
int ds;
ds = n2alg->digest_size;
if (unlikely(req->nbytes == 0)) {
memcpy(req->result, n2alg->hash_zero, ds);
return 0;
}
memcpy(&rctx->u, n2alg->hash_init, n2alg->hw_op_hashsz);
return n2_do_async_digest(req, n2alg->auth_type,
n2alg->hw_op_hashsz, ds,
&rctx->u, 0UL, 0);
}
static int n2_hmac_async_digest(struct ahash_request *req)
{
struct n2_hmac_alg *n2alg = n2_hmac_alg(req->base.tfm);
struct n2_hash_req_ctx *rctx = ahash_request_ctx(req);
struct crypto_ahash *tfm = crypto_ahash_reqtfm(req);
struct n2_hmac_ctx *ctx = crypto_ahash_ctx(tfm);
int ds;
ds = n2alg->derived.digest_size;
if (unlikely(req->nbytes == 0) ||
unlikely(ctx->hash_key_len > N2_HASH_KEY_MAX)) {
struct n2_hash_req_ctx *rctx = ahash_request_ctx(req);
struct n2_hash_ctx *ctx = crypto_ahash_ctx(tfm);
ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback_tfm);
rctx->fallback_req.base.flags =
req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP;
rctx->fallback_req.nbytes = req->nbytes;
rctx->fallback_req.src = req->src;
rctx->fallback_req.result = req->result;
return crypto_ahash_digest(&rctx->fallback_req);
}
memcpy(&rctx->u, n2alg->derived.hash_init,
n2alg->derived.hw_op_hashsz);
return n2_do_async_digest(req, n2alg->derived.hmac_type,
n2alg->derived.hw_op_hashsz, ds,
&rctx->u,
__pa(&ctx->hash_key),
ctx->hash_key_len);
}
struct n2_cipher_context {
int key_len;
int enc_type;
union {
u8 aes[AES_MAX_KEY_SIZE];
u8 des[DES_KEY_SIZE];
u8 des3[3 * DES_KEY_SIZE];
u8 arc4[258]; /* S-box, X, Y */
} key;
};
#define N2_CHUNK_ARR_LEN 16
struct n2_crypto_chunk {
struct list_head entry;
unsigned long iv_paddr : 44;
unsigned long arr_len : 20;
unsigned long dest_paddr;
unsigned long dest_final;
struct {
unsigned long src_paddr : 44;
unsigned long src_len : 20;
} arr[N2_CHUNK_ARR_LEN];
};
struct n2_request_context {
struct ablkcipher_walk walk;
struct list_head chunk_list;
struct n2_crypto_chunk chunk;
u8 temp_iv[16];
};
/* The SPU allows some level of flexibility for partial cipher blocks
* being specified in a descriptor.
*
* It merely requires that every descriptor's length field is at least
* as large as the cipher block size. This means that a cipher block
* can span at most 2 descriptors. However, this does not allow a
* partial block to span into the final descriptor as that would
* violate the rule (since every descriptor's length must be at lest
* the block size). So, for example, assuming an 8 byte block size:
*
* 0xe --> 0xa --> 0x8
*
* is a valid length sequence, whereas:
*
* 0xe --> 0xb --> 0x7
*
* is not a valid sequence.
