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
|
/* $OpenBSD: ccd.c,v 1.68 2006/11/29 15:03:42 krw Exp $ */
/* $NetBSD: ccd.c,v 1.33 1996/05/05 04:21:14 thorpej Exp $ */
/*-
* Copyright (c) 1996 The NetBSD Foundation, Inc.
* Copyright (c) 1997 Niklas Hallqvist.
* Copyright (c) 2005 Michael Shalayeff.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Jason R. Thorpe.
* Niklas Hallqvist redid the buffer policy for better performance.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the NetBSD
* Foundation, Inc. and its contributors.
* 4. Neither the name of The NetBSD Foundation nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
* ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
* TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE
* LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGE.
*/
/*
* Copyright (c) 1988 University of Utah.
* Copyright (c) 1990, 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* the Systems Programming Group of the University of Utah Computer
* Science Department.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* from: Utah $Hdr: cd.c 1.6 90/11/28$
*
* @(#)cd.c 8.2 (Berkeley) 11/16/93
*/
/*
* "Concatenated" disk driver.
*
* Dynamic configuration and disklabel support by:
* Jason R. Thorpe <thorpej@nas.nasa.gov>
* Numerical Aerodynamic Simulation Facility
* Mail Stop 258-6
* NASA Ames Research Center
* Moffett Field, CA 94035
*
* Mirroring support based on code written by Satoshi Asami
* and Nisha Talagala.
*
* Buffer scatter/gather policy by Niklas Hallqvist.
*/
/* #define CCDDEBUG */
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/proc.h>
#include <sys/errno.h>
#include <sys/buf.h>
#include <sys/malloc.h>
#include <sys/pool.h>
#include <sys/namei.h>
#include <sys/stat.h>
#include <sys/ioctl.h>
#include <sys/disklabel.h>
#include <sys/device.h>
#include <sys/disk.h>
#include <sys/syslog.h>
#include <sys/fcntl.h>
#include <sys/vnode.h>
#include <sys/conf.h>
#include <dev/ccdvar.h>
#ifdef __GNUC__
#define INLINE static __inline
#else
#define INLINE
#endif
/*
* Overridable value telling how many kvm spaces of MAXBSIZE we need for
* component I/O operations.
*/
#ifndef CCD_CLUSTERS
#define CCD_CLUSTERS 16
#endif
#ifdef CCDDEBUG
#define CCD_DCALL(m,c) if (ccddebug & (m)) c
#define CCD_DPRINTF(m,a) CCD_DCALL(m, printf a)
#define CCDB_FOLLOW 0x01
#define CCDB_INIT 0x02
#define CCDB_IO 0x04
#define CCDB_LABEL 0x08
#define CCDB_VNODE 0x10
int ccddebug = 0x00;
#else
#define CCD_DCALL(m,c) /* m, c */
#define CCD_DPRINTF(m,a) /* m, a */
#endif
#define ccdunit(x) DISKUNIT(x)
#define CCDLABELDEV(dev) \
(MAKEDISKDEV(major((dev)), ccdunit((dev)), RAW_PART))
struct ccdbuf {
struct buf cb_buf; /* new I/O buf */
struct buf *cb_obp; /* ptr. to original I/O buf */
struct ccd_softc*cb_sc; /* point back to the device */
struct ccdbuf *cb_dep; /* mutual ptrs for mirror part */
int cb_comp; /* target component */
int cb_flags; /* misc. flags */
#define CBF_MIRROR 0x01 /* we're for a mirror component */
#define CBF_OLD 0x02 /* use old I/O protocol */
#define CBF_DONE 0x04 /* this buffer is done */
int cb_sgcnt; /* scatter/gather segment count */
#define CCD_SGMAX (MAXBSIZE >> PAGE_SHIFT)
struct ccdseg {
caddr_t cs_sgaddr; /* scatter/gather segment addresses */
long cs_sglen; /* scatter/gather segment lengths */
} cb_sg[1];
};
/* called by main() at boot time */
void ccdattach(int);
/* called by biodone() at interrupt time */
void ccdiodone(struct buf *);
int ccdsize(dev_t);
void ccdstart(struct ccd_softc *, struct buf *);
void ccdinterleave(struct ccd_softc *);
void ccdintr(struct ccd_softc *, struct buf *);
int ccdinit(struct ccddevice *, char **, struct proc *);
int ccdlookup(char *, struct proc *p, struct vnode **);
long ccdbuffer(struct ccd_softc *, struct buf *, daddr_t, caddr_t,
long, struct ccdbuf **, int);
void ccdgetdisklabel(dev_t, struct ccd_softc *, struct disklabel *,
struct cpu_disklabel *, int);
int ccdlock(struct ccd_softc *);
void ccdunlock(struct ccd_softc *);
INLINE struct ccdbuf *getccdbuf(void);
INLINE void putccdbuf(struct ccdbuf *);
#ifdef CCDDEBUG
void printiinfo(struct ccdiinfo *);
#endif
/* Non-private for the benefit of libkvm. */
struct ccd_softc *ccd_softc;
struct ccddevice *ccddevs;
int numccd = 0;
int ccdbufsizeof;
/*
* A separate map so that locking on kernel_map won't happen in interrupts
* (XXX due to fragmentation this might fail easy and panic the kernel)
*/
struct vm_map *ccdmap;
/*
* Set when a process need some kvm.
* XXX should we fallback to old I/O policy instead when out of ccd kvm?
*/
int ccd_need_kvm = 0;
/*
* struct ccdbuf allocator
*/
struct pool ccdbufpl;
/*
* Manage the ccd buffer structures.
*/
INLINE struct ccdbuf *
getccdbuf(void)
{
struct ccdbuf *cbp;
if ((cbp = pool_get(&ccdbufpl, PR_WAITOK)))
bzero(cbp, ccdbufsizeof);
return (cbp);
}
INLINE void
putccdbuf(struct ccdbuf *cbp)
{
pool_put(&ccdbufpl, cbp);
}
/*
* Called by main() during pseudo-device attachment. All we need
* to do is allocate enough space for devices to be configured later.
