summaryrefslogtreecommitdiffstats
path: root/drivers/dma/pxa_dma.c
blob: debca824bed676f0ecacc9e71476ecc33a477298 (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
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
/*
 * Copyright 2015 Robert Jarzmik <robert.jarzmik@free.fr>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 */

#include <linux/err.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/types.h>
#include <linux/interrupt.h>
#include <linux/dma-mapping.h>
#include <linux/slab.h>
#include <linux/dmaengine.h>
#include <linux/platform_device.h>
#include <linux/device.h>
#include <linux/platform_data/mmp_dma.h>
#include <linux/dmapool.h>
#include <linux/of_device.h>
#include <linux/of_dma.h>
#include <linux/of.h>
#include <linux/dma/pxa-dma.h>

#include "dmaengine.h"
#include "virt-dma.h"

#define DCSR(n)		(0x0000 + ((n) << 2))
#define DALGN(n)	0x00a0
#define DINT		0x00f0
#define DDADR(n)	(0x0200 + ((n) << 4))
#define DSADR(n)	(0x0204 + ((n) << 4))
#define DTADR(n)	(0x0208 + ((n) << 4))
#define DCMD(n)		(0x020c + ((n) << 4))

#define PXA_DCSR_RUN		BIT(31)	/* Run Bit (read / write) */
#define PXA_DCSR_NODESC		BIT(30)	/* No-Descriptor Fetch (read / write) */
#define PXA_DCSR_STOPIRQEN	BIT(29)	/* Stop Interrupt Enable (R/W) */
#define PXA_DCSR_REQPEND	BIT(8)	/* Request Pending (read-only) */
#define PXA_DCSR_STOPSTATE	BIT(3)	/* Stop State (read-only) */
#define PXA_DCSR_ENDINTR	BIT(2)	/* End Interrupt (read / write) */
#define PXA_DCSR_STARTINTR	BIT(1)	/* Start Interrupt (read / write) */
#define PXA_DCSR_BUSERR		BIT(0)	/* Bus Error Interrupt (read / write) */

#define PXA_DCSR_EORIRQEN	BIT(28)	/* End of Receive IRQ Enable (R/W) */
#define PXA_DCSR_EORJMPEN	BIT(27)	/* Jump to next descriptor on EOR */
#define PXA_DCSR_EORSTOPEN	BIT(26)	/* STOP on an EOR */
#define PXA_DCSR_SETCMPST	BIT(25)	/* Set Descriptor Compare Status */
#define PXA_DCSR_CLRCMPST	BIT(24)	/* Clear Descriptor Compare Status */
#define PXA_DCSR_CMPST		BIT(10)	/* The Descriptor Compare Status */
#define PXA_DCSR_EORINTR	BIT(9)	/* The end of Receive */

#define DRCMR_MAPVLD	BIT(7)	/* Map Valid (read / write) */
#define DRCMR_CHLNUM	0x1f	/* mask for Channel Number (read / write) */

#define DDADR_DESCADDR	0xfffffff0	/* Address of next descriptor (mask) */
#define DDADR_STOP	BIT(0)	/* Stop (read / write) */

#define PXA_DCMD_INCSRCADDR	BIT(31)	/* Source Address Increment Setting. */
#define PXA_DCMD_INCTRGADDR	BIT(30)	/* Target Address Increment Setting. */
#define PXA_DCMD_FLOWSRC	BIT(29)	/* Flow Control by the source. */
#define PXA_DCMD_FLOWTRG	BIT(28)	/* Flow Control by the target. */
#define PXA_DCMD_STARTIRQEN	BIT(22)	/* Start Interrupt Enable */
#define PXA_DCMD_ENDIRQEN	BIT(21)	/* End Interrupt Enable */
#define PXA_DCMD_ENDIAN		BIT(18)	/* Device Endian-ness. */
#define PXA_DCMD_BURST8		(1 << 16)	/* 8 byte burst */
#define PXA_DCMD_BURST16	(2 << 16)	/* 16 byte burst */
#define PXA_DCMD_BURST32	(3 << 16)	/* 32 byte burst */
#define PXA_DCMD_WIDTH1		(1 << 14)	/* 1 byte width */
#define PXA_DCMD_WIDTH2		(2 << 14)	/* 2 byte width (HalfWord) */
#define PXA_DCMD_WIDTH4		(3 << 14)	/* 4 byte width (Word) */
#define PXA_DCMD_LENGTH		0x01fff		/* length mask (max = 8K - 1) */

#define PDMA_ALIGNMENT		3
#define PDMA_MAX_DESC_BYTES	(PXA_DCMD_LENGTH & ~((1 << PDMA_ALIGNMENT) - 1))

struct pxad_desc_hw {
	u32 ddadr;	/* Points to the next descriptor + flags */
	u32 dsadr;	/* DSADR value for the current transfer */
	u32 dtadr;	/* DTADR value for the current transfer */
	u32 dcmd;	/* DCMD value for the current transfer */
} __aligned(16);

struct pxad_desc_sw {
	struct virt_dma_desc	vd;		/* Virtual descriptor */
	int			nb_desc;	/* Number of hw. descriptors */
	size_t			len;		/* Number of bytes xfered */
	dma_addr_t		first;		/* First descriptor's addr */

	/* At least one descriptor has an src/dst address not multiple of 8 */
	bool			misaligned;
	bool			cyclic;
	struct dma_pool		*desc_pool;	/* Channel's used allocator */

	struct pxad_desc_hw	*hw_desc[];	/* DMA coherent descriptors */
};

struct pxad_phy {
	int			idx;
	void __iomem		*base;
	struct pxad_chan	*vchan;
};

struct pxad_chan {
	struct virt_dma_chan	vc;		/* Virtual channel */
	u32			drcmr;		/* Requestor of the channel */
	enum pxad_chan_prio	prio;		/* Required priority of phy */
	/*
	 * At least one desc_sw in submitted or issued transfers on this channel
	 * has one address such as: addr % 8 != 0. This implies the DALGN
	 * setting on the phy.
	 */
	bool			misaligned;
	struct dma_slave_config	cfg;		/* Runtime config */

	/* protected by vc->lock */
	struct pxad_phy		*phy;
	struct dma_pool		*desc_pool;	/* Descriptors pool */
};

struct pxad_device {
	struct dma_device		slave;
	int				nr_chans;
	void __iomem			*base;
	struct pxad_phy			*phys;
	spinlock_t			phy_lock;	/* Phy association */
#ifdef CONFIG_DEBUG_FS
	struct dentry			*dbgfs_root;
	struct dentry			*dbgfs_state;
	struct dentry			**dbgfs_chan;
#endif
};

#define tx_to_pxad_desc(tx)					\
	container_of(tx, struct pxad_desc_sw, async_tx)
#define to_pxad_chan(dchan)					\
	container_of(dchan, struct pxad_chan, vc.chan)
#define to_pxad_dev(dmadev)					\
	container_of(dmadev, struct pxad_device, slave)
#define to_pxad_sw_desc(_vd)				\
	container_of((_vd), struct pxad_desc_sw, vd)

