summaryrefslogtreecommitdiffstats
path: root/drivers/dma/sa11x0-dma.c
blob: 01e656c69e6c1f79657f09d94d676d885d5a1b2f (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
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
// SPDX-License-Identifier: GPL-2.0-only
/*
 * SA11x0 DMAengine support
 *
 * Copyright (C) 2012 Russell King
 *   Derived in part from arch/arm/mach-sa1100/dma.c,
 *   Copyright (C) 2000, 2001 by Nicolas Pitre
 */
#include <linux/sched.h>
#include <linux/device.h>
#include <linux/dmaengine.h>
#include <linux/init.h>
#include <linux/interrupt.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/platform_device.h>
#include <linux/slab.h>
#include <linux/spinlock.h>

#include "virt-dma.h"

#define NR_PHY_CHAN	6
#define DMA_ALIGN	3
#define DMA_MAX_SIZE	0x1fff
#define DMA_CHUNK_SIZE	0x1000

#define DMA_DDAR	0x00
#define DMA_DCSR_S	0x04
#define DMA_DCSR_C	0x08
#define DMA_DCSR_R	0x0c
#define DMA_DBSA	0x10
#define DMA_DBTA	0x14
#define DMA_DBSB	0x18
#define DMA_DBTB	0x1c
#define DMA_SIZE	0x20

#define DCSR_RUN	(1 << 0)
#define DCSR_IE		(1 << 1)
#define DCSR_ERROR	(1 << 2)
#define DCSR_DONEA	(1 << 3)
#define DCSR_STRTA	(1 << 4)
#define DCSR_DONEB	(1 << 5)
#define DCSR_STRTB	(1 << 6)
#define DCSR_BIU	(1 << 7)

#define DDAR_RW		(1 << 0)	/* 0 = W, 1 = R */
#define DDAR_E		(1 << 1)	/* 0 = LE, 1 = BE */
#define DDAR_BS		(1 << 2)	/* 0 = BS4, 1 = BS8 */
#define DDAR_DW		(1 << 3)	/* 0 = 8b, 1 = 16b */
#define DDAR_Ser0UDCTr	(0x0 << 4)
#define DDAR_Ser0UDCRc	(0x1 << 4)
#define DDAR_Ser1SDLCTr	(0x2 << 4)
#define DDAR_Ser1SDLCRc	(0x3 << 4)
#define DDAR_Ser1UARTTr	(0x4 << 4)
#define DDAR_Ser1UARTRc	(0x5 << 4)
#define DDAR_Ser2ICPTr	(0x6 << 4)
#define DDAR_Ser2ICPRc	(0x7 << 4)
#define DDAR_Ser3UARTTr	(0x8 << 4)
#define DDAR_Ser3UARTRc	(0x9 << 4)
#define DDAR_Ser4MCP0Tr	(0xa << 4)
#define DDAR_Ser4MCP0Rc	(0xb << 4)
#define DDAR_Ser4MCP1Tr	(0xc << 4)
#define DDAR_Ser4MCP1Rc	(0xd << 4)
#define DDAR_Ser4SSPTr	(0xe << 4)
#define DDAR_Ser4SSPRc	(0xf << 4)

struct sa11x0_dma_sg {
	u32			addr;
	u32			len;
};

struct sa11x0_dma_desc {
	struct virt_dma_desc	vd;

	u32			ddar;
	size_t			size;
	unsigned		period;
	bool			cyclic;

	unsigned		sglen;
	struct sa11x0_dma_sg	sg[] __counted_by(sglen);
};

struct sa11x0_dma_phy;

struct sa11x0_dma_chan {
	struct virt_dma_chan	vc;

	/* protected by c->vc.lock */
	struct sa11x0_dma_phy	*phy;
	enum dma_status		status;

	/* protected by d->lock */
	struct list_head	node;

	u32			ddar;
	const char		*name;
};

struct sa11x0_dma_phy {
	void __iomem		*base;
	struct sa11x0_dma_dev	*dev;
	unsigned		num;

	struct sa11x0_dma_chan	*vchan;

	/* Protected by c->vc.lock */
	unsigned		sg_load;
	struct sa11x0_dma_desc	*txd_load;
	unsigned		sg_done;
	struct sa11x0_dma_desc	*txd_done;
	u32			dbs[2];
	u32			dbt[2];
	u32			dcsr;
};

struct sa11x0_dma_dev {
	struct dma_device	slave;
	void __iomem		*base;
	spinlock_t		lock;
	struct tasklet_struct	task;
	struct list_head	chan_pending;
	struct sa11x0_dma_phy	phy[NR_PHY_CHAN];
};

static struct sa11x0_dma_chan *to_sa11x0_dma_chan(struct dma_chan *chan)
{
	return container_of(chan, struct sa11x0_dma_chan, vc.chan);
}

static struct sa11x0_dma_dev *to_sa11x0_dma(struct dma_device *dmadev)
{
	return container_of(dmadev, struct sa11x0_dma_dev, slave);
}