*/
void
ccdattach(int num)
{
if (num <= 0) {
#ifdef DIAGNOSTIC
panic("ccdattach: count <= 0");
#endif
return;
}
ccd_softc = (struct ccd_softc *)malloc(num * sizeof(struct ccd_softc),
M_DEVBUF, M_NOWAIT);
ccddevs = (struct ccddevice *)malloc(num * sizeof(struct ccddevice),
M_DEVBUF, M_NOWAIT);
if ((ccd_softc == NULL) || (ccddevs == NULL)) {
printf("WARNING: no memory for concatenated disks\n");
if (ccd_softc != NULL)
free(ccd_softc, M_DEVBUF);
if (ccddevs != NULL)
free(ccddevs, M_DEVBUF);
return;
}
numccd = num;
bzero(ccd_softc, num * sizeof(struct ccd_softc));
bzero(ccddevs, num * sizeof(struct ccddevice));
ccdbufsizeof = sizeof(struct ccdbuf) +
(CCD_SGMAX - 1) * sizeof(struct ccdseg);
pool_init(&ccdbufpl, ccdbufsizeof, 0, 0, 0, "ccdbufpl", NULL);
pool_setlowat(&ccdbufpl, 16);
pool_sethiwat(&ccdbufpl, 1024);
}
int
ccdinit(struct ccddevice *ccd, char **cpaths, struct proc *p)
{
struct ccd_softc *cs = &ccd_softc[ccd->ccd_unit];
struct ccdcinfo *ci = NULL;
size_t size;
int ix, rpm;
struct vnode *vp;
struct vattr va;
size_t minsize;
int maxsecsize;
struct partinfo dpart;
struct ccdgeom *ccg = &cs->sc_geom;
char tmppath[MAXPATHLEN];
int error;
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT, ("ccdinit: unit %d cflags %b\n",
ccd->ccd_unit, ccd->ccd_flags, CCDF_BITS));
cs->sc_size = 0;
cs->sc_ileave = ccd->ccd_interleave;
cs->sc_nccdisks = ccd->ccd_ndev;
if (snprintf(cs->sc_xname, sizeof(cs->sc_xname), "ccd%d",
ccd->ccd_unit) >= sizeof(cs->sc_xname)) {
printf("ccdinit: device name too long.\n");
return(ENXIO);
}
/* Allocate space for the component info. */
cs->sc_cinfo = malloc(cs->sc_nccdisks * sizeof(struct ccdcinfo),
M_DEVBUF, M_WAITOK);
bzero(cs->sc_cinfo, cs->sc_nccdisks * sizeof(struct ccdcinfo));
/*
* Verify that each component piece exists and record
* relevant information about it.
*/
maxsecsize = 0;
minsize = 0;
rpm = 0;
for (ix = 0; ix < cs->sc_nccdisks; ix++) {
vp = ccd->ccd_vpp[ix];
ci = &cs->sc_cinfo[ix];
ci->ci_vp = vp;
/*
* Copy in the pathname of the component.
*/
bzero(tmppath, sizeof(tmppath)); /* sanity */
error = copyinstr(cpaths[ix], tmppath,
MAXPATHLEN, &ci->ci_pathlen);
if (error) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("%s: can't copy path, error = %d\n",
cs->sc_xname, error));
free(cs->sc_cinfo, M_DEVBUF);
return (error);
}
ci->ci_path = malloc(ci->ci_pathlen, M_DEVBUF, M_WAITOK);
bcopy(tmppath, ci->ci_path, ci->ci_pathlen);
/*
* XXX: Cache the component's dev_t.
*/
if ((error = VOP_GETATTR(vp, &va, p->p_ucred, p)) != 0) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("%s: %s: getattr failed error = %d\n",
cs->sc_xname, ci->ci_path, error));
free(ci->ci_path, M_DEVBUF);
free(cs->sc_cinfo, M_DEVBUF);
return (error);
}
ci->ci_dev = va.va_rdev;
/*
* Get partition information for the component.
*/
error = VOP_IOCTL(vp, DIOCGPART, (caddr_t)&dpart,
FREAD, p->p_ucred, p);
if (error) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("%s: %s: ioctl failed, error = %d\n",
cs->sc_xname, ci->ci_path, error));
free(ci->ci_path, M_DEVBUF);
free(cs->sc_cinfo, M_DEVBUF);
return (error);
}
if (dpart.part->p_fstype == FS_CCD ||
dpart.part->p_fstype == FS_BSDFFS) {
maxsecsize =
((dpart.disklab->d_secsize > maxsecsize) ?
dpart.disklab->d_secsize : maxsecsize);
size = dpart.part->p_size;
} else {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("%s: %s: incorrect partition type\n",
cs->sc_xname, ci->ci_path));
free(ci->ci_path, M_DEVBUF);
free(cs->sc_cinfo, M_DEVBUF);
return (EFTYPE);
}
/*
* Calculate the size, truncating to an interleave
* boundary if necessary.
*/
if (cs->sc_ileave > 1)
size -= size % cs->sc_ileave;
if (size == 0) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("%s: %s: size == 0\n", cs->sc_xname, ci->ci_path));
free(ci->ci_path, M_DEVBUF);
free(cs->sc_cinfo, M_DEVBUF);
return (ENODEV);
}
if (minsize == 0 || size < minsize)
minsize = size;
ci->ci_size = size;
cs->sc_size += size;
rpm += dpart.disklab->d_rpm;
}
ccg->ccg_rpm = rpm / cs->sc_nccdisks;
/*
* Don't allow the interleave to be smaller than
* the biggest component sector.
*/
if ((cs->sc_ileave > 0) &&
(cs->sc_ileave < (maxsecsize / DEV_BSIZE))) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("%s: interleave must be at least %d\n",
cs->sc_xname, (maxsecsize / DEV_BSIZE)));
free(ci->ci_path, M_DEVBUF);
free(cs->sc_cinfo, M_DEVBUF);
return (EINVAL);
}
/*
* Mirroring support requires uniform interleave and
* and even number of components.
