blk-sysfs.c 20.0 KB
Newer Older
1 2 3 4
/*
 * Functions related to sysfs handling
 */
#include <linux/kernel.h>
5
#include <linux/slab.h>
6 7 8
#include <linux/module.h>
#include <linux/bio.h>
#include <linux/blkdev.h>
9
#include <linux/backing-dev.h>
10
#include <linux/blktrace_api.h>
11
#include <linux/blk-mq.h>
12
#include <linux/blk-cgroup.h>
13 14

#include "blk.h"
15
#include "blk-mq.h"
16 17 18 19 20 21 22 23

struct queue_sysfs_entry {
	struct attribute attr;
	ssize_t (*show)(struct request_queue *, char *);
	ssize_t (*store)(struct request_queue *, const char *, size_t);
};

static ssize_t
24
queue_var_show(unsigned long var, char *page)
25
{
26
	return sprintf(page, "%lu\n", var);
27 28 29 30 31
}

static ssize_t
queue_var_store(unsigned long *var, const char *page, size_t count)
{
32 33 34
	int err;
	unsigned long v;

35
	err = kstrtoul(page, 10, &v);
36 37 38 39
	if (err || v > UINT_MAX)
		return -EINVAL;

	*var = v;
40 41 42 43 44 45 46 47 48 49 50 51 52

	return count;
}

static ssize_t queue_requests_show(struct request_queue *q, char *page)
{
	return queue_var_show(q->nr_requests, (page));
}

static ssize_t
queue_requests_store(struct request_queue *q, const char *page, size_t count)
{
	unsigned long nr;
53
	int ret, err;
54

55
	if (!q->request_fn && !q->mq_ops)
56 57 58
		return -EINVAL;

	ret = queue_var_store(&nr, page, count);
59 60 61
	if (ret < 0)
		return ret;

62 63 64
	if (nr < BLKDEV_MIN_RQ)
		nr = BLKDEV_MIN_RQ;

65 66 67 68 69 70 71 72
	if (q->request_fn)
		err = blk_update_nr_requests(q, nr);
	else
		err = blk_mq_update_nr_requests(q, nr);

	if (err)
		return err;

73 74 75 76 77
	return ret;
}

static ssize_t queue_ra_show(struct request_queue *q, char *page)
{
78
	unsigned long ra_kb = q->backing_dev_info.ra_pages <<
79
					(PAGE_SHIFT - 10);
80 81 82 83 84 85 86 87 88 89

	return queue_var_show(ra_kb, (page));
}

static ssize_t
queue_ra_store(struct request_queue *q, const char *page, size_t count)
{
	unsigned long ra_kb;
	ssize_t ret = queue_var_store(&ra_kb, page, count);

90 91 92
	if (ret < 0)
		return ret;

93
	q->backing_dev_info.ra_pages = ra_kb >> (PAGE_SHIFT - 10);
94 95 96 97 98 99

	return ret;
}

static ssize_t queue_max_sectors_show(struct request_queue *q, char *page)
{
100
	int max_sectors_kb = queue_max_sectors(q) >> 1;
101 102 103 104

	return queue_var_show(max_sectors_kb, (page));
}

105 106 107 108 109
static ssize_t queue_max_segments_show(struct request_queue *q, char *page)
{
	return queue_var_show(queue_max_segments(q), (page));
}

110 111 112 113 114
static ssize_t queue_max_integrity_segments_show(struct request_queue *q, char *page)
{
	return queue_var_show(q->limits.max_integrity_segments, (page));
}

115 116
static ssize_t queue_max_segment_size_show(struct request_queue *q, char *page)
{
117
	if (blk_queue_cluster(q))
118 119
		return queue_var_show(queue_max_segment_size(q), (page));

120
	return queue_var_show(PAGE_SIZE, (page));
121 122
}

123
static ssize_t queue_logical_block_size_show(struct request_queue *q, char *page)
M
Martin K. Petersen 已提交
124
{
125
	return queue_var_show(queue_logical_block_size(q), page);
M
Martin K. Petersen 已提交
126 127
}

128 129 130 131 132
static ssize_t queue_physical_block_size_show(struct request_queue *q, char *page)
{
	return queue_var_show(queue_physical_block_size(q), page);
}