#define _phy_readl_relaxed(phy, _reg)					\
	readl_relaxed((phy)->base + _reg((phy)->idx))
#define phy_readl_relaxed(phy, _reg)					\
	({								\
		u32 _v;							\
		_v = readl_relaxed((phy)->base + _reg((phy)->idx));	\
		dev_vdbg(&phy->vchan->vc.chan.dev->device,		\
			 "%s(): readl(%s): 0x%08x\n", __func__, #_reg,	\
			  _v);						\
		_v;							\
	})
#define phy_writel(phy, val, _reg)					\
	do {								\
		writel((val), (phy)->base + _reg((phy)->idx));		\
		dev_vdbg(&phy->vchan->vc.chan.dev->device,		\
			 "%s(): writel(0x%08x, %s)\n",			\
			 __func__, (u32)(val), #_reg);			\
	} while (0)
#define phy_writel_relaxed(phy, val, _reg)				\
	do {								\
		writel_relaxed((val), (phy)->base + _reg((phy)->idx));	\
		dev_vdbg(&phy->vchan->vc.chan.dev->device,		\
			 "%s(): writel_relaxed(0x%08x, %s)\n",		\
			 __func__, (u32)(val), #_reg);			\
	} while (0)

static unsigned int pxad_drcmr(unsigned int line)
{
	if (line < 64)
		return 0x100 + line * 4;
	return 0x1000 + line * 4;
}

/*
 * Debug fs
 */
#ifdef CONFIG_DEBUG_FS
#include <linux/debugfs.h>
#include <linux/uaccess.h>
#include <linux/seq_file.h>

static int dbg_show_requester_chan(struct seq_file *s, void *p)
{
	struct pxad_phy *phy = s->private;
	int i;
	u32 drcmr;

	seq_printf(s, "DMA channel %d requester :\n", phy->idx);
	for (i = 0; i < 70; i++) {
		drcmr = readl_relaxed(phy->base + pxad_drcmr(i));
		if ((drcmr & DRCMR_CHLNUM) == phy->idx)
			seq_printf(s, "\tRequester %d (MAPVLD=%d)\n", i,
				   !!(drcmr & DRCMR_MAPVLD));
	}
	return 0;
}

static inline int dbg_burst_from_dcmd(u32 dcmd)
{
	int burst = (dcmd >> 16) & 0x3;

	return burst ? 4 << burst : 0;
}

static int is_phys_valid(unsigned long addr)
{
	return pfn_valid(__phys_to_pfn(addr));
}

#define PXA_DCSR_STR(flag) (dcsr & PXA_DCSR_##flag ? #flag" " : "")
#define PXA_DCMD_STR(flag) (dcmd & PXA_DCMD_##flag ? #flag" " : "")

static int dbg_show_descriptors(struct seq_file *s, void *p)
{
	struct pxad_phy *phy = s->private;
	int i, max_show = 20, burst, width;
	u32 dcmd;
	unsigned long phys_desc, ddadr;
	struct pxad_desc_hw *desc;

	phys_desc = ddadr = _phy_readl_relaxed(phy, DDADR);

	seq_printf(s, "DMA channel %d descriptors :\n", phy->idx);
	seq_printf(s, "[%03d] First descriptor unknown\n", 0);
	for (i = 1; i < max_show && is_phys_valid(phys_desc); i++) {
		desc = phys_to_virt(phys_desc);
		dcmd = desc->dcmd;
		burst = dbg_burst_from_dcmd(dcmd);
		width = (1 << ((dcmd >> 14) & 0x3)) >> 1;

		seq_printf(s, "[%03d] Desc at %08lx(virt %p)\n",
			   i, phys_desc, desc);
		seq_printf(s, "\tDDADR = %08x\n", desc->ddadr);
		seq_printf(s, "\tDSADR = %08x\n", desc->dsadr);
		seq_printf(s, "\tDTADR = %08x\n", desc->dtadr);
		seq_printf(s, "\tDCMD  = %08x (%s%s%s%s%s%s%sburst=%d width=%d len=%d)\n",
			   dcmd,
			   PXA_DCMD_STR(INCSRCADDR), PXA_DCMD_STR(INCTRGADDR),
			   PXA_DCMD_STR(FLOWSRC), PXA_DCMD_STR(FLOWTRG),
			   PXA_DCMD_STR(STARTIRQEN), PXA_DCMD_STR(ENDIRQEN),
			   PXA_DCMD_STR(ENDIAN), burst, width,
			   dcmd & PXA_DCMD_LENGTH);
		phys_desc = desc->ddadr;
	}
	if (i == max_show)
		seq_printf(s, "[%03d] Desc at %08lx ... max display reached\n",
			   i, phys_desc);
	else
		seq_printf(s, "[%03d] Desc at %08lx is %s\n",
			   i, phys_desc, phys_desc == DDADR_STOP ?
			   "DDADR_STOP" : "invalid");

	return 0;
}

static int dbg_show_chan_state(struct seq_file *s, void *p)
{
	struct pxad_phy *phy = s->private;
	u32 dcsr, dcmd;
	int burst, width;
	static const char * const str_prio[] = {
		"high", "normal", "low", "invalid"
	};

	dcsr = _phy_readl_relaxed(phy, DCSR);
	dcmd = _phy_readl_relaxed(phy, DCMD);
	burst = dbg_burst_from_dcmd(dcmd);
	width = (1 << ((dcmd >> 14) & 0x3)) >> 1;

	seq_printf(s, "DMA channel %d\n", phy->idx);
	seq_printf(s, "\tPriority : %s\n",
			  str_prio[(phy->idx & 0xf) / 4]);
	seq_printf(s, "\tUnaligned transfer bit: %s\n",
			  _phy_readl_relaxed(phy, DALGN) & BIT(phy->idx) ?
			  "yes" : "no");
	seq_printf(s, "\tDCSR  = %08x (%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s)\n",
		   dcsr, PXA_DCSR_STR(RUN), PXA_DCSR_STR(NODESC),
		   PXA_DCSR_STR(STOPIRQEN), PXA_DCSR_STR(EORIRQEN),
		   PXA_DCSR_STR(EORJMPEN), PXA_DCSR_STR(EORSTOPEN),
		   PXA_DCSR_STR(SETCMPST), PXA_DCSR_STR(CLRCMPST),
		   PXA_DCSR_STR(CMPST), PXA_DCSR_STR(EORINTR),
		   PXA_DCSR_STR(REQPEND), PXA_DCSR_STR(STOPSTATE),
		   PXA_DCSR_STR(ENDINTR), PXA_DCSR_STR(STARTINTR),
		   PXA_DCSR_STR(BUSERR));

	seq_printf(s, "\tDCMD  = %08x (%s%s%s%s%s%s%sburst=%d width=%d len=%d)\n",
		   dcmd,
		   PXA_DCMD_STR(INCSRCADDR), PXA_DCMD_STR(INCTRGADDR),
		   PXA_DCMD_STR(FLOWSRC), PXA_DCMD_STR(FLOWTRG),
		   PXA_DCMD_STR(STARTIRQEN), PXA_DCMD_STR(ENDIRQEN),
		   PXA_DCMD_STR(ENDIAN), burst, width, dcmd & PXA_DCMD_LENGTH);
	seq_printf(s, "\tDSADR = %08x\n", _phy_readl_relaxed(phy, DSADR));
	seq_printf(s, "\tDTADR = %08x\n", _phy_readl_relaxed(phy, DTADR));
	seq_printf(s, "\tDDADR = %08x\n", _phy_readl_relaxed(phy, DDADR));

	return 0;
}

static int dbg_show_state(struct seq_file *s, void *p)
{
	struct pxad_device *pdev = s->private;