*/
struct n2_cipher_alg {
struct list_head entry;
u8 enc_type;
struct crypto_alg alg;
};
static inline struct n2_cipher_alg *n2_cipher_alg(struct crypto_tfm *tfm)
{
struct crypto_alg *alg = tfm->__crt_alg;
return container_of(alg, struct n2_cipher_alg, alg);
}
struct n2_cipher_request_context {
struct ablkcipher_walk walk;
};
static int n2_aes_setkey(struct crypto_ablkcipher *cipher, const u8 *key,
unsigned int keylen)
{
struct crypto_tfm *tfm = crypto_ablkcipher_tfm(cipher);
struct n2_cipher_context *ctx = crypto_tfm_ctx(tfm);
struct n2_cipher_alg *n2alg = n2_cipher_alg(tfm);
ctx->enc_type = (n2alg->enc_type & ENC_TYPE_CHAINING_MASK);
switch (keylen) {
case AES_KEYSIZE_128:
ctx->enc_type |= ENC_TYPE_ALG_AES128;
break;
case AES_KEYSIZE_192:
ctx->enc_type |= ENC_TYPE_ALG_AES192;
break;
case AES_KEYSIZE_256:
ctx->enc_type |= ENC_TYPE_ALG_AES256;
break;
default:
crypto_ablkcipher_set_flags(cipher, CRYPTO_TFM_RES_BAD_KEY_LEN);
return -EINVAL;
}
ctx->key_len = keylen;
memcpy(ctx->key.aes, key, keylen);
return 0;
}
static int n2_des_setkey(struct crypto_ablkcipher *cipher, const u8 *key,
unsigned int keylen)
{
struct crypto_tfm *tfm = crypto_ablkcipher_tfm(cipher);
struct n2_cipher_context *ctx = crypto_tfm_ctx(tfm);
struct n2_cipher_alg *n2alg = n2_cipher_alg(tfm);
u32 tmp[DES_EXPKEY_WORDS];
int err;
ctx->enc_type = n2alg->enc_type;
if (keylen != DES_KEY_SIZE) {
crypto_ablkcipher_set_flags(cipher, CRYPTO_TFM_RES_BAD_KEY_LEN);
return -EINVAL;
}
err = des_ekey(tmp, key);
if (err == 0 && (tfm->crt_flags & CRYPTO_TFM_REQ_WEAK_KEY)) {
tfm->crt_flags |= CRYPTO_TFM_RES_WEAK_KEY;
return -EINVAL;
}
ctx->key_len = keylen;
memcpy(ctx->key.des, key, keylen);
return 0;
}
static int n2_3des_setkey(struct crypto_ablkcipher *cipher, const u8 *key,
unsigned int keylen)
{
struct crypto_tfm *tfm = crypto_ablkcipher_tfm(cipher);
struct n2_cipher_context *ctx = crypto_tfm_ctx(tfm);
struct n2_cipher_alg *n2alg = n2_cipher_alg(tfm);
ctx->enc_type = n2alg->enc_type;
if (keylen != (3 * DES_KEY_SIZE)) {
crypto_ablkcipher_set_flags(cipher, CRYPTO_TFM_RES_BAD_KEY_LEN);
return -EINVAL;
}
ctx->key_len = keylen;
memcpy(ctx->key.des3, key, keylen);
return 0;
}
static int n2_arc4_setkey(struct crypto_ablkcipher *cipher, const u8 *key,
unsigned int keylen)
{
struct crypto_tfm *tfm = crypto_ablkcipher_tfm(cipher);
struct n2_cipher_context *ctx = crypto_tfm_ctx(tfm);
struct n2_cipher_alg *n2alg = n2_cipher_alg(tfm);
u8 *s = ctx->key.arc4;
u8 *x = s + 256;
u8 *y = x + 1;
int i, j, k;
ctx->enc_type = n2alg->enc_type;
j = k = 0;
*x = 0;
*y = 0;
for (i = 0; i < 256; i++)
s[i] = i;
for (i = 0; i < 256; i++) {
u8 a = s[i];
j = (j + key[k] + a) & 0xff;
s[i] = s[j];
s[j] = a;
if (++k >= keylen)
k = 0;
}
return 0;
}
static inline int cipher_descriptor_len(int nbytes, unsigned int block_size)
{
int this_len = nbytes;
this_len -= (nbytes & (block_size - 1));
return this_len > (1 << 16) ? (1 << 16) : this_len;
}
static int __n2_crypt_chunk(struct crypto_tfm *tfm, struct n2_crypto_chunk *cp,
struct spu_queue *qp, bool encrypt)
{
struct n2_cipher_context *ctx = crypto_tfm_ctx(tfm);
struct cwq_initial_entry *ent;
bool in_place;
int i;
ent = spu_queue_alloc(qp, cp->arr_len);
if (!ent) {
pr_info("queue_alloc() of %d fails\n",
cp->arr_len);
return -EBUSY;
}
in_place = (cp->dest_paddr == cp->arr[0].src_paddr);
ent->control = control_word_base(cp->arr[0].src_len,
0, ctx->enc_type, 0, 0,
false, true, false, encrypt,
OPCODE_ENCRYPT |
(in_place ? OPCODE_INPLACE_BIT : 0));
