aboutsummaryrefslogtreecommitdiff
path: root/drivers/ide/ide-io.c
blob: f01b103e55f0d300784103f2c14cc1898d8a891f (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
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
/*
 *	IDE I/O functions
 *
 *	Basic PIO and command management functionality.
 *
 * This code was split off from ide.c. See ide.c for history and original
 * copyrights.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License as published by the
 * Free Software Foundation; either version 2, or (at your option) any
 * later version.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * For the avoidance of doubt the "preferred form" of this code is one which
 * is in an open non patent encumbered format. Where cryptographic key signing
 * forms part of the process of creating an executable the information
 * including keys needed to generate an equivalently functional executable
 * are deemed to be part of the source code.
 */
 
 
#include <linux/module.h>
#include <linux/types.h>
#include <linux/string.h>
#include <linux/kernel.h>
#include <linux/timer.h>
#include <linux/mm.h>
#include <linux/interrupt.h>
#include <linux/major.h>
#include <linux/errno.h>
#include <linux/genhd.h>
#include <linux/blkpg.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/pci.h>
#include <linux/delay.h>
#include <linux/ide.h>
#include <linux/completion.h>
#include <linux/reboot.h>
#include <linux/cdrom.h>
#include <linux/seq_file.h>
#include <linux/device.h>
#include <linux/kmod.h>
#include <linux/scatterlist.h>
#include <linux/bitops.h>

#include <asm/byteorder.h>
#include <asm/irq.h>
#include <asm/uaccess.h>
#include <asm/io.h>

static int __ide_end_request(ide_drive_t *drive, struct request *rq,
			     int uptodate, unsigned int nr_bytes, int dequeue)
{
	int ret = 1;

	/*
	 * if failfast is set on a request, override number of sectors and
	 * complete the whole request right now
	 */
	if (blk_noretry_request(rq) && end_io_error(uptodate))
		nr_bytes = rq->hard_nr_sectors << 9;

	if (!blk_fs_request(rq) && end_io_error(uptodate) && !rq->errors)
		rq->errors = -EIO;

	/*
	 * decide whether to reenable DMA -- 3 is a random magic for now,
	 * if we DMA timeout more than 3 times, just stay in PIO
	 */
	if (drive->state == DMA_PIO_RETRY && drive->retry_pio <= 3) {
		drive->state = 0;
		ide_dma_on(drive);
	}

	if (!end_that_request_chunk(rq, uptodate, nr_bytes)) {
		add_disk_randomness(rq->rq_disk);
		if (dequeue) {
			if (!list_empty(&rq->queuelist))
				blkdev_dequeue_request(rq);
			HWGROUP(drive)->rq = NULL;
		}
		end_that_request_last(rq, uptodate);
		ret = 0;
	}

	return ret;
}

/**
 *	ide_end_request		-	complete an IDE I/O
 *	@drive: IDE device for the I/O
 *	@uptodate:
 *	@nr_sectors: number of sectors completed
 *
 *	This is our end_request wrapper function. We complete the I/O
 *	update random number input and dequeue the request, which if
 *	it was tagged may be out of order.
 */

int ide_end_request (ide_drive_t *drive, int uptodate, int nr_sectors)
{
	unsigned int nr_bytes = nr_sectors << 9;
	struct request *rq;
	unsigned long flags;
	int ret = 1;

	/*
	 * room for locking improvements here, the calls below don't
	 * need the queue lock held at all
	 */
	spin_lock_irqsave(&ide_lock, flags);
	rq = HWGROUP(drive)->rq;

	if (!nr_bytes) {
		if (blk_pc_request(rq))
			nr_bytes = rq->data_len;
		else
			nr_bytes = rq->hard_cur_sectors << 9;
	}

	ret = __ide_end_request(drive, rq, uptodate, nr_bytes, 1);

	spin_unlock_irqrestore(&ide_lock, flags);
	return ret;
}
EXPORT_SYMBOL(ide_end_request);

/*
 * Power Management state machine. This one is rather trivial for now,
 * we should probably add more, like switching back to PIO on suspend
 * to help some BIOSes, re-do the door locking on resume, etc...
 */

enum {
	ide_pm_flush_cache	= ide_pm_state_start_suspend,
	idedisk_pm_standby,

	idedisk_pm_restore_pio	= ide_pm_state_start_resume,
	idedisk_pm_idle,
	ide_pm_restore_dma,
};

static void ide_complete_power_step(ide_drive_t *drive, struct request *rq, u8 stat, u8 error)
{
	struct request_pm_state *pm = rq->data;

	if (drive->media != ide_disk)
		return;

	switch (pm->pm_step) {
	case ide_pm_flush_cache:	/* Suspend step 1 (flush cache) complete */
		if (pm->pm_state == PM_EVENT_FREEZE)
			pm->pm_step = ide_pm_state_completed;
		else
			pm->pm_step = idedisk_pm_standby;
		break;
	case idedisk_pm_standby:	/* Suspend step 2 (standby) complete */
		pm->pm_step = ide_pm_state_completed;
		break;
	case idedisk_pm_restore_pio:	/* Resume step 1 complete */
		pm->pm_step = idedisk_pm_idle;
		break;
	case idedisk_pm_idle:		/* Resume step 2 (idle) complete */
		pm->pm_step = ide_pm_restore_dma;
		break;
	}
}

static ide_startstop_t ide_start_power_step(ide_drive_t *drive, struct request *rq)
{
	struct request_pm_state *pm = rq->data;
	ide_task_t *args = rq->special;

	memset(args, 0, sizeof(*args));

	switch (pm->pm_step) {
	case ide_pm_flush_cache:	/* Suspend step 1 (flush cache) */
		if (drive->media != ide_disk)
			break;
		/* Not supported? Switch to next step now. */
		if (!drive->wcache || !ide_id_has_flush_cache(drive->id)) {
			ide_complete_power_step(drive, rq, 0, 0);
			return ide_stopped;
		}
		if (ide_id_has_flush_cache_ext(drive->id))
			args->tf.command = WIN_FLUSH_CACHE_EXT;
		else
			args->tf.command = WIN_FLUSH_CACHE;
		goto out_do_tf;

	case idedisk_pm_standby:	/* Suspend step 2 (standby) */
		args->tf.command = WIN_STANDBYNOW1;
		goto out_do_tf;

	case idedisk_pm_restore_pio:	/* Resume step 1 (restore PIO) */
		ide_set_max_pio(drive);
		/*
		 * skip idedisk_pm_idle for ATAPI devices
		 */
		if (drive->media != ide_disk)
			pm->pm_step = ide_pm_restore_dma;
		else
			ide_complete_power_step(drive, rq, 0, 0);
		return ide_stopped;

	case idedisk_pm_idle:		/* Resume step 2 (idle) */
		args->tf.command = WIN_IDLEIMMEDIATE;
		goto out_do_tf;

	case ide_pm_restore_dma:	/* Resume step 3 (restore DMA) */
		/*
		 * Right now, all we do is call ide_set_dma(drive),
		 * we could be smarter and check for current xfer_speed
		 * in struct drive etc...
		 */
		if (drive->hwif->dma_host_set == NULL)
			break;
		/*
		 * TODO: respect ->using_dma setting
		 */
		ide_set_dma(drive);
		break;
	}
	pm->pm_step = ide_pm_state_completed;
	return ide_stopped;

out_do_tf:
	args->tf_flags	 = IDE_TFLAG_OUT_TF | IDE_TFLAG_OUT_DEVICE;
	args->data_phase = TASKFILE_NO_DATA;
	return do_rw_taskfile(drive, args);
}