133 134 135 136 137
static ssize_t queue_chunk_sectors_show(struct request_queue *q, char *page)
{
	return queue_var_show(q->limits.chunk_sectors, page);
}

138 139 140 141 142 143 144 145
static ssize_t queue_io_min_show(struct request_queue *q, char *page)
{
	return queue_var_show(queue_io_min(q), page);
}

static ssize_t queue_io_opt_show(struct request_queue *q, char *page)
{
	return queue_var_show(queue_io_opt(q), page);
M
Martin K. Petersen 已提交
146 147
}

148 149 150 151 152
static ssize_t queue_discard_granularity_show(struct request_queue *q, char *page)
{
	return queue_var_show(q->limits.discard_granularity, page);
}

153 154 155
static ssize_t queue_discard_max_hw_show(struct request_queue *q, char *page)
{

156 157
	return sprintf(page, "%llu\n",
		(unsigned long long)q->limits.max_hw_discard_sectors << 9);
158 159
}

160 161
static ssize_t queue_discard_max_show(struct request_queue *q, char *page)
{
162 163
	return sprintf(page, "%llu\n",
		       (unsigned long long)q->limits.max_discard_sectors << 9);
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
static ssize_t queue_discard_max_store(struct request_queue *q,
				       const char *page, size_t count)
{
	unsigned long max_discard;
	ssize_t ret = queue_var_store(&max_discard, page, count);

	if (ret < 0)
		return ret;

	if (max_discard & (q->limits.discard_granularity - 1))
		return -EINVAL;

	max_discard >>= 9;
	if (max_discard > UINT_MAX)
		return -EINVAL;

	if (max_discard > q->limits.max_hw_discard_sectors)
		max_discard = q->limits.max_hw_discard_sectors;

	q->limits.max_discard_sectors = max_discard;
	return ret;
}

189 190 191 192 193
static ssize_t queue_discard_zeroes_data_show(struct request_queue *q, char *page)
{
	return queue_var_show(queue_discard_zeroes_data(q), page);
}

194 195 196 197 198 199 200
static ssize_t queue_write_same_max_show(struct request_queue *q, char *page)
{
	return sprintf(page, "%llu\n",
		(unsigned long long)q->limits.max_write_same_sectors << 9);
}


201 202 203 204
static ssize_t
queue_max_sectors_store(struct request_queue *q, const char *page, size_t count)
{
	unsigned long max_sectors_kb,
205
		max_hw_sectors_kb = queue_max_hw_sectors(q) >> 1,
206
			page_kb = 1 << (PAGE_SHIFT - 10);
207 208
	ssize_t ret = queue_var_store(&max_sectors_kb, page, count);

209 210 211
	if (ret < 0)
		return ret;

212 213 214
	max_hw_sectors_kb = min_not_zero(max_hw_sectors_kb, (unsigned long)
					 q->limits.max_dev_sectors >> 1);

215 216
	if (max_sectors_kb > max_hw_sectors_kb || max_sectors_kb < page_kb)
		return -EINVAL;
217

218
	spin_lock_irq(q->queue_lock);
219
	q->limits.max_sectors = max_sectors_kb << 1;
220 221 222 223 224 225 226
	spin_unlock_irq(q->queue_lock);

	return ret;
}

static ssize_t queue_max_hw_sectors_show(struct request_queue *q, char *page)
{
227
	int max_hw_sectors_kb = queue_max_hw_sectors(q) >> 1;
228 229 230 231

	return queue_var_show(max_hw_sectors_kb, (page));
}

232 233 234 235 236 237 238 239 240 241 242 243 244 245
#define QUEUE_SYSFS_BIT_FNS(name, flag, neg)				\
static ssize_t								\
queue_show_##name(struct request_queue *q, char *page)			\
{									\
	int bit;							\
	bit = test_bit(QUEUE_FLAG_##flag, &q->queue_flags);		\
	return queue_var_show(neg ? !bit : bit, page);			\
}									\
static ssize_t								\
queue_store_##name(struct request_queue *q, const char *page, size_t count) \
{									\
	unsigned long val;						\
	ssize_t ret;							\
	ret = queue_var_store(&val, page, count);			\
246 247
	if (ret < 0)							\
		 return ret;						\
248 249 250 251 252 253 254 255 256 257
	if (neg)							\
		val = !val;						\
									\
	spin_lock_irq(q->queue_lock);					\
	if (val)							\
		queue_flag_set(QUEUE_FLAG_##flag, q);			\
	else								\
		queue_flag_clear(QUEUE_FLAG_##flag, q);			\
	spin_unlock_irq(q->queue_lock);					\
	return ret;							\
258 259
}