	/* basic device status */
	seq_puts(s, "DMA engine status\n");
	seq_printf(s, "\tChannel number: %d\n", pdev->nr_chans);

	return 0;
}

#define DBGFS_FUNC_DECL(name) \
static int dbg_open_##name(struct inode *inode, struct file *file) \
{ \
	return single_open(file, dbg_show_##name, inode->i_private); \
} \
static const struct file_operations dbg_fops_##name = { \
	.owner		= THIS_MODULE, \
	.open		= dbg_open_##name, \
	.llseek		= seq_lseek, \
	.read		= seq_read, \
	.release	= single_release, \
}

DBGFS_FUNC_DECL(state);
DBGFS_FUNC_DECL(chan_state);
DBGFS_FUNC_DECL(descriptors);
DBGFS_FUNC_DECL(requester_chan);

static struct dentry *pxad_dbg_alloc_chan(struct pxad_device *pdev,
					     int ch, struct dentry *chandir)
{
	char chan_name[11];
	struct dentry *chan, *chan_state = NULL, *chan_descr = NULL;
	struct dentry *chan_reqs = NULL;
	void *dt;

	scnprintf(chan_name, sizeof(chan_name), "%d", ch);
	chan = debugfs_create_dir(chan_name, chandir);
	dt = (void *)&pdev->phys[ch];

	if (chan)
		chan_state = debugfs_create_file("state", 0400, chan, dt,
						 &dbg_fops_chan_state);
	if (chan_state)
		chan_descr = debugfs_create_file("descriptors", 0400, chan, dt,
						 &dbg_fops_descriptors);
	if (chan_descr)
		chan_reqs = debugfs_create_file("requesters", 0400, chan, dt,
						&dbg_fops_requester_chan);
	if (!chan_reqs)
		goto err_state;

	return chan;

err_state:
	debugfs_remove_recursive(chan);
	return NULL;
}

static void pxad_init_debugfs(struct pxad_device *pdev)
{
	int i;
	struct dentry *chandir;

	pdev->dbgfs_root = debugfs_create_dir(dev_name(pdev->slave.dev), NULL);
	if (IS_ERR(pdev->dbgfs_root) || !pdev->dbgfs_root)
		goto err_root;

	pdev->dbgfs_state = debugfs_create_file("state", 0400, pdev->dbgfs_root,
						pdev, &dbg_fops_state);
	if (!pdev->dbgfs_state)
		goto err_state;

	pdev->dbgfs_chan =
		kmalloc_array(pdev->nr_chans, sizeof(*pdev->dbgfs_state),
			      GFP_KERNEL);
	if (!pdev->dbgfs_chan)
		goto err_alloc;

	chandir = debugfs_create_dir("channels", pdev->dbgfs_root);
	if (!chandir)
		goto err_chandir;

	for (i = 0; i < pdev->nr_chans; i++) {
		pdev->dbgfs_chan[i] = pxad_dbg_alloc_chan(pdev, i, chandir);
		if (!pdev->dbgfs_chan[i])
			goto err_chans;
	}

	return;
err_chans:
err_chandir:
	kfree(pdev->dbgfs_chan);
err_alloc:
err_state:
	debugfs_remove_recursive(pdev->dbgfs_root);
err_root:
	pr_err("pxad: debugfs is not available\n");
}

static void pxad_cleanup_debugfs(struct pxad_device *pdev)
{
	debugfs_remove_recursive(pdev->dbgfs_root);
}
#else
static inline void pxad_init_debugfs(struct pxad_device *pdev) {}
static inline void pxad_cleanup_debugfs(struct pxad_device *pdev) {}
#endif

/*
 * In the transition phase where legacy pxa handling is done at the same time as
 * mmp_dma, the DMA physical channel split between the 2 DMA providers is done
 * through legacy_reserved. Legacy code reserves DMA channels by settings
 * corresponding bits in legacy_reserved.
 */
static u32 legacy_reserved;
static u32 legacy_unavailable;

static struct pxad_phy *lookup_phy(struct pxad_chan *pchan)
{
	int prio, i;
	struct pxad_device *pdev = to_pxad_dev(pchan->vc.chan.device);
	struct pxad_phy *phy, *found = NULL;
	unsigned long flags;

	/*
	 * dma channel priorities
	 * ch 0 - 3,  16 - 19  <--> (0)
	 * ch 4 - 7,  20 - 23  <--> (1)
	 * ch 8 - 11, 24 - 27  <--> (2)
	 * ch 12 - 15, 28 - 31  <--> (3)
	 */

	spin_lock_irqsave(&pdev->phy_lock, flags);
	for (prio = pchan->prio; prio >= PXAD_PRIO_HIGHEST; prio--) {
		for (i = 0; i < pdev->nr_chans; i++) {
			if (prio != (i & 0xf) >> 2)
				continue;
			if ((i < 32) && (legacy_reserved & BIT(i)))
				continue;
			phy = &pdev->phys[i];
			if (!phy->vchan) {
				phy->vchan = pchan;
				found = phy;
				if (i < 32)
					legacy_unavailable |= BIT(i);
				goto out_unlock;
			}
		}
	}

out_unlock:
	spin_unlock_irqrestore(&pdev->phy_lock, flags);
	dev_dbg(&pchan->vc.chan.dev->device,
		"%s(): phy=%p(%d)\n", __func__, found,
		found ? found->idx : -1);

	return found;
}

static void pxad_free_phy(struct pxad_chan *chan)
{
	struct pxad_device *pdev = to_pxad_dev(chan->vc.chan.device);
	unsigned long flags;
	u32 reg;
	int i;

	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): freeing\n", __func__);
	if (!chan->phy)
		return;

	/* clear the channel mapping in DRCMR */
	if (chan->drcmr <= DRCMR_CHLNUM) {
		reg = pxad_drcmr(chan->drcmr);
		writel_relaxed(0, chan->phy->base + reg);
	}

	spin_lock_irqsave(&pdev->phy_lock, flags);
	for (i = 0; i < 32; i++)
		if (chan->phy == &pdev->phys[i])
			legacy_unavailable &= ~BIT(i);
	chan->phy->vchan = NULL;
	chan->phy = NULL;
	spin_unlock_irqrestore(&pdev->phy_lock, flags);
}

static bool is_chan_running(struct pxad_chan *chan)
{
	u32 dcsr;
	struct pxad_phy *phy = chan->phy;

	if (!phy)
		return false;
	dcsr = phy_readl_relaxed(phy, DCSR);
	return dcsr & PXA_DCSR_RUN;
}

static bool is_running_chan_misaligned(struct pxad_chan *chan)
{
	u32 dalgn;

