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
|
/*
* Copyright (c) 1982, 1986, 1988, 1993
* The Regents of the University of California. All rights reserved.
* (c) UNIX System Laboratories, Inc.
* All or some portions of this file are derived from material licensed
* to the University of California by American Telephone and Telegraph
* Co. or Unix System Laboratories, Inc. and are reproduced herein with
* the permission of UNIX System Laboratories, Inc.
*
* 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 University of
* California, Berkeley and its contributors.
* 4. 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.
*
* @(#)ufs_disksubr.c 8.5 (Berkeley) 1/21/94
* $Id: ufs_disksubr.c,v 1.37 1998/10/16 10:14:21 jkh Exp $
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/buf.h>
#include <sys/conf.h>
#include <sys/disklabel.h>
#include <sys/diskslice.h>
#include <sys/syslog.h>
/*
* Seek sort for disks.
*
* The buf_queue keep two queues, sorted in ascending block order. The first
* queue holds those requests which are positioned after the current block
* (in the first request); the second, which starts at queue->switch_point,
* holds requests which came in after their block number was passed. Thus
* we implement a one way scan, retracting after reaching the end of the drive
* to the first request on the second queue, at which time it becomes the
* first queue.
*
* A one-way scan is natural because of the way UNIX read-ahead blocks are
* allocated.
*/
void
bufqdisksort(bufq, bp)
struct buf_queue_head *bufq;
struct buf *bp;
{
struct buf *bq;
struct buf *bn;
struct buf *be;
be = TAILQ_LAST(&bufq->queue, buf_queue);
/*
* If the queue is empty or we are an
* ordered transaction, then it's easy.
*/
if ((bq = bufq_first(bufq)) == NULL
|| (bp->b_flags & B_ORDERED) != 0) {
bufq_insert_tail(bufq, bp);
return;
} else if (bufq->insert_point != NULL) {
/*
* A certain portion of the list is
* "locked" to preserve ordering, so
* we can only insert after the insert
* point.
*/
bq = bufq->insert_point;
} else {
/*
* If we lie before the last removed (currently active)
* request, and are not inserting ourselves into the
* "locked" portion of the list, then we must add ourselves
* to the second request list.
*/
if (bp->b_pblkno < bufq->last_pblkno) {
bq = bufq->switch_point;
/*
* If we are starting a new secondary list,
* then it's easy.
*/
if (bq == NULL) {
bufq->switch_point = bp;
bufq_insert_tail(bufq, bp);
return;
}
/*
* If we lie ahead of the current switch point,
* insert us before the switch point and move
* the switch point.
*/
if (bp->b_pblkno < bq->b_pblkno) {
bufq->switch_point = bp;
TAILQ_INSERT_BEFORE(bq, bp, b_act);
return;
}
} else {
if (bufq->switch_point != NULL)
be = TAILQ_PREV(bufq->switch_point,
buf_queue, b_act);
/*
* If we lie between last_pblkno and bq,
* insert before bq.
*/
if (bp->b_pblkno < bq->b_pblkno) {
TAILQ_INSERT_BEFORE(bq, bp, b_act);
return;
}
}
}
/*
* Request is at/after our current position in the list.
* Optimize for sequential I/O by seeing if we go at the tail.
*/
if (bp->b_pblkno > be->b_pblkno) {
TAILQ_INSERT_AFTER(&bufq->queue, be, bp, b_act);
return;
}
/* Otherwise, insertion sort */
while ((bn = TAILQ_NEXT(bq, b_act)) != NULL) {
/*
* We want to go after the current request if it is the end
* of the first request list, or if the next request is a
* larger cylinder than our request.
*/
if (bn == bufq->switch_point
|| bp->b_pblkno < bn->b_pblkno)
break;
bq = bn;
}
TAILQ_INSERT_AFTER(&bufq->queue, bq, bp, b_act);
}
/*
* Attempt to read a disk label from a device using the indicated strategy
* routine. The label must be partly set up before this: secpercyl, secsize
* and anything required in the strategy routine (e.g., dummy bounds for the
* partition containing the label) must be filled in before calling us.