static struct sa11x0_dma_desc *sa11x0_dma_next_desc(struct sa11x0_dma_chan *c)
{
	struct virt_dma_desc *vd = vchan_next_desc(&c->vc);

	return vd ? container_of(vd, struct sa11x0_dma_desc, vd) : NULL;
}

static void sa11x0_dma_free_desc(struct virt_dma_desc *vd)
{
	kfree(container_of(vd, struct sa11x0_dma_desc, vd));
}

static void sa11x0_dma_start_desc(struct sa11x0_dma_phy *p, struct sa11x0_dma_desc *txd)
{
	list_del(&txd->vd.node);
	p->txd_load = txd;
	p->sg_load = 0;

	dev_vdbg(p->dev->slave.dev, "pchan %u: txd %p[%x]: starting: DDAR:%x\n",
		p->num, &txd->vd, txd->vd.tx.cookie, txd->ddar);
}

static void noinline sa11x0_dma_start_sg(struct sa11x0_dma_phy *p,
	struct sa11x0_dma_chan *c)
{
	struct sa11x0_dma_desc *txd = p->txd_load;
	struct sa11x0_dma_sg *sg;
	void __iomem *base = p->base;
	unsigned dbsx, dbtx;
	u32 dcsr;

	if (!txd)
		return;

	dcsr = readl_relaxed(base + DMA_DCSR_R);

	/* Don't try to load the next transfer if both buffers are started */
	if ((dcsr & (DCSR_STRTA | DCSR_STRTB)) == (DCSR_STRTA | DCSR_STRTB))
		return;

	if (p->sg_load == txd->sglen) {
		if (!txd->cyclic) {
			struct sa11x0_dma_desc *txn = sa11x0_dma_next_desc(c);

			/*
			 * We have reached the end of the current descriptor.
			 * Peek at the next descriptor, and if compatible with
			 * the current, start processing it.
			 */
			if (txn && txn->ddar == txd->ddar) {
				txd = txn;
				sa11x0_dma_start_desc(p, txn);
			} else {
				p->txd_load = NULL;
				return;
			}
		} else {
			/* Cyclic: reset back to beginning */
			p->sg_load = 0;
		}
	}

	sg = &txd->sg[p->sg_load++];

	/* Select buffer to load according to channel status */
	if (((dcsr & (DCSR_BIU | DCSR_STRTB)) == (DCSR_BIU | DCSR_STRTB)) ||
	    ((dcsr & (DCSR_BIU | DCSR_STRTA)) == 0)) {
		dbsx = DMA_DBSA;
		dbtx = DMA_DBTA;
		dcsr = DCSR_STRTA | DCSR_IE | DCSR_RUN;
	} else {
		dbsx = DMA_DBSB;
		dbtx = DMA_DBTB;
		dcsr = DCSR_STRTB | DCSR_IE | DCSR_RUN;
	}

	writel_relaxed(sg->addr, base + dbsx);
	writel_relaxed(sg->len, base + dbtx);
	writel(dcsr, base + DMA_DCSR_S);

	dev_dbg(p->dev->slave.dev, "pchan %u: load: DCSR:%02x DBS%c:%08x DBT%c:%08x\n",
		p->num, dcsr,
		'A' + (dbsx == DMA_DBSB), sg->addr,
		'A' + (dbtx == DMA_DBTB), sg->len);
}

static void noinline sa11x0_dma_complete(struct sa11x0_dma_phy *p,
	struct sa11x0_dma_chan *c)
{
	struct sa11x0_dma_desc *txd = p->txd_done;

	if (++p->sg_done == txd->sglen) {
		if (!txd->cyclic) {
			vchan_cookie_complete(&txd->vd);

			p->sg_done = 0;
			p->txd_done = p->txd_load;

			if (!p->txd_done)
				tasklet_schedule(&p->dev->task);
		} else {
			if ((p->sg_done % txd->period) == 0)
				vchan_cyclic_callback(&txd->vd);

			/* Cyclic: reset back to beginning */
			p->sg_done = 0;
		}
	}

	sa11x0_dma_start_sg(p, c);
}

static irqreturn_t sa11x0_dma_irq(int irq, void *dev_id)
{
	struct sa11x0_dma_phy *p = dev_id;
	struct sa11x0_dma_dev *d = p->dev;
	struct sa11x0_dma_chan *c;
	u32 dcsr;

	dcsr = readl_relaxed(p->base + DMA_DCSR_R);
	if (!(dcsr & (DCSR_ERROR | DCSR_DONEA | DCSR_DONEB)))
		return IRQ_NONE;