/**
 *	ide_end_dequeued_request	-	complete an IDE I/O
 *	@drive: IDE device for the I/O
 *	@uptodate:
 *	@nr_sectors: number of sectors completed
 *
 *	Complete an I/O that is no longer on the request queue. This
 *	typically occurs when we pull the request and issue a REQUEST_SENSE.
 *	We must still finish the old request but we must not tamper with the
 *	queue in the meantime.
 *
 *	NOTE: This path does not handle barrier, but barrier is not supported
 *	on ide-cd anyway.
 */

int ide_end_dequeued_request(ide_drive_t *drive, struct request *rq,
			     int uptodate, int nr_sectors)
{
	unsigned long flags;
	int ret;

	spin_lock_irqsave(&ide_lock, flags);
	BUG_ON(!blk_rq_started(rq));
	ret = __ide_end_request(drive, rq, uptodate, nr_sectors << 9, 0);
	spin_unlock_irqrestore(&ide_lock, flags);

	return ret;
}
EXPORT_SYMBOL_GPL(ide_end_dequeued_request);


/**
 *	ide_complete_pm_request - end the current Power Management request
 *	@drive: target drive
 *	@rq: request
 *
 *	This function cleans up the current PM request and stops the queue
 *	if necessary.
 */
static void ide_complete_pm_request (ide_drive_t *drive, struct request *rq)
{
	unsigned long flags;

#ifdef DEBUG_PM
	printk("%s: completing PM request, %s\n", drive->name,
	       blk_pm_suspend_request(rq) ? "suspend" : "resume");
#endif
	spin_lock_irqsave(&ide_lock, flags);
	if (blk_pm_suspend_request(rq)) {
		blk_stop_queue(drive->queue);
	} else {
		drive->blocked = 0;
		blk_start_queue(drive->queue);
	}
	blkdev_dequeue_request(rq);
	HWGROUP(drive)->rq = NULL;
	end_that_request_last(rq, 1);
	spin_unlock_irqrestore(&ide_lock, flags);
}

void ide_tf_read(ide_drive_t *drive, ide_task_t *task)
{
	ide_hwif_t *hwif = drive->hwif;
	struct ide_taskfile *tf = &task->tf;

	if (task->tf_flags & IDE_TFLAG_IN_DATA) {
		u16 data = hwif->INW(IDE_DATA_REG);

		tf->data = data & 0xff;
		tf->hob_data = (data >> 8) & 0xff;
	}

	/* be sure we're looking at the low order bits */
	hwif->OUTB(drive->ctl & ~0x80, IDE_CONTROL_REG);

	if (task->tf_flags & IDE_TFLAG_IN_NSECT)
		tf->nsect  = hwif->INB(IDE_NSECTOR_REG);
	if (task->tf_flags & IDE_TFLAG_IN_LBAL)
		tf->lbal   = hwif->INB(IDE_SECTOR_REG);
	if (task->tf_flags & IDE_TFLAG_IN_LBAM)
		tf->lbam   = hwif->INB(IDE_LCYL_REG);
	if (task->tf_flags & IDE_TFLAG_IN_LBAH)
		tf->lbah   = hwif->INB(IDE_HCYL_REG);
	if (task->tf_flags & IDE_TFLAG_IN_DEVICE)
		tf->device = hwif->INB(IDE_SELECT_REG);

	if (task->tf_flags & IDE_TFLAG_LBA48) {
		hwif->OUTB(drive->ctl | 0x80, IDE_CONTROL_REG);

		if (task->tf_flags & IDE_TFLAG_IN_HOB_FEATURE)
			tf->hob_feature = hwif->INB(IDE_FEATURE_REG);
		if (task->tf_flags & IDE_TFLAG_IN_HOB_NSECT)
			tf->hob_nsect   = hwif->INB(IDE_NSECTOR_REG);
		if (task->tf_flags & IDE_TFLAG_IN_HOB_LBAL)
			tf->hob_lbal    = hwif->INB(IDE_SECTOR_REG);
		if (task->tf_flags & IDE_TFLAG_IN_HOB_LBAM)
			tf->hob_lbam    = hwif->INB(IDE_LCYL_REG);
		if (task->tf_flags & IDE_TFLAG_IN_HOB_LBAH)
			tf->hob_lbah    = hwif->INB(IDE_HCYL_REG);
	}
}

/**
 *	ide_end_drive_cmd	-	end an explicit drive command
 *	@drive: command 
 *	@stat: status bits
 *	@err: error bits
 *
 *	Clean up after success/failure of an explicit drive command.
 *	These get thrown onto the queue so they are synchronized with
 *	real I/O operations on the drive.
 *
 *	In LBA48 mode we have to read the register set twice to get
 *	all the extra information out.
 */
 
void ide_end_drive_cmd (ide_drive_t *drive, u8 stat, u8 err)
{
	ide_hwif_t *hwif = HWIF(drive);
	unsigned long flags;
	struct request *rq;

	spin_lock_irqsave(&ide_lock, flags);
	rq = HWGROUP(drive)->rq;
	spin_unlock_irqrestore(&ide_lock, flags);

	if (rq->cmd_type == REQ_TYPE_ATA_CMD) {
		u8 *args = (u8 *) rq->buffer;
		if (rq->errors == 0)
			rq->errors = !OK_STAT(stat,READY_STAT,BAD_STAT);

		if (args) {
			args[0] = stat;
			args[1] = err;
			/* be sure we're looking at the low order bits */
			hwif->OUTB(drive->ctl & ~0x80, IDE_CONTROL_REG);
			args[2] = hwif->INB(IDE_NSECTOR_REG);
		}
	} else if (rq->cmd_type == REQ_TYPE_ATA_TASKFILE) {
		ide_task_t *args = (ide_task_t *) rq->special;
		if (rq->errors == 0)
			rq->errors = !OK_STAT(stat,READY_STAT,BAD_STAT);
			
		if (args) {
			struct ide_taskfile *tf = &args->tf;

			tf->error = err;
			tf->status = stat;

			args->tf_flags |= (IDE_TFLAG_IN_TF|IDE_TFLAG_IN_DEVICE);
			if (args->tf_flags & IDE_TFLAG_LBA48)
				args->tf_flags |= IDE_TFLAG_IN_HOB;

			ide_tf_read(drive, args);
		}
	} else if (blk_pm_request(rq)) {
		struct request_pm_state *pm = rq->data;
#ifdef DEBUG_PM
		printk("%s: complete_power_step(step: %d, stat: %x, err: %x)\n",
			drive->name, rq->pm->pm_step, stat, err);
#endif
		ide_complete_power_step(drive, rq, stat, err);
		if (pm->pm_step == ide_pm_state_completed)
			ide_complete_pm_request(drive, rq);
		return;
	}

	spin_lock_irqsave(&ide_lock, flags);
	blkdev_dequeue_request(rq);
	HWGROUP(drive)->rq = NULL;
	rq->errors = err;
	end_that_request_last(rq, !rq->errors);
	spin_unlock_irqrestore(&ide_lock, flags);
}

EXPORT_SYMBOL(ide_end_drive_cmd);