	BUG_ON(!chan->phy);
	dalgn = phy_readl_relaxed(chan->phy, DALGN);
	return dalgn & (BIT(chan->phy->idx));
}

static void phy_enable(struct pxad_phy *phy, bool misaligned)
{
	u32 reg, dalgn;

	if (!phy->vchan)
		return;

	dev_dbg(&phy->vchan->vc.chan.dev->device,
		"%s(); phy=%p(%d) misaligned=%d\n", __func__,
		phy, phy->idx, misaligned);

	if (phy->vchan->drcmr <= DRCMR_CHLNUM) {
		reg = pxad_drcmr(phy->vchan->drcmr);
		writel_relaxed(DRCMR_MAPVLD | phy->idx, phy->base + reg);
	}

	dalgn = phy_readl_relaxed(phy, DALGN);
	if (misaligned)
		dalgn |= BIT(phy->idx);
	else
		dalgn &= ~BIT(phy->idx);
	phy_writel_relaxed(phy, dalgn, DALGN);

	phy_writel(phy, PXA_DCSR_STOPIRQEN | PXA_DCSR_ENDINTR |
		   PXA_DCSR_BUSERR | PXA_DCSR_RUN, DCSR);
}

static void phy_disable(struct pxad_phy *phy)
{
	u32 dcsr;

	if (!phy)
		return;

	dcsr = phy_readl_relaxed(phy, DCSR);
	dev_dbg(&phy->vchan->vc.chan.dev->device,
		"%s(): phy=%p(%d)\n", __func__, phy, phy->idx);
	phy_writel(phy, dcsr & ~PXA_DCSR_RUN & ~PXA_DCSR_STOPIRQEN, DCSR);
}

static void pxad_launch_chan(struct pxad_chan *chan,
				 struct pxad_desc_sw *desc)
{
	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): desc=%p\n", __func__, desc);
	if (!chan->phy) {
		chan->phy = lookup_phy(chan);
		if (!chan->phy) {
			dev_dbg(&chan->vc.chan.dev->device,
				"%s(): no free dma channel\n", __func__);
			return;
		}
	}

	/*
	 * Program the descriptor's address into the DMA controller,
	 * then start the DMA transaction
	 */
	phy_writel(chan->phy, desc->first, DDADR);
	phy_enable(chan->phy, chan->misaligned);
}

static void set_updater_desc(struct pxad_desc_sw *sw_desc,
			     unsigned long flags)
{
	struct pxad_desc_hw *updater =
		sw_desc->hw_desc[sw_desc->nb_desc - 1];
	dma_addr_t dma = sw_desc->hw_desc[sw_desc->nb_desc - 2]->ddadr;

	updater->ddadr = DDADR_STOP;
	updater->dsadr = dma;
	updater->dtadr = dma + 8;
	updater->dcmd = PXA_DCMD_WIDTH4 | PXA_DCMD_BURST32 |
		(PXA_DCMD_LENGTH & sizeof(u32));
	if (flags & DMA_PREP_INTERRUPT)
		updater->dcmd |= PXA_DCMD_ENDIRQEN;
	if (sw_desc->cyclic)
		sw_desc->hw_desc[sw_desc->nb_desc - 2]->ddadr = sw_desc->first;
}

static bool is_desc_completed(struct virt_dma_desc *vd)
{
	struct pxad_desc_sw *sw_desc = to_pxad_sw_desc(vd);
	struct pxad_desc_hw *updater =
		sw_desc->hw_desc[sw_desc->nb_desc - 1];

	return updater->dtadr != (updater->dsadr + 8);
}

static void pxad_desc_chain(struct virt_dma_desc *vd1,
				struct virt_dma_desc *vd2)
{
	struct pxad_desc_sw *desc1 = to_pxad_sw_desc(vd1);
	struct pxad_desc_sw *desc2 = to_pxad_sw_desc(vd2);
	dma_addr_t dma_to_chain;

	dma_to_chain = desc2->first;
	desc1->hw_desc[desc1->nb_desc - 1]->ddadr = dma_to_chain;
}

static bool pxad_try_hotchain(struct virt_dma_chan *vc,
				  struct virt_dma_desc *vd)
{
	struct virt_dma_desc *vd_last_issued = NULL;
	struct pxad_chan *chan = to_pxad_chan(&vc->chan);

	/*
	 * Attempt to hot chain the tx if the phy is still running. This is
	 * considered successful only if either the channel is still running
	 * after the chaining, or if the chained transfer is completed after
	 * having been hot chained.
	 * A change of alignment is not allowed, and forbids hotchaining.
	 */
	if (is_chan_running(chan)) {
		BUG_ON(list_empty(&vc->desc_issued));

		if (!is_running_chan_misaligned(chan) &&
		    to_pxad_sw_desc(vd)->misaligned)
			return false;

		vd_last_issued = list_entry(vc->desc_issued.prev,
					    struct virt_dma_desc, node);
		pxad_desc_chain(vd_last_issued, vd);
		if (is_chan_running(chan) || is_desc_completed(vd_last_issued))
			return true;
	}

	return false;
}

static unsigned int clear_chan_irq(struct pxad_phy *phy)
{
	u32 dcsr;
	u32 dint = readl(phy->base + DINT);

	if (!(dint & BIT(phy->idx)))
		return PXA_DCSR_RUN;

	/* clear irq */
	dcsr = phy_readl_relaxed(phy, DCSR);
	phy_writel(phy, dcsr, DCSR);
	if ((dcsr & PXA_DCSR_BUSERR) && (phy->vchan))
		dev_warn(&phy->vchan->vc.chan.dev->device,
			 "%s(chan=%p): PXA_DCSR_BUSERR\n",
			 __func__, &phy->vchan);

	return dcsr & ~PXA_DCSR_RUN;
}

static irqreturn_t pxad_chan_handler(int irq, void *dev_id)
{
	struct pxad_phy *phy = dev_id;
	struct pxad_chan *chan = phy->vchan;
	struct virt_dma_desc *vd, *tmp;
	unsigned int dcsr;
	unsigned long flags;

	BUG_ON(!chan);

	dcsr = clear_chan_irq(phy);
	if (dcsr & PXA_DCSR_RUN)
		return IRQ_NONE;

	spin_lock_irqsave(&chan->vc.lock, flags);
	list_for_each_entry_safe(vd, tmp, &chan->vc.desc_issued, node) {
		dev_dbg(&chan->vc.chan.dev->device,
			"%s(): checking txd %p[%x]: completed=%d\n",
			__func__, vd, vd->tx.cookie, is_desc_completed(vd));
		if (to_pxad_sw_desc(vd)->cyclic) {
			vchan_cyclic_callback(vd);
			break;
		}
		if (is_desc_completed(vd)) {
			list_del(&vd->node);
			vchan_cookie_complete(vd);
		} else {
			break;
		}
	}

	if (dcsr & PXA_DCSR_STOPSTATE) {
		dev_dbg(&chan->vc.chan.dev->device,
		"%s(): channel stopped, submitted_empty=%d issued_empty=%d",
			__func__,
			list_empty(&chan->vc.desc_submitted),
			list_empty(&chan->vc.desc_issued));
		phy_writel_relaxed(phy, dcsr & ~PXA_DCSR_STOPIRQEN, DCSR);