	/* Clear reported status bits */
	writel_relaxed(dcsr & (DCSR_ERROR | DCSR_DONEA | DCSR_DONEB),
		p->base + DMA_DCSR_C);

	dev_dbg(d->slave.dev, "pchan %u: irq: DCSR:%02x\n", p->num, dcsr);

	if (dcsr & DCSR_ERROR) {
		dev_err(d->slave.dev, "pchan %u: error. DCSR:%02x DDAR:%08x DBSA:%08x DBTA:%08x DBSB:%08x DBTB:%08x\n",
			p->num, dcsr,
			readl_relaxed(p->base + DMA_DDAR),
			readl_relaxed(p->base + DMA_DBSA),
			readl_relaxed(p->base + DMA_DBTA),
			readl_relaxed(p->base + DMA_DBSB),
			readl_relaxed(p->base + DMA_DBTB));
	}

	c = p->vchan;
	if (c) {
		unsigned long flags;

		spin_lock_irqsave(&c->vc.lock, flags);
		/*
		 * Now that we're holding the lock, check that the vchan
		 * really is associated with this pchan before touching the
		 * hardware.  This should always succeed, because we won't
		 * change p->vchan or c->phy while the channel is actively
		 * transferring.
		 */
		if (c->phy == p) {
			if (dcsr & DCSR_DONEA)
				sa11x0_dma_complete(p, c);
			if (dcsr & DCSR_DONEB)
				sa11x0_dma_complete(p, c);
		}
		spin_unlock_irqrestore(&c->vc.lock, flags);
	}

	return IRQ_HANDLED;
}

static void sa11x0_dma_start_txd(struct sa11x0_dma_chan *c)
{
	struct sa11x0_dma_desc *txd = sa11x0_dma_next_desc(c);

	/* If the issued list is empty, we have no further txds to process */
	if (txd) {
		struct sa11x0_dma_phy *p = c->phy;

		sa11x0_dma_start_desc(p, txd);
		p->txd_done = txd;
		p->sg_done = 0;

		/* The channel should not have any transfers started */
		WARN_ON(readl_relaxed(p->base + DMA_DCSR_R) &
				      (DCSR_STRTA | DCSR_STRTB));

		/* Clear the run and start bits before changing DDAR */
		writel_relaxed(DCSR_RUN | DCSR_STRTA | DCSR_STRTB,
			       p->base + DMA_DCSR_C);
		writel_relaxed(txd->ddar, p->base + DMA_DDAR);

		/* Try to start both buffers */
		sa11x0_dma_start_sg(p, c);
		sa11x0_dma_start_sg(p, c);
	}
}

static void sa11x0_dma_tasklet(struct tasklet_struct *t)
{
	struct sa11x0_dma_dev *d = from_tasklet(d, t, task);
	struct sa11x0_dma_phy *p;
	struct sa11x0_dma_chan *c;
	unsigned pch, pch_alloc = 0;

	dev_dbg(d->slave.dev, "tasklet enter\n");

	list_for_each_entry(c, &d->slave.channels, vc.chan.device_node) {
		spin_lock_irq(&c->vc.lock);
		p = c->phy;
		if (p && !p->txd_done) {
			sa11x0_dma_start_txd(c);
			if (!p->txd_done) {
				/* No current txd associated with this channel */
				dev_dbg(d->slave.dev, "pchan %u: free\n", p->num);

				/* Mark this channel free */
				c->phy = NULL;
				p->vchan = NULL;
			}
		}
		spin_unlock_irq(&c->vc.lock);
	}

	spin_lock_irq(&d->lock);
	for (pch = 0; pch < NR_PHY_CHAN; pch++) {
		p = &d->phy[pch];

		if (p->vchan == NULL && !list_empty(&d->chan_pending)) {
			c = list_first_entry(&d->chan_pending,
				struct sa11x0_dma_chan, node);
			list_del_init(&c->node);

			pch_alloc |= 1 << pch;

			/* Mark this channel allocated */
			p->vchan = c;

			dev_dbg(d->slave.dev, "pchan %u: alloc vchan %p\n", pch, &c->vc);
		}
	}
	spin_unlock_irq(&d->lock);

	for (pch = 0; pch < NR_PHY_CHAN; pch++) {
		if (pch_alloc & (1 << pch)) {
			p = &d->phy[pch];
			c = p->vchan;

			spin_lock_irq(&c->vc.lock);
			c->phy = p;

			sa11x0_dma_start_txd(c);
			spin_unlock_irq(&c->vc.lock);
		}
	}

	dev_dbg(d->slave.dev, "tasklet exit\n");
}


static void sa11x0_dma_free_chan_resources(struct dma_chan *chan)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
	unsigned long flags;

	spin_lock_irqsave(&d->lock, flags);
	list_del_init(&c->node);
	spin_unlock_irqrestore(&d->lock, flags);

	vchan_free_chan_resources(&c->vc);
}

static dma_addr_t sa11x0_dma_pos(struct sa11x0_dma_phy *p)
{
	unsigned reg;
	u32 dcsr;

	dcsr = readl_relaxed(p->base + DMA_DCSR_R);