/**
 *	try_to_flush_leftover_data	-	flush junk
 *	@drive: drive to flush
 *
 *	try_to_flush_leftover_data() is invoked in response to a drive
 *	unexpectedly having its DRQ_STAT bit set.  As an alternative to
 *	resetting the drive, this routine tries to clear the condition
 *	by read a sector's worth of data from the drive.  Of course,
 *	this may not help if the drive is *waiting* for data from *us*.
 */
static void try_to_flush_leftover_data (ide_drive_t *drive)
{
	int i = (drive->mult_count ? drive->mult_count : 1) * SECTOR_WORDS;

	if (drive->media != ide_disk)
		return;
	while (i > 0) {
		u32 buffer[16];
		u32 wcount = (i > 16) ? 16 : i;

		i -= wcount;
		HWIF(drive)->ata_input_data(drive, buffer, wcount);
	}
}

static void ide_kill_rq(ide_drive_t *drive, struct request *rq)
{
	if (rq->rq_disk) {
		ide_driver_t *drv;

		drv = *(ide_driver_t **)rq->rq_disk->private_data;
		drv->end_request(drive, 0, 0);
	} else
		ide_end_request(drive, 0, 0);
}

static ide_startstop_t ide_ata_error(ide_drive_t *drive, struct request *rq, u8 stat, u8 err)
{
	ide_hwif_t *hwif = drive->hwif;

	if (stat & BUSY_STAT || ((stat & WRERR_STAT) && !drive->nowerr)) {
		/* other bits are useless when BUSY */
		rq->errors |= ERROR_RESET;
	} else if (stat & ERR_STAT) {
		/* err has different meaning on cdrom and tape */
		if (err == ABRT_ERR) {
			if (drive->select.b.lba &&
			    /* some newer drives don't support WIN_SPECIFY */
			    hwif->INB(IDE_COMMAND_REG) == WIN_SPECIFY)
				return ide_stopped;
		} else if ((err & BAD_CRC) == BAD_CRC) {
			/* UDMA crc error, just retry the operation */
			drive->crc_count++;
		} else if (err & (BBD_ERR | ECC_ERR)) {
			/* retries won't help these */
			rq->errors = ERROR_MAX;
		} else if (err & TRK0_ERR) {
			/* help it find track zero */
			rq->errors |= ERROR_RECAL;
		}
	}

	if ((stat & DRQ_STAT) && rq_data_dir(rq) == READ &&
	    (hwif->host_flags & IDE_HFLAG_ERROR_STOPS_FIFO) == 0)
		try_to_flush_leftover_data(drive);

	if (rq->errors >= ERROR_MAX || blk_noretry_request(rq)) {
		ide_kill_rq(drive, rq);
		return ide_stopped;
	}

	if (hwif->INB(IDE_STATUS_REG) & (BUSY_STAT|DRQ_STAT))
		rq->errors |= ERROR_RESET;

	if ((rq->errors & ERROR_RESET) == ERROR_RESET) {
		++rq->errors;
		return ide_do_reset(drive);
	}

	if ((rq->errors & ERROR_RECAL) == ERROR_RECAL)
		drive->special.b.recalibrate = 1;

	++rq->errors;

	return ide_stopped;
}

static ide_startstop_t ide_atapi_error(ide_drive_t *drive, struct request *rq, u8 stat, u8 err)
{
	ide_hwif_t *hwif = drive->hwif;

	if (stat & BUSY_STAT || ((stat & WRERR_STAT) && !drive->nowerr)) {
		/* other bits are useless when BUSY */
		rq->errors |= ERROR_RESET;
	} else {
		/* add decoding error stuff */
	}

	if (hwif->INB(IDE_STATUS_REG) & (BUSY_STAT|DRQ_STAT))
		/* force an abort */
		hwif->OUTB(WIN_IDLEIMMEDIATE, IDE_COMMAND_REG);

	if (rq->errors >= ERROR_MAX) {
		ide_kill_rq(drive, rq);
	} else {
		if ((rq->errors & ERROR_RESET) == ERROR_RESET) {
			++rq->errors;
			return ide_do_reset(drive);
		}
		++rq->errors;
	}

	return ide_stopped;
}

ide_startstop_t
__ide_error(ide_drive_t *drive, struct request *rq, u8 stat, u8 err)
{
	if (drive->media == ide_disk)
		return ide_ata_error(drive, rq, stat, err);
	return ide_atapi_error(drive, rq, stat, err);
}

EXPORT_SYMBOL_GPL(__ide_error);

/**
 *	ide_error	-	handle an error on the IDE
 *	@drive: drive the error occurred on
 *	@msg: message to report
 *	@stat: status bits
 *
 *	ide_error() takes action based on the error returned by the drive.
 *	For normal I/O that may well include retries. We deal with
 *	both new-style (taskfile) and old style command handling here.
 *	In the case of taskfile command handling there is work left to
 *	do
 */
 
ide_startstop_t ide_error (ide_drive_t *drive, const char *msg, u8 stat)
{
	struct request *rq;
	u8 err;

	err = ide_dump_status(drive, msg, stat);

	if ((rq = HWGROUP(drive)->rq) == NULL)
		return ide_stopped;

	/* retry only "normal" I/O: */
	if (!blk_fs_request(rq)) {
		rq->errors = 1;
		ide_end_drive_cmd(drive, stat, err);
		return ide_stopped;
	}

	if (rq->rq_disk) {
		ide_driver_t *drv;

		drv = *(ide_driver_t **)rq->rq_disk->private_data;
		return drv->error(drive, rq, stat, err);
	} else
		return __ide_error(drive, rq, stat, err);
}

EXPORT_SYMBOL_GPL(ide_error);

ide_startstop_t __ide_abort(ide_drive_t *drive, struct request *rq)
{
	if (drive->media != ide_disk)
		rq->errors |= ERROR_RESET;

	ide_kill_rq(drive, rq);

	return ide_stopped;
}

EXPORT_SYMBOL_GPL(__ide_abort);

/**
 *	ide_abort	-	abort pending IDE operations
 *	@drive: drive the error occurred on
 *	@msg: message to report
 *
 *	ide_abort kills and cleans up when we are about to do a 
 *	host initiated reset on active commands. Longer term we
 *	want handlers to have sensible abort handling themselves
 *
 *	This differs fundamentally from ide_error because in 
 *	this case the command is doing just fine when we
 *	blow it away.
 */
 
ide_startstop_t ide_abort(ide_drive_t *drive, const char *msg)
{
	struct request *rq;

	if (drive == NULL || (rq = HWGROUP(drive)->rq) == NULL)
		return ide_stopped;

	/* retry only "normal" I/O: */
	if (!blk_fs_request(rq)) {
		rq->errors = 1;
		ide_end_drive_cmd(drive, BUSY_STAT, 0);
		return ide_stopped;
	}

	if (rq->rq_disk) {
		ide_driver_t *drv;

		drv = *(ide_driver_t **)rq->rq_disk->private_data;
		return drv->abort(drive, rq);
	} else
		return __ide_abort(drive, rq);
}

/**
 *	drive_cmd_intr		- 	drive command completion interrupt
 *	@drive: drive the completion interrupt occurred on
 *
 *	drive_cmd_intr() is invoked on completion of a special DRIVE_CMD.
 *	We do any necessary data reading and then wait for the drive to
 *	go non busy. At that point we may read the error data and complete
 *	the request
 */
 
static ide_startstop_t drive_cmd_intr (ide_drive_t *drive)
{
	struct request *rq = HWGROUP(drive)->rq;
	ide_hwif_t *hwif = HWIF(drive);
	u8 *args = (u8 *) rq->buffer;
	u8 stat = hwif->INB(IDE_STATUS_REG);
	int retries = 10;

	local_irq_enable_in_hardirq();
	if (rq->cmd_type == REQ_TYPE_ATA_CMD &&
	    (stat & DRQ_STAT) && args && args[3]) {
		u8 io_32bit = drive->io_32bit;
		drive->io_32bit = 0;
		hwif->ata_input_data(drive, &args[4], args[3] * SECTOR_WORDS);
		drive->io_32bit = io_32bit;
		while (((stat = hwif->INB(IDE_STATUS_REG)) & BUSY_STAT) && retries--)
			udelay(100);
	}

	if (!OK_STAT(stat, READY_STAT, BAD_STAT))
		return ide_error(drive, "drive_cmd", stat);
		/* calls ide_end_drive_cmd */
	ide_end_drive_cmd(drive, stat, hwif->INB(IDE_ERROR_REG));
	return ide_stopped;
}

static void ide_tf_set_specify_cmd(ide_drive_t *drive, struct ide_taskfile *tf)
{
	tf->nsect   = drive->sect;
	tf->lbal    = drive->sect;
	tf->lbam    = drive->cyl;
	tf->lbah    = drive->cyl >> 8;
	tf->device  = ((drive->head - 1) | drive->select.all) & ~ATA_LBA;
	tf->command = WIN_SPECIFY;
}