ent->src_addr = cp->arr[0].src_paddr;
ent->auth_key_addr = 0UL;
ent->auth_iv_addr = 0UL;
ent->final_auth_state_addr = 0UL;
ent->enc_key_addr = __pa(&ctx->key);
ent->enc_iv_addr = cp->iv_paddr;
ent->dest_addr = (in_place ? 0UL : cp->dest_paddr);
for (i = 1; i < cp->arr_len; i++) {
ent = spu_queue_next(qp, ent);
ent->control = cp->arr[i].src_len - 1;
ent->src_addr = cp->arr[i].src_paddr;
ent->auth_key_addr = 0UL;
ent->auth_iv_addr = 0UL;
ent->final_auth_state_addr = 0UL;
ent->enc_key_addr = 0UL;
ent->enc_iv_addr = 0UL;
ent->dest_addr = 0UL;
}
ent->control |= CONTROL_END_OF_BLOCK;
return (spu_queue_submit(qp, ent) != HV_EOK) ? -EINVAL : 0;
}
static int n2_compute_chunks(struct ablkcipher_request *req)
{
struct n2_request_context *rctx = ablkcipher_request_ctx(req);
struct ablkcipher_walk *walk = &rctx->walk;
struct n2_crypto_chunk *chunk;
unsigned long dest_prev;
unsigned int tot_len;
bool prev_in_place;
int err, nbytes;
ablkcipher_walk_init(walk, req->dst, req->src, req->nbytes);
err = ablkcipher_walk_phys(req, walk);
if (err)
return err;
INIT_LIST_HEAD(&rctx->chunk_list);
chunk = &rctx->chunk;
INIT_LIST_HEAD(&chunk->entry);
chunk->iv_paddr = 0UL;
chunk->arr_len = 0;
chunk->dest_paddr = 0UL;
prev_in_place = false;
dest_prev = ~0UL;
tot_len = 0;
while ((nbytes = walk->nbytes) != 0) {
unsigned long dest_paddr, src_paddr;
bool in_place;
int this_len;
src_paddr = (page_to_phys(walk->src.page) +
walk->src.offset);
dest_paddr = (page_to_phys(walk->dst.page) +
walk->dst.offset);
in_place = (src_paddr == dest_paddr);
this_len = cipher_descriptor_len(nbytes, walk->blocksize);
if (chunk->arr_len != 0) {
if (in_place != prev_in_place ||
(!prev_in_place &&
dest_paddr != dest_prev) ||
chunk->arr_len == N2_CHUNK_ARR_LEN ||
tot_len + this_len > (1 << 16)) {
chunk->dest_final = dest_prev;
list_add_tail(&chunk->entry,
&rctx->chunk_list);
chunk = kzalloc(sizeof(*chunk), GFP_ATOMIC);
if (!chunk) {
err = -ENOMEM;
break;
}
INIT_LIST_HEAD(&chunk->entry);
}
}
if (chunk->arr_len == 0) {
chunk->dest_paddr = dest_paddr;
tot_len = 0;
}
chunk->arr[chunk->arr_len].src_paddr = src_paddr;
chunk->arr[chunk->arr_len].src_len = this_len;
chunk->arr_len++;
dest_prev = dest_paddr + this_len;
prev_in_place = in_place;
tot_len += this_len;
err = ablkcipher_walk_done(req, walk, nbytes - this_len);
if (err)
break;
}
if (!err && chunk->arr_len != 0) {
chunk->dest_final = dest_prev;
list_add_tail(&chunk->entry, &rctx->chunk_list);
}
return err;
}
static void n2_chunk_complete(struct ablkcipher_request *req, void *final_iv)
{
struct n2_request_context *rctx = ablkcipher_request_ctx(req);
struct n2_crypto_chunk *c, *tmp;
if (final_iv)
memcpy(rctx->walk.iv, final_iv, rctx->walk.blocksize);
ablkcipher_walk_complete(&rctx->walk);
list_for_each_entry_safe(c, tmp, &rctx->chunk_list, entry) {
list_del(&c->entry);
if (unlikely(c != &rctx->chunk))
kfree(c);
}
}
static int n2_do_ecb(struct ablkcipher_request *req, bool encrypt)
{
struct n2_request_context *rctx = ablkcipher_request_ctx(req);
struct crypto_tfm *tfm = req->base.tfm;
int err = n2_compute_chunks(req);
struct n2_crypto_chunk *c, *tmp;
unsigned long flags, hv_ret;
struct spu_queue *qp;
if (err)
return err;
qp = cpu_to_cwq[get_cpu()];
err = -ENODEV;
if (!qp)
goto out;
spin_lock_irqsave(&qp->lock, flags);
list_for_each_entry_safe(c, tmp, &rctx->chunk_list, entry) {
err = __n2_crypt_chunk(tfm, c, qp, encrypt);
if (err)
break;
list_del(&c->entry);
if (unlikely(c != &rctx->chunk))
kfree(c);
}