	if ((dcsr & (DCSR_BIU | DCSR_STRTA)) == DCSR_STRTA ||
	    (dcsr & (DCSR_BIU | DCSR_STRTB)) == DCSR_BIU)
		reg = DMA_DBSA;
	else
		reg = DMA_DBSB;

	return readl_relaxed(p->base + reg);
}

static enum dma_status sa11x0_dma_tx_status(struct dma_chan *chan,
	dma_cookie_t cookie, struct dma_tx_state *state)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
	struct sa11x0_dma_phy *p;
	struct virt_dma_desc *vd;
	unsigned long flags;
	enum dma_status ret;

	ret = dma_cookie_status(&c->vc.chan, cookie, state);
	if (ret == DMA_COMPLETE)
		return ret;

	if (!state)
		return c->status;

	spin_lock_irqsave(&c->vc.lock, flags);
	p = c->phy;

	/*
	 * If the cookie is on our issue queue, then the residue is
	 * its total size.
	 */
	vd = vchan_find_desc(&c->vc, cookie);
	if (vd) {
		state->residue = container_of(vd, struct sa11x0_dma_desc, vd)->size;
	} else if (!p) {
		state->residue = 0;
	} else {
		struct sa11x0_dma_desc *txd;
		size_t bytes = 0;

		if (p->txd_done && p->txd_done->vd.tx.cookie == cookie)
			txd = p->txd_done;
		else if (p->txd_load && p->txd_load->vd.tx.cookie == cookie)
			txd = p->txd_load;
		else
			txd = NULL;

		ret = c->status;
		if (txd) {
			dma_addr_t addr = sa11x0_dma_pos(p);
			unsigned i;

			dev_vdbg(d->slave.dev, "tx_status: addr:%pad\n", &addr);

			for (i = 0; i < txd->sglen; i++) {
				dev_vdbg(d->slave.dev, "tx_status: [%u] %x+%x\n",
					i, txd->sg[i].addr, txd->sg[i].len);
				if (addr >= txd->sg[i].addr &&
				    addr < txd->sg[i].addr + txd->sg[i].len) {
					unsigned len;

					len = txd->sg[i].len -
						(addr - txd->sg[i].addr);
					dev_vdbg(d->slave.dev, "tx_status: [%u] +%x\n",
						i, len);
					bytes += len;
					i++;
					break;
				}
			}
			for (; i < txd->sglen; i++) {
				dev_vdbg(d->slave.dev, "tx_status: [%u] %x+%x ++\n",
					i, txd->sg[i].addr, txd->sg[i].len);
				bytes += txd->sg[i].len;
			}
		}
		state->residue = bytes;
	}
	spin_unlock_irqrestore(&c->vc.lock, flags);

	dev_vdbg(d->slave.dev, "tx_status: bytes 0x%x\n", state->residue);

	return ret;
}

/*
 * Move pending txds to the issued list, and re-init pending list.
 * If not already pending, add this channel to the list of pending
 * channels and trigger the tasklet to run.
 */
static void sa11x0_dma_issue_pending(struct dma_chan *chan)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
	unsigned long flags;

	spin_lock_irqsave(&c->vc.lock, flags);
	if (vchan_issue_pending(&c->vc)) {
		if (!c->phy) {
			spin_lock(&d->lock);
			if (list_empty(&c->node)) {
				list_add_tail(&c->node, &d->chan_pending);
				tasklet_schedule(&d->task);
				dev_dbg(d->slave.dev, "vchan %p: issued\n", &c->vc);
			}
			spin_unlock(&d->lock);
		}
	} else
		dev_dbg(d->slave.dev, "vchan %p: nothing to issue\n", &c->vc);
	spin_unlock_irqrestore(&c->vc.lock, flags);
}

static struct dma_async_tx_descriptor *sa11x0_dma_prep_slave_sg(
	struct dma_chan *chan, struct scatterlist *sg, unsigned int sglen,
	enum dma_transfer_direction dir, unsigned long flags, void *context)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	struct sa11x0_dma_desc *txd;
	struct scatterlist *sgent;
	unsigned i, j = sglen;
	size_t size = 0;

	/* SA11x0 channels can only operate in their native direction */
	if (dir != (c->ddar & DDAR_RW ? DMA_DEV_TO_MEM : DMA_MEM_TO_DEV)) {
		dev_err(chan->device->dev, "vchan %p: bad DMA direction: DDAR:%08x dir:%u\n",
			&c->vc, c->ddar, dir);
		return NULL;
	}