260 261 262 263
QUEUE_SYSFS_BIT_FNS(nonrot, NONROT, 1);
QUEUE_SYSFS_BIT_FNS(random, ADD_RANDOM, 0);
QUEUE_SYSFS_BIT_FNS(iostats, IO_STAT, 0);
#undef QUEUE_SYSFS_BIT_FNS
264

D
Damien Le Moal 已提交
265 266 267 268 269 270 271 272 273 274 275 276
static ssize_t queue_zoned_show(struct request_queue *q, char *page)
{
	switch (blk_queue_zoned_model(q)) {
	case BLK_ZONED_HA:
		return sprintf(page, "host-aware\n");
	case BLK_ZONED_HM:
		return sprintf(page, "host-managed\n");
	default:
		return sprintf(page, "none\n");
	}
}

277 278
static ssize_t queue_nomerges_show(struct request_queue *q, char *page)
{
279 280
	return queue_var_show((blk_queue_nomerges(q) << 1) |
			       blk_queue_noxmerges(q), page);
281 282 283 284 285 286 287 288
}

static ssize_t queue_nomerges_store(struct request_queue *q, const char *page,
				    size_t count)
{
	unsigned long nm;
	ssize_t ret = queue_var_store(&nm, page, count);

289 290 291
	if (ret < 0)
		return ret;

292
	spin_lock_irq(q->queue_lock);
293 294 295
	queue_flag_clear(QUEUE_FLAG_NOMERGES, q);
	queue_flag_clear(QUEUE_FLAG_NOXMERGES, q);
	if (nm == 2)
296
		queue_flag_set(QUEUE_FLAG_NOMERGES, q);
297 298
	else if (nm)
		queue_flag_set(QUEUE_FLAG_NOXMERGES, q);
299
	spin_unlock_irq(q->queue_lock);
300

301 302 303
	return ret;
}

304 305
static ssize_t queue_rq_affinity_show(struct request_queue *q, char *page)
{
306
	bool set = test_bit(QUEUE_FLAG_SAME_COMP, &q->queue_flags);
D
Dan Williams 已提交
307
	bool force = test_bit(QUEUE_FLAG_SAME_FORCE, &q->queue_flags);
308

D
Dan Williams 已提交
309
	return queue_var_show(set << force, page);
310 311 312 313 314 315
}

static ssize_t
queue_rq_affinity_store(struct request_queue *q, const char *page, size_t count)
{
	ssize_t ret = -EINVAL;
316
#ifdef CONFIG_SMP
317 318 319
	unsigned long val;

	ret = queue_var_store(&val, page, count);
320 321 322
	if (ret < 0)
		return ret;

323
	spin_lock_irq(q->queue_lock);
324
	if (val == 2) {
325
		queue_flag_set(QUEUE_FLAG_SAME_COMP, q);
326 327 328 329 330
		queue_flag_set(QUEUE_FLAG_SAME_FORCE, q);
	} else if (val == 1) {
		queue_flag_set(QUEUE_FLAG_SAME_COMP, q);
		queue_flag_clear(QUEUE_FLAG_SAME_FORCE, q);
	} else if (val == 0) {
D
Dan Williams 已提交
331 332 333
		queue_flag_clear(QUEUE_FLAG_SAME_COMP, q);
		queue_flag_clear(QUEUE_FLAG_SAME_FORCE, q);
	}
334 335 336 337
	spin_unlock_irq(q->queue_lock);
#endif
	return ret;
}
338