* Returns NULL on success and an error string on failure.
*/
char *
readdisklabel(dev, strat, lp)
dev_t dev;
d_strategy_t *strat;
register struct disklabel *lp;
{
register struct buf *bp;
struct disklabel *dlp;
char *msg = NULL;
bp = geteblk((int)lp->d_secsize);
bp->b_dev = dev;
bp->b_blkno = LABELSECTOR * ((int)lp->d_secsize/DEV_BSIZE);
bp->b_bcount = lp->d_secsize;
bp->b_flags &= ~B_INVAL;
bp->b_flags |= B_BUSY | B_READ;
(*strat)(bp);
if (biowait(bp))
msg = "I/O error";
else for (dlp = (struct disklabel *)bp->b_data;
dlp <= (struct disklabel *)((char *)bp->b_data +
lp->d_secsize - sizeof(*dlp));
dlp = (struct disklabel *)((char *)dlp + sizeof(long))) {
if (dlp->d_magic != DISKMAGIC || dlp->d_magic2 != DISKMAGIC) {
if (msg == NULL)
msg = "no disk label";
} else if (dlp->d_npartitions > MAXPARTITIONS ||
dkcksum(dlp) != 0)
msg = "disk label corrupted";
else {
*lp = *dlp;
msg = NULL;
break;
}
}
bp->b_flags |= B_INVAL | B_AGE;
brelse(bp);
return (msg);
}
/*
* Check new disk label for sensibility before setting it.
*/
int
setdisklabel(olp, nlp, openmask)
register struct disklabel *olp, *nlp;
u_long openmask;
{
register int i;
register struct partition *opp, *npp;
/*
* Check it is actually a disklabel we are looking at.
*/
if (nlp->d_magic != DISKMAGIC || nlp->d_magic2 != DISKMAGIC ||
dkcksum(nlp) != 0)
return (EINVAL);
/*
* For each partition that we think is open,
*/
while ((i = ffs((long)openmask)) != 0) {
i--;
/*
* Check it is not changing....
*/
openmask &= ~(1 << i);
if (nlp->d_npartitions <= i)
return (EBUSY);
opp = &olp->d_partitions[i];
npp = &nlp->d_partitions[i];
if (npp->p_offset != opp->p_offset || npp->p_size < opp->p_size)
return (EBUSY);
/*
* Copy internally-set partition information
* if new label doesn't include it. XXX
* (If we are using it then we had better stay the same type)
* This is possibly dubious, as someone else noted (XXX)
*/
if (npp->p_fstype == FS_UNUSED && opp->p_fstype != FS_UNUSED) {
npp->p_fstype = opp->p_fstype;
npp->p_fsize = opp->p_fsize;
npp->p_frag = opp->p_frag;
npp->p_cpg = opp->p_cpg;
}
}
nlp->d_checksum = 0;
nlp->d_checksum = dkcksum(nlp);
*olp = *nlp;
return (0);
}
/*
* Write disk label back to device after modification.
*/
int
writedisklabel(dev, strat, lp)
dev_t dev;
d_strategy_t *strat;
register struct disklabel *lp;
{
struct buf *bp;
struct disklabel *dlp;
int error = 0;
if (lp->d_partitions[RAW_PART].p_offset != 0)
return (EXDEV); /* not quite right */
bp = geteblk((int)lp->d_secsize);
bp->b_dev = dkmodpart(dev, RAW_PART);
bp->b_blkno = LABELSECTOR * ((int)lp->d_secsize/DEV_BSIZE);
bp->b_bcount = lp->d_secsize;
#if 1
/*
* We read the label first to see if it's there,
* in which case we will put ours at the same offset into the block..
* (I think this is stupid [Julian])
* Note that you can't write a label out over a corrupted label!
* (also stupid.. how do you write the first one? by raw writes?)
