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
|
/*-
* Copyright (c) 2002 Poul-Henning Kamp
* Copyright (c) 2002 Networks Associates Technology, Inc.
* All rights reserved.
*
* This software was developed for the FreeBSD Project by Poul-Henning Kamp
* and NAI Labs, the Security Research Division of Network Associates, Inc.
* under DARPA/SPAWAR contract N66001-01-C-8035 ("CBOSS"), as part of the
* DARPA CHATS research program.
*
* 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. The names of the authors may not be used to endorse or promote
* products derived from this software without specific prior written
* permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR 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 AUTHOR 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.
*
* $FreeBSD$
*/
#include <sys/param.h>
#include <sys/stdint.h>
#ifndef _KERNEL
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <signal.h>
#include <err.h>
#include <sched.h>
#else
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/bio.h>
#endif
#include <sys/errno.h>
#include <geom/geom.h>
#include <geom/geom_int.h>
#include <geom/geom_stats.h>
static struct g_bioq g_bio_run_down;
static struct g_bioq g_bio_run_up;
static struct g_bioq g_bio_run_task;
static struct g_bioq g_bio_idle;
static u_int pace;
#include <machine/atomic.h>
static void
g_bioq_lock(struct g_bioq *bq)
{
mtx_lock(&bq->bio_queue_lock);
}
static void
g_bioq_unlock(struct g_bioq *bq)
{
mtx_unlock(&bq->bio_queue_lock);
}
#if 0
static void
g_bioq_destroy(struct g_bioq *bq)
{
mtx_destroy(&bq->bio_queue_lock);
}
#endif
static void
g_bioq_init(struct g_bioq *bq)
{
TAILQ_INIT(&bq->bio_queue);
mtx_init(&bq->bio_queue_lock, "bio queue", NULL, MTX_DEF);
}
static struct bio *
g_bioq_first(struct g_bioq *bq)
{
struct bio *bp;
bp = TAILQ_FIRST(&bq->bio_queue);
if (bp != NULL) {
TAILQ_REMOVE(&bq->bio_queue, bp, bio_queue);
bq->bio_queue_length--;
}
return (bp);
}
static void
g_bioq_enqueue_tail(struct bio *bp, struct g_bioq *rq)
{
g_bioq_lock(rq);
TAILQ_INSERT_TAIL(&rq->bio_queue, bp, bio_queue);
rq->bio_queue_length++;
g_bioq_unlock(rq);
}
struct bio *
g_new_bio(void)
{
struct bio *bp;
g_bioq_lock(&g_bio_idle);
bp = g_bioq_first(&g_bio_idle);
g_bioq_unlock(&g_bio_idle);
if (bp == NULL)
bp = g_malloc(sizeof *bp, M_NOWAIT | M_ZERO);
/* g_trace(G_T_BIO, "g_new_bio() = %p", bp); */
return (bp);
}
void
g_destroy_bio(struct bio *bp)
{
/* g_trace(G_T_BIO, "g_destroy_bio(%p)", bp); */
bzero(bp, sizeof *bp);
g_bioq_enqueue_tail(bp, &g_bio_idle);
}
struct bio *
g_clone_bio(struct bio *bp)
{
struct bio *bp2;
bp2 = g_new_bio();
if (bp2 != NULL) {
bp2->bio_parent = bp;
bp2->bio_cmd = bp->bio_cmd;
bp2->bio_length = bp->bio_length;
bp2->bio_offset = bp->bio_offset;
bp2->bio_data = bp->bio_data;
bp2->bio_attribute = bp->bio_attribute;
bp->bio_children++;
}
/* g_trace(G_T_BIO, "g_clone_bio(%p) = %p", bp, bp2); */
return(bp2);
}
void
g_io_init()
{
g_bioq_init(&g_bio_run_down);
g_bioq_init(&g_bio_run_up);
g_bioq_init(&g_bio_run_task);
g_bioq_init(&g_bio_idle);
}
int
g_io_setattr(const char *attr, struct g_consumer *cp, int len, void *ptr)
{
struct bio *bp;
int error;
g_trace(G_T_BIO, "bio_setattr(%s)", attr);
bp = g_new_bio();
bp->bio_cmd = BIO_SETATTR;
bp->bio_done = NULL;
bp->bio_attribute = attr;
bp->bio_length = len;
bp->bio_data = ptr;
g_io_request(bp, cp);
error = biowait(bp, "gsetattr");
g_destroy_bio(bp);
return (error);
}
int
g_io_getattr(const char *attr, struct g_consumer *cp, int *len, void *ptr)
{
struct bio *bp;
int error;