*/
if (ccd->ccd_flags & CCDF_MIRROR) {
ccd->ccd_flags |= CCDF_UNIFORM;
if (cs->sc_ileave == 0) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("%s: mirroring requires interleave\n",
cs->sc_xname));
free(ci->ci_path, M_DEVBUF);
free(cs->sc_cinfo, M_DEVBUF);
return (EINVAL);
}
if (cs->sc_nccdisks % 2) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("%s: mirroring requires even # of components\n",
cs->sc_xname));
free(ci->ci_path, M_DEVBUF);
free(cs->sc_cinfo, M_DEVBUF);
return (EINVAL);
}
}
/*
* If uniform interleave is desired set all sizes to that of
* the smallest component.
*/
ccg->ccg_ntracks = cs->sc_nccunits = cs->sc_nccdisks;
if (ccd->ccd_flags & CCDF_UNIFORM) {
for (ci = cs->sc_cinfo;
ci < &cs->sc_cinfo[cs->sc_nccdisks]; ci++)
ci->ci_size = minsize;
if (ccd->ccd_flags & CCDF_MIRROR)
cs->sc_nccunits = ccg->ccg_ntracks /= 2;
cs->sc_size = ccg->ccg_ntracks * minsize;
}
cs->sc_cflags = ccd->ccd_flags; /* So we can find out later... */
/*
* Construct the interleave table.
*/
ccdinterleave(cs);
/*
* Create pseudo-geometry based on 1MB cylinders. It's
* pretty close.
*/
ccg->ccg_secsize = DEV_BSIZE;
ccg->ccg_nsectors = cs->sc_ileave? cs->sc_ileave :
1024 * (1024 / ccg->ccg_secsize);
ccg->ccg_ncylinders = cs->sc_size / ccg->ccg_ntracks /
ccg->ccg_nsectors;
cs->sc_flags |= CCDF_INITED;
return (0);
}
void
ccdinterleave(struct ccd_softc *cs)
{
struct ccdcinfo *ci, *smallci;
struct ccdiinfo *ii;
daddr_t bn, lbn;
int ix;
u_long size;
CCD_DPRINTF(CCDB_INIT,
("ccdinterleave(%p): ileave %d\n", cs, cs->sc_ileave));
/*
* Allocate an interleave table.
* Chances are this is too big, but we don't care.
*/
size = (cs->sc_nccdisks + 1) * sizeof(struct ccdiinfo);
cs->sc_itable = (struct ccdiinfo *)malloc(size, M_DEVBUF, M_WAITOK);
bzero((caddr_t)cs->sc_itable, size);
/*
* Trivial case: no interleave (actually interleave of disk size).
* Each table entry represents a single component in its entirety.
*/
if (cs->sc_ileave == 0) {
bn = 0;
ii = cs->sc_itable;
for (ix = 0; ix < cs->sc_nccdisks; ix++) {
/* Allocate space for ii_index. */
ii->ii_index = malloc(sizeof(int), M_DEVBUF, M_WAITOK);
ii->ii_ndisk = 1;
ii->ii_startblk = bn;
ii->ii_startoff = 0;
ii->ii_index[0] = ix;
bn += cs->sc_cinfo[ix].ci_size;
ii++;
}
ii->ii_ndisk = 0;
CCD_DCALL(CCDB_INIT, printiinfo(cs->sc_itable));
return;
}
/*
* The following isn't fast or pretty; it doesn't have to be.
*/
size = 0;
bn = lbn = 0;
for (ii = cs->sc_itable; ; ii++) {
/* Allocate space for ii_index. */
ii->ii_index = malloc((sizeof(int) * cs->sc_nccdisks),
M_DEVBUF, M_WAITOK);
/*
* Locate the smallest of the remaining components
*/
smallci = NULL;
for (ci = cs->sc_cinfo;
ci < &cs->sc_cinfo[cs->sc_nccdisks]; ci++)
if (ci->ci_size > size &&
(smallci == NULL ||
ci->ci_size < smallci->ci_size))
smallci = ci;
/*
* Nobody left, all done
*/
if (smallci == NULL) {
ii->ii_ndisk = 0;
break;
}
/*
* Record starting logical block and component offset
*/
ii->ii_startblk = bn / cs->sc_ileave;
ii->ii_startoff = lbn;
/*
* Determine how many disks take part in this interleave
* and record their indices.
*/
ix = 0;
for (ci = cs->sc_cinfo;
ci < &cs->sc_cinfo[cs->sc_nccunits]; ci++)
if (ci->ci_size >= smallci->ci_size)
ii->ii_index[ix++] = ci - cs->sc_cinfo;
ii->ii_ndisk = ix;
bn += ix * (smallci->ci_size - size);
lbn = smallci->ci_size / cs->sc_ileave;
size = smallci->ci_size;
}
CCD_DCALL(CCDB_INIT, printiinfo(cs->sc_itable));
}
/* ARGSUSED */
int
ccdopen(dev_t dev, int flags, int fmt, struct proc *p)
{
int unit = ccdunit(dev);
struct ccd_softc *cs;
struct disklabel *lp;
int error = 0, part, pmask;
CCD_DPRINTF(CCDB_FOLLOW, ("ccdopen(%x, %x)\n", dev, flags));
if (unit >= numccd)
return (ENXIO);
cs = &ccd_softc[unit];
if ((error = ccdlock(cs)) != 0)
return (error);
lp = cs->sc_dkdev.dk_label;
part = DISKPART(dev);
pmask = (1 << part);
/*
* If we're initialized, check to see if there are any other
* open partitions. If not, then it's safe to update
* the in-core disklabel.