J
Jens Axboe 已提交
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
static ssize_t queue_poll_show(struct request_queue *q, char *page)
{
	return queue_var_show(test_bit(QUEUE_FLAG_POLL, &q->queue_flags), page);
}

static ssize_t queue_poll_store(struct request_queue *q, const char *page,
				size_t count)
{
	unsigned long poll_on;
	ssize_t ret;

	if (!q->mq_ops || !q->mq_ops->poll)
		return -EINVAL;

	ret = queue_var_store(&poll_on, page, count);
	if (ret < 0)
		return ret;

	spin_lock_irq(q->queue_lock);
	if (poll_on)
		queue_flag_set(QUEUE_FLAG_POLL, q);
	else
		queue_flag_clear(QUEUE_FLAG_POLL, q);
	spin_unlock_irq(q->queue_lock);

	return ret;
}

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
static ssize_t queue_wc_show(struct request_queue *q, char *page)
{
	if (test_bit(QUEUE_FLAG_WC, &q->queue_flags))
		return sprintf(page, "write back\n");

	return sprintf(page, "write through\n");
}

static ssize_t queue_wc_store(struct request_queue *q, const char *page,
			      size_t count)
{
	int set = -1;

	if (!strncmp(page, "write back", 10))
		set = 1;
	else if (!strncmp(page, "write through", 13) ||
		 !strncmp(page, "none", 4))
		set = 0;

	if (set == -1)
		return -EINVAL;

	spin_lock_irq(q->queue_lock);
	if (set)
		queue_flag_set(QUEUE_FLAG_WC, q);
	else
		queue_flag_clear(QUEUE_FLAG_WC, q);
	spin_unlock_irq(q->queue_lock);

	return count;
}

399 400 401 402 403
static ssize_t queue_dax_show(struct request_queue *q, char *page)
{
	return queue_var_show(blk_queue_dax(q), page);
}

404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423
static ssize_t print_stat(char *page, struct blk_rq_stat *stat, const char *pre)
{
	return sprintf(page, "%s samples=%llu, mean=%lld, min=%lld, max=%lld\n",
			pre, (long long) stat->nr_samples,
			(long long) stat->mean, (long long) stat->min,
			(long long) stat->max);
}

static ssize_t queue_stats_show(struct request_queue *q, char *page)
{
	struct blk_rq_stat stat[2];
	ssize_t ret;

	blk_queue_stat_get(q, stat);

	ret = print_stat(page, &stat[BLK_STAT_READ], "read :");
	ret += print_stat(page + ret, &stat[BLK_STAT_WRITE], "write:");
	return ret;
}

424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446
static struct queue_sysfs_entry queue_requests_entry = {
	.attr = {.name = "nr_requests", .mode = S_IRUGO | S_IWUSR },
	.show = queue_requests_show,
	.store = queue_requests_store,
};

static struct queue_sysfs_entry queue_ra_entry = {
	.attr = {.name = "read_ahead_kb", .mode = S_IRUGO | S_IWUSR },
	.show = queue_ra_show,
	.store = queue_ra_store,
};

static struct queue_sysfs_entry queue_max_sectors_entry = {
	.attr = {.name = "max_sectors_kb", .mode = S_IRUGO | S_IWUSR },
	.show = queue_max_sectors_show,
	.store = queue_max_sectors_store,
};

static struct queue_sysfs_entry queue_max_hw_sectors_entry = {
	.attr = {.name = "max_hw_sectors_kb", .mode = S_IRUGO },
	.show = queue_max_hw_sectors_show,
};

447 448 449 450 451
static struct queue_sysfs_entry queue_max_segments_entry = {
	.attr = {.name = "max_segments", .mode = S_IRUGO },
	.show = queue_max_segments_show,
};

452 453 454 455 456
static struct queue_sysfs_entry queue_max_integrity_segments_entry = {
	.attr = {.name = "max_integrity_segments", .mode = S_IRUGO },
	.show = queue_max_integrity_segments_show,
};

457 458 459 460 461
static struct queue_sysfs_entry queue_max_segment_size_entry = {
	.attr = {.name = "max_segment_size", .mode = S_IRUGO },
	.show = queue_max_segment_size_show,
};

462 463 464 465 466 467
static struct queue_sysfs_entry queue_iosched_entry = {
	.attr = {.name = "scheduler", .mode = S_IRUGO | S_IWUSR },
	.show = elv_iosched_show,
	.store = elv_iosched_store,
};

M
Martin K. Petersen 已提交
468 469
static struct queue_sysfs_entry queue_hw_sector_size_entry = {
	.attr = {.name = "hw_sector_size", .mode = S_IRUGO },
470 471 472 473 474 475
	.show = queue_logical_block_size_show,
};

static struct queue_sysfs_entry queue_logical_block_size_entry = {
	.attr = {.name = "logical_block_size", .mode = S_IRUGO },
	.show = queue_logical_block_size_show,
M
Martin K. Petersen 已提交
476 477
};