		if (list_empty(&chan->vc.desc_issued)) {
			chan->misaligned =
				!list_empty(&chan->vc.desc_submitted);
		} else {
			vd = list_first_entry(&chan->vc.desc_issued,
					      struct virt_dma_desc, node);
			pxad_launch_chan(chan, to_pxad_sw_desc(vd));
		}
	}
	spin_unlock_irqrestore(&chan->vc.lock, flags);

	return IRQ_HANDLED;
}

static irqreturn_t pxad_int_handler(int irq, void *dev_id)
{
	struct pxad_device *pdev = dev_id;
	struct pxad_phy *phy;
	u32 dint = readl(pdev->base + DINT);
	int i, ret = IRQ_NONE;

	while (dint) {
		i = __ffs(dint);
		dint &= (dint - 1);
		phy = &pdev->phys[i];
		if ((i < 32) && (legacy_reserved & BIT(i)))
			continue;
		if (pxad_chan_handler(irq, phy) == IRQ_HANDLED)
			ret = IRQ_HANDLED;
	}

	return ret;
}

static int pxad_alloc_chan_resources(struct dma_chan *dchan)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);
	struct pxad_device *pdev = to_pxad_dev(chan->vc.chan.device);

	if (chan->desc_pool)
		return 1;

	chan->desc_pool = dma_pool_create(dma_chan_name(dchan),
					  pdev->slave.dev,
					  sizeof(struct pxad_desc_hw),
					  __alignof__(struct pxad_desc_hw),
					  0);
	if (!chan->desc_pool) {
		dev_err(&chan->vc.chan.dev->device,
			"%s(): unable to allocate descriptor pool\n",
			__func__);
		return -ENOMEM;
	}

	return 1;
}

static void pxad_free_chan_resources(struct dma_chan *dchan)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);

	vchan_free_chan_resources(&chan->vc);
	dma_pool_destroy(chan->desc_pool);
	chan->desc_pool = NULL;

}

static void pxad_free_desc(struct virt_dma_desc *vd)
{
	int i;
	dma_addr_t dma;
	struct pxad_desc_sw *sw_desc = to_pxad_sw_desc(vd);

	BUG_ON(sw_desc->nb_desc == 0);
	for (i = sw_desc->nb_desc - 1; i >= 0; i--) {
		if (i > 0)
			dma = sw_desc->hw_desc[i - 1]->ddadr;
		else
			dma = sw_desc->first;
		dma_pool_free(sw_desc->desc_pool,
			      sw_desc->hw_desc[i], dma);
	}
	sw_desc->nb_desc = 0;
	kfree(sw_desc);
}

static struct pxad_desc_sw *
pxad_alloc_desc(struct pxad_chan *chan, unsigned int nb_hw_desc)
{
	struct pxad_desc_sw *sw_desc;
	dma_addr_t dma;
	int i;

	sw_desc = kzalloc(sizeof(*sw_desc) +
			  nb_hw_desc * sizeof(struct pxad_desc_hw *),
			  GFP_NOWAIT);
	if (!sw_desc)
		return NULL;
	sw_desc->desc_pool = chan->desc_pool;

	for (i = 0; i < nb_hw_desc; i++) {
		sw_desc->hw_desc[i] = dma_pool_alloc(sw_desc->desc_pool,
						     GFP_NOWAIT, &dma);
		if (!sw_desc->hw_desc[i]) {
			dev_err(&chan->vc.chan.dev->device,
				"%s(): Couldn't allocate the %dth hw_desc from dma_pool %p\n",
				__func__, i, sw_desc->desc_pool);
			goto err;
		}

		if (i == 0)
			sw_desc->first = dma;
		else
			sw_desc->hw_desc[i - 1]->ddadr = dma;
		sw_desc->nb_desc++;
	}

	return sw_desc;
err:
	pxad_free_desc(&sw_desc->vd);
	return NULL;
}

static dma_cookie_t pxad_tx_submit(struct dma_async_tx_descriptor *tx)
{
	struct virt_dma_chan *vc = to_virt_chan(tx->chan);
	struct pxad_chan *chan = to_pxad_chan(&vc->chan);
	struct virt_dma_desc *vd_chained = NULL,
		*vd = container_of(tx, struct virt_dma_desc, tx);
	dma_cookie_t cookie;
	unsigned long flags;

	set_updater_desc(to_pxad_sw_desc(vd), tx->flags);

	spin_lock_irqsave(&vc->lock, flags);
	cookie = dma_cookie_assign(tx);

	if (list_empty(&vc->desc_submitted) && pxad_try_hotchain(vc, vd)) {
		list_move_tail(&vd->node, &vc->desc_issued);
		dev_dbg(&chan->vc.chan.dev->device,
			"%s(): txd %p[%x]: submitted (hot linked)\n",
			__func__, vd, cookie);
		goto out;
	}

	/*
	 * Fallback to placing the tx in the submitted queue
	 */
	if (!list_empty(&vc->desc_submitted)) {
		vd_chained = list_entry(vc->desc_submitted.prev,
					struct virt_dma_desc, node);
		/*
		 * Only chain the descriptors if no new misalignment is
		 * introduced. If a new misalignment is chained, let the channel
		 * stop, and be relaunched in misalign mode from the irq
		 * handler.
		 */
		if (chan->misaligned || !to_pxad_sw_desc(vd)->misaligned)
			pxad_desc_chain(vd_chained, vd);
		else
			vd_chained = NULL;
	}
	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): txd %p[%x]: submitted (%s linked)\n",
		__func__, vd, cookie, vd_chained ? "cold" : "not");
	list_move_tail(&vd->node, &vc->desc_submitted);
	chan->misaligned |= to_pxad_sw_desc(vd)->misaligned;

out:
	spin_unlock_irqrestore(&vc->lock, flags);
	return cookie;
}

static void pxad_issue_pending(struct dma_chan *dchan)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);
	struct virt_dma_desc *vd_first;
	unsigned long flags;

	spin_lock_irqsave(&chan->vc.lock, flags);
	if (list_empty(&chan->vc.desc_submitted))
		goto out;

	vd_first = list_first_entry(&chan->vc.desc_submitted,
				    struct virt_dma_desc, node);
	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): txd %p[%x]", __func__, vd_first, vd_first->tx.cookie);

	vchan_issue_pending(&chan->vc);
	if (!pxad_try_hotchain(&chan->vc, vd_first))
		pxad_launch_chan(chan, to_pxad_sw_desc(vd_first));
out:
	spin_unlock_irqrestore(&chan->vc.lock, flags);
}

static inline struct dma_async_tx_descriptor *
pxad_tx_prep(struct virt_dma_chan *vc, struct virt_dma_desc *vd,
		 unsigned long tx_flags)
{
	struct dma_async_tx_descriptor *tx;
	struct pxad_chan *chan = container_of(vc, struct pxad_chan, vc);