*/
if ((cs->sc_flags & CCDF_INITED) && (cs->sc_dkdev.dk_openmask == 0))
ccdgetdisklabel(dev, cs, lp, cs->sc_dkdev.dk_cpulabel, 0);
/* Check that the partition exists. */
if (part != RAW_PART) {
if (((cs->sc_flags & CCDF_INITED) == 0) ||
((part >= lp->d_npartitions) ||
(lp->d_partitions[part].p_fstype == FS_UNUSED))) {
error = ENXIO;
goto done;
}
}
/* Prevent our unit from being unconfigured while open. */
switch (fmt) {
case S_IFCHR:
cs->sc_dkdev.dk_copenmask |= pmask;
break;
case S_IFBLK:
cs->sc_dkdev.dk_bopenmask |= pmask;
break;
}
cs->sc_dkdev.dk_openmask =
cs->sc_dkdev.dk_copenmask | cs->sc_dkdev.dk_bopenmask;
done:
ccdunlock(cs);
return (error);
}
/* ARGSUSED */
int
ccdclose(dev_t dev, int flags, int fmt, struct proc *p)
{
int unit = ccdunit(dev);
struct ccd_softc *cs;
int error = 0, part;
CCD_DPRINTF(CCDB_FOLLOW, ("ccdclose(%x, %x)\n", dev, flags));
if (unit >= numccd)
return (ENXIO);
cs = &ccd_softc[unit];
if ((error = ccdlock(cs)) != 0)
return (error);
part = DISKPART(dev);
/* ...that much closer to allowing unconfiguration... */
switch (fmt) {
case S_IFCHR:
cs->sc_dkdev.dk_copenmask &= ~(1 << part);
break;
case S_IFBLK:
cs->sc_dkdev.dk_bopenmask &= ~(1 << part);
break;
}
cs->sc_dkdev.dk_openmask =
cs->sc_dkdev.dk_copenmask | cs->sc_dkdev.dk_bopenmask;
ccdunlock(cs);
return (0);
}
void
ccdstrategy(struct buf *bp)
{
int unit = ccdunit(bp->b_dev);
struct ccd_softc *cs = &ccd_softc[unit];
int s;
int wlabel;
struct disklabel *lp;
CCD_DPRINTF(CCDB_FOLLOW, ("ccdstrategy(%p): unit %d\n", bp, unit));
if ((cs->sc_flags & CCDF_INITED) == 0) {
bp->b_error = ENXIO;
bp->b_resid = bp->b_bcount;
bp->b_flags |= B_ERROR;
goto done;
}
/* If it's a nil transfer, wake up the top half now. */
if (bp->b_bcount == 0)
goto done;
lp = cs->sc_dkdev.dk_label;
/*
* Do bounds checking and adjust transfer. If there's an
* error, the bounds check will flag that for us.
*/
wlabel = cs->sc_flags & (CCDF_WLABEL|CCDF_LABELLING);
if (DISKPART(bp->b_dev) != RAW_PART &&
bounds_check_with_label(bp, lp, cs->sc_dkdev.dk_cpulabel,
wlabel) <= 0)
goto done;
bp->b_resid = bp->b_bcount;
/*
* "Start" the unit.
*/
s = splbio();
ccdstart(cs, bp);
splx(s);
return;
done:
s = splbio();
biodone(bp);
splx(s);
}
void
ccdstart(struct ccd_softc *cs, struct buf *bp)
{
long bcount, rcount;
struct ccdbuf **cbpp, *cbp;
caddr_t addr;
daddr_t bn;
struct partition *pp;
int i, old_io = cs->sc_cflags & CCDF_OLD;
CCD_DPRINTF(CCDB_FOLLOW, ("ccdstart(%p, %p, %s)\n", cs, bp,
bp->b_flags & B_READ? "read" : "write"));
/* Instrumentation. */
disk_busy(&cs->sc_dkdev);
/*
* Translate the partition-relative block number to an absolute.
*/
bn = bp->b_blkno;
if (DISKPART(bp->b_dev) != RAW_PART) {
pp = &cs->sc_dkdev.dk_label->d_partitions[DISKPART(bp->b_dev)];
bn += pp->p_offset;
}
/*
* Allocate component buffers
*/
cbpp = malloc(2 * cs->sc_nccdisks * sizeof(struct ccdbuf *), M_DEVBUF,
M_WAITOK);
bzero(cbpp, 2 * cs->sc_nccdisks * sizeof(struct ccdbuf *));
addr = bp->b_data;
old_io = old_io || ((vaddr_t)addr & PAGE_MASK);
for (bcount = bp->b_bcount; bcount > 0; bcount -= rcount) {
rcount = ccdbuffer(cs, bp, bn, addr, bcount, cbpp, old_io);
/*
* This is the old, slower, but less restrictive, mode of
* operation. It allows interleaves which are not multiples
* of PAGE_SIZE and mirroring.
*/
if (old_io) {
if ((cbpp[0]->cb_buf.b_flags & B_READ) == 0)
cbpp[0]->cb_buf.b_vp->v_numoutput++;
VOP_STRATEGY(&cbpp[0]->cb_buf);
if ((cs->sc_cflags & CCDF_MIRROR) &&
((cbpp[0]->cb_buf.b_flags & B_READ) == 0)) {
cbpp[1]->cb_buf.b_vp->v_numoutput++;
VOP_STRATEGY(&cbpp[1]->cb_buf);
}
}
bn += btodb(rcount);
addr += rcount;
}
/* The new leaner mode of operation */
if (!old_io)
/*
* Fire off the requests
*/
for (i = 0; i < 2*cs->sc_nccdisks; i++) {
cbp = cbpp[i];
if (cbp) {
if ((cbp->cb_buf.b_flags & B_READ) == 0)
cbp->cb_buf.b_vp->v_numoutput++;
VOP_STRATEGY(&cbp->cb_buf);
}
}
free(cbpp, M_DEVBUF);
}
/*
* Build a component buffer header.
*/
long
ccdbuffer(struct ccd_softc *cs, struct buf *bp, daddr_t bn, caddr_t addr,
long bcount, struct ccdbuf **cbpp, int old_io)
{
struct ccdcinfo *ci, *ci2 = NULL;
struct ccdbuf *cbp;
daddr_t cbn, cboff, sblk;
int ccdisk, ccdisk2, off;
long old_bcount, cnt;
struct ccdiinfo *ii;
struct buf *nbp;
CCD_DPRINTF(CCDB_IO, ("ccdbuffer(%p, %p, %d, %p, %ld, %p)\n",
cs, bp, bn, addr, bcount, cbpp));
/*
* Determine which component bn falls in.