static void ide_tf_set_restore_cmd(ide_drive_t *drive, struct ide_taskfile *tf)
{
	tf->nsect   = drive->sect;
	tf->command = WIN_RESTORE;
}

static void ide_tf_set_setmult_cmd(ide_drive_t *drive, struct ide_taskfile *tf)
{
	tf->nsect   = drive->mult_req;
	tf->command = WIN_SETMULT;
}

static ide_startstop_t ide_disk_special(ide_drive_t *drive)
{
	special_t *s = &drive->special;
	ide_task_t args;

	memset(&args, 0, sizeof(ide_task_t));
	args.data_phase = TASKFILE_NO_DATA;

	if (s->b.set_geometry) {
		s->b.set_geometry = 0;
		ide_tf_set_specify_cmd(drive, &args.tf);
	} else if (s->b.recalibrate) {
		s->b.recalibrate = 0;
		ide_tf_set_restore_cmd(drive, &args.tf);
	} else if (s->b.set_multmode) {
		s->b.set_multmode = 0;
		if (drive->mult_req > drive->id->max_multsect)
			drive->mult_req = drive->id->max_multsect;
		ide_tf_set_setmult_cmd(drive, &args.tf);
	} else if (s->all) {
		int special = s->all;
		s->all = 0;
		printk(KERN_ERR "%s: bad special flag: 0x%02x\n", drive->name, special);
		return ide_stopped;
	}

	args.tf_flags = IDE_TFLAG_OUT_TF | IDE_TFLAG_OUT_DEVICE |
			IDE_TFLAG_CUSTOM_HANDLER;

	do_rw_taskfile(drive, &args);

	return ide_started;
}

/*
 * handle HDIO_SET_PIO_MODE ioctl abusers here, eventually it will go away
 */
static int set_pio_mode_abuse(ide_hwif_t *hwif, u8 req_pio)
{
	switch (req_pio) {
	case 202:
	case 201:
	case 200:
	case 102:
	case 101:
	case 100:
		return (hwif->host_flags & IDE_HFLAG_ABUSE_DMA_MODES) ? 1 : 0;
	case 9:
	case 8:
		return (hwif->host_flags & IDE_HFLAG_ABUSE_PREFETCH) ? 1 : 0;
	case 7:
	case 6:
		return (hwif->host_flags & IDE_HFLAG_ABUSE_FAST_DEVSEL) ? 1 : 0;
	default:
		return 0;
	}
}

/**
 *	do_special		-	issue some special commands
 *	@drive: drive the command is for
 *
 *	do_special() is used to issue WIN_SPECIFY, WIN_RESTORE, and WIN_SETMULT
 *	commands to a drive.  It used to do much more, but has been scaled
 *	back.
 */

static ide_startstop_t do_special (ide_drive_t *drive)
{
	special_t *s = &drive->special;

#ifdef DEBUG
	printk("%s: do_special: 0x%02x\n", drive->name, s->all);
#endif
	if (s->b.set_tune) {
		ide_hwif_t *hwif = drive->hwif;
		u8 req_pio = drive->tune_req;

		s->b.set_tune = 0;

		if (set_pio_mode_abuse(drive->hwif, req_pio)) {

			if (hwif->set_pio_mode == NULL)
				return ide_stopped;

			/*
			 * take ide_lock for drive->[no_]unmask/[no_]io_32bit
			 */
			if (req_pio == 8 || req_pio == 9) {
				unsigned long flags;

				spin_lock_irqsave(&ide_lock, flags);
				hwif->set_pio_mode(drive, req_pio);
				spin_unlock_irqrestore(&ide_lock, flags);
			} else
				hwif->set_pio_mode(drive, req_pio);
		} else {
			int keep_dma = drive->using_dma;

			ide_set_pio(drive, req_pio);

			if (hwif->host_flags & IDE_HFLAG_SET_PIO_MODE_KEEP_DMA) {
				if (keep_dma)
					ide_dma_on(drive);
			}
		}

		return ide_stopped;
	} else {
		if (drive->media == ide_disk)
			return ide_disk_special(drive);

		s->all = 0;
		drive->mult_req = 0;
		return ide_stopped;
	}
}

void ide_map_sg(ide_drive_t *drive, struct request *rq)
{
	ide_hwif_t *hwif = drive->hwif;
	struct scatterlist *sg = hwif->sg_table;

	if (hwif->sg_mapped)	/* needed by ide-scsi */
		return;

	if (rq->cmd_type != REQ_TYPE_ATA_TASKFILE) {
		hwif->sg_nents = blk_rq_map_sg(drive->queue, rq, sg);
	} else {
		sg_init_one(sg, rq->buffer, rq->nr_sectors * SECTOR_SIZE);
		hwif->sg_nents = 1;
	}
}

EXPORT_SYMBOL_GPL(ide_map_sg);

void ide_init_sg_cmd(ide_drive_t *drive, struct request *rq)
{
	ide_hwif_t *hwif = drive->hwif;

	hwif->nsect = hwif->nleft = rq->nr_sectors;
	hwif->cursg_ofs = 0;
	hwif->cursg = NULL;
}

EXPORT_SYMBOL_GPL(ide_init_sg_cmd);

/**
 *	execute_drive_command	-	issue special drive command
 *	@drive: the drive to issue the command on
 *	@rq: the request structure holding the command
 *
 *	execute_drive_cmd() issues a special drive command,  usually 
 *	initiated by ioctl() from the external hdparm program. The
 *	command can be a drive command, drive task or taskfile 
 *	operation. Weirdly you can call it with NULL to wait for
 *	all commands to finish. Don't do this as that is due to change
 */

static ide_startstop_t execute_drive_cmd (ide_drive_t *drive,
		struct request *rq)
{
	ide_hwif_t *hwif = HWIF(drive);
	u8 *args = rq->buffer;
	ide_task_t ltask;
	struct ide_taskfile *tf = &ltask.tf;

	if (rq->cmd_type == REQ_TYPE_ATA_TASKFILE) {
		ide_task_t *task = rq->special;
 
		if (task == NULL)
			goto done;

		hwif->data_phase = task->data_phase;

		switch (hwif->data_phase) {
		case TASKFILE_MULTI_OUT:
		case TASKFILE_OUT:
		case TASKFILE_MULTI_IN:
		case TASKFILE_IN:
			ide_init_sg_cmd(drive, rq);
			ide_map_sg(drive, rq);
		default:
			break;
		}

		return do_rw_taskfile(drive, task);
	}

	if (args == NULL)
		goto done;

	memset(&ltask, 0, sizeof(ltask));
	if (rq->cmd_type == REQ_TYPE_ATA_CMD) {
#ifdef DEBUG
		printk("%s: DRIVE_CMD\n", drive->name);
#endif
		tf->feature = args[2];
		if (args[0] == WIN_SMART) {
			tf->nsect = args[3];
			tf->lbal  = args[1];
			tf->lbam  = 0x4f;
			tf->lbah  = 0xc2;
			ltask.tf_flags = IDE_TFLAG_OUT_TF;
		} else {
			tf->nsect = args[1];
			ltask.tf_flags = IDE_TFLAG_OUT_FEATURE |
					 IDE_TFLAG_OUT_NSECT;
		}
 	}
	tf->command = args[0];
	ide_tf_load(drive, &ltask);
	ide_execute_command(drive, args[0], &drive_cmd_intr, WAIT_WORSTCASE, NULL);
	return ide_started;

done:
 	/*
 	 * NULL is actually a valid way of waiting for
 	 * all current requests to be flushed from the queue.
 	 */
#ifdef DEBUG
 	printk("%s: DRIVE_CMD (null)\n", drive->name);
#endif
 	ide_end_drive_cmd(drive,
			hwif->INB(IDE_STATUS_REG),
			hwif->INB(IDE_ERROR_REG));
 	return ide_stopped;
}