	INIT_LIST_HEAD(&vd->node);
	tx = vchan_tx_prep(vc, vd, tx_flags);
	tx->tx_submit = pxad_tx_submit;
	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): vc=%p txd=%p[%x] flags=0x%lx\n", __func__,
		vc, vd, vd->tx.cookie,
		tx_flags);

	return tx;
}

static void pxad_get_config(struct pxad_chan *chan,
			    enum dma_transfer_direction dir,
			    u32 *dcmd, u32 *dev_src, u32 *dev_dst)
{
	u32 maxburst = 0, dev_addr = 0;
	enum dma_slave_buswidth width = DMA_SLAVE_BUSWIDTH_UNDEFINED;

	*dcmd = 0;
	if (dir == DMA_DEV_TO_MEM) {
		maxburst = chan->cfg.src_maxburst;
		width = chan->cfg.src_addr_width;
		dev_addr = chan->cfg.src_addr;
		*dev_src = dev_addr;
		*dcmd |= PXA_DCMD_INCTRGADDR;
		if (chan->drcmr <= DRCMR_CHLNUM)
			*dcmd |= PXA_DCMD_FLOWSRC;
	}
	if (dir == DMA_MEM_TO_DEV) {
		maxburst = chan->cfg.dst_maxburst;
		width = chan->cfg.dst_addr_width;
		dev_addr = chan->cfg.dst_addr;
		*dev_dst = dev_addr;
		*dcmd |= PXA_DCMD_INCSRCADDR;
		if (chan->drcmr <= DRCMR_CHLNUM)
			*dcmd |= PXA_DCMD_FLOWTRG;
	}
	if (dir == DMA_MEM_TO_MEM)
		*dcmd |= PXA_DCMD_BURST32 | PXA_DCMD_INCTRGADDR |
			PXA_DCMD_INCSRCADDR;

	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): dev_addr=0x%x maxburst=%d width=%d  dir=%d\n",
		__func__, dev_addr, maxburst, width, dir);

	if (width == DMA_SLAVE_BUSWIDTH_1_BYTE)
		*dcmd |= PXA_DCMD_WIDTH1;
	else if (width == DMA_SLAVE_BUSWIDTH_2_BYTES)
		*dcmd |= PXA_DCMD_WIDTH2;
	else if (width == DMA_SLAVE_BUSWIDTH_4_BYTES)
		*dcmd |= PXA_DCMD_WIDTH4;

	if (maxburst == 8)
		*dcmd |= PXA_DCMD_BURST8;
	else if (maxburst == 16)
		*dcmd |= PXA_DCMD_BURST16;
	else if (maxburst == 32)
		*dcmd |= PXA_DCMD_BURST32;

	/* FIXME: drivers should be ported over to use the filter
	 * function. Once that's done, the following two lines can
	 * be removed.
	 */
	if (chan->cfg.slave_id)
		chan->drcmr = chan->cfg.slave_id;
}

static struct dma_async_tx_descriptor *
pxad_prep_memcpy(struct dma_chan *dchan,
		 dma_addr_t dma_dst, dma_addr_t dma_src,
		 size_t len, unsigned long flags)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);
	struct pxad_desc_sw *sw_desc;
	struct pxad_desc_hw *hw_desc;
	u32 dcmd;
	unsigned int i, nb_desc = 0;
	size_t copy;

	if (!dchan || !len)
		return NULL;

	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): dma_dst=0x%lx dma_src=0x%lx len=%zu flags=%lx\n",
		__func__, (unsigned long)dma_dst, (unsigned long)dma_src,
		len, flags);
	pxad_get_config(chan, DMA_MEM_TO_MEM, &dcmd, NULL, NULL);

	nb_desc = DIV_ROUND_UP(len, PDMA_MAX_DESC_BYTES);
	sw_desc = pxad_alloc_desc(chan, nb_desc + 1);
	if (!sw_desc)
		return NULL;
	sw_desc->len = len;

	if (!IS_ALIGNED(dma_src, 1 << PDMA_ALIGNMENT) ||
	    !IS_ALIGNED(dma_dst, 1 << PDMA_ALIGNMENT))
		sw_desc->misaligned = true;

	i = 0;
	do {
		hw_desc = sw_desc->hw_desc[i++];
		copy = min_t(size_t, len, PDMA_MAX_DESC_BYTES);
		hw_desc->dcmd = dcmd | (PXA_DCMD_LENGTH & copy);
		hw_desc->dsadr = dma_src;
		hw_desc->dtadr = dma_dst;
		len -= copy;
		dma_src += copy;
		dma_dst += copy;
	} while (len);
	set_updater_desc(sw_desc, flags);

	return pxad_tx_prep(&chan->vc, &sw_desc->vd, flags);
}

static struct dma_async_tx_descriptor *
pxad_prep_slave_sg(struct dma_chan *dchan, struct scatterlist *sgl,
		   unsigned int sg_len, enum dma_transfer_direction dir,
		   unsigned long flags, void *context)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);
	struct pxad_desc_sw *sw_desc;
	size_t len, avail;
	struct scatterlist *sg;
	dma_addr_t dma;
	u32 dcmd, dsadr = 0, dtadr = 0;
	unsigned int nb_desc = 0, i, j = 0;

	if ((sgl == NULL) || (sg_len == 0))
		return NULL;

	pxad_get_config(chan, dir, &dcmd, &dsadr, &dtadr);
	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): dir=%d flags=%lx\n", __func__, dir, flags);

	for_each_sg(sgl, sg, sg_len, i)
		nb_desc += DIV_ROUND_UP(sg_dma_len(sg), PDMA_MAX_DESC_BYTES);
	sw_desc = pxad_alloc_desc(chan, nb_desc + 1);
	if (!sw_desc)
		return NULL;

	for_each_sg(sgl, sg, sg_len, i) {
		dma = sg_dma_address(sg);
		avail = sg_dma_len(sg);
		sw_desc->len += avail;

		do {
			len = min_t(size_t, avail, PDMA_MAX_DESC_BYTES);
			if (dma & 0x7)
				sw_desc->misaligned = true;

			sw_desc->hw_desc[j]->dcmd =
				dcmd | (PXA_DCMD_LENGTH & len);
			sw_desc->hw_desc[j]->dsadr = dsadr ? dsadr : dma;
			sw_desc->hw_desc[j++]->dtadr = dtadr ? dtadr : dma;

			dma += len;
			avail -= len;
		} while (avail);
	}
	set_updater_desc(sw_desc, flags);

	return pxad_tx_prep(&chan->vc, &sw_desc->vd, flags);
}

static struct dma_async_tx_descriptor *
pxad_prep_dma_cyclic(struct dma_chan *dchan,
		     dma_addr_t buf_addr, size_t len, size_t period_len,
		     enum dma_transfer_direction dir, unsigned long flags)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);
	struct pxad_desc_sw *sw_desc;
	struct pxad_desc_hw **phw_desc;
	dma_addr_t dma;
	u32 dcmd, dsadr = 0, dtadr = 0;
	unsigned int nb_desc = 0;