*/
cbn = bn;
cboff = 0;
if (cs->sc_ileave == 0) {
/*
* Serially concatenated
*/
sblk = 0;
for (ccdisk = 0, ci = &cs->sc_cinfo[ccdisk];
cbn >= sblk + ci->ci_size;
ccdisk++, ci = &cs->sc_cinfo[ccdisk])
sblk += ci->ci_size;
cbn -= sblk;
} else {
/*
* Interleaved
*/
cboff = cbn % cs->sc_ileave;
cbn /= cs->sc_ileave;
for (ii = cs->sc_itable; ii->ii_ndisk; ii++)
if (ii->ii_startblk > cbn)
break;
ii--;
off = cbn - ii->ii_startblk;
if (ii->ii_ndisk == 1) {
ccdisk = ii->ii_index[0];
cbn = ii->ii_startoff + off;
} else {
ccdisk = ii->ii_index[off % ii->ii_ndisk];
cbn = ii->ii_startoff + off / ii->ii_ndisk;
}
if (cs->sc_cflags & CCDF_MIRROR) {
/* Mirrored data */
ccdisk2 = ccdisk + ii->ii_ndisk;
ci2 = &cs->sc_cinfo[ccdisk2];
/* spread the read over both parts */
if (bp->b_flags & B_READ &&
bcount > bp->b_bcount / 2 &&
(!(ci2->ci_flags & CCIF_FAILED) ||
ci->ci_flags & CCIF_FAILED))
ccdisk = ccdisk2;
}
cbn *= cs->sc_ileave;
ci = &cs->sc_cinfo[ccdisk];
CCD_DPRINTF(CCDB_IO, ("ccdisk %d cbn %d ci %p ci2 %p\n",
ccdisk, cbn, ci, ci2));
}
/* Limit the operation at next component border */
if (cs->sc_ileave == 0)
cnt = dbtob(ci->ci_size - cbn);
else
cnt = dbtob(cs->sc_ileave - cboff);
if (cnt < bcount)
bcount = cnt;
if (old_io || cbpp[ccdisk] == NULL) {
/*
* Setup new component buffer.
*/
cbp = cbpp[old_io ? 0 : ccdisk] = getccdbuf();
cbp->cb_flags = old_io ? CBF_OLD : 0;
nbp = &cbp->cb_buf;
nbp->b_flags = bp->b_flags | B_CALL;
nbp->b_iodone = ccdiodone;
nbp->b_proc = bp->b_proc;
nbp->b_dev = ci->ci_dev; /* XXX */
nbp->b_blkno = cbn + cboff;
nbp->b_vp = ci->ci_vp;
nbp->b_bcount = bcount;
LIST_INIT(&nbp->b_dep);
/*
* context for ccdiodone
*/
cbp->cb_obp = bp;
cbp->cb_sc = cs;
cbp->cb_comp = ccdisk;
/* Deal with the different algorithms */
if (old_io)
nbp->b_data = addr;
else {
do {
nbp->b_data = (caddr_t) uvm_km_valloc(ccdmap,
bp->b_bcount);
/*
* XXX Instead of sleeping, we might revert
* XXX to old I/O policy for this buffer set.
*/
if (nbp->b_data == NULL) {
ccd_need_kvm++;
tsleep(ccdmap, PRIBIO, "ccdbuffer", 0);
}
} while (nbp->b_data == NULL);
cbp->cb_sgcnt = 0;
old_bcount = 0;
}
/*
* Mirrors have an additional write operation that is nearly
* identical to the first.
*/
if ((cs->sc_cflags & CCDF_MIRROR) &&
!(ci2->ci_flags & CCIF_FAILED) &&
((cbp->cb_buf.b_flags & B_READ) == 0)) {
struct ccdbuf *cbp2;
cbpp[old_io? 1 : ccdisk2] = cbp2 = getccdbuf();
*cbp2 = *cbp;
cbp2->cb_flags = CBF_MIRROR | (old_io ? CBF_OLD : 0);
cbp2->cb_buf.b_dev = ci2->ci_dev; /* XXX */
cbp2->cb_buf.b_vp = ci2->ci_vp;
LIST_INIT(&cbp2->cb_buf.b_dep);
cbp2->cb_comp = ccdisk2;
cbp2->cb_dep = cbp;
cbp->cb_dep = cbp2;
}
} else {
/*
* Continue on an already started component buffer
*/
cbp = cbpp[ccdisk];
nbp = &cbp->cb_buf;
/*
* Map the new pages at the end of the buffer.
*/
old_bcount = nbp->b_bcount;
nbp->b_bcount += bcount;
}
if (!old_io) {
CCD_DPRINTF(CCDB_IO, ("ccdbuffer: sg %d (%p/%x) off %x\n",
cbp->cb_sgcnt, addr, bcount, old_bcount));
pagemove(addr, nbp->b_data + old_bcount, round_page(bcount));
nbp->b_bufsize += round_page(bcount);
cbp->cb_sg[cbp->cb_sgcnt].cs_sgaddr = addr;
cbp->cb_sg[cbp->cb_sgcnt].cs_sglen = bcount;
cbp->cb_sgcnt++;
}
CCD_DPRINTF(CCDB_IO, (" dev %x(u%d): cbp %p bn %d addr %p bcnt %ld\n",
ci->ci_dev, ci-cs->sc_cinfo, cbp, bp->b_blkno,
bp->b_data, bp->b_bcount));
return (bcount);
}
void
ccdintr(struct ccd_softc *cs, struct buf *bp)
{
splassert(IPL_BIO);
CCD_DPRINTF(CCDB_FOLLOW, ("ccdintr(%p, %p)\n", cs, bp));
/*
* Request is done for better or worse, wakeup the top half.
*/
if (bp->b_flags & B_ERROR)
bp->b_resid = bp->b_bcount;
disk_unbusy(&cs->sc_dkdev, (bp->b_bcount - bp->b_resid),
(bp->b_flags & B_READ));
biodone(bp);
}
/*
* Called at interrupt time.
* Mark the component as done and if all components are done,
* take a ccd interrupt.
*/
void
ccdiodone(struct buf *vbp)
{
struct ccdbuf *cbp = (struct ccdbuf *)vbp;
struct buf *bp = cbp->cb_obp;
struct ccd_softc *cs = cbp->cb_sc;
int old_io = cbp->cb_flags & CBF_OLD;
int i;
long count = bp->b_bcount, off;
char *comptype;
splassert(IPL_BIO);
CCD_DPRINTF(CCDB_FOLLOW, ("ccdiodone(%p)\n", cbp));
CCD_DPRINTF(CCDB_IO, (cbp->cb_flags & CBF_MIRROR?