478 479 480 481 482
static struct queue_sysfs_entry queue_physical_block_size_entry = {
	.attr = {.name = "physical_block_size", .mode = S_IRUGO },
	.show = queue_physical_block_size_show,
};

483 484 485 486 487
static struct queue_sysfs_entry queue_chunk_sectors_entry = {
	.attr = {.name = "chunk_sectors", .mode = S_IRUGO },
	.show = queue_chunk_sectors_show,
};

488 489 490 491 492 493 494 495
static struct queue_sysfs_entry queue_io_min_entry = {
	.attr = {.name = "minimum_io_size", .mode = S_IRUGO },
	.show = queue_io_min_show,
};

static struct queue_sysfs_entry queue_io_opt_entry = {
	.attr = {.name = "optimal_io_size", .mode = S_IRUGO },
	.show = queue_io_opt_show,
M
Martin K. Petersen 已提交
496 497
};

498 499 500 501 502
static struct queue_sysfs_entry queue_discard_granularity_entry = {
	.attr = {.name = "discard_granularity", .mode = S_IRUGO },
	.show = queue_discard_granularity_show,
};

503 504 505 506 507
static struct queue_sysfs_entry queue_discard_max_hw_entry = {
	.attr = {.name = "discard_max_hw_bytes", .mode = S_IRUGO },
	.show = queue_discard_max_hw_show,
};

508
static struct queue_sysfs_entry queue_discard_max_entry = {
509
	.attr = {.name = "discard_max_bytes", .mode = S_IRUGO | S_IWUSR },
510
	.show = queue_discard_max_show,
511
	.store = queue_discard_max_store,
512 513
};

514 515 516 517 518
static struct queue_sysfs_entry queue_discard_zeroes_data_entry = {
	.attr = {.name = "discard_zeroes_data", .mode = S_IRUGO },
	.show = queue_discard_zeroes_data_show,
};

519 520 521 522 523
static struct queue_sysfs_entry queue_write_same_max_entry = {
	.attr = {.name = "write_same_max_bytes", .mode = S_IRUGO },
	.show = queue_write_same_max_show,
};

524 525
static struct queue_sysfs_entry queue_nonrot_entry = {
	.attr = {.name = "rotational", .mode = S_IRUGO | S_IWUSR },
526 527
	.show = queue_show_nonrot,
	.store = queue_store_nonrot,
528 529
};

D
Damien Le Moal 已提交
530 531 532 533 534
static struct queue_sysfs_entry queue_zoned_entry = {
	.attr = {.name = "zoned", .mode = S_IRUGO },
	.show = queue_zoned_show,
};

535 536 537 538 539 540
static struct queue_sysfs_entry queue_nomerges_entry = {
	.attr = {.name = "nomerges", .mode = S_IRUGO | S_IWUSR },
	.show = queue_nomerges_show,
	.store = queue_nomerges_store,
};

541 542 543 544 545 546
static struct queue_sysfs_entry queue_rq_affinity_entry = {
	.attr = {.name = "rq_affinity", .mode = S_IRUGO | S_IWUSR },
	.show = queue_rq_affinity_show,
	.store = queue_rq_affinity_store,
};

547 548
static struct queue_sysfs_entry queue_iostats_entry = {
	.attr = {.name = "iostats", .mode = S_IRUGO | S_IWUSR },
549 550
	.show = queue_show_iostats,
	.store = queue_store_iostats,
551 552
};

553 554
static struct queue_sysfs_entry queue_random_entry = {
	.attr = {.name = "add_random", .mode = S_IRUGO | S_IWUSR },
555 556
	.show = queue_show_random,
	.store = queue_store_random,
557 558
};

J
Jens Axboe 已提交
559 560 561 562 563 564
static struct queue_sysfs_entry queue_poll_entry = {
	.attr = {.name = "io_poll", .mode = S_IRUGO | S_IWUSR },
	.show = queue_poll_show,
	.store = queue_poll_store,
};

565 566 567 568 569 570
static struct queue_sysfs_entry queue_wc_entry = {
	.attr = {.name = "write_cache", .mode = S_IRUGO | S_IWUSR },
	.show = queue_wc_show,
	.store = queue_wc_store,
};