static void ide_check_pm_state(ide_drive_t *drive, struct request *rq)
{
	struct request_pm_state *pm = rq->data;

	if (blk_pm_suspend_request(rq) &&
	    pm->pm_step == ide_pm_state_start_suspend)
		/* Mark drive blocked when starting the suspend sequence. */
		drive->blocked = 1;
	else if (blk_pm_resume_request(rq) &&
		 pm->pm_step == ide_pm_state_start_resume) {
		/* 
		 * The first thing we do on wakeup is to wait for BSY bit to
		 * go away (with a looong timeout) as a drive on this hwif may
		 * just be POSTing itself.
		 * We do that before even selecting as the "other" device on
		 * the bus may be broken enough to walk on our toes at this
		 * point.
		 */
		int rc;
#ifdef DEBUG_PM
		printk("%s: Wakeup request inited, waiting for !BSY...\n", drive->name);
#endif
		rc = ide_wait_not_busy(HWIF(drive), 35000);
		if (rc)
			printk(KERN_WARNING "%s: bus not ready on wakeup\n", drive->name);
		SELECT_DRIVE(drive);
		ide_set_irq(drive, 1);
		rc = ide_wait_not_busy(HWIF(drive), 100000);
		if (rc)
			printk(KERN_WARNING "%s: drive not ready on wakeup\n", drive->name);
	}
}

/**
 *	start_request	-	start of I/O and command issuing for IDE
 *
 *	start_request() initiates handling of a new I/O request. It
 *	accepts commands and I/O (read/write) requests. It also does
 *	the final remapping for weird stuff like EZDrive. Once 
 *	device mapper can work sector level the EZDrive stuff can go away
 *
 *	FIXME: this function needs a rename
 */
 
static ide_startstop_t start_request (ide_drive_t *drive, struct request *rq)
{
	ide_startstop_t startstop;
	sector_t block;

	BUG_ON(!blk_rq_started(rq));

#ifdef DEBUG
	printk("%s: start_request: current=0x%08lx\n",
		HWIF(drive)->name, (unsigned long) rq);
#endif

	/* bail early if we've exceeded max_failures */
	if (drive->max_failures && (drive->failures > drive->max_failures)) {
		rq->cmd_flags |= REQ_FAILED;
		goto kill_rq;
	}

	block    = rq->sector;
	if (blk_fs_request(rq) &&
	    (drive->media == ide_disk || drive->media == ide_floppy)) {
		block += drive->sect0;
	}
	/* Yecch - this will shift the entire interval,
	   possibly killing some innocent following sector */
	if (block == 0 && drive->remap_0_to_1 == 1)
		block = 1;  /* redirect MBR access to EZ-Drive partn table */

	if (blk_pm_request(rq))
		ide_check_pm_state(drive, rq);

	SELECT_DRIVE(drive);
	if (ide_wait_stat(&startstop, drive, drive->ready_stat, BUSY_STAT|DRQ_STAT, WAIT_READY)) {
		printk(KERN_ERR "%s: drive not ready for command\n", drive->name);
		return startstop;
	}
	if (!drive->special.all) {
		ide_driver_t *drv;

		/*
		 * We reset the drive so we need to issue a SETFEATURES.
		 * Do it _after_ do_special() restored device parameters.
		 */
		if (drive->current_speed == 0xff)
			ide_config_drive_speed(drive, drive->desired_speed);

		if (rq->cmd_type == REQ_TYPE_ATA_CMD ||
		    rq->cmd_type == REQ_TYPE_ATA_TASKFILE)
			return execute_drive_cmd(drive, rq);
		else if (blk_pm_request(rq)) {
			struct request_pm_state *pm = rq->data;
#ifdef DEBUG_PM
			printk("%s: start_power_step(step: %d)\n",
				drive->name, rq->pm->pm_step);
#endif
			startstop = ide_start_power_step(drive, rq);
			if (startstop == ide_stopped &&
			    pm->pm_step == ide_pm_state_completed)
				ide_complete_pm_request(drive, rq);
			return startstop;
		}

		drv = *(ide_driver_t **)rq->rq_disk->private_data;
		return drv->do_request(drive, rq, block);
	}
	return do_special(drive);
kill_rq:
	ide_kill_rq(drive, rq);
	return ide_stopped;
}

/**
 *	ide_stall_queue		-	pause an IDE device
 *	@drive: drive to stall
 *	@timeout: time to stall for (jiffies)
 *
 *	ide_stall_queue() can be used by a drive to give excess bandwidth back
 *	to the hwgroup by sleeping for timeout jiffies.
 */
 
void ide_stall_queue (ide_drive_t *drive, unsigned long timeout)
{
	if (timeout > WAIT_WORSTCASE)
		timeout = WAIT_WORSTCASE;
	drive->sleep = timeout + jiffies;
	drive->sleeping = 1;
}

EXPORT_SYMBOL(ide_stall_queue);

#define WAKEUP(drive)	((drive)->service_start + 2 * (drive)->service_time)

/**
 *	choose_drive		-	select a drive to service
 *	@hwgroup: hardware group to select on
 *
 *	choose_drive() selects the next drive which will be serviced.
 *	This is necessary because the IDE layer can't issue commands
 *	to both drives on the same cable, unlike SCSI.
 */
 
static inline ide_drive_t *choose_drive (ide_hwgroup_t *hwgroup)
{
	ide_drive_t *drive, *best;

repeat:	
	best = NULL;
	drive = hwgroup->drive;

	/*
	 * drive is doing pre-flush, ordered write, post-flush sequence. even
	 * though that is 3 requests, it must be seen as a single transaction.
	 * we must not preempt this drive until that is complete
	 */
	if (blk_queue_flushing(drive->queue)) {
		/*
		 * small race where queue could get replugged during
		 * the 3-request flush cycle, just yank the plug since
		 * we want it to finish asap
		 */
		blk_remove_plug(drive->queue);
		return drive;
	}

	do {
		if ((!drive->sleeping || time_after_eq(jiffies, drive->sleep))
		    && !elv_queue_empty(drive->queue)) {
			if (!best
			 || (drive->sleeping && (!best->sleeping || time_before(drive->sleep, best->sleep)))
			 || (!best->sleeping && time_before(WAKEUP(drive), WAKEUP(best))))
			{
				if (!blk_queue_plugged(drive->queue))
					best = drive;
			}
		}
	} while ((drive = drive->next) != hwgroup->drive);
	if (best && best->nice1 && !best->sleeping && best != hwgroup->drive && best->service_time > WAIT_MIN_SLEEP) {
		long t = (signed long)(WAKEUP(best) - jiffies);
		if (t >= WAIT_MIN_SLEEP) {
		/*
		 * We *may* have some time to spare, but first let's see if
		 * someone can potentially benefit from our nice mood today..
		 */
			drive = best->next;
			do {
				if (!drive->sleeping
				 && time_before(jiffies - best->service_time, WAKEUP(drive))
				 && time_before(WAKEUP(drive), jiffies + t))
				{
					ide_stall_queue(best, min_t(long, t, 10 * WAIT_MIN_SLEEP));
					goto repeat;
				}
			} while ((drive = drive->next) != best);
		}
	}
	return best;
}