"ccdiodone: mirror component\n" :
"ccdiodone: bp %p bcount %ld resid %ld\n",
bp, bp->b_bcount, bp->b_resid));
CCD_DPRINTF(CCDB_IO, (" dev %x(u%d), cbp %p bn %d addr %p bcnt %ld\n",
vbp->b_dev, cbp->cb_comp, cbp, vbp->b_blkno,
vbp->b_data, vbp->b_bcount));
if (vbp->b_flags & B_ERROR) {
cs->sc_cinfo[cbp->cb_comp].ci_flags |= CCIF_FAILED;
if (cbp->cb_flags & CBF_MIRROR)
comptype = " (mirror)";
else {
bp->b_flags |= B_ERROR;
bp->b_error = vbp->b_error ?
vbp->b_error : EIO;
comptype = "";
}
printf("%s: error %d on component %d%s\n",
cs->sc_xname, bp->b_error, cbp->cb_comp, comptype);
}
cbp->cb_flags |= CBF_DONE;
if (cbp->cb_dep &&
(cbp->cb_dep->cb_flags & CBF_DONE) != (cbp->cb_flags & CBF_DONE))
return;
if (cbp->cb_flags & CBF_MIRROR &&
!(cbp->cb_dep->cb_flags & CBF_MIRROR)) {
cbp = cbp->cb_dep;
vbp = (struct buf *)cbp;
}
if (!old_io) {
/*
* Gather all the pieces and put them where they should be.
*/
for (i = 0, off = 0; i < cbp->cb_sgcnt; i++) {
CCD_DPRINTF(CCDB_IO,
("ccdiodone: sg %d (%p/%x) off %x\n", i,
cbp->cb_sg[i].cs_sgaddr,
cbp->cb_sg[i].cs_sglen, off));
pagemove(vbp->b_data + off, cbp->cb_sg[i].cs_sgaddr,
round_page(cbp->cb_sg[i].cs_sglen));
off += cbp->cb_sg[i].cs_sglen;
}
uvm_km_free(ccdmap, (vaddr_t)vbp->b_data, count);
if (ccd_need_kvm) {
ccd_need_kvm = 0;
wakeup(ccdmap);
}
}
count = vbp->b_bcount;
putccdbuf(cbp);
if (cbp->cb_dep)
putccdbuf(cbp->cb_dep);
/*
* If all done, "interrupt".
*
* Note that mirror component buffers aren't counted against
* the original I/O buffer.
*/
if (count > bp->b_resid)
panic("ccdiodone: count");
bp->b_resid -= count;
if (bp->b_resid == 0)
ccdintr(cs, bp);
}
/* ARGSUSED */
int
ccdread(dev_t dev, struct uio *uio, int flags)
{
int unit = ccdunit(dev);
struct ccd_softc *cs;
CCD_DPRINTF(CCDB_FOLLOW, ("ccdread(%x, %p)\n", dev, uio));
if (unit >= numccd)
return (ENXIO);
cs = &ccd_softc[unit];
if ((cs->sc_flags & CCDF_INITED) == 0)
return (ENXIO);
/*
* XXX: It's not clear that using minphys() is completely safe,
* in particular, for raw I/O. Underlying devices might have some
* non-obvious limits, because of the copy to user-space.
*/
return (physio(ccdstrategy, NULL, dev, B_READ, minphys, uio));
}
/* ARGSUSED */
int
ccdwrite(dev_t dev, struct uio *uio, int flags)
{
int unit = ccdunit(dev);
struct ccd_softc *cs;
CCD_DPRINTF(CCDB_FOLLOW, ("ccdwrite(%x, %p)\n", dev, uio));
if (unit >= numccd)
return (ENXIO);
cs = &ccd_softc[unit];
if ((cs->sc_flags & CCDF_INITED) == 0)
return (ENXIO);
/*
* XXX: It's not clear that using minphys() is completely safe,
* in particular, for raw I/O. Underlying devices might have some
* non-obvious limits, because of the copy to user-space.
*/
return (physio(ccdstrategy, NULL, dev, B_WRITE, minphys, uio));
}
int
ccdioctl(dev_t dev, u_long cmd, caddr_t data, int flag, struct proc *p)
{
int unit = ccdunit(dev);
int i, j, lookedup = 0, error = 0;
int part, pmask, s;
struct ccd_softc *cs;
struct ccd_ioctl *ccio = (struct ccd_ioctl *)data;
struct ccddevice ccd;
char **cpp;
struct vnode **vpp;
vaddr_t min, max;
if (unit >= numccd)
return (ENXIO);
cs = &ccd_softc[unit];
if (cmd != CCDIOCSET && !(cs->sc_flags & CCDF_INITED))
return (ENXIO);
/* access control */
switch (cmd) {
case CCDIOCSET:
case CCDIOCCLR:
case DIOCWDINFO:
case DIOCSDINFO:
case DIOCWLABEL:
if ((flag & FWRITE) == 0)
return (EBADF);
}
bzero(&ccd, sizeof(ccd));
switch (cmd) {
case CCDIOCSET:
if (cs->sc_flags & CCDF_INITED)
return (EBUSY);
if (ccio->ccio_ndisks == 0 || ccio->ccio_ndisks > INT_MAX ||
ccio->ccio_ileave <= 0)
return (EINVAL);
if ((error = ccdlock(cs)) != 0)
return (error);
/* Fill in some important bits. */
ccd.ccd_unit = unit;
ccd.ccd_interleave = ccio->ccio_ileave;
ccd.ccd_flags = ccio->ccio_flags & CCDF_USERMASK;
/* XXX the new code is unstable still */
ccd.ccd_flags |= CCDF_OLD;
/*
* Interleaving which is not a multiple of the click size
* must use the old I/O code (by design)
*/
if (ccio->ccio_ileave % (PAGE_SIZE / DEV_BSIZE) != 0)
ccd.ccd_flags |= CCDF_OLD;
/*
* Allocate space for and copy in the array of
* componet pathnames and device numbers.