	/* Do not allow zero-sized txds */
	if (sglen == 0)
		return NULL;

	for_each_sg(sg, sgent, sglen, i) {
		dma_addr_t addr = sg_dma_address(sgent);
		unsigned int len = sg_dma_len(sgent);

		if (len > DMA_MAX_SIZE)
			j += DIV_ROUND_UP(len, DMA_MAX_SIZE & ~DMA_ALIGN) - 1;
		if (addr & DMA_ALIGN) {
			dev_dbg(chan->device->dev, "vchan %p: bad buffer alignment: %pad\n",
				&c->vc, &addr);
			return NULL;
		}
	}

	txd = kzalloc(struct_size(txd, sg, j), GFP_ATOMIC);
	if (!txd) {
		dev_dbg(chan->device->dev, "vchan %p: kzalloc failed\n", &c->vc);
		return NULL;
	}
	txd->sglen = j;

	j = 0;
	for_each_sg(sg, sgent, sglen, i) {
		dma_addr_t addr = sg_dma_address(sgent);
		unsigned len = sg_dma_len(sgent);

		size += len;

		do {
			unsigned tlen = len;

			/*
			 * Check whether the transfer will fit.  If not, try
			 * to split the transfer up such that we end up with
			 * equal chunks - but make sure that we preserve the
			 * alignment.  This avoids small segments.
			 */
			if (tlen > DMA_MAX_SIZE) {
				unsigned mult = DIV_ROUND_UP(tlen,
					DMA_MAX_SIZE & ~DMA_ALIGN);

				tlen = (tlen / mult) & ~DMA_ALIGN;
			}

			txd->sg[j].addr = addr;
			txd->sg[j].len = tlen;

			addr += tlen;
			len -= tlen;
			j++;
		} while (len);
	}

	txd->ddar = c->ddar;
	txd->size = size;

	dev_dbg(chan->device->dev, "vchan %p: txd %p: size %zu nr %u\n",
		&c->vc, &txd->vd, txd->size, txd->sglen);

	return vchan_tx_prep(&c->vc, &txd->vd, flags);
}

static struct dma_async_tx_descriptor *sa11x0_dma_prep_dma_cyclic(
	struct dma_chan *chan, dma_addr_t addr, size_t size, size_t period,
	enum dma_transfer_direction dir, unsigned long flags)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	struct sa11x0_dma_desc *txd;
	unsigned i, j, k, sglen, sgperiod;

	/* SA11x0 channels can only operate in their native direction */
	if (dir != (c->ddar & DDAR_RW ? DMA_DEV_TO_MEM : DMA_MEM_TO_DEV)) {
		dev_err(chan->device->dev, "vchan %p: bad DMA direction: DDAR:%08x dir:%u\n",
			&c->vc, c->ddar, dir);
		return NULL;
	}

	sgperiod = DIV_ROUND_UP(period, DMA_MAX_SIZE & ~DMA_ALIGN);
	sglen = size * sgperiod / period;

	/* Do not allow zero-sized txds */
	if (sglen == 0)
		return NULL;

	txd = kzalloc(struct_size(txd, sg, sglen), GFP_ATOMIC);
	if (!txd) {
		dev_dbg(chan->device->dev, "vchan %p: kzalloc failed\n", &c->vc);
		return NULL;
	}
	txd->sglen = sglen;

	for (i = k = 0; i < size / period; i++) {
		size_t tlen, len = period;

		for (j = 0; j < sgperiod; j++, k++) {
			tlen = len;

			if (tlen > DMA_MAX_SIZE) {
				unsigned mult = DIV_ROUND_UP(tlen, DMA_MAX_SIZE & ~DMA_ALIGN);
				tlen = (tlen / mult) & ~DMA_ALIGN;
			}

			txd->sg[k].addr = addr;
			txd->sg[k].len = tlen;
			addr += tlen;
			len -= tlen;
		}

		WARN_ON(len != 0);
	}

	WARN_ON(k != sglen);

	txd->ddar = c->ddar;
	txd->size = size;
	txd->cyclic = 1;
	txd->period = sgperiod;

	return vchan_tx_prep(&c->vc, &txd->vd, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
}

static int sa11x0_dma_device_config(struct dma_chan *chan,
				    struct dma_slave_config *cfg)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	u32 ddar = c->ddar & ((0xf << 4) | DDAR_RW);
	dma_addr_t addr;
	enum dma_slave_buswidth width;
	u32 maxburst;

	if (ddar & DDAR_RW) {
		addr = cfg->src_addr;
		width = cfg->src_addr_width;
		maxburst = cfg->src_maxburst;
	} else {
		addr = cfg->dst_addr;
		width = cfg->dst_addr_width;
		maxburst = cfg->dst_maxburst;
	}

	if ((width != DMA_SLAVE_BUSWIDTH_1_BYTE &&
	     width != DMA_SLAVE_BUSWIDTH_2_BYTES) ||
	    (maxburst != 4 && maxburst != 8))
		return -EINVAL;

	if (width == DMA_SLAVE_BUSWIDTH_2_BYTES)
		ddar |= DDAR_DW;
	if (maxburst == 8)
		ddar |= DDAR_BS;

	dev_dbg(c->vc.chan.device->dev, "vchan %p: dma_slave_config addr %pad width %u burst %u\n",
		&c->vc, &addr, width, maxburst);

	c->ddar = ddar | (addr & 0xf0000000) | (addr & 0x003ffffc) << 6;