	if (!dchan || !len || !period_len)
		return NULL;
	if ((dir != DMA_DEV_TO_MEM) && (dir != DMA_MEM_TO_DEV)) {
		dev_err(&chan->vc.chan.dev->device,
			"Unsupported direction for cyclic DMA\n");
		return NULL;
	}
	/* the buffer length must be a multiple of period_len */
	if (len % period_len != 0 || period_len > PDMA_MAX_DESC_BYTES ||
	    !IS_ALIGNED(period_len, 1 << PDMA_ALIGNMENT))
		return NULL;

	pxad_get_config(chan, dir, &dcmd, &dsadr, &dtadr);
	dcmd |= PXA_DCMD_ENDIRQEN | (PXA_DCMD_LENGTH & period_len);
	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): buf_addr=0x%lx len=%zu period=%zu dir=%d flags=%lx\n",
		__func__, (unsigned long)buf_addr, len, period_len, dir, flags);

	nb_desc = DIV_ROUND_UP(period_len, PDMA_MAX_DESC_BYTES);
	nb_desc *= DIV_ROUND_UP(len, period_len);
	sw_desc = pxad_alloc_desc(chan, nb_desc + 1);
	if (!sw_desc)
		return NULL;
	sw_desc->cyclic = true;
	sw_desc->len = len;

	phw_desc = sw_desc->hw_desc;
	dma = buf_addr;
	do {
		phw_desc[0]->dsadr = dsadr ? dsadr : dma;
		phw_desc[0]->dtadr = dtadr ? dtadr : dma;
		phw_desc[0]->dcmd = dcmd;
		phw_desc++;
		dma += period_len;
		len -= period_len;
	} while (len);
	set_updater_desc(sw_desc, flags);

	return pxad_tx_prep(&chan->vc, &sw_desc->vd, flags);
}

static int pxad_config(struct dma_chan *dchan,
		       struct dma_slave_config *cfg)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);

	if (!dchan)
		return -EINVAL;

	chan->cfg = *cfg;
	return 0;
}

static int pxad_terminate_all(struct dma_chan *dchan)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);
	struct pxad_device *pdev = to_pxad_dev(chan->vc.chan.device);
	struct virt_dma_desc *vd = NULL;
	unsigned long flags;
	struct pxad_phy *phy;
	LIST_HEAD(head);

	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): vchan %p: terminate all\n", __func__, &chan->vc);

	spin_lock_irqsave(&chan->vc.lock, flags);
	vchan_get_all_descriptors(&chan->vc, &head);

	list_for_each_entry(vd, &head, node) {
		dev_dbg(&chan->vc.chan.dev->device,
			"%s(): cancelling txd %p[%x] (completed=%d)", __func__,
			vd, vd->tx.cookie, is_desc_completed(vd));
	}

	phy = chan->phy;
	if (phy) {
		phy_disable(chan->phy);
		pxad_free_phy(chan);
		chan->phy = NULL;
		spin_lock(&pdev->phy_lock);
		phy->vchan = NULL;
		spin_unlock(&pdev->phy_lock);
	}
	spin_unlock_irqrestore(&chan->vc.lock, flags);
	vchan_dma_desc_free_list(&chan->vc, &head);

	return 0;
}

static unsigned int pxad_residue(struct pxad_chan *chan,
				 dma_cookie_t cookie)
{
	struct virt_dma_desc *vd = NULL;
	struct pxad_desc_sw *sw_desc = NULL;
	struct pxad_desc_hw *hw_desc = NULL;
	u32 curr, start, len, end, residue = 0;
	unsigned long flags;
	bool passed = false;
	int i;

	/*
	 * If the channel does not have a phy pointer anymore, it has already
	 * been completed. Therefore, its residue is 0.
	 */
	if (!chan->phy)
		return 0;

	spin_lock_irqsave(&chan->vc.lock, flags);

	vd = vchan_find_desc(&chan->vc, cookie);
	if (!vd)
		goto out;

	sw_desc = to_pxad_sw_desc(vd);
	if (sw_desc->hw_desc[0]->dcmd & PXA_DCMD_INCSRCADDR)
		curr = phy_readl_relaxed(chan->phy, DSADR);
	else
		curr = phy_readl_relaxed(chan->phy, DTADR);

	/*
	 * curr has to be actually read before checking descriptor
	 * completion, so that a curr inside a status updater
	 * descriptor implies the following test returns true, and
	 * preventing reordering of curr load and the test.
	 */
	rmb();
	if (is_desc_completed(vd))
		goto out;

	for (i = 0; i < sw_desc->nb_desc - 1; i++) {
		hw_desc = sw_desc->hw_desc[i];
		if (sw_desc->hw_desc[0]->dcmd & PXA_DCMD_INCSRCADDR)
			start = hw_desc->dsadr;
		else
			start = hw_desc->dtadr;
		len = hw_desc->dcmd & PXA_DCMD_LENGTH;
		end = start + len;

		/*
		 * 'passed' will be latched once we found the descriptor
		 * which lies inside the boundaries of the curr
		 * pointer. All descriptors that occur in the list
		 * _after_ we found that partially handled descriptor
		 * are still to be processed and are hence added to the
		 * residual bytes counter.
		 */

		if (passed) {
			residue += len;
		} else if (curr >= start && curr <= end) {
			residue += end - curr;
			passed = true;
		}
	}
	if (!passed)
		residue = sw_desc->len;

out:
	spin_unlock_irqrestore(&chan->vc.lock, flags);
	dev_dbg(&chan->vc.chan.dev->device,
		"%s(): txd %p[%x] sw_desc=%p: %d\n",
		__func__, vd, cookie, sw_desc, residue);
	return residue;
}

static enum dma_status pxad_tx_status(struct dma_chan *dchan,
				      dma_cookie_t cookie,
				      struct dma_tx_state *txstate)
{
	struct pxad_chan *chan = to_pxad_chan(dchan);
	enum dma_status ret;

	ret = dma_cookie_status(dchan, cookie, txstate);
	if (likely(txstate && (ret != DMA_ERROR)))
		dma_set_residue(txstate, pxad_residue(chan, cookie));

	return ret;
}

static void pxad_free_channels(struct dma_device *dmadev)
{
	struct pxad_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);
	}
}

static int pxad_remove(struct platform_device *op)
{
	struct pxad_device *pdev = platform_get_drvdata(op);

	pxad_cleanup_debugfs(pdev);
	pxad_free_channels(&pdev->slave);
	dma_async_device_unregister(&pdev->slave);
	return 0;
}

static int pxad_init_phys(struct platform_device *op,
			  struct pxad_device *pdev,
			  unsigned int nb_phy_chans)
{
	int irq0, irq, nr_irq = 0, i, ret;
	struct pxad_phy *phy;

	irq0 = platform_get_irq(op, 0);
	if (irq0 < 0)
		return irq0;

	pdev->phys = devm_kcalloc(&op->dev, nb_phy_chans,
				  sizeof(pdev->phys[0]), GFP_KERNEL);
	if (!pdev->phys)
		return -ENOMEM;

	for (i = 0; i < nb_phy_chans; i++)
		if (platform_get_irq(op, i) > 0)
			nr_irq++;