/*
 * Issue a new request to a drive from hwgroup
 * Caller must have already done spin_lock_irqsave(&ide_lock, ..);
 *
 * A hwgroup is a serialized group of IDE interfaces.  Usually there is
 * exactly one hwif (interface) per hwgroup, but buggy controllers (eg. CMD640)
 * may have both interfaces in a single hwgroup to "serialize" access.
 * Or possibly multiple ISA interfaces can share a common IRQ by being grouped
 * together into one hwgroup for serialized access.
 *
 * Note also that several hwgroups can end up sharing a single IRQ,
 * possibly along with many other devices.  This is especially common in
 * PCI-based systems with off-board IDE controller cards.
 *
 * The IDE driver uses the single global ide_lock spinlock to protect
 * access to the request queues, and to protect the hwgroup->busy flag.
 *
 * The first thread into the driver for a particular hwgroup sets the
 * hwgroup->busy flag to indicate that this hwgroup is now active,
 * and then initiates processing of the top request from the request queue.
 *
 * Other threads attempting entry notice the busy setting, and will simply
 * queue their new requests and exit immediately.  Note that hwgroup->busy
 * remains set even when the driver is merely awaiting the next interrupt.
 * Thus, the meaning is "this hwgroup is busy processing a request".
 *
 * When processing of a request completes, the completing thread or IRQ-handler
 * will start the next request from the queue.  If no more work remains,
 * the driver will clear the hwgroup->busy flag and exit.
 *
 * The ide_lock (spinlock) is used to protect all access to the
 * hwgroup->busy flag, but is otherwise not needed for most processing in
 * the driver.  This makes the driver much more friendlier to shared IRQs
 * than previous designs, while remaining 100% (?) SMP safe and capable.
 */
static void ide_do_request (ide_hwgroup_t *hwgroup, int masked_irq)
{
	ide_drive_t	*drive;
	ide_hwif_t	*hwif;
	struct request	*rq;
	ide_startstop_t	startstop;
	int             loops = 0;

	/* for atari only: POSSIBLY BROKEN HERE(?) */
	ide_get_lock(ide_intr, hwgroup);

	/* caller must own ide_lock */
	BUG_ON(!irqs_disabled());

	while (!hwgroup->busy) {
		hwgroup->busy = 1;
		drive = choose_drive(hwgroup);
		if (drive == NULL) {
			int sleeping = 0;
			unsigned long sleep = 0; /* shut up, gcc */
			hwgroup->rq = NULL;
			drive = hwgroup->drive;
			do {
				if (drive->sleeping && (!sleeping || time_before(drive->sleep, sleep))) {
					sleeping = 1;
					sleep = drive->sleep;
				}
			} while ((drive = drive->next) != hwgroup->drive);
			if (sleeping) {
		/*
		 * Take a short snooze, and then wake up this hwgroup again.
		 * This gives other hwgroups on the same a chance to
		 * play fairly with us, just in case there are big differences
		 * in relative throughputs.. don't want to hog the cpu too much.
		 */
				if (time_before(sleep, jiffies + WAIT_MIN_SLEEP))
					sleep = jiffies + WAIT_MIN_SLEEP;
#if 1
				if (timer_pending(&hwgroup->timer))
					printk(KERN_CRIT "ide_set_handler: timer already active\n");
#endif
				/* so that ide_timer_expiry knows what to do */
				hwgroup->sleeping = 1;
				hwgroup->req_gen_timer = hwgroup->req_gen;
				mod_timer(&hwgroup->timer, sleep);
				/* we purposely leave hwgroup->busy==1
				 * while sleeping */
			} else {
				/* Ugly, but how can we sleep for the lock
				 * otherwise? perhaps from tq_disk?
				 */

				/* for atari only */
				ide_release_lock();
				hwgroup->busy = 0;
			}

			/* no more work for this hwgroup (for now) */
			return;
		}
	again:
		hwif = HWIF(drive);
		if (hwgroup->hwif->sharing_irq && hwif != hwgroup->hwif) {
			/*
			 * set nIEN for previous hwif, drives in the
			 * quirk_list may not like intr setups/cleanups
			 */
			if (drive->quirk_list != 1)
				ide_set_irq(drive, 0);
		}
		hwgroup->hwif = hwif;
		hwgroup->drive = drive;
		drive->sleeping = 0;
		drive->service_start = jiffies;

		if (blk_queue_plugged(drive->queue)) {
			printk(KERN_ERR "ide: huh? queue was plugged!\n");
			break;
		}

		/*
		 * we know that the queue isn't empty, but this can happen
		 * if the q->prep_rq_fn() decides to kill a request
		 */
		rq = elv_next_request(drive->queue);
		if (!rq) {
			hwgroup->busy = 0;
			break;
		}

		/*
		 * Sanity: don't accept a request that isn't a PM request
		 * if we are currently power managed. This is very important as
		 * blk_stop_queue() doesn't prevent the elv_next_request()
		 * above to return us whatever is in the queue. Since we call
		 * ide_do_request() ourselves, we end up taking requests while
		 * the queue is blocked...
		 * 
		 * We let requests forced at head of queue with ide-preempt
		 * though. I hope that doesn't happen too much, hopefully not
		 * unless the subdriver triggers such a thing in its own PM
		 * state machine.
		 *
		 * We count how many times we loop here to make sure we service
		 * all drives in the hwgroup without looping for ever
		 */
		if (drive->blocked && !blk_pm_request(rq) && !(rq->cmd_flags & REQ_PREEMPT)) {
			drive = drive->next ? drive->next : hwgroup->drive;
			if (loops++ < 4 && !blk_queue_plugged(drive->queue))
				goto again;
			/* We clear busy, there should be no pending ATA command at this point. */
			hwgroup->busy = 0;
			break;
		}

		hwgroup->rq = rq;

		/*
		 * Some systems have trouble with IDE IRQs arriving while
		 * the driver is still setting things up.  So, here we disable
		 * the IRQ used by this interface while the request is being started.
		 * This may look bad at first, but pretty much the same thing
		 * happens anyway when any interrupt comes in, IDE or otherwise
		 *  -- the kernel masks the IRQ while it is being handled.
		 */
		if (masked_irq != IDE_NO_IRQ && hwif->irq != masked_irq)
			disable_irq_nosync(hwif->irq);
		spin_unlock(&ide_lock);
		local_irq_enable_in_hardirq();
			/* allow other IRQs while we start this request */
		startstop = start_request(drive, rq);
		spin_lock_irq(&ide_lock);
		if (masked_irq != IDE_NO_IRQ && hwif->irq != masked_irq)
			enable_irq(hwif->irq);
		if (startstop == ide_stopped)
			hwgroup->busy = 0;
	}
}

/*
 * Passes the stuff to ide_do_request
 */
void do_ide_request(struct request_queue *q)
{
	ide_drive_t *drive = q->queuedata;

	ide_do_request(HWGROUP(drive), IDE_NO_IRQ);
}

/*
 * un-busy the hwgroup etc, and clear any pending DMA status. we want to
 * retry the current request in pio mode instead of risking tossing it
 * all away
 */
static ide_startstop_t ide_dma_timeout_retry(ide_drive_t *drive, int error)
{
	ide_hwif_t *hwif = HWIF(drive);
	struct request *rq;
	ide_startstop_t ret = ide_stopped;

	/*
	 * end current dma transaction
	 */

	if (error < 0) {
		printk(KERN_WARNING "%s: DMA timeout error\n", drive->name);
		(void)HWIF(drive)->ide_dma_end(drive);
		ret = ide_error(drive, "dma timeout error",
						hwif->INB(IDE_STATUS_REG));
	} else {
		printk(KERN_WARNING "%s: DMA timeout retry\n", drive->name);
		hwif->dma_timeout(drive);
	}

	/*
	 * disable dma for now, but remember that we did so because of
	 * a timeout -- we'll reenable after we finish this next request
	 * (or rather the first chunk of it) in pio.
	 */
	drive->retry_pio++;
	drive->state = DMA_PIO_RETRY;
	ide_dma_off_quietly(drive);

	/*
	 * un-busy drive etc (hwgroup->busy is cleared on return) and
	 * make sure request is sane
	 */
	rq = HWGROUP(drive)->rq;

	if (!rq)
		goto out;