*/
bp->b_flags &= ~B_INVAL;
bp->b_flags |= B_BUSY | B_READ;
(*strat)(bp);
error = biowait(bp);
if (error)
goto done;
for (dlp = (struct disklabel *)bp->b_data;
dlp <= (struct disklabel *)
((char *)bp->b_data + lp->d_secsize - sizeof(*dlp));
dlp = (struct disklabel *)((char *)dlp + sizeof(long))) {
if (dlp->d_magic == DISKMAGIC && dlp->d_magic2 == DISKMAGIC &&
dkcksum(dlp) == 0) {
*dlp = *lp;
bp->b_flags &= ~(B_DONE | B_READ);
bp->b_flags |= B_BUSY | B_WRITE;
#ifdef __alpha__
alpha_fix_srm_checksum(bp);
#endif
(*strat)(bp);
error = biowait(bp);
goto done;
}
}
error = ESRCH;
done:
#else
bzero(bp->b_data, lp->d_secsize);
dlp = (struct disklabel *)bp->b_data;
*dlp = *lp;
bp->b_flags &= ~B_INVAL;
bp->b_flags |= B_BUSY | B_WRITE;
(*strat)(bp);
error = biowait(bp);
#endif
bp->b_flags |= B_INVAL | B_AGE;
brelse(bp);
return (error);
}
/*
* Compute checksum for disk label.
*/
u_int
dkcksum(lp)
register struct disklabel *lp;
{
register u_short *start, *end;
register u_short sum = 0;
start = (u_short *)lp;
end = (u_short *)&lp->d_partitions[lp->d_npartitions];
while (start < end)
sum ^= *start++;
return (sum);
}
/*
* Disk error is the preface to plaintive error messages
* about failing disk transfers. It prints messages of the form
hp0g: hard error reading fsbn 12345 of 12344-12347 (hp0 bn %d cn %d tn %d sn %d)
* if the offset of the error in the transfer and a disk label
* are both available. blkdone should be -1 if the position of the error
* is unknown; the disklabel pointer may be null from drivers that have not
* been converted to use them. The message is printed with printf
* if pri is LOG_PRINTF, otherwise it uses log at the specified priority.
* The message should be completed (with at least a newline) with printf
* or addlog, respectively. There is no trailing space.
*/
void
diskerr(bp, dname, what, pri, blkdone, lp)
register struct buf *bp;
char *dname, *what;
int pri, blkdone;
register struct disklabel *lp;
{
int unit = dkunit(bp->b_dev);
int slice = dkslice(bp->b_dev);
int part = dkpart(bp->b_dev);
register int (*pr) __P((const char *, ...));
char partname[2];
char *sname;
daddr_t sn;
if (pri != LOG_PRINTF) {
log(pri, "");
pr = addlog;
} else
pr = printf;
sname = dsname(dname, unit, slice, part, partname);
(*pr)("%s%s: %s %sing fsbn ", sname, partname, what,
bp->b_flags & B_READ ? "read" : "writ");
sn = bp->b_blkno;
if (bp->b_bcount <= DEV_BSIZE)
(*pr)("%ld", (long)sn);
else {
if (blkdone >= 0) {
sn += blkdone;
(*pr)("%ld of ", (long)sn);
}
(*pr)("%ld-%ld", (long)bp->b_blkno,
(long)(bp->b_blkno + (bp->b_bcount - 1) / DEV_BSIZE));
}
if (lp && (blkdone >= 0 || bp->b_bcount <= lp->d_secsize)) {
#ifdef tahoe
sn *= DEV_BSIZE / lp->d_secsize; /* XXX */
#endif
sn += lp->d_partitions[part].p_offset;
/*
* XXX should add slice offset and not print the slice,
* but we don't know the slice pointer.
* XXX should print bp->b_pblkno so that this will work
* independent of slices, labels and bad sector remapping,
* but some drivers don't set bp->b_pblkno.
*/
(*pr)(" (%s bn %ld; cn %ld", sname, (long)sn,
(long)(sn / lp->d_secpercyl));
sn %= (long)lp->d_secpercyl;
(*pr)(" tn %ld sn %ld)", (long)(sn / lp->d_nsectors),
(long)(sn % lp->d_nsectors));
}
}
|