	for (i = 0; i < nb_phy_chans; i++) {
		phy = &pdev->phys[i];
		phy->base = pdev->base;
		phy->idx = i;
		irq = platform_get_irq(op, i);
		if ((nr_irq > 1) && (irq > 0))
			ret = devm_request_irq(&op->dev, irq,
					       pxad_chan_handler,
					       IRQF_SHARED, "pxa-dma", phy);
		if ((nr_irq == 1) && (i == 0))
			ret = devm_request_irq(&op->dev, irq0,
					       pxad_int_handler,
					       IRQF_SHARED, "pxa-dma", pdev);
		if (ret) {
			dev_err(pdev->slave.dev,
				"%s(): can't request irq %d:%d\n", __func__,
				irq, ret);
			return ret;
		}
	}

	return 0;
}

static const struct of_device_id const pxad_dt_ids[] = {
	{ .compatible = "marvell,pdma-1.0", },
	{}
};
MODULE_DEVICE_TABLE(of, pxad_dt_ids);

static struct dma_chan *pxad_dma_xlate(struct of_phandle_args *dma_spec,
					   struct of_dma *ofdma)
{
	struct pxad_device *d = ofdma->of_dma_data;
	struct dma_chan *chan;

	chan = dma_get_any_slave_channel(&d->slave);
	if (!chan)
		return NULL;

	to_pxad_chan(chan)->drcmr = dma_spec->args[0];
	to_pxad_chan(chan)->prio = dma_spec->args[1];

	return chan;
}

static int pxad_init_dmadev(struct platform_device *op,
			    struct pxad_device *pdev,
			    unsigned int nr_phy_chans)
{
	int ret;
	unsigned int i;
	struct pxad_chan *c;

	pdev->nr_chans = nr_phy_chans;
	INIT_LIST_HEAD(&pdev->slave.channels);
	pdev->slave.device_alloc_chan_resources = pxad_alloc_chan_resources;
	pdev->slave.device_free_chan_resources = pxad_free_chan_resources;
	pdev->slave.device_tx_status = pxad_tx_status;
	pdev->slave.device_issue_pending = pxad_issue_pending;
	pdev->slave.device_config = pxad_config;
	pdev->slave.device_terminate_all = pxad_terminate_all;

	if (op->dev.coherent_dma_mask)
		dma_set_mask(&op->dev, op->dev.coherent_dma_mask);
	else
		dma_set_mask(&op->dev, DMA_BIT_MASK(32));

	ret = pxad_init_phys(op, pdev, nr_phy_chans);
	if (ret)
		return ret;

	for (i = 0; i < nr_phy_chans; i++) {
		c = devm_kzalloc(&op->dev, sizeof(*c), GFP_KERNEL);
		if (!c)
			return -ENOMEM;
		c->vc.desc_free = pxad_free_desc;
		vchan_init(&c->vc, &pdev->slave);
	}

	return dma_async_device_register(&pdev->slave);
}

static int pxad_probe(struct platform_device *op)
{
	struct pxad_device *pdev;
	const struct of_device_id *of_id;
	struct mmp_dma_platdata *pdata = dev_get_platdata(&op->dev);
	struct resource *iores;
	int ret, dma_channels = 0;
	const enum dma_slave_buswidth widths =
		DMA_SLAVE_BUSWIDTH_1_BYTE   | DMA_SLAVE_BUSWIDTH_2_BYTES |
		DMA_SLAVE_BUSWIDTH_4_BYTES;

	pdev = devm_kzalloc(&op->dev, sizeof(*pdev), GFP_KERNEL);
	if (!pdev)
		return -ENOMEM;

	spin_lock_init(&pdev->phy_lock);

	iores = platform_get_resource(op, IORESOURCE_MEM, 0);
	pdev->base = devm_ioremap_resource(&op->dev, iores);
	if (IS_ERR(pdev->base))
		return PTR_ERR(pdev->base);

	of_id = of_match_device(pxad_dt_ids, &op->dev);
	if (of_id)
		of_property_read_u32(op->dev.of_node, "#dma-channels",
				     &dma_channels);
	else if (pdata && pdata->dma_channels)
		dma_channels = pdata->dma_channels;
	else
		dma_channels = 32;	/* default 32 channel */

	dma_cap_set(DMA_SLAVE, pdev->slave.cap_mask);
	dma_cap_set(DMA_MEMCPY, pdev->slave.cap_mask);
	dma_cap_set(DMA_CYCLIC, pdev->slave.cap_mask);
	dma_cap_set(DMA_PRIVATE, pdev->slave.cap_mask);
	pdev->slave.device_prep_dma_memcpy = pxad_prep_memcpy;
	pdev->slave.device_prep_slave_sg = pxad_prep_slave_sg;
	pdev->slave.device_prep_dma_cyclic = pxad_prep_dma_cyclic;

	pdev->slave.copy_align = PDMA_ALIGNMENT;
	pdev->slave.src_addr_widths = widths;
	pdev->slave.dst_addr_widths = widths;
	pdev->slave.directions = BIT(DMA_MEM_TO_DEV) | BIT(DMA_DEV_TO_MEM);
	pdev->slave.residue_granularity = DMA_RESIDUE_GRANULARITY_DESCRIPTOR;
	pdev->slave.descriptor_reuse = true;

	pdev->slave.dev = &op->dev;
	ret = pxad_init_dmadev(op, pdev, dma_channels);
	if (ret) {
		dev_err(pdev->slave.dev, "unable to register\n");
		return ret;
	}

	if (op->dev.of_node) {
		/* Device-tree DMA controller registration */
		ret = of_dma_controller_register(op->dev.of_node,
						 pxad_dma_xlate, pdev);
		if (ret < 0) {
			dev_err(pdev->slave.dev,
				"of_dma_controller_register failed\n");
			return ret;
		}
	}

	platform_set_drvdata(op, pdev);
	pxad_init_debugfs(pdev);
	dev_info(pdev->slave.dev, "initialized %d channels\n", dma_channels);
	return 0;
}

static const struct platform_device_id pxad_id_table[] = {
	{ "pxa-dma", },
	{ },
};

static struct platform_driver pxad_driver = {
	.driver		= {
		.name	= "pxa-dma",
		.of_match_table = pxad_dt_ids,
	},
	.id_table	= pxad_id_table,
	.probe		= pxad_probe,
	.remove		= pxad_remove,
};

bool pxad_filter_fn(struct dma_chan *chan, void *param)
{
	struct pxad_chan *c = to_pxad_chan(chan);
	struct pxad_param *p = param;

	if (chan->device->dev->driver != &pxad_driver.driver)
		return false;

	c->drcmr = p->drcmr;
	c->prio = p->prio;

	return true;
}
EXPORT_SYMBOL_GPL(pxad_filter_fn);

int pxad_toggle_reserved_channel(int legacy_channel)
{
	if (legacy_unavailable & (BIT(legacy_channel)))
		return -EBUSY;
	legacy_reserved ^= BIT(legacy_channel);
	return 0;
}
EXPORT_SYMBOL_GPL(pxad_toggle_reserved_channel);

module_platform_driver(pxad_driver);

MODULE_DESCRIPTION("Marvell PXA Peripheral DMA Driver");
MODULE_AUTHOR("Robert Jarzmik <robert.jarzmik@free.fr>");
MODULE_LICENSE("GPL v2");