	HWGROUP(drive)->rq = NULL;

	rq->errors = 0;

	if (!rq->bio)
		goto out;

	rq->sector = rq->bio->bi_sector;
	rq->current_nr_sectors = bio_iovec(rq->bio)->bv_len >> 9;
	rq->hard_cur_sectors = rq->current_nr_sectors;
	rq->buffer = bio_data(rq->bio);
out:
	return ret;
}

/**
 *	ide_timer_expiry	-	handle lack of an IDE interrupt
 *	@data: timer callback magic (hwgroup)
 *
 *	An IDE command has timed out before the expected drive return
 *	occurred. At this point we attempt to clean up the current
 *	mess. If the current handler includes an expiry handler then
 *	we invoke the expiry handler, and providing it is happy the
 *	work is done. If that fails we apply generic recovery rules
 *	invoking the handler and checking the drive DMA status. We
 *	have an excessively incestuous relationship with the DMA
 *	logic that wants cleaning up.
 */
 
void ide_timer_expiry (unsigned long data)
{
	ide_hwgroup_t	*hwgroup = (ide_hwgroup_t *) data;
	ide_handler_t	*handler;
	ide_expiry_t	*expiry;
	unsigned long	flags;
	unsigned long	wait = -1;

	spin_lock_irqsave(&ide_lock, flags);

	if (((handler = hwgroup->handler) == NULL) ||
	    (hwgroup->req_gen != hwgroup->req_gen_timer)) {
		/*
		 * Either a marginal timeout occurred
		 * (got the interrupt just as timer expired),
		 * or we were "sleeping" to give other devices a chance.
		 * Either way, we don't really want to complain about anything.
		 */
		if (hwgroup->sleeping) {
			hwgroup->sleeping = 0;
			hwgroup->busy = 0;
		}
	} else {
		ide_drive_t *drive = hwgroup->drive;
		if (!drive) {
			printk(KERN_ERR "ide_timer_expiry: hwgroup->drive was NULL\n");
			hwgroup->handler = NULL;
		} else {
			ide_hwif_t *hwif;
			ide_startstop_t startstop = ide_stopped;
			if (!hwgroup->busy) {
				hwgroup->busy = 1;	/* paranoia */
				printk(KERN_ERR "%s: ide_timer_expiry: hwgroup->busy was 0 ??\n", drive->name);
			}
			if ((expiry = hwgroup->expiry) != NULL) {
				/* continue */
				if ((wait = expiry(drive)) > 0) {
					/* reset timer */
					hwgroup->timer.expires  = jiffies + wait;
					hwgroup->req_gen_timer = hwgroup->req_gen;
					add_timer(&hwgroup->timer);
					spin_unlock_irqrestore(&ide_lock, flags);
					return;
				}
			}
			hwgroup->handler = NULL;
			/*
			 * We need to simulate a real interrupt when invoking
			 * the handler() function, which means we need to
			 * globally mask the specific IRQ:
			 */
			spin_unlock(&ide_lock);
			hwif  = HWIF(drive);
			/* disable_irq_nosync ?? */
			disable_irq(hwif->irq);
			/* local CPU only,
			 * as if we were handling an interrupt */
			local_irq_disable();
			if (hwgroup->polling) {
				startstop = handler(drive);
			} else if (drive_is_ready(drive)) {
				if (drive->waiting_for_dma)
					hwgroup->hwif->dma_lost_irq(drive);
				(void)ide_ack_intr(hwif);
				printk(KERN_WARNING "%s: lost interrupt\n", drive->name);
				startstop = handler(drive);
			} else {
				if (drive->waiting_for_dma) {
					startstop = ide_dma_timeout_retry(drive, wait);
				} else
					startstop =
					ide_error(drive, "irq timeout", hwif->INB(IDE_STATUS_REG));
			}
			drive->service_time = jiffies - drive->service_start;
			spin_lock_irq(&ide_lock);
			enable_irq(hwif->irq);
			if (startstop == ide_stopped)
				hwgroup->busy = 0;
		}
	}
	ide_do_request(hwgroup, IDE_NO_IRQ);
	spin_unlock_irqrestore(&ide_lock, flags);
}

/**
 *	unexpected_intr		-	handle an unexpected IDE interrupt
 *	@irq: interrupt line
 *	@hwgroup: hwgroup being processed
 *
 *	There's nothing really useful we can do with an unexpected interrupt,
 *	other than reading the status register (to clear it), and logging it.
 *	There should be no way that an irq can happen before we're ready for it,
 *	so we needn't worry much about losing an "important" interrupt here.
 *
 *	On laptops (and "green" PCs), an unexpected interrupt occurs whenever
 *	the drive enters "idle", "standby", or "sleep" mode, so if the status
 *	looks "good", we just ignore the interrupt completely.
 *
 *	This routine assumes __cli() is in effect when called.
 *
 *	If an unexpected interrupt happens on irq15 while we are handling irq14
 *	and if the two interfaces are "serialized" (CMD640), then it looks like
 *	we could screw up by interfering with a new request being set up for 
 *	irq15.
 *
 *	In reality, this is a non-issue.  The new command is not sent unless 
 *	the drive is ready to accept one, in which case we know the drive is
 *	not trying to interrupt us.  And ide_set_handler() is always invoked
 *	before completing the issuance of any new drive command, so we will not
 *	be accidentally invoked as a result of any valid command completion
 *	interrupt.
 *
 *	Note that we must walk the entire hwgroup here. We know which hwif
 *	is doing the current command, but we don't know which hwif burped
 *	mysteriously.
 */
 
static void unexpected_intr (int irq, ide_hwgroup_t *hwgroup)
{
	u8 stat;
	ide_hwif_t *hwif = hwgroup->hwif;

	/*
	 * handle the unexpected interrupt
	 */
	do {
		if (hwif->irq == irq) {
			stat = hwif->INB(hwif->io_ports[IDE_STATUS_OFFSET]);
			if (!OK_STAT(stat, READY_STAT, BAD_STAT)) {
				/* Try to not flood the console with msgs */
				static unsigned long last_msgtime, count;
				++count;
				if (time_after(jiffies, last_msgtime + HZ)) {
					last_msgtime = jiffies;
					printk(KERN_ERR "%s%s: unexpected interrupt, "
						"status=0x%02x, count=%ld\n",
						hwif->name,
						(hwif->next==hwgroup->hwif) ? "" : "(?)", stat, count);
				}
			}
		}
	} while ((hwif = hwif->next) != hwgroup->hwif);
}

/**
 *	ide_intr	-	default IDE interrupt handler
 *	@irq: interrupt number
 *	@dev_id: hwif group
 *	@regs: unused weirdness from the kernel irq layer
 *
 *	This is the default IRQ handler for the IDE layer. You should
 *	not need to override it. If you do be aware it is subtle in
 *	places
 *
 *	hwgroup->hwif is the interface in the group currently performing
 *	a command. hwgroup->drive is the drive and hwgroup->handler is
 *	the IRQ handler to call. As we issue a command the handlers
 *	step through multiple states, reassigning the handler to the
 *	next step in the process. Unlike a smart SCSI controller IDE
 *	expects the main processor to sequence the various transfer
 *	stages. We also manage a poll timer to catch up with most
 *	timeout situations. There are still a few where the handlers
 *	don't ever decide to give up.
 *
 *	The handler eventually returns ide_stopped to indicate the
 *	request completed. At this point we issue the next request
 *	on the hwgroup and the process begins again.
 */
 
irqreturn_t ide_intr (int irq, void *dev_id)
{
	unsigned long flags;
	ide_hwgroup_t *hwgroup = (ide_hwgroup_t *)dev_id;
	ide_hwif_t *hwif;
	ide_drive_t *drive;
	ide_handler_t *handler;
	ide_startstop_t startstop;

	spin_lock_irqsave(&ide_lock, flags);
	hwif = hwgroup->hwif;