*/
cpp = malloc(ccio->ccio_ndisks * sizeof(char *),
M_DEVBUF, M_WAITOK);
vpp = malloc(ccio->ccio_ndisks * sizeof(struct vnode *),
M_DEVBUF, M_WAITOK);
error = copyin((caddr_t)ccio->ccio_disks, (caddr_t)cpp,
ccio->ccio_ndisks * sizeof(char **));
if (error) {
free(vpp, M_DEVBUF);
free(cpp, M_DEVBUF);
ccdunlock(cs);
return (error);
}
for (i = 0; i < ccio->ccio_ndisks; ++i) {
CCD_DPRINTF(CCDB_INIT,
("ccdioctl: component %d: %p, lookedup = %d\n",
i, cpp[i], lookedup));
if ((error = ccdlookup(cpp[i], p, &vpp[i])) != 0) {
for (j = 0; j < lookedup; ++j)
(void)vn_close(vpp[j], FREAD|FWRITE,
p->p_ucred, p);
free(vpp, M_DEVBUF);
free(cpp, M_DEVBUF);
ccdunlock(cs);
return (error);
}
++lookedup;
}
ccd.ccd_cpp = cpp;
ccd.ccd_vpp = vpp;
ccd.ccd_ndev = ccio->ccio_ndisks;
/*
* Initialize the ccd. Fills in the softc for us.
*/
if ((error = ccdinit(&ccd, cpp, p)) != 0) {
for (j = 0; j < lookedup; ++j)
(void)vn_close(vpp[j], FREAD|FWRITE,
p->p_ucred, p);
bzero(&ccd_softc[unit], sizeof(struct ccd_softc));
free(vpp, M_DEVBUF);
free(cpp, M_DEVBUF);
ccdunlock(cs);
return (error);
}
/*
* The ccd has been successfully initialized, so
* we can place it into the array. Don't try to
* read the disklabel until the disk has been attached,
* because space for the disklabel is allocated
* in disk_attach();
*/
bcopy(&ccd, &ccddevs[unit], sizeof(ccd));
ccio->ccio_unit = unit;
ccio->ccio_size = cs->sc_size;
/*
* If we use the optimized protocol we need some kvm space
* for the component buffers. Allocate it here.
*
* XXX I'd like to have a more dynamic way of acquiring kvm
* XXX space, but that is problematic as we are not allowed
* XXX to lock the kernel_map in interrupt context. It is
* XXX doable via a freelist implementation though.
*/
if (!ccdmap && !(ccd.ccd_flags & CCDF_OLD)) {
min = vm_map_min(kernel_map);
ccdmap = uvm_km_suballoc(kernel_map, &min, &max,
CCD_CLUSTERS * MAXBSIZE, VM_MAP_INTRSAFE,
FALSE, NULL);
}
/* Attach the disk. */
cs->sc_dkdev.dk_name = cs->sc_xname;
disk_attach(&cs->sc_dkdev);
/* Try and read the disklabel. */
ccdgetdisklabel(dev, cs, cs->sc_dkdev.dk_label,
cs->sc_dkdev.dk_cpulabel, 0);
ccdunlock(cs);
break;
case CCDIOCCLR:
if ((error = ccdlock(cs)) != 0)
return (error);
/*
* Don't unconfigure if any other partitions are open
* or if both the character and block flavors of this
* partition are open.
*/
part = DISKPART(dev);
pmask = (1 << part);
if ((cs->sc_dkdev.dk_openmask & ~pmask) ||
((cs->sc_dkdev.dk_bopenmask & pmask) &&
(cs->sc_dkdev.dk_copenmask & pmask))) {
ccdunlock(cs);
return (EBUSY);
}
/*
* Free ccd_softc information and clear entry.
*/
/* Close the components and free their pathnames. */
for (i = 0; i < cs->sc_nccdisks; ++i) {
/*
* XXX: this close could potentially fail and
* cause Bad Things. Maybe we need to force
* the close to happen?
*/
#ifdef DIAGNOSTIC
CCD_DCALL(CCDB_VNODE, vprint("CCDIOCCLR: vnode info",
cs->sc_cinfo[i].ci_vp));
#endif
(void)vn_close(cs->sc_cinfo[i].ci_vp, FREAD|FWRITE,
p->p_ucred, p);
free(cs->sc_cinfo[i].ci_path, M_DEVBUF);
}
/* Free interleave index. */
for (i = 0; cs->sc_itable[i].ii_ndisk; ++i)
free(cs->sc_itable[i].ii_index, M_DEVBUF);
/* Free component info and interleave table. */
free(cs->sc_cinfo, M_DEVBUF);
free(cs->sc_itable, M_DEVBUF);
cs->sc_flags &= ~CCDF_INITED;
/*
* Free ccddevice information and clear entry.
*/
free(ccddevs[unit].ccd_cpp, M_DEVBUF);
free(ccddevs[unit].ccd_vpp, M_DEVBUF);
bcopy(&ccd, &ccddevs[unit], sizeof(ccd));
/* Detatch the disk. */
disk_detach(&cs->sc_dkdev);
/* This must be atomic. */
s = splhigh();
ccdunlock(cs);
bzero(cs, sizeof(struct ccd_softc));
splx(s);
break;
case DIOCGPDINFO: {
struct cpu_disklabel osdep;
if ((error = ccdlock(cs)) != 0)
return (error);
ccdgetdisklabel(dev, cs, (struct disklabel *)data,
&osdep, 1);
ccdunlock(cs);
break;
}
case DIOCGDINFO:
*(struct disklabel *)data = *(cs->sc_dkdev.dk_label);
break;
case DIOCGPART:
((struct partinfo *)data)->disklab = cs->sc_dkdev.dk_label;
((struct partinfo *)data)->part =
&cs->sc_dkdev.dk_label->d_partitions[DISKPART(dev)];
break;
case DIOCWDINFO:
case DIOCSDINFO:
if ((error = ccdlock(cs)) != 0)
return (error);
cs->sc_flags |= CCDF_LABELLING;
error = setdisklabel(cs->sc_dkdev.dk_label,
(struct disklabel *)data, 0, cs->sc_dkdev.dk_cpulabel);
if (error == 0) {
if (cmd == DIOCWDINFO)
error = writedisklabel(CCDLABELDEV(dev),
ccdstrategy, cs->sc_dkdev.dk_label,
cs->sc_dkdev.dk_cpulabel);
}
cs->sc_flags &= ~CCDF_LABELLING;
ccdunlock(cs);
if (error)
return (error);
break;
case DIOCWLABEL:
if (*(int *)data != 0)
cs->sc_flags |= CCDF_WLABEL;
else
cs->sc_flags &= ~CCDF_WLABEL;
break;
default:
return (ENOTTY);
}
return (0);
}
int
ccdsize(dev_t dev)
{
struct ccd_softc *cs;
int part, size, unit;
unit = ccdunit(dev);
if (unit >= numccd)
return (-1);
cs = &ccd_softc[unit];
if ((cs->sc_flags & CCDF_INITED) == 0)
return (-1);
if (ccdopen(dev, 0, S_IFBLK, curproc))
return (-1);
part = DISKPART(dev);
if (cs->sc_dkdev.dk_label->d_partitions[part].p_fstype != FS_SWAP)
size = -1;
else
size = cs->sc_dkdev.dk_label->d_partitions[part].p_size;
if (ccdclose(dev, 0, S_IFBLK, curproc))
return (-1);
return (size);
}
int
ccddump(dev_t dev, daddr_t blkno, caddr_t va, size_t size)
{
/* Not implemented. */
return ENXIO;
}
/*
* Lookup the provided name in the filesystem. If the file exists,
* is a valid block device, and isn't being used by anyone else,
* set *vpp to the file's vnode.