g_trace(G_T_BIO, "bio_getattr(%s)", attr);
bp = g_new_bio();
bp->bio_cmd = BIO_GETATTR;
bp->bio_done = NULL;
bp->bio_attribute = attr;
bp->bio_length = *len;
bp->bio_data = ptr;
g_io_request(bp, cp);
error = biowait(bp, "ggetattr");
*len = bp->bio_completed;
g_destroy_bio(bp);
return (error);
}
static int
g_io_check(struct bio *bp)
{
struct g_consumer *cp;
struct g_provider *pp;
cp = bp->bio_from;
pp = bp->bio_to;
/* Fail if access counters dont allow the operation */
switch(bp->bio_cmd) {
case BIO_READ:
case BIO_GETATTR:
if (cp->acr == 0)
return (EPERM);
break;
case BIO_WRITE:
case BIO_DELETE:
case BIO_SETATTR:
if (cp->acw == 0)
return (EPERM);
break;
default:
return (EPERM);
}
/* if provider is marked for error, don't disturb. */
if (pp->error)
return (pp->error);
switch(bp->bio_cmd) {
case BIO_READ:
case BIO_WRITE:
case BIO_DELETE:
/* Reject I/O not on sector boundary */
if (bp->bio_offset % pp->sectorsize)
return (EINVAL);
/* Reject I/O not integral sector long */
if (bp->bio_length % pp->sectorsize)
return (EINVAL);
/* Reject requests past the end of media. */
if (bp->bio_offset > pp->mediasize)
return (EIO);
break;
default:
break;
}
return (0);
}
void
g_io_request(struct bio *bp, struct g_consumer *cp)
{
struct g_provider *pp;
struct bintime bt;
pp = cp->provider;
KASSERT(cp != NULL, ("NULL cp in g_io_request"));
KASSERT(bp != NULL, ("NULL bp in g_io_request"));
KASSERT(bp->bio_data != NULL, ("NULL bp->data in g_io_request"));
KASSERT(pp != NULL, ("consumer not attached in g_io_request"));
bp->bio_from = cp;
bp->bio_to = pp;
bp->bio_error = 0;
bp->bio_completed = 0;
if (g_collectstats) {
binuptime(&bt);
bp->bio_t0 = bt;
if (cp->nstart == cp->nend)
cp->stat->wentbusy = bt; /* Consumer is idle */
if (pp->nstart == pp->nend)
pp->stat->wentbusy = bt; /* Provider is idle */
cp->stat->nop++;
pp->stat->nop++;
}
cp->nstart++;
pp->nstart++;
/* Pass it on down. */
g_trace(G_T_BIO, "bio_request(%p) from %p(%s) to %p(%s) cmd %d",
bp, cp, cp->geom->name, pp, pp->name, bp->bio_cmd);
g_bioq_enqueue_tail(bp, &g_bio_run_down);
wakeup(&g_wait_down);
}
void
g_io_deliver(struct bio *bp, int error)
{
struct g_consumer *cp;
struct g_provider *pp;
struct bintime t1, dt;
int idx;
cp = bp->bio_from;
pp = bp->bio_to;
KASSERT(bp != NULL, ("NULL bp in g_io_deliver"));
KASSERT(cp != NULL, ("NULL bio_from in g_io_deliver"));
KASSERT(cp->geom != NULL, ("NULL bio_from->geom in g_io_deliver"));
KASSERT(pp != NULL, ("NULL bio_to in g_io_deliver"));
g_trace(G_T_BIO,
"g_io_deliver(%p) from %p(%s) to %p(%s) cmd %d error %d off %jd len %jd",
bp, cp, cp->geom->name, pp, pp->name, bp->bio_cmd, error,
(intmax_t)bp->bio_offset, (intmax_t)bp->bio_length);
if (g_collectstats) {
switch (bp->bio_cmd) {
case BIO_READ: idx = G_STAT_IDX_READ; break;
case BIO_WRITE: idx = G_STAT_IDX_WRITE; break;
case BIO_DELETE: idx = G_STAT_IDX_DELETE; break;
case BIO_GETATTR: idx = -1; break;
case BIO_SETATTR: idx = -1; break;
default:
panic("unknown bio_cmd in g_io_deliver");
break;
}
binuptime(&t1);
/* Raise the "inconsistent" flag for userland */
atomic_add_acq_int(&cp->stat->seq0, 1);
atomic_add_acq_int(&pp->stat->seq0, 1);
if (idx >= 0) {
/* Account the service time */
dt = t1;
bintime_sub(&dt, &bp->bio_t0);
bintime_add(&cp->stat->ops[idx].dt, &dt);
bintime_add(&pp->stat->ops[idx].dt, &dt);
/* ... and the metrics */
pp->stat->ops[idx].nbyte += bp->bio_completed;
cp->stat->ops[idx].nbyte += bp->bio_completed;
pp->stat->ops[idx].nop++;
cp->stat->ops[idx].nop++;
/* ... and any errors */
if (error == ENOMEM) {
cp->stat->ops[idx].nmem++;
pp->stat->ops[idx].nmem++;