571 572 573 574 575
static struct queue_sysfs_entry queue_dax_entry = {
	.attr = {.name = "dax", .mode = S_IRUGO },
	.show = queue_dax_show,
};

576 577 578 579 580
static struct queue_sysfs_entry queue_stats_entry = {
	.attr = {.name = "stats", .mode = S_IRUGO },
	.show = queue_stats_show,
};

581 582 583 584 585
static struct attribute *default_attrs[] = {
	&queue_requests_entry.attr,
	&queue_ra_entry.attr,
	&queue_max_hw_sectors_entry.attr,
	&queue_max_sectors_entry.attr,
586
	&queue_max_segments_entry.attr,
587
	&queue_max_integrity_segments_entry.attr,
588
	&queue_max_segment_size_entry.attr,
589
	&queue_iosched_entry.attr,
M
Martin K. Petersen 已提交
590
	&queue_hw_sector_size_entry.attr,
591
	&queue_logical_block_size_entry.attr,
592
	&queue_physical_block_size_entry.attr,
593
	&queue_chunk_sectors_entry.attr,
594 595
	&queue_io_min_entry.attr,
	&queue_io_opt_entry.attr,
596 597
	&queue_discard_granularity_entry.attr,
	&queue_discard_max_entry.attr,
598
	&queue_discard_max_hw_entry.attr,
599
	&queue_discard_zeroes_data_entry.attr,
600
	&queue_write_same_max_entry.attr,
601
	&queue_nonrot_entry.attr,
D
Damien Le Moal 已提交
602
	&queue_zoned_entry.attr,
603
	&queue_nomerges_entry.attr,
604
	&queue_rq_affinity_entry.attr,
605
	&queue_iostats_entry.attr,
606
	&queue_random_entry.attr,
J
Jens Axboe 已提交
607
	&queue_poll_entry.attr,
608
	&queue_wc_entry.attr,
609
	&queue_dax_entry.attr,
610
	&queue_stats_entry.attr,
611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626
	NULL,
};

#define to_queue(atr) container_of((atr), struct queue_sysfs_entry, attr)

static ssize_t
queue_attr_show(struct kobject *kobj, struct attribute *attr, char *page)
{
	struct queue_sysfs_entry *entry = to_queue(attr);
	struct request_queue *q =
		container_of(kobj, struct request_queue, kobj);
	ssize_t res;

	if (!entry->show)
		return -EIO;
	mutex_lock(&q->sysfs_lock);
B
Bart Van Assche 已提交
627
	if (blk_queue_dying(q)) {
628 629 630 631 632 633 634 635 636 637 638 639 640
		mutex_unlock(&q->sysfs_lock);
		return -ENOENT;
	}
	res = entry->show(q, page);
	mutex_unlock(&q->sysfs_lock);
	return res;
}

static ssize_t
queue_attr_store(struct kobject *kobj, struct attribute *attr,
		    const char *page, size_t length)
{
	struct queue_sysfs_entry *entry = to_queue(attr);
641
	struct request_queue *q;
642 643 644 645
	ssize_t res;

	if (!entry->store)
		return -EIO;
646 647

	q = container_of(kobj, struct request_queue, kobj);
648
	mutex_lock(&q->sysfs_lock);
B
Bart Van Assche 已提交
649
	if (blk_queue_dying(q)) {
650 651 652 653 654 655 656 657
		mutex_unlock(&q->sysfs_lock);
		return -ENOENT;
	}
	res = entry->store(q, page, length);
	mutex_unlock(&q->sysfs_lock);
	return res;
}

T
Tejun Heo 已提交
658 659 660 661 662 663 664
static void blk_free_queue_rcu(struct rcu_head *rcu_head)
{
	struct request_queue *q = container_of(rcu_head, struct request_queue,
					       rcu_head);
	kmem_cache_free(blk_requestq_cachep, q);
}

665
/**
666 667
 * blk_release_queue: - release a &struct request_queue when it is no longer needed
 * @kobj:    the kobj belonging to the request queue to be released
668 669
 *
 * Description:
670
 *     blk_release_queue is the pair to blk_init_queue() or
671 672 673 674 675
 *     blk_queue_make_request().  It should be called when a request queue is
 *     being released; typically when a block device is being de-registered.
 *     Currently, its primary task it to free all the &struct request
 *     structures that were allocated to the queue and the queue itself.
 *
B
Bart Van Assche 已提交
676 677 678
 * Note:
 *     The low level driver must have finished any outstanding requests first
 *     via blk_cleanup_queue().
679 680 681 682 683 684
 **/
static void blk_release_queue(struct kobject *kobj)
{
	struct request_queue *q =
		container_of(kobj, struct request_queue, kobj);