*/
int
ccdlookup(char *path, struct proc *p, struct vnode **vpp)
{
struct nameidata nd;
struct vnode *vp;
struct vattr va;
int error;
NDINIT(&nd, LOOKUP, FOLLOW, UIO_USERSPACE, path, p);
if ((error = vn_open(&nd, FREAD|FWRITE, 0)) != 0) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("ccdlookup: vn_open error = %d\n", error));
return (error);
}
vp = nd.ni_vp;
if (vp->v_usecount > 1) {
VOP_UNLOCK(vp, 0, p);
(void)vn_close(vp, FREAD|FWRITE, p->p_ucred, p);
return (EBUSY);
}
if ((error = VOP_GETATTR(vp, &va, p->p_ucred, p)) != 0) {
CCD_DPRINTF(CCDB_FOLLOW | CCDB_INIT,
("ccdlookup: getattr error = %d\n", error));
VOP_UNLOCK(vp, 0, p);
(void)vn_close(vp, FREAD|FWRITE, p->p_ucred, p);
return (error);
}
/* XXX: eventually we should handle VREG, too. */
if (va.va_type != VBLK) {
VOP_UNLOCK(vp, 0, p);
(void)vn_close(vp, FREAD|FWRITE, p->p_ucred, p);
return (ENOTBLK);
}
#ifdef DIAGNOSTIC
CCD_DCALL(CCDB_VNODE, vprint("ccdlookup: vnode info", vp));
#endif
VOP_UNLOCK(vp, 0, p);
*vpp = vp;
return (0);
}
/*
* Read the disklabel from the ccd. If one is not present, fake one
* up.
*/
void
ccdgetdisklabel(dev_t dev, struct ccd_softc *cs, struct disklabel *lp,
struct cpu_disklabel *clp, int spoofonly)
{
struct ccdgeom *ccg = &cs->sc_geom;
char *errstring;
bzero(lp, sizeof(*lp));
bzero(clp, sizeof(*clp));
lp->d_secperunit = cs->sc_size;
lp->d_secsize = ccg->ccg_secsize;
lp->d_nsectors = ccg->ccg_nsectors;
lp->d_ntracks = ccg->ccg_ntracks;
lp->d_ncylinders = ccg->ccg_ncylinders;
lp->d_secpercyl = lp->d_ntracks * lp->d_nsectors;
lp->d_rpm = ccg->ccg_rpm;
strncpy(lp->d_typename, "ccd", sizeof(lp->d_typename));
lp->d_type = DTYPE_CCD;
strncpy(lp->d_packname, "fictitious", sizeof(lp->d_packname));
lp->d_interleave = 1;
lp->d_flags = 0;
lp->d_partitions[RAW_PART].p_offset = 0;
lp->d_partitions[RAW_PART].p_size = lp->d_secperunit;
lp->d_partitions[RAW_PART].p_fstype = FS_UNUSED;
lp->d_npartitions = RAW_PART + 1;
lp->d_magic = DISKMAGIC;
lp->d_magic2 = DISKMAGIC;
lp->d_checksum = dkcksum(cs->sc_dkdev.dk_label);
/*
* Call the generic disklabel extraction routine.
*/
errstring = readdisklabel(CCDLABELDEV(dev), ccdstrategy,
cs->sc_dkdev.dk_label, cs->sc_dkdev.dk_cpulabel, spoofonly);
/* It's actually extremely common to have unlabeled ccds. */
if (errstring != NULL)
CCD_DPRINTF(CCDB_LABEL, ("%s: %s\n", cs->sc_xname, errstring));
}
/*
* Wait interruptibly for an exclusive lock.
*
* XXX
* Several drivers do this; it should be abstracted and made MP-safe.
*/
int
ccdlock(struct ccd_softc *cs)
{
int error;
while ((cs->sc_flags & CCDF_LOCKED) != 0) {
cs->sc_flags |= CCDF_WANTED;
if ((error = tsleep(cs, PRIBIO | PCATCH, "ccdlck", 0)) != 0)
return (error);
}
cs->sc_flags |= CCDF_LOCKED;
return (0);
}
/*
* Unlock and wake up any waiters.
*/
void
ccdunlock(struct ccd_softc *cs)
{
cs->sc_flags &= ~CCDF_LOCKED;
if ((cs->sc_flags & CCDF_WANTED) != 0) {
cs->sc_flags &= ~CCDF_WANTED;
wakeup(cs);
}
}
#ifdef CCDDEBUG
void
printiinfo(struct ccdiinfo *ii)
{
int ix, i;
for (ix = 0; ii->ii_ndisk; ix++, ii++) {
printf(" itab[%d]: #dk %d sblk %d soff %d",
ix, ii->ii_ndisk, ii->ii_startblk, ii->ii_startoff);
for (i = 0; i < ii->ii_ndisk; i++)
printf(" %d", ii->ii_index[i]);
printf("\n");
}
}
#endif
|