	return 0;
}

static int sa11x0_dma_device_pause(struct dma_chan *chan)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
	struct sa11x0_dma_phy *p;
	unsigned long flags;

	dev_dbg(d->slave.dev, "vchan %p: pause\n", &c->vc);
	spin_lock_irqsave(&c->vc.lock, flags);
	if (c->status == DMA_IN_PROGRESS) {
		c->status = DMA_PAUSED;

		p = c->phy;
		if (p) {
			writel(DCSR_RUN | DCSR_IE, p->base + DMA_DCSR_C);
		} else {
			spin_lock(&d->lock);
			list_del_init(&c->node);
			spin_unlock(&d->lock);
		}
	}
	spin_unlock_irqrestore(&c->vc.lock, flags);

	return 0;
}

static int sa11x0_dma_device_resume(struct dma_chan *chan)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
	struct sa11x0_dma_phy *p;
	unsigned long flags;

	dev_dbg(d->slave.dev, "vchan %p: resume\n", &c->vc);
	spin_lock_irqsave(&c->vc.lock, flags);
	if (c->status == DMA_PAUSED) {
		c->status = DMA_IN_PROGRESS;

		p = c->phy;
		if (p) {
			writel(DCSR_RUN | DCSR_IE, p->base + DMA_DCSR_S);
		} else if (!list_empty(&c->vc.desc_issued)) {
			spin_lock(&d->lock);
			list_add_tail(&c->node, &d->chan_pending);
			spin_unlock(&d->lock);
		}
	}
	spin_unlock_irqrestore(&c->vc.lock, flags);

	return 0;
}

static int sa11x0_dma_device_terminate_all(struct dma_chan *chan)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
	struct sa11x0_dma_phy *p;
	LIST_HEAD(head);
	unsigned long flags;

	dev_dbg(d->slave.dev, "vchan %p: terminate all\n", &c->vc);
	/* Clear the tx descriptor lists */
	spin_lock_irqsave(&c->vc.lock, flags);
	vchan_get_all_descriptors(&c->vc, &head);

	p = c->phy;
	if (p) {
		dev_dbg(d->slave.dev, "pchan %u: terminating\n", p->num);
		/* vchan is assigned to a pchan - stop the channel */
		writel(DCSR_RUN | DCSR_IE |
		       DCSR_STRTA | DCSR_DONEA |
		       DCSR_STRTB | DCSR_DONEB,
		       p->base + DMA_DCSR_C);

		if (p->txd_load) {
			if (p->txd_load != p->txd_done)
				list_add_tail(&p->txd_load->vd.node, &head);
			p->txd_load = NULL;
		}
		if (p->txd_done) {
			list_add_tail(&p->txd_done->vd.node, &head);
			p->txd_done = NULL;
		}
		c->phy = NULL;
		spin_lock(&d->lock);
		p->vchan = NULL;
		spin_unlock(&d->lock);
		tasklet_schedule(&d->task);
	}
	spin_unlock_irqrestore(&c->vc.lock, flags);
	vchan_dma_desc_free_list(&c->vc, &head);

	return 0;
}

struct sa11x0_dma_channel_desc {
	u32 ddar;
	const char *name;
};

#define CD(d1, d2) { .ddar = DDAR_##d1 | d2, .name = #d1 }
static const struct sa11x0_dma_channel_desc chan_desc[] = {
	CD(Ser0UDCTr, 0),
	CD(Ser0UDCRc, DDAR_RW),
	CD(Ser1SDLCTr, 0),
	CD(Ser1SDLCRc, DDAR_RW),
	CD(Ser1UARTTr, 0),
	CD(Ser1UARTRc, DDAR_RW),
	CD(Ser2ICPTr, 0),
	CD(Ser2ICPRc, DDAR_RW),
	CD(Ser3UARTTr, 0),
	CD(Ser3UARTRc, DDAR_RW),
	CD(Ser4MCP0Tr, 0),
	CD(Ser4MCP0Rc, DDAR_RW),
	CD(Ser4MCP1Tr, 0),
	CD(Ser4MCP1Rc, DDAR_RW),
	CD(Ser4SSPTr, 0),
	CD(Ser4SSPRc, DDAR_RW),
};

static const struct dma_slave_map sa11x0_dma_map[] = {
	{ "sa11x0-ir", "tx", "Ser2ICPTr" },
	{ "sa11x0-ir", "rx", "Ser2ICPRc" },
	{ "sa11x0-ssp", "tx", "Ser4SSPTr" },
	{ "sa11x0-ssp", "rx", "Ser4SSPRc" },
};

static bool sa11x0_dma_filter_fn(struct dma_chan *chan, void *param)
{
	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
	const char *p = param;

	return !strcmp(c->name, p);
}

static int sa11x0_dma_init_dmadev(struct dma_device *dmadev,
	struct device *dev)
{
	unsigned i;