	if (!ide_ack_intr(hwif)) {
		spin_unlock_irqrestore(&ide_lock, flags);
		return IRQ_NONE;
	}

	if ((handler = hwgroup->handler) == NULL || hwgroup->polling) {
		/*
		 * Not expecting an interrupt from this drive.
		 * That means this could be:
		 *	(1) an interrupt from another PCI device
		 *	sharing the same PCI INT# as us.
		 * or	(2) a drive just entered sleep or standby mode,
		 *	and is interrupting to let us know.
		 * or	(3) a spurious interrupt of unknown origin.
		 *
		 * For PCI, we cannot tell the difference,
		 * so in that case we just ignore it and hope it goes away.
		 *
		 * FIXME: unexpected_intr should be hwif-> then we can
		 * remove all the ifdef PCI crap
		 */
#ifdef CONFIG_BLK_DEV_IDEPCI
		if (hwif->pci_dev && !hwif->pci_dev->vendor)
#endif	/* CONFIG_BLK_DEV_IDEPCI */
		{
			/*
			 * Probably not a shared PCI interrupt,
			 * so we can safely try to do something about it:
			 */
			unexpected_intr(irq, hwgroup);
#ifdef CONFIG_BLK_DEV_IDEPCI
		} else {
			/*
			 * Whack the status register, just in case
			 * we have a leftover pending IRQ.
			 */
			(void) hwif->INB(hwif->io_ports[IDE_STATUS_OFFSET]);
#endif /* CONFIG_BLK_DEV_IDEPCI */
		}
		spin_unlock_irqrestore(&ide_lock, flags);
		return IRQ_NONE;
	}
	drive = hwgroup->drive;
	if (!drive) {
		/*
		 * This should NEVER happen, and there isn't much
		 * we could do about it here.
		 *
		 * [Note - this can occur if the drive is hot unplugged]
		 */
		spin_unlock_irqrestore(&ide_lock, flags);
		return IRQ_HANDLED;
	}
	if (!drive_is_ready(drive)) {
		/*
		 * This happens regularly when we share a PCI IRQ with
		 * another device.  Unfortunately, it can also happen
		 * with some buggy drives that trigger the IRQ before
		 * their status register is up to date.  Hopefully we have
		 * enough advance overhead that the latter isn't a problem.
		 */
		spin_unlock_irqrestore(&ide_lock, flags);
		return IRQ_NONE;
	}
	if (!hwgroup->busy) {
		hwgroup->busy = 1;	/* paranoia */
		printk(KERN_ERR "%s: ide_intr: hwgroup->busy was 0 ??\n", drive->name);
	}
	hwgroup->handler = NULL;
	hwgroup->req_gen++;
	del_timer(&hwgroup->timer);
	spin_unlock(&ide_lock);

	/* Some controllers might set DMA INTR no matter DMA or PIO;
	 * bmdma status might need to be cleared even for
	 * PIO interrupts to prevent spurious/lost irq.
	 */
	if (hwif->ide_dma_clear_irq && !(drive->waiting_for_dma))
		/* ide_dma_end() needs bmdma status for error checking.
		 * So, skip clearing bmdma status here and leave it
		 * to ide_dma_end() if this is dma interrupt.
		 */
		hwif->ide_dma_clear_irq(drive);

	if (drive->unmask)
		local_irq_enable_in_hardirq();
	/* service this interrupt, may set handler for next interrupt */
	startstop = handler(drive);
	spin_lock_irq(&ide_lock);

	/*
	 * Note that handler() may have set things up for another
	 * interrupt to occur soon, but it cannot happen until
	 * we exit from this routine, because it will be the
	 * same irq as is currently being serviced here, and Linux
	 * won't allow another of the same (on any CPU) until we return.
	 */
	drive->service_time = jiffies - drive->service_start;
	if (startstop == ide_stopped) {
		if (hwgroup->handler == NULL) {	/* paranoia */
			hwgroup->busy = 0;
			ide_do_request(hwgroup, hwif->irq);
		} else {
			printk(KERN_ERR "%s: ide_intr: huh? expected NULL handler "
				"on exit\n", drive->name);
		}
	}
	spin_unlock_irqrestore(&ide_lock, flags);
	return IRQ_HANDLED;
}

/**
 *	ide_init_drive_cmd	-	initialize a drive command request
 *	@rq: request object
 *
 *	Initialize a request before we fill it in and send it down to
 *	ide_do_drive_cmd. Commands must be set up by this function. Right
 *	now it doesn't do a lot, but if that changes abusers will have a
 *	nasty surprise.
 */

void ide_init_drive_cmd (struct request *rq)
{
	memset(rq, 0, sizeof(*rq));
	rq->cmd_type = REQ_TYPE_ATA_CMD;
	rq->ref_count = 1;
}

EXPORT_SYMBOL(ide_init_drive_cmd);

/**
 *	ide_do_drive_cmd	-	issue IDE special command
 *	@drive: device to issue command
 *	@rq: request to issue
 *	@action: action for processing
 *
 *	This function issues a special IDE device request
 *	onto the request queue.
 *
 *	If action is ide_wait, then the rq is queued at the end of the
 *	request queue, and the function sleeps until it has been processed.
 *	This is for use when invoked from an ioctl handler.
 *
 *	If action is ide_preempt, then the rq is queued at the head of
 *	the request queue, displacing the currently-being-processed
 *	request and this function returns immediately without waiting
 *	for the new rq to be completed.  This is VERY DANGEROUS, and is
 *	intended for careful use by the ATAPI tape/cdrom driver code.
 *
 *	If action is ide_end, then the rq is queued at the end of the
 *	request queue, and the function returns immediately without waiting
 *	for the new rq to be completed. This is again intended for careful
 *	use by the ATAPI tape/cdrom driver code.
 */
 
int ide_do_drive_cmd (ide_drive_t *drive, struct request *rq, ide_action_t action)
{
	unsigned long flags;
	ide_hwgroup_t *hwgroup = HWGROUP(drive);
	DECLARE_COMPLETION_ONSTACK(wait);
	int where = ELEVATOR_INSERT_BACK, err;
	int must_wait = (action == ide_wait || action == ide_head_wait);

	rq->errors = 0;

	/*
	 * we need to hold an extra reference to request for safe inspection
	 * after completion
	 */
	if (must_wait) {
		rq->ref_count++;
		rq->end_io_data = &wait;
		rq->end_io = blk_end_sync_rq;
	}

	spin_lock_irqsave(&ide_lock, flags);
	if (action == ide_preempt)
		hwgroup->rq = NULL;
	if (action == ide_preempt || action == ide_head_wait) {
		where = ELEVATOR_INSERT_FRONT;
		rq->cmd_flags |= REQ_PREEMPT;
	}
	__elv_add_request(drive->queue, rq, where, 0);
	ide_do_request(hwgroup, IDE_NO_IRQ);
	spin_unlock_irqrestore(&ide_lock, flags);

	err = 0;
	if (must_wait) {
		wait_for_completion(&wait);
		if (rq->errors)
			err = -EIO;

		blk_put_request(rq);
	}

	return err;
}

EXPORT_SYMBOL(ide_do_drive_cmd);

void ide_pktcmd_tf_load(ide_drive_t *drive, u32 tf_flags, u16 bcount, u8 dma)
{
	ide_task_t task;

	memset(&task, 0, sizeof(task));
	task.tf_flags = IDE_TFLAG_OUT_LBAH | IDE_TFLAG_OUT_LBAM |
			IDE_TFLAG_OUT_FEATURE | tf_flags;
	task.tf.feature = dma;		/* Use PIO/DMA */
	task.tf.lbam    = bcount & 0xff;
	task.tf.lbah    = (bcount >> 8) & 0xff;

	ide_tf_load(drive, &task);
}

EXPORT_SYMBOL_GPL(ide_pktcmd_tf_load);