} else if (error != 0) {
cp->stat->ops[idx].nerr++;
pp->stat->ops[idx].nerr++;
}
}
/* Account for busy time on the consumer */
dt = t1;
bintime_sub(&dt, &cp->stat->wentbusy);
bintime_add(&cp->stat->bt, &dt);
cp->stat->wentbusy = t1;
/* Account for busy time on the provider */
dt = t1;
bintime_sub(&dt, &pp->stat->wentbusy);
bintime_add(&pp->stat->bt, &dt);
pp->stat->wentbusy = t1;
/* Mark the structures as consistent again */
atomic_add_acq_int(&cp->stat->seq1, 1);
atomic_add_acq_int(&pp->stat->seq1, 1);
cp->stat->nend++;
pp->stat->nend++;
}
cp->nend++;
pp->nend++;
if (error == ENOMEM) {
printf("ENOMEM %p on %p(%s)\n", bp, pp, pp->name);
g_io_request(bp, cp);
pace++;
return;
}
bp->bio_error = error;
g_bioq_enqueue_tail(bp, &g_bio_run_up);
wakeup(&g_wait_up);
}
void
g_io_schedule_down(struct thread *tp __unused)
{
struct bio *bp;
off_t excess;
int error;
struct mtx mymutex;
bzero(&mymutex, sizeof mymutex);
mtx_init(&mymutex, "g_xdown", MTX_DEF, 0);
for(;;) {
g_bioq_lock(&g_bio_run_down);
bp = g_bioq_first(&g_bio_run_down);
if (bp == NULL) {
msleep(&g_wait_down, &g_bio_run_down.bio_queue_lock,
PRIBIO | PDROP, "g_down", hz/10);
continue;
}
g_bioq_unlock(&g_bio_run_down);
error = g_io_check(bp);
if (error) {
g_io_deliver(bp, error);
continue;
}
switch (bp->bio_cmd) {
case BIO_READ:
case BIO_WRITE:
case BIO_DELETE:
/* Truncate requests to the end of providers media. */
excess = bp->bio_offset + bp->bio_length;
if (excess > bp->bio_to->mediasize) {
excess -= bp->bio_to->mediasize;
bp->bio_length -= excess;
}
/* Deliver zero length transfers right here. */
if (bp->bio_length == 0) {
g_io_deliver(bp, 0);
continue;
}
break;
default:
break;
}
mtx_lock(&mymutex);
bp->bio_to->geom->start(bp);
mtx_unlock(&mymutex);
if (pace) {
pace--;
break;
}
}
}
void
bio_taskqueue(struct bio *bp, bio_task_t *func, void *arg)
{
bp->bio_task = func;
bp->bio_task_arg = arg;
/*
* The taskqueue is actually just a second queue off the "up"
* queue, so we use the same lock.
*/
g_bioq_lock(&g_bio_run_up);
TAILQ_INSERT_TAIL(&g_bio_run_task.bio_queue, bp, bio_queue);
g_bio_run_task.bio_queue_length++;
wakeup(&g_wait_up);
g_bioq_unlock(&g_bio_run_up);
}
void
g_io_schedule_up(struct thread *tp __unused)
{
struct bio *bp;
struct mtx mymutex;
bzero(&mymutex, sizeof mymutex);
mtx_init(&mymutex, "g_xup", MTX_DEF, 0);
for(;;) {
g_bioq_lock(&g_bio_run_up);
bp = g_bioq_first(&g_bio_run_task);
if (bp != NULL) {
g_bioq_unlock(&g_bio_run_up);
mtx_lock(&mymutex);
bp->bio_task(bp, bp->bio_task_arg);
mtx_unlock(&mymutex);
continue;
}
bp = g_bioq_first(&g_bio_run_up);
if (bp != NULL) {
g_bioq_unlock(&g_bio_run_up);
mtx_lock(&mymutex);
biodone(bp);
mtx_unlock(&mymutex);
continue;
}
msleep(&g_wait_up, &g_bio_run_up.bio_queue_lock,
PRIBIO | PDROP, "g_up", hz/10);
}
}
void *
g_read_data(struct g_consumer *cp, off_t offset, off_t length, int *error)
{
struct bio *bp;
void *ptr;
int errorc;
bp = g_new_bio();
bp->bio_cmd = BIO_READ;
bp->bio_done = NULL;
bp->bio_offset = offset;
bp->bio_length = length;
ptr = g_malloc(length, M_WAITOK);
bp->bio_data = ptr;
g_io_request(bp, cp);
errorc = biowait(bp, "gread");
if (error != NULL)
*error = errorc;
g_destroy_bio(bp);
if (errorc) {
g_free(ptr);
ptr = NULL;
}
return (ptr);
}
int
g_write_data(struct g_consumer *cp, off_t offset, void *ptr, off_t length)
{
struct bio *bp;
int error;
bp = g_new_bio();
bp->bio_cmd = BIO_WRITE;
bp->bio_done = NULL;
bp->bio_offset = offset;
bp->bio_length = length;
bp->bio_data = ptr;
g_io_request(bp, cp);
error = biowait(bp, "gwrite");
g_destroy_bio(bp);
return (error);
}
|