685
	bdi_exit(&q->backing_dev_info);
686 687
	blkcg_exit_queue(q);

688 689 690 691
	if (q->elevator) {
		spin_lock_irq(q->queue_lock);
		ioc_clear_queue(q);
		spin_unlock_irq(q->queue_lock);
692
		elevator_exit(q->elevator);
693
	}
694

695
	blk_exit_rl(&q->root_rl);
696 697 698 699

	if (q->queue_tags)
		__blk_queue_free_tags(q);

B
Bart Van Assche 已提交
700
	if (!q->mq_ops)
701
		blk_free_flush_queue(q->fq);
702 703
	else
		blk_mq_release(q);
704

705 706
	blk_trace_shutdown(q);

707 708 709
	if (q->bio_split)
		bioset_free(q->bio_split);

710
	ida_simple_remove(&blk_queue_ida, q->id);
T
Tejun Heo 已提交
711
	call_rcu(&q->rcu_head, blk_free_queue_rcu);
712 713
}

714
static const struct sysfs_ops queue_sysfs_ops = {
715 716 717 718 719 720 721 722 723 724 725 726 727
	.show	= queue_attr_show,
	.store	= queue_attr_store,
};

struct kobj_type blk_queue_ktype = {
	.sysfs_ops	= &queue_sysfs_ops,
	.default_attrs	= default_attrs,
	.release	= blk_release_queue,
};

int blk_register_queue(struct gendisk *disk)
{
	int ret;
728
	struct device *dev = disk_to_dev(disk);
729 730
	struct request_queue *q = disk->queue;

731
	if (WARN_ON(!q))
732 733
		return -ENXIO;

734
	/*
735 736 737 738 739 740 741
	 * SCSI probing may synchronously create and destroy a lot of
	 * request_queues for non-existent devices.  Shutting down a fully
	 * functional queue takes measureable wallclock time as RCU grace
	 * periods are involved.  To avoid excessive latency in these
	 * cases, a request_queue starts out in a degraded mode which is
	 * faster to shut down and is made fully functional here as
	 * request_queues for non-existent devices never get registered.
742
	 */
743 744
	if (!blk_queue_init_done(q)) {
		queue_flag_set_unlocked(QUEUE_FLAG_INIT_DONE, q);
745
		percpu_ref_switch_to_percpu(&q->q_usage_counter);
746 747
		blk_queue_bypass_end(q);
	}
748

749 750 751 752
	ret = blk_trace_init_sysfs(dev);
	if (ret)
		return ret;

753
	ret = kobject_add(&q->kobj, kobject_get(&dev->kobj), "%s", "queue");
754 755
	if (ret < 0) {
		blk_trace_remove_sysfs(dev);
756
		return ret;
757
	}
758 759 760

	kobject_uevent(&q->kobj, KOBJ_ADD);

761
	if (q->mq_ops)
762
		blk_mq_register_dev(dev, q);
763

764 765 766
	if (!q->request_fn)
		return 0;

767 768 769 770
	ret = elv_register_queue(q);
	if (ret) {
		kobject_uevent(&q->kobj, KOBJ_REMOVE);
		kobject_del(&q->kobj);
771
		blk_trace_remove_sysfs(dev);
772
		kobject_put(&dev->kobj);
773 774 775 776 777 778 779 780 781 782
		return ret;
	}

	return 0;
}

void blk_unregister_queue(struct gendisk *disk)
{
	struct request_queue *q = disk->queue;

783 784 785
	if (WARN_ON(!q))
		return;

786
	if (q->mq_ops)
787
		blk_mq_unregister_dev(disk_to_dev(disk), q);
788

789
	if (q->request_fn)
790 791
		elv_unregister_queue(q);

792 793 794 795
	kobject_uevent(&q->kobj, KOBJ_REMOVE);
	kobject_del(&q->kobj);
	blk_trace_remove_sysfs(disk_to_dev(disk));
	kobject_put(&disk_to_dev(disk)->kobj);
796
}