	INIT_LIST_HEAD(&dmadev->channels);
	dmadev->dev = dev;
	dmadev->device_free_chan_resources = sa11x0_dma_free_chan_resources;
	dmadev->device_config = sa11x0_dma_device_config;
	dmadev->device_pause = sa11x0_dma_device_pause;
	dmadev->device_resume = sa11x0_dma_device_resume;
	dmadev->device_terminate_all = sa11x0_dma_device_terminate_all;
	dmadev->device_tx_status = sa11x0_dma_tx_status;
	dmadev->device_issue_pending = sa11x0_dma_issue_pending;

	for (i = 0; i < ARRAY_SIZE(chan_desc); i++) {
		struct sa11x0_dma_chan *c;

		c = kzalloc(sizeof(*c), GFP_KERNEL);
		if (!c) {
			dev_err(dev, "no memory for channel %u\n", i);
			return -ENOMEM;
		}

		c->status = DMA_IN_PROGRESS;
		c->ddar = chan_desc[i].ddar;
		c->name = chan_desc[i].name;
		INIT_LIST_HEAD(&c->node);

		c->vc.desc_free = sa11x0_dma_free_desc;
		vchan_init(&c->vc, dmadev);
	}

	return dma_async_device_register(dmadev);
}

static int sa11x0_dma_request_irq(struct platform_device *pdev, int nr,
	void *data)
{
	int irq = platform_get_irq(pdev, nr);

	if (irq <= 0)
		return -ENXIO;

	return request_irq(irq, sa11x0_dma_irq, 0, dev_name(&pdev->dev), data);
}

static void sa11x0_dma_free_irq(struct platform_device *pdev, int nr,
	void *data)
{
	int irq = platform_get_irq(pdev, nr);
	if (irq > 0)
		free_irq(irq, data);
}

static void sa11x0_dma_free_channels(struct dma_device *dmadev)
{
	struct sa11x0_dma_chan *c, *cn;

	list_for_each_entry_safe(c, cn, &dmadev->channels, vc.chan.device_node) {
		list_del(&c->vc.chan.device_node);
		tasklet_kill(&c->vc.task);
		kfree(c);
	}
}

static int sa11x0_dma_probe(struct platform_device *pdev)
{
	struct sa11x0_dma_dev *d;
	struct resource *res;
	unsigned i;
	int ret;

	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
	if (!res)
		return -ENXIO;

	d = kzalloc(sizeof(*d), GFP_KERNEL);
	if (!d) {
		ret = -ENOMEM;
		goto err_alloc;
	}

	spin_lock_init(&d->lock);
	INIT_LIST_HEAD(&d->chan_pending);

	d->slave.filter.fn = sa11x0_dma_filter_fn;
	d->slave.filter.mapcnt = ARRAY_SIZE(sa11x0_dma_map);
	d->slave.filter.map = sa11x0_dma_map;

	d->base = ioremap(res->start, resource_size(res));
	if (!d->base) {
		ret = -ENOMEM;
		goto err_ioremap;
	}

	tasklet_setup(&d->task, sa11x0_dma_tasklet);

	for (i = 0; i < NR_PHY_CHAN; i++) {
		struct sa11x0_dma_phy *p = &d->phy[i];

		p->dev = d;
		p->num = i;
		p->base = d->base + i * DMA_SIZE;
		writel_relaxed(DCSR_RUN | DCSR_IE | DCSR_ERROR |
			DCSR_DONEA | DCSR_STRTA | DCSR_DONEB | DCSR_STRTB,
			p->base + DMA_DCSR_C);
		writel_relaxed(0, p->base + DMA_DDAR);

		ret = sa11x0_dma_request_irq(pdev, i, p);
		if (ret) {
			while (i) {
				i--;
				sa11x0_dma_free_irq(pdev, i, &d->phy[i]);
			}
			goto err_irq;
		}
	}

	dma_cap_set(DMA_SLAVE, d->slave.cap_mask);
	dma_cap_set(DMA_CYCLIC, d->slave.cap_mask);
	d->slave.device_prep_slave_sg = sa11x0_dma_prep_slave_sg;
	d->slave.device_prep_dma_cyclic = sa11x0_dma_prep_dma_cyclic;
	d->slave.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
	d->slave.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
	d->slave.src_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |
				   BIT(DMA_SLAVE_BUSWIDTH_2_BYTES);
	d->slave.dst_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |
				   BIT(DMA_SLAVE_BUSWIDTH_2_BYTES);
	ret = sa11x0_dma_init_dmadev(&d->slave, &pdev->dev);
	if (ret) {
		dev_warn(d->slave.dev, "failed to register slave async device: %d\n",
			ret);
		goto err_slave_reg;
	}

	platform_set_drvdata(pdev, d);
	return 0;

 err_slave_reg:
	sa11x0_dma_free_channels(&d->slave);
	for (i = 0; i < NR_PHY_CHAN; i++)
		sa11x0_dma_free_irq(pdev, i, &d->phy[i]);
 err_irq:
	tasklet_kill(&d->task);
	iounmap(d->base);
 err_ioremap:
	kfree(d);
 err_alloc:
	return ret;
}

static void sa11x0_dma_remove(struct platform_device *pdev)
{
	struct sa11x0_dma_dev *d = platform_get_drvdata(pdev);
	unsigned pch;

	dma_async_device_unregister(&d->slave);

	sa11x0_dma_free_channels(&d->slave);
	for (pch = 0; pch < NR_PHY_CHAN; pch++)
		sa11x0_dma_free_irq(pdev, pch, &d->phy[pch]);
	tasklet_kill(&d->task);
	iounmap(d->base);
	kfree(d);
}

static __maybe_unused int sa11x0_dma_suspend(struct device *dev)
{
	struct sa11x0_dma_dev *d = dev_get_drvdata(dev);
	unsigned pch;

	for (pch = 0; pch < NR_PHY_CHAN; pch++) {
		struct sa11x0_dma_phy *p = &d->phy[pch];
		u32 dcsr, saved_dcsr;

		dcsr = saved_dcsr = readl_relaxed(p->base + DMA_DCSR_R);
		if (dcsr & DCSR_RUN) {
			writel(DCSR_RUN | DCSR_IE, p->base + DMA_DCSR_C);
			dcsr = readl_relaxed(p->base + DMA_DCSR_R);
		}

		saved_dcsr &= DCSR_RUN | DCSR_IE;
		if (dcsr & DCSR_BIU) {
			p->dbs[0] = readl_relaxed(p->base + DMA_DBSB);
			p->dbt[0] = readl_relaxed(p->base + DMA_DBTB);
			p->dbs[1] = readl_relaxed(p->base + DMA_DBSA);
			p->dbt[1] = readl_relaxed(p->base + DMA_DBTA);
			saved_dcsr |= (dcsr & DCSR_STRTA ? DCSR_STRTB : 0) |
				      (dcsr & DCSR_STRTB ? DCSR_STRTA : 0);
		} else {
			p->dbs[0] = readl_relaxed(p->base + DMA_DBSA);
			p->dbt[0] = readl_relaxed(p->base + DMA_DBTA);
			p->dbs[1] = readl_relaxed(p->base + DMA_DBSB);
			p->dbt[1] = readl_relaxed(p->base + DMA_DBTB);
			saved_dcsr |= dcsr & (DCSR_STRTA | DCSR_STRTB);
		}
		p->dcsr = saved_dcsr;

		writel(DCSR_STRTA | DCSR_STRTB, p->base + DMA_DCSR_C);
	}

	return 0;
}

static __maybe_unused int sa11x0_dma_resume(struct device *dev)
{
	struct sa11x0_dma_dev *d = dev_get_drvdata(dev);
	unsigned pch;

	for (pch = 0; pch < NR_PHY_CHAN; pch++) {
		struct sa11x0_dma_phy *p = &d->phy[pch];
		struct sa11x0_dma_desc *txd = NULL;
		u32 dcsr = readl_relaxed(p->base + DMA_DCSR_R);

		WARN_ON(dcsr & (DCSR_BIU | DCSR_STRTA | DCSR_STRTB | DCSR_RUN));

		if (p->txd_done)
			txd = p->txd_done;
		else if (p->txd_load)
			txd = p->txd_load;

		if (!txd)
			continue;

		writel_relaxed(txd->ddar, p->base + DMA_DDAR);

		writel_relaxed(p->dbs[0], p->base + DMA_DBSA);
		writel_relaxed(p->dbt[0], p->base + DMA_DBTA);
		writel_relaxed(p->dbs[1], p->base + DMA_DBSB);
		writel_relaxed(p->dbt[1], p->base + DMA_DBTB);
		writel_relaxed(p->dcsr, p->base + DMA_DCSR_S);
	}

	return 0;
}

static const struct dev_pm_ops sa11x0_dma_pm_ops = {
	SET_NOIRQ_SYSTEM_SLEEP_PM_OPS(sa11x0_dma_suspend, sa11x0_dma_resume)
};

static struct platform_driver sa11x0_dma_driver = {
	.driver = {
		.name	= "sa11x0-dma",
		.pm	= &sa11x0_dma_pm_ops,
	},
	.probe		= sa11x0_dma_probe,
	.remove_new	= sa11x0_dma_remove,
};

static int __init sa11x0_dma_init(void)
{
	return platform_driver_register(&sa11x0_dma_driver);
}
subsys_initcall(sa11x0_dma_init);

static void __exit sa11x0_dma_exit(void)
{
	platform_driver_unregister(&sa11x0_dma_driver);
}
module_exit(sa11x0_dma_exit);

MODULE_AUTHOR("Russell King");
MODULE_DESCRIPTION("SA-11x0 DMA driver");
MODULE_LICENSE("GPL v2");
MODULE_ALIAS("platform:sa